Upload folder using huggingface_hub
Browse files- checkpoint-400/config.json +79 -0
- checkpoint-400/generation_config.json +6 -0
- checkpoint-400/model-00001-of-00002.safetensors +3 -0
- checkpoint-400/model-00002-of-00002.safetensors +3 -0
- checkpoint-400/model.safetensors.index.json +0 -0
- checkpoint-400/optimizer.pt +3 -0
- checkpoint-400/rng_state_0.pth +3 -0
- checkpoint-400/rng_state_1.pth +3 -0
- checkpoint-400/rng_state_2.pth +3 -0
- checkpoint-400/rng_state_3.pth +3 -0
- checkpoint-400/scheduler.pt +3 -0
- checkpoint-400/trainer_state.json +2833 -0
- checkpoint-400/training_args.bin +3 -0
checkpoint-400/config.json
ADDED
|
@@ -0,0 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"ar_steps": 1,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DiffVLMBaseline"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"condition_layer": -1,
|
| 9 |
+
"eos_token_id": 151645,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 1536,
|
| 12 |
+
"image_token_id": 151655,
|
| 13 |
+
"img_cross_attention_dim": 2048,
|
| 14 |
+
"img_diffuser_depth": 6,
|
| 15 |
+
"img_ffn_dim_multiplier": null,
|
| 16 |
+
"img_hidden_size": 1536,
|
| 17 |
+
"img_multiple_of": 256,
|
| 18 |
+
"img_norm_eps": 1e-05,
|
| 19 |
+
"img_num_attention_heads": 12,
|
| 20 |
+
"img_num_kv_heads": 12,
|
| 21 |
+
"img_qk_norm": true,
|
| 22 |
+
"in_channels": 512,
|
| 23 |
+
"initializer_range": 0.02,
|
| 24 |
+
"inject_img_diffuser": false,
|
| 25 |
+
"input_size": 64,
|
| 26 |
+
"intermediate_size": 8960,
|
| 27 |
+
"layer_group_size": 7,
|
| 28 |
+
"layerwise_start_idx": 0,
|
| 29 |
+
"lora_alpha": 256,
|
| 30 |
+
"lora_bias": "none",
|
| 31 |
+
"lora_dropout": 0.05,
|
| 32 |
+
"lora_enable": false,
|
| 33 |
+
"lora_r": 128,
|
| 34 |
+
"max_position_embeddings": 32768,
|
| 35 |
+
"max_window_layers": 28,
|
| 36 |
+
"model_type": "qwen2_vl",
|
| 37 |
+
"non_linearity": 1,
|
| 38 |
+
"norm_elementwise_affine": true,
|
| 39 |
+
"num_attention_heads": 12,
|
| 40 |
+
"num_hidden_layers": 28,
|
| 41 |
+
"num_key_value_heads": 2,
|
| 42 |
+
"patch_size": 1,
|
| 43 |
+
"repa_coeff": 0.1,
|
| 44 |
+
"repa_layers": "2,9",
|
| 45 |
+
"repa_shared": false,
|
| 46 |
+
"rms_norm_eps": 1e-06,
|
| 47 |
+
"rope_scaling": {
|
| 48 |
+
"mrope_section": [
|
| 49 |
+
16,
|
| 50 |
+
24,
|
| 51 |
+
24
|
| 52 |
+
],
|
| 53 |
+
"rope_type": "default",
|
| 54 |
+
"type": "default"
|
| 55 |
+
},
|
| 56 |
+
"rope_theta": 1000000.0,
|
| 57 |
+
"sample_size": 128,
|
| 58 |
+
"sampling_steps": 28,
|
| 59 |
+
"sliding_window": null,
|
| 60 |
+
"tie_word_embeddings": true,
|
| 61 |
+
"torch_dtype": "bfloat16",
|
| 62 |
+
"transformers_version": "4.47.0",
|
| 63 |
+
"use_cache": true,
|
| 64 |
+
"use_repa": false,
|
| 65 |
+
"use_residual_attn": false,
|
| 66 |
+
"use_sliding_window": false,
|
| 67 |
+
"vae_path": "mit-han-lab/dc-ae-f128c512-mix-1.0-diffusers",
|
| 68 |
+
"video_token_id": 151656,
|
| 69 |
+
"vision_config": {
|
| 70 |
+
"hidden_size": 1536,
|
| 71 |
+
"in_chans": 3,
|
| 72 |
+
"model_type": "qwen2_vl",
|
| 73 |
+
"spatial_patch_size": 14
|
| 74 |
+
},
|
| 75 |
+
"vision_end_token_id": 151653,
|
| 76 |
+
"vision_start_token_id": 151652,
|
| 77 |
+
"vision_token_id": 151654,
|
| 78 |
+
"vocab_size": 151936
|
| 79 |
+
}
|
checkpoint-400/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151645,
|
| 5 |
+
"transformers_version": "4.47.0"
|
| 6 |
+
}
|
checkpoint-400/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c422412bdf535f6931dccfd08564d2bbf0558a42a220865eceeb3ca5b2f2cc0c
|
| 3 |
+
size 4950516240
|
checkpoint-400/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e816987aee1254b1aaf8dbb1b50e9e09b329c1f317bcf57a406bdbccd360a5ff
|
| 3 |
+
size 1664603880
|
checkpoint-400/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-400/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60a03cfdea99a2c8b6db855c7b7bd560ac645b1fadfd42b863a01c4f8d968958
|
| 3 |
+
size 7549751118
|
checkpoint-400/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fbe95690f3a0904434402353c5fc6381b2c82d4402b33dc54f4bdf2cec70fdd3
|
| 3 |
+
size 15024
|
checkpoint-400/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58db16b45e0bce3f2a68182cd8ffb586be10d1f0f7c4dbeec5687e3fadd9954f
|
| 3 |
+
size 15024
|
checkpoint-400/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c04eb51520aaefd08d8770389e5e759719ce98e40c8834d5ae6573704d3616e8
|
| 3 |
+
size 15024
|
checkpoint-400/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a1f42c1c8100a5fc33d5b867c0fcc821a3eb11e9e281ca60ca3b67cb33793403
|
| 3 |
+
size 15024
|
checkpoint-400/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a3dedf52aaea6588501a76bbcc0db7c74d91e0049f19cb86f239096b664186d
|
| 3 |
+
size 1064
|
checkpoint-400/trainer_state.json
ADDED
|
@@ -0,0 +1,2833 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.019981018032868773,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 400,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 4.9952545082171936e-05,
|
| 13 |
+
"grad_norm": 0.87890625,
|
| 14 |
+
"learning_rate": 0.0002,
|
| 15 |
+
"loss": 3.1785,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 9.990509016434387e-05,
|
| 20 |
+
"grad_norm": 0.92578125,
|
| 21 |
+
"learning_rate": 0.0002,
|
| 22 |
+
"loss": 2.9941,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0001498576352465158,
|
| 27 |
+
"grad_norm": 1.0859375,
|
| 28 |
+
"learning_rate": 0.0002,
|
| 29 |
+
"loss": 2.7646,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.00019981018032868775,
|
| 34 |
+
"grad_norm": 1.0,
|
| 35 |
+
"learning_rate": 0.0002,
|
| 36 |
+
"loss": 2.6073,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.0002497627254108597,
|
| 41 |
+
"grad_norm": 0.984375,
|
| 42 |
+
"learning_rate": 0.0002,
|
| 43 |
+
"loss": 2.3667,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0002997152704930316,
|
| 48 |
+
"grad_norm": 0.734375,
|
| 49 |
+
"learning_rate": 0.0002,
|
| 50 |
+
"loss": 2.0751,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0003496678155752036,
|
| 55 |
+
"grad_norm": 0.54296875,
|
| 56 |
+
"learning_rate": 0.0002,
|
| 57 |
+
"loss": 2.0617,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0003996203606573755,
|
| 62 |
+
"grad_norm": 0.423828125,
|
| 63 |
+
"learning_rate": 0.0002,
|
| 64 |
+
"loss": 1.9628,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0004495729057395474,
|
| 69 |
+
"grad_norm": 0.39453125,
|
| 70 |
+
"learning_rate": 0.0002,
|
| 71 |
+
"loss": 2.0122,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0004995254508217194,
|
| 76 |
+
"grad_norm": 0.3203125,
|
| 77 |
+
"learning_rate": 0.0002,
|
| 78 |
+
"loss": 1.9424,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.0005494779959038913,
|
| 83 |
+
"grad_norm": 0.30859375,
|
| 84 |
+
"learning_rate": 0.0002,
|
| 85 |
+
"loss": 1.9736,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0005994305409860632,
|
| 90 |
+
"grad_norm": 0.25390625,
|
| 91 |
+
"learning_rate": 0.0002,
|
| 92 |
+
"loss": 1.9844,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.0006493830860682351,
|
| 97 |
+
"grad_norm": 0.25390625,
|
| 98 |
+
"learning_rate": 0.0002,
|
| 99 |
+
"loss": 1.9095,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0006993356311504072,
|
| 104 |
+
"grad_norm": 0.21875,
|
| 105 |
+
"learning_rate": 0.0002,
|
| 106 |
+
"loss": 1.9302,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.0007492881762325791,
|
| 111 |
+
"grad_norm": 0.2119140625,
|
| 112 |
+
"learning_rate": 0.0002,
|
| 113 |
+
"loss": 1.9764,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.000799240721314751,
|
| 118 |
+
"grad_norm": 0.1923828125,
|
| 119 |
+
"learning_rate": 0.0002,
|
| 120 |
+
"loss": 1.9615,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.0008491932663969229,
|
| 125 |
+
"grad_norm": 0.19140625,
|
| 126 |
+
"learning_rate": 0.0002,
|
| 127 |
+
"loss": 1.9642,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0008991458114790948,
|
| 132 |
+
"grad_norm": 0.18359375,
|
| 133 |
+
"learning_rate": 0.0002,
|
| 134 |
+
"loss": 1.9293,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.0009490983565612668,
|
| 139 |
+
"grad_norm": 0.2041015625,
|
| 140 |
+
"learning_rate": 0.0002,
|
| 141 |
+
"loss": 1.9398,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.0009990509016434388,
|
| 146 |
+
"grad_norm": 0.16015625,
|
| 147 |
+
"learning_rate": 0.0002,
|
| 148 |
+
"loss": 1.9432,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0010490034467256107,
|
| 153 |
+
"grad_norm": 0.1962890625,
|
| 154 |
+
"learning_rate": 0.0002,
|
| 155 |
+
"loss": 1.959,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.0010989559918077826,
|
| 160 |
+
"grad_norm": 0.1611328125,
|
| 161 |
+
"learning_rate": 0.0002,
|
| 162 |
+
"loss": 1.9206,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.0011489085368899545,
|
| 167 |
+
"grad_norm": 0.25390625,
|
| 168 |
+
"learning_rate": 0.0002,
|
| 169 |
+
"loss": 1.9494,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.0011988610819721264,
|
| 174 |
+
"grad_norm": 0.158203125,
|
| 175 |
+
"learning_rate": 0.0002,
|
| 176 |
+
"loss": 1.9419,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.0012488136270542983,
|
| 181 |
+
"grad_norm": 0.208984375,
|
| 182 |
+
"learning_rate": 0.0002,
|
| 183 |
+
"loss": 1.924,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.0012987661721364703,
|
| 188 |
+
"grad_norm": 0.17578125,
|
| 189 |
+
"learning_rate": 0.0002,
|
| 190 |
+
"loss": 1.9857,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.0013487187172186424,
|
| 195 |
+
"grad_norm": 0.1533203125,
|
| 196 |
+
"learning_rate": 0.0002,
|
| 197 |
+
"loss": 1.91,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.0013986712623008143,
|
| 202 |
+
"grad_norm": 0.19140625,
|
| 203 |
+
"learning_rate": 0.0002,
|
| 204 |
+
"loss": 1.9722,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.0014486238073829862,
|
| 209 |
+
"grad_norm": 0.1591796875,
|
| 210 |
+
"learning_rate": 0.0002,
|
| 211 |
+
"loss": 1.9358,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.0014985763524651581,
|
| 216 |
+
"grad_norm": 0.138671875,
|
| 217 |
+
"learning_rate": 0.0002,
|
| 218 |
+
"loss": 1.927,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.00154852889754733,
|
| 223 |
+
"grad_norm": 0.1298828125,
|
| 224 |
+
"learning_rate": 0.0002,
|
| 225 |
+
"loss": 1.9234,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.001598481442629502,
|
| 230 |
+
"grad_norm": 0.1689453125,
|
| 231 |
+
"learning_rate": 0.0002,
|
| 232 |
+
"loss": 1.9374,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.0016484339877116739,
|
| 237 |
+
"grad_norm": 0.134765625,
|
| 238 |
+
"learning_rate": 0.0002,
|
| 239 |
+
"loss": 1.9535,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.0016983865327938458,
|
| 244 |
+
"grad_norm": 0.130859375,
|
| 245 |
+
"learning_rate": 0.0002,
|
| 246 |
+
"loss": 1.9122,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.0017483390778760177,
|
| 251 |
+
"grad_norm": 0.1455078125,
|
| 252 |
+
"learning_rate": 0.0002,
|
| 253 |
+
"loss": 1.9185,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.0017982916229581896,
|
| 258 |
+
"grad_norm": 0.1474609375,
|
| 259 |
+
"learning_rate": 0.0002,
|
| 260 |
+
"loss": 1.9548,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.0018482441680403618,
|
| 265 |
+
"grad_norm": 0.1484375,
|
| 266 |
+
"learning_rate": 0.0002,
|
| 267 |
+
"loss": 1.9392,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.0018981967131225337,
|
| 272 |
+
"grad_norm": 0.1533203125,
|
| 273 |
+
"learning_rate": 0.0002,
|
| 274 |
+
"loss": 1.9088,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.0019481492582047056,
|
| 279 |
+
"grad_norm": 0.1416015625,
|
| 280 |
+
"learning_rate": 0.0002,
|
| 281 |
+
"loss": 1.9547,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.0019981018032868775,
|
| 286 |
+
"grad_norm": 0.1630859375,
|
| 287 |
+
"learning_rate": 0.0002,
|
| 288 |
+
"loss": 1.9569,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.0020480543483690494,
|
| 293 |
+
"grad_norm": 0.171875,
|
| 294 |
+
"learning_rate": 0.0002,
|
| 295 |
+
"loss": 1.9443,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.0020980068934512213,
|
| 300 |
+
"grad_norm": 0.1376953125,
|
| 301 |
+
"learning_rate": 0.0002,
|
| 302 |
+
"loss": 1.9093,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.0021479594385333933,
|
| 307 |
+
"grad_norm": 0.1748046875,
|
| 308 |
+
"learning_rate": 0.0002,
|
| 309 |
+
"loss": 1.9484,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.002197911983615565,
|
| 314 |
+
"grad_norm": 0.140625,
|
| 315 |
+
"learning_rate": 0.0002,
|
| 316 |
+
"loss": 1.9196,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.002247864528697737,
|
| 321 |
+
"grad_norm": 0.154296875,
|
| 322 |
+
"learning_rate": 0.0002,
|
| 323 |
+
"loss": 1.9028,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.002297817073779909,
|
| 328 |
+
"grad_norm": 0.1591796875,
|
| 329 |
+
"learning_rate": 0.0002,
|
| 330 |
+
"loss": 1.9247,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.002347769618862081,
|
| 335 |
+
"grad_norm": 0.1591796875,
|
| 336 |
+
"learning_rate": 0.0002,
|
| 337 |
+
"loss": 1.9197,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.002397722163944253,
|
| 342 |
+
"grad_norm": 0.1669921875,
|
| 343 |
+
"learning_rate": 0.0002,
|
| 344 |
+
"loss": 1.929,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.0024476747090264248,
|
| 349 |
+
"grad_norm": 0.169921875,
|
| 350 |
+
"learning_rate": 0.0002,
|
| 351 |
+
"loss": 1.9066,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.0024976272541085967,
|
| 356 |
+
"grad_norm": 0.1708984375,
|
| 357 |
+
"learning_rate": 0.0002,
|
| 358 |
+
"loss": 1.954,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.0025475797991907686,
|
| 363 |
+
"grad_norm": 0.1669921875,
|
| 364 |
+
"learning_rate": 0.0002,
|
| 365 |
+
"loss": 1.9443,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.0025975323442729405,
|
| 370 |
+
"grad_norm": 0.1552734375,
|
| 371 |
+
"learning_rate": 0.0002,
|
| 372 |
+
"loss": 1.9144,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.002647484889355113,
|
| 377 |
+
"grad_norm": 0.1748046875,
|
| 378 |
+
"learning_rate": 0.0002,
|
| 379 |
+
"loss": 1.8864,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.0026974374344372848,
|
| 384 |
+
"grad_norm": 0.162109375,
|
| 385 |
+
"learning_rate": 0.0002,
|
| 386 |
+
"loss": 1.9331,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.0027473899795194567,
|
| 391 |
+
"grad_norm": 0.15234375,
|
| 392 |
+
"learning_rate": 0.0002,
|
| 393 |
+
"loss": 1.9029,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.0027973425246016286,
|
| 398 |
+
"grad_norm": 0.1513671875,
|
| 399 |
+
"learning_rate": 0.0002,
|
| 400 |
+
"loss": 1.9292,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.0028472950696838005,
|
| 405 |
+
"grad_norm": 0.154296875,
|
| 406 |
+
"learning_rate": 0.0002,
|
| 407 |
+
"loss": 1.9331,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.0028972476147659724,
|
| 412 |
+
"grad_norm": 0.1572265625,
|
| 413 |
+
"learning_rate": 0.0002,
|
| 414 |
+
"loss": 1.9132,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.0029472001598481443,
|
| 419 |
+
"grad_norm": 0.181640625,
|
| 420 |
+
"learning_rate": 0.0002,
|
| 421 |
+
"loss": 1.8793,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.0029971527049303163,
|
| 426 |
+
"grad_norm": 0.2490234375,
|
| 427 |
+
"learning_rate": 0.0002,
|
| 428 |
+
"loss": 1.8864,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.003047105250012488,
|
| 433 |
+
"grad_norm": 0.18359375,
|
| 434 |
+
"learning_rate": 0.0002,
|
| 435 |
+
"loss": 1.8755,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.00309705779509466,
|
| 440 |
+
"grad_norm": 0.171875,
|
| 441 |
+
"learning_rate": 0.0002,
|
| 442 |
+
"loss": 1.9099,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.003147010340176832,
|
| 447 |
+
"grad_norm": 0.1982421875,
|
| 448 |
+
"learning_rate": 0.0002,
|
| 449 |
+
"loss": 1.9434,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.003196962885259004,
|
| 454 |
+
"grad_norm": 0.171875,
|
| 455 |
+
"learning_rate": 0.0002,
|
| 456 |
+
"loss": 1.9597,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.003246915430341176,
|
| 461 |
+
"grad_norm": 0.169921875,
|
| 462 |
+
"learning_rate": 0.0002,
|
| 463 |
+
"loss": 1.8813,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.0032968679754233478,
|
| 468 |
+
"grad_norm": 0.1806640625,
|
| 469 |
+
"learning_rate": 0.0002,
|
| 470 |
+
"loss": 1.9121,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.0033468205205055197,
|
| 475 |
+
"grad_norm": 0.1845703125,
|
| 476 |
+
"learning_rate": 0.0002,
|
| 477 |
+
"loss": 1.8903,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.0033967730655876916,
|
| 482 |
+
"grad_norm": 0.208984375,
|
| 483 |
+
"learning_rate": 0.0002,
|
| 484 |
+
"loss": 1.892,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.0034467256106698635,
|
| 489 |
+
"grad_norm": 0.16796875,
|
| 490 |
+
"learning_rate": 0.0002,
|
| 491 |
+
"loss": 1.9102,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.0034966781557520354,
|
| 496 |
+
"grad_norm": 0.25390625,
|
| 497 |
+
"learning_rate": 0.0002,
|
| 498 |
+
"loss": 1.9265,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.0035466307008342073,
|
| 503 |
+
"grad_norm": 0.2060546875,
|
| 504 |
+
"learning_rate": 0.0002,
|
| 505 |
+
"loss": 1.9059,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.0035965832459163793,
|
| 510 |
+
"grad_norm": 0.2451171875,
|
| 511 |
+
"learning_rate": 0.0002,
|
| 512 |
+
"loss": 1.8916,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.003646535790998551,
|
| 517 |
+
"grad_norm": 0.2041015625,
|
| 518 |
+
"learning_rate": 0.0002,
|
| 519 |
+
"loss": 1.9525,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.0036964883360807235,
|
| 524 |
+
"grad_norm": 0.2197265625,
|
| 525 |
+
"learning_rate": 0.0002,
|
| 526 |
+
"loss": 1.912,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.0037464408811628954,
|
| 531 |
+
"grad_norm": 0.1669921875,
|
| 532 |
+
"learning_rate": 0.0002,
|
| 533 |
+
"loss": 1.8936,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.0037963934262450674,
|
| 538 |
+
"grad_norm": 0.2001953125,
|
| 539 |
+
"learning_rate": 0.0002,
|
| 540 |
+
"loss": 1.8205,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.0038463459713272393,
|
| 545 |
+
"grad_norm": 0.2275390625,
|
| 546 |
+
"learning_rate": 0.0002,
|
| 547 |
+
"loss": 1.8902,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.003896298516409411,
|
| 552 |
+
"grad_norm": 0.1884765625,
|
| 553 |
+
"learning_rate": 0.0002,
|
| 554 |
+
"loss": 1.8934,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.003946251061491583,
|
| 559 |
+
"grad_norm": 0.2138671875,
|
| 560 |
+
"learning_rate": 0.0002,
|
| 561 |
+
"loss": 1.8697,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.003996203606573755,
|
| 566 |
+
"grad_norm": 0.240234375,
|
| 567 |
+
"learning_rate": 0.0002,
|
| 568 |
+
"loss": 1.8818,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.004046156151655927,
|
| 573 |
+
"grad_norm": 0.2041015625,
|
| 574 |
+
"learning_rate": 0.0002,
|
| 575 |
+
"loss": 1.8684,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.004096108696738099,
|
| 580 |
+
"grad_norm": 0.2001953125,
|
| 581 |
+
"learning_rate": 0.0002,
|
| 582 |
+
"loss": 1.8652,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.004146061241820271,
|
| 587 |
+
"grad_norm": 0.1826171875,
|
| 588 |
+
"learning_rate": 0.0002,
|
| 589 |
+
"loss": 1.9427,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.004196013786902443,
|
| 594 |
+
"grad_norm": 0.197265625,
|
| 595 |
+
"learning_rate": 0.0002,
|
| 596 |
+
"loss": 1.8898,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.004245966331984615,
|
| 601 |
+
"grad_norm": 0.1953125,
|
| 602 |
+
"learning_rate": 0.0002,
|
| 603 |
+
"loss": 1.8865,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.0042959188770667865,
|
| 608 |
+
"grad_norm": 0.208984375,
|
| 609 |
+
"learning_rate": 0.0002,
|
| 610 |
+
"loss": 1.9,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.004345871422148958,
|
| 615 |
+
"grad_norm": 0.224609375,
|
| 616 |
+
"learning_rate": 0.0002,
|
| 617 |
+
"loss": 1.8938,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.00439582396723113,
|
| 622 |
+
"grad_norm": 0.1962890625,
|
| 623 |
+
"learning_rate": 0.0002,
|
| 624 |
+
"loss": 1.8586,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.004445776512313302,
|
| 629 |
+
"grad_norm": 0.2041015625,
|
| 630 |
+
"learning_rate": 0.0002,
|
| 631 |
+
"loss": 1.9017,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.004495729057395474,
|
| 636 |
+
"grad_norm": 0.216796875,
|
| 637 |
+
"learning_rate": 0.0002,
|
| 638 |
+
"loss": 1.8907,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.004545681602477646,
|
| 643 |
+
"grad_norm": 0.2021484375,
|
| 644 |
+
"learning_rate": 0.0002,
|
| 645 |
+
"loss": 1.9174,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.004595634147559818,
|
| 650 |
+
"grad_norm": 0.2021484375,
|
| 651 |
+
"learning_rate": 0.0002,
|
| 652 |
+
"loss": 1.8972,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.00464558669264199,
|
| 657 |
+
"grad_norm": 0.236328125,
|
| 658 |
+
"learning_rate": 0.0002,
|
| 659 |
+
"loss": 1.8409,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.004695539237724162,
|
| 664 |
+
"grad_norm": 0.19921875,
|
| 665 |
+
"learning_rate": 0.0002,
|
| 666 |
+
"loss": 1.9078,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.004745491782806334,
|
| 671 |
+
"grad_norm": 0.197265625,
|
| 672 |
+
"learning_rate": 0.0002,
|
| 673 |
+
"loss": 1.8524,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.004795444327888506,
|
| 678 |
+
"grad_norm": 0.205078125,
|
| 679 |
+
"learning_rate": 0.0002,
|
| 680 |
+
"loss": 1.8636,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.004845396872970678,
|
| 685 |
+
"grad_norm": 0.189453125,
|
| 686 |
+
"learning_rate": 0.0002,
|
| 687 |
+
"loss": 1.86,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.0048953494180528495,
|
| 692 |
+
"grad_norm": 0.193359375,
|
| 693 |
+
"learning_rate": 0.0002,
|
| 694 |
+
"loss": 1.8506,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.004945301963135021,
|
| 699 |
+
"grad_norm": 0.205078125,
|
| 700 |
+
"learning_rate": 0.0002,
|
| 701 |
+
"loss": 1.8516,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.004995254508217193,
|
| 706 |
+
"grad_norm": 0.1953125,
|
| 707 |
+
"learning_rate": 0.0002,
|
| 708 |
+
"loss": 1.8619,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.005045207053299365,
|
| 713 |
+
"grad_norm": 0.23828125,
|
| 714 |
+
"learning_rate": 0.0002,
|
| 715 |
+
"loss": 1.8473,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.005095159598381537,
|
| 720 |
+
"grad_norm": 0.2177734375,
|
| 721 |
+
"learning_rate": 0.0002,
|
| 722 |
+
"loss": 1.8933,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.005145112143463709,
|
| 727 |
+
"grad_norm": 0.220703125,
|
| 728 |
+
"learning_rate": 0.0002,
|
| 729 |
+
"loss": 1.9207,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.005195064688545881,
|
| 734 |
+
"grad_norm": 0.2216796875,
|
| 735 |
+
"learning_rate": 0.0002,
|
| 736 |
+
"loss": 1.9016,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.005245017233628053,
|
| 741 |
+
"grad_norm": 0.20703125,
|
| 742 |
+
"learning_rate": 0.0002,
|
| 743 |
+
"loss": 1.8581,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.005294969778710226,
|
| 748 |
+
"grad_norm": 0.2392578125,
|
| 749 |
+
"learning_rate": 0.0002,
|
| 750 |
+
"loss": 1.8559,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.005344922323792398,
|
| 755 |
+
"grad_norm": 0.400390625,
|
| 756 |
+
"learning_rate": 0.0002,
|
| 757 |
+
"loss": 1.8552,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.0053948748688745695,
|
| 762 |
+
"grad_norm": 0.271484375,
|
| 763 |
+
"learning_rate": 0.0002,
|
| 764 |
+
"loss": 1.8718,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.0054448274139567415,
|
| 769 |
+
"grad_norm": 0.2255859375,
|
| 770 |
+
"learning_rate": 0.0002,
|
| 771 |
+
"loss": 1.8216,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.005494779959038913,
|
| 776 |
+
"grad_norm": 0.212890625,
|
| 777 |
+
"learning_rate": 0.0002,
|
| 778 |
+
"loss": 1.8263,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.005544732504121085,
|
| 783 |
+
"grad_norm": 0.244140625,
|
| 784 |
+
"learning_rate": 0.0002,
|
| 785 |
+
"loss": 1.9171,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.005594685049203257,
|
| 790 |
+
"grad_norm": 0.216796875,
|
| 791 |
+
"learning_rate": 0.0002,
|
| 792 |
+
"loss": 1.8821,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.005644637594285429,
|
| 797 |
+
"grad_norm": 0.2470703125,
|
| 798 |
+
"learning_rate": 0.0002,
|
| 799 |
+
"loss": 1.8437,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.005694590139367601,
|
| 804 |
+
"grad_norm": 0.2294921875,
|
| 805 |
+
"learning_rate": 0.0002,
|
| 806 |
+
"loss": 1.8886,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.005744542684449773,
|
| 811 |
+
"grad_norm": 0.2119140625,
|
| 812 |
+
"learning_rate": 0.0002,
|
| 813 |
+
"loss": 1.8356,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.005794495229531945,
|
| 818 |
+
"grad_norm": 0.255859375,
|
| 819 |
+
"learning_rate": 0.0002,
|
| 820 |
+
"loss": 1.8809,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.005844447774614117,
|
| 825 |
+
"grad_norm": 0.2470703125,
|
| 826 |
+
"learning_rate": 0.0002,
|
| 827 |
+
"loss": 1.869,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.005894400319696289,
|
| 832 |
+
"grad_norm": 0.2392578125,
|
| 833 |
+
"learning_rate": 0.0002,
|
| 834 |
+
"loss": 1.8319,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.005944352864778461,
|
| 839 |
+
"grad_norm": 0.234375,
|
| 840 |
+
"learning_rate": 0.0002,
|
| 841 |
+
"loss": 1.874,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.0059943054098606325,
|
| 846 |
+
"grad_norm": 0.275390625,
|
| 847 |
+
"learning_rate": 0.0002,
|
| 848 |
+
"loss": 1.8399,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.0060442579549428044,
|
| 853 |
+
"grad_norm": 0.236328125,
|
| 854 |
+
"learning_rate": 0.0002,
|
| 855 |
+
"loss": 1.8748,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.006094210500024976,
|
| 860 |
+
"grad_norm": 0.259765625,
|
| 861 |
+
"learning_rate": 0.0002,
|
| 862 |
+
"loss": 1.8189,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.006144163045107148,
|
| 867 |
+
"grad_norm": 0.234375,
|
| 868 |
+
"learning_rate": 0.0002,
|
| 869 |
+
"loss": 1.826,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.00619411559018932,
|
| 874 |
+
"grad_norm": 0.265625,
|
| 875 |
+
"learning_rate": 0.0002,
|
| 876 |
+
"loss": 1.8255,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.006244068135271492,
|
| 881 |
+
"grad_norm": 0.2734375,
|
| 882 |
+
"learning_rate": 0.0002,
|
| 883 |
+
"loss": 1.8599,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.006294020680353664,
|
| 888 |
+
"grad_norm": 0.298828125,
|
| 889 |
+
"learning_rate": 0.0002,
|
| 890 |
+
"loss": 1.8133,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.006343973225435836,
|
| 895 |
+
"grad_norm": 0.30078125,
|
| 896 |
+
"learning_rate": 0.0002,
|
| 897 |
+
"loss": 1.8326,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.006393925770518008,
|
| 902 |
+
"grad_norm": 0.26953125,
|
| 903 |
+
"learning_rate": 0.0002,
|
| 904 |
+
"loss": 1.8662,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.00644387831560018,
|
| 909 |
+
"grad_norm": 0.275390625,
|
| 910 |
+
"learning_rate": 0.0002,
|
| 911 |
+
"loss": 1.866,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.006493830860682352,
|
| 916 |
+
"grad_norm": 0.29296875,
|
| 917 |
+
"learning_rate": 0.0002,
|
| 918 |
+
"loss": 1.796,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.006543783405764524,
|
| 923 |
+
"grad_norm": 0.35546875,
|
| 924 |
+
"learning_rate": 0.0002,
|
| 925 |
+
"loss": 1.8509,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.0065937359508466955,
|
| 930 |
+
"grad_norm": 0.30859375,
|
| 931 |
+
"learning_rate": 0.0002,
|
| 932 |
+
"loss": 1.8754,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.0066436884959288674,
|
| 937 |
+
"grad_norm": 0.30859375,
|
| 938 |
+
"learning_rate": 0.0002,
|
| 939 |
+
"loss": 1.7899,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.006693641041011039,
|
| 944 |
+
"grad_norm": 0.337890625,
|
| 945 |
+
"learning_rate": 0.0002,
|
| 946 |
+
"loss": 1.8174,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.006743593586093211,
|
| 951 |
+
"grad_norm": 0.29296875,
|
| 952 |
+
"learning_rate": 0.0002,
|
| 953 |
+
"loss": 1.7652,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.006793546131175383,
|
| 958 |
+
"grad_norm": 0.287109375,
|
| 959 |
+
"learning_rate": 0.0002,
|
| 960 |
+
"loss": 1.79,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.006843498676257555,
|
| 965 |
+
"grad_norm": 0.341796875,
|
| 966 |
+
"learning_rate": 0.0002,
|
| 967 |
+
"loss": 1.7966,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.006893451221339727,
|
| 972 |
+
"grad_norm": 0.35546875,
|
| 973 |
+
"learning_rate": 0.0002,
|
| 974 |
+
"loss": 1.805,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.006943403766421899,
|
| 979 |
+
"grad_norm": 0.30078125,
|
| 980 |
+
"learning_rate": 0.0002,
|
| 981 |
+
"loss": 1.8051,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.006993356311504071,
|
| 986 |
+
"grad_norm": 0.31640625,
|
| 987 |
+
"learning_rate": 0.0002,
|
| 988 |
+
"loss": 1.7455,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.007043308856586243,
|
| 993 |
+
"grad_norm": 0.353515625,
|
| 994 |
+
"learning_rate": 0.0002,
|
| 995 |
+
"loss": 1.8044,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.007093261401668415,
|
| 1000 |
+
"grad_norm": 0.302734375,
|
| 1001 |
+
"learning_rate": 0.0002,
|
| 1002 |
+
"loss": 1.7841,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.007143213946750587,
|
| 1007 |
+
"grad_norm": 0.34765625,
|
| 1008 |
+
"learning_rate": 0.0002,
|
| 1009 |
+
"loss": 1.7681,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.0071931664918327585,
|
| 1014 |
+
"grad_norm": 0.34375,
|
| 1015 |
+
"learning_rate": 0.0002,
|
| 1016 |
+
"loss": 1.7841,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.00724311903691493,
|
| 1021 |
+
"grad_norm": 0.328125,
|
| 1022 |
+
"learning_rate": 0.0002,
|
| 1023 |
+
"loss": 1.7377,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.007293071581997102,
|
| 1028 |
+
"grad_norm": 0.490234375,
|
| 1029 |
+
"learning_rate": 0.0002,
|
| 1030 |
+
"loss": 1.7701,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.007343024127079274,
|
| 1035 |
+
"grad_norm": 0.3515625,
|
| 1036 |
+
"learning_rate": 0.0002,
|
| 1037 |
+
"loss": 1.7723,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.007392976672161447,
|
| 1042 |
+
"grad_norm": 0.279296875,
|
| 1043 |
+
"learning_rate": 0.0002,
|
| 1044 |
+
"loss": 1.7951,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.007442929217243619,
|
| 1049 |
+
"grad_norm": 0.6484375,
|
| 1050 |
+
"learning_rate": 0.0002,
|
| 1051 |
+
"loss": 1.807,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.007492881762325791,
|
| 1056 |
+
"grad_norm": 0.34765625,
|
| 1057 |
+
"learning_rate": 0.0002,
|
| 1058 |
+
"loss": 1.6923,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.007542834307407963,
|
| 1063 |
+
"grad_norm": 0.486328125,
|
| 1064 |
+
"learning_rate": 0.0002,
|
| 1065 |
+
"loss": 1.7309,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.007592786852490135,
|
| 1070 |
+
"grad_norm": 0.53125,
|
| 1071 |
+
"learning_rate": 0.0002,
|
| 1072 |
+
"loss": 1.7541,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.007642739397572307,
|
| 1077 |
+
"grad_norm": 0.38671875,
|
| 1078 |
+
"learning_rate": 0.0002,
|
| 1079 |
+
"loss": 1.7762,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.0076926919426544785,
|
| 1084 |
+
"grad_norm": 0.388671875,
|
| 1085 |
+
"learning_rate": 0.0002,
|
| 1086 |
+
"loss": 1.7433,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.0077426444877366505,
|
| 1091 |
+
"grad_norm": 0.384765625,
|
| 1092 |
+
"learning_rate": 0.0002,
|
| 1093 |
+
"loss": 1.7226,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.007792597032818822,
|
| 1098 |
+
"grad_norm": 0.458984375,
|
| 1099 |
+
"learning_rate": 0.0002,
|
| 1100 |
+
"loss": 1.7588,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.007842549577900993,
|
| 1105 |
+
"grad_norm": 0.412109375,
|
| 1106 |
+
"learning_rate": 0.0002,
|
| 1107 |
+
"loss": 1.7293,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.007892502122983166,
|
| 1112 |
+
"grad_norm": 0.484375,
|
| 1113 |
+
"learning_rate": 0.0002,
|
| 1114 |
+
"loss": 1.727,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.007942454668065337,
|
| 1119 |
+
"grad_norm": 0.4765625,
|
| 1120 |
+
"learning_rate": 0.0002,
|
| 1121 |
+
"loss": 1.7792,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.00799240721314751,
|
| 1126 |
+
"grad_norm": 0.408203125,
|
| 1127 |
+
"learning_rate": 0.0002,
|
| 1128 |
+
"loss": 1.7624,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.008042359758229681,
|
| 1133 |
+
"grad_norm": 0.515625,
|
| 1134 |
+
"learning_rate": 0.0002,
|
| 1135 |
+
"loss": 1.7513,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.008092312303311854,
|
| 1140 |
+
"grad_norm": 0.423828125,
|
| 1141 |
+
"learning_rate": 0.0002,
|
| 1142 |
+
"loss": 1.7936,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.008142264848394025,
|
| 1147 |
+
"grad_norm": 0.5078125,
|
| 1148 |
+
"learning_rate": 0.0002,
|
| 1149 |
+
"loss": 1.7459,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.008192217393476198,
|
| 1154 |
+
"grad_norm": 0.482421875,
|
| 1155 |
+
"learning_rate": 0.0002,
|
| 1156 |
+
"loss": 1.7248,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.008242169938558369,
|
| 1161 |
+
"grad_norm": 0.33984375,
|
| 1162 |
+
"learning_rate": 0.0002,
|
| 1163 |
+
"loss": 1.7496,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.008292122483640542,
|
| 1168 |
+
"grad_norm": 0.365234375,
|
| 1169 |
+
"learning_rate": 0.0002,
|
| 1170 |
+
"loss": 1.6761,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.008342075028722713,
|
| 1175 |
+
"grad_norm": 0.40234375,
|
| 1176 |
+
"learning_rate": 0.0002,
|
| 1177 |
+
"loss": 1.743,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.008392027573804885,
|
| 1182 |
+
"grad_norm": 0.375,
|
| 1183 |
+
"learning_rate": 0.0002,
|
| 1184 |
+
"loss": 1.7075,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.008441980118887058,
|
| 1189 |
+
"grad_norm": 0.34375,
|
| 1190 |
+
"learning_rate": 0.0002,
|
| 1191 |
+
"loss": 1.7036,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.00849193266396923,
|
| 1196 |
+
"grad_norm": 0.357421875,
|
| 1197 |
+
"learning_rate": 0.0002,
|
| 1198 |
+
"loss": 1.7552,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.008541885209051402,
|
| 1203 |
+
"grad_norm": 0.318359375,
|
| 1204 |
+
"learning_rate": 0.0002,
|
| 1205 |
+
"loss": 1.704,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.008591837754133573,
|
| 1210 |
+
"grad_norm": 0.31640625,
|
| 1211 |
+
"learning_rate": 0.0002,
|
| 1212 |
+
"loss": 1.6707,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.008641790299215746,
|
| 1217 |
+
"grad_norm": 0.318359375,
|
| 1218 |
+
"learning_rate": 0.0002,
|
| 1219 |
+
"loss": 1.6477,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.008691742844297917,
|
| 1224 |
+
"grad_norm": 0.333984375,
|
| 1225 |
+
"learning_rate": 0.0002,
|
| 1226 |
+
"loss": 1.6907,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.00874169538938009,
|
| 1231 |
+
"grad_norm": 0.33984375,
|
| 1232 |
+
"learning_rate": 0.0002,
|
| 1233 |
+
"loss": 1.6488,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.00879164793446226,
|
| 1238 |
+
"grad_norm": 0.30078125,
|
| 1239 |
+
"learning_rate": 0.0002,
|
| 1240 |
+
"loss": 1.6818,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.008841600479544433,
|
| 1245 |
+
"grad_norm": 0.326171875,
|
| 1246 |
+
"learning_rate": 0.0002,
|
| 1247 |
+
"loss": 1.6745,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.008891553024626605,
|
| 1252 |
+
"grad_norm": 0.298828125,
|
| 1253 |
+
"learning_rate": 0.0002,
|
| 1254 |
+
"loss": 1.686,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.008941505569708777,
|
| 1259 |
+
"grad_norm": 0.3125,
|
| 1260 |
+
"learning_rate": 0.0002,
|
| 1261 |
+
"loss": 1.6602,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.008991458114790948,
|
| 1266 |
+
"grad_norm": 0.32421875,
|
| 1267 |
+
"learning_rate": 0.0002,
|
| 1268 |
+
"loss": 1.6997,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.009041410659873121,
|
| 1273 |
+
"grad_norm": 0.291015625,
|
| 1274 |
+
"learning_rate": 0.0002,
|
| 1275 |
+
"loss": 1.6798,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.009091363204955292,
|
| 1280 |
+
"grad_norm": 0.5546875,
|
| 1281 |
+
"learning_rate": 0.0002,
|
| 1282 |
+
"loss": 1.6876,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.009141315750037465,
|
| 1287 |
+
"grad_norm": 0.55078125,
|
| 1288 |
+
"learning_rate": 0.0002,
|
| 1289 |
+
"loss": 1.7072,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.009191268295119636,
|
| 1294 |
+
"grad_norm": 0.3203125,
|
| 1295 |
+
"learning_rate": 0.0002,
|
| 1296 |
+
"loss": 1.6438,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.009241220840201809,
|
| 1301 |
+
"grad_norm": 0.37109375,
|
| 1302 |
+
"learning_rate": 0.0002,
|
| 1303 |
+
"loss": 1.646,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.00929117338528398,
|
| 1308 |
+
"grad_norm": 0.30078125,
|
| 1309 |
+
"learning_rate": 0.0002,
|
| 1310 |
+
"loss": 1.6515,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.009341125930366153,
|
| 1315 |
+
"grad_norm": 0.28125,
|
| 1316 |
+
"learning_rate": 0.0002,
|
| 1317 |
+
"loss": 1.6847,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.009391078475448324,
|
| 1322 |
+
"grad_norm": 0.31640625,
|
| 1323 |
+
"learning_rate": 0.0002,
|
| 1324 |
+
"loss": 1.6676,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.009441031020530496,
|
| 1329 |
+
"grad_norm": 0.283203125,
|
| 1330 |
+
"learning_rate": 0.0002,
|
| 1331 |
+
"loss": 1.5783,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.009490983565612668,
|
| 1336 |
+
"grad_norm": 0.318359375,
|
| 1337 |
+
"learning_rate": 0.0002,
|
| 1338 |
+
"loss": 1.6475,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.00954093611069484,
|
| 1343 |
+
"grad_norm": 0.333984375,
|
| 1344 |
+
"learning_rate": 0.0002,
|
| 1345 |
+
"loss": 1.6159,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.009590888655777011,
|
| 1350 |
+
"grad_norm": 0.34765625,
|
| 1351 |
+
"learning_rate": 0.0002,
|
| 1352 |
+
"loss": 1.63,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.009640841200859184,
|
| 1357 |
+
"grad_norm": 0.416015625,
|
| 1358 |
+
"learning_rate": 0.0002,
|
| 1359 |
+
"loss": 1.6873,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.009690793745941355,
|
| 1364 |
+
"grad_norm": 0.30078125,
|
| 1365 |
+
"learning_rate": 0.0002,
|
| 1366 |
+
"loss": 1.5986,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.009740746291023528,
|
| 1371 |
+
"grad_norm": 0.427734375,
|
| 1372 |
+
"learning_rate": 0.0002,
|
| 1373 |
+
"loss": 1.6463,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.009790698836105699,
|
| 1378 |
+
"grad_norm": 0.349609375,
|
| 1379 |
+
"learning_rate": 0.0002,
|
| 1380 |
+
"loss": 1.6545,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.009840651381187872,
|
| 1385 |
+
"grad_norm": 0.326171875,
|
| 1386 |
+
"learning_rate": 0.0002,
|
| 1387 |
+
"loss": 1.6332,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.009890603926270043,
|
| 1392 |
+
"grad_norm": 0.314453125,
|
| 1393 |
+
"learning_rate": 0.0002,
|
| 1394 |
+
"loss": 1.6229,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.009940556471352216,
|
| 1399 |
+
"grad_norm": 0.283203125,
|
| 1400 |
+
"learning_rate": 0.0002,
|
| 1401 |
+
"loss": 1.6052,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.009990509016434387,
|
| 1406 |
+
"grad_norm": 0.2890625,
|
| 1407 |
+
"learning_rate": 0.0002,
|
| 1408 |
+
"loss": 1.6108,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.01004046156151656,
|
| 1413 |
+
"grad_norm": 0.302734375,
|
| 1414 |
+
"learning_rate": 0.0002,
|
| 1415 |
+
"loss": 1.5585,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.01009041410659873,
|
| 1420 |
+
"grad_norm": 0.3203125,
|
| 1421 |
+
"learning_rate": 0.0002,
|
| 1422 |
+
"loss": 1.654,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.010140366651680903,
|
| 1427 |
+
"grad_norm": 0.328125,
|
| 1428 |
+
"learning_rate": 0.0002,
|
| 1429 |
+
"loss": 1.6022,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.010190319196763074,
|
| 1434 |
+
"grad_norm": 0.51171875,
|
| 1435 |
+
"learning_rate": 0.0002,
|
| 1436 |
+
"loss": 1.635,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.010240271741845247,
|
| 1441 |
+
"grad_norm": 0.3359375,
|
| 1442 |
+
"learning_rate": 0.0002,
|
| 1443 |
+
"loss": 1.6118,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.010290224286927418,
|
| 1448 |
+
"grad_norm": 0.60546875,
|
| 1449 |
+
"learning_rate": 0.0002,
|
| 1450 |
+
"loss": 1.66,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.010340176832009591,
|
| 1455 |
+
"grad_norm": 0.470703125,
|
| 1456 |
+
"learning_rate": 0.0002,
|
| 1457 |
+
"loss": 1.6917,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.010390129377091762,
|
| 1462 |
+
"grad_norm": 0.3125,
|
| 1463 |
+
"learning_rate": 0.0002,
|
| 1464 |
+
"loss": 1.6195,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.010440081922173935,
|
| 1469 |
+
"grad_norm": 0.369140625,
|
| 1470 |
+
"learning_rate": 0.0002,
|
| 1471 |
+
"loss": 1.6828,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.010490034467256106,
|
| 1476 |
+
"grad_norm": 0.376953125,
|
| 1477 |
+
"learning_rate": 0.0002,
|
| 1478 |
+
"loss": 1.5738,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.010539987012338279,
|
| 1483 |
+
"grad_norm": 0.291015625,
|
| 1484 |
+
"learning_rate": 0.0002,
|
| 1485 |
+
"loss": 1.5971,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.010589939557420451,
|
| 1490 |
+
"grad_norm": 0.3046875,
|
| 1491 |
+
"learning_rate": 0.0002,
|
| 1492 |
+
"loss": 1.6603,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.010639892102502622,
|
| 1497 |
+
"grad_norm": 0.291015625,
|
| 1498 |
+
"learning_rate": 0.0002,
|
| 1499 |
+
"loss": 1.652,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.010689844647584795,
|
| 1504 |
+
"grad_norm": 0.3828125,
|
| 1505 |
+
"learning_rate": 0.0002,
|
| 1506 |
+
"loss": 1.5818,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.010739797192666966,
|
| 1511 |
+
"grad_norm": 0.318359375,
|
| 1512 |
+
"learning_rate": 0.0002,
|
| 1513 |
+
"loss": 1.5863,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.010789749737749139,
|
| 1518 |
+
"grad_norm": 0.318359375,
|
| 1519 |
+
"learning_rate": 0.0002,
|
| 1520 |
+
"loss": 1.6022,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.01083970228283131,
|
| 1525 |
+
"grad_norm": 0.27734375,
|
| 1526 |
+
"learning_rate": 0.0002,
|
| 1527 |
+
"loss": 1.6294,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.010889654827913483,
|
| 1532 |
+
"grad_norm": 0.30859375,
|
| 1533 |
+
"learning_rate": 0.0002,
|
| 1534 |
+
"loss": 1.6128,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.010939607372995654,
|
| 1539 |
+
"grad_norm": 0.337890625,
|
| 1540 |
+
"learning_rate": 0.0002,
|
| 1541 |
+
"loss": 1.5839,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.010989559918077827,
|
| 1546 |
+
"grad_norm": 0.306640625,
|
| 1547 |
+
"learning_rate": 0.0002,
|
| 1548 |
+
"loss": 1.5474,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.011039512463159998,
|
| 1553 |
+
"grad_norm": 0.333984375,
|
| 1554 |
+
"learning_rate": 0.0002,
|
| 1555 |
+
"loss": 1.5484,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.01108946500824217,
|
| 1560 |
+
"grad_norm": 0.2890625,
|
| 1561 |
+
"learning_rate": 0.0002,
|
| 1562 |
+
"loss": 1.5365,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.011139417553324342,
|
| 1567 |
+
"grad_norm": 0.48046875,
|
| 1568 |
+
"learning_rate": 0.0002,
|
| 1569 |
+
"loss": 1.5604,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.011189370098406514,
|
| 1574 |
+
"grad_norm": 0.44140625,
|
| 1575 |
+
"learning_rate": 0.0002,
|
| 1576 |
+
"loss": 1.5072,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.011239322643488685,
|
| 1581 |
+
"grad_norm": 0.29296875,
|
| 1582 |
+
"learning_rate": 0.0002,
|
| 1583 |
+
"loss": 1.5259,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.011289275188570858,
|
| 1588 |
+
"grad_norm": 0.625,
|
| 1589 |
+
"learning_rate": 0.0002,
|
| 1590 |
+
"loss": 1.6972,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.01133922773365303,
|
| 1595 |
+
"grad_norm": 0.80078125,
|
| 1596 |
+
"learning_rate": 0.0002,
|
| 1597 |
+
"loss": 1.6205,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.011389180278735202,
|
| 1602 |
+
"grad_norm": 0.35546875,
|
| 1603 |
+
"learning_rate": 0.0002,
|
| 1604 |
+
"loss": 1.6108,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.011439132823817373,
|
| 1609 |
+
"grad_norm": 0.28515625,
|
| 1610 |
+
"learning_rate": 0.0002,
|
| 1611 |
+
"loss": 1.6078,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.011489085368899546,
|
| 1616 |
+
"grad_norm": 0.2890625,
|
| 1617 |
+
"learning_rate": 0.0002,
|
| 1618 |
+
"loss": 1.6216,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.011539037913981717,
|
| 1623 |
+
"grad_norm": 0.353515625,
|
| 1624 |
+
"learning_rate": 0.0002,
|
| 1625 |
+
"loss": 1.5713,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.01158899045906389,
|
| 1630 |
+
"grad_norm": 0.361328125,
|
| 1631 |
+
"learning_rate": 0.0002,
|
| 1632 |
+
"loss": 1.5866,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.01163894300414606,
|
| 1637 |
+
"grad_norm": 0.482421875,
|
| 1638 |
+
"learning_rate": 0.0002,
|
| 1639 |
+
"loss": 1.6194,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.011688895549228234,
|
| 1644 |
+
"grad_norm": 0.84765625,
|
| 1645 |
+
"learning_rate": 0.0002,
|
| 1646 |
+
"loss": 1.615,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.011738848094310405,
|
| 1651 |
+
"grad_norm": 0.392578125,
|
| 1652 |
+
"learning_rate": 0.0002,
|
| 1653 |
+
"loss": 1.5403,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.011788800639392577,
|
| 1658 |
+
"grad_norm": 0.4296875,
|
| 1659 |
+
"learning_rate": 0.0002,
|
| 1660 |
+
"loss": 1.5468,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.011838753184474748,
|
| 1665 |
+
"grad_norm": 0.384765625,
|
| 1666 |
+
"learning_rate": 0.0002,
|
| 1667 |
+
"loss": 1.5983,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.011888705729556921,
|
| 1672 |
+
"grad_norm": 0.466796875,
|
| 1673 |
+
"learning_rate": 0.0002,
|
| 1674 |
+
"loss": 1.4359,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.011938658274639092,
|
| 1679 |
+
"grad_norm": 0.51953125,
|
| 1680 |
+
"learning_rate": 0.0002,
|
| 1681 |
+
"loss": 1.6278,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.011988610819721265,
|
| 1686 |
+
"grad_norm": 0.52734375,
|
| 1687 |
+
"learning_rate": 0.0002,
|
| 1688 |
+
"loss": 1.6101,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.012038563364803436,
|
| 1693 |
+
"grad_norm": 0.58984375,
|
| 1694 |
+
"learning_rate": 0.0002,
|
| 1695 |
+
"loss": 1.5788,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.012088515909885609,
|
| 1700 |
+
"grad_norm": 1.328125,
|
| 1701 |
+
"learning_rate": 0.0002,
|
| 1702 |
+
"loss": 1.6447,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.01213846845496778,
|
| 1707 |
+
"grad_norm": 0.431640625,
|
| 1708 |
+
"learning_rate": 0.0002,
|
| 1709 |
+
"loss": 1.5342,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.012188421000049953,
|
| 1714 |
+
"grad_norm": 0.96484375,
|
| 1715 |
+
"learning_rate": 0.0002,
|
| 1716 |
+
"loss": 1.5754,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.012238373545132124,
|
| 1721 |
+
"grad_norm": 0.302734375,
|
| 1722 |
+
"learning_rate": 0.0002,
|
| 1723 |
+
"loss": 1.6004,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.012288326090214297,
|
| 1728 |
+
"grad_norm": 0.93359375,
|
| 1729 |
+
"learning_rate": 0.0002,
|
| 1730 |
+
"loss": 1.6194,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.012338278635296468,
|
| 1735 |
+
"grad_norm": 0.36328125,
|
| 1736 |
+
"learning_rate": 0.0002,
|
| 1737 |
+
"loss": 1.5661,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.01238823118037864,
|
| 1742 |
+
"grad_norm": 0.5,
|
| 1743 |
+
"learning_rate": 0.0002,
|
| 1744 |
+
"loss": 1.6007,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.012438183725460811,
|
| 1749 |
+
"grad_norm": 0.328125,
|
| 1750 |
+
"learning_rate": 0.0002,
|
| 1751 |
+
"loss": 1.6114,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.012488136270542984,
|
| 1756 |
+
"grad_norm": 0.43359375,
|
| 1757 |
+
"learning_rate": 0.0002,
|
| 1758 |
+
"loss": 1.5924,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.012538088815625155,
|
| 1763 |
+
"grad_norm": 0.443359375,
|
| 1764 |
+
"learning_rate": 0.0002,
|
| 1765 |
+
"loss": 1.594,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.012588041360707328,
|
| 1770 |
+
"grad_norm": 0.296875,
|
| 1771 |
+
"learning_rate": 0.0002,
|
| 1772 |
+
"loss": 1.5704,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.0126379939057895,
|
| 1777 |
+
"grad_norm": 0.83203125,
|
| 1778 |
+
"learning_rate": 0.0002,
|
| 1779 |
+
"loss": 1.6637,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.012687946450871672,
|
| 1784 |
+
"grad_norm": 0.7421875,
|
| 1785 |
+
"learning_rate": 0.0002,
|
| 1786 |
+
"loss": 1.6046,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.012737898995953845,
|
| 1791 |
+
"grad_norm": 0.3515625,
|
| 1792 |
+
"learning_rate": 0.0002,
|
| 1793 |
+
"loss": 1.6011,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.012787851541036016,
|
| 1798 |
+
"grad_norm": 0.412109375,
|
| 1799 |
+
"learning_rate": 0.0002,
|
| 1800 |
+
"loss": 1.5344,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.012837804086118189,
|
| 1805 |
+
"grad_norm": 0.36328125,
|
| 1806 |
+
"learning_rate": 0.0002,
|
| 1807 |
+
"loss": 1.5352,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.01288775663120036,
|
| 1812 |
+
"grad_norm": 0.263671875,
|
| 1813 |
+
"learning_rate": 0.0002,
|
| 1814 |
+
"loss": 1.5997,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.012937709176282532,
|
| 1819 |
+
"grad_norm": 0.337890625,
|
| 1820 |
+
"learning_rate": 0.0002,
|
| 1821 |
+
"loss": 1.5496,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.012987661721364703,
|
| 1826 |
+
"grad_norm": 0.34375,
|
| 1827 |
+
"learning_rate": 0.0002,
|
| 1828 |
+
"loss": 1.5487,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.013037614266446876,
|
| 1833 |
+
"grad_norm": 0.265625,
|
| 1834 |
+
"learning_rate": 0.0002,
|
| 1835 |
+
"loss": 1.5494,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.013087566811529047,
|
| 1840 |
+
"grad_norm": 0.3125,
|
| 1841 |
+
"learning_rate": 0.0002,
|
| 1842 |
+
"loss": 1.4822,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.01313751935661122,
|
| 1847 |
+
"grad_norm": 0.2890625,
|
| 1848 |
+
"learning_rate": 0.0002,
|
| 1849 |
+
"loss": 1.6164,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.013187471901693391,
|
| 1854 |
+
"grad_norm": 0.294921875,
|
| 1855 |
+
"learning_rate": 0.0002,
|
| 1856 |
+
"loss": 1.4879,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.013237424446775564,
|
| 1861 |
+
"grad_norm": 0.263671875,
|
| 1862 |
+
"learning_rate": 0.0002,
|
| 1863 |
+
"loss": 1.6085,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.013287376991857735,
|
| 1868 |
+
"grad_norm": 0.294921875,
|
| 1869 |
+
"learning_rate": 0.0002,
|
| 1870 |
+
"loss": 1.5786,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.013337329536939908,
|
| 1875 |
+
"grad_norm": 0.46484375,
|
| 1876 |
+
"learning_rate": 0.0002,
|
| 1877 |
+
"loss": 1.5585,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.013387282082022079,
|
| 1882 |
+
"grad_norm": 0.267578125,
|
| 1883 |
+
"learning_rate": 0.0002,
|
| 1884 |
+
"loss": 1.5732,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.013437234627104251,
|
| 1889 |
+
"grad_norm": 0.26953125,
|
| 1890 |
+
"learning_rate": 0.0002,
|
| 1891 |
+
"loss": 1.6396,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.013487187172186423,
|
| 1896 |
+
"grad_norm": 0.435546875,
|
| 1897 |
+
"learning_rate": 0.0002,
|
| 1898 |
+
"loss": 1.5882,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.013537139717268595,
|
| 1903 |
+
"grad_norm": 0.26953125,
|
| 1904 |
+
"learning_rate": 0.0002,
|
| 1905 |
+
"loss": 1.4639,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.013587092262350766,
|
| 1910 |
+
"grad_norm": 0.345703125,
|
| 1911 |
+
"learning_rate": 0.0002,
|
| 1912 |
+
"loss": 1.5626,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.01363704480743294,
|
| 1917 |
+
"grad_norm": 0.4140625,
|
| 1918 |
+
"learning_rate": 0.0002,
|
| 1919 |
+
"loss": 1.6264,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.01368699735251511,
|
| 1924 |
+
"grad_norm": 0.474609375,
|
| 1925 |
+
"learning_rate": 0.0002,
|
| 1926 |
+
"loss": 1.4873,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.013736949897597283,
|
| 1931 |
+
"grad_norm": 0.298828125,
|
| 1932 |
+
"learning_rate": 0.0002,
|
| 1933 |
+
"loss": 1.5403,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.013786902442679454,
|
| 1938 |
+
"grad_norm": 0.28515625,
|
| 1939 |
+
"learning_rate": 0.0002,
|
| 1940 |
+
"loss": 1.5397,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.013836854987761627,
|
| 1945 |
+
"grad_norm": 0.333984375,
|
| 1946 |
+
"learning_rate": 0.0002,
|
| 1947 |
+
"loss": 1.471,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.013886807532843798,
|
| 1952 |
+
"grad_norm": 0.322265625,
|
| 1953 |
+
"learning_rate": 0.0002,
|
| 1954 |
+
"loss": 1.5636,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.01393676007792597,
|
| 1959 |
+
"grad_norm": 0.341796875,
|
| 1960 |
+
"learning_rate": 0.0002,
|
| 1961 |
+
"loss": 1.4521,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.013986712623008142,
|
| 1966 |
+
"grad_norm": 0.400390625,
|
| 1967 |
+
"learning_rate": 0.0002,
|
| 1968 |
+
"loss": 1.5053,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.014036665168090314,
|
| 1973 |
+
"grad_norm": 0.263671875,
|
| 1974 |
+
"learning_rate": 0.0002,
|
| 1975 |
+
"loss": 1.5575,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.014086617713172486,
|
| 1980 |
+
"grad_norm": 0.4921875,
|
| 1981 |
+
"learning_rate": 0.0002,
|
| 1982 |
+
"loss": 1.6155,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.014136570258254658,
|
| 1987 |
+
"grad_norm": 0.337890625,
|
| 1988 |
+
"learning_rate": 0.0002,
|
| 1989 |
+
"loss": 1.4641,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.01418652280333683,
|
| 1994 |
+
"grad_norm": 0.396484375,
|
| 1995 |
+
"learning_rate": 0.0002,
|
| 1996 |
+
"loss": 1.5408,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.014236475348419002,
|
| 2001 |
+
"grad_norm": 0.298828125,
|
| 2002 |
+
"learning_rate": 0.0002,
|
| 2003 |
+
"loss": 1.4735,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.014286427893501173,
|
| 2008 |
+
"grad_norm": 0.298828125,
|
| 2009 |
+
"learning_rate": 0.0002,
|
| 2010 |
+
"loss": 1.5263,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.014336380438583346,
|
| 2015 |
+
"grad_norm": 0.306640625,
|
| 2016 |
+
"learning_rate": 0.0002,
|
| 2017 |
+
"loss": 1.5367,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.014386332983665517,
|
| 2022 |
+
"grad_norm": 0.279296875,
|
| 2023 |
+
"learning_rate": 0.0002,
|
| 2024 |
+
"loss": 1.5946,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.01443628552874769,
|
| 2029 |
+
"grad_norm": 0.296875,
|
| 2030 |
+
"learning_rate": 0.0002,
|
| 2031 |
+
"loss": 1.4774,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.01448623807382986,
|
| 2036 |
+
"grad_norm": 0.26171875,
|
| 2037 |
+
"learning_rate": 0.0002,
|
| 2038 |
+
"loss": 1.4944,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.014536190618912034,
|
| 2043 |
+
"grad_norm": 0.2392578125,
|
| 2044 |
+
"learning_rate": 0.0002,
|
| 2045 |
+
"loss": 1.4827,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.014586143163994205,
|
| 2050 |
+
"grad_norm": 0.267578125,
|
| 2051 |
+
"learning_rate": 0.0002,
|
| 2052 |
+
"loss": 1.5828,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.014636095709076377,
|
| 2057 |
+
"grad_norm": 0.3046875,
|
| 2058 |
+
"learning_rate": 0.0002,
|
| 2059 |
+
"loss": 1.6001,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.014686048254158549,
|
| 2064 |
+
"grad_norm": 0.2734375,
|
| 2065 |
+
"learning_rate": 0.0002,
|
| 2066 |
+
"loss": 1.6227,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.014736000799240721,
|
| 2071 |
+
"grad_norm": 0.251953125,
|
| 2072 |
+
"learning_rate": 0.0002,
|
| 2073 |
+
"loss": 1.4904,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.014785953344322894,
|
| 2078 |
+
"grad_norm": 0.251953125,
|
| 2079 |
+
"learning_rate": 0.0002,
|
| 2080 |
+
"loss": 1.5204,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.014835905889405065,
|
| 2085 |
+
"grad_norm": 0.27734375,
|
| 2086 |
+
"learning_rate": 0.0002,
|
| 2087 |
+
"loss": 1.485,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.014885858434487238,
|
| 2092 |
+
"grad_norm": 0.251953125,
|
| 2093 |
+
"learning_rate": 0.0002,
|
| 2094 |
+
"loss": 1.6064,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.014935810979569409,
|
| 2099 |
+
"grad_norm": 0.279296875,
|
| 2100 |
+
"learning_rate": 0.0002,
|
| 2101 |
+
"loss": 1.4897,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.014985763524651582,
|
| 2106 |
+
"grad_norm": 0.29296875,
|
| 2107 |
+
"learning_rate": 0.0002,
|
| 2108 |
+
"loss": 1.5404,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.015035716069733753,
|
| 2113 |
+
"grad_norm": 0.26953125,
|
| 2114 |
+
"learning_rate": 0.0002,
|
| 2115 |
+
"loss": 1.5293,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.015085668614815926,
|
| 2120 |
+
"grad_norm": 0.28125,
|
| 2121 |
+
"learning_rate": 0.0002,
|
| 2122 |
+
"loss": 1.4464,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.015135621159898097,
|
| 2127 |
+
"grad_norm": 0.28125,
|
| 2128 |
+
"learning_rate": 0.0002,
|
| 2129 |
+
"loss": 1.4699,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.01518557370498027,
|
| 2134 |
+
"grad_norm": 0.2490234375,
|
| 2135 |
+
"learning_rate": 0.0002,
|
| 2136 |
+
"loss": 1.6375,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.01523552625006244,
|
| 2141 |
+
"grad_norm": 0.26171875,
|
| 2142 |
+
"learning_rate": 0.0002,
|
| 2143 |
+
"loss": 1.5147,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.015285478795144613,
|
| 2148 |
+
"grad_norm": 0.234375,
|
| 2149 |
+
"learning_rate": 0.0002,
|
| 2150 |
+
"loss": 1.5629,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.015335431340226784,
|
| 2155 |
+
"grad_norm": 0.234375,
|
| 2156 |
+
"learning_rate": 0.0002,
|
| 2157 |
+
"loss": 1.5722,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.015385383885308957,
|
| 2162 |
+
"grad_norm": 0.279296875,
|
| 2163 |
+
"learning_rate": 0.0002,
|
| 2164 |
+
"loss": 1.4906,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.015435336430391128,
|
| 2169 |
+
"grad_norm": 0.248046875,
|
| 2170 |
+
"learning_rate": 0.0002,
|
| 2171 |
+
"loss": 1.4308,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.015485288975473301,
|
| 2176 |
+
"grad_norm": 0.345703125,
|
| 2177 |
+
"learning_rate": 0.0002,
|
| 2178 |
+
"loss": 1.6079,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.015535241520555472,
|
| 2183 |
+
"grad_norm": 0.357421875,
|
| 2184 |
+
"learning_rate": 0.0002,
|
| 2185 |
+
"loss": 1.4713,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.015585194065637645,
|
| 2190 |
+
"grad_norm": 0.359375,
|
| 2191 |
+
"learning_rate": 0.0002,
|
| 2192 |
+
"loss": 1.3843,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.015635146610719818,
|
| 2197 |
+
"grad_norm": 0.2734375,
|
| 2198 |
+
"learning_rate": 0.0002,
|
| 2199 |
+
"loss": 1.5325,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.015685099155801987,
|
| 2204 |
+
"grad_norm": 0.3046875,
|
| 2205 |
+
"learning_rate": 0.0002,
|
| 2206 |
+
"loss": 1.5379,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.01573505170088416,
|
| 2211 |
+
"grad_norm": 0.369140625,
|
| 2212 |
+
"learning_rate": 0.0002,
|
| 2213 |
+
"loss": 1.4443,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.015785004245966332,
|
| 2218 |
+
"grad_norm": 0.2734375,
|
| 2219 |
+
"learning_rate": 0.0002,
|
| 2220 |
+
"loss": 1.4831,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.015834956791048505,
|
| 2225 |
+
"grad_norm": 0.318359375,
|
| 2226 |
+
"learning_rate": 0.0002,
|
| 2227 |
+
"loss": 1.578,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.015884909336130675,
|
| 2232 |
+
"grad_norm": 0.271484375,
|
| 2233 |
+
"learning_rate": 0.0002,
|
| 2234 |
+
"loss": 1.5371,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.015934861881212847,
|
| 2239 |
+
"grad_norm": 0.322265625,
|
| 2240 |
+
"learning_rate": 0.0002,
|
| 2241 |
+
"loss": 1.5143,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.01598481442629502,
|
| 2246 |
+
"grad_norm": 0.390625,
|
| 2247 |
+
"learning_rate": 0.0002,
|
| 2248 |
+
"loss": 1.5948,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.016034766971377193,
|
| 2253 |
+
"grad_norm": 0.380859375,
|
| 2254 |
+
"learning_rate": 0.0002,
|
| 2255 |
+
"loss": 1.4966,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.016084719516459362,
|
| 2260 |
+
"grad_norm": 0.298828125,
|
| 2261 |
+
"learning_rate": 0.0002,
|
| 2262 |
+
"loss": 1.5,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.016134672061541535,
|
| 2267 |
+
"grad_norm": 0.6328125,
|
| 2268 |
+
"learning_rate": 0.0002,
|
| 2269 |
+
"loss": 1.5381,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.016184624606623708,
|
| 2274 |
+
"grad_norm": 0.359375,
|
| 2275 |
+
"learning_rate": 0.0002,
|
| 2276 |
+
"loss": 1.468,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.01623457715170588,
|
| 2281 |
+
"grad_norm": 0.6875,
|
| 2282 |
+
"learning_rate": 0.0002,
|
| 2283 |
+
"loss": 1.6085,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.01628452969678805,
|
| 2288 |
+
"grad_norm": 0.82421875,
|
| 2289 |
+
"learning_rate": 0.0002,
|
| 2290 |
+
"loss": 1.5044,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.016334482241870223,
|
| 2295 |
+
"grad_norm": 0.48828125,
|
| 2296 |
+
"learning_rate": 0.0002,
|
| 2297 |
+
"loss": 1.4576,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.016384434786952395,
|
| 2302 |
+
"grad_norm": 1.03125,
|
| 2303 |
+
"learning_rate": 0.0002,
|
| 2304 |
+
"loss": 1.555,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.016434387332034568,
|
| 2309 |
+
"grad_norm": 0.267578125,
|
| 2310 |
+
"learning_rate": 0.0002,
|
| 2311 |
+
"loss": 1.4362,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.016484339877116738,
|
| 2316 |
+
"grad_norm": 0.703125,
|
| 2317 |
+
"learning_rate": 0.0002,
|
| 2318 |
+
"loss": 1.5176,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.01653429242219891,
|
| 2323 |
+
"grad_norm": 0.27734375,
|
| 2324 |
+
"learning_rate": 0.0002,
|
| 2325 |
+
"loss": 1.4686,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.016584244967281083,
|
| 2330 |
+
"grad_norm": 0.52734375,
|
| 2331 |
+
"learning_rate": 0.0002,
|
| 2332 |
+
"loss": 1.7015,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.016634197512363256,
|
| 2337 |
+
"grad_norm": 0.28125,
|
| 2338 |
+
"learning_rate": 0.0002,
|
| 2339 |
+
"loss": 1.472,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.016684150057445425,
|
| 2344 |
+
"grad_norm": 0.447265625,
|
| 2345 |
+
"learning_rate": 0.0002,
|
| 2346 |
+
"loss": 1.453,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.016734102602527598,
|
| 2351 |
+
"grad_norm": 0.2412109375,
|
| 2352 |
+
"learning_rate": 0.0002,
|
| 2353 |
+
"loss": 1.4794,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.01678405514760977,
|
| 2358 |
+
"grad_norm": 0.318359375,
|
| 2359 |
+
"learning_rate": 0.0002,
|
| 2360 |
+
"loss": 1.4526,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.016834007692691944,
|
| 2365 |
+
"grad_norm": 0.32421875,
|
| 2366 |
+
"learning_rate": 0.0002,
|
| 2367 |
+
"loss": 1.5541,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.016883960237774116,
|
| 2372 |
+
"grad_norm": 0.25390625,
|
| 2373 |
+
"learning_rate": 0.0002,
|
| 2374 |
+
"loss": 1.5544,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.016933912782856286,
|
| 2379 |
+
"grad_norm": 0.32421875,
|
| 2380 |
+
"learning_rate": 0.0002,
|
| 2381 |
+
"loss": 1.5339,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.01698386532793846,
|
| 2386 |
+
"grad_norm": 0.255859375,
|
| 2387 |
+
"learning_rate": 0.0002,
|
| 2388 |
+
"loss": 1.4796,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.01703381787302063,
|
| 2393 |
+
"grad_norm": 0.26171875,
|
| 2394 |
+
"learning_rate": 0.0002,
|
| 2395 |
+
"loss": 1.4852,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.017083770418102804,
|
| 2400 |
+
"grad_norm": 0.2470703125,
|
| 2401 |
+
"learning_rate": 0.0002,
|
| 2402 |
+
"loss": 1.4896,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.017133722963184973,
|
| 2407 |
+
"grad_norm": 0.27734375,
|
| 2408 |
+
"learning_rate": 0.0002,
|
| 2409 |
+
"loss": 1.5921,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.017183675508267146,
|
| 2414 |
+
"grad_norm": 0.322265625,
|
| 2415 |
+
"learning_rate": 0.0002,
|
| 2416 |
+
"loss": 1.5798,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.01723362805334932,
|
| 2421 |
+
"grad_norm": 0.4765625,
|
| 2422 |
+
"learning_rate": 0.0002,
|
| 2423 |
+
"loss": 1.5625,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.01728358059843149,
|
| 2428 |
+
"grad_norm": 0.53125,
|
| 2429 |
+
"learning_rate": 0.0002,
|
| 2430 |
+
"loss": 1.5253,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.01733353314351366,
|
| 2435 |
+
"grad_norm": 0.6796875,
|
| 2436 |
+
"learning_rate": 0.0002,
|
| 2437 |
+
"loss": 1.5444,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.017383485688595834,
|
| 2442 |
+
"grad_norm": 1.09375,
|
| 2443 |
+
"learning_rate": 0.0002,
|
| 2444 |
+
"loss": 1.4605,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.017433438233678007,
|
| 2449 |
+
"grad_norm": 0.546875,
|
| 2450 |
+
"learning_rate": 0.0002,
|
| 2451 |
+
"loss": 1.4357,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.01748339077876018,
|
| 2456 |
+
"grad_norm": 1.28125,
|
| 2457 |
+
"learning_rate": 0.0002,
|
| 2458 |
+
"loss": 1.5956,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.01753334332384235,
|
| 2463 |
+
"grad_norm": 0.94921875,
|
| 2464 |
+
"learning_rate": 0.0002,
|
| 2465 |
+
"loss": 1.5409,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.01758329586892452,
|
| 2470 |
+
"grad_norm": 2.796875,
|
| 2471 |
+
"learning_rate": 0.0002,
|
| 2472 |
+
"loss": 1.5776,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.017633248414006694,
|
| 2477 |
+
"grad_norm": 2.953125,
|
| 2478 |
+
"learning_rate": 0.0002,
|
| 2479 |
+
"loss": 1.5703,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.017683200959088867,
|
| 2484 |
+
"grad_norm": 0.57421875,
|
| 2485 |
+
"learning_rate": 0.0002,
|
| 2486 |
+
"loss": 1.4604,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.017733153504171036,
|
| 2491 |
+
"grad_norm": 2.5,
|
| 2492 |
+
"learning_rate": 0.0002,
|
| 2493 |
+
"loss": 1.6811,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.01778310604925321,
|
| 2498 |
+
"grad_norm": 1.9921875,
|
| 2499 |
+
"learning_rate": 0.0002,
|
| 2500 |
+
"loss": 1.5956,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.017833058594335382,
|
| 2505 |
+
"grad_norm": 0.63671875,
|
| 2506 |
+
"learning_rate": 0.0002,
|
| 2507 |
+
"loss": 1.4853,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.017883011139417555,
|
| 2512 |
+
"grad_norm": 0.95703125,
|
| 2513 |
+
"learning_rate": 0.0002,
|
| 2514 |
+
"loss": 1.5039,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.017932963684499724,
|
| 2519 |
+
"grad_norm": 2.46875,
|
| 2520 |
+
"learning_rate": 0.0002,
|
| 2521 |
+
"loss": 1.5961,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.017982916229581897,
|
| 2526 |
+
"grad_norm": 0.76171875,
|
| 2527 |
+
"learning_rate": 0.0002,
|
| 2528 |
+
"loss": 1.6429,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.01803286877466407,
|
| 2533 |
+
"grad_norm": 1.7421875,
|
| 2534 |
+
"learning_rate": 0.0002,
|
| 2535 |
+
"loss": 1.5974,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.018082821319746242,
|
| 2540 |
+
"grad_norm": 1.234375,
|
| 2541 |
+
"learning_rate": 0.0002,
|
| 2542 |
+
"loss": 1.4604,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.01813277386482841,
|
| 2547 |
+
"grad_norm": 0.51171875,
|
| 2548 |
+
"learning_rate": 0.0002,
|
| 2549 |
+
"loss": 1.521,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.018182726409910584,
|
| 2554 |
+
"grad_norm": 0.61328125,
|
| 2555 |
+
"learning_rate": 0.0002,
|
| 2556 |
+
"loss": 1.4228,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.018232678954992757,
|
| 2561 |
+
"grad_norm": 1.25,
|
| 2562 |
+
"learning_rate": 0.0002,
|
| 2563 |
+
"loss": 1.5559,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.01828263150007493,
|
| 2568 |
+
"grad_norm": 0.67578125,
|
| 2569 |
+
"learning_rate": 0.0002,
|
| 2570 |
+
"loss": 1.4612,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.0183325840451571,
|
| 2575 |
+
"grad_norm": 0.99609375,
|
| 2576 |
+
"learning_rate": 0.0002,
|
| 2577 |
+
"loss": 1.5216,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.018382536590239272,
|
| 2582 |
+
"grad_norm": 0.7109375,
|
| 2583 |
+
"learning_rate": 0.0002,
|
| 2584 |
+
"loss": 1.4677,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.018432489135321445,
|
| 2589 |
+
"grad_norm": 0.490234375,
|
| 2590 |
+
"learning_rate": 0.0002,
|
| 2591 |
+
"loss": 1.4851,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.018482441680403618,
|
| 2596 |
+
"grad_norm": 0.62890625,
|
| 2597 |
+
"learning_rate": 0.0002,
|
| 2598 |
+
"loss": 1.5265,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.018532394225485787,
|
| 2603 |
+
"grad_norm": 0.6953125,
|
| 2604 |
+
"learning_rate": 0.0002,
|
| 2605 |
+
"loss": 1.5865,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.01858234677056796,
|
| 2610 |
+
"grad_norm": 0.345703125,
|
| 2611 |
+
"learning_rate": 0.0002,
|
| 2612 |
+
"loss": 1.5052,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.018632299315650133,
|
| 2617 |
+
"grad_norm": 0.376953125,
|
| 2618 |
+
"learning_rate": 0.0002,
|
| 2619 |
+
"loss": 1.4114,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.018682251860732305,
|
| 2624 |
+
"grad_norm": 0.79296875,
|
| 2625 |
+
"learning_rate": 0.0002,
|
| 2626 |
+
"loss": 1.4811,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.018732204405814475,
|
| 2631 |
+
"grad_norm": 0.328125,
|
| 2632 |
+
"learning_rate": 0.0002,
|
| 2633 |
+
"loss": 1.431,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.018782156950896647,
|
| 2638 |
+
"grad_norm": 0.5234375,
|
| 2639 |
+
"learning_rate": 0.0002,
|
| 2640 |
+
"loss": 1.4495,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.01883210949597882,
|
| 2645 |
+
"grad_norm": 0.49609375,
|
| 2646 |
+
"learning_rate": 0.0002,
|
| 2647 |
+
"loss": 1.5299,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.018882062041060993,
|
| 2652 |
+
"grad_norm": 0.330078125,
|
| 2653 |
+
"learning_rate": 0.0002,
|
| 2654 |
+
"loss": 1.6144,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.018932014586143162,
|
| 2659 |
+
"grad_norm": 0.486328125,
|
| 2660 |
+
"learning_rate": 0.0002,
|
| 2661 |
+
"loss": 1.4537,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.018981967131225335,
|
| 2666 |
+
"grad_norm": 0.65234375,
|
| 2667 |
+
"learning_rate": 0.0002,
|
| 2668 |
+
"loss": 1.4466,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.019031919676307508,
|
| 2673 |
+
"grad_norm": 0.26953125,
|
| 2674 |
+
"learning_rate": 0.0002,
|
| 2675 |
+
"loss": 1.4474,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.01908187222138968,
|
| 2680 |
+
"grad_norm": 0.53515625,
|
| 2681 |
+
"learning_rate": 0.0002,
|
| 2682 |
+
"loss": 1.5835,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.019131824766471853,
|
| 2687 |
+
"grad_norm": 0.36328125,
|
| 2688 |
+
"learning_rate": 0.0002,
|
| 2689 |
+
"loss": 1.5146,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.019181777311554023,
|
| 2694 |
+
"grad_norm": 0.310546875,
|
| 2695 |
+
"learning_rate": 0.0002,
|
| 2696 |
+
"loss": 1.4287,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.019231729856636195,
|
| 2701 |
+
"grad_norm": 0.384765625,
|
| 2702 |
+
"learning_rate": 0.0002,
|
| 2703 |
+
"loss": 1.5497,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.01928168240171837,
|
| 2708 |
+
"grad_norm": 0.40234375,
|
| 2709 |
+
"learning_rate": 0.0002,
|
| 2710 |
+
"loss": 1.4779,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.01933163494680054,
|
| 2715 |
+
"grad_norm": 0.279296875,
|
| 2716 |
+
"learning_rate": 0.0002,
|
| 2717 |
+
"loss": 1.4196,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.01938158749188271,
|
| 2722 |
+
"grad_norm": 0.275390625,
|
| 2723 |
+
"learning_rate": 0.0002,
|
| 2724 |
+
"loss": 1.4635,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.019431540036964883,
|
| 2729 |
+
"grad_norm": 0.345703125,
|
| 2730 |
+
"learning_rate": 0.0002,
|
| 2731 |
+
"loss": 1.4386,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.019481492582047056,
|
| 2736 |
+
"grad_norm": 0.275390625,
|
| 2737 |
+
"learning_rate": 0.0002,
|
| 2738 |
+
"loss": 1.5325,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.01953144512712923,
|
| 2743 |
+
"grad_norm": 0.3125,
|
| 2744 |
+
"learning_rate": 0.0002,
|
| 2745 |
+
"loss": 1.5064,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.019581397672211398,
|
| 2750 |
+
"grad_norm": 0.220703125,
|
| 2751 |
+
"learning_rate": 0.0002,
|
| 2752 |
+
"loss": 1.5197,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.01963135021729357,
|
| 2757 |
+
"grad_norm": 0.208984375,
|
| 2758 |
+
"learning_rate": 0.0002,
|
| 2759 |
+
"loss": 1.554,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.019681302762375744,
|
| 2764 |
+
"grad_norm": 0.326171875,
|
| 2765 |
+
"learning_rate": 0.0002,
|
| 2766 |
+
"loss": 1.6138,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.019731255307457916,
|
| 2771 |
+
"grad_norm": 0.2333984375,
|
| 2772 |
+
"learning_rate": 0.0002,
|
| 2773 |
+
"loss": 1.5034,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.019781207852540086,
|
| 2778 |
+
"grad_norm": 0.294921875,
|
| 2779 |
+
"learning_rate": 0.0002,
|
| 2780 |
+
"loss": 1.5005,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.01983116039762226,
|
| 2785 |
+
"grad_norm": 0.2333984375,
|
| 2786 |
+
"learning_rate": 0.0002,
|
| 2787 |
+
"loss": 1.4746,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.01988111294270443,
|
| 2792 |
+
"grad_norm": 0.259765625,
|
| 2793 |
+
"learning_rate": 0.0002,
|
| 2794 |
+
"loss": 1.4957,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.019931065487786604,
|
| 2799 |
+
"grad_norm": 0.255859375,
|
| 2800 |
+
"learning_rate": 0.0002,
|
| 2801 |
+
"loss": 1.4178,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.019981018032868773,
|
| 2806 |
+
"grad_norm": 0.2216796875,
|
| 2807 |
+
"learning_rate": 0.0002,
|
| 2808 |
+
"loss": 1.5933,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
}
|
| 2811 |
+
],
|
| 2812 |
+
"logging_steps": 1.0,
|
| 2813 |
+
"max_steps": 12011400,
|
| 2814 |
+
"num_input_tokens_seen": 0,
|
| 2815 |
+
"num_train_epochs": 600,
|
| 2816 |
+
"save_steps": 50,
|
| 2817 |
+
"stateful_callbacks": {
|
| 2818 |
+
"TrainerControl": {
|
| 2819 |
+
"args": {
|
| 2820 |
+
"should_epoch_stop": false,
|
| 2821 |
+
"should_evaluate": false,
|
| 2822 |
+
"should_log": false,
|
| 2823 |
+
"should_save": true,
|
| 2824 |
+
"should_training_stop": false
|
| 2825 |
+
},
|
| 2826 |
+
"attributes": {}
|
| 2827 |
+
}
|
| 2828 |
+
},
|
| 2829 |
+
"total_flos": 8181056945520640.0,
|
| 2830 |
+
"train_batch_size": 16,
|
| 2831 |
+
"trial_name": null,
|
| 2832 |
+
"trial_params": null
|
| 2833 |
+
}
|
checkpoint-400/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c93d29dee29f05e035594dfb4a5fc4ff03df196cd27c500781551c8b657a625
|
| 3 |
+
size 5944
|