diff --git a/checkpoint-13000/config.json b/checkpoint-13000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..60a257df017ccff177cc7be1151b89b4bd290c54 --- /dev/null +++ b/checkpoint-13000/config.json @@ -0,0 +1,79 @@ +{ + "ar_steps": 1, + "architectures": [ + "DiffVLMDiffusion" + ], + "attention_dropout": 0.0, + "bos_token_id": 151643, + "condition_layer": -1, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "image_token_id": 151655, + "img_cross_attention_dim": 2048, + "img_diffuser_depth": 2, + "img_ffn_dim_multiplier": null, + "img_hidden_size": 1536, + "img_multiple_of": 256, + "img_norm_eps": 1e-05, + "img_num_attention_heads": 12, + "img_num_kv_heads": 12, + "img_qk_norm": true, + "in_channels": 32, + "initializer_range": 0.02, + "inject_img_diffuser": false, + "input_size": 32, + "intermediate_size": 8960, + "layer_group_size": 7, + "layerwise_start_idx": 0, + "lora_alpha": 128, + "lora_bias": "none", + "lora_dropout": 0.05, + "lora_enable": false, + "lora_r": 64, + "max_position_embeddings": 32768, + "max_window_layers": 28, + "model_type": "qwen2_vl", + "non_linearity": 1, + "norm_elementwise_affine": true, + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "patch_size": 2, + "repa_coeff": 0.1, + "repa_layers": "2", + "repa_shared": false, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sample_size": 128, + "sampling_steps": 28, + "sliding_window": null, + "tie_word_embeddings": true, + "torch_dtype": "bfloat16", + "transformers_version": "4.47.0", + "use_cache": true, + "use_repa": false, + "use_residual_attn": true, + "use_sliding_window": false, + "vae_path": "mit-han-lab/dc-ae-f32c32-sana-1.1-diffusers", + "video_token_id": 151656, + "vision_config": { + "hidden_size": 1536, + "in_chans": 3, + "model_type": "qwen2_vl", + "spatial_patch_size": 14 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 151936 +} diff --git a/checkpoint-13000/generation_config.json b/checkpoint-13000/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0e3d465da90cb0ab59d8ba7babdefb0e88fbfa6b --- /dev/null +++ b/checkpoint-13000/generation_config.json @@ -0,0 +1,6 @@ +{ + "_from_model_config": true, + "bos_token_id": 151643, + "eos_token_id": 151645, + "transformers_version": "4.47.0" +} diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_0_tench, Tinca tinca.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_0_tench, Tinca tinca.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4a327e4f96242576533b41d56ea309e53118351e Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_0_tench, Tinca tinca.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_10_brambling, Fringilla montifringilla.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_10_brambling, Fringilla montifringilla.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4f160df6633bbec6880fc6e659626ed7fbb23078 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_10_brambling, Fringilla montifringilla.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_11_goldfinch, Carduelis carduelis.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_11_goldfinch, Carduelis carduelis.jpg new file mode 100644 index 0000000000000000000000000000000000000000..59e49344871abe324590a53794a32f57e7ac5ba9 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_11_goldfinch, Carduelis carduelis.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_12_house finch, linnet, Carpodacus mexicanus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_12_house finch, linnet, Carpodacus mexicanus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..414ac12428f6edc31a3ed53071d5b5ccb0de55ad Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_12_house finch, linnet, Carpodacus mexicanus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_13_junco, snowbird.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_13_junco, snowbird.jpg new file mode 100644 index 0000000000000000000000000000000000000000..5fc618d11ed121739afb19eb623ceee170616844 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_13_junco, snowbird.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_14_indigo bunting, indigo finch, indigo bird, Passerina cyanea.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_14_indigo bunting, indigo finch, indigo bird, Passerina cyanea.jpg new file mode 100644 index 0000000000000000000000000000000000000000..0270824ce291592411b2021afd2c48b366a94c7e Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_14_indigo bunting, indigo finch, indigo bird, Passerina cyanea.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_15_robin, American robin, Turdus migratorius.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_15_robin, American robin, Turdus migratorius.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a32d26172cbfd11b64e301cb83959b958fc1c15e Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_15_robin, American robin, Turdus migratorius.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_16_bulbul.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_16_bulbul.jpg new file mode 100644 index 0000000000000000000000000000000000000000..40889c4fc00e803adebfed35e4e8ee8fc47759bf Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_16_bulbul.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_17_jay.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_17_jay.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f85d0695ad02005b2af8834285a960a369698d76 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_17_jay.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_18_magpie.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_18_magpie.jpg new file mode 100644 index 0000000000000000000000000000000000000000..7a511de3dc79f6a6db20c8eaf48c300f072f8bf2 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_18_magpie.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_19_chickadee.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_19_chickadee.jpg new file mode 100644 index 0000000000000000000000000000000000000000..0ffbd31eb30e7b05972f1511318a35e0645dcd6a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_19_chickadee.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_1_goldfish, Carassius auratus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_1_goldfish, Carassius auratus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..3408788482486d6f26afc049186780718edfdb44 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_1_goldfish, Carassius auratus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_20_water ouzel, dipper.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_20_water ouzel, dipper.jpg new file mode 100644 index 0000000000000000000000000000000000000000..49ef26d86c838d222f6da7360ccb1a7ce3cb256a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_20_water ouzel, dipper.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_21_kite.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_21_kite.jpg new file mode 100644 index 0000000000000000000000000000000000000000..35fe56df5ebe88c99f05905e650db7bb28289f88 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_21_kite.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_22_bald eagle, American eagle, Haliaeetus leucocephalus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_22_bald eagle, American eagle, Haliaeetus leucocephalus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..53ba23d8af453040f9f7b17b1849c83781063b05 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_22_bald eagle, American eagle, Haliaeetus leucocephalus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_23_vulture.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_23_vulture.jpg new file mode 100644 index 0000000000000000000000000000000000000000..5960545a71605011e0407e0399644fe290643d4b Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_23_vulture.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_24_great grey owl, great gray owl, Strix nebulosa.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_24_great grey owl, great gray owl, Strix nebulosa.jpg new file mode 100644 index 0000000000000000000000000000000000000000..39fbbca352de030868ef16ea8171487a4ee80983 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_24_great grey owl, great gray owl, Strix nebulosa.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_25_European fire salamander, Salamandra salamandra.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_25_European fire salamander, Salamandra salamandra.jpg new file mode 100644 index 0000000000000000000000000000000000000000..c5c00e02608d7a80557c79f3ad56204f2f3dc5e0 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_25_European fire salamander, Salamandra salamandra.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_26_common newt, Triturus vulgaris.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_26_common newt, Triturus vulgaris.jpg new file mode 100644 index 0000000000000000000000000000000000000000..97c6fc1d9182f1a21e114fbd2ea364bfc85efdb0 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_26_common newt, Triturus vulgaris.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_27_eft.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_27_eft.jpg new file mode 100644 index 0000000000000000000000000000000000000000..7e8415b6c4ecd72d16c11efdbac8c2027c4e2571 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_27_eft.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_28_spotted salamander, Ambystoma maculatum.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_28_spotted salamander, Ambystoma maculatum.jpg new file mode 100644 index 0000000000000000000000000000000000000000..3338cda1a124af25c33af7a578a5be60cd2b3bcf Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_28_spotted salamander, Ambystoma maculatum.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_29_axolotl, mud puppy, Ambystoma mexicanum.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_29_axolotl, mud puppy, Ambystoma mexicanum.jpg new file mode 100644 index 0000000000000000000000000000000000000000..224924147c01470f724eb31aeb39a954f20c22ac Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_29_axolotl, mud puppy, Ambystoma mexicanum.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_2_great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_2_great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f4d5ea0b4d07bd55da467345b40b5056b13e26e7 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_2_great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_30_bullfrog, Rana catesbeiana.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_30_bullfrog, Rana catesbeiana.jpg new file mode 100644 index 0000000000000000000000000000000000000000..2c9d526b6b5aa0c93f42549a6dc5cff855d4ec03 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_30_bullfrog, Rana catesbeiana.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_31_tree frog, tree-frog.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_31_tree frog, tree-frog.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a91ace4de89710162afe52f16549437d7fc45f7a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_31_tree frog, tree-frog.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_32_tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_32_tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui.jpg new file mode 100644 index 0000000000000000000000000000000000000000..8b78226c39d2792cef2069964c5cb7c497b3054d Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_32_tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_33_loggerhead, loggerhead turtle, Caretta caretta.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_33_loggerhead, loggerhead turtle, Caretta caretta.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4f2f5ca0e00d8a2f772c02773c0d62eee0a7788a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_33_loggerhead, loggerhead turtle, Caretta caretta.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_34_leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_34_leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea.jpg new file mode 100644 index 0000000000000000000000000000000000000000..8737f8b716ff76cf0fd153509a6856a657c7e401 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_34_leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_35_mud turtle.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_35_mud turtle.jpg new file mode 100644 index 0000000000000000000000000000000000000000..568405ffbf43fc7baafbcf7febacbd8b6c4b8de8 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_35_mud turtle.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_36_terrapin.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_36_terrapin.jpg new file mode 100644 index 0000000000000000000000000000000000000000..8c51f811be6995d3c29f2a8607c44a2a875e7d7d Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_36_terrapin.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_37_box turtle, box tortoise.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_37_box turtle, box tortoise.jpg new file mode 100644 index 0000000000000000000000000000000000000000..40c2b3190707100fcf27866c9118fd7a980eea29 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_37_box turtle, box tortoise.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_38_banded gecko.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_38_banded gecko.jpg new file mode 100644 index 0000000000000000000000000000000000000000..22e36758e56802b6456de3b3ea5df6bb647b2994 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_38_banded gecko.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_39_common iguana, iguana, Iguana iguana.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_39_common iguana, iguana, Iguana iguana.jpg new file mode 100644 index 0000000000000000000000000000000000000000..6b07f7b734a6e57715eb0d829a95b040fb688322 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_39_common iguana, iguana, Iguana iguana.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_3_tiger shark, Galeocerdo cuvieri.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_3_tiger shark, Galeocerdo cuvieri.jpg new file mode 100644 index 0000000000000000000000000000000000000000..ba7dae64919993dccf3b5fca2488a95fd3be8a6d Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_3_tiger shark, Galeocerdo cuvieri.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_40_American chameleon, anole, Anolis carolinensis.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_40_American chameleon, anole, Anolis carolinensis.jpg new file mode 100644 index 0000000000000000000000000000000000000000..c2817f81cb817a90b0b8ae7eb8a627400d80adb8 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_40_American chameleon, anole, Anolis carolinensis.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_41_whiptail, whiptail lizard.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_41_whiptail, whiptail lizard.jpg new file mode 100644 index 0000000000000000000000000000000000000000..bda7982ceee5f87c89f22151f6e669272cc90872 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_41_whiptail, whiptail lizard.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_42_agama.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_42_agama.jpg new file mode 100644 index 0000000000000000000000000000000000000000..0380352704b58d0f33add50dda8bea37f8639293 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_42_agama.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_43_frilled lizard, Chlamydosaurus kingi.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_43_frilled lizard, Chlamydosaurus kingi.jpg new file mode 100644 index 0000000000000000000000000000000000000000..8773363f2bb36f78dbbf4c205f43637f054373cf Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_43_frilled lizard, Chlamydosaurus kingi.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_44_alligator lizard.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_44_alligator lizard.jpg new file mode 100644 index 0000000000000000000000000000000000000000..ea28b4518525cd939c3a337730803936f157d3a9 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_44_alligator lizard.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_45_Gila monster, Heloderma suspectum.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_45_Gila monster, Heloderma suspectum.jpg new file mode 100644 index 0000000000000000000000000000000000000000..eeaa2942913e45ff1ef81cbe00abdf4081f30448 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_45_Gila monster, Heloderma suspectum.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_46_green lizard, Lacerta viridis.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_46_green lizard, Lacerta viridis.jpg new file mode 100644 index 0000000000000000000000000000000000000000..d08c8d0dfc46fc564906ae5bcb0c3f1b221a543a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_46_green lizard, Lacerta viridis.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_47_African chameleon, Chamaeleo chamaeleon.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_47_African chameleon, Chamaeleo chamaeleon.jpg new file mode 100644 index 0000000000000000000000000000000000000000..2ed8fb41129a56af8d4187cf052201d4106dcd9a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_47_African chameleon, Chamaeleo chamaeleon.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_48_Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_48_Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis.jpg new file mode 100644 index 0000000000000000000000000000000000000000..ec7b00642ab177622b313fdbef8144a9b8757bf3 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_48_Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_49_African crocodile, Nile crocodile, Crocodylus niloticus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_49_African crocodile, Nile crocodile, Crocodylus niloticus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f24f0b8e659e4e5499c3f057a7660746b3e04e31 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_49_African crocodile, Nile crocodile, Crocodylus niloticus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_4_hammerhead, hammerhead shark.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_4_hammerhead, hammerhead shark.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4a1910b9f4e49a03d7f86aacc18324199e0ecef8 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_4_hammerhead, hammerhead shark.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_50_American alligator, Alligator mississipiensis.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_50_American alligator, Alligator mississipiensis.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a5afcdec8fbcfe92db1976bc7dfc1698e3405b12 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_50_American alligator, Alligator mississipiensis.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_51_triceratops.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_51_triceratops.jpg new file mode 100644 index 0000000000000000000000000000000000000000..609d6d1f538c0275ce18da087fdeea7852156ada Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_51_triceratops.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_52_thunder snake, worm snake, Carphophis amoenus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_52_thunder snake, worm snake, Carphophis amoenus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..47f3dc9ff73c888e33db36b8b43e319f14007848 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_52_thunder snake, worm snake, Carphophis amoenus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_53_ringneck snake, ring-necked snake, ring snake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_53_ringneck snake, ring-necked snake, ring snake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..24ef46bbe56a5fd03bb932043a20600c3271c32f Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_53_ringneck snake, ring-necked snake, ring snake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_54_hognose snake, puff adder, sand viper.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_54_hognose snake, puff adder, sand viper.jpg new file mode 100644 index 0000000000000000000000000000000000000000..c168a5a8fa2f6a9a2705358bb770887d30b13589 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_54_hognose snake, puff adder, sand viper.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_55_green snake, grass snake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_55_green snake, grass snake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..73e39d291d3fc1e47afad202fb11c3b31dc306e5 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_55_green snake, grass snake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_56_king snake, kingsnake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_56_king snake, kingsnake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..83077df49a13f5f49251b4ac82a316b5e65e43a2 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_56_king snake, kingsnake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_57_garter snake, grass snake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_57_garter snake, grass snake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..24d5ff97711aac5fbe61e7e616b2e44995f2174f Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_57_garter snake, grass snake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_58_water snake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_58_water snake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..330d58a206e041738a5bed913ae7144ac58e5365 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_58_water snake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_59_vine snake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_59_vine snake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..73a1c346f3c5b70a98eb4f5184d8cfe9bea8b492 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_59_vine snake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_5_electric ray, crampfish, numbfish, torpedo.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_5_electric ray, crampfish, numbfish, torpedo.jpg new file mode 100644 index 0000000000000000000000000000000000000000..8d7a63723ac46e3f68bc205649c16b609fb4768a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_5_electric ray, crampfish, numbfish, torpedo.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_60_night snake, Hypsiglena torquata.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_60_night snake, Hypsiglena torquata.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f3b1c2277ea70a79dc55f2c4291f995bbbf4a022 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_60_night snake, Hypsiglena torquata.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_61_boa constrictor, Constrictor constrictor.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_61_boa constrictor, Constrictor constrictor.jpg new file mode 100644 index 0000000000000000000000000000000000000000..6c55827206bcf5fde0dec291429e2da09130afdf Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_61_boa constrictor, Constrictor constrictor.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_62_rock python, rock snake, Python sebae.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_62_rock python, rock snake, Python sebae.jpg new file mode 100644 index 0000000000000000000000000000000000000000..8fec69b0cac6c7f582331b60d622e9f68f54abe7 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_62_rock python, rock snake, Python sebae.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_63_Indian cobra, Naja naja.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_63_Indian cobra, Naja naja.jpg new file mode 100644 index 0000000000000000000000000000000000000000..8b5d96163742875ee8b4a2772c3b2ba51ace32b5 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_63_Indian cobra, Naja naja.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_64_green mamba.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_64_green mamba.jpg new file mode 100644 index 0000000000000000000000000000000000000000..3b855a6aa7c26fed0d0c52d35def94234655b152 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_64_green mamba.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_65_sea snake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_65_sea snake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..e1feab84a8fabfad8dedec883ee0e93657f9fda6 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_65_sea snake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_66_horned viper, cerastes, sand viper, horned asp, Cerastes cornutus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_66_horned viper, cerastes, sand viper, horned asp, Cerastes cornutus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..5923522ddce83ad1c45e43315ae00a06fc8e536f Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_66_horned viper, cerastes, sand viper, horned asp, Cerastes cornutus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_67_diamondback, diamondback rattlesnake, Crotalus adamanteus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_67_diamondback, diamondback rattlesnake, Crotalus adamanteus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..83913dae8b9eebf9fb6047d1bc4b3eda102c0d49 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_67_diamondback, diamondback rattlesnake, Crotalus adamanteus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_68_sidewinder, horned rattlesnake, Crotalus cerastes.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_68_sidewinder, horned rattlesnake, Crotalus cerastes.jpg new file mode 100644 index 0000000000000000000000000000000000000000..1188a6293e76d4c49f558c37b899dc3b46737ef7 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_68_sidewinder, horned rattlesnake, Crotalus cerastes.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_69_trilobite.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_69_trilobite.jpg new file mode 100644 index 0000000000000000000000000000000000000000..34751ec3f93fb864e17947e3da6bdfdc11d9c44a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_69_trilobite.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_6_stingray.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_6_stingray.jpg new file mode 100644 index 0000000000000000000000000000000000000000..fe110b9cc0f25eb3c029204b29a26ef6163e0e90 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_6_stingray.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_70_harvestman, daddy longlegs, Phalangium opilio.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_70_harvestman, daddy longlegs, Phalangium opilio.jpg new file mode 100644 index 0000000000000000000000000000000000000000..b03c2eeee5633b407f49f16ed9a2b191d01b0e48 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_70_harvestman, daddy longlegs, Phalangium opilio.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_71_scorpion.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_71_scorpion.jpg new file mode 100644 index 0000000000000000000000000000000000000000..eb287bec087291785e739639c0ff0fe7814dfa4e Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_71_scorpion.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_72_black and gold garden spider, Argiope aurantia.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_72_black and gold garden spider, Argiope aurantia.jpg new file mode 100644 index 0000000000000000000000000000000000000000..5a9a81bc56950928fc7e5075b4e88fc0e2b403ae Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_72_black and gold garden spider, Argiope aurantia.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_73_barn spider, Araneus cavaticus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_73_barn spider, Araneus cavaticus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..96e7bc977c055dbbc47d6b7a39dc326c933e68a1 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_73_barn spider, Araneus cavaticus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_74_garden spider, Aranea diademata.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_74_garden spider, Aranea diademata.jpg new file mode 100644 index 0000000000000000000000000000000000000000..566c76dd69e4ef1f942fb7f116313a02d237dfa4 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_74_garden spider, Aranea diademata.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_75_black widow, Latrodectus mactans.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_75_black widow, Latrodectus mactans.jpg new file mode 100644 index 0000000000000000000000000000000000000000..7566b7e300b4bb80fff970dc583613c0463b2200 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_75_black widow, Latrodectus mactans.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_76_tarantula.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_76_tarantula.jpg new file mode 100644 index 0000000000000000000000000000000000000000..b619e7990b7d3938db012d76b3d5edd825755c65 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_76_tarantula.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_77_wolf spider, hunting spider.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_77_wolf spider, hunting spider.jpg new file mode 100644 index 0000000000000000000000000000000000000000..df87001bc39f71f96372ea497933365338407cc3 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_77_wolf spider, hunting spider.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_78_tick.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_78_tick.jpg new file mode 100644 index 0000000000000000000000000000000000000000..533d2c1778afee9786bebcd348b9a85f95daa2cb Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_78_tick.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_79_centipede.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_79_centipede.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a0fdfad71b3a9accbd690238e7e7f1800bfe175f Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_79_centipede.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_7_cock.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_7_cock.jpg new file mode 100644 index 0000000000000000000000000000000000000000..4899a6c06ca12ce06a26bd138f50849d1788aebe Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_7_cock.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_80_black grouse.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_80_black grouse.jpg new file mode 100644 index 0000000000000000000000000000000000000000..ecb930362d74e8e4f037c2d07fdc0dff819b6742 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_80_black grouse.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_81_ptarmigan.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_81_ptarmigan.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a55736446c3763e6c0e85de6bc7e28a93990cafe Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_81_ptarmigan.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_82_ruffed grouse, partridge, Bonasa umbellus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_82_ruffed grouse, partridge, Bonasa umbellus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..27bfd4d318372bc5fc5a93f9e89f6e60994fb9d4 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_82_ruffed grouse, partridge, Bonasa umbellus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_83_prairie chicken, prairie grouse, prairie fowl.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_83_prairie chicken, prairie grouse, prairie fowl.jpg new file mode 100644 index 0000000000000000000000000000000000000000..09db04e0fb488400098fd7556cc00d76ce3b5f15 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_83_prairie chicken, prairie grouse, prairie fowl.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_84_peacock.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_84_peacock.jpg new file mode 100644 index 0000000000000000000000000000000000000000..89ede6800d3e009651d4cf3aebffc7c64c1d0bdb Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_84_peacock.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_85_quail.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_85_quail.jpg new file mode 100644 index 0000000000000000000000000000000000000000..3fd41b703334bdaa0cc63c5bcf76713e9b96d633 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_85_quail.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_86_partridge.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_86_partridge.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f9416c9dfd3700e89c12fcaabd08e468cb1dc740 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_86_partridge.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_87_African grey, African gray, Psittacus erithacus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_87_African grey, African gray, Psittacus erithacus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f99bcbc78b26ba7d02ebd8ba141cf90d7d7c6958 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_87_African grey, African gray, Psittacus erithacus.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_88_macaw.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_88_macaw.jpg new file mode 100644 index 0000000000000000000000000000000000000000..a6c30087d872601f1a9985832e432517c5a9f5c2 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_88_macaw.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_89_sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_89_sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita.jpg new file mode 100644 index 0000000000000000000000000000000000000000..b60489591affc41046e16c1054aa0e1368c0fedc Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_89_sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_8_hen.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_8_hen.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f707b8ffa98c72e1483633f6c0608aeb74ecb976 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_8_hen.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_90_lorikeet.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_90_lorikeet.jpg new file mode 100644 index 0000000000000000000000000000000000000000..7dacef82137e35383c5d09ebaafa91a3043af786 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_90_lorikeet.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_91_coucal.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_91_coucal.jpg new file mode 100644 index 0000000000000000000000000000000000000000..9583e05faebf1a7cce1adf8ba05adaa5ca157f7a Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_91_coucal.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_92_bee eater.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_92_bee eater.jpg new file mode 100644 index 0000000000000000000000000000000000000000..fee2c6f2011d75bda94ae6779e4ff86f8e120563 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_92_bee eater.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_93_hornbill.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_93_hornbill.jpg new file mode 100644 index 0000000000000000000000000000000000000000..d1664a6a4bd21f1ed6e30ca00663a2623cda8034 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_93_hornbill.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_94_hummingbird.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_94_hummingbird.jpg new file mode 100644 index 0000000000000000000000000000000000000000..de4eb84a5743891d4c04971a5b94b41897df4829 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_94_hummingbird.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_95_jacamar.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_95_jacamar.jpg new file mode 100644 index 0000000000000000000000000000000000000000..054b8d7f644618504adcd337db7b36fa0997c008 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_95_jacamar.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_96_toucan.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_96_toucan.jpg new file mode 100644 index 0000000000000000000000000000000000000000..f3cfa8cae2d50f1a3d1253f7f65aebe003d6483c Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_96_toucan.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_97_drake.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_97_drake.jpg new file mode 100644 index 0000000000000000000000000000000000000000..1ec8e27e9cc745b93bdf67ba60211c32c5883a99 Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_97_drake.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_98_red-breasted merganser, Mergus serrator.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_98_red-breasted merganser, Mergus serrator.jpg new file mode 100644 index 0000000000000000000000000000000000000000..99e197251e9991e076579cca477a43f52c615ead Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_98_red-breasted merganser, Mergus serrator.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_99_goose.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_99_goose.jpg new file mode 100644 index 0000000000000000000000000000000000000000..598f0d5ea4ac3e12f7a935d7ec37de918a854e1b Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_99_goose.jpg differ diff --git a/checkpoint-13000/images_steps-28_cfg-4/imagenet_9_ostrich, Struthio camelus.jpg b/checkpoint-13000/images_steps-28_cfg-4/imagenet_9_ostrich, Struthio camelus.jpg new file mode 100644 index 0000000000000000000000000000000000000000..aa8f81e43c986448544d1c0de9906caa93da97ca Binary files /dev/null and b/checkpoint-13000/images_steps-28_cfg-4/imagenet_9_ostrich, Struthio camelus.jpg differ diff --git a/checkpoint-13000/model.safetensors b/checkpoint-13000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2feae196d0b254b4b6706b1e3599ca856f039bf8 --- /dev/null +++ b/checkpoint-13000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:782e23ef3d535365d8999ef82ac3ac3e2746b170b2140b454166ca35dc7d696f +size 4539487822 diff --git a/checkpoint-13000/optimizer.pt b/checkpoint-13000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..23f87e7f9293ce70dfc7aa5e00f2a04ab87d8ca0 --- /dev/null +++ b/checkpoint-13000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7f18f2fcf8da6b318c38298b8054fb55e1173fa215db22e5aaf52ea7880da314 +size 6632345482 diff --git a/checkpoint-13000/rng_state_0.pth b/checkpoint-13000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..6fdcc94193c59f1984cf63fcb9df7bb16469d66c --- /dev/null +++ b/checkpoint-13000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e0570cb5955463ec9040fdb0bf0d8878ff413fe82286e88f19d08f550aab714a +size 15984 diff --git a/checkpoint-13000/rng_state_1.pth b/checkpoint-13000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..e7640a6f8794f4952ec0115389e21db432198854 --- /dev/null +++ b/checkpoint-13000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dfb6e27c9bc148d7ddbbfdfed67fedce53ff5fcfb12715e5906c768cd8564322 +size 15984 diff --git a/checkpoint-13000/rng_state_2.pth b/checkpoint-13000/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..8e5ada59a65efd2a389c3fc49a4b99ee485531b3 --- /dev/null +++ b/checkpoint-13000/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d555b49b6e198fe390f5e0c87aec35789a8efcfff65bb19e6a4f4a62353922c3 +size 15984 diff --git a/checkpoint-13000/rng_state_3.pth b/checkpoint-13000/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..d1f5325308552e51504ceff9c52603b0e62bdb01 --- /dev/null +++ b/checkpoint-13000/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d7ebbaddae17202df878e77b9be2422473d0e9a364aebf1b023e7e9c89f4519 +size 15984 diff --git a/checkpoint-13000/rng_state_4.pth b/checkpoint-13000/rng_state_4.pth new file mode 100644 index 0000000000000000000000000000000000000000..997fe43181197747db3d0835eb08f23247e9ee17 --- /dev/null +++ b/checkpoint-13000/rng_state_4.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8bd13e67eff2d2e666d95c1d373f402c6bdab590c60e4f50421d33b1cd3a7118 +size 15984 diff --git a/checkpoint-13000/rng_state_5.pth b/checkpoint-13000/rng_state_5.pth new file mode 100644 index 0000000000000000000000000000000000000000..c1965bfcb3c7f967f352252a4329d422ec5f7fee --- /dev/null +++ b/checkpoint-13000/rng_state_5.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c689b116a48533f7824fffe1ca2a6908600b5d69c1d456e6721a703fa9c5d36 +size 15984 diff --git a/checkpoint-13000/rng_state_6.pth b/checkpoint-13000/rng_state_6.pth new file mode 100644 index 0000000000000000000000000000000000000000..313ca8f4aaee764f4910c4a66996e236163418aa --- /dev/null +++ b/checkpoint-13000/rng_state_6.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f395e4edad0fe6fe15c1e2fd5016d6a25a71b3720e1b51490d126c105b38834d +size 15984 diff --git a/checkpoint-13000/rng_state_7.pth b/checkpoint-13000/rng_state_7.pth new file mode 100644 index 0000000000000000000000000000000000000000..1dc181867b353077bef0ff9da8b0de08b01fa895 --- /dev/null +++ b/checkpoint-13000/rng_state_7.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a0b13904328379742f01fc308fb5d335dbc1d1a4b76c7fca493388b0ed00836 +size 15984 diff --git a/checkpoint-13000/scheduler.pt b/checkpoint-13000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..253edfb5a75b6cd349e9b478283ee1148e9eeaa0 --- /dev/null +++ b/checkpoint-13000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91edfb41dddbfeb14ac5bde19835d54b92dd2350ba592cab3e1c4993b44b7dcd +size 1064 diff --git a/checkpoint-13000/trainer_state.json b/checkpoint-13000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..b2c80e161fd59d6dc4760b127029e2b112082870 --- /dev/null +++ b/checkpoint-13000/trainer_state.json @@ -0,0 +1,91033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 20.766773162939298, + "eval_steps": 500, + "global_step": 13000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001597444089456869, + "grad_norm": 3.390625, + "learning_rate": 0.0005, + "loss": 3.6303, + "step": 1 + }, + { + "epoch": 0.003194888178913738, + "grad_norm": 17.5, + "learning_rate": 0.0005, + "loss": 5.5153, + "step": 2 + }, + { + "epoch": 0.004792332268370607, + "grad_norm": 7.84375, + "learning_rate": 0.0005, + "loss": 2.7811, + "step": 3 + }, + { + "epoch": 0.006389776357827476, + "grad_norm": 7.65625, + "learning_rate": 0.0005, + "loss": 2.8695, + "step": 4 + }, + { + "epoch": 0.007987220447284345, + "grad_norm": 3.734375, + "learning_rate": 0.0005, + "loss": 2.3246, + "step": 5 + }, + { + "epoch": 0.009584664536741214, + "grad_norm": 3.640625, + "learning_rate": 0.0005, + "loss": 2.0843, + "step": 6 + }, + { + "epoch": 0.011182108626198083, + "grad_norm": 1.015625, + "learning_rate": 0.0005, + "loss": 1.7376, + "step": 7 + }, + { + "epoch": 0.012779552715654952, + "grad_norm": 4.65625, + "learning_rate": 0.0005, + "loss": 1.863, + "step": 8 + }, + { + "epoch": 0.01437699680511182, + "grad_norm": 1.3828125, + "learning_rate": 0.0005, + "loss": 1.7027, + "step": 9 + }, + { + "epoch": 0.01597444089456869, + "grad_norm": 4.90625, + "learning_rate": 0.0005, + "loss": 1.8852, + "step": 10 + }, + { + "epoch": 0.01757188498402556, + "grad_norm": 4.46875, + "learning_rate": 0.0005, + "loss": 1.8597, + "step": 11 + }, + { + "epoch": 0.019169329073482427, + "grad_norm": 1.015625, + "learning_rate": 0.0005, + "loss": 1.6276, + "step": 12 + }, + { + "epoch": 0.020766773162939296, + "grad_norm": 2.640625, + "learning_rate": 0.0005, + "loss": 1.6508, + "step": 13 + }, + { + "epoch": 0.022364217252396165, + "grad_norm": 2.578125, + "learning_rate": 0.0005, + "loss": 1.6636, + "step": 14 + }, + { + "epoch": 0.023961661341853034, + "grad_norm": 0.703125, + "learning_rate": 0.0005, + "loss": 1.5623, + "step": 15 + }, + { + "epoch": 0.025559105431309903, + "grad_norm": 1.34375, + "learning_rate": 0.0005, + "loss": 1.5314, + "step": 16 + }, + { + "epoch": 0.027156549520766772, + "grad_norm": 1.0546875, + "learning_rate": 0.0005, + "loss": 1.4973, + "step": 17 + }, + { + "epoch": 0.02875399361022364, + "grad_norm": 0.412109375, + "learning_rate": 0.0005, + "loss": 1.4629, + "step": 18 + }, + { + "epoch": 0.03035143769968051, + "grad_norm": 0.640625, + "learning_rate": 0.0005, + "loss": 1.445, + "step": 19 + }, + { + "epoch": 0.03194888178913738, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.4098, + "step": 20 + }, + { + "epoch": 0.03354632587859425, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.3997, + "step": 21 + }, + { + "epoch": 0.03514376996805112, + "grad_norm": 0.376953125, + "learning_rate": 0.0005, + "loss": 1.401, + "step": 22 + }, + { + "epoch": 0.036741214057507986, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.378, + "step": 23 + }, + { + "epoch": 0.038338658146964855, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.3717, + "step": 24 + }, + { + "epoch": 0.039936102236421724, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.3598, + "step": 25 + }, + { + "epoch": 0.04153354632587859, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.3511, + "step": 26 + }, + { + "epoch": 0.04313099041533546, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.3562, + "step": 27 + }, + { + "epoch": 0.04472843450479233, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.3464, + "step": 28 + }, + { + "epoch": 0.0463258785942492, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.3385, + "step": 29 + }, + { + "epoch": 0.04792332268370607, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.3363, + "step": 30 + }, + { + "epoch": 0.04952076677316294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.3294, + "step": 31 + }, + { + "epoch": 0.051118210862619806, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.3254, + "step": 32 + }, + { + "epoch": 0.052715654952076675, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.315, + "step": 33 + }, + { + "epoch": 0.054313099041533544, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.3153, + "step": 34 + }, + { + "epoch": 0.05591054313099041, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.3213, + "step": 35 + }, + { + "epoch": 0.05750798722044728, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.309, + "step": 36 + }, + { + "epoch": 0.05910543130990415, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.3139, + "step": 37 + }, + { + "epoch": 0.06070287539936102, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.3085, + "step": 38 + }, + { + "epoch": 0.06230031948881789, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.2991, + "step": 39 + }, + { + "epoch": 0.06389776357827476, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.3113, + "step": 40 + }, + { + "epoch": 0.06549520766773163, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.3159, + "step": 41 + }, + { + "epoch": 0.0670926517571885, + "grad_norm": 0.2373046875, + "learning_rate": 0.0005, + "loss": 1.295, + "step": 42 + }, + { + "epoch": 0.06869009584664537, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.2948, + "step": 43 + }, + { + "epoch": 0.07028753993610223, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.303, + "step": 44 + }, + { + "epoch": 0.07188498402555911, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.3037, + "step": 45 + }, + { + "epoch": 0.07348242811501597, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.29, + "step": 46 + }, + { + "epoch": 0.07507987220447285, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.2882, + "step": 47 + }, + { + "epoch": 0.07667731629392971, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.2929, + "step": 48 + }, + { + "epoch": 0.07827476038338659, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.2836, + "step": 49 + }, + { + "epoch": 0.07987220447284345, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.2844, + "step": 50 + }, + { + "epoch": 0.08146964856230032, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.2863, + "step": 51 + }, + { + "epoch": 0.08306709265175719, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.2772, + "step": 52 + }, + { + "epoch": 0.08466453674121406, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.2777, + "step": 53 + }, + { + "epoch": 0.08626198083067092, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.2746, + "step": 54 + }, + { + "epoch": 0.0878594249201278, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.2789, + "step": 55 + }, + { + "epoch": 0.08945686900958466, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.2659, + "step": 56 + }, + { + "epoch": 0.09105431309904154, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.2669, + "step": 57 + }, + { + "epoch": 0.0926517571884984, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.2729, + "step": 58 + }, + { + "epoch": 0.09424920127795527, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.269, + "step": 59 + }, + { + "epoch": 0.09584664536741214, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.2707, + "step": 60 + }, + { + "epoch": 0.09744408945686901, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.2673, + "step": 61 + }, + { + "epoch": 0.09904153354632587, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.2676, + "step": 62 + }, + { + "epoch": 0.10063897763578275, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.2581, + "step": 63 + }, + { + "epoch": 0.10223642172523961, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.2608, + "step": 64 + }, + { + "epoch": 0.10383386581469649, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.2584, + "step": 65 + }, + { + "epoch": 0.10543130990415335, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.2597, + "step": 66 + }, + { + "epoch": 0.10702875399361023, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.2582, + "step": 67 + }, + { + "epoch": 0.10862619808306709, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.2544, + "step": 68 + }, + { + "epoch": 0.11022364217252396, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.2545, + "step": 69 + }, + { + "epoch": 0.11182108626198083, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.2556, + "step": 70 + }, + { + "epoch": 0.1134185303514377, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.2508, + "step": 71 + }, + { + "epoch": 0.11501597444089456, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.2514, + "step": 72 + }, + { + "epoch": 0.11661341853035144, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.2509, + "step": 73 + }, + { + "epoch": 0.1182108626198083, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.2508, + "step": 74 + }, + { + "epoch": 0.11980830670926518, + "grad_norm": 0.3125, + "learning_rate": 0.0005, + "loss": 1.2531, + "step": 75 + }, + { + "epoch": 0.12140575079872204, + "grad_norm": 0.5390625, + "learning_rate": 0.0005, + "loss": 1.2642, + "step": 76 + }, + { + "epoch": 0.12300319488817892, + "grad_norm": 0.9453125, + "learning_rate": 0.0005, + "loss": 1.2656, + "step": 77 + }, + { + "epoch": 0.12460063897763578, + "grad_norm": 0.9609375, + "learning_rate": 0.0005, + "loss": 1.3036, + "step": 78 + }, + { + "epoch": 0.12619808306709265, + "grad_norm": 0.498046875, + "learning_rate": 0.0005, + "loss": 1.2778, + "step": 79 + }, + { + "epoch": 0.12779552715654952, + "grad_norm": 1.03125, + "learning_rate": 0.0005, + "loss": 1.2695, + "step": 80 + }, + { + "epoch": 0.12939297124600638, + "grad_norm": 0.578125, + "learning_rate": 0.0005, + "loss": 1.2628, + "step": 81 + }, + { + "epoch": 0.13099041533546327, + "grad_norm": 0.39453125, + "learning_rate": 0.0005, + "loss": 1.2666, + "step": 82 + }, + { + "epoch": 0.13258785942492013, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.2624, + "step": 83 + }, + { + "epoch": 0.134185303514377, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.2575, + "step": 84 + }, + { + "epoch": 0.13578274760383385, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.253, + "step": 85 + }, + { + "epoch": 0.13738019169329074, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.2532, + "step": 86 + }, + { + "epoch": 0.1389776357827476, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.2481, + "step": 87 + }, + { + "epoch": 0.14057507987220447, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.2461, + "step": 88 + }, + { + "epoch": 0.14217252396166133, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.2395, + "step": 89 + }, + { + "epoch": 0.14376996805111822, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.2436, + "step": 90 + }, + { + "epoch": 0.14536741214057508, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.2458, + "step": 91 + }, + { + "epoch": 0.14696485623003194, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.2398, + "step": 92 + }, + { + "epoch": 0.1485623003194888, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.2416, + "step": 93 + }, + { + "epoch": 0.1501597444089457, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.2359, + "step": 94 + }, + { + "epoch": 0.15175718849840256, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.2355, + "step": 95 + }, + { + "epoch": 0.15335463258785942, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.2457, + "step": 96 + }, + { + "epoch": 0.15495207667731628, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.2436, + "step": 97 + }, + { + "epoch": 0.15654952076677317, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.2392, + "step": 98 + }, + { + "epoch": 0.15814696485623003, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.2442, + "step": 99 + }, + { + "epoch": 0.1597444089456869, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.2386, + "step": 100 + }, + { + "epoch": 0.16134185303514376, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.2429, + "step": 101 + }, + { + "epoch": 0.16293929712460065, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.2238, + "step": 102 + }, + { + "epoch": 0.1645367412140575, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.238, + "step": 103 + }, + { + "epoch": 0.16613418530351437, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.2289, + "step": 104 + }, + { + "epoch": 0.16773162939297126, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.2384, + "step": 105 + }, + { + "epoch": 0.16932907348242812, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.2244, + "step": 106 + }, + { + "epoch": 0.17092651757188498, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.2248, + "step": 107 + }, + { + "epoch": 0.17252396166134185, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.2196, + "step": 108 + }, + { + "epoch": 0.17412140575079874, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.2263, + "step": 109 + }, + { + "epoch": 0.1757188498402556, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.2254, + "step": 110 + }, + { + "epoch": 0.17731629392971246, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.2265, + "step": 111 + }, + { + "epoch": 0.17891373801916932, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.2184, + "step": 112 + }, + { + "epoch": 0.1805111821086262, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.224, + "step": 113 + }, + { + "epoch": 0.18210862619808307, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.2227, + "step": 114 + }, + { + "epoch": 0.18370607028753994, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.2152, + "step": 115 + }, + { + "epoch": 0.1853035143769968, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.2217, + "step": 116 + }, + { + "epoch": 0.1869009584664537, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.2197, + "step": 117 + }, + { + "epoch": 0.18849840255591055, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.215, + "step": 118 + }, + { + "epoch": 0.1900958466453674, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.2158, + "step": 119 + }, + { + "epoch": 0.19169329073482427, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.219, + "step": 120 + }, + { + "epoch": 0.19329073482428116, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.2184, + "step": 121 + }, + { + "epoch": 0.19488817891373802, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.2155, + "step": 122 + }, + { + "epoch": 0.1964856230031949, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.218, + "step": 123 + }, + { + "epoch": 0.19808306709265175, + "grad_norm": 0.5703125, + "learning_rate": 0.0005, + "loss": 1.2236, + "step": 124 + }, + { + "epoch": 0.19968051118210864, + "grad_norm": 1.2265625, + "learning_rate": 0.0005, + "loss": 1.2427, + "step": 125 + }, + { + "epoch": 0.2012779552715655, + "grad_norm": 0.83203125, + "learning_rate": 0.0005, + "loss": 1.2307, + "step": 126 + }, + { + "epoch": 0.20287539936102236, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.2214, + "step": 127 + }, + { + "epoch": 0.20447284345047922, + "grad_norm": 0.416015625, + "learning_rate": 0.0005, + "loss": 1.2235, + "step": 128 + }, + { + "epoch": 0.20607028753993611, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.2175, + "step": 129 + }, + { + "epoch": 0.20766773162939298, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.2159, + "step": 130 + }, + { + "epoch": 0.20926517571884984, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.2158, + "step": 131 + }, + { + "epoch": 0.2108626198083067, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.2192, + "step": 132 + }, + { + "epoch": 0.2124600638977636, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.2172, + "step": 133 + }, + { + "epoch": 0.21405750798722045, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.2174, + "step": 134 + }, + { + "epoch": 0.21565495207667731, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.214, + "step": 135 + }, + { + "epoch": 0.21725239616613418, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.2048, + "step": 136 + }, + { + "epoch": 0.21884984025559107, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.215, + "step": 137 + }, + { + "epoch": 0.22044728434504793, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.2119, + "step": 138 + }, + { + "epoch": 0.2220447284345048, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.2138, + "step": 139 + }, + { + "epoch": 0.22364217252396165, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.208, + "step": 140 + }, + { + "epoch": 0.22523961661341854, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.2094, + "step": 141 + }, + { + "epoch": 0.2268370607028754, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.1994, + "step": 142 + }, + { + "epoch": 0.22843450479233227, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.2043, + "step": 143 + }, + { + "epoch": 0.23003194888178913, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.2063, + "step": 144 + }, + { + "epoch": 0.23162939297124602, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.2027, + "step": 145 + }, + { + "epoch": 0.23322683706070288, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.2019, + "step": 146 + }, + { + "epoch": 0.23482428115015974, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.2065, + "step": 147 + }, + { + "epoch": 0.2364217252396166, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.2055, + "step": 148 + }, + { + "epoch": 0.2380191693290735, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.2036, + "step": 149 + }, + { + "epoch": 0.23961661341853036, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.2094, + "step": 150 + }, + { + "epoch": 0.24121405750798722, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.207, + "step": 151 + }, + { + "epoch": 0.24281150159744408, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.2027, + "step": 152 + }, + { + "epoch": 0.24440894568690097, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1961, + "step": 153 + }, + { + "epoch": 0.24600638977635783, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1957, + "step": 154 + }, + { + "epoch": 0.2476038338658147, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.2024, + "step": 155 + }, + { + "epoch": 0.24920127795527156, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.1964, + "step": 156 + }, + { + "epoch": 0.2507987220447284, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.2021, + "step": 157 + }, + { + "epoch": 0.2523961661341853, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1907, + "step": 158 + }, + { + "epoch": 0.2539936102236422, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.1935, + "step": 159 + }, + { + "epoch": 0.25559105431309903, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1975, + "step": 160 + }, + { + "epoch": 0.2571884984025559, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1983, + "step": 161 + }, + { + "epoch": 0.25878594249201275, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1941, + "step": 162 + }, + { + "epoch": 0.26038338658146964, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1872, + "step": 163 + }, + { + "epoch": 0.26198083067092653, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.2013, + "step": 164 + }, + { + "epoch": 0.26357827476038337, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1971, + "step": 165 + }, + { + "epoch": 0.26517571884984026, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.2061, + "step": 166 + }, + { + "epoch": 0.26677316293929715, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1886, + "step": 167 + }, + { + "epoch": 0.268370607028754, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1969, + "step": 168 + }, + { + "epoch": 0.26996805111821087, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.194, + "step": 169 + }, + { + "epoch": 0.2715654952076677, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1917, + "step": 170 + }, + { + "epoch": 0.2731629392971246, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1935, + "step": 171 + }, + { + "epoch": 0.2747603833865815, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1936, + "step": 172 + }, + { + "epoch": 0.2763578274760383, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1906, + "step": 173 + }, + { + "epoch": 0.2779552715654952, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1991, + "step": 174 + }, + { + "epoch": 0.2795527156549521, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.1884, + "step": 175 + }, + { + "epoch": 0.28115015974440893, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.1936, + "step": 176 + }, + { + "epoch": 0.2827476038338658, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1833, + "step": 177 + }, + { + "epoch": 0.28434504792332266, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.187, + "step": 178 + }, + { + "epoch": 0.28594249201277955, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.1916, + "step": 179 + }, + { + "epoch": 0.28753993610223644, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.1884, + "step": 180 + }, + { + "epoch": 0.28913738019169327, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1856, + "step": 181 + }, + { + "epoch": 0.29073482428115016, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1918, + "step": 182 + }, + { + "epoch": 0.29233226837060705, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1908, + "step": 183 + }, + { + "epoch": 0.2939297124600639, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1946, + "step": 184 + }, + { + "epoch": 0.2955271565495208, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1867, + "step": 185 + }, + { + "epoch": 0.2971246006389776, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1944, + "step": 186 + }, + { + "epoch": 0.2987220447284345, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1837, + "step": 187 + }, + { + "epoch": 0.3003194888178914, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1921, + "step": 188 + }, + { + "epoch": 0.3019169329073482, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1931, + "step": 189 + }, + { + "epoch": 0.3035143769968051, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1835, + "step": 190 + }, + { + "epoch": 0.305111821086262, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.1885, + "step": 191 + }, + { + "epoch": 0.30670926517571884, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.1809, + "step": 192 + }, + { + "epoch": 0.3083067092651757, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.1812, + "step": 193 + }, + { + "epoch": 0.30990415335463256, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.1909, + "step": 194 + }, + { + "epoch": 0.31150159744408945, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1774, + "step": 195 + }, + { + "epoch": 0.31309904153354634, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.1777, + "step": 196 + }, + { + "epoch": 0.3146964856230032, + "grad_norm": 0.578125, + "learning_rate": 0.0005, + "loss": 1.1873, + "step": 197 + }, + { + "epoch": 0.31629392971246006, + "grad_norm": 2.203125, + "learning_rate": 0.0005, + "loss": 1.2371, + "step": 198 + }, + { + "epoch": 0.31789137380191695, + "grad_norm": 0.828125, + "learning_rate": 0.0005, + "loss": 1.2053, + "step": 199 + }, + { + "epoch": 0.3194888178913738, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.1954, + "step": 200 + }, + { + "epoch": 0.3210862619808307, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.1942, + "step": 201 + }, + { + "epoch": 0.3226837060702875, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.192, + "step": 202 + }, + { + "epoch": 0.3242811501597444, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1883, + "step": 203 + }, + { + "epoch": 0.3258785942492013, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1916, + "step": 204 + }, + { + "epoch": 0.3274760383386581, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1866, + "step": 205 + }, + { + "epoch": 0.329073482428115, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1961, + "step": 206 + }, + { + "epoch": 0.3306709265175719, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1856, + "step": 207 + }, + { + "epoch": 0.33226837060702874, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.1813, + "step": 208 + }, + { + "epoch": 0.33386581469648563, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1809, + "step": 209 + }, + { + "epoch": 0.3354632587859425, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.18, + "step": 210 + }, + { + "epoch": 0.33706070287539935, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1808, + "step": 211 + }, + { + "epoch": 0.33865814696485624, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.1839, + "step": 212 + }, + { + "epoch": 0.3402555910543131, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.1841, + "step": 213 + }, + { + "epoch": 0.34185303514376997, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1847, + "step": 214 + }, + { + "epoch": 0.34345047923322686, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.1796, + "step": 215 + }, + { + "epoch": 0.3450479233226837, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1798, + "step": 216 + }, + { + "epoch": 0.3466453674121406, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1732, + "step": 217 + }, + { + "epoch": 0.34824281150159747, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1827, + "step": 218 + }, + { + "epoch": 0.3498402555910543, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1789, + "step": 219 + }, + { + "epoch": 0.3514376996805112, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1765, + "step": 220 + }, + { + "epoch": 0.35303514376996803, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1797, + "step": 221 + }, + { + "epoch": 0.3546325878594249, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1816, + "step": 222 + }, + { + "epoch": 0.3562300319488818, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1773, + "step": 223 + }, + { + "epoch": 0.35782747603833864, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.173, + "step": 224 + }, + { + "epoch": 0.35942492012779553, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1787, + "step": 225 + }, + { + "epoch": 0.3610223642172524, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.187, + "step": 226 + }, + { + "epoch": 0.36261980830670926, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.178, + "step": 227 + }, + { + "epoch": 0.36421725239616615, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1788, + "step": 228 + }, + { + "epoch": 0.365814696485623, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1764, + "step": 229 + }, + { + "epoch": 0.36741214057507987, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1771, + "step": 230 + }, + { + "epoch": 0.36900958466453676, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1738, + "step": 231 + }, + { + "epoch": 0.3706070287539936, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1827, + "step": 232 + }, + { + "epoch": 0.3722044728434505, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.1733, + "step": 233 + }, + { + "epoch": 0.3738019169329074, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1692, + "step": 234 + }, + { + "epoch": 0.3753993610223642, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.1743, + "step": 235 + }, + { + "epoch": 0.3769968051118211, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1739, + "step": 236 + }, + { + "epoch": 0.37859424920127793, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1723, + "step": 237 + }, + { + "epoch": 0.3801916932907348, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1715, + "step": 238 + }, + { + "epoch": 0.3817891373801917, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1725, + "step": 239 + }, + { + "epoch": 0.38338658146964855, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1695, + "step": 240 + }, + { + "epoch": 0.38498402555910544, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1688, + "step": 241 + }, + { + "epoch": 0.3865814696485623, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1654, + "step": 242 + }, + { + "epoch": 0.38817891373801916, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1729, + "step": 243 + }, + { + "epoch": 0.38977635782747605, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1694, + "step": 244 + }, + { + "epoch": 0.3913738019169329, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1674, + "step": 245 + }, + { + "epoch": 0.3929712460063898, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.169, + "step": 246 + }, + { + "epoch": 0.39456869009584666, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.177, + "step": 247 + }, + { + "epoch": 0.3961661341853035, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1646, + "step": 248 + }, + { + "epoch": 0.3977635782747604, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1635, + "step": 249 + }, + { + "epoch": 0.3993610223642173, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1784, + "step": 250 + }, + { + "epoch": 0.4009584664536741, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1592, + "step": 251 + }, + { + "epoch": 0.402555910543131, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1734, + "step": 252 + }, + { + "epoch": 0.40415335463258784, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1726, + "step": 253 + }, + { + "epoch": 0.4057507987220447, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1565, + "step": 254 + }, + { + "epoch": 0.4073482428115016, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.167, + "step": 255 + }, + { + "epoch": 0.40894568690095845, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1711, + "step": 256 + }, + { + "epoch": 0.41054313099041534, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1626, + "step": 257 + }, + { + "epoch": 0.41214057507987223, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.171, + "step": 258 + }, + { + "epoch": 0.41373801916932906, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.1686, + "step": 259 + }, + { + "epoch": 0.41533546325878595, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1719, + "step": 260 + }, + { + "epoch": 0.4169329073482428, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1659, + "step": 261 + }, + { + "epoch": 0.4185303514376997, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.1659, + "step": 262 + }, + { + "epoch": 0.42012779552715657, + "grad_norm": 0.427734375, + "learning_rate": 0.0005, + "loss": 1.1693, + "step": 263 + }, + { + "epoch": 0.4217252396166134, + "grad_norm": 1.4765625, + "learning_rate": 0.0005, + "loss": 1.1913, + "step": 264 + }, + { + "epoch": 0.4233226837060703, + "grad_norm": 2.140625, + "learning_rate": 0.0005, + "loss": 1.2367, + "step": 265 + }, + { + "epoch": 0.4249201277955272, + "grad_norm": 0.66015625, + "learning_rate": 0.0005, + "loss": 1.1874, + "step": 266 + }, + { + "epoch": 0.426517571884984, + "grad_norm": 2.4375, + "learning_rate": 0.0005, + "loss": 1.2668, + "step": 267 + }, + { + "epoch": 0.4281150159744409, + "grad_norm": 0.81640625, + "learning_rate": 0.0005, + "loss": 1.2247, + "step": 268 + }, + { + "epoch": 0.42971246006389774, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.2079, + "step": 269 + }, + { + "epoch": 0.43130990415335463, + "grad_norm": 0.37890625, + "learning_rate": 0.0005, + "loss": 1.2128, + "step": 270 + }, + { + "epoch": 0.4329073482428115, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.2106, + "step": 271 + }, + { + "epoch": 0.43450479233226835, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.2115, + "step": 272 + }, + { + "epoch": 0.43610223642172524, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.2015, + "step": 273 + }, + { + "epoch": 0.43769968051118213, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1962, + "step": 274 + }, + { + "epoch": 0.43929712460063897, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.194, + "step": 275 + }, + { + "epoch": 0.44089456869009586, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1941, + "step": 276 + }, + { + "epoch": 0.4424920127795527, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.1882, + "step": 277 + }, + { + "epoch": 0.4440894568690096, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1838, + "step": 278 + }, + { + "epoch": 0.44568690095846647, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1878, + "step": 279 + }, + { + "epoch": 0.4472843450479233, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.1924, + "step": 280 + }, + { + "epoch": 0.4488817891373802, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.1842, + "step": 281 + }, + { + "epoch": 0.4504792332268371, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1834, + "step": 282 + }, + { + "epoch": 0.4520766773162939, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.176, + "step": 283 + }, + { + "epoch": 0.4536741214057508, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1806, + "step": 284 + }, + { + "epoch": 0.45527156549520764, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1785, + "step": 285 + }, + { + "epoch": 0.45686900958466453, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.178, + "step": 286 + }, + { + "epoch": 0.4584664536741214, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1791, + "step": 287 + }, + { + "epoch": 0.46006389776357826, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.176, + "step": 288 + }, + { + "epoch": 0.46166134185303515, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1777, + "step": 289 + }, + { + "epoch": 0.46325878594249204, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1672, + "step": 290 + }, + { + "epoch": 0.46485623003194887, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1717, + "step": 291 + }, + { + "epoch": 0.46645367412140576, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1705, + "step": 292 + }, + { + "epoch": 0.4680511182108626, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1731, + "step": 293 + }, + { + "epoch": 0.4696485623003195, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1703, + "step": 294 + }, + { + "epoch": 0.4712460063897764, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1692, + "step": 295 + }, + { + "epoch": 0.4728434504792332, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1634, + "step": 296 + }, + { + "epoch": 0.4744408945686901, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1687, + "step": 297 + }, + { + "epoch": 0.476038338658147, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1705, + "step": 298 + }, + { + "epoch": 0.4776357827476038, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1623, + "step": 299 + }, + { + "epoch": 0.4792332268370607, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.1619, + "step": 300 + }, + { + "epoch": 0.48083067092651754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1682, + "step": 301 + }, + { + "epoch": 0.48242811501597443, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1659, + "step": 302 + }, + { + "epoch": 0.4840255591054313, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1653, + "step": 303 + }, + { + "epoch": 0.48562300319488816, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1651, + "step": 304 + }, + { + "epoch": 0.48722044728434505, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1712, + "step": 305 + }, + { + "epoch": 0.48881789137380194, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1699, + "step": 306 + }, + { + "epoch": 0.4904153354632588, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 307 + }, + { + "epoch": 0.49201277955271566, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1681, + "step": 308 + }, + { + "epoch": 0.4936102236421725, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.163, + "step": 309 + }, + { + "epoch": 0.4952076677316294, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1573, + "step": 310 + }, + { + "epoch": 0.4968051118210863, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1596, + "step": 311 + }, + { + "epoch": 0.4984025559105431, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1613, + "step": 312 + }, + { + "epoch": 0.5, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1677, + "step": 313 + }, + { + "epoch": 0.5015974440894568, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 314 + }, + { + "epoch": 0.5031948881789138, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.155, + "step": 315 + }, + { + "epoch": 0.5047923322683706, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.173, + "step": 316 + }, + { + "epoch": 0.5063897763578274, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1627, + "step": 317 + }, + { + "epoch": 0.5079872204472844, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 318 + }, + { + "epoch": 0.5095846645367412, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1573, + "step": 319 + }, + { + "epoch": 0.5111821086261981, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.16, + "step": 320 + }, + { + "epoch": 0.512779552715655, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1642, + "step": 321 + }, + { + "epoch": 0.5143769968051118, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1613, + "step": 322 + }, + { + "epoch": 0.5159744408945687, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1608, + "step": 323 + }, + { + "epoch": 0.5175718849840255, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1655, + "step": 324 + }, + { + "epoch": 0.5191693290734825, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1624, + "step": 325 + }, + { + "epoch": 0.5207667731629393, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1572, + "step": 326 + }, + { + "epoch": 0.5223642172523961, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1549, + "step": 327 + }, + { + "epoch": 0.5239616613418531, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1572, + "step": 328 + }, + { + "epoch": 0.5255591054313099, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1581, + "step": 329 + }, + { + "epoch": 0.5271565495207667, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.1651, + "step": 330 + }, + { + "epoch": 0.5287539936102237, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 331 + }, + { + "epoch": 0.5303514376996805, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1526, + "step": 332 + }, + { + "epoch": 0.5319488817891374, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1547, + "step": 333 + }, + { + "epoch": 0.5335463258785943, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1605, + "step": 334 + }, + { + "epoch": 0.5351437699680511, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1533, + "step": 335 + }, + { + "epoch": 0.536741214057508, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.1524, + "step": 336 + }, + { + "epoch": 0.5383386581469649, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.152, + "step": 337 + }, + { + "epoch": 0.5399361022364217, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1595, + "step": 338 + }, + { + "epoch": 0.5415335463258786, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1675, + "step": 339 + }, + { + "epoch": 0.5431309904153354, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.1577, + "step": 340 + }, + { + "epoch": 0.5447284345047924, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1598, + "step": 341 + }, + { + "epoch": 0.5463258785942492, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1569, + "step": 342 + }, + { + "epoch": 0.547923322683706, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1548, + "step": 343 + }, + { + "epoch": 0.549520766773163, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1543, + "step": 344 + }, + { + "epoch": 0.5511182108626198, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1538, + "step": 345 + }, + { + "epoch": 0.5527156549520766, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.1539, + "step": 346 + }, + { + "epoch": 0.5543130990415336, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1618, + "step": 347 + }, + { + "epoch": 0.5559105431309904, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1558, + "step": 348 + }, + { + "epoch": 0.5575079872204473, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1596, + "step": 349 + }, + { + "epoch": 0.5591054313099042, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.1602, + "step": 350 + }, + { + "epoch": 0.560702875399361, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1512, + "step": 351 + }, + { + "epoch": 0.5623003194888179, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.1569, + "step": 352 + }, + { + "epoch": 0.5638977635782748, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.1529, + "step": 353 + }, + { + "epoch": 0.5654952076677316, + "grad_norm": 0.419921875, + "learning_rate": 0.0005, + "loss": 1.1527, + "step": 354 + }, + { + "epoch": 0.5670926517571885, + "grad_norm": 1.3125, + "learning_rate": 0.0005, + "loss": 1.1742, + "step": 355 + }, + { + "epoch": 0.5686900958466453, + "grad_norm": 1.5625, + "learning_rate": 0.0005, + "loss": 1.2129, + "step": 356 + }, + { + "epoch": 0.5702875399361023, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.1702, + "step": 357 + }, + { + "epoch": 0.5718849840255591, + "grad_norm": 0.8203125, + "learning_rate": 0.0005, + "loss": 1.1896, + "step": 358 + }, + { + "epoch": 0.5734824281150159, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.1831, + "step": 359 + }, + { + "epoch": 0.5750798722044729, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.1721, + "step": 360 + }, + { + "epoch": 0.5766773162939297, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.1761, + "step": 361 + }, + { + "epoch": 0.5782747603833865, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.174, + "step": 362 + }, + { + "epoch": 0.5798722044728435, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.1736, + "step": 363 + }, + { + "epoch": 0.5814696485623003, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.1748, + "step": 364 + }, + { + "epoch": 0.5830670926517572, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.1559, + "step": 365 + }, + { + "epoch": 0.5846645367412141, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1733, + "step": 366 + }, + { + "epoch": 0.5862619808306709, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.1667, + "step": 367 + }, + { + "epoch": 0.5878594249201278, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1627, + "step": 368 + }, + { + "epoch": 0.5894568690095847, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1644, + "step": 369 + }, + { + "epoch": 0.5910543130990416, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1565, + "step": 370 + }, + { + "epoch": 0.5926517571884984, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1622, + "step": 371 + }, + { + "epoch": 0.5942492012779552, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1589, + "step": 372 + }, + { + "epoch": 0.5958466453674122, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1561, + "step": 373 + }, + { + "epoch": 0.597444089456869, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1561, + "step": 374 + }, + { + "epoch": 0.5990415335463258, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1579, + "step": 375 + }, + { + "epoch": 0.6006389776357828, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.161, + "step": 376 + }, + { + "epoch": 0.6022364217252396, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1574, + "step": 377 + }, + { + "epoch": 0.6038338658146964, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1564, + "step": 378 + }, + { + "epoch": 0.6054313099041534, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1547, + "step": 379 + }, + { + "epoch": 0.6070287539936102, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1603, + "step": 380 + }, + { + "epoch": 0.6086261980830671, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1552, + "step": 381 + }, + { + "epoch": 0.610223642172524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1622, + "step": 382 + }, + { + "epoch": 0.6118210862619808, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1624, + "step": 383 + }, + { + "epoch": 0.6134185303514377, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1534, + "step": 384 + }, + { + "epoch": 0.6150159744408946, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1592, + "step": 385 + }, + { + "epoch": 0.6166134185303515, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1529, + "step": 386 + }, + { + "epoch": 0.6182108626198083, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.1488, + "step": 387 + }, + { + "epoch": 0.6198083067092651, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1631, + "step": 388 + }, + { + "epoch": 0.6214057507987221, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1545, + "step": 389 + }, + { + "epoch": 0.6230031948881789, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1534, + "step": 390 + }, + { + "epoch": 0.6246006389776357, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1502, + "step": 391 + }, + { + "epoch": 0.6261980830670927, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.16, + "step": 392 + }, + { + "epoch": 0.6277955271565495, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1615, + "step": 393 + }, + { + "epoch": 0.6293929712460063, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.152, + "step": 394 + }, + { + "epoch": 0.6309904153354633, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1425, + "step": 395 + }, + { + "epoch": 0.6325878594249201, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1518, + "step": 396 + }, + { + "epoch": 0.634185303514377, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1514, + "step": 397 + }, + { + "epoch": 0.6357827476038339, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1516, + "step": 398 + }, + { + "epoch": 0.6373801916932907, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1415, + "step": 399 + }, + { + "epoch": 0.6389776357827476, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1537, + "step": 400 + }, + { + "epoch": 0.6405750798722045, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.154, + "step": 401 + }, + { + "epoch": 0.6421725239616614, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1529, + "step": 402 + }, + { + "epoch": 0.6437699680511182, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.144, + "step": 403 + }, + { + "epoch": 0.645367412140575, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1508, + "step": 404 + }, + { + "epoch": 0.646964856230032, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1469, + "step": 405 + }, + { + "epoch": 0.6485623003194888, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.154, + "step": 406 + }, + { + "epoch": 0.6501597444089456, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1507, + "step": 407 + }, + { + "epoch": 0.6517571884984026, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1458, + "step": 408 + }, + { + "epoch": 0.6533546325878594, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1458, + "step": 409 + }, + { + "epoch": 0.6549520766773163, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1489, + "step": 410 + }, + { + "epoch": 0.6565495207667732, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1467, + "step": 411 + }, + { + "epoch": 0.65814696485623, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1494, + "step": 412 + }, + { + "epoch": 0.6597444089456869, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1538, + "step": 413 + }, + { + "epoch": 0.6613418530351438, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1482, + "step": 414 + }, + { + "epoch": 0.6629392971246006, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1509, + "step": 415 + }, + { + "epoch": 0.6645367412140575, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1487, + "step": 416 + }, + { + "epoch": 0.6661341853035144, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1535, + "step": 417 + }, + { + "epoch": 0.6677316293929713, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1516, + "step": 418 + }, + { + "epoch": 0.6693290734824281, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1538, + "step": 419 + }, + { + "epoch": 0.670926517571885, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1428, + "step": 420 + }, + { + "epoch": 0.6725239616613419, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 421 + }, + { + "epoch": 0.6741214057507987, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1504, + "step": 422 + }, + { + "epoch": 0.6757188498402555, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1445, + "step": 423 + }, + { + "epoch": 0.6773162939297125, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 424 + }, + { + "epoch": 0.6789137380191693, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1373, + "step": 425 + }, + { + "epoch": 0.6805111821086262, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1476, + "step": 426 + }, + { + "epoch": 0.6821086261980831, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1445, + "step": 427 + }, + { + "epoch": 0.6837060702875399, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1463, + "step": 428 + }, + { + "epoch": 0.6853035143769968, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1443, + "step": 429 + }, + { + "epoch": 0.6869009584664537, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 430 + }, + { + "epoch": 0.6884984025559105, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 431 + }, + { + "epoch": 0.6900958466453674, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1411, + "step": 432 + }, + { + "epoch": 0.6916932907348243, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1499, + "step": 433 + }, + { + "epoch": 0.6932907348242812, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1478, + "step": 434 + }, + { + "epoch": 0.694888178913738, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 435 + }, + { + "epoch": 0.6964856230031949, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.145, + "step": 436 + }, + { + "epoch": 0.6980830670926518, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1501, + "step": 437 + }, + { + "epoch": 0.6996805111821086, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1392, + "step": 438 + }, + { + "epoch": 0.7012779552715654, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1467, + "step": 439 + }, + { + "epoch": 0.7028753993610224, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1399, + "step": 440 + }, + { + "epoch": 0.7044728434504792, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1474, + "step": 441 + }, + { + "epoch": 0.7060702875399361, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1481, + "step": 442 + }, + { + "epoch": 0.707667731629393, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1448, + "step": 443 + }, + { + "epoch": 0.7092651757188498, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1421, + "step": 444 + }, + { + "epoch": 0.7108626198083067, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1441, + "step": 445 + }, + { + "epoch": 0.7124600638977636, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1408, + "step": 446 + }, + { + "epoch": 0.7140575079872205, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1486, + "step": 447 + }, + { + "epoch": 0.7156549520766773, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1442, + "step": 448 + }, + { + "epoch": 0.7172523961661342, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.1455, + "step": 449 + }, + { + "epoch": 0.7188498402555911, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.1424, + "step": 450 + }, + { + "epoch": 0.7204472843450479, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1436, + "step": 451 + }, + { + "epoch": 0.7220447284345048, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1496, + "step": 452 + }, + { + "epoch": 0.7236421725239617, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.1472, + "step": 453 + }, + { + "epoch": 0.7252396166134185, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1452, + "step": 454 + }, + { + "epoch": 0.7268370607028753, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1463, + "step": 455 + }, + { + "epoch": 0.7284345047923323, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1481, + "step": 456 + }, + { + "epoch": 0.7300319488817891, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 457 + }, + { + "epoch": 0.731629392971246, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1487, + "step": 458 + }, + { + "epoch": 0.7332268370607029, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 459 + }, + { + "epoch": 0.7348242811501597, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1434, + "step": 460 + }, + { + "epoch": 0.7364217252396166, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1468, + "step": 461 + }, + { + "epoch": 0.7380191693290735, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 462 + }, + { + "epoch": 0.7396166134185304, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1406, + "step": 463 + }, + { + "epoch": 0.7412140575079872, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1433, + "step": 464 + }, + { + "epoch": 0.7428115015974441, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1493, + "step": 465 + }, + { + "epoch": 0.744408945686901, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1488, + "step": 466 + }, + { + "epoch": 0.7460063897763578, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1449, + "step": 467 + }, + { + "epoch": 0.7476038338658147, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1379, + "step": 468 + }, + { + "epoch": 0.7492012779552716, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1415, + "step": 469 + }, + { + "epoch": 0.7507987220447284, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.1439, + "step": 470 + }, + { + "epoch": 0.7523961661341853, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.1436, + "step": 471 + }, + { + "epoch": 0.7539936102236422, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1468, + "step": 472 + }, + { + "epoch": 0.755591054313099, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1449, + "step": 473 + }, + { + "epoch": 0.7571884984025559, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1393, + "step": 474 + }, + { + "epoch": 0.7587859424920128, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1366, + "step": 475 + }, + { + "epoch": 0.7603833865814696, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.1453, + "step": 476 + }, + { + "epoch": 0.7619808306709265, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1398, + "step": 477 + }, + { + "epoch": 0.7635782747603834, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 478 + }, + { + "epoch": 0.7651757188498403, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1498, + "step": 479 + }, + { + "epoch": 0.7667731629392971, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1414, + "step": 480 + }, + { + "epoch": 0.768370607028754, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.143, + "step": 481 + }, + { + "epoch": 0.7699680511182109, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 482 + }, + { + "epoch": 0.7715654952076677, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1431, + "step": 483 + }, + { + "epoch": 0.7731629392971247, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.1494, + "step": 484 + }, + { + "epoch": 0.7747603833865815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1452, + "step": 485 + }, + { + "epoch": 0.7763578274760383, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1342, + "step": 486 + }, + { + "epoch": 0.7779552715654952, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 487 + }, + { + "epoch": 0.7795527156549521, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1376, + "step": 488 + }, + { + "epoch": 0.7811501597444089, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1412, + "step": 489 + }, + { + "epoch": 0.7827476038338658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1347, + "step": 490 + }, + { + "epoch": 0.7843450479233227, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.1423, + "step": 491 + }, + { + "epoch": 0.7859424920127795, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1385, + "step": 492 + }, + { + "epoch": 0.7875399361022364, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.1343, + "step": 493 + }, + { + "epoch": 0.7891373801916933, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1404, + "step": 494 + }, + { + "epoch": 0.7907348242811502, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.1346, + "step": 495 + }, + { + "epoch": 0.792332268370607, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1386, + "step": 496 + }, + { + "epoch": 0.7939297124600639, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1481, + "step": 497 + }, + { + "epoch": 0.7955271565495208, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1382, + "step": 498 + }, + { + "epoch": 0.7971246006389776, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1337, + "step": 499 + }, + { + "epoch": 0.7987220447284346, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1437, + "step": 500 + }, + { + "epoch": 0.8003194888178914, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.139, + "step": 501 + }, + { + "epoch": 0.8019169329073482, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1396, + "step": 502 + }, + { + "epoch": 0.8035143769968051, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1437, + "step": 503 + }, + { + "epoch": 0.805111821086262, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1411, + "step": 504 + }, + { + "epoch": 0.8067092651757188, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.1381, + "step": 505 + }, + { + "epoch": 0.8083067092651757, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1341, + "step": 506 + }, + { + "epoch": 0.8099041533546326, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1376, + "step": 507 + }, + { + "epoch": 0.8115015974440895, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.1377, + "step": 508 + }, + { + "epoch": 0.8130990415335463, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1346, + "step": 509 + }, + { + "epoch": 0.8146964856230032, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 510 + }, + { + "epoch": 0.8162939297124601, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 511 + }, + { + "epoch": 0.8178913738019169, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1397, + "step": 512 + }, + { + "epoch": 0.8194888178913738, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1431, + "step": 513 + }, + { + "epoch": 0.8210862619808307, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1336, + "step": 514 + }, + { + "epoch": 0.8226837060702875, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 515 + }, + { + "epoch": 0.8242811501597445, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.1342, + "step": 516 + }, + { + "epoch": 0.8258785942492013, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1366, + "step": 517 + }, + { + "epoch": 0.8274760383386581, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1398, + "step": 518 + }, + { + "epoch": 0.829073482428115, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1421, + "step": 519 + }, + { + "epoch": 0.8306709265175719, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.1385, + "step": 520 + }, + { + "epoch": 0.8322683706070287, + "grad_norm": 0.2255859375, + "learning_rate": 0.0005, + "loss": 1.1344, + "step": 521 + }, + { + "epoch": 0.8338658146964856, + "grad_norm": 0.466796875, + "learning_rate": 0.0005, + "loss": 1.1325, + "step": 522 + }, + { + "epoch": 0.8354632587859425, + "grad_norm": 1.5859375, + "learning_rate": 0.0005, + "loss": 1.1728, + "step": 523 + }, + { + "epoch": 0.8370607028753994, + "grad_norm": 1.8125, + "learning_rate": 0.0005, + "loss": 1.1896, + "step": 524 + }, + { + "epoch": 0.8386581469648562, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.1537, + "step": 525 + }, + { + "epoch": 0.8402555910543131, + "grad_norm": 1.796875, + "learning_rate": 0.0005, + "loss": 1.2008, + "step": 526 + }, + { + "epoch": 0.84185303514377, + "grad_norm": 0.82421875, + "learning_rate": 0.0005, + "loss": 1.1803, + "step": 527 + }, + { + "epoch": 0.8434504792332268, + "grad_norm": 0.859375, + "learning_rate": 0.0005, + "loss": 1.1776, + "step": 528 + }, + { + "epoch": 0.8450479233226837, + "grad_norm": 0.408203125, + "learning_rate": 0.0005, + "loss": 1.1669, + "step": 529 + }, + { + "epoch": 0.8466453674121406, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.1631, + "step": 530 + }, + { + "epoch": 0.8482428115015974, + "grad_norm": 0.4921875, + "learning_rate": 0.0005, + "loss": 1.1665, + "step": 531 + }, + { + "epoch": 0.8498402555910544, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.149, + "step": 532 + }, + { + "epoch": 0.8514376996805112, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1572, + "step": 533 + }, + { + "epoch": 0.853035143769968, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.1569, + "step": 534 + }, + { + "epoch": 0.854632587859425, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1521, + "step": 535 + }, + { + "epoch": 0.8562300319488818, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1544, + "step": 536 + }, + { + "epoch": 0.8578274760383386, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.1528, + "step": 537 + }, + { + "epoch": 0.8594249201277955, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.1478, + "step": 538 + }, + { + "epoch": 0.8610223642172524, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.1565, + "step": 539 + }, + { + "epoch": 0.8626198083067093, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1381, + "step": 540 + }, + { + "epoch": 0.8642172523961661, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.14, + "step": 541 + }, + { + "epoch": 0.865814696485623, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1485, + "step": 542 + }, + { + "epoch": 0.8674121405750799, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1363, + "step": 543 + }, + { + "epoch": 0.8690095846645367, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.1419, + "step": 544 + }, + { + "epoch": 0.8706070287539937, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1422, + "step": 545 + }, + { + "epoch": 0.8722044728434505, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1422, + "step": 546 + }, + { + "epoch": 0.8738019169329073, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1393, + "step": 547 + }, + { + "epoch": 0.8753993610223643, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1409, + "step": 548 + }, + { + "epoch": 0.8769968051118211, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1389, + "step": 549 + }, + { + "epoch": 0.8785942492012779, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1445, + "step": 550 + }, + { + "epoch": 0.8801916932907349, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1483, + "step": 551 + }, + { + "epoch": 0.8817891373801917, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.136, + "step": 552 + }, + { + "epoch": 0.8833865814696485, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1369, + "step": 553 + }, + { + "epoch": 0.8849840255591054, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1393, + "step": 554 + }, + { + "epoch": 0.8865814696485623, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1409, + "step": 555 + }, + { + "epoch": 0.8881789137380192, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1412, + "step": 556 + }, + { + "epoch": 0.889776357827476, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1403, + "step": 557 + }, + { + "epoch": 0.8913738019169329, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1394, + "step": 558 + }, + { + "epoch": 0.8929712460063898, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.142, + "step": 559 + }, + { + "epoch": 0.8945686900958466, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1377, + "step": 560 + }, + { + "epoch": 0.8961661341853036, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1366, + "step": 561 + }, + { + "epoch": 0.8977635782747604, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.132, + "step": 562 + }, + { + "epoch": 0.8993610223642172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 563 + }, + { + "epoch": 0.9009584664536742, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1332, + "step": 564 + }, + { + "epoch": 0.902555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1398, + "step": 565 + }, + { + "epoch": 0.9041533546325878, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 566 + }, + { + "epoch": 0.9057507987220448, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1395, + "step": 567 + }, + { + "epoch": 0.9073482428115016, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1408, + "step": 568 + }, + { + "epoch": 0.9089456869009584, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1376, + "step": 569 + }, + { + "epoch": 0.9105431309904153, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1392, + "step": 570 + }, + { + "epoch": 0.9121405750798722, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1396, + "step": 571 + }, + { + "epoch": 0.9137380191693291, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1382, + "step": 572 + }, + { + "epoch": 0.9153354632587859, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 573 + }, + { + "epoch": 0.9169329073482428, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1336, + "step": 574 + }, + { + "epoch": 0.9185303514376997, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 575 + }, + { + "epoch": 0.9201277955271565, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 576 + }, + { + "epoch": 0.9217252396166135, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1305, + "step": 577 + }, + { + "epoch": 0.9233226837060703, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 578 + }, + { + "epoch": 0.9249201277955271, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1411, + "step": 579 + }, + { + "epoch": 0.9265175718849841, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1365, + "step": 580 + }, + { + "epoch": 0.9281150159744409, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 581 + }, + { + "epoch": 0.9297124600638977, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1318, + "step": 582 + }, + { + "epoch": 0.9313099041533547, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1374, + "step": 583 + }, + { + "epoch": 0.9329073482428115, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.137, + "step": 584 + }, + { + "epoch": 0.9345047923322684, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1343, + "step": 585 + }, + { + "epoch": 0.9361022364217252, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1339, + "step": 586 + }, + { + "epoch": 0.9376996805111821, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1307, + "step": 587 + }, + { + "epoch": 0.939297124600639, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1364, + "step": 588 + }, + { + "epoch": 0.9408945686900958, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1364, + "step": 589 + }, + { + "epoch": 0.9424920127795527, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 590 + }, + { + "epoch": 0.9440894568690096, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1352, + "step": 591 + }, + { + "epoch": 0.9456869009584664, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1364, + "step": 592 + }, + { + "epoch": 0.9472843450479234, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1382, + "step": 593 + }, + { + "epoch": 0.9488817891373802, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1333, + "step": 594 + }, + { + "epoch": 0.950479233226837, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1314, + "step": 595 + }, + { + "epoch": 0.952076677316294, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 596 + }, + { + "epoch": 0.9536741214057508, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1303, + "step": 597 + }, + { + "epoch": 0.9552715654952076, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 598 + }, + { + "epoch": 0.9568690095846646, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1291, + "step": 599 + }, + { + "epoch": 0.9584664536741214, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1397, + "step": 600 + }, + { + "epoch": 0.9600638977635783, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.1302, + "step": 601 + }, + { + "epoch": 0.9616613418530351, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.129, + "step": 602 + }, + { + "epoch": 0.963258785942492, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1343, + "step": 603 + }, + { + "epoch": 0.9648562300319489, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 604 + }, + { + "epoch": 0.9664536741214057, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1325, + "step": 605 + }, + { + "epoch": 0.9680511182108626, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1328, + "step": 606 + }, + { + "epoch": 0.9696485623003195, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1349, + "step": 607 + }, + { + "epoch": 0.9712460063897763, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 608 + }, + { + "epoch": 0.9728434504792333, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 609 + }, + { + "epoch": 0.9744408945686901, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 610 + }, + { + "epoch": 0.9760383386581469, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1284, + "step": 611 + }, + { + "epoch": 0.9776357827476039, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1313, + "step": 612 + }, + { + "epoch": 0.9792332268370607, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 613 + }, + { + "epoch": 0.9808306709265175, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1303, + "step": 614 + }, + { + "epoch": 0.9824281150159745, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1289, + "step": 615 + }, + { + "epoch": 0.9840255591054313, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1326, + "step": 616 + }, + { + "epoch": 0.9856230031948882, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1346, + "step": 617 + }, + { + "epoch": 0.987220447284345, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 618 + }, + { + "epoch": 0.9888178913738019, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1292, + "step": 619 + }, + { + "epoch": 0.9904153354632588, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1349, + "step": 620 + }, + { + "epoch": 0.9920127795527156, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1301, + "step": 621 + }, + { + "epoch": 0.9936102236421726, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1256, + "step": 622 + }, + { + "epoch": 0.9952076677316294, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1375, + "step": 623 + }, + { + "epoch": 0.9968051118210862, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 624 + }, + { + "epoch": 0.9984025559105432, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 625 + }, + { + "epoch": 1.0, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 626 + }, + { + "epoch": 1.0015974440894568, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.131, + "step": 627 + }, + { + "epoch": 1.0031948881789137, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1355, + "step": 628 + }, + { + "epoch": 1.0047923322683705, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 629 + }, + { + "epoch": 1.0063897763578276, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1337, + "step": 630 + }, + { + "epoch": 1.0079872204472844, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1331, + "step": 631 + }, + { + "epoch": 1.0095846645367412, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1246, + "step": 632 + }, + { + "epoch": 1.011182108626198, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1276, + "step": 633 + }, + { + "epoch": 1.012779552715655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1298, + "step": 634 + }, + { + "epoch": 1.0143769968051117, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1247, + "step": 635 + }, + { + "epoch": 1.0159744408945688, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1303, + "step": 636 + }, + { + "epoch": 1.0175718849840256, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 637 + }, + { + "epoch": 1.0191693290734825, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 638 + }, + { + "epoch": 1.0207667731629393, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 639 + }, + { + "epoch": 1.0223642172523961, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 640 + }, + { + "epoch": 1.023961661341853, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 641 + }, + { + "epoch": 1.0255591054313098, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1305, + "step": 642 + }, + { + "epoch": 1.0271565495207668, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 643 + }, + { + "epoch": 1.0287539936102237, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1328, + "step": 644 + }, + { + "epoch": 1.0303514376996805, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 645 + }, + { + "epoch": 1.0319488817891374, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 646 + }, + { + "epoch": 1.0335463258785942, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 647 + }, + { + "epoch": 1.035143769968051, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.1291, + "step": 648 + }, + { + "epoch": 1.036741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1218, + "step": 649 + }, + { + "epoch": 1.038338658146965, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 650 + }, + { + "epoch": 1.0399361022364217, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 651 + }, + { + "epoch": 1.0415335463258786, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 652 + }, + { + "epoch": 1.0431309904153354, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1283, + "step": 653 + }, + { + "epoch": 1.0447284345047922, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.132, + "step": 654 + }, + { + "epoch": 1.0463258785942493, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.128, + "step": 655 + }, + { + "epoch": 1.0479233226837061, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 656 + }, + { + "epoch": 1.049520766773163, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1214, + "step": 657 + }, + { + "epoch": 1.0511182108626198, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1275, + "step": 658 + }, + { + "epoch": 1.0527156549520766, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 659 + }, + { + "epoch": 1.0543130990415335, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1321, + "step": 660 + }, + { + "epoch": 1.0559105431309903, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 661 + }, + { + "epoch": 1.0575079872204474, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 662 + }, + { + "epoch": 1.0591054313099042, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1292, + "step": 663 + }, + { + "epoch": 1.060702875399361, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1309, + "step": 664 + }, + { + "epoch": 1.0623003194888179, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 665 + }, + { + "epoch": 1.0638977635782747, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1296, + "step": 666 + }, + { + "epoch": 1.0654952076677315, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1286, + "step": 667 + }, + { + "epoch": 1.0670926517571886, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 668 + }, + { + "epoch": 1.0686900958466454, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1271, + "step": 669 + }, + { + "epoch": 1.0702875399361023, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1238, + "step": 670 + }, + { + "epoch": 1.071884984025559, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1271, + "step": 671 + }, + { + "epoch": 1.073482428115016, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.1297, + "step": 672 + }, + { + "epoch": 1.0750798722044728, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.1168, + "step": 673 + }, + { + "epoch": 1.0766773162939298, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 674 + }, + { + "epoch": 1.0782747603833867, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 675 + }, + { + "epoch": 1.0798722044728435, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 676 + }, + { + "epoch": 1.0814696485623003, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1294, + "step": 677 + }, + { + "epoch": 1.0830670926517572, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1248, + "step": 678 + }, + { + "epoch": 1.084664536741214, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1313, + "step": 679 + }, + { + "epoch": 1.0862619808306708, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 680 + }, + { + "epoch": 1.0878594249201279, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 681 + }, + { + "epoch": 1.0894568690095847, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 682 + }, + { + "epoch": 1.0910543130990416, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1255, + "step": 683 + }, + { + "epoch": 1.0926517571884984, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 684 + }, + { + "epoch": 1.0942492012779552, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 685 + }, + { + "epoch": 1.095846645367412, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1322, + "step": 686 + }, + { + "epoch": 1.097444089456869, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 687 + }, + { + "epoch": 1.099041533546326, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1314, + "step": 688 + }, + { + "epoch": 1.1006389776357828, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.128, + "step": 689 + }, + { + "epoch": 1.1022364217252396, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 690 + }, + { + "epoch": 1.1038338658146964, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1262, + "step": 691 + }, + { + "epoch": 1.1054313099041533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 692 + }, + { + "epoch": 1.1070287539936103, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1208, + "step": 693 + }, + { + "epoch": 1.1086261980830672, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.123, + "step": 694 + }, + { + "epoch": 1.110223642172524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1287, + "step": 695 + }, + { + "epoch": 1.1118210862619808, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 696 + }, + { + "epoch": 1.1134185303514377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 697 + }, + { + "epoch": 1.1150159744408945, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1319, + "step": 698 + }, + { + "epoch": 1.1166134185303513, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 699 + }, + { + "epoch": 1.1182108626198084, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1214, + "step": 700 + }, + { + "epoch": 1.1198083067092652, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1261, + "step": 701 + }, + { + "epoch": 1.121405750798722, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1327, + "step": 702 + }, + { + "epoch": 1.123003194888179, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 703 + }, + { + "epoch": 1.1246006389776357, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 704 + }, + { + "epoch": 1.1261980830670926, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1237, + "step": 705 + }, + { + "epoch": 1.1277955271565494, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1306, + "step": 706 + }, + { + "epoch": 1.1293929712460065, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1246, + "step": 707 + }, + { + "epoch": 1.1309904153354633, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1263, + "step": 708 + }, + { + "epoch": 1.1325878594249201, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 709 + }, + { + "epoch": 1.134185303514377, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.1217, + "step": 710 + }, + { + "epoch": 1.1357827476038338, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1193, + "step": 711 + }, + { + "epoch": 1.1373801916932909, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1253, + "step": 712 + }, + { + "epoch": 1.1389776357827477, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1311, + "step": 713 + }, + { + "epoch": 1.1405750798722045, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 714 + }, + { + "epoch": 1.1421725239616614, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1302, + "step": 715 + }, + { + "epoch": 1.1437699680511182, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1235, + "step": 716 + }, + { + "epoch": 1.145367412140575, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 717 + }, + { + "epoch": 1.1469648562300319, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.1266, + "step": 718 + }, + { + "epoch": 1.148562300319489, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 719 + }, + { + "epoch": 1.1501597444089458, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 720 + }, + { + "epoch": 1.1517571884984026, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1209, + "step": 721 + }, + { + "epoch": 1.1533546325878594, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.1327, + "step": 722 + }, + { + "epoch": 1.1549520766773163, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1228, + "step": 723 + }, + { + "epoch": 1.156549520766773, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.1277, + "step": 724 + }, + { + "epoch": 1.15814696485623, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 725 + }, + { + "epoch": 1.159744408945687, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 726 + }, + { + "epoch": 1.1613418530351438, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.1213, + "step": 727 + }, + { + "epoch": 1.1629392971246006, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 728 + }, + { + "epoch": 1.1645367412140575, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.1277, + "step": 729 + }, + { + "epoch": 1.1661341853035143, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.1318, + "step": 730 + }, + { + "epoch": 1.1677316293929714, + "grad_norm": 0.33203125, + "learning_rate": 0.0005, + "loss": 1.1302, + "step": 731 + }, + { + "epoch": 1.1693290734824282, + "grad_norm": 1.046875, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 732 + }, + { + "epoch": 1.170926517571885, + "grad_norm": 2.453125, + "learning_rate": 0.0005, + "loss": 1.234, + "step": 733 + }, + { + "epoch": 1.1725239616613419, + "grad_norm": 0.7578125, + "learning_rate": 0.0005, + "loss": 1.1564, + "step": 734 + }, + { + "epoch": 1.1741214057507987, + "grad_norm": 1.328125, + "learning_rate": 0.0005, + "loss": 1.1934, + "step": 735 + }, + { + "epoch": 1.1757188498402555, + "grad_norm": 0.71875, + "learning_rate": 0.0005, + "loss": 1.1747, + "step": 736 + }, + { + "epoch": 1.1773162939297124, + "grad_norm": 0.80078125, + "learning_rate": 0.0005, + "loss": 1.1646, + "step": 737 + }, + { + "epoch": 1.1789137380191694, + "grad_norm": 0.921875, + "learning_rate": 0.0005, + "loss": 1.2053, + "step": 738 + }, + { + "epoch": 1.1805111821086263, + "grad_norm": 0.93359375, + "learning_rate": 0.0005, + "loss": 1.1907, + "step": 739 + }, + { + "epoch": 1.182108626198083, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.1779, + "step": 740 + }, + { + "epoch": 1.18370607028754, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.168, + "step": 741 + }, + { + "epoch": 1.1853035143769968, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.1682, + "step": 742 + }, + { + "epoch": 1.1869009584664536, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1609, + "step": 743 + }, + { + "epoch": 1.1884984025559104, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.1537, + "step": 744 + }, + { + "epoch": 1.1900958466453675, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.1637, + "step": 745 + }, + { + "epoch": 1.1916932907348243, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.16, + "step": 746 + }, + { + "epoch": 1.1932907348242812, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.156, + "step": 747 + }, + { + "epoch": 1.194888178913738, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1511, + "step": 748 + }, + { + "epoch": 1.1964856230031948, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.1443, + "step": 749 + }, + { + "epoch": 1.1980830670926517, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1474, + "step": 750 + }, + { + "epoch": 1.1996805111821087, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.1467, + "step": 751 + }, + { + "epoch": 1.2012779552715656, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1418, + "step": 752 + }, + { + "epoch": 1.2028753993610224, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1434, + "step": 753 + }, + { + "epoch": 1.2044728434504792, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1402, + "step": 754 + }, + { + "epoch": 1.206070287539936, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1406, + "step": 755 + }, + { + "epoch": 1.207667731629393, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1351, + "step": 756 + }, + { + "epoch": 1.20926517571885, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 757 + }, + { + "epoch": 1.2108626198083068, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1402, + "step": 758 + }, + { + "epoch": 1.2124600638977636, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1326, + "step": 759 + }, + { + "epoch": 1.2140575079872205, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1347, + "step": 760 + }, + { + "epoch": 1.2156549520766773, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1353, + "step": 761 + }, + { + "epoch": 1.2172523961661341, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.14, + "step": 762 + }, + { + "epoch": 1.218849840255591, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.1326, + "step": 763 + }, + { + "epoch": 1.220447284345048, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.136, + "step": 764 + }, + { + "epoch": 1.2220447284345048, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 765 + }, + { + "epoch": 1.2236421725239617, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1325, + "step": 766 + }, + { + "epoch": 1.2252396166134185, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1351, + "step": 767 + }, + { + "epoch": 1.2268370607028753, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.131, + "step": 768 + }, + { + "epoch": 1.2284345047923322, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1289, + "step": 769 + }, + { + "epoch": 1.230031948881789, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 770 + }, + { + "epoch": 1.231629392971246, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1219, + "step": 771 + }, + { + "epoch": 1.233226837060703, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1227, + "step": 772 + }, + { + "epoch": 1.2348242811501597, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 773 + }, + { + "epoch": 1.2364217252396166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 774 + }, + { + "epoch": 1.2380191693290734, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 775 + }, + { + "epoch": 1.2396166134185305, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1308, + "step": 776 + }, + { + "epoch": 1.2412140575079873, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1356, + "step": 777 + }, + { + "epoch": 1.2428115015974441, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1287, + "step": 778 + }, + { + "epoch": 1.244408945686901, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 779 + }, + { + "epoch": 1.2460063897763578, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.1331, + "step": 780 + }, + { + "epoch": 1.2476038338658146, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.127, + "step": 781 + }, + { + "epoch": 1.2492012779552715, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 782 + }, + { + "epoch": 1.2507987220447285, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 783 + }, + { + "epoch": 1.2523961661341854, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1241, + "step": 784 + }, + { + "epoch": 1.2539936102236422, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1203, + "step": 785 + }, + { + "epoch": 1.255591054313099, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 786 + }, + { + "epoch": 1.2571884984025559, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 787 + }, + { + "epoch": 1.2587859424920127, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1204, + "step": 788 + }, + { + "epoch": 1.2603833865814695, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1212, + "step": 789 + }, + { + "epoch": 1.2619808306709266, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1245, + "step": 790 + }, + { + "epoch": 1.2635782747603834, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1229, + "step": 791 + }, + { + "epoch": 1.2651757188498403, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.122, + "step": 792 + }, + { + "epoch": 1.266773162939297, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 793 + }, + { + "epoch": 1.268370607028754, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.1196, + "step": 794 + }, + { + "epoch": 1.269968051118211, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1263, + "step": 795 + }, + { + "epoch": 1.2715654952076676, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1289, + "step": 796 + }, + { + "epoch": 1.2731629392971247, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 797 + }, + { + "epoch": 1.2747603833865815, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1288, + "step": 798 + }, + { + "epoch": 1.2763578274760383, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1161, + "step": 799 + }, + { + "epoch": 1.2779552715654952, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1276, + "step": 800 + }, + { + "epoch": 1.279552715654952, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1221, + "step": 801 + }, + { + "epoch": 1.281150159744409, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1276, + "step": 802 + }, + { + "epoch": 1.2827476038338659, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 803 + }, + { + "epoch": 1.2843450479233227, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1208, + "step": 804 + }, + { + "epoch": 1.2859424920127795, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1251, + "step": 805 + }, + { + "epoch": 1.2875399361022364, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 806 + }, + { + "epoch": 1.2891373801916932, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1278, + "step": 807 + }, + { + "epoch": 1.29073482428115, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 808 + }, + { + "epoch": 1.292332268370607, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 809 + }, + { + "epoch": 1.293929712460064, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 810 + }, + { + "epoch": 1.2955271565495208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 811 + }, + { + "epoch": 1.2971246006389776, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1252, + "step": 812 + }, + { + "epoch": 1.2987220447284344, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.118, + "step": 813 + }, + { + "epoch": 1.3003194888178915, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.1198, + "step": 814 + }, + { + "epoch": 1.3019169329073481, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1241, + "step": 815 + }, + { + "epoch": 1.3035143769968052, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 816 + }, + { + "epoch": 1.305111821086262, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1231, + "step": 817 + }, + { + "epoch": 1.3067092651757188, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1227, + "step": 818 + }, + { + "epoch": 1.3083067092651757, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 819 + }, + { + "epoch": 1.3099041533546325, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1242, + "step": 820 + }, + { + "epoch": 1.3115015974440896, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1278, + "step": 821 + }, + { + "epoch": 1.3130990415335464, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 822 + }, + { + "epoch": 1.3146964856230032, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 823 + }, + { + "epoch": 1.31629392971246, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1217, + "step": 824 + }, + { + "epoch": 1.317891373801917, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1279, + "step": 825 + }, + { + "epoch": 1.3194888178913737, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.1239, + "step": 826 + }, + { + "epoch": 1.3210862619808306, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 827 + }, + { + "epoch": 1.3226837060702876, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1231, + "step": 828 + }, + { + "epoch": 1.3242811501597445, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 829 + }, + { + "epoch": 1.3258785942492013, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1159, + "step": 830 + }, + { + "epoch": 1.3274760383386581, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 831 + }, + { + "epoch": 1.329073482428115, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1199, + "step": 832 + }, + { + "epoch": 1.330670926517572, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 833 + }, + { + "epoch": 1.3322683706070286, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1239, + "step": 834 + }, + { + "epoch": 1.3338658146964857, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1207, + "step": 835 + }, + { + "epoch": 1.3354632587859425, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.123, + "step": 836 + }, + { + "epoch": 1.3370607028753994, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1115, + "step": 837 + }, + { + "epoch": 1.3386581469648562, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1162, + "step": 838 + }, + { + "epoch": 1.340255591054313, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 839 + }, + { + "epoch": 1.34185303514377, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 840 + }, + { + "epoch": 1.343450479233227, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 841 + }, + { + "epoch": 1.3450479233226837, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 842 + }, + { + "epoch": 1.3466453674121406, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 843 + }, + { + "epoch": 1.3482428115015974, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1272, + "step": 844 + }, + { + "epoch": 1.3498402555910542, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1135, + "step": 845 + }, + { + "epoch": 1.351437699680511, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 846 + }, + { + "epoch": 1.3530351437699681, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1167, + "step": 847 + }, + { + "epoch": 1.354632587859425, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1239, + "step": 848 + }, + { + "epoch": 1.3562300319488818, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1149, + "step": 849 + }, + { + "epoch": 1.3578274760383386, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 850 + }, + { + "epoch": 1.3594249201277955, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 851 + }, + { + "epoch": 1.3610223642172525, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1243, + "step": 852 + }, + { + "epoch": 1.3626198083067091, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.124, + "step": 853 + }, + { + "epoch": 1.3642172523961662, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1211, + "step": 854 + }, + { + "epoch": 1.365814696485623, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 855 + }, + { + "epoch": 1.3674121405750799, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 856 + }, + { + "epoch": 1.3690095846645367, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 857 + }, + { + "epoch": 1.3706070287539935, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 858 + }, + { + "epoch": 1.3722044728434506, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1202, + "step": 859 + }, + { + "epoch": 1.3738019169329074, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 860 + }, + { + "epoch": 1.3753993610223643, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1233, + "step": 861 + }, + { + "epoch": 1.376996805111821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 862 + }, + { + "epoch": 1.378594249201278, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1202, + "step": 863 + }, + { + "epoch": 1.3801916932907348, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1258, + "step": 864 + }, + { + "epoch": 1.3817891373801916, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.123, + "step": 865 + }, + { + "epoch": 1.3833865814696487, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1204, + "step": 866 + }, + { + "epoch": 1.3849840255591055, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1237, + "step": 867 + }, + { + "epoch": 1.3865814696485623, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 868 + }, + { + "epoch": 1.3881789137380192, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1146, + "step": 869 + }, + { + "epoch": 1.389776357827476, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1288, + "step": 870 + }, + { + "epoch": 1.3913738019169328, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.118, + "step": 871 + }, + { + "epoch": 1.3929712460063897, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 872 + }, + { + "epoch": 1.3945686900958467, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 873 + }, + { + "epoch": 1.3961661341853036, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1256, + "step": 874 + }, + { + "epoch": 1.3977635782747604, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 875 + }, + { + "epoch": 1.3993610223642172, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 876 + }, + { + "epoch": 1.400958466453674, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.12, + "step": 877 + }, + { + "epoch": 1.4025559105431311, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 878 + }, + { + "epoch": 1.4041533546325877, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 879 + }, + { + "epoch": 1.4057507987220448, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1214, + "step": 880 + }, + { + "epoch": 1.4073482428115016, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 881 + }, + { + "epoch": 1.4089456869009584, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 882 + }, + { + "epoch": 1.4105431309904153, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 883 + }, + { + "epoch": 1.4121405750798721, + "grad_norm": 0.53515625, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 884 + }, + { + "epoch": 1.4137380191693292, + "grad_norm": 1.3359375, + "learning_rate": 0.0005, + "loss": 1.1437, + "step": 885 + }, + { + "epoch": 1.415335463258786, + "grad_norm": 1.7578125, + "learning_rate": 0.0005, + "loss": 1.1625, + "step": 886 + }, + { + "epoch": 1.4169329073482428, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.1297, + "step": 887 + }, + { + "epoch": 1.4185303514376997, + "grad_norm": 1.09375, + "learning_rate": 0.0005, + "loss": 1.147, + "step": 888 + }, + { + "epoch": 1.4201277955271565, + "grad_norm": 0.6328125, + "learning_rate": 0.0005, + "loss": 1.1441, + "step": 889 + }, + { + "epoch": 1.4217252396166133, + "grad_norm": 0.443359375, + "learning_rate": 0.0005, + "loss": 1.1385, + "step": 890 + }, + { + "epoch": 1.4233226837060702, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 891 + }, + { + "epoch": 1.4249201277955272, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.133, + "step": 892 + }, + { + "epoch": 1.426517571884984, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.1278, + "step": 893 + }, + { + "epoch": 1.428115015974441, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.134, + "step": 894 + }, + { + "epoch": 1.4297124600638977, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.1251, + "step": 895 + }, + { + "epoch": 1.4313099041533546, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.1338, + "step": 896 + }, + { + "epoch": 1.4329073482428116, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 897 + }, + { + "epoch": 1.4345047923322682, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.132, + "step": 898 + }, + { + "epoch": 1.4361022364217253, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 899 + }, + { + "epoch": 1.4376996805111821, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 900 + }, + { + "epoch": 1.439297124600639, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 901 + }, + { + "epoch": 1.4408945686900958, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.1308, + "step": 902 + }, + { + "epoch": 1.4424920127795526, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1225, + "step": 903 + }, + { + "epoch": 1.4440894568690097, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 904 + }, + { + "epoch": 1.4456869009584665, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1258, + "step": 905 + }, + { + "epoch": 1.4472843450479234, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.1212, + "step": 906 + }, + { + "epoch": 1.4488817891373802, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 907 + }, + { + "epoch": 1.450479233226837, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.1262, + "step": 908 + }, + { + "epoch": 1.4520766773162939, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 909 + }, + { + "epoch": 1.4536741214057507, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 910 + }, + { + "epoch": 1.4552715654952078, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1197, + "step": 911 + }, + { + "epoch": 1.4568690095846646, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.1187, + "step": 912 + }, + { + "epoch": 1.4584664536741214, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.1165, + "step": 913 + }, + { + "epoch": 1.4600638977635783, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 914 + }, + { + "epoch": 1.461661341853035, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.1211, + "step": 915 + }, + { + "epoch": 1.4632587859424921, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.1258, + "step": 916 + }, + { + "epoch": 1.4648562300319488, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 917 + }, + { + "epoch": 1.4664536741214058, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1159, + "step": 918 + }, + { + "epoch": 1.4680511182108626, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 919 + }, + { + "epoch": 1.4696485623003195, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1264, + "step": 920 + }, + { + "epoch": 1.4712460063897763, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1217, + "step": 921 + }, + { + "epoch": 1.4728434504792332, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 922 + }, + { + "epoch": 1.4744408945686902, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 923 + }, + { + "epoch": 1.476038338658147, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.1215, + "step": 924 + }, + { + "epoch": 1.4776357827476039, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1255, + "step": 925 + }, + { + "epoch": 1.4792332268370607, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 926 + }, + { + "epoch": 1.4808306709265175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 927 + }, + { + "epoch": 1.4824281150159744, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1196, + "step": 928 + }, + { + "epoch": 1.4840255591054312, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 929 + }, + { + "epoch": 1.4856230031948883, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 930 + }, + { + "epoch": 1.487220447284345, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.12, + "step": 931 + }, + { + "epoch": 1.488817891373802, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 932 + }, + { + "epoch": 1.4904153354632588, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 933 + }, + { + "epoch": 1.4920127795527156, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 934 + }, + { + "epoch": 1.4936102236421724, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 935 + }, + { + "epoch": 1.4952076677316293, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 936 + }, + { + "epoch": 1.4968051118210863, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1183, + "step": 937 + }, + { + "epoch": 1.4984025559105432, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1204, + "step": 938 + }, + { + "epoch": 1.5, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1229, + "step": 939 + }, + { + "epoch": 1.5015974440894568, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 940 + }, + { + "epoch": 1.5031948881789137, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1152, + "step": 941 + }, + { + "epoch": 1.5047923322683707, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 942 + }, + { + "epoch": 1.5063897763578273, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1171, + "step": 943 + }, + { + "epoch": 1.5079872204472844, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1181, + "step": 944 + }, + { + "epoch": 1.5095846645367412, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 945 + }, + { + "epoch": 1.511182108626198, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 946 + }, + { + "epoch": 1.5127795527156551, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 947 + }, + { + "epoch": 1.5143769968051117, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1215, + "step": 948 + }, + { + "epoch": 1.5159744408945688, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1109, + "step": 949 + }, + { + "epoch": 1.5175718849840254, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1147, + "step": 950 + }, + { + "epoch": 1.5191693290734825, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 951 + }, + { + "epoch": 1.5207667731629393, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 952 + }, + { + "epoch": 1.5223642172523961, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1136, + "step": 953 + }, + { + "epoch": 1.5239616613418532, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1182, + "step": 954 + }, + { + "epoch": 1.5255591054313098, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1186, + "step": 955 + }, + { + "epoch": 1.5271565495207668, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1253, + "step": 956 + }, + { + "epoch": 1.5287539936102237, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1107, + "step": 957 + }, + { + "epoch": 1.5303514376996805, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 958 + }, + { + "epoch": 1.5319488817891374, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 959 + }, + { + "epoch": 1.5335463258785942, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 960 + }, + { + "epoch": 1.5351437699680512, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 961 + }, + { + "epoch": 1.5367412140575079, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1119, + "step": 962 + }, + { + "epoch": 1.538338658146965, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1248, + "step": 963 + }, + { + "epoch": 1.5399361022364217, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1137, + "step": 964 + }, + { + "epoch": 1.5415335463258786, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 965 + }, + { + "epoch": 1.5431309904153354, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 966 + }, + { + "epoch": 1.5447284345047922, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 967 + }, + { + "epoch": 1.5463258785942493, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 968 + }, + { + "epoch": 1.547923322683706, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 969 + }, + { + "epoch": 1.549520766773163, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1157, + "step": 970 + }, + { + "epoch": 1.5511182108626198, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.1186, + "step": 971 + }, + { + "epoch": 1.5527156549520766, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 972 + }, + { + "epoch": 1.5543130990415337, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.115, + "step": 973 + }, + { + "epoch": 1.5559105431309903, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 974 + }, + { + "epoch": 1.5575079872204474, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 975 + }, + { + "epoch": 1.5591054313099042, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1171, + "step": 976 + }, + { + "epoch": 1.560702875399361, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 977 + }, + { + "epoch": 1.5623003194888179, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1189, + "step": 978 + }, + { + "epoch": 1.5638977635782747, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1177, + "step": 979 + }, + { + "epoch": 1.5654952076677318, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1184, + "step": 980 + }, + { + "epoch": 1.5670926517571884, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 981 + }, + { + "epoch": 1.5686900958466454, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 982 + }, + { + "epoch": 1.5702875399361023, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 983 + }, + { + "epoch": 1.571884984025559, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 984 + }, + { + "epoch": 1.573482428115016, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.116, + "step": 985 + }, + { + "epoch": 1.5750798722044728, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1218, + "step": 986 + }, + { + "epoch": 1.5766773162939298, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 987 + }, + { + "epoch": 1.5782747603833864, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1125, + "step": 988 + }, + { + "epoch": 1.5798722044728435, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 989 + }, + { + "epoch": 1.5814696485623003, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 990 + }, + { + "epoch": 1.5830670926517572, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1174, + "step": 991 + }, + { + "epoch": 1.5846645367412142, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1094, + "step": 992 + }, + { + "epoch": 1.5862619808306708, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 993 + }, + { + "epoch": 1.5878594249201279, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.118, + "step": 994 + }, + { + "epoch": 1.5894568690095847, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.115, + "step": 995 + }, + { + "epoch": 1.5910543130990416, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1146, + "step": 996 + }, + { + "epoch": 1.5926517571884984, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 997 + }, + { + "epoch": 1.5942492012779552, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1151, + "step": 998 + }, + { + "epoch": 1.5958466453674123, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 999 + }, + { + "epoch": 1.5974440894568689, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 1000 + }, + { + "epoch": 1.599041533546326, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1128, + "step": 1001 + }, + { + "epoch": 1.6006389776357828, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 1002 + }, + { + "epoch": 1.6022364217252396, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1112, + "step": 1003 + }, + { + "epoch": 1.6038338658146964, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.121, + "step": 1004 + }, + { + "epoch": 1.6054313099041533, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1143, + "step": 1005 + }, + { + "epoch": 1.6070287539936103, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 1006 + }, + { + "epoch": 1.608626198083067, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 1007 + }, + { + "epoch": 1.610223642172524, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1008 + }, + { + "epoch": 1.6118210862619808, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 1009 + }, + { + "epoch": 1.6134185303514377, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 1010 + }, + { + "epoch": 1.6150159744408947, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1197, + "step": 1011 + }, + { + "epoch": 1.6166134185303513, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.1135, + "step": 1012 + }, + { + "epoch": 1.6182108626198084, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1107, + "step": 1013 + }, + { + "epoch": 1.619808306709265, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 1014 + }, + { + "epoch": 1.621405750798722, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1171, + "step": 1015 + }, + { + "epoch": 1.623003194888179, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1016 + }, + { + "epoch": 1.6246006389776357, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1017 + }, + { + "epoch": 1.6261980830670928, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 1018 + }, + { + "epoch": 1.6277955271565494, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 1019 + }, + { + "epoch": 1.6293929712460065, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1173, + "step": 1020 + }, + { + "epoch": 1.6309904153354633, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1021 + }, + { + "epoch": 1.6325878594249201, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1022 + }, + { + "epoch": 1.634185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1161, + "step": 1023 + }, + { + "epoch": 1.6357827476038338, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1173, + "step": 1024 + }, + { + "epoch": 1.6373801916932909, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1175, + "step": 1025 + }, + { + "epoch": 1.6389776357827475, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1026 + }, + { + "epoch": 1.6405750798722045, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 1027 + }, + { + "epoch": 1.6421725239616614, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1028 + }, + { + "epoch": 1.6437699680511182, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 1029 + }, + { + "epoch": 1.645367412140575, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1118, + "step": 1030 + }, + { + "epoch": 1.6469648562300319, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 1031 + }, + { + "epoch": 1.648562300319489, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 1032 + }, + { + "epoch": 1.6501597444089455, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 1033 + }, + { + "epoch": 1.6517571884984026, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1114, + "step": 1034 + }, + { + "epoch": 1.6533546325878594, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1136, + "step": 1035 + }, + { + "epoch": 1.6549520766773163, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 1036 + }, + { + "epoch": 1.6565495207667733, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1167, + "step": 1037 + }, + { + "epoch": 1.65814696485623, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1038 + }, + { + "epoch": 1.659744408945687, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 1039 + }, + { + "epoch": 1.6613418530351438, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1040 + }, + { + "epoch": 1.6629392971246006, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1041 + }, + { + "epoch": 1.6645367412140575, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 1042 + }, + { + "epoch": 1.6661341853035143, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.1077, + "step": 1043 + }, + { + "epoch": 1.6677316293929714, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1044 + }, + { + "epoch": 1.669329073482428, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1041, + "step": 1045 + }, + { + "epoch": 1.670926517571885, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 1046 + }, + { + "epoch": 1.6725239616613419, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1102, + "step": 1047 + }, + { + "epoch": 1.6741214057507987, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 1048 + }, + { + "epoch": 1.6757188498402555, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1049 + }, + { + "epoch": 1.6773162939297124, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1148, + "step": 1050 + }, + { + "epoch": 1.6789137380191694, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1125, + "step": 1051 + }, + { + "epoch": 1.680511182108626, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1147, + "step": 1052 + }, + { + "epoch": 1.682108626198083, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1053 + }, + { + "epoch": 1.68370607028754, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1157, + "step": 1054 + }, + { + "epoch": 1.6853035143769968, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 1055 + }, + { + "epoch": 1.6869009584664538, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1146, + "step": 1056 + }, + { + "epoch": 1.6884984025559104, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 1057 + }, + { + "epoch": 1.6900958466453675, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1149, + "step": 1058 + }, + { + "epoch": 1.6916932907348243, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1059 + }, + { + "epoch": 1.6932907348242812, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1122, + "step": 1060 + }, + { + "epoch": 1.694888178913738, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1164, + "step": 1061 + }, + { + "epoch": 1.6964856230031948, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1062 + }, + { + "epoch": 1.6980830670926519, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 1063 + }, + { + "epoch": 1.6996805111821085, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1119, + "step": 1064 + }, + { + "epoch": 1.7012779552715656, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 1065 + }, + { + "epoch": 1.7028753993610224, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1165, + "step": 1066 + }, + { + "epoch": 1.7044728434504792, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1067 + }, + { + "epoch": 1.706070287539936, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.116, + "step": 1068 + }, + { + "epoch": 1.707667731629393, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1131, + "step": 1069 + }, + { + "epoch": 1.70926517571885, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.1163, + "step": 1070 + }, + { + "epoch": 1.7108626198083066, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1066, + "step": 1071 + }, + { + "epoch": 1.7124600638977636, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1114, + "step": 1072 + }, + { + "epoch": 1.7140575079872205, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.1142, + "step": 1073 + }, + { + "epoch": 1.7156549520766773, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1177, + "step": 1074 + }, + { + "epoch": 1.7172523961661343, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1075 + }, + { + "epoch": 1.718849840255591, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1121, + "step": 1076 + }, + { + "epoch": 1.720447284345048, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1077 + }, + { + "epoch": 1.7220447284345048, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1112, + "step": 1078 + }, + { + "epoch": 1.7236421725239617, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1079 + }, + { + "epoch": 1.7252396166134185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 1080 + }, + { + "epoch": 1.7268370607028753, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1105, + "step": 1081 + }, + { + "epoch": 1.7284345047923324, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 1082 + }, + { + "epoch": 1.730031948881789, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 1083 + }, + { + "epoch": 1.731629392971246, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 1084 + }, + { + "epoch": 1.733226837060703, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 1085 + }, + { + "epoch": 1.7348242811501597, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 1086 + }, + { + "epoch": 1.7364217252396166, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1087 + }, + { + "epoch": 1.7380191693290734, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1088 + }, + { + "epoch": 1.7396166134185305, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 1089 + }, + { + "epoch": 1.741214057507987, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 1090 + }, + { + "epoch": 1.7428115015974441, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1095, + "step": 1091 + }, + { + "epoch": 1.744408945686901, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1092 + }, + { + "epoch": 1.7460063897763578, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1111, + "step": 1093 + }, + { + "epoch": 1.7476038338658149, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1113, + "step": 1094 + }, + { + "epoch": 1.7492012779552715, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1095 + }, + { + "epoch": 1.7507987220447285, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1096 + }, + { + "epoch": 1.7523961661341851, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1097 + }, + { + "epoch": 1.7539936102236422, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1098 + }, + { + "epoch": 1.755591054313099, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 1099 + }, + { + "epoch": 1.7571884984025559, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 1100 + }, + { + "epoch": 1.758785942492013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1176, + "step": 1101 + }, + { + "epoch": 1.7603833865814695, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1102 + }, + { + "epoch": 1.7619808306709266, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1166, + "step": 1103 + }, + { + "epoch": 1.7635782747603834, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 1104 + }, + { + "epoch": 1.7651757188498403, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1105 + }, + { + "epoch": 1.766773162939297, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1106 + }, + { + "epoch": 1.768370607028754, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1107 + }, + { + "epoch": 1.769968051118211, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 1108 + }, + { + "epoch": 1.7715654952076676, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1109 + }, + { + "epoch": 1.7731629392971247, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1138, + "step": 1110 + }, + { + "epoch": 1.7747603833865815, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1111 + }, + { + "epoch": 1.7763578274760383, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 1112 + }, + { + "epoch": 1.7779552715654952, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1113 + }, + { + "epoch": 1.779552715654952, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1123, + "step": 1114 + }, + { + "epoch": 1.781150159744409, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1105, + "step": 1115 + }, + { + "epoch": 1.7827476038338657, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1116 + }, + { + "epoch": 1.7843450479233227, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 1117 + }, + { + "epoch": 1.7859424920127795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1126, + "step": 1118 + }, + { + "epoch": 1.7875399361022364, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1119 + }, + { + "epoch": 1.7891373801916934, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 1120 + }, + { + "epoch": 1.79073482428115, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1207, + "step": 1121 + }, + { + "epoch": 1.792332268370607, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1122 + }, + { + "epoch": 1.793929712460064, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1123 + }, + { + "epoch": 1.7955271565495208, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 1124 + }, + { + "epoch": 1.7971246006389776, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1125 + }, + { + "epoch": 1.7987220447284344, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1126 + }, + { + "epoch": 1.8003194888178915, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1127 + }, + { + "epoch": 1.8019169329073481, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1128 + }, + { + "epoch": 1.8035143769968052, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1137, + "step": 1129 + }, + { + "epoch": 1.805111821086262, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1162, + "step": 1130 + }, + { + "epoch": 1.8067092651757188, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1131 + }, + { + "epoch": 1.8083067092651757, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1132 + }, + { + "epoch": 1.8099041533546325, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1133 + }, + { + "epoch": 1.8115015974440896, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 1134 + }, + { + "epoch": 1.8130990415335462, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 1135 + }, + { + "epoch": 1.8146964856230032, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1136 + }, + { + "epoch": 1.81629392971246, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1137 + }, + { + "epoch": 1.817891373801917, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1138 + }, + { + "epoch": 1.819488817891374, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1108, + "step": 1139 + }, + { + "epoch": 1.8210862619808306, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1115, + "step": 1140 + }, + { + "epoch": 1.8226837060702876, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1141 + }, + { + "epoch": 1.8242811501597445, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1135, + "step": 1142 + }, + { + "epoch": 1.8258785942492013, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1143 + }, + { + "epoch": 1.8274760383386581, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1144 + }, + { + "epoch": 1.829073482428115, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1048, + "step": 1145 + }, + { + "epoch": 1.830670926517572, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1102, + "step": 1146 + }, + { + "epoch": 1.8322683706070286, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1128, + "step": 1147 + }, + { + "epoch": 1.8338658146964857, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1148 + }, + { + "epoch": 1.8354632587859425, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1149 + }, + { + "epoch": 1.8370607028753994, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1150 + }, + { + "epoch": 1.8386581469648562, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1151 + }, + { + "epoch": 1.840255591054313, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 1152 + }, + { + "epoch": 1.84185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1153 + }, + { + "epoch": 1.8434504792332267, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 1154 + }, + { + "epoch": 1.8450479233226837, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1155 + }, + { + "epoch": 1.8466453674121406, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1156 + }, + { + "epoch": 1.8482428115015974, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1095, + "step": 1157 + }, + { + "epoch": 1.8498402555910545, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1158 + }, + { + "epoch": 1.851437699680511, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 1159 + }, + { + "epoch": 1.8530351437699681, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1160 + }, + { + "epoch": 1.854632587859425, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1119, + "step": 1161 + }, + { + "epoch": 1.8562300319488818, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1162 + }, + { + "epoch": 1.8578274760383386, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1071, + "step": 1163 + }, + { + "epoch": 1.8594249201277955, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1164 + }, + { + "epoch": 1.8610223642172525, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 1165 + }, + { + "epoch": 1.8626198083067091, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1166 + }, + { + "epoch": 1.8642172523961662, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 1167 + }, + { + "epoch": 1.865814696485623, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1168 + }, + { + "epoch": 1.8674121405750799, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1169 + }, + { + "epoch": 1.8690095846645367, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 1170 + }, + { + "epoch": 1.8706070287539935, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1171 + }, + { + "epoch": 1.8722044728434506, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1076, + "step": 1172 + }, + { + "epoch": 1.8738019169329072, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 1173 + }, + { + "epoch": 1.8753993610223643, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1174 + }, + { + "epoch": 1.876996805111821, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1175 + }, + { + "epoch": 1.878594249201278, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1176 + }, + { + "epoch": 1.880191693290735, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1177 + }, + { + "epoch": 1.8817891373801916, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1178 + }, + { + "epoch": 1.8833865814696487, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1179 + }, + { + "epoch": 1.8849840255591053, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1159, + "step": 1180 + }, + { + "epoch": 1.8865814696485623, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1181 + }, + { + "epoch": 1.8881789137380192, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1182 + }, + { + "epoch": 1.889776357827476, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1183 + }, + { + "epoch": 1.891373801916933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1184 + }, + { + "epoch": 1.8929712460063897, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1185 + }, + { + "epoch": 1.8945686900958467, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1186 + }, + { + "epoch": 1.8961661341853036, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1187 + }, + { + "epoch": 1.8977635782747604, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1188 + }, + { + "epoch": 1.8993610223642172, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1189 + }, + { + "epoch": 1.900958466453674, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 1190 + }, + { + "epoch": 1.9025559105431311, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1191 + }, + { + "epoch": 1.9041533546325877, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1192 + }, + { + "epoch": 1.9057507987220448, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 1193 + }, + { + "epoch": 1.9073482428115016, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1194 + }, + { + "epoch": 1.9089456869009584, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 1195 + }, + { + "epoch": 1.9105431309904153, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 1196 + }, + { + "epoch": 1.9121405750798721, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1197 + }, + { + "epoch": 1.9137380191693292, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1198 + }, + { + "epoch": 1.9153354632587858, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1199 + }, + { + "epoch": 1.9169329073482428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1200 + }, + { + "epoch": 1.9185303514376997, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1201 + }, + { + "epoch": 1.9201277955271565, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 1202 + }, + { + "epoch": 1.9217252396166136, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1125, + "step": 1203 + }, + { + "epoch": 1.9233226837060702, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1152, + "step": 1204 + }, + { + "epoch": 1.9249201277955272, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1107, + "step": 1205 + }, + { + "epoch": 1.926517571884984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 1206 + }, + { + "epoch": 1.928115015974441, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1109, + "step": 1207 + }, + { + "epoch": 1.9297124600638977, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1208 + }, + { + "epoch": 1.9313099041533546, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1209 + }, + { + "epoch": 1.9329073482428116, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1210 + }, + { + "epoch": 1.9345047923322682, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1211 + }, + { + "epoch": 1.9361022364217253, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1086, + "step": 1212 + }, + { + "epoch": 1.9376996805111821, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1213 + }, + { + "epoch": 1.939297124600639, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 1214 + }, + { + "epoch": 1.9408945686900958, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1215 + }, + { + "epoch": 1.9424920127795526, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1143, + "step": 1216 + }, + { + "epoch": 1.9440894568690097, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1217 + }, + { + "epoch": 1.9456869009584663, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.114, + "step": 1218 + }, + { + "epoch": 1.9472843450479234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 1219 + }, + { + "epoch": 1.9488817891373802, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1220 + }, + { + "epoch": 1.950479233226837, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 1221 + }, + { + "epoch": 1.952076677316294, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1222 + }, + { + "epoch": 1.9536741214057507, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 1223 + }, + { + "epoch": 1.9552715654952078, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1224 + }, + { + "epoch": 1.9568690095846646, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1225 + }, + { + "epoch": 1.9584664536741214, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 1226 + }, + { + "epoch": 1.9600638977635783, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 1227 + }, + { + "epoch": 1.961661341853035, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1228 + }, + { + "epoch": 1.9632587859424921, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1066, + "step": 1229 + }, + { + "epoch": 1.9648562300319488, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.1033, + "step": 1230 + }, + { + "epoch": 1.9664536741214058, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0984, + "step": 1231 + }, + { + "epoch": 1.9680511182108626, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.1071, + "step": 1232 + }, + { + "epoch": 1.9696485623003195, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1233 + }, + { + "epoch": 1.9712460063897763, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1234 + }, + { + "epoch": 1.9728434504792332, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1235 + }, + { + "epoch": 1.9744408945686902, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1066, + "step": 1236 + }, + { + "epoch": 1.9760383386581468, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 1237 + }, + { + "epoch": 1.9776357827476039, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1238 + }, + { + "epoch": 1.9792332268370607, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1048, + "step": 1239 + }, + { + "epoch": 1.9808306709265175, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1240 + }, + { + "epoch": 1.9824281150159746, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1241 + }, + { + "epoch": 1.9840255591054312, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 1242 + }, + { + "epoch": 1.9856230031948883, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1052, + "step": 1243 + }, + { + "epoch": 1.9872204472843449, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1111, + "step": 1244 + }, + { + "epoch": 1.988817891373802, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1245 + }, + { + "epoch": 1.9904153354632588, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1246 + }, + { + "epoch": 1.9920127795527156, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 1247 + }, + { + "epoch": 1.9936102236421727, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1248 + }, + { + "epoch": 1.9952076677316293, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1043, + "step": 1249 + }, + { + "epoch": 1.9968051118210863, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1043, + "step": 1250 + }, + { + "epoch": 1.9984025559105432, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 1251 + }, + { + "epoch": 2.0, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1252 + }, + { + "epoch": 2.001597444089457, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1253 + }, + { + "epoch": 2.0031948881789137, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.1086, + "step": 1254 + }, + { + "epoch": 2.0047923322683707, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1255 + }, + { + "epoch": 2.0063897763578273, + "grad_norm": 0.73828125, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1256 + }, + { + "epoch": 2.0079872204472844, + "grad_norm": 1.9609375, + "learning_rate": 0.0005, + "loss": 1.1489, + "step": 1257 + }, + { + "epoch": 2.009584664536741, + "grad_norm": 0.431640625, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 1258 + }, + { + "epoch": 2.011182108626198, + "grad_norm": 0.5, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 1259 + }, + { + "epoch": 2.012779552715655, + "grad_norm": 0.921875, + "learning_rate": 0.0005, + "loss": 1.1174, + "step": 1260 + }, + { + "epoch": 2.0143769968051117, + "grad_norm": 0.671875, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 1261 + }, + { + "epoch": 2.015974440894569, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.1147, + "step": 1262 + }, + { + "epoch": 2.0175718849840254, + "grad_norm": 0.8359375, + "learning_rate": 0.0005, + "loss": 1.1288, + "step": 1263 + }, + { + "epoch": 2.0191693290734825, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1264 + }, + { + "epoch": 2.0207667731629395, + "grad_norm": 0.55859375, + "learning_rate": 0.0005, + "loss": 1.1187, + "step": 1265 + }, + { + "epoch": 2.022364217252396, + "grad_norm": 0.52734375, + "learning_rate": 0.0005, + "loss": 1.1175, + "step": 1266 + }, + { + "epoch": 2.023961661341853, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.1108, + "step": 1267 + }, + { + "epoch": 2.02555910543131, + "grad_norm": 0.75, + "learning_rate": 0.0005, + "loss": 1.1191, + "step": 1268 + }, + { + "epoch": 2.027156549520767, + "grad_norm": 0.6875, + "learning_rate": 0.0005, + "loss": 1.1229, + "step": 1269 + }, + { + "epoch": 2.0287539936102235, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.1143, + "step": 1270 + }, + { + "epoch": 2.0303514376996805, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1271 + }, + { + "epoch": 2.0319488817891376, + "grad_norm": 0.326171875, + "learning_rate": 0.0005, + "loss": 1.1166, + "step": 1272 + }, + { + "epoch": 2.033546325878594, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1273 + }, + { + "epoch": 2.0351437699680512, + "grad_norm": 0.392578125, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1274 + }, + { + "epoch": 2.036741214057508, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1111, + "step": 1275 + }, + { + "epoch": 2.038338658146965, + "grad_norm": 0.2216796875, + "learning_rate": 0.0005, + "loss": 1.1128, + "step": 1276 + }, + { + "epoch": 2.0399361022364215, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1277 + }, + { + "epoch": 2.0415335463258786, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.1175, + "step": 1278 + }, + { + "epoch": 2.0431309904153356, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1279 + }, + { + "epoch": 2.0447284345047922, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 1280 + }, + { + "epoch": 2.0463258785942493, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1281 + }, + { + "epoch": 2.047923322683706, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 1282 + }, + { + "epoch": 2.049520766773163, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1283 + }, + { + "epoch": 2.0511182108626196, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1284 + }, + { + "epoch": 2.0527156549520766, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 1285 + }, + { + "epoch": 2.0543130990415337, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 1286 + }, + { + "epoch": 2.0559105431309903, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1043, + "step": 1287 + }, + { + "epoch": 2.0575079872204474, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1288 + }, + { + "epoch": 2.059105431309904, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1289 + }, + { + "epoch": 2.060702875399361, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1290 + }, + { + "epoch": 2.062300319488818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1082, + "step": 1291 + }, + { + "epoch": 2.0638977635782747, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1076, + "step": 1292 + }, + { + "epoch": 2.0654952076677318, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1293 + }, + { + "epoch": 2.0670926517571884, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1294 + }, + { + "epoch": 2.0686900958466454, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1295 + }, + { + "epoch": 2.070287539936102, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1296 + }, + { + "epoch": 2.071884984025559, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1297 + }, + { + "epoch": 2.073482428115016, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1095, + "step": 1298 + }, + { + "epoch": 2.0750798722044728, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1131, + "step": 1299 + }, + { + "epoch": 2.07667731629393, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1300 + }, + { + "epoch": 2.0782747603833864, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1301 + }, + { + "epoch": 2.0798722044728435, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1302 + }, + { + "epoch": 2.0814696485623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1041, + "step": 1303 + }, + { + "epoch": 2.083067092651757, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1304 + }, + { + "epoch": 2.084664536741214, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1305 + }, + { + "epoch": 2.086261980830671, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1306 + }, + { + "epoch": 2.087859424920128, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1307 + }, + { + "epoch": 2.0894568690095845, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1308 + }, + { + "epoch": 2.0910543130990416, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.1025, + "step": 1309 + }, + { + "epoch": 2.0926517571884986, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1310 + }, + { + "epoch": 2.094249201277955, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1053, + "step": 1311 + }, + { + "epoch": 2.0958466453674123, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1312 + }, + { + "epoch": 2.097444089456869, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1313 + }, + { + "epoch": 2.099041533546326, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1314 + }, + { + "epoch": 2.1006389776357826, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 1315 + }, + { + "epoch": 2.1022364217252396, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1316 + }, + { + "epoch": 2.1038338658146967, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1317 + }, + { + "epoch": 2.1054313099041533, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1318 + }, + { + "epoch": 2.1070287539936103, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 1319 + }, + { + "epoch": 2.108626198083067, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1320 + }, + { + "epoch": 2.110223642172524, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1321 + }, + { + "epoch": 2.1118210862619806, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 1322 + }, + { + "epoch": 2.1134185303514377, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1323 + }, + { + "epoch": 2.1150159744408947, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1324 + }, + { + "epoch": 2.1166134185303513, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.102, + "step": 1325 + }, + { + "epoch": 2.1182108626198084, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1326 + }, + { + "epoch": 2.119808306709265, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1327 + }, + { + "epoch": 2.121405750798722, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1328 + }, + { + "epoch": 2.123003194888179, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1329 + }, + { + "epoch": 2.1246006389776357, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1330 + }, + { + "epoch": 2.126198083067093, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1331 + }, + { + "epoch": 2.1277955271565494, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1332 + }, + { + "epoch": 2.1293929712460065, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1333 + }, + { + "epoch": 2.130990415335463, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1334 + }, + { + "epoch": 2.13258785942492, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 1335 + }, + { + "epoch": 2.134185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 1336 + }, + { + "epoch": 2.135782747603834, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1337 + }, + { + "epoch": 2.137380191693291, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1338 + }, + { + "epoch": 2.1389776357827475, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1036, + "step": 1339 + }, + { + "epoch": 2.1405750798722045, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1340 + }, + { + "epoch": 2.142172523961661, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1341 + }, + { + "epoch": 2.143769968051118, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1077, + "step": 1342 + }, + { + "epoch": 2.1453674121405752, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1343 + }, + { + "epoch": 2.146964856230032, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1344 + }, + { + "epoch": 2.148562300319489, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1345 + }, + { + "epoch": 2.1501597444089455, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1346 + }, + { + "epoch": 2.1517571884984026, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1347 + }, + { + "epoch": 2.1533546325878596, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1348 + }, + { + "epoch": 2.1549520766773163, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1349 + }, + { + "epoch": 2.1565495207667733, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1033, + "step": 1350 + }, + { + "epoch": 2.15814696485623, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1351 + }, + { + "epoch": 2.159744408945687, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1352 + }, + { + "epoch": 2.1613418530351436, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1353 + }, + { + "epoch": 2.1629392971246006, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1052, + "step": 1354 + }, + { + "epoch": 2.1645367412140577, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1355 + }, + { + "epoch": 2.1661341853035143, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1356 + }, + { + "epoch": 2.1677316293929714, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1357 + }, + { + "epoch": 2.169329073482428, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.1053, + "step": 1358 + }, + { + "epoch": 2.170926517571885, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1359 + }, + { + "epoch": 2.1725239616613417, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1360 + }, + { + "epoch": 2.1741214057507987, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1361 + }, + { + "epoch": 2.1757188498402558, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1362 + }, + { + "epoch": 2.1773162939297124, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.103, + "step": 1363 + }, + { + "epoch": 2.1789137380191694, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1041, + "step": 1364 + }, + { + "epoch": 2.180511182108626, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 1365 + }, + { + "epoch": 2.182108626198083, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1366 + }, + { + "epoch": 2.18370607028754, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1367 + }, + { + "epoch": 2.1853035143769968, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1368 + }, + { + "epoch": 2.186900958466454, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1369 + }, + { + "epoch": 2.1884984025559104, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1370 + }, + { + "epoch": 2.1900958466453675, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1371 + }, + { + "epoch": 2.191693290734824, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1372 + }, + { + "epoch": 2.193290734824281, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1373 + }, + { + "epoch": 2.194888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1374 + }, + { + "epoch": 2.196485623003195, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1375 + }, + { + "epoch": 2.198083067092652, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1376 + }, + { + "epoch": 2.1996805111821085, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1377 + }, + { + "epoch": 2.2012779552715656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1082, + "step": 1378 + }, + { + "epoch": 2.202875399361022, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1379 + }, + { + "epoch": 2.2044728434504792, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1380 + }, + { + "epoch": 2.2060702875399363, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1071, + "step": 1381 + }, + { + "epoch": 2.207667731629393, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1382 + }, + { + "epoch": 2.20926517571885, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.1006, + "step": 1383 + }, + { + "epoch": 2.2108626198083066, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1384 + }, + { + "epoch": 2.2124600638977636, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1385 + }, + { + "epoch": 2.2140575079872207, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1386 + }, + { + "epoch": 2.2156549520766773, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1021, + "step": 1387 + }, + { + "epoch": 2.2172523961661343, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1388 + }, + { + "epoch": 2.218849840255591, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1389 + }, + { + "epoch": 2.220447284345048, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.103, + "step": 1390 + }, + { + "epoch": 2.2220447284345046, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1391 + }, + { + "epoch": 2.2236421725239617, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1392 + }, + { + "epoch": 2.2252396166134187, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1393 + }, + { + "epoch": 2.2268370607028753, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1394 + }, + { + "epoch": 2.2284345047923324, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1395 + }, + { + "epoch": 2.230031948881789, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1396 + }, + { + "epoch": 2.231629392971246, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.1036, + "step": 1397 + }, + { + "epoch": 2.2332268370607027, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1398 + }, + { + "epoch": 2.2348242811501597, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 1399 + }, + { + "epoch": 2.236421725239617, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1400 + }, + { + "epoch": 2.2380191693290734, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1401 + }, + { + "epoch": 2.2396166134185305, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1402 + }, + { + "epoch": 2.241214057507987, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1403 + }, + { + "epoch": 2.242811501597444, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1404 + }, + { + "epoch": 2.244408945686901, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1405 + }, + { + "epoch": 2.246006389776358, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1406 + }, + { + "epoch": 2.247603833865815, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1407 + }, + { + "epoch": 2.2492012779552715, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1408 + }, + { + "epoch": 2.2507987220447285, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1409 + }, + { + "epoch": 2.252396166134185, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1410 + }, + { + "epoch": 2.253993610223642, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1411 + }, + { + "epoch": 2.255591054313099, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1412 + }, + { + "epoch": 2.257188498402556, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 1413 + }, + { + "epoch": 2.258785942492013, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1414 + }, + { + "epoch": 2.2603833865814695, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1415 + }, + { + "epoch": 2.2619808306709266, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1416 + }, + { + "epoch": 2.263578274760383, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1417 + }, + { + "epoch": 2.2651757188498403, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1418 + }, + { + "epoch": 2.2667731629392973, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1419 + }, + { + "epoch": 2.268370607028754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1420 + }, + { + "epoch": 2.269968051118211, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 1421 + }, + { + "epoch": 2.2715654952076676, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1422 + }, + { + "epoch": 2.2731629392971247, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1423 + }, + { + "epoch": 2.2747603833865817, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 1424 + }, + { + "epoch": 2.2763578274760383, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1425 + }, + { + "epoch": 2.2779552715654954, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1138, + "step": 1426 + }, + { + "epoch": 2.279552715654952, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1427 + }, + { + "epoch": 2.281150159744409, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1428 + }, + { + "epoch": 2.2827476038338657, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1429 + }, + { + "epoch": 2.2843450479233227, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1430 + }, + { + "epoch": 2.2859424920127793, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1431 + }, + { + "epoch": 2.2875399361022364, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1432 + }, + { + "epoch": 2.2891373801916934, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1433 + }, + { + "epoch": 2.29073482428115, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1434 + }, + { + "epoch": 2.292332268370607, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1435 + }, + { + "epoch": 2.2939297124600637, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1436 + }, + { + "epoch": 2.2955271565495208, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1437 + }, + { + "epoch": 2.297124600638978, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.103, + "step": 1438 + }, + { + "epoch": 2.2987220447284344, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1439 + }, + { + "epoch": 2.3003194888178915, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 1440 + }, + { + "epoch": 2.301916932907348, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1441 + }, + { + "epoch": 2.303514376996805, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1442 + }, + { + "epoch": 2.3051118210862622, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1443 + }, + { + "epoch": 2.306709265175719, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1444 + }, + { + "epoch": 2.308306709265176, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1445 + }, + { + "epoch": 2.3099041533546325, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1446 + }, + { + "epoch": 2.3115015974440896, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1447 + }, + { + "epoch": 2.313099041533546, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1448 + }, + { + "epoch": 2.3146964856230032, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1449 + }, + { + "epoch": 2.31629392971246, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1450 + }, + { + "epoch": 2.317891373801917, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1451 + }, + { + "epoch": 2.319488817891374, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 1452 + }, + { + "epoch": 2.3210862619808306, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1453 + }, + { + "epoch": 2.3226837060702876, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1454 + }, + { + "epoch": 2.3242811501597442, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 1455 + }, + { + "epoch": 2.3258785942492013, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1456 + }, + { + "epoch": 2.3274760383386583, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 1457 + }, + { + "epoch": 2.329073482428115, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1458 + }, + { + "epoch": 2.330670926517572, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1459 + }, + { + "epoch": 2.3322683706070286, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1460 + }, + { + "epoch": 2.3338658146964857, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 1461 + }, + { + "epoch": 2.3354632587859427, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1462 + }, + { + "epoch": 2.3370607028753994, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1463 + }, + { + "epoch": 2.3386581469648564, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 1464 + }, + { + "epoch": 2.340255591054313, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1465 + }, + { + "epoch": 2.34185303514377, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1466 + }, + { + "epoch": 2.3434504792332267, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1467 + }, + { + "epoch": 2.3450479233226837, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.102, + "step": 1468 + }, + { + "epoch": 2.3466453674121404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1469 + }, + { + "epoch": 2.3482428115015974, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1470 + }, + { + "epoch": 2.3498402555910545, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.104, + "step": 1471 + }, + { + "epoch": 2.351437699680511, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1025, + "step": 1472 + }, + { + "epoch": 2.353035143769968, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1473 + }, + { + "epoch": 2.3546325878594248, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1474 + }, + { + "epoch": 2.356230031948882, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1475 + }, + { + "epoch": 2.357827476038339, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1476 + }, + { + "epoch": 2.3594249201277955, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1477 + }, + { + "epoch": 2.3610223642172525, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1478 + }, + { + "epoch": 2.362619808306709, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1479 + }, + { + "epoch": 2.364217252396166, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1480 + }, + { + "epoch": 2.365814696485623, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1481 + }, + { + "epoch": 2.36741214057508, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 1482 + }, + { + "epoch": 2.369009584664537, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1483 + }, + { + "epoch": 2.3706070287539935, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1484 + }, + { + "epoch": 2.3722044728434506, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1485 + }, + { + "epoch": 2.373801916932907, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1486 + }, + { + "epoch": 2.3753993610223643, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1487 + }, + { + "epoch": 2.376996805111821, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1488 + }, + { + "epoch": 2.378594249201278, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1489 + }, + { + "epoch": 2.380191693290735, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1490 + }, + { + "epoch": 2.3817891373801916, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1491 + }, + { + "epoch": 2.3833865814696487, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1492 + }, + { + "epoch": 2.3849840255591053, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 1493 + }, + { + "epoch": 2.3865814696485623, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1494 + }, + { + "epoch": 2.3881789137380194, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0984, + "step": 1495 + }, + { + "epoch": 2.389776357827476, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1496 + }, + { + "epoch": 2.391373801916933, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 1497 + }, + { + "epoch": 2.3929712460063897, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1498 + }, + { + "epoch": 2.3945686900958467, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1499 + }, + { + "epoch": 2.3961661341853033, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1500 + }, + { + "epoch": 2.3977635782747604, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1501 + }, + { + "epoch": 2.3993610223642174, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 1502 + }, + { + "epoch": 2.400958466453674, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1503 + }, + { + "epoch": 2.402555910543131, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 1504 + }, + { + "epoch": 2.4041533546325877, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1505 + }, + { + "epoch": 2.405750798722045, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1506 + }, + { + "epoch": 2.4073482428115014, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1507 + }, + { + "epoch": 2.4089456869009584, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1508 + }, + { + "epoch": 2.4105431309904155, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 1509 + }, + { + "epoch": 2.412140575079872, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1510 + }, + { + "epoch": 2.413738019169329, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1511 + }, + { + "epoch": 2.415335463258786, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 1512 + }, + { + "epoch": 2.416932907348243, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1513 + }, + { + "epoch": 2.4185303514377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1514 + }, + { + "epoch": 2.4201277955271565, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1515 + }, + { + "epoch": 2.4217252396166136, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1516 + }, + { + "epoch": 2.42332268370607, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1517 + }, + { + "epoch": 2.4249201277955272, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 1518 + }, + { + "epoch": 2.426517571884984, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1519 + }, + { + "epoch": 2.428115015974441, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1520 + }, + { + "epoch": 2.4297124600638975, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1521 + }, + { + "epoch": 2.4313099041533546, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1522 + }, + { + "epoch": 2.4329073482428116, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 1523 + }, + { + "epoch": 2.4345047923322682, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1524 + }, + { + "epoch": 2.4361022364217253, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 1525 + }, + { + "epoch": 2.437699680511182, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1526 + }, + { + "epoch": 2.439297124600639, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1527 + }, + { + "epoch": 2.440894568690096, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1528 + }, + { + "epoch": 2.4424920127795526, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.102, + "step": 1529 + }, + { + "epoch": 2.4440894568690097, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1530 + }, + { + "epoch": 2.4456869009584663, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1531 + }, + { + "epoch": 2.4472843450479234, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 1532 + }, + { + "epoch": 2.4488817891373804, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1533 + }, + { + "epoch": 2.450479233226837, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1534 + }, + { + "epoch": 2.452076677316294, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1535 + }, + { + "epoch": 2.4536741214057507, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1536 + }, + { + "epoch": 2.4552715654952078, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1537 + }, + { + "epoch": 2.4568690095846644, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0997, + "step": 1538 + }, + { + "epoch": 2.4584664536741214, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1539 + }, + { + "epoch": 2.460063897763578, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1540 + }, + { + "epoch": 2.461661341853035, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1541 + }, + { + "epoch": 2.463258785942492, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1542 + }, + { + "epoch": 2.4648562300319488, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1543 + }, + { + "epoch": 2.466453674121406, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1544 + }, + { + "epoch": 2.4680511182108624, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1545 + }, + { + "epoch": 2.4696485623003195, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1546 + }, + { + "epoch": 2.4712460063897765, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1547 + }, + { + "epoch": 2.472843450479233, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1548 + }, + { + "epoch": 2.47444089456869, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1549 + }, + { + "epoch": 2.476038338658147, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1550 + }, + { + "epoch": 2.477635782747604, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 1551 + }, + { + "epoch": 2.479233226837061, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1552 + }, + { + "epoch": 2.4808306709265175, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1553 + }, + { + "epoch": 2.4824281150159746, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1554 + }, + { + "epoch": 2.484025559105431, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1555 + }, + { + "epoch": 2.4856230031948883, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1556 + }, + { + "epoch": 2.487220447284345, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 1557 + }, + { + "epoch": 2.488817891373802, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1558 + }, + { + "epoch": 2.4904153354632586, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 1559 + }, + { + "epoch": 2.4920127795527156, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1560 + }, + { + "epoch": 2.4936102236421727, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1561 + }, + { + "epoch": 2.4952076677316293, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 1562 + }, + { + "epoch": 2.4968051118210863, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1563 + }, + { + "epoch": 2.498402555910543, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1564 + }, + { + "epoch": 2.5, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1565 + }, + { + "epoch": 2.501597444089457, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1566 + }, + { + "epoch": 2.5031948881789137, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1567 + }, + { + "epoch": 2.5047923322683707, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1568 + }, + { + "epoch": 2.5063897763578273, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 1569 + }, + { + "epoch": 2.5079872204472844, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1570 + }, + { + "epoch": 2.5095846645367414, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1003, + "step": 1571 + }, + { + "epoch": 2.511182108626198, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1572 + }, + { + "epoch": 2.512779552715655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.096, + "step": 1573 + }, + { + "epoch": 2.5143769968051117, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1574 + }, + { + "epoch": 2.515974440894569, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1575 + }, + { + "epoch": 2.5175718849840254, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1576 + }, + { + "epoch": 2.5191693290734825, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1577 + }, + { + "epoch": 2.520766773162939, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1578 + }, + { + "epoch": 2.522364217252396, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1579 + }, + { + "epoch": 2.523961661341853, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1003, + "step": 1580 + }, + { + "epoch": 2.52555910543131, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1581 + }, + { + "epoch": 2.527156549520767, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 1582 + }, + { + "epoch": 2.5287539936102235, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1583 + }, + { + "epoch": 2.5303514376996805, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1584 + }, + { + "epoch": 2.5319488817891376, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1585 + }, + { + "epoch": 2.533546325878594, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1586 + }, + { + "epoch": 2.5351437699680512, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1587 + }, + { + "epoch": 2.536741214057508, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1588 + }, + { + "epoch": 2.538338658146965, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1589 + }, + { + "epoch": 2.539936102236422, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1590 + }, + { + "epoch": 2.5415335463258786, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1591 + }, + { + "epoch": 2.543130990415335, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1592 + }, + { + "epoch": 2.5447284345047922, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1593 + }, + { + "epoch": 2.5463258785942493, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1594 + }, + { + "epoch": 2.547923322683706, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1595 + }, + { + "epoch": 2.549520766773163, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1596 + }, + { + "epoch": 2.5511182108626196, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1597 + }, + { + "epoch": 2.5527156549520766, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1598 + }, + { + "epoch": 2.5543130990415337, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1599 + }, + { + "epoch": 2.5559105431309903, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 1600 + }, + { + "epoch": 2.5575079872204474, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1601 + }, + { + "epoch": 2.559105431309904, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1602 + }, + { + "epoch": 2.560702875399361, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1603 + }, + { + "epoch": 2.562300319488818, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.1003, + "step": 1604 + }, + { + "epoch": 2.5638977635782747, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1605 + }, + { + "epoch": 2.5654952076677318, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1606 + }, + { + "epoch": 2.5670926517571884, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1607 + }, + { + "epoch": 2.5686900958466454, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1006, + "step": 1608 + }, + { + "epoch": 2.5702875399361025, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.1053, + "step": 1609 + }, + { + "epoch": 2.571884984025559, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1610 + }, + { + "epoch": 2.5734824281150157, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1611 + }, + { + "epoch": 2.5750798722044728, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 1612 + }, + { + "epoch": 2.57667731629393, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1613 + }, + { + "epoch": 2.5782747603833864, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1614 + }, + { + "epoch": 2.5798722044728435, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1615 + }, + { + "epoch": 2.5814696485623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1616 + }, + { + "epoch": 2.583067092651757, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1617 + }, + { + "epoch": 2.584664536741214, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1618 + }, + { + "epoch": 2.586261980830671, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1619 + }, + { + "epoch": 2.587859424920128, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1620 + }, + { + "epoch": 2.5894568690095845, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1036, + "step": 1621 + }, + { + "epoch": 2.5910543130990416, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1622 + }, + { + "epoch": 2.5926517571884986, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1623 + }, + { + "epoch": 2.594249201277955, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1624 + }, + { + "epoch": 2.5958466453674123, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1625 + }, + { + "epoch": 2.597444089456869, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1025, + "step": 1626 + }, + { + "epoch": 2.599041533546326, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1627 + }, + { + "epoch": 2.600638977635783, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0972, + "step": 1628 + }, + { + "epoch": 2.6022364217252396, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1629 + }, + { + "epoch": 2.6038338658146962, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1630 + }, + { + "epoch": 2.6054313099041533, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1631 + }, + { + "epoch": 2.6070287539936103, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 1632 + }, + { + "epoch": 2.608626198083067, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 1633 + }, + { + "epoch": 2.610223642172524, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1634 + }, + { + "epoch": 2.6118210862619806, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1635 + }, + { + "epoch": 2.6134185303514377, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 1636 + }, + { + "epoch": 2.6150159744408947, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1637 + }, + { + "epoch": 2.6166134185303513, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1638 + }, + { + "epoch": 2.6182108626198084, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1639 + }, + { + "epoch": 2.619808306709265, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1640 + }, + { + "epoch": 2.621405750798722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 1641 + }, + { + "epoch": 2.623003194888179, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1642 + }, + { + "epoch": 2.6246006389776357, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1643 + }, + { + "epoch": 2.626198083067093, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1644 + }, + { + "epoch": 2.6277955271565494, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1645 + }, + { + "epoch": 2.6293929712460065, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1646 + }, + { + "epoch": 2.6309904153354635, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1647 + }, + { + "epoch": 2.63258785942492, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1648 + }, + { + "epoch": 2.6341853035143767, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1649 + }, + { + "epoch": 2.635782747603834, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1650 + }, + { + "epoch": 2.637380191693291, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0997, + "step": 1651 + }, + { + "epoch": 2.6389776357827475, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1652 + }, + { + "epoch": 2.6405750798722045, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1653 + }, + { + "epoch": 2.642172523961661, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1654 + }, + { + "epoch": 2.643769968051118, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1655 + }, + { + "epoch": 2.6453674121405752, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1656 + }, + { + "epoch": 2.646964856230032, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1657 + }, + { + "epoch": 2.648562300319489, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0925, + "step": 1658 + }, + { + "epoch": 2.6501597444089455, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1659 + }, + { + "epoch": 2.6517571884984026, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1660 + }, + { + "epoch": 2.6533546325878596, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1661 + }, + { + "epoch": 2.6549520766773163, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1662 + }, + { + "epoch": 2.6565495207667733, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0992, + "step": 1663 + }, + { + "epoch": 2.65814696485623, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1664 + }, + { + "epoch": 2.659744408945687, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1665 + }, + { + "epoch": 2.661341853035144, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1666 + }, + { + "epoch": 2.6629392971246006, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1667 + }, + { + "epoch": 2.6645367412140573, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1668 + }, + { + "epoch": 2.6661341853035143, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1669 + }, + { + "epoch": 2.6677316293929714, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1670 + }, + { + "epoch": 2.669329073482428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1671 + }, + { + "epoch": 2.670926517571885, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1672 + }, + { + "epoch": 2.6725239616613417, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1673 + }, + { + "epoch": 2.6741214057507987, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1674 + }, + { + "epoch": 2.6757188498402558, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1675 + }, + { + "epoch": 2.6773162939297124, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1676 + }, + { + "epoch": 2.6789137380191694, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1677 + }, + { + "epoch": 2.680511182108626, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1678 + }, + { + "epoch": 2.682108626198083, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1679 + }, + { + "epoch": 2.68370607028754, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1680 + }, + { + "epoch": 2.6853035143769968, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1681 + }, + { + "epoch": 2.686900958466454, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 1682 + }, + { + "epoch": 2.6884984025559104, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 1683 + }, + { + "epoch": 2.6900958466453675, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 1684 + }, + { + "epoch": 2.6916932907348246, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1685 + }, + { + "epoch": 2.693290734824281, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1686 + }, + { + "epoch": 2.6948881789137378, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1687 + }, + { + "epoch": 2.696485623003195, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1688 + }, + { + "epoch": 2.698083067092652, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 1689 + }, + { + "epoch": 2.6996805111821085, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1690 + }, + { + "epoch": 2.7012779552715656, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 1691 + }, + { + "epoch": 2.702875399361022, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1692 + }, + { + "epoch": 2.7044728434504792, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1693 + }, + { + "epoch": 2.7060702875399363, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1694 + }, + { + "epoch": 2.707667731629393, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1695 + }, + { + "epoch": 2.70926517571885, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1696 + }, + { + "epoch": 2.7108626198083066, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1697 + }, + { + "epoch": 2.7124600638977636, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1698 + }, + { + "epoch": 2.7140575079872207, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1699 + }, + { + "epoch": 2.7156549520766773, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1700 + }, + { + "epoch": 2.7172523961661343, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1701 + }, + { + "epoch": 2.718849840255591, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1702 + }, + { + "epoch": 2.720447284345048, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 1703 + }, + { + "epoch": 2.722044728434505, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0963, + "step": 1704 + }, + { + "epoch": 2.7236421725239617, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1705 + }, + { + "epoch": 2.7252396166134183, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1706 + }, + { + "epoch": 2.7268370607028753, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0972, + "step": 1707 + }, + { + "epoch": 2.7284345047923324, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1708 + }, + { + "epoch": 2.730031948881789, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1709 + }, + { + "epoch": 2.731629392971246, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1710 + }, + { + "epoch": 2.7332268370607027, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1711 + }, + { + "epoch": 2.7348242811501597, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1712 + }, + { + "epoch": 2.736421725239617, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 1713 + }, + { + "epoch": 2.7380191693290734, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1714 + }, + { + "epoch": 2.7396166134185305, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1715 + }, + { + "epoch": 2.741214057507987, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1716 + }, + { + "epoch": 2.742811501597444, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1717 + }, + { + "epoch": 2.744408945686901, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 1718 + }, + { + "epoch": 2.746006389776358, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 1719 + }, + { + "epoch": 2.747603833865815, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1720 + }, + { + "epoch": 2.7492012779552715, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1721 + }, + { + "epoch": 2.7507987220447285, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0971, + "step": 1722 + }, + { + "epoch": 2.752396166134185, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 1723 + }, + { + "epoch": 2.753993610223642, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1724 + }, + { + "epoch": 2.755591054313099, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1725 + }, + { + "epoch": 2.757188498402556, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1726 + }, + { + "epoch": 2.758785942492013, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1727 + }, + { + "epoch": 2.7603833865814695, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1728 + }, + { + "epoch": 2.7619808306709266, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1729 + }, + { + "epoch": 2.763578274760383, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0963, + "step": 1730 + }, + { + "epoch": 2.7651757188498403, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1731 + }, + { + "epoch": 2.7667731629392973, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 1732 + }, + { + "epoch": 2.768370607028754, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1733 + }, + { + "epoch": 2.769968051118211, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 1734 + }, + { + "epoch": 2.7715654952076676, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1735 + }, + { + "epoch": 2.7731629392971247, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1736 + }, + { + "epoch": 2.7747603833865817, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1737 + }, + { + "epoch": 2.7763578274760383, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 1738 + }, + { + "epoch": 2.777955271565495, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 1739 + }, + { + "epoch": 2.779552715654952, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1740 + }, + { + "epoch": 2.781150159744409, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 1741 + }, + { + "epoch": 2.7827476038338657, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1742 + }, + { + "epoch": 2.7843450479233227, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1743 + }, + { + "epoch": 2.7859424920127793, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 1744 + }, + { + "epoch": 2.7875399361022364, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1745 + }, + { + "epoch": 2.7891373801916934, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1746 + }, + { + "epoch": 2.79073482428115, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1747 + }, + { + "epoch": 2.792332268370607, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1748 + }, + { + "epoch": 2.7939297124600637, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1749 + }, + { + "epoch": 2.7955271565495208, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 1750 + }, + { + "epoch": 2.797124600638978, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1751 + }, + { + "epoch": 2.7987220447284344, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1752 + }, + { + "epoch": 2.8003194888178915, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1753 + }, + { + "epoch": 2.801916932907348, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 1754 + }, + { + "epoch": 2.803514376996805, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 1755 + }, + { + "epoch": 2.8051118210862622, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1756 + }, + { + "epoch": 2.806709265175719, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1757 + }, + { + "epoch": 2.8083067092651754, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1758 + }, + { + "epoch": 2.8099041533546325, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1759 + }, + { + "epoch": 2.8115015974440896, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1760 + }, + { + "epoch": 2.813099041533546, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1761 + }, + { + "epoch": 2.8146964856230032, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1762 + }, + { + "epoch": 2.81629392971246, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.096, + "step": 1763 + }, + { + "epoch": 2.817891373801917, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1764 + }, + { + "epoch": 2.819488817891374, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 1765 + }, + { + "epoch": 2.8210862619808306, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1766 + }, + { + "epoch": 2.8226837060702876, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1767 + }, + { + "epoch": 2.8242811501597442, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1768 + }, + { + "epoch": 2.8258785942492013, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1769 + }, + { + "epoch": 2.8274760383386583, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1770 + }, + { + "epoch": 2.829073482428115, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1771 + }, + { + "epoch": 2.830670926517572, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1772 + }, + { + "epoch": 2.8322683706070286, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1773 + }, + { + "epoch": 2.8338658146964857, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1774 + }, + { + "epoch": 2.8354632587859427, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1775 + }, + { + "epoch": 2.8370607028753994, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 1776 + }, + { + "epoch": 2.838658146964856, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1777 + }, + { + "epoch": 2.840255591054313, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1778 + }, + { + "epoch": 2.84185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1779 + }, + { + "epoch": 2.8434504792332267, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1780 + }, + { + "epoch": 2.8450479233226837, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1781 + }, + { + "epoch": 2.8466453674121404, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1782 + }, + { + "epoch": 2.8482428115015974, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1783 + }, + { + "epoch": 2.8498402555910545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1784 + }, + { + "epoch": 2.851437699680511, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1785 + }, + { + "epoch": 2.853035143769968, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1786 + }, + { + "epoch": 2.8546325878594248, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 1787 + }, + { + "epoch": 2.856230031948882, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1788 + }, + { + "epoch": 2.857827476038339, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1789 + }, + { + "epoch": 2.8594249201277955, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1790 + }, + { + "epoch": 2.8610223642172525, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 1791 + }, + { + "epoch": 2.862619808306709, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 1792 + }, + { + "epoch": 2.864217252396166, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 1793 + }, + { + "epoch": 2.8658146964856233, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1794 + }, + { + "epoch": 2.86741214057508, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1795 + }, + { + "epoch": 2.8690095846645365, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1796 + }, + { + "epoch": 2.8706070287539935, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1797 + }, + { + "epoch": 2.8722044728434506, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1798 + }, + { + "epoch": 2.873801916932907, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1799 + }, + { + "epoch": 2.8753993610223643, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1800 + }, + { + "epoch": 2.876996805111821, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 1801 + }, + { + "epoch": 2.878594249201278, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1802 + }, + { + "epoch": 2.880191693290735, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0997, + "step": 1803 + }, + { + "epoch": 2.8817891373801916, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1804 + }, + { + "epoch": 2.8833865814696487, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 1805 + }, + { + "epoch": 2.8849840255591053, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 1806 + }, + { + "epoch": 2.8865814696485623, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 1807 + }, + { + "epoch": 2.8881789137380194, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1808 + }, + { + "epoch": 2.889776357827476, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1809 + }, + { + "epoch": 2.891373801916933, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1810 + }, + { + "epoch": 2.8929712460063897, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1811 + }, + { + "epoch": 2.8945686900958467, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0966, + "step": 1812 + }, + { + "epoch": 2.8961661341853038, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1813 + }, + { + "epoch": 2.8977635782747604, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1814 + }, + { + "epoch": 2.899361022364217, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 1815 + }, + { + "epoch": 2.900958466453674, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1816 + }, + { + "epoch": 2.902555910543131, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 1817 + }, + { + "epoch": 2.9041533546325877, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1818 + }, + { + "epoch": 2.905750798722045, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 1819 + }, + { + "epoch": 2.9073482428115014, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 1820 + }, + { + "epoch": 2.9089456869009584, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1821 + }, + { + "epoch": 2.9105431309904155, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 1822 + }, + { + "epoch": 2.912140575079872, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1823 + }, + { + "epoch": 2.913738019169329, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 1824 + }, + { + "epoch": 2.915335463258786, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 1825 + }, + { + "epoch": 2.916932907348243, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1826 + }, + { + "epoch": 2.9185303514377, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 1827 + }, + { + "epoch": 2.9201277955271565, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1828 + }, + { + "epoch": 2.9217252396166136, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 1829 + }, + { + "epoch": 2.92332268370607, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 1830 + }, + { + "epoch": 2.9249201277955272, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1831 + }, + { + "epoch": 2.9265175718849843, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1832 + }, + { + "epoch": 2.928115015974441, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1833 + }, + { + "epoch": 2.9297124600638975, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 1834 + }, + { + "epoch": 2.9313099041533546, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1835 + }, + { + "epoch": 2.9329073482428116, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 1836 + }, + { + "epoch": 2.9345047923322682, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1837 + }, + { + "epoch": 2.9361022364217253, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1838 + }, + { + "epoch": 2.937699680511182, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0966, + "step": 1839 + }, + { + "epoch": 2.939297124600639, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1840 + }, + { + "epoch": 2.940894568690096, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 1841 + }, + { + "epoch": 2.9424920127795526, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1842 + }, + { + "epoch": 2.9440894568690097, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1843 + }, + { + "epoch": 2.9456869009584663, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1844 + }, + { + "epoch": 2.9472843450479234, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 1845 + }, + { + "epoch": 2.9488817891373804, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 1846 + }, + { + "epoch": 2.950479233226837, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 1847 + }, + { + "epoch": 2.952076677316294, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 1848 + }, + { + "epoch": 2.9536741214057507, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1849 + }, + { + "epoch": 2.9552715654952078, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 1850 + }, + { + "epoch": 2.956869009584665, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0925, + "step": 1851 + }, + { + "epoch": 2.9584664536741214, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1852 + }, + { + "epoch": 2.960063897763578, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1853 + }, + { + "epoch": 2.961661341853035, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1854 + }, + { + "epoch": 2.963258785942492, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1855 + }, + { + "epoch": 2.9648562300319488, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1856 + }, + { + "epoch": 2.966453674121406, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1857 + }, + { + "epoch": 2.9680511182108624, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1858 + }, + { + "epoch": 2.9696485623003195, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 1859 + }, + { + "epoch": 2.9712460063897765, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1860 + }, + { + "epoch": 2.972843450479233, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1861 + }, + { + "epoch": 2.97444089456869, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1862 + }, + { + "epoch": 2.976038338658147, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1863 + }, + { + "epoch": 2.977635782747604, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 1864 + }, + { + "epoch": 2.979233226837061, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1865 + }, + { + "epoch": 2.9808306709265175, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1866 + }, + { + "epoch": 2.9824281150159746, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1867 + }, + { + "epoch": 2.984025559105431, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1868 + }, + { + "epoch": 2.9856230031948883, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 1869 + }, + { + "epoch": 2.987220447284345, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 1870 + }, + { + "epoch": 2.988817891373802, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 1871 + }, + { + "epoch": 2.9904153354632586, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 1872 + }, + { + "epoch": 2.9920127795527156, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1873 + }, + { + "epoch": 2.9936102236421727, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1874 + }, + { + "epoch": 2.9952076677316293, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1875 + }, + { + "epoch": 2.9968051118210863, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1876 + }, + { + "epoch": 2.998402555910543, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1877 + }, + { + "epoch": 3.0, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1878 + }, + { + "epoch": 3.001597444089457, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 1879 + }, + { + "epoch": 3.0031948881789137, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0925, + "step": 1880 + }, + { + "epoch": 3.0047923322683707, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1881 + }, + { + "epoch": 3.0063897763578273, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1882 + }, + { + "epoch": 3.0079872204472844, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1883 + }, + { + "epoch": 3.009584664536741, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 1884 + }, + { + "epoch": 3.011182108626198, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1885 + }, + { + "epoch": 3.012779552715655, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 1886 + }, + { + "epoch": 3.0143769968051117, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1887 + }, + { + "epoch": 3.015974440894569, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1888 + }, + { + "epoch": 3.0175718849840254, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0919, + "step": 1889 + }, + { + "epoch": 3.0191693290734825, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1890 + }, + { + "epoch": 3.0207667731629395, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 1891 + }, + { + "epoch": 3.022364217252396, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1892 + }, + { + "epoch": 3.023961661341853, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1893 + }, + { + "epoch": 3.02555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 1894 + }, + { + "epoch": 3.027156549520767, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1895 + }, + { + "epoch": 3.0287539936102235, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1896 + }, + { + "epoch": 3.0303514376996805, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1897 + }, + { + "epoch": 3.0319488817891376, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 1898 + }, + { + "epoch": 3.033546325878594, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 1899 + }, + { + "epoch": 3.0351437699680512, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1900 + }, + { + "epoch": 3.036741214057508, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 1901 + }, + { + "epoch": 3.038338658146965, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1902 + }, + { + "epoch": 3.0399361022364215, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1903 + }, + { + "epoch": 3.0415335463258786, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1904 + }, + { + "epoch": 3.0431309904153356, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1905 + }, + { + "epoch": 3.0447284345047922, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1906 + }, + { + "epoch": 3.0463258785942493, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 1907 + }, + { + "epoch": 3.047923322683706, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 1908 + }, + { + "epoch": 3.049520766773163, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 1909 + }, + { + "epoch": 3.0511182108626196, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 1910 + }, + { + "epoch": 3.0527156549520766, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1911 + }, + { + "epoch": 3.0543130990415337, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1912 + }, + { + "epoch": 3.0559105431309903, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1913 + }, + { + "epoch": 3.0575079872204474, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1914 + }, + { + "epoch": 3.059105431309904, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1915 + }, + { + "epoch": 3.060702875399361, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1916 + }, + { + "epoch": 3.062300319488818, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0963, + "step": 1917 + }, + { + "epoch": 3.0638977635782747, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.096, + "step": 1918 + }, + { + "epoch": 3.0654952076677318, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1919 + }, + { + "epoch": 3.0670926517571884, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 1920 + }, + { + "epoch": 3.0686900958466454, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1921 + }, + { + "epoch": 3.070287539936102, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 1922 + }, + { + "epoch": 3.071884984025559, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 1923 + }, + { + "epoch": 3.073482428115016, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1924 + }, + { + "epoch": 3.0750798722044728, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0929, + "step": 1925 + }, + { + "epoch": 3.07667731629393, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1926 + }, + { + "epoch": 3.0782747603833864, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1927 + }, + { + "epoch": 3.0798722044728435, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1928 + }, + { + "epoch": 3.0814696485623, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1929 + }, + { + "epoch": 3.083067092651757, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 1930 + }, + { + "epoch": 3.084664536741214, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1931 + }, + { + "epoch": 3.086261980830671, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 1932 + }, + { + "epoch": 3.087859424920128, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1933 + }, + { + "epoch": 3.0894568690095845, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1934 + }, + { + "epoch": 3.0910543130990416, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1935 + }, + { + "epoch": 3.0926517571884986, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 1936 + }, + { + "epoch": 3.094249201277955, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1937 + }, + { + "epoch": 3.0958466453674123, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 1938 + }, + { + "epoch": 3.097444089456869, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1939 + }, + { + "epoch": 3.099041533546326, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 1940 + }, + { + "epoch": 3.1006389776357826, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 1941 + }, + { + "epoch": 3.1022364217252396, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 1942 + }, + { + "epoch": 3.1038338658146967, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 1943 + }, + { + "epoch": 3.1054313099041533, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1944 + }, + { + "epoch": 3.1070287539936103, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1945 + }, + { + "epoch": 3.108626198083067, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1946 + }, + { + "epoch": 3.110223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1947 + }, + { + "epoch": 3.1118210862619806, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 1948 + }, + { + "epoch": 3.1134185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1949 + }, + { + "epoch": 3.1150159744408947, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 1950 + }, + { + "epoch": 3.1166134185303513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1951 + }, + { + "epoch": 3.1182108626198084, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1952 + }, + { + "epoch": 3.119808306709265, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 1953 + }, + { + "epoch": 3.121405750798722, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1954 + }, + { + "epoch": 3.123003194888179, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1955 + }, + { + "epoch": 3.1246006389776357, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1956 + }, + { + "epoch": 3.126198083067093, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 1957 + }, + { + "epoch": 3.1277955271565494, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1958 + }, + { + "epoch": 3.1293929712460065, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 1959 + }, + { + "epoch": 3.130990415335463, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 1960 + }, + { + "epoch": 3.13258785942492, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 1961 + }, + { + "epoch": 3.134185303514377, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 1962 + }, + { + "epoch": 3.135782747603834, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1963 + }, + { + "epoch": 3.137380191693291, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1964 + }, + { + "epoch": 3.1389776357827475, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1965 + }, + { + "epoch": 3.1405750798722045, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 1966 + }, + { + "epoch": 3.142172523961661, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 1967 + }, + { + "epoch": 3.143769968051118, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1968 + }, + { + "epoch": 3.1453674121405752, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 1969 + }, + { + "epoch": 3.146964856230032, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 1970 + }, + { + "epoch": 3.148562300319489, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1971 + }, + { + "epoch": 3.1501597444089455, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 1972 + }, + { + "epoch": 3.1517571884984026, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 1973 + }, + { + "epoch": 3.1533546325878596, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1974 + }, + { + "epoch": 3.1549520766773163, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 1975 + }, + { + "epoch": 3.1565495207667733, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1976 + }, + { + "epoch": 3.15814696485623, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1977 + }, + { + "epoch": 3.159744408945687, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1978 + }, + { + "epoch": 3.1613418530351436, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 1979 + }, + { + "epoch": 3.1629392971246006, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1980 + }, + { + "epoch": 3.1645367412140577, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1981 + }, + { + "epoch": 3.1661341853035143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 1982 + }, + { + "epoch": 3.1677316293929714, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 1983 + }, + { + "epoch": 3.169329073482428, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1984 + }, + { + "epoch": 3.170926517571885, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1985 + }, + { + "epoch": 3.1725239616613417, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1986 + }, + { + "epoch": 3.1741214057507987, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1987 + }, + { + "epoch": 3.1757188498402558, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 1988 + }, + { + "epoch": 3.1773162939297124, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1989 + }, + { + "epoch": 3.1789137380191694, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1990 + }, + { + "epoch": 3.180511182108626, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 1991 + }, + { + "epoch": 3.182108626198083, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 1992 + }, + { + "epoch": 3.18370607028754, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1993 + }, + { + "epoch": 3.1853035143769968, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1994 + }, + { + "epoch": 3.186900958466454, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0929, + "step": 1995 + }, + { + "epoch": 3.1884984025559104, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 1996 + }, + { + "epoch": 3.1900958466453675, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1997 + }, + { + "epoch": 3.191693290734824, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 1998 + }, + { + "epoch": 3.193290734824281, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 1999 + }, + { + "epoch": 3.194888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2000 + }, + { + "epoch": 3.196485623003195, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 2001 + }, + { + "epoch": 3.198083067092652, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 2002 + }, + { + "epoch": 3.1996805111821085, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 2003 + }, + { + "epoch": 3.2012779552715656, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 2004 + }, + { + "epoch": 3.202875399361022, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2005 + }, + { + "epoch": 3.2044728434504792, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2006 + }, + { + "epoch": 3.2060702875399363, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 2007 + }, + { + "epoch": 3.207667731629393, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2008 + }, + { + "epoch": 3.20926517571885, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 2009 + }, + { + "epoch": 3.2108626198083066, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2010 + }, + { + "epoch": 3.2124600638977636, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 2011 + }, + { + "epoch": 3.2140575079872207, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2012 + }, + { + "epoch": 3.2156549520766773, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2013 + }, + { + "epoch": 3.2172523961661343, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2014 + }, + { + "epoch": 3.218849840255591, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2015 + }, + { + "epoch": 3.220447284345048, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 2016 + }, + { + "epoch": 3.2220447284345046, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 2017 + }, + { + "epoch": 3.2236421725239617, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 2018 + }, + { + "epoch": 3.2252396166134187, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2019 + }, + { + "epoch": 3.2268370607028753, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2020 + }, + { + "epoch": 3.2284345047923324, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2021 + }, + { + "epoch": 3.230031948881789, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2022 + }, + { + "epoch": 3.231629392971246, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2023 + }, + { + "epoch": 3.2332268370607027, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2024 + }, + { + "epoch": 3.2348242811501597, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2025 + }, + { + "epoch": 3.236421725239617, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 2026 + }, + { + "epoch": 3.2380191693290734, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2027 + }, + { + "epoch": 3.2396166134185305, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2028 + }, + { + "epoch": 3.241214057507987, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2029 + }, + { + "epoch": 3.242811501597444, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2030 + }, + { + "epoch": 3.244408945686901, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 2031 + }, + { + "epoch": 3.246006389776358, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2032 + }, + { + "epoch": 3.247603833865815, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2033 + }, + { + "epoch": 3.2492012779552715, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2034 + }, + { + "epoch": 3.2507987220447285, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2035 + }, + { + "epoch": 3.252396166134185, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2036 + }, + { + "epoch": 3.253993610223642, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2037 + }, + { + "epoch": 3.255591054313099, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2038 + }, + { + "epoch": 3.257188498402556, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2039 + }, + { + "epoch": 3.258785942492013, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2040 + }, + { + "epoch": 3.2603833865814695, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2041 + }, + { + "epoch": 3.2619808306709266, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 2042 + }, + { + "epoch": 3.263578274760383, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2043 + }, + { + "epoch": 3.2651757188498403, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2044 + }, + { + "epoch": 3.2667731629392973, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2045 + }, + { + "epoch": 3.268370607028754, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2046 + }, + { + "epoch": 3.269968051118211, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 2047 + }, + { + "epoch": 3.2715654952076676, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2048 + }, + { + "epoch": 3.2731629392971247, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2049 + }, + { + "epoch": 3.2747603833865817, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 2050 + }, + { + "epoch": 3.2763578274760383, + "grad_norm": 0.404296875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 2051 + }, + { + "epoch": 3.2779552715654954, + "grad_norm": 0.8984375, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 2052 + }, + { + "epoch": 3.279552715654952, + "grad_norm": 2.359375, + "learning_rate": 0.0005, + "loss": 1.1479, + "step": 2053 + }, + { + "epoch": 3.281150159744409, + "grad_norm": 0.58984375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 2054 + }, + { + "epoch": 3.2827476038338657, + "grad_norm": 2.0, + "learning_rate": 0.0005, + "loss": 1.1742, + "step": 2055 + }, + { + "epoch": 3.2843450479233227, + "grad_norm": 0.85546875, + "learning_rate": 0.0005, + "loss": 1.1357, + "step": 2056 + }, + { + "epoch": 3.2859424920127793, + "grad_norm": 0.55078125, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 2057 + }, + { + "epoch": 3.2875399361022364, + "grad_norm": 0.5546875, + "learning_rate": 0.0005, + "loss": 1.1345, + "step": 2058 + }, + { + "epoch": 3.2891373801916934, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 2059 + }, + { + "epoch": 3.29073482428115, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 2060 + }, + { + "epoch": 3.292332268370607, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.1142, + "step": 2061 + }, + { + "epoch": 3.2939297124600637, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 2062 + }, + { + "epoch": 3.2955271565495208, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.1149, + "step": 2063 + }, + { + "epoch": 3.297124600638978, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.1106, + "step": 2064 + }, + { + "epoch": 3.2987220447284344, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 2065 + }, + { + "epoch": 3.3003194888178915, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1077, + "step": 2066 + }, + { + "epoch": 3.301916932907348, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 2067 + }, + { + "epoch": 3.303514376996805, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 2068 + }, + { + "epoch": 3.3051118210862622, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 2069 + }, + { + "epoch": 3.306709265175719, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1052, + "step": 2070 + }, + { + "epoch": 3.308306709265176, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1082, + "step": 2071 + }, + { + "epoch": 3.3099041533546325, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 2072 + }, + { + "epoch": 3.3115015974440896, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2073 + }, + { + "epoch": 3.313099041533546, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 2074 + }, + { + "epoch": 3.3146964856230032, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 2075 + }, + { + "epoch": 3.31629392971246, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 2076 + }, + { + "epoch": 3.317891373801917, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 2077 + }, + { + "epoch": 3.319488817891374, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 2078 + }, + { + "epoch": 3.3210862619808306, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 2079 + }, + { + "epoch": 3.3226837060702876, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 2080 + }, + { + "epoch": 3.3242811501597442, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 2081 + }, + { + "epoch": 3.3258785942492013, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 2082 + }, + { + "epoch": 3.3274760383386583, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 2083 + }, + { + "epoch": 3.329073482428115, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2084 + }, + { + "epoch": 3.330670926517572, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2085 + }, + { + "epoch": 3.3322683706070286, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 2086 + }, + { + "epoch": 3.3338658146964857, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 2087 + }, + { + "epoch": 3.3354632587859427, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 2088 + }, + { + "epoch": 3.3370607028753994, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 2089 + }, + { + "epoch": 3.3386581469648564, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2090 + }, + { + "epoch": 3.340255591054313, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2091 + }, + { + "epoch": 3.34185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2092 + }, + { + "epoch": 3.3434504792332267, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2093 + }, + { + "epoch": 3.3450479233226837, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 2094 + }, + { + "epoch": 3.3466453674121404, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2095 + }, + { + "epoch": 3.3482428115015974, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2096 + }, + { + "epoch": 3.3498402555910545, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 2097 + }, + { + "epoch": 3.351437699680511, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2098 + }, + { + "epoch": 3.353035143769968, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 2099 + }, + { + "epoch": 3.3546325878594248, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2100 + }, + { + "epoch": 3.356230031948882, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2101 + }, + { + "epoch": 3.357827476038339, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2102 + }, + { + "epoch": 3.3594249201277955, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 2103 + }, + { + "epoch": 3.3610223642172525, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2104 + }, + { + "epoch": 3.362619808306709, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 2105 + }, + { + "epoch": 3.364217252396166, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 2106 + }, + { + "epoch": 3.365814696485623, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2107 + }, + { + "epoch": 3.36741214057508, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2108 + }, + { + "epoch": 3.369009584664537, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2109 + }, + { + "epoch": 3.3706070287539935, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2110 + }, + { + "epoch": 3.3722044728434506, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2111 + }, + { + "epoch": 3.373801916932907, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2112 + }, + { + "epoch": 3.3753993610223643, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2113 + }, + { + "epoch": 3.376996805111821, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2114 + }, + { + "epoch": 3.378594249201278, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2115 + }, + { + "epoch": 3.380191693290735, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2116 + }, + { + "epoch": 3.3817891373801916, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 2117 + }, + { + "epoch": 3.3833865814696487, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2118 + }, + { + "epoch": 3.3849840255591053, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2119 + }, + { + "epoch": 3.3865814696485623, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2120 + }, + { + "epoch": 3.3881789137380194, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2121 + }, + { + "epoch": 3.389776357827476, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 2122 + }, + { + "epoch": 3.391373801916933, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2123 + }, + { + "epoch": 3.3929712460063897, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 2124 + }, + { + "epoch": 3.3945686900958467, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 2125 + }, + { + "epoch": 3.3961661341853033, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 2126 + }, + { + "epoch": 3.3977635782747604, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2127 + }, + { + "epoch": 3.3993610223642174, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 2128 + }, + { + "epoch": 3.400958466453674, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2129 + }, + { + "epoch": 3.402555910543131, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 2130 + }, + { + "epoch": 3.4041533546325877, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2131 + }, + { + "epoch": 3.405750798722045, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2132 + }, + { + "epoch": 3.4073482428115014, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2133 + }, + { + "epoch": 3.4089456869009584, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0919, + "step": 2134 + }, + { + "epoch": 3.4105431309904155, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2135 + }, + { + "epoch": 3.412140575079872, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2136 + }, + { + "epoch": 3.413738019169329, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2137 + }, + { + "epoch": 3.415335463258786, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2138 + }, + { + "epoch": 3.416932907348243, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2139 + }, + { + "epoch": 3.4185303514377, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 2140 + }, + { + "epoch": 3.4201277955271565, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2141 + }, + { + "epoch": 3.4217252396166136, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2142 + }, + { + "epoch": 3.42332268370607, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2143 + }, + { + "epoch": 3.4249201277955272, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2144 + }, + { + "epoch": 3.426517571884984, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2145 + }, + { + "epoch": 3.428115015974441, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2146 + }, + { + "epoch": 3.4297124600638975, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2147 + }, + { + "epoch": 3.4313099041533546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2148 + }, + { + "epoch": 3.4329073482428116, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2149 + }, + { + "epoch": 3.4345047923322682, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2150 + }, + { + "epoch": 3.4361022364217253, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2151 + }, + { + "epoch": 3.437699680511182, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 2152 + }, + { + "epoch": 3.439297124600639, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2153 + }, + { + "epoch": 3.440894568690096, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2154 + }, + { + "epoch": 3.4424920127795526, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2155 + }, + { + "epoch": 3.4440894568690097, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2156 + }, + { + "epoch": 3.4456869009584663, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2157 + }, + { + "epoch": 3.4472843450479234, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2158 + }, + { + "epoch": 3.4488817891373804, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 2159 + }, + { + "epoch": 3.450479233226837, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 2160 + }, + { + "epoch": 3.452076677316294, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2161 + }, + { + "epoch": 3.4536741214057507, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2162 + }, + { + "epoch": 3.4552715654952078, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2163 + }, + { + "epoch": 3.4568690095846644, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 2164 + }, + { + "epoch": 3.4584664536741214, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 2165 + }, + { + "epoch": 3.460063897763578, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2166 + }, + { + "epoch": 3.461661341853035, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2167 + }, + { + "epoch": 3.463258785942492, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2168 + }, + { + "epoch": 3.4648562300319488, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2169 + }, + { + "epoch": 3.466453674121406, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2170 + }, + { + "epoch": 3.4680511182108624, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 2171 + }, + { + "epoch": 3.4696485623003195, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2172 + }, + { + "epoch": 3.4712460063897765, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2173 + }, + { + "epoch": 3.472843450479233, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2174 + }, + { + "epoch": 3.47444089456869, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2175 + }, + { + "epoch": 3.476038338658147, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0929, + "step": 2176 + }, + { + "epoch": 3.477635782747604, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2177 + }, + { + "epoch": 3.479233226837061, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2178 + }, + { + "epoch": 3.4808306709265175, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2179 + }, + { + "epoch": 3.4824281150159746, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2180 + }, + { + "epoch": 3.484025559105431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 2181 + }, + { + "epoch": 3.4856230031948883, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2182 + }, + { + "epoch": 3.487220447284345, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 2183 + }, + { + "epoch": 3.488817891373802, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2184 + }, + { + "epoch": 3.4904153354632586, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 2185 + }, + { + "epoch": 3.4920127795527156, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2186 + }, + { + "epoch": 3.4936102236421727, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2187 + }, + { + "epoch": 3.4952076677316293, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 2188 + }, + { + "epoch": 3.4968051118210863, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2189 + }, + { + "epoch": 3.498402555910543, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2190 + }, + { + "epoch": 3.5, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2191 + }, + { + "epoch": 3.501597444089457, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 2192 + }, + { + "epoch": 3.5031948881789137, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 2193 + }, + { + "epoch": 3.5047923322683707, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2194 + }, + { + "epoch": 3.5063897763578273, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 2195 + }, + { + "epoch": 3.5079872204472844, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 2196 + }, + { + "epoch": 3.5095846645367414, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2197 + }, + { + "epoch": 3.511182108626198, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2198 + }, + { + "epoch": 3.512779552715655, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2199 + }, + { + "epoch": 3.5143769968051117, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2200 + }, + { + "epoch": 3.515974440894569, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2201 + }, + { + "epoch": 3.5175718849840254, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2202 + }, + { + "epoch": 3.5191693290734825, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 2203 + }, + { + "epoch": 3.520766773162939, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2204 + }, + { + "epoch": 3.522364217252396, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2205 + }, + { + "epoch": 3.523961661341853, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2206 + }, + { + "epoch": 3.52555910543131, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1006, + "step": 2207 + }, + { + "epoch": 3.527156549520767, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2208 + }, + { + "epoch": 3.5287539936102235, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2209 + }, + { + "epoch": 3.5303514376996805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2210 + }, + { + "epoch": 3.5319488817891376, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2211 + }, + { + "epoch": 3.533546325878594, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2212 + }, + { + "epoch": 3.5351437699680512, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2213 + }, + { + "epoch": 3.536741214057508, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2214 + }, + { + "epoch": 3.538338658146965, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2215 + }, + { + "epoch": 3.539936102236422, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2216 + }, + { + "epoch": 3.5415335463258786, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2217 + }, + { + "epoch": 3.543130990415335, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2218 + }, + { + "epoch": 3.5447284345047922, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 2219 + }, + { + "epoch": 3.5463258785942493, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2220 + }, + { + "epoch": 3.547923322683706, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2221 + }, + { + "epoch": 3.549520766773163, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2222 + }, + { + "epoch": 3.5511182108626196, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2223 + }, + { + "epoch": 3.5527156549520766, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2224 + }, + { + "epoch": 3.5543130990415337, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2225 + }, + { + "epoch": 3.5559105431309903, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 2226 + }, + { + "epoch": 3.5575079872204474, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 2227 + }, + { + "epoch": 3.559105431309904, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2228 + }, + { + "epoch": 3.560702875399361, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2229 + }, + { + "epoch": 3.562300319488818, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 2230 + }, + { + "epoch": 3.5638977635782747, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 2231 + }, + { + "epoch": 3.5654952076677318, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2232 + }, + { + "epoch": 3.5670926517571884, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2233 + }, + { + "epoch": 3.5686900958466454, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2234 + }, + { + "epoch": 3.5702875399361025, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2235 + }, + { + "epoch": 3.571884984025559, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2236 + }, + { + "epoch": 3.5734824281150157, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2237 + }, + { + "epoch": 3.5750798722044728, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2238 + }, + { + "epoch": 3.57667731629393, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2239 + }, + { + "epoch": 3.5782747603833864, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2240 + }, + { + "epoch": 3.5798722044728435, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2241 + }, + { + "epoch": 3.5814696485623, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2242 + }, + { + "epoch": 3.583067092651757, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2243 + }, + { + "epoch": 3.584664536741214, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2244 + }, + { + "epoch": 3.586261980830671, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 2245 + }, + { + "epoch": 3.587859424920128, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2246 + }, + { + "epoch": 3.5894568690095845, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2247 + }, + { + "epoch": 3.5910543130990416, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 2248 + }, + { + "epoch": 3.5926517571884986, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2249 + }, + { + "epoch": 3.594249201277955, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2250 + }, + { + "epoch": 3.5958466453674123, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2251 + }, + { + "epoch": 3.597444089456869, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2252 + }, + { + "epoch": 3.599041533546326, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2253 + }, + { + "epoch": 3.600638977635783, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2254 + }, + { + "epoch": 3.6022364217252396, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2255 + }, + { + "epoch": 3.6038338658146962, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2256 + }, + { + "epoch": 3.6054313099041533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2257 + }, + { + "epoch": 3.6070287539936103, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2258 + }, + { + "epoch": 3.608626198083067, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 2259 + }, + { + "epoch": 3.610223642172524, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2260 + }, + { + "epoch": 3.6118210862619806, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2261 + }, + { + "epoch": 3.6134185303514377, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2262 + }, + { + "epoch": 3.6150159744408947, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 2263 + }, + { + "epoch": 3.6166134185303513, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2264 + }, + { + "epoch": 3.6182108626198084, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2265 + }, + { + "epoch": 3.619808306709265, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2266 + }, + { + "epoch": 3.621405750798722, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 2267 + }, + { + "epoch": 3.623003194888179, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2268 + }, + { + "epoch": 3.6246006389776357, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2269 + }, + { + "epoch": 3.626198083067093, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2270 + }, + { + "epoch": 3.6277955271565494, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2271 + }, + { + "epoch": 3.6293929712460065, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2272 + }, + { + "epoch": 3.6309904153354635, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2273 + }, + { + "epoch": 3.63258785942492, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 2274 + }, + { + "epoch": 3.6341853035143767, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2275 + }, + { + "epoch": 3.635782747603834, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2276 + }, + { + "epoch": 3.637380191693291, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2277 + }, + { + "epoch": 3.6389776357827475, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2278 + }, + { + "epoch": 3.6405750798722045, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2279 + }, + { + "epoch": 3.642172523961661, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2280 + }, + { + "epoch": 3.643769968051118, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2281 + }, + { + "epoch": 3.6453674121405752, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2282 + }, + { + "epoch": 3.646964856230032, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2283 + }, + { + "epoch": 3.648562300319489, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2284 + }, + { + "epoch": 3.6501597444089455, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2285 + }, + { + "epoch": 3.6517571884984026, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 2286 + }, + { + "epoch": 3.6533546325878596, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2287 + }, + { + "epoch": 3.6549520766773163, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2288 + }, + { + "epoch": 3.6565495207667733, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2289 + }, + { + "epoch": 3.65814696485623, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2290 + }, + { + "epoch": 3.659744408945687, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2291 + }, + { + "epoch": 3.661341853035144, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2292 + }, + { + "epoch": 3.6629392971246006, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2293 + }, + { + "epoch": 3.6645367412140573, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2294 + }, + { + "epoch": 3.6661341853035143, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2295 + }, + { + "epoch": 3.6677316293929714, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2296 + }, + { + "epoch": 3.669329073482428, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2297 + }, + { + "epoch": 3.670926517571885, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2298 + }, + { + "epoch": 3.6725239616613417, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2299 + }, + { + "epoch": 3.6741214057507987, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2300 + }, + { + "epoch": 3.6757188498402558, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2301 + }, + { + "epoch": 3.6773162939297124, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 2302 + }, + { + "epoch": 3.6789137380191694, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2303 + }, + { + "epoch": 3.680511182108626, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2304 + }, + { + "epoch": 3.682108626198083, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2305 + }, + { + "epoch": 3.68370607028754, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2306 + }, + { + "epoch": 3.6853035143769968, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2307 + }, + { + "epoch": 3.686900958466454, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2308 + }, + { + "epoch": 3.6884984025559104, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2309 + }, + { + "epoch": 3.6900958466453675, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 2310 + }, + { + "epoch": 3.6916932907348246, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2311 + }, + { + "epoch": 3.693290734824281, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2312 + }, + { + "epoch": 3.6948881789137378, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2313 + }, + { + "epoch": 3.696485623003195, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2314 + }, + { + "epoch": 3.698083067092652, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2315 + }, + { + "epoch": 3.6996805111821085, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2316 + }, + { + "epoch": 3.7012779552715656, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2317 + }, + { + "epoch": 3.702875399361022, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2318 + }, + { + "epoch": 3.7044728434504792, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2319 + }, + { + "epoch": 3.7060702875399363, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2320 + }, + { + "epoch": 3.707667731629393, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2321 + }, + { + "epoch": 3.70926517571885, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2322 + }, + { + "epoch": 3.7108626198083066, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2323 + }, + { + "epoch": 3.7124600638977636, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 2324 + }, + { + "epoch": 3.7140575079872207, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2325 + }, + { + "epoch": 3.7156549520766773, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 2326 + }, + { + "epoch": 3.7172523961661343, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2327 + }, + { + "epoch": 3.718849840255591, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2328 + }, + { + "epoch": 3.720447284345048, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2329 + }, + { + "epoch": 3.722044728434505, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 2330 + }, + { + "epoch": 3.7236421725239617, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2331 + }, + { + "epoch": 3.7252396166134183, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2332 + }, + { + "epoch": 3.7268370607028753, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2333 + }, + { + "epoch": 3.7284345047923324, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 2334 + }, + { + "epoch": 3.730031948881789, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2335 + }, + { + "epoch": 3.731629392971246, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2336 + }, + { + "epoch": 3.7332268370607027, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2337 + }, + { + "epoch": 3.7348242811501597, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2338 + }, + { + "epoch": 3.736421725239617, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2339 + }, + { + "epoch": 3.7380191693290734, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2340 + }, + { + "epoch": 3.7396166134185305, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2341 + }, + { + "epoch": 3.741214057507987, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2342 + }, + { + "epoch": 3.742811501597444, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 2343 + }, + { + "epoch": 3.744408945686901, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2344 + }, + { + "epoch": 3.746006389776358, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2345 + }, + { + "epoch": 3.747603833865815, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2346 + }, + { + "epoch": 3.7492012779552715, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2347 + }, + { + "epoch": 3.7507987220447285, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 2348 + }, + { + "epoch": 3.752396166134185, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2349 + }, + { + "epoch": 3.753993610223642, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2350 + }, + { + "epoch": 3.755591054313099, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2351 + }, + { + "epoch": 3.757188498402556, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2352 + }, + { + "epoch": 3.758785942492013, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2353 + }, + { + "epoch": 3.7603833865814695, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2354 + }, + { + "epoch": 3.7619808306709266, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2355 + }, + { + "epoch": 3.763578274760383, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2356 + }, + { + "epoch": 3.7651757188498403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 2357 + }, + { + "epoch": 3.7667731629392973, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 2358 + }, + { + "epoch": 3.768370607028754, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 2359 + }, + { + "epoch": 3.769968051118211, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2360 + }, + { + "epoch": 3.7715654952076676, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2361 + }, + { + "epoch": 3.7731629392971247, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2362 + }, + { + "epoch": 3.7747603833865817, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2363 + }, + { + "epoch": 3.7763578274760383, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2364 + }, + { + "epoch": 3.777955271565495, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2365 + }, + { + "epoch": 3.779552715654952, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2366 + }, + { + "epoch": 3.781150159744409, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2367 + }, + { + "epoch": 3.7827476038338657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2368 + }, + { + "epoch": 3.7843450479233227, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2369 + }, + { + "epoch": 3.7859424920127793, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 2370 + }, + { + "epoch": 3.7875399361022364, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2371 + }, + { + "epoch": 3.7891373801916934, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2372 + }, + { + "epoch": 3.79073482428115, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2373 + }, + { + "epoch": 3.792332268370607, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2374 + }, + { + "epoch": 3.7939297124600637, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2375 + }, + { + "epoch": 3.7955271565495208, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 2376 + }, + { + "epoch": 3.797124600638978, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2377 + }, + { + "epoch": 3.7987220447284344, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2378 + }, + { + "epoch": 3.8003194888178915, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2379 + }, + { + "epoch": 3.801916932907348, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2380 + }, + { + "epoch": 3.803514376996805, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2381 + }, + { + "epoch": 3.8051118210862622, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2382 + }, + { + "epoch": 3.806709265175719, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2383 + }, + { + "epoch": 3.8083067092651754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2384 + }, + { + "epoch": 3.8099041533546325, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2385 + }, + { + "epoch": 3.8115015974440896, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2386 + }, + { + "epoch": 3.813099041533546, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2387 + }, + { + "epoch": 3.8146964856230032, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2388 + }, + { + "epoch": 3.81629392971246, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2389 + }, + { + "epoch": 3.817891373801917, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2390 + }, + { + "epoch": 3.819488817891374, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2391 + }, + { + "epoch": 3.8210862619808306, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2392 + }, + { + "epoch": 3.8226837060702876, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2393 + }, + { + "epoch": 3.8242811501597442, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2394 + }, + { + "epoch": 3.8258785942492013, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2395 + }, + { + "epoch": 3.8274760383386583, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2396 + }, + { + "epoch": 3.829073482428115, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2397 + }, + { + "epoch": 3.830670926517572, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2398 + }, + { + "epoch": 3.8322683706070286, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2399 + }, + { + "epoch": 3.8338658146964857, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2400 + }, + { + "epoch": 3.8354632587859427, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2401 + }, + { + "epoch": 3.8370607028753994, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2402 + }, + { + "epoch": 3.838658146964856, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2403 + }, + { + "epoch": 3.840255591054313, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2404 + }, + { + "epoch": 3.84185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2405 + }, + { + "epoch": 3.8434504792332267, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2406 + }, + { + "epoch": 3.8450479233226837, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2407 + }, + { + "epoch": 3.8466453674121404, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2408 + }, + { + "epoch": 3.8482428115015974, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 2409 + }, + { + "epoch": 3.8498402555910545, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2410 + }, + { + "epoch": 3.851437699680511, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2411 + }, + { + "epoch": 3.853035143769968, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2412 + }, + { + "epoch": 3.8546325878594248, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2413 + }, + { + "epoch": 3.856230031948882, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2414 + }, + { + "epoch": 3.857827476038339, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2415 + }, + { + "epoch": 3.8594249201277955, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2416 + }, + { + "epoch": 3.8610223642172525, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2417 + }, + { + "epoch": 3.862619808306709, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2418 + }, + { + "epoch": 3.864217252396166, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2419 + }, + { + "epoch": 3.8658146964856233, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2420 + }, + { + "epoch": 3.86741214057508, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2421 + }, + { + "epoch": 3.8690095846645365, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2422 + }, + { + "epoch": 3.8706070287539935, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2423 + }, + { + "epoch": 3.8722044728434506, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2424 + }, + { + "epoch": 3.873801916932907, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2425 + }, + { + "epoch": 3.8753993610223643, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2426 + }, + { + "epoch": 3.876996805111821, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2427 + }, + { + "epoch": 3.878594249201278, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2428 + }, + { + "epoch": 3.880191693290735, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2429 + }, + { + "epoch": 3.8817891373801916, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2430 + }, + { + "epoch": 3.8833865814696487, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2431 + }, + { + "epoch": 3.8849840255591053, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2432 + }, + { + "epoch": 3.8865814696485623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2433 + }, + { + "epoch": 3.8881789137380194, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2434 + }, + { + "epoch": 3.889776357827476, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2435 + }, + { + "epoch": 3.891373801916933, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2436 + }, + { + "epoch": 3.8929712460063897, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2437 + }, + { + "epoch": 3.8945686900958467, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2438 + }, + { + "epoch": 3.8961661341853038, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2439 + }, + { + "epoch": 3.8977635782747604, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2440 + }, + { + "epoch": 3.899361022364217, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2441 + }, + { + "epoch": 3.900958466453674, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2442 + }, + { + "epoch": 3.902555910543131, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2443 + }, + { + "epoch": 3.9041533546325877, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2444 + }, + { + "epoch": 3.905750798722045, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2445 + }, + { + "epoch": 3.9073482428115014, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2446 + }, + { + "epoch": 3.9089456869009584, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2447 + }, + { + "epoch": 3.9105431309904155, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 2448 + }, + { + "epoch": 3.912140575079872, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2449 + }, + { + "epoch": 3.913738019169329, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 2450 + }, + { + "epoch": 3.915335463258786, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2451 + }, + { + "epoch": 3.916932907348243, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2452 + }, + { + "epoch": 3.9185303514377, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2453 + }, + { + "epoch": 3.9201277955271565, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2454 + }, + { + "epoch": 3.9217252396166136, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2455 + }, + { + "epoch": 3.92332268370607, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2456 + }, + { + "epoch": 3.9249201277955272, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2457 + }, + { + "epoch": 3.9265175718849843, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2458 + }, + { + "epoch": 3.928115015974441, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2459 + }, + { + "epoch": 3.9297124600638975, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 2460 + }, + { + "epoch": 3.9313099041533546, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2461 + }, + { + "epoch": 3.9329073482428116, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2462 + }, + { + "epoch": 3.9345047923322682, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2463 + }, + { + "epoch": 3.9361022364217253, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2464 + }, + { + "epoch": 3.937699680511182, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 2465 + }, + { + "epoch": 3.939297124600639, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2466 + }, + { + "epoch": 3.940894568690096, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2467 + }, + { + "epoch": 3.9424920127795526, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2468 + }, + { + "epoch": 3.9440894568690097, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2469 + }, + { + "epoch": 3.9456869009584663, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2470 + }, + { + "epoch": 3.9472843450479234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2471 + }, + { + "epoch": 3.9488817891373804, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2472 + }, + { + "epoch": 3.950479233226837, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2473 + }, + { + "epoch": 3.952076677316294, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2474 + }, + { + "epoch": 3.9536741214057507, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 2475 + }, + { + "epoch": 3.9552715654952078, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2476 + }, + { + "epoch": 3.956869009584665, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2477 + }, + { + "epoch": 3.9584664536741214, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 2478 + }, + { + "epoch": 3.960063897763578, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2479 + }, + { + "epoch": 3.961661341853035, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2480 + }, + { + "epoch": 3.963258785942492, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2481 + }, + { + "epoch": 3.9648562300319488, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2482 + }, + { + "epoch": 3.966453674121406, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 2483 + }, + { + "epoch": 3.9680511182108624, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2484 + }, + { + "epoch": 3.9696485623003195, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2485 + }, + { + "epoch": 3.9712460063897765, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2486 + }, + { + "epoch": 3.972843450479233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2487 + }, + { + "epoch": 3.97444089456869, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2488 + }, + { + "epoch": 3.976038338658147, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2489 + }, + { + "epoch": 3.977635782747604, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2490 + }, + { + "epoch": 3.979233226837061, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 2491 + }, + { + "epoch": 3.9808306709265175, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 2492 + }, + { + "epoch": 3.9824281150159746, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2493 + }, + { + "epoch": 3.984025559105431, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2494 + }, + { + "epoch": 3.9856230031948883, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2495 + }, + { + "epoch": 3.987220447284345, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2496 + }, + { + "epoch": 3.988817891373802, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2497 + }, + { + "epoch": 3.9904153354632586, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2498 + }, + { + "epoch": 3.9920127795527156, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2499 + }, + { + "epoch": 3.9936102236421727, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2500 + }, + { + "epoch": 3.9952076677316293, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2501 + }, + { + "epoch": 3.9968051118210863, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2502 + }, + { + "epoch": 3.998402555910543, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2503 + }, + { + "epoch": 4.0, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2504 + }, + { + "epoch": 4.001597444089457, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2505 + }, + { + "epoch": 4.003194888178914, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2506 + }, + { + "epoch": 4.00479233226837, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2507 + }, + { + "epoch": 4.006389776357827, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2508 + }, + { + "epoch": 4.007987220447284, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2509 + }, + { + "epoch": 4.0095846645367414, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2510 + }, + { + "epoch": 4.0111821086261985, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2511 + }, + { + "epoch": 4.012779552715655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2512 + }, + { + "epoch": 4.014376996805112, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2513 + }, + { + "epoch": 4.015974440894569, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2514 + }, + { + "epoch": 4.017571884984026, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2515 + }, + { + "epoch": 4.019169329073482, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2516 + }, + { + "epoch": 4.020766773162939, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2517 + }, + { + "epoch": 4.022364217252396, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2518 + }, + { + "epoch": 4.023961661341853, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 2519 + }, + { + "epoch": 4.02555910543131, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2520 + }, + { + "epoch": 4.027156549520766, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2521 + }, + { + "epoch": 4.0287539936102235, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2522 + }, + { + "epoch": 4.0303514376996805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2523 + }, + { + "epoch": 4.031948881789138, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 2524 + }, + { + "epoch": 4.033546325878595, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2525 + }, + { + "epoch": 4.035143769968051, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2526 + }, + { + "epoch": 4.036741214057508, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2527 + }, + { + "epoch": 4.038338658146965, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2528 + }, + { + "epoch": 4.039936102236422, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2529 + }, + { + "epoch": 4.041533546325879, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2530 + }, + { + "epoch": 4.043130990415335, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 2531 + }, + { + "epoch": 4.044728434504792, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 2532 + }, + { + "epoch": 4.046325878594249, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2533 + }, + { + "epoch": 4.047923322683706, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2534 + }, + { + "epoch": 4.0495207667731625, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2535 + }, + { + "epoch": 4.05111821086262, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2536 + }, + { + "epoch": 4.052715654952077, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2537 + }, + { + "epoch": 4.054313099041534, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2538 + }, + { + "epoch": 4.055910543130991, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2539 + }, + { + "epoch": 4.057507987220447, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2540 + }, + { + "epoch": 4.059105431309904, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2541 + }, + { + "epoch": 4.060702875399361, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2542 + }, + { + "epoch": 4.062300319488818, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 2543 + }, + { + "epoch": 4.063897763578275, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2544 + }, + { + "epoch": 4.065495207667731, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 2545 + }, + { + "epoch": 4.067092651757188, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2546 + }, + { + "epoch": 4.068690095846645, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2547 + }, + { + "epoch": 4.0702875399361025, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2548 + }, + { + "epoch": 4.0718849840255595, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2549 + }, + { + "epoch": 4.073482428115016, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2550 + }, + { + "epoch": 4.075079872204473, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2551 + }, + { + "epoch": 4.07667731629393, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2552 + }, + { + "epoch": 4.078274760383387, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2553 + }, + { + "epoch": 4.079872204472843, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 2554 + }, + { + "epoch": 4.0814696485623, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2555 + }, + { + "epoch": 4.083067092651757, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2556 + }, + { + "epoch": 4.084664536741214, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2557 + }, + { + "epoch": 4.086261980830671, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2558 + }, + { + "epoch": 4.087859424920127, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2559 + }, + { + "epoch": 4.0894568690095845, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2560 + }, + { + "epoch": 4.0910543130990416, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2561 + }, + { + "epoch": 4.092651757188499, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2562 + }, + { + "epoch": 4.094249201277956, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2563 + }, + { + "epoch": 4.095846645367412, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2564 + }, + { + "epoch": 4.097444089456869, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 2565 + }, + { + "epoch": 4.099041533546326, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2566 + }, + { + "epoch": 4.100638977635783, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2567 + }, + { + "epoch": 4.102236421725239, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 2568 + }, + { + "epoch": 4.103833865814696, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2569 + }, + { + "epoch": 4.105431309904153, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 2570 + }, + { + "epoch": 4.10702875399361, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2571 + }, + { + "epoch": 4.108626198083067, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2572 + }, + { + "epoch": 4.110223642172524, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2573 + }, + { + "epoch": 4.111821086261981, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2574 + }, + { + "epoch": 4.113418530351438, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2575 + }, + { + "epoch": 4.115015974440895, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2576 + }, + { + "epoch": 4.116613418530352, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 2577 + }, + { + "epoch": 4.118210862619808, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2578 + }, + { + "epoch": 4.119808306709265, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2579 + }, + { + "epoch": 4.121405750798722, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2580 + }, + { + "epoch": 4.123003194888179, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2581 + }, + { + "epoch": 4.124600638977636, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2582 + }, + { + "epoch": 4.126198083067092, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2583 + }, + { + "epoch": 4.127795527156549, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 2584 + }, + { + "epoch": 4.1293929712460065, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2585 + }, + { + "epoch": 4.1309904153354635, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2586 + }, + { + "epoch": 4.13258785942492, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2587 + }, + { + "epoch": 4.134185303514377, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2588 + }, + { + "epoch": 4.135782747603834, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2589 + }, + { + "epoch": 4.137380191693291, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2590 + }, + { + "epoch": 4.138977635782748, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 2591 + }, + { + "epoch": 4.140575079872204, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2592 + }, + { + "epoch": 4.142172523961661, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2593 + }, + { + "epoch": 4.143769968051118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2594 + }, + { + "epoch": 4.145367412140575, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2595 + }, + { + "epoch": 4.146964856230032, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2596 + }, + { + "epoch": 4.1485623003194885, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2597 + }, + { + "epoch": 4.1501597444089455, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 2598 + }, + { + "epoch": 4.151757188498403, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2599 + }, + { + "epoch": 4.15335463258786, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 2600 + }, + { + "epoch": 4.154952076677317, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2601 + }, + { + "epoch": 4.156549520766773, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2602 + }, + { + "epoch": 4.15814696485623, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2603 + }, + { + "epoch": 4.159744408945687, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2604 + }, + { + "epoch": 4.161341853035144, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2605 + }, + { + "epoch": 4.1629392971246, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2606 + }, + { + "epoch": 4.164536741214057, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2607 + }, + { + "epoch": 4.166134185303514, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2608 + }, + { + "epoch": 4.167731629392971, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2609 + }, + { + "epoch": 4.169329073482428, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2610 + }, + { + "epoch": 4.170926517571885, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2611 + }, + { + "epoch": 4.172523961661342, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2612 + }, + { + "epoch": 4.174121405750799, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2613 + }, + { + "epoch": 4.175718849840256, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 2614 + }, + { + "epoch": 4.177316293929713, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2615 + }, + { + "epoch": 4.178913738019169, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2616 + }, + { + "epoch": 4.180511182108626, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2617 + }, + { + "epoch": 4.182108626198083, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2618 + }, + { + "epoch": 4.18370607028754, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2619 + }, + { + "epoch": 4.185303514376997, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2620 + }, + { + "epoch": 4.186900958466453, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2621 + }, + { + "epoch": 4.18849840255591, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2622 + }, + { + "epoch": 4.1900958466453675, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2623 + }, + { + "epoch": 4.1916932907348246, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2624 + }, + { + "epoch": 4.193290734824281, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2625 + }, + { + "epoch": 4.194888178913738, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 2626 + }, + { + "epoch": 4.196485623003195, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2627 + }, + { + "epoch": 4.198083067092652, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2628 + }, + { + "epoch": 4.199680511182109, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2629 + }, + { + "epoch": 4.201277955271565, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2630 + }, + { + "epoch": 4.202875399361022, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2631 + }, + { + "epoch": 4.204472843450479, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2632 + }, + { + "epoch": 4.206070287539936, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2633 + }, + { + "epoch": 4.207667731629393, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2634 + }, + { + "epoch": 4.2092651757188495, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2635 + }, + { + "epoch": 4.210862619808307, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2636 + }, + { + "epoch": 4.212460063897764, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2637 + }, + { + "epoch": 4.214057507987221, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2638 + }, + { + "epoch": 4.215654952076678, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2639 + }, + { + "epoch": 4.217252396166134, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2640 + }, + { + "epoch": 4.218849840255591, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2641 + }, + { + "epoch": 4.220447284345048, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2642 + }, + { + "epoch": 4.222044728434505, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2643 + }, + { + "epoch": 4.223642172523961, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2644 + }, + { + "epoch": 4.225239616613418, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2645 + }, + { + "epoch": 4.226837060702875, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 2646 + }, + { + "epoch": 4.228434504792332, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2647 + }, + { + "epoch": 4.2300319488817895, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2648 + }, + { + "epoch": 4.231629392971246, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2649 + }, + { + "epoch": 4.233226837060703, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2650 + }, + { + "epoch": 4.23482428115016, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2651 + }, + { + "epoch": 4.236421725239617, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2652 + }, + { + "epoch": 4.238019169329074, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2653 + }, + { + "epoch": 4.23961661341853, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 2654 + }, + { + "epoch": 4.241214057507987, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2655 + }, + { + "epoch": 4.242811501597444, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2656 + }, + { + "epoch": 4.244408945686901, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2657 + }, + { + "epoch": 4.246006389776358, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2658 + }, + { + "epoch": 4.247603833865814, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2659 + }, + { + "epoch": 4.2492012779552715, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 2660 + }, + { + "epoch": 4.2507987220447285, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2661 + }, + { + "epoch": 4.252396166134186, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2662 + }, + { + "epoch": 4.253993610223642, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2663 + }, + { + "epoch": 4.255591054313099, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 2664 + }, + { + "epoch": 4.257188498402556, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2665 + }, + { + "epoch": 4.258785942492013, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2666 + }, + { + "epoch": 4.26038338658147, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2667 + }, + { + "epoch": 4.261980830670926, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2668 + }, + { + "epoch": 4.263578274760383, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2669 + }, + { + "epoch": 4.26517571884984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2670 + }, + { + "epoch": 4.266773162939297, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2671 + }, + { + "epoch": 4.268370607028754, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 2672 + }, + { + "epoch": 4.2699680511182105, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 2673 + }, + { + "epoch": 4.271565495207668, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2674 + }, + { + "epoch": 4.273162939297125, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 2675 + }, + { + "epoch": 4.274760383386582, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2676 + }, + { + "epoch": 4.276357827476039, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2677 + }, + { + "epoch": 4.277955271565495, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2678 + }, + { + "epoch": 4.279552715654952, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2679 + }, + { + "epoch": 4.281150159744409, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2680 + }, + { + "epoch": 4.282747603833866, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2681 + }, + { + "epoch": 4.284345047923322, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2682 + }, + { + "epoch": 4.285942492012779, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2683 + }, + { + "epoch": 4.287539936102236, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2684 + }, + { + "epoch": 4.289137380191693, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2685 + }, + { + "epoch": 4.2907348242811505, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2686 + }, + { + "epoch": 4.292332268370607, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2687 + }, + { + "epoch": 4.293929712460064, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2688 + }, + { + "epoch": 4.295527156549521, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2689 + }, + { + "epoch": 4.297124600638978, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2690 + }, + { + "epoch": 4.298722044728435, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2691 + }, + { + "epoch": 4.300319488817891, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2692 + }, + { + "epoch": 4.301916932907348, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2693 + }, + { + "epoch": 4.303514376996805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2694 + }, + { + "epoch": 4.305111821086262, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2695 + }, + { + "epoch": 4.306709265175719, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 2696 + }, + { + "epoch": 4.3083067092651754, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2697 + }, + { + "epoch": 4.3099041533546325, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2698 + }, + { + "epoch": 4.31150159744409, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2699 + }, + { + "epoch": 4.313099041533547, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2700 + }, + { + "epoch": 4.314696485623003, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2701 + }, + { + "epoch": 4.31629392971246, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 2702 + }, + { + "epoch": 4.317891373801917, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 2703 + }, + { + "epoch": 4.319488817891374, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2704 + }, + { + "epoch": 4.321086261980831, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2705 + }, + { + "epoch": 4.322683706070287, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2706 + }, + { + "epoch": 4.324281150159744, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2707 + }, + { + "epoch": 4.325878594249201, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2708 + }, + { + "epoch": 4.327476038338658, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 2709 + }, + { + "epoch": 4.329073482428115, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 2710 + }, + { + "epoch": 4.330670926517572, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2711 + }, + { + "epoch": 4.332268370607029, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2712 + }, + { + "epoch": 4.333865814696486, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2713 + }, + { + "epoch": 4.335463258785943, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2714 + }, + { + "epoch": 4.3370607028754, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2715 + }, + { + "epoch": 4.338658146964856, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2716 + }, + { + "epoch": 4.340255591054313, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2717 + }, + { + "epoch": 4.34185303514377, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 2718 + }, + { + "epoch": 4.343450479233227, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2719 + }, + { + "epoch": 4.345047923322683, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2720 + }, + { + "epoch": 4.34664536741214, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2721 + }, + { + "epoch": 4.348242811501597, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2722 + }, + { + "epoch": 4.3498402555910545, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2723 + }, + { + "epoch": 4.3514376996805115, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2724 + }, + { + "epoch": 4.353035143769968, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2725 + }, + { + "epoch": 4.354632587859425, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2726 + }, + { + "epoch": 4.356230031948882, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2727 + }, + { + "epoch": 4.357827476038339, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 2728 + }, + { + "epoch": 4.359424920127796, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2729 + }, + { + "epoch": 4.361022364217252, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2730 + }, + { + "epoch": 4.362619808306709, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2731 + }, + { + "epoch": 4.364217252396166, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2732 + }, + { + "epoch": 4.365814696485623, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2733 + }, + { + "epoch": 4.36741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 2734 + }, + { + "epoch": 4.3690095846645365, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2735 + }, + { + "epoch": 4.3706070287539935, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 2736 + }, + { + "epoch": 4.372204472843451, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2737 + }, + { + "epoch": 4.373801916932908, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2738 + }, + { + "epoch": 4.375399361022364, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2739 + }, + { + "epoch": 4.376996805111821, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 2740 + }, + { + "epoch": 4.378594249201278, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2741 + }, + { + "epoch": 4.380191693290735, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2742 + }, + { + "epoch": 4.381789137380192, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2743 + }, + { + "epoch": 4.383386581469648, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2744 + }, + { + "epoch": 4.384984025559105, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2745 + }, + { + "epoch": 4.386581469648562, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2746 + }, + { + "epoch": 4.388178913738019, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2747 + }, + { + "epoch": 4.389776357827476, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 2748 + }, + { + "epoch": 4.391373801916933, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2749 + }, + { + "epoch": 4.39297124600639, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2750 + }, + { + "epoch": 4.394568690095847, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2751 + }, + { + "epoch": 4.396166134185304, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2752 + }, + { + "epoch": 4.397763578274761, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2753 + }, + { + "epoch": 4.399361022364217, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2754 + }, + { + "epoch": 4.400958466453674, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2755 + }, + { + "epoch": 4.402555910543131, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2756 + }, + { + "epoch": 4.404153354632588, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2757 + }, + { + "epoch": 4.405750798722044, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2758 + }, + { + "epoch": 4.407348242811501, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2759 + }, + { + "epoch": 4.4089456869009584, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2760 + }, + { + "epoch": 4.4105431309904155, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2761 + }, + { + "epoch": 4.412140575079873, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2762 + }, + { + "epoch": 4.413738019169329, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2763 + }, + { + "epoch": 4.415335463258786, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2764 + }, + { + "epoch": 4.416932907348243, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2765 + }, + { + "epoch": 4.4185303514377, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2766 + }, + { + "epoch": 4.420127795527157, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2767 + }, + { + "epoch": 4.421725239616613, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 2768 + }, + { + "epoch": 4.42332268370607, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2769 + }, + { + "epoch": 4.424920127795527, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2770 + }, + { + "epoch": 4.426517571884984, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 2771 + }, + { + "epoch": 4.428115015974441, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2772 + }, + { + "epoch": 4.4297124600638975, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2773 + }, + { + "epoch": 4.431309904153355, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2774 + }, + { + "epoch": 4.432907348242812, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2775 + }, + { + "epoch": 4.434504792332269, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 2776 + }, + { + "epoch": 4.436102236421725, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 2777 + }, + { + "epoch": 4.437699680511182, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2778 + }, + { + "epoch": 4.439297124600639, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2779 + }, + { + "epoch": 4.440894568690096, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2780 + }, + { + "epoch": 4.442492012779553, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2781 + }, + { + "epoch": 4.444089456869009, + "grad_norm": 0.458984375, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2782 + }, + { + "epoch": 4.445686900958466, + "grad_norm": 0.95703125, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2783 + }, + { + "epoch": 4.447284345047923, + "grad_norm": 1.859375, + "learning_rate": 0.0005, + "loss": 1.135, + "step": 2784 + }, + { + "epoch": 4.44888178913738, + "grad_norm": 0.37109375, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 2785 + }, + { + "epoch": 4.4504792332268375, + "grad_norm": 1.6015625, + "learning_rate": 0.0005, + "loss": 1.1374, + "step": 2786 + }, + { + "epoch": 4.452076677316294, + "grad_norm": 0.298828125, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 2787 + }, + { + "epoch": 4.453674121405751, + "grad_norm": 0.828125, + "learning_rate": 0.0005, + "loss": 1.1162, + "step": 2788 + }, + { + "epoch": 4.455271565495208, + "grad_norm": 0.4140625, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 2789 + }, + { + "epoch": 4.456869009584665, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 2790 + }, + { + "epoch": 4.458466453674122, + "grad_norm": 0.35546875, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 2791 + }, + { + "epoch": 4.460063897763578, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 2792 + }, + { + "epoch": 4.461661341853035, + "grad_norm": 0.384765625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2793 + }, + { + "epoch": 4.463258785942492, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2794 + }, + { + "epoch": 4.464856230031949, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2795 + }, + { + "epoch": 4.466453674121405, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 2796 + }, + { + "epoch": 4.468051118210862, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2797 + }, + { + "epoch": 4.4696485623003195, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2798 + }, + { + "epoch": 4.4712460063897765, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2799 + }, + { + "epoch": 4.472843450479234, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2800 + }, + { + "epoch": 4.47444089456869, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2801 + }, + { + "epoch": 4.476038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2802 + }, + { + "epoch": 4.477635782747604, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2803 + }, + { + "epoch": 4.479233226837061, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2804 + }, + { + "epoch": 4.480830670926517, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2805 + }, + { + "epoch": 4.482428115015974, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2806 + }, + { + "epoch": 4.484025559105431, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2807 + }, + { + "epoch": 4.485623003194888, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2808 + }, + { + "epoch": 4.487220447284345, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2809 + }, + { + "epoch": 4.488817891373802, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2810 + }, + { + "epoch": 4.4904153354632586, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2811 + }, + { + "epoch": 4.492012779552716, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2812 + }, + { + "epoch": 4.493610223642173, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2813 + }, + { + "epoch": 4.49520766773163, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2814 + }, + { + "epoch": 4.496805111821086, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2815 + }, + { + "epoch": 4.498402555910543, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2816 + }, + { + "epoch": 4.5, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2817 + }, + { + "epoch": 4.501597444089457, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 2818 + }, + { + "epoch": 4.503194888178914, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2819 + }, + { + "epoch": 4.50479233226837, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2820 + }, + { + "epoch": 4.506389776357827, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 2821 + }, + { + "epoch": 4.507987220447284, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2822 + }, + { + "epoch": 4.5095846645367414, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2823 + }, + { + "epoch": 4.511182108626198, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2824 + }, + { + "epoch": 4.512779552715655, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2825 + }, + { + "epoch": 4.514376996805112, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2826 + }, + { + "epoch": 4.515974440894569, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 2827 + }, + { + "epoch": 4.517571884984026, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2828 + }, + { + "epoch": 4.519169329073483, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2829 + }, + { + "epoch": 4.520766773162939, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2830 + }, + { + "epoch": 4.522364217252396, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2831 + }, + { + "epoch": 4.523961661341853, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2832 + }, + { + "epoch": 4.52555910543131, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2833 + }, + { + "epoch": 4.527156549520766, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2834 + }, + { + "epoch": 4.5287539936102235, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2835 + }, + { + "epoch": 4.5303514376996805, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2836 + }, + { + "epoch": 4.531948881789138, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 2837 + }, + { + "epoch": 4.533546325878595, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 2838 + }, + { + "epoch": 4.535143769968051, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2839 + }, + { + "epoch": 4.536741214057508, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2840 + }, + { + "epoch": 4.538338658146965, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 2841 + }, + { + "epoch": 4.539936102236422, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2842 + }, + { + "epoch": 4.541533546325878, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 2843 + }, + { + "epoch": 4.543130990415335, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2844 + }, + { + "epoch": 4.544728434504792, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2845 + }, + { + "epoch": 4.546325878594249, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2846 + }, + { + "epoch": 4.547923322683706, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2847 + }, + { + "epoch": 4.549520766773163, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2848 + }, + { + "epoch": 4.55111821086262, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2849 + }, + { + "epoch": 4.552715654952077, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2850 + }, + { + "epoch": 4.554313099041534, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2851 + }, + { + "epoch": 4.555910543130991, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2852 + }, + { + "epoch": 4.557507987220447, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 2853 + }, + { + "epoch": 4.559105431309904, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2854 + }, + { + "epoch": 4.560702875399361, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2855 + }, + { + "epoch": 4.562300319488818, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2856 + }, + { + "epoch": 4.563897763578275, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2857 + }, + { + "epoch": 4.565495207667731, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2858 + }, + { + "epoch": 4.567092651757188, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2859 + }, + { + "epoch": 4.568690095846645, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2860 + }, + { + "epoch": 4.5702875399361025, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2861 + }, + { + "epoch": 4.571884984025559, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2862 + }, + { + "epoch": 4.573482428115016, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2863 + }, + { + "epoch": 4.575079872204473, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2864 + }, + { + "epoch": 4.57667731629393, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2865 + }, + { + "epoch": 4.578274760383387, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2866 + }, + { + "epoch": 4.579872204472844, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2867 + }, + { + "epoch": 4.5814696485623, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2868 + }, + { + "epoch": 4.583067092651757, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 2869 + }, + { + "epoch": 4.584664536741214, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 2870 + }, + { + "epoch": 4.586261980830671, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 2871 + }, + { + "epoch": 4.587859424920127, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2872 + }, + { + "epoch": 4.5894568690095845, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 2873 + }, + { + "epoch": 4.5910543130990416, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2874 + }, + { + "epoch": 4.592651757188499, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2875 + }, + { + "epoch": 4.594249201277956, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2876 + }, + { + "epoch": 4.595846645367412, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 2877 + }, + { + "epoch": 4.597444089456869, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2878 + }, + { + "epoch": 4.599041533546326, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2879 + }, + { + "epoch": 4.600638977635783, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2880 + }, + { + "epoch": 4.602236421725239, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2881 + }, + { + "epoch": 4.603833865814696, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2882 + }, + { + "epoch": 4.605431309904153, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2883 + }, + { + "epoch": 4.60702875399361, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2884 + }, + { + "epoch": 4.608626198083067, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 2885 + }, + { + "epoch": 4.6102236421725244, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2886 + }, + { + "epoch": 4.611821086261981, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 2887 + }, + { + "epoch": 4.613418530351438, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2888 + }, + { + "epoch": 4.615015974440895, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2889 + }, + { + "epoch": 4.616613418530352, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2890 + }, + { + "epoch": 4.618210862619808, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2891 + }, + { + "epoch": 4.619808306709265, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2892 + }, + { + "epoch": 4.621405750798722, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2893 + }, + { + "epoch": 4.623003194888179, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2894 + }, + { + "epoch": 4.624600638977636, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2895 + }, + { + "epoch": 4.626198083067092, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 2896 + }, + { + "epoch": 4.627795527156549, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2897 + }, + { + "epoch": 4.6293929712460065, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2898 + }, + { + "epoch": 4.6309904153354635, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2899 + }, + { + "epoch": 4.63258785942492, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2900 + }, + { + "epoch": 4.634185303514377, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2901 + }, + { + "epoch": 4.635782747603834, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2902 + }, + { + "epoch": 4.637380191693291, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2903 + }, + { + "epoch": 4.638977635782748, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2904 + }, + { + "epoch": 4.640575079872205, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2905 + }, + { + "epoch": 4.642172523961661, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 2906 + }, + { + "epoch": 4.643769968051118, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2907 + }, + { + "epoch": 4.645367412140575, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 2908 + }, + { + "epoch": 4.646964856230032, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2909 + }, + { + "epoch": 4.6485623003194885, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2910 + }, + { + "epoch": 4.6501597444089455, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2911 + }, + { + "epoch": 4.651757188498403, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2912 + }, + { + "epoch": 4.65335463258786, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 2913 + }, + { + "epoch": 4.654952076677317, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2914 + }, + { + "epoch": 4.656549520766773, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2915 + }, + { + "epoch": 4.65814696485623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2916 + }, + { + "epoch": 4.659744408945687, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 2917 + }, + { + "epoch": 4.661341853035144, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2918 + }, + { + "epoch": 4.6629392971246, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2919 + }, + { + "epoch": 4.664536741214057, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 2920 + }, + { + "epoch": 4.666134185303514, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2921 + }, + { + "epoch": 4.667731629392971, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2922 + }, + { + "epoch": 4.669329073482428, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2923 + }, + { + "epoch": 4.6709265175718855, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 2924 + }, + { + "epoch": 4.672523961661342, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2925 + }, + { + "epoch": 4.674121405750799, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2926 + }, + { + "epoch": 4.675718849840256, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 2927 + }, + { + "epoch": 4.677316293929713, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2928 + }, + { + "epoch": 4.678913738019169, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2929 + }, + { + "epoch": 4.680511182108626, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2930 + }, + { + "epoch": 4.682108626198083, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 2931 + }, + { + "epoch": 4.68370607028754, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2932 + }, + { + "epoch": 4.685303514376997, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 2933 + }, + { + "epoch": 4.686900958466453, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2934 + }, + { + "epoch": 4.68849840255591, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2935 + }, + { + "epoch": 4.6900958466453675, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2936 + }, + { + "epoch": 4.6916932907348246, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 2937 + }, + { + "epoch": 4.693290734824281, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2938 + }, + { + "epoch": 4.694888178913738, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2939 + }, + { + "epoch": 4.696485623003195, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2940 + }, + { + "epoch": 4.698083067092652, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 2941 + }, + { + "epoch": 4.699680511182109, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 2942 + }, + { + "epoch": 4.701277955271565, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2943 + }, + { + "epoch": 4.702875399361022, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2944 + }, + { + "epoch": 4.704472843450479, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2945 + }, + { + "epoch": 4.706070287539936, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2946 + }, + { + "epoch": 4.707667731629393, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2947 + }, + { + "epoch": 4.7092651757188495, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2948 + }, + { + "epoch": 4.710862619808307, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 2949 + }, + { + "epoch": 4.712460063897764, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2950 + }, + { + "epoch": 4.714057507987221, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2951 + }, + { + "epoch": 4.715654952076678, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2952 + }, + { + "epoch": 4.717252396166134, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2953 + }, + { + "epoch": 4.718849840255591, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2954 + }, + { + "epoch": 4.720447284345048, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 2955 + }, + { + "epoch": 4.722044728434505, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 2956 + }, + { + "epoch": 4.723642172523961, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 2957 + }, + { + "epoch": 4.725239616613418, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 2958 + }, + { + "epoch": 4.726837060702875, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2959 + }, + { + "epoch": 4.728434504792332, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2960 + }, + { + "epoch": 4.7300319488817895, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 2961 + }, + { + "epoch": 4.731629392971246, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 2962 + }, + { + "epoch": 4.733226837060703, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2963 + }, + { + "epoch": 4.73482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2964 + }, + { + "epoch": 4.736421725239617, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2965 + }, + { + "epoch": 4.738019169329074, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2966 + }, + { + "epoch": 4.73961661341853, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2967 + }, + { + "epoch": 4.741214057507987, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 2968 + }, + { + "epoch": 4.742811501597444, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2969 + }, + { + "epoch": 4.744408945686901, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2970 + }, + { + "epoch": 4.746006389776358, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2971 + }, + { + "epoch": 4.747603833865814, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2972 + }, + { + "epoch": 4.7492012779552715, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2973 + }, + { + "epoch": 4.7507987220447285, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2974 + }, + { + "epoch": 4.752396166134186, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2975 + }, + { + "epoch": 4.753993610223642, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2976 + }, + { + "epoch": 4.755591054313099, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2977 + }, + { + "epoch": 4.757188498402556, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2978 + }, + { + "epoch": 4.758785942492013, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2979 + }, + { + "epoch": 4.76038338658147, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2980 + }, + { + "epoch": 4.761980830670926, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2981 + }, + { + "epoch": 4.763578274760383, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2982 + }, + { + "epoch": 4.76517571884984, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2983 + }, + { + "epoch": 4.766773162939297, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2984 + }, + { + "epoch": 4.768370607028754, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2985 + }, + { + "epoch": 4.7699680511182105, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 2986 + }, + { + "epoch": 4.771565495207668, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2987 + }, + { + "epoch": 4.773162939297125, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 2988 + }, + { + "epoch": 4.774760383386582, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2989 + }, + { + "epoch": 4.776357827476039, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2990 + }, + { + "epoch": 4.777955271565495, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2991 + }, + { + "epoch": 4.779552715654952, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2992 + }, + { + "epoch": 4.781150159744409, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2993 + }, + { + "epoch": 4.782747603833866, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2994 + }, + { + "epoch": 4.784345047923322, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2995 + }, + { + "epoch": 4.785942492012779, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2996 + }, + { + "epoch": 4.787539936102236, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2997 + }, + { + "epoch": 4.789137380191693, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2998 + }, + { + "epoch": 4.7907348242811505, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2999 + }, + { + "epoch": 4.792332268370607, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3000 + }, + { + "epoch": 4.793929712460064, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3001 + }, + { + "epoch": 4.795527156549521, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3002 + }, + { + "epoch": 4.797124600638978, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3003 + }, + { + "epoch": 4.798722044728435, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 3004 + }, + { + "epoch": 4.800319488817891, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3005 + }, + { + "epoch": 4.801916932907348, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3006 + }, + { + "epoch": 4.803514376996805, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3007 + }, + { + "epoch": 4.805111821086262, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 3008 + }, + { + "epoch": 4.806709265175719, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 3009 + }, + { + "epoch": 4.8083067092651754, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3010 + }, + { + "epoch": 4.8099041533546325, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3011 + }, + { + "epoch": 4.81150159744409, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3012 + }, + { + "epoch": 4.813099041533547, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3013 + }, + { + "epoch": 4.814696485623003, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3014 + }, + { + "epoch": 4.81629392971246, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 3015 + }, + { + "epoch": 4.817891373801917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3016 + }, + { + "epoch": 4.819488817891374, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 3017 + }, + { + "epoch": 4.821086261980831, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3018 + }, + { + "epoch": 4.822683706070287, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3019 + }, + { + "epoch": 4.824281150159744, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3020 + }, + { + "epoch": 4.825878594249201, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3021 + }, + { + "epoch": 4.827476038338658, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3022 + }, + { + "epoch": 4.8290734824281145, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3023 + }, + { + "epoch": 4.830670926517572, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3024 + }, + { + "epoch": 4.832268370607029, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 3025 + }, + { + "epoch": 4.833865814696486, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3026 + }, + { + "epoch": 4.835463258785943, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3027 + }, + { + "epoch": 4.8370607028754, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3028 + }, + { + "epoch": 4.838658146964856, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3029 + }, + { + "epoch": 4.840255591054313, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3030 + }, + { + "epoch": 4.84185303514377, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3031 + }, + { + "epoch": 4.843450479233227, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3032 + }, + { + "epoch": 4.845047923322683, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3033 + }, + { + "epoch": 4.84664536741214, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3034 + }, + { + "epoch": 4.848242811501597, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3035 + }, + { + "epoch": 4.8498402555910545, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 3036 + }, + { + "epoch": 4.8514376996805115, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3037 + }, + { + "epoch": 4.853035143769968, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 3038 + }, + { + "epoch": 4.854632587859425, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3039 + }, + { + "epoch": 4.856230031948882, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3040 + }, + { + "epoch": 4.857827476038339, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3041 + }, + { + "epoch": 4.859424920127795, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3042 + }, + { + "epoch": 4.861022364217252, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3043 + }, + { + "epoch": 4.862619808306709, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3044 + }, + { + "epoch": 4.864217252396166, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3045 + }, + { + "epoch": 4.865814696485623, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3046 + }, + { + "epoch": 4.86741214057508, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3047 + }, + { + "epoch": 4.8690095846645365, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3048 + }, + { + "epoch": 4.8706070287539935, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3049 + }, + { + "epoch": 4.872204472843451, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 3050 + }, + { + "epoch": 4.873801916932908, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 3051 + }, + { + "epoch": 4.875399361022364, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3052 + }, + { + "epoch": 4.876996805111821, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3053 + }, + { + "epoch": 4.878594249201278, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3054 + }, + { + "epoch": 4.880191693290735, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3055 + }, + { + "epoch": 4.881789137380192, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 3056 + }, + { + "epoch": 4.883386581469648, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3057 + }, + { + "epoch": 4.884984025559105, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3058 + }, + { + "epoch": 4.886581469648562, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3059 + }, + { + "epoch": 4.888178913738019, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3060 + }, + { + "epoch": 4.8897763578274756, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3061 + }, + { + "epoch": 4.891373801916933, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3062 + }, + { + "epoch": 4.89297124600639, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3063 + }, + { + "epoch": 4.894568690095847, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3064 + }, + { + "epoch": 4.896166134185304, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 3065 + }, + { + "epoch": 4.897763578274761, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3066 + }, + { + "epoch": 4.899361022364217, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 3067 + }, + { + "epoch": 4.900958466453674, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3068 + }, + { + "epoch": 4.902555910543131, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3069 + }, + { + "epoch": 4.904153354632588, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3070 + }, + { + "epoch": 4.905750798722044, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 3071 + }, + { + "epoch": 4.907348242811501, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3072 + }, + { + "epoch": 4.9089456869009584, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3073 + }, + { + "epoch": 4.9105431309904155, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3074 + }, + { + "epoch": 4.912140575079873, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3075 + }, + { + "epoch": 4.913738019169329, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 3076 + }, + { + "epoch": 4.915335463258786, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3077 + }, + { + "epoch": 4.916932907348243, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3078 + }, + { + "epoch": 4.9185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3079 + }, + { + "epoch": 4.920127795527156, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 3080 + }, + { + "epoch": 4.921725239616613, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3081 + }, + { + "epoch": 4.92332268370607, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 3082 + }, + { + "epoch": 4.924920127795527, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 3083 + }, + { + "epoch": 4.926517571884984, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3084 + }, + { + "epoch": 4.928115015974441, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3085 + }, + { + "epoch": 4.9297124600638975, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3086 + }, + { + "epoch": 4.931309904153355, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3087 + }, + { + "epoch": 4.932907348242812, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3088 + }, + { + "epoch": 4.934504792332269, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3089 + }, + { + "epoch": 4.936102236421725, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3090 + }, + { + "epoch": 4.937699680511182, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3091 + }, + { + "epoch": 4.939297124600639, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 3092 + }, + { + "epoch": 4.940894568690096, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 3093 + }, + { + "epoch": 4.942492012779553, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3094 + }, + { + "epoch": 4.944089456869009, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3095 + }, + { + "epoch": 4.945686900958466, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3096 + }, + { + "epoch": 4.947284345047923, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3097 + }, + { + "epoch": 4.94888178913738, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3098 + }, + { + "epoch": 4.950479233226837, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3099 + }, + { + "epoch": 4.952076677316294, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3100 + }, + { + "epoch": 4.953674121405751, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3101 + }, + { + "epoch": 4.955271565495208, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3102 + }, + { + "epoch": 4.956869009584665, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3103 + }, + { + "epoch": 4.958466453674122, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3104 + }, + { + "epoch": 4.960063897763578, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3105 + }, + { + "epoch": 4.961661341853035, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 3106 + }, + { + "epoch": 4.963258785942492, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3107 + }, + { + "epoch": 4.964856230031949, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3108 + }, + { + "epoch": 4.966453674121405, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 3109 + }, + { + "epoch": 4.968051118210862, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3110 + }, + { + "epoch": 4.9696485623003195, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 3111 + }, + { + "epoch": 4.9712460063897765, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3112 + }, + { + "epoch": 4.972843450479234, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3113 + }, + { + "epoch": 4.97444089456869, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3114 + }, + { + "epoch": 4.976038338658147, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 3115 + }, + { + "epoch": 4.977635782747604, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 3116 + }, + { + "epoch": 4.979233226837061, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3117 + }, + { + "epoch": 4.980830670926517, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3118 + }, + { + "epoch": 4.982428115015974, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3119 + }, + { + "epoch": 4.984025559105431, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3120 + }, + { + "epoch": 4.985623003194888, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3121 + }, + { + "epoch": 4.987220447284345, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3122 + }, + { + "epoch": 4.988817891373802, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3123 + }, + { + "epoch": 4.9904153354632586, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3124 + }, + { + "epoch": 4.992012779552716, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3125 + }, + { + "epoch": 4.993610223642173, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3126 + }, + { + "epoch": 4.99520766773163, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3127 + }, + { + "epoch": 4.996805111821086, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3128 + }, + { + "epoch": 4.998402555910543, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3129 + }, + { + "epoch": 5.0, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3130 + }, + { + "epoch": 5.001597444089457, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3131 + }, + { + "epoch": 5.003194888178914, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3132 + }, + { + "epoch": 5.00479233226837, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3133 + }, + { + "epoch": 5.006389776357827, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3134 + }, + { + "epoch": 5.007987220447284, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3135 + }, + { + "epoch": 5.0095846645367414, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3136 + }, + { + "epoch": 5.0111821086261985, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3137 + }, + { + "epoch": 5.012779552715655, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 3138 + }, + { + "epoch": 5.014376996805112, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3139 + }, + { + "epoch": 5.015974440894569, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3140 + }, + { + "epoch": 5.017571884984026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3141 + }, + { + "epoch": 5.019169329073482, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3142 + }, + { + "epoch": 5.020766773162939, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3143 + }, + { + "epoch": 5.022364217252396, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3144 + }, + { + "epoch": 5.023961661341853, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 3145 + }, + { + "epoch": 5.02555910543131, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3146 + }, + { + "epoch": 5.027156549520766, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3147 + }, + { + "epoch": 5.0287539936102235, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3148 + }, + { + "epoch": 5.0303514376996805, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3149 + }, + { + "epoch": 5.031948881789138, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3150 + }, + { + "epoch": 5.033546325878595, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3151 + }, + { + "epoch": 5.035143769968051, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3152 + }, + { + "epoch": 5.036741214057508, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3153 + }, + { + "epoch": 5.038338658146965, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3154 + }, + { + "epoch": 5.039936102236422, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3155 + }, + { + "epoch": 5.041533546325879, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3156 + }, + { + "epoch": 5.043130990415335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3157 + }, + { + "epoch": 5.044728434504792, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 3158 + }, + { + "epoch": 5.046325878594249, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 3159 + }, + { + "epoch": 5.047923322683706, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3160 + }, + { + "epoch": 5.0495207667731625, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3161 + }, + { + "epoch": 5.05111821086262, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3162 + }, + { + "epoch": 5.052715654952077, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3163 + }, + { + "epoch": 5.054313099041534, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3164 + }, + { + "epoch": 5.055910543130991, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3165 + }, + { + "epoch": 5.057507987220447, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3166 + }, + { + "epoch": 5.059105431309904, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3167 + }, + { + "epoch": 5.060702875399361, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3168 + }, + { + "epoch": 5.062300319488818, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3169 + }, + { + "epoch": 5.063897763578275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3170 + }, + { + "epoch": 5.065495207667731, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3171 + }, + { + "epoch": 5.067092651757188, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3172 + }, + { + "epoch": 5.068690095846645, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 3173 + }, + { + "epoch": 5.0702875399361025, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3174 + }, + { + "epoch": 5.0718849840255595, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3175 + }, + { + "epoch": 5.073482428115016, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 3176 + }, + { + "epoch": 5.075079872204473, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 3177 + }, + { + "epoch": 5.07667731629393, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3178 + }, + { + "epoch": 5.078274760383387, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 3179 + }, + { + "epoch": 5.079872204472843, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3180 + }, + { + "epoch": 5.0814696485623, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3181 + }, + { + "epoch": 5.083067092651757, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3182 + }, + { + "epoch": 5.084664536741214, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3183 + }, + { + "epoch": 5.086261980830671, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 3184 + }, + { + "epoch": 5.087859424920127, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 3185 + }, + { + "epoch": 5.0894568690095845, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3186 + }, + { + "epoch": 5.0910543130990416, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3187 + }, + { + "epoch": 5.092651757188499, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3188 + }, + { + "epoch": 5.094249201277956, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 3189 + }, + { + "epoch": 5.095846645367412, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3190 + }, + { + "epoch": 5.097444089456869, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3191 + }, + { + "epoch": 5.099041533546326, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3192 + }, + { + "epoch": 5.100638977635783, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3193 + }, + { + "epoch": 5.102236421725239, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3194 + }, + { + "epoch": 5.103833865814696, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3195 + }, + { + "epoch": 5.105431309904153, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3196 + }, + { + "epoch": 5.10702875399361, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3197 + }, + { + "epoch": 5.108626198083067, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3198 + }, + { + "epoch": 5.110223642172524, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 3199 + }, + { + "epoch": 5.111821086261981, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3200 + }, + { + "epoch": 5.113418530351438, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 3201 + }, + { + "epoch": 5.115015974440895, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 3202 + }, + { + "epoch": 5.116613418530352, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3203 + }, + { + "epoch": 5.118210862619808, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3204 + }, + { + "epoch": 5.119808306709265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 3205 + }, + { + "epoch": 5.121405750798722, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3206 + }, + { + "epoch": 5.123003194888179, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3207 + }, + { + "epoch": 5.124600638977636, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3208 + }, + { + "epoch": 5.126198083067092, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3209 + }, + { + "epoch": 5.127795527156549, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3210 + }, + { + "epoch": 5.1293929712460065, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3211 + }, + { + "epoch": 5.1309904153354635, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3212 + }, + { + "epoch": 5.13258785942492, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3213 + }, + { + "epoch": 5.134185303514377, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3214 + }, + { + "epoch": 5.135782747603834, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3215 + }, + { + "epoch": 5.137380191693291, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3216 + }, + { + "epoch": 5.138977635782748, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3217 + }, + { + "epoch": 5.140575079872204, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 3218 + }, + { + "epoch": 5.142172523961661, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3219 + }, + { + "epoch": 5.143769968051118, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3220 + }, + { + "epoch": 5.145367412140575, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3221 + }, + { + "epoch": 5.146964856230032, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3222 + }, + { + "epoch": 5.1485623003194885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3223 + }, + { + "epoch": 5.1501597444089455, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3224 + }, + { + "epoch": 5.151757188498403, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3225 + }, + { + "epoch": 5.15335463258786, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3226 + }, + { + "epoch": 5.154952076677317, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3227 + }, + { + "epoch": 5.156549520766773, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3228 + }, + { + "epoch": 5.15814696485623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3229 + }, + { + "epoch": 5.159744408945687, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3230 + }, + { + "epoch": 5.161341853035144, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3231 + }, + { + "epoch": 5.1629392971246, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3232 + }, + { + "epoch": 5.164536741214057, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 3233 + }, + { + "epoch": 5.166134185303514, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3234 + }, + { + "epoch": 5.167731629392971, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3235 + }, + { + "epoch": 5.169329073482428, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3236 + }, + { + "epoch": 5.170926517571885, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 3237 + }, + { + "epoch": 5.172523961661342, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3238 + }, + { + "epoch": 5.174121405750799, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3239 + }, + { + "epoch": 5.175718849840256, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3240 + }, + { + "epoch": 5.177316293929713, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3241 + }, + { + "epoch": 5.178913738019169, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 3242 + }, + { + "epoch": 5.180511182108626, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3243 + }, + { + "epoch": 5.182108626198083, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3244 + }, + { + "epoch": 5.18370607028754, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3245 + }, + { + "epoch": 5.185303514376997, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 3246 + }, + { + "epoch": 5.186900958466453, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3247 + }, + { + "epoch": 5.18849840255591, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3248 + }, + { + "epoch": 5.1900958466453675, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3249 + }, + { + "epoch": 5.1916932907348246, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3250 + }, + { + "epoch": 5.193290734824281, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3251 + }, + { + "epoch": 5.194888178913738, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3252 + }, + { + "epoch": 5.196485623003195, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 3253 + }, + { + "epoch": 5.198083067092652, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3254 + }, + { + "epoch": 5.199680511182109, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3255 + }, + { + "epoch": 5.201277955271565, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3256 + }, + { + "epoch": 5.202875399361022, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3257 + }, + { + "epoch": 5.204472843450479, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3258 + }, + { + "epoch": 5.206070287539936, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3259 + }, + { + "epoch": 5.207667731629393, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3260 + }, + { + "epoch": 5.2092651757188495, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3261 + }, + { + "epoch": 5.210862619808307, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3262 + }, + { + "epoch": 5.212460063897764, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3263 + }, + { + "epoch": 5.214057507987221, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3264 + }, + { + "epoch": 5.215654952076678, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 3265 + }, + { + "epoch": 5.217252396166134, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3266 + }, + { + "epoch": 5.218849840255591, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3267 + }, + { + "epoch": 5.220447284345048, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3268 + }, + { + "epoch": 5.222044728434505, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3269 + }, + { + "epoch": 5.223642172523961, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 3270 + }, + { + "epoch": 5.225239616613418, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3271 + }, + { + "epoch": 5.226837060702875, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3272 + }, + { + "epoch": 5.228434504792332, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3273 + }, + { + "epoch": 5.2300319488817895, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3274 + }, + { + "epoch": 5.231629392971246, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3275 + }, + { + "epoch": 5.233226837060703, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3276 + }, + { + "epoch": 5.23482428115016, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3277 + }, + { + "epoch": 5.236421725239617, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3278 + }, + { + "epoch": 5.238019169329074, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3279 + }, + { + "epoch": 5.23961661341853, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3280 + }, + { + "epoch": 5.241214057507987, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3281 + }, + { + "epoch": 5.242811501597444, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 3282 + }, + { + "epoch": 5.244408945686901, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3283 + }, + { + "epoch": 5.246006389776358, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3284 + }, + { + "epoch": 5.247603833865814, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3285 + }, + { + "epoch": 5.2492012779552715, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3286 + }, + { + "epoch": 5.2507987220447285, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3287 + }, + { + "epoch": 5.252396166134186, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3288 + }, + { + "epoch": 5.253993610223642, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3289 + }, + { + "epoch": 5.255591054313099, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3290 + }, + { + "epoch": 5.257188498402556, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3291 + }, + { + "epoch": 5.258785942492013, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3292 + }, + { + "epoch": 5.26038338658147, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3293 + }, + { + "epoch": 5.261980830670926, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3294 + }, + { + "epoch": 5.263578274760383, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3295 + }, + { + "epoch": 5.26517571884984, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3296 + }, + { + "epoch": 5.266773162939297, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3297 + }, + { + "epoch": 5.268370607028754, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3298 + }, + { + "epoch": 5.2699680511182105, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3299 + }, + { + "epoch": 5.271565495207668, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3300 + }, + { + "epoch": 5.273162939297125, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 3301 + }, + { + "epoch": 5.274760383386582, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 3302 + }, + { + "epoch": 5.276357827476039, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3303 + }, + { + "epoch": 5.277955271565495, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3304 + }, + { + "epoch": 5.279552715654952, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 3305 + }, + { + "epoch": 5.281150159744409, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 3306 + }, + { + "epoch": 5.282747603833866, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3307 + }, + { + "epoch": 5.284345047923322, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3308 + }, + { + "epoch": 5.285942492012779, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3309 + }, + { + "epoch": 5.287539936102236, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3310 + }, + { + "epoch": 5.289137380191693, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3311 + }, + { + "epoch": 5.2907348242811505, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3312 + }, + { + "epoch": 5.292332268370607, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3313 + }, + { + "epoch": 5.293929712460064, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3314 + }, + { + "epoch": 5.295527156549521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3315 + }, + { + "epoch": 5.297124600638978, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3316 + }, + { + "epoch": 5.298722044728435, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3317 + }, + { + "epoch": 5.300319488817891, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3318 + }, + { + "epoch": 5.301916932907348, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3319 + }, + { + "epoch": 5.303514376996805, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3320 + }, + { + "epoch": 5.305111821086262, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3321 + }, + { + "epoch": 5.306709265175719, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3322 + }, + { + "epoch": 5.3083067092651754, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3323 + }, + { + "epoch": 5.3099041533546325, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3324 + }, + { + "epoch": 5.31150159744409, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3325 + }, + { + "epoch": 5.313099041533547, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3326 + }, + { + "epoch": 5.314696485623003, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3327 + }, + { + "epoch": 5.31629392971246, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3328 + }, + { + "epoch": 5.317891373801917, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3329 + }, + { + "epoch": 5.319488817891374, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3330 + }, + { + "epoch": 5.321086261980831, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 3331 + }, + { + "epoch": 5.322683706070287, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3332 + }, + { + "epoch": 5.324281150159744, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3333 + }, + { + "epoch": 5.325878594249201, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3334 + }, + { + "epoch": 5.327476038338658, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3335 + }, + { + "epoch": 5.329073482428115, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 3336 + }, + { + "epoch": 5.330670926517572, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3337 + }, + { + "epoch": 5.332268370607029, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3338 + }, + { + "epoch": 5.333865814696486, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3339 + }, + { + "epoch": 5.335463258785943, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3340 + }, + { + "epoch": 5.3370607028754, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3341 + }, + { + "epoch": 5.338658146964856, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3342 + }, + { + "epoch": 5.340255591054313, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3343 + }, + { + "epoch": 5.34185303514377, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3344 + }, + { + "epoch": 5.343450479233227, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3345 + }, + { + "epoch": 5.345047923322683, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3346 + }, + { + "epoch": 5.34664536741214, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3347 + }, + { + "epoch": 5.348242811501597, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3348 + }, + { + "epoch": 5.3498402555910545, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3349 + }, + { + "epoch": 5.3514376996805115, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3350 + }, + { + "epoch": 5.353035143769968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3351 + }, + { + "epoch": 5.354632587859425, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3352 + }, + { + "epoch": 5.356230031948882, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3353 + }, + { + "epoch": 5.357827476038339, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3354 + }, + { + "epoch": 5.359424920127796, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3355 + }, + { + "epoch": 5.361022364217252, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3356 + }, + { + "epoch": 5.362619808306709, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3357 + }, + { + "epoch": 5.364217252396166, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3358 + }, + { + "epoch": 5.365814696485623, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3359 + }, + { + "epoch": 5.36741214057508, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3360 + }, + { + "epoch": 5.3690095846645365, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3361 + }, + { + "epoch": 5.3706070287539935, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 3362 + }, + { + "epoch": 5.372204472843451, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3363 + }, + { + "epoch": 5.373801916932908, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3364 + }, + { + "epoch": 5.375399361022364, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3365 + }, + { + "epoch": 5.376996805111821, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3366 + }, + { + "epoch": 5.378594249201278, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3367 + }, + { + "epoch": 5.380191693290735, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3368 + }, + { + "epoch": 5.381789137380192, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3369 + }, + { + "epoch": 5.383386581469648, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3370 + }, + { + "epoch": 5.384984025559105, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3371 + }, + { + "epoch": 5.386581469648562, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3372 + }, + { + "epoch": 5.388178913738019, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3373 + }, + { + "epoch": 5.389776357827476, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3374 + }, + { + "epoch": 5.391373801916933, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 3375 + }, + { + "epoch": 5.39297124600639, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 3376 + }, + { + "epoch": 5.394568690095847, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3377 + }, + { + "epoch": 5.396166134185304, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3378 + }, + { + "epoch": 5.397763578274761, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3379 + }, + { + "epoch": 5.399361022364217, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3380 + }, + { + "epoch": 5.400958466453674, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3381 + }, + { + "epoch": 5.402555910543131, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3382 + }, + { + "epoch": 5.404153354632588, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3383 + }, + { + "epoch": 5.405750798722044, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3384 + }, + { + "epoch": 5.407348242811501, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3385 + }, + { + "epoch": 5.4089456869009584, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3386 + }, + { + "epoch": 5.4105431309904155, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3387 + }, + { + "epoch": 5.412140575079873, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3388 + }, + { + "epoch": 5.413738019169329, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3389 + }, + { + "epoch": 5.415335463258786, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 3390 + }, + { + "epoch": 5.416932907348243, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3391 + }, + { + "epoch": 5.4185303514377, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3392 + }, + { + "epoch": 5.420127795527157, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 3393 + }, + { + "epoch": 5.421725239616613, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3394 + }, + { + "epoch": 5.42332268370607, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3395 + }, + { + "epoch": 5.424920127795527, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3396 + }, + { + "epoch": 5.426517571884984, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3397 + }, + { + "epoch": 5.428115015974441, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3398 + }, + { + "epoch": 5.4297124600638975, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3399 + }, + { + "epoch": 5.431309904153355, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3400 + }, + { + "epoch": 5.432907348242812, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3401 + }, + { + "epoch": 5.434504792332269, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3402 + }, + { + "epoch": 5.436102236421725, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3403 + }, + { + "epoch": 5.437699680511182, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3404 + }, + { + "epoch": 5.439297124600639, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3405 + }, + { + "epoch": 5.440894568690096, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3406 + }, + { + "epoch": 5.442492012779553, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3407 + }, + { + "epoch": 5.444089456869009, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3408 + }, + { + "epoch": 5.445686900958466, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3409 + }, + { + "epoch": 5.447284345047923, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3410 + }, + { + "epoch": 5.44888178913738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3411 + }, + { + "epoch": 5.4504792332268375, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3412 + }, + { + "epoch": 5.452076677316294, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3413 + }, + { + "epoch": 5.453674121405751, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3414 + }, + { + "epoch": 5.455271565495208, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 3415 + }, + { + "epoch": 5.456869009584665, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3416 + }, + { + "epoch": 5.458466453674122, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3417 + }, + { + "epoch": 5.460063897763578, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3418 + }, + { + "epoch": 5.461661341853035, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 3419 + }, + { + "epoch": 5.463258785942492, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3420 + }, + { + "epoch": 5.464856230031949, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3421 + }, + { + "epoch": 5.466453674121405, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3422 + }, + { + "epoch": 5.468051118210862, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3423 + }, + { + "epoch": 5.4696485623003195, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3424 + }, + { + "epoch": 5.4712460063897765, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3425 + }, + { + "epoch": 5.472843450479234, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 3426 + }, + { + "epoch": 5.47444089456869, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3427 + }, + { + "epoch": 5.476038338658147, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3428 + }, + { + "epoch": 5.477635782747604, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3429 + }, + { + "epoch": 5.479233226837061, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 3430 + }, + { + "epoch": 5.480830670926517, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3431 + }, + { + "epoch": 5.482428115015974, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3432 + }, + { + "epoch": 5.484025559105431, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3433 + }, + { + "epoch": 5.485623003194888, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3434 + }, + { + "epoch": 5.487220447284345, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 3435 + }, + { + "epoch": 5.488817891373802, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3436 + }, + { + "epoch": 5.4904153354632586, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3437 + }, + { + "epoch": 5.492012779552716, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3438 + }, + { + "epoch": 5.493610223642173, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3439 + }, + { + "epoch": 5.49520766773163, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3440 + }, + { + "epoch": 5.496805111821086, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3441 + }, + { + "epoch": 5.498402555910543, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3442 + }, + { + "epoch": 5.5, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3443 + }, + { + "epoch": 5.501597444089457, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3444 + }, + { + "epoch": 5.503194888178914, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3445 + }, + { + "epoch": 5.50479233226837, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3446 + }, + { + "epoch": 5.506389776357827, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3447 + }, + { + "epoch": 5.507987220447284, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3448 + }, + { + "epoch": 5.5095846645367414, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3449 + }, + { + "epoch": 5.511182108626198, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 3450 + }, + { + "epoch": 5.512779552715655, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3451 + }, + { + "epoch": 5.514376996805112, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3452 + }, + { + "epoch": 5.515974440894569, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3453 + }, + { + "epoch": 5.517571884984026, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3454 + }, + { + "epoch": 5.519169329073483, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3455 + }, + { + "epoch": 5.520766773162939, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3456 + }, + { + "epoch": 5.522364217252396, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3457 + }, + { + "epoch": 5.523961661341853, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3458 + }, + { + "epoch": 5.52555910543131, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 3459 + }, + { + "epoch": 5.527156549520766, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3460 + }, + { + "epoch": 5.5287539936102235, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3461 + }, + { + "epoch": 5.5303514376996805, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3462 + }, + { + "epoch": 5.531948881789138, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 3463 + }, + { + "epoch": 5.533546325878595, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3464 + }, + { + "epoch": 5.535143769968051, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3465 + }, + { + "epoch": 5.536741214057508, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 3466 + }, + { + "epoch": 5.538338658146965, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3467 + }, + { + "epoch": 5.539936102236422, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3468 + }, + { + "epoch": 5.541533546325878, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3469 + }, + { + "epoch": 5.543130990415335, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3470 + }, + { + "epoch": 5.544728434504792, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3471 + }, + { + "epoch": 5.546325878594249, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3472 + }, + { + "epoch": 5.547923322683706, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3473 + }, + { + "epoch": 5.549520766773163, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 3474 + }, + { + "epoch": 5.55111821086262, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3475 + }, + { + "epoch": 5.552715654952077, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 3476 + }, + { + "epoch": 5.554313099041534, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3477 + }, + { + "epoch": 5.555910543130991, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3478 + }, + { + "epoch": 5.557507987220447, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3479 + }, + { + "epoch": 5.559105431309904, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3480 + }, + { + "epoch": 5.560702875399361, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3481 + }, + { + "epoch": 5.562300319488818, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3482 + }, + { + "epoch": 5.563897763578275, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3483 + }, + { + "epoch": 5.565495207667731, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3484 + }, + { + "epoch": 5.567092651757188, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3485 + }, + { + "epoch": 5.568690095846645, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3486 + }, + { + "epoch": 5.5702875399361025, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3487 + }, + { + "epoch": 5.571884984025559, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3488 + }, + { + "epoch": 5.573482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3489 + }, + { + "epoch": 5.575079872204473, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3490 + }, + { + "epoch": 5.57667731629393, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3491 + }, + { + "epoch": 5.578274760383387, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 3492 + }, + { + "epoch": 5.579872204472844, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3493 + }, + { + "epoch": 5.5814696485623, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 3494 + }, + { + "epoch": 5.583067092651757, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3495 + }, + { + "epoch": 5.584664536741214, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3496 + }, + { + "epoch": 5.586261980830671, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3497 + }, + { + "epoch": 5.587859424920127, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3498 + }, + { + "epoch": 5.5894568690095845, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3499 + }, + { + "epoch": 5.5910543130990416, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3500 + }, + { + "epoch": 5.592651757188499, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3501 + }, + { + "epoch": 5.594249201277956, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3502 + }, + { + "epoch": 5.595846645367412, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3503 + }, + { + "epoch": 5.597444089456869, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3504 + }, + { + "epoch": 5.599041533546326, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3505 + }, + { + "epoch": 5.600638977635783, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3506 + }, + { + "epoch": 5.602236421725239, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3507 + }, + { + "epoch": 5.603833865814696, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3508 + }, + { + "epoch": 5.605431309904153, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3509 + }, + { + "epoch": 5.60702875399361, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3510 + }, + { + "epoch": 5.608626198083067, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3511 + }, + { + "epoch": 5.6102236421725244, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 3512 + }, + { + "epoch": 5.611821086261981, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3513 + }, + { + "epoch": 5.613418530351438, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3514 + }, + { + "epoch": 5.615015974440895, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3515 + }, + { + "epoch": 5.616613418530352, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3516 + }, + { + "epoch": 5.618210862619808, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3517 + }, + { + "epoch": 5.619808306709265, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 3518 + }, + { + "epoch": 5.621405750798722, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3519 + }, + { + "epoch": 5.623003194888179, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3520 + }, + { + "epoch": 5.624600638977636, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3521 + }, + { + "epoch": 5.626198083067092, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3522 + }, + { + "epoch": 5.627795527156549, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3523 + }, + { + "epoch": 5.6293929712460065, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3524 + }, + { + "epoch": 5.6309904153354635, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3525 + }, + { + "epoch": 5.63258785942492, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3526 + }, + { + "epoch": 5.634185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 3527 + }, + { + "epoch": 5.635782747603834, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3528 + }, + { + "epoch": 5.637380191693291, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3529 + }, + { + "epoch": 5.638977635782748, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3530 + }, + { + "epoch": 5.640575079872205, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 3531 + }, + { + "epoch": 5.642172523961661, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3532 + }, + { + "epoch": 5.643769968051118, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 3533 + }, + { + "epoch": 5.645367412140575, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3534 + }, + { + "epoch": 5.646964856230032, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3535 + }, + { + "epoch": 5.6485623003194885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3536 + }, + { + "epoch": 5.6501597444089455, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3537 + }, + { + "epoch": 5.651757188498403, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3538 + }, + { + "epoch": 5.65335463258786, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3539 + }, + { + "epoch": 5.654952076677317, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3540 + }, + { + "epoch": 5.656549520766773, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3541 + }, + { + "epoch": 5.65814696485623, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3542 + }, + { + "epoch": 5.659744408945687, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3543 + }, + { + "epoch": 5.661341853035144, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3544 + }, + { + "epoch": 5.6629392971246, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3545 + }, + { + "epoch": 5.664536741214057, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3546 + }, + { + "epoch": 5.666134185303514, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3547 + }, + { + "epoch": 5.667731629392971, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3548 + }, + { + "epoch": 5.669329073482428, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3549 + }, + { + "epoch": 5.6709265175718855, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3550 + }, + { + "epoch": 5.672523961661342, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3551 + }, + { + "epoch": 5.674121405750799, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3552 + }, + { + "epoch": 5.675718849840256, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3553 + }, + { + "epoch": 5.677316293929713, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3554 + }, + { + "epoch": 5.678913738019169, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3555 + }, + { + "epoch": 5.680511182108626, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 3556 + }, + { + "epoch": 5.682108626198083, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3557 + }, + { + "epoch": 5.68370607028754, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3558 + }, + { + "epoch": 5.685303514376997, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3559 + }, + { + "epoch": 5.686900958466453, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3560 + }, + { + "epoch": 5.68849840255591, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3561 + }, + { + "epoch": 5.6900958466453675, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 3562 + }, + { + "epoch": 5.6916932907348246, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 3563 + }, + { + "epoch": 5.693290734824281, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3564 + }, + { + "epoch": 5.694888178913738, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3565 + }, + { + "epoch": 5.696485623003195, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3566 + }, + { + "epoch": 5.698083067092652, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3567 + }, + { + "epoch": 5.699680511182109, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3568 + }, + { + "epoch": 5.701277955271565, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3569 + }, + { + "epoch": 5.702875399361022, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3570 + }, + { + "epoch": 5.704472843450479, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3571 + }, + { + "epoch": 5.706070287539936, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3572 + }, + { + "epoch": 5.707667731629393, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3573 + }, + { + "epoch": 5.7092651757188495, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3574 + }, + { + "epoch": 5.710862619808307, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3575 + }, + { + "epoch": 5.712460063897764, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3576 + }, + { + "epoch": 5.714057507987221, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3577 + }, + { + "epoch": 5.715654952076678, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3578 + }, + { + "epoch": 5.717252396166134, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3579 + }, + { + "epoch": 5.718849840255591, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3580 + }, + { + "epoch": 5.720447284345048, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3581 + }, + { + "epoch": 5.722044728434505, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3582 + }, + { + "epoch": 5.723642172523961, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3583 + }, + { + "epoch": 5.725239616613418, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 3584 + }, + { + "epoch": 5.726837060702875, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3585 + }, + { + "epoch": 5.728434504792332, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3586 + }, + { + "epoch": 5.7300319488817895, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3587 + }, + { + "epoch": 5.731629392971246, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3588 + }, + { + "epoch": 5.733226837060703, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3589 + }, + { + "epoch": 5.73482428115016, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3590 + }, + { + "epoch": 5.736421725239617, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3591 + }, + { + "epoch": 5.738019169329074, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3592 + }, + { + "epoch": 5.73961661341853, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3593 + }, + { + "epoch": 5.741214057507987, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 3594 + }, + { + "epoch": 5.742811501597444, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3595 + }, + { + "epoch": 5.744408945686901, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3596 + }, + { + "epoch": 5.746006389776358, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3597 + }, + { + "epoch": 5.747603833865814, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3598 + }, + { + "epoch": 5.7492012779552715, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3599 + }, + { + "epoch": 5.7507987220447285, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3600 + }, + { + "epoch": 5.752396166134186, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3601 + }, + { + "epoch": 5.753993610223642, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3602 + }, + { + "epoch": 5.755591054313099, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3603 + }, + { + "epoch": 5.757188498402556, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3604 + }, + { + "epoch": 5.758785942492013, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3605 + }, + { + "epoch": 5.76038338658147, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3606 + }, + { + "epoch": 5.761980830670926, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3607 + }, + { + "epoch": 5.763578274760383, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3608 + }, + { + "epoch": 5.76517571884984, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3609 + }, + { + "epoch": 5.766773162939297, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3610 + }, + { + "epoch": 5.768370607028754, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3611 + }, + { + "epoch": 5.7699680511182105, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3612 + }, + { + "epoch": 5.771565495207668, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3613 + }, + { + "epoch": 5.773162939297125, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3614 + }, + { + "epoch": 5.774760383386582, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3615 + }, + { + "epoch": 5.776357827476039, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3616 + }, + { + "epoch": 5.777955271565495, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3617 + }, + { + "epoch": 5.779552715654952, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3618 + }, + { + "epoch": 5.781150159744409, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3619 + }, + { + "epoch": 5.782747603833866, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3620 + }, + { + "epoch": 5.784345047923322, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3621 + }, + { + "epoch": 5.785942492012779, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3622 + }, + { + "epoch": 5.787539936102236, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3623 + }, + { + "epoch": 5.789137380191693, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 3624 + }, + { + "epoch": 5.7907348242811505, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3625 + }, + { + "epoch": 5.792332268370607, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 3626 + }, + { + "epoch": 5.793929712460064, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3627 + }, + { + "epoch": 5.795527156549521, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3628 + }, + { + "epoch": 5.797124600638978, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3629 + }, + { + "epoch": 5.798722044728435, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 3630 + }, + { + "epoch": 5.800319488817891, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 3631 + }, + { + "epoch": 5.801916932907348, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3632 + }, + { + "epoch": 5.803514376996805, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3633 + }, + { + "epoch": 5.805111821086262, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 3634 + }, + { + "epoch": 5.806709265175719, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3635 + }, + { + "epoch": 5.8083067092651754, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3636 + }, + { + "epoch": 5.8099041533546325, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3637 + }, + { + "epoch": 5.81150159744409, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3638 + }, + { + "epoch": 5.813099041533547, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3639 + }, + { + "epoch": 5.814696485623003, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3640 + }, + { + "epoch": 5.81629392971246, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3641 + }, + { + "epoch": 5.817891373801917, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 3642 + }, + { + "epoch": 5.819488817891374, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 3643 + }, + { + "epoch": 5.821086261980831, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3644 + }, + { + "epoch": 5.822683706070287, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3645 + }, + { + "epoch": 5.824281150159744, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3646 + }, + { + "epoch": 5.825878594249201, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3647 + }, + { + "epoch": 5.827476038338658, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 3648 + }, + { + "epoch": 5.8290734824281145, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3649 + }, + { + "epoch": 5.830670926517572, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3650 + }, + { + "epoch": 5.832268370607029, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3651 + }, + { + "epoch": 5.833865814696486, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 3652 + }, + { + "epoch": 5.835463258785943, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 3653 + }, + { + "epoch": 5.8370607028754, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3654 + }, + { + "epoch": 5.838658146964856, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3655 + }, + { + "epoch": 5.840255591054313, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3656 + }, + { + "epoch": 5.84185303514377, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3657 + }, + { + "epoch": 5.843450479233227, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3658 + }, + { + "epoch": 5.845047923322683, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3659 + }, + { + "epoch": 5.84664536741214, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3660 + }, + { + "epoch": 5.848242811501597, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3661 + }, + { + "epoch": 5.8498402555910545, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3662 + }, + { + "epoch": 5.8514376996805115, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 3663 + }, + { + "epoch": 5.853035143769968, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 3664 + }, + { + "epoch": 5.854632587859425, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3665 + }, + { + "epoch": 5.856230031948882, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3666 + }, + { + "epoch": 5.857827476038339, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3667 + }, + { + "epoch": 5.859424920127795, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3668 + }, + { + "epoch": 5.861022364217252, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3669 + }, + { + "epoch": 5.862619808306709, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3670 + }, + { + "epoch": 5.864217252396166, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 3671 + }, + { + "epoch": 5.865814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3672 + }, + { + "epoch": 5.86741214057508, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3673 + }, + { + "epoch": 5.8690095846645365, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3674 + }, + { + "epoch": 5.8706070287539935, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 3675 + }, + { + "epoch": 5.872204472843451, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 3676 + }, + { + "epoch": 5.873801916932908, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3677 + }, + { + "epoch": 5.875399361022364, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3678 + }, + { + "epoch": 5.876996805111821, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3679 + }, + { + "epoch": 5.878594249201278, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3680 + }, + { + "epoch": 5.880191693290735, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3681 + }, + { + "epoch": 5.881789137380192, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3682 + }, + { + "epoch": 5.883386581469648, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3683 + }, + { + "epoch": 5.884984025559105, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3684 + }, + { + "epoch": 5.886581469648562, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3685 + }, + { + "epoch": 5.888178913738019, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3686 + }, + { + "epoch": 5.8897763578274756, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3687 + }, + { + "epoch": 5.891373801916933, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3688 + }, + { + "epoch": 5.89297124600639, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3689 + }, + { + "epoch": 5.894568690095847, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3690 + }, + { + "epoch": 5.896166134185304, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3691 + }, + { + "epoch": 5.897763578274761, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 3692 + }, + { + "epoch": 5.899361022364217, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3693 + }, + { + "epoch": 5.900958466453674, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 3694 + }, + { + "epoch": 5.902555910543131, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 3695 + }, + { + "epoch": 5.904153354632588, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 3696 + }, + { + "epoch": 5.905750798722044, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3697 + }, + { + "epoch": 5.907348242811501, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3698 + }, + { + "epoch": 5.9089456869009584, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3699 + }, + { + "epoch": 5.9105431309904155, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3700 + }, + { + "epoch": 5.912140575079873, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3701 + }, + { + "epoch": 5.913738019169329, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3702 + }, + { + "epoch": 5.915335463258786, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 3703 + }, + { + "epoch": 5.916932907348243, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3704 + }, + { + "epoch": 5.9185303514377, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3705 + }, + { + "epoch": 5.920127795527156, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3706 + }, + { + "epoch": 5.921725239616613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3707 + }, + { + "epoch": 5.92332268370607, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3708 + }, + { + "epoch": 5.924920127795527, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3709 + }, + { + "epoch": 5.926517571884984, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 3710 + }, + { + "epoch": 5.928115015974441, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3711 + }, + { + "epoch": 5.9297124600638975, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 3712 + }, + { + "epoch": 5.931309904153355, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3713 + }, + { + "epoch": 5.932907348242812, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3714 + }, + { + "epoch": 5.934504792332269, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3715 + }, + { + "epoch": 5.936102236421725, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3716 + }, + { + "epoch": 5.937699680511182, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3717 + }, + { + "epoch": 5.939297124600639, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3718 + }, + { + "epoch": 5.940894568690096, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3719 + }, + { + "epoch": 5.942492012779553, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3720 + }, + { + "epoch": 5.944089456869009, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3721 + }, + { + "epoch": 5.945686900958466, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3722 + }, + { + "epoch": 5.947284345047923, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3723 + }, + { + "epoch": 5.94888178913738, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3724 + }, + { + "epoch": 5.950479233226837, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 3725 + }, + { + "epoch": 5.952076677316294, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3726 + }, + { + "epoch": 5.953674121405751, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3727 + }, + { + "epoch": 5.955271565495208, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 3728 + }, + { + "epoch": 5.956869009584665, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3729 + }, + { + "epoch": 5.958466453674122, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3730 + }, + { + "epoch": 5.960063897763578, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3731 + }, + { + "epoch": 5.961661341853035, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3732 + }, + { + "epoch": 5.963258785942492, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 3733 + }, + { + "epoch": 5.964856230031949, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3734 + }, + { + "epoch": 5.966453674121405, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 3735 + }, + { + "epoch": 5.968051118210862, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 3736 + }, + { + "epoch": 5.9696485623003195, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3737 + }, + { + "epoch": 5.9712460063897765, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3738 + }, + { + "epoch": 5.972843450479234, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3739 + }, + { + "epoch": 5.97444089456869, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3740 + }, + { + "epoch": 5.976038338658147, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3741 + }, + { + "epoch": 5.977635782747604, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3742 + }, + { + "epoch": 5.979233226837061, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3743 + }, + { + "epoch": 5.980830670926517, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3744 + }, + { + "epoch": 5.982428115015974, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3745 + }, + { + "epoch": 5.984025559105431, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 3746 + }, + { + "epoch": 5.985623003194888, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3747 + }, + { + "epoch": 5.987220447284345, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3748 + }, + { + "epoch": 5.988817891373802, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3749 + }, + { + "epoch": 5.9904153354632586, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3750 + }, + { + "epoch": 5.992012779552716, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3751 + }, + { + "epoch": 5.993610223642173, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3752 + }, + { + "epoch": 5.99520766773163, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3753 + }, + { + "epoch": 5.996805111821086, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3754 + }, + { + "epoch": 5.998402555910543, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3755 + }, + { + "epoch": 6.0, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3756 + }, + { + "epoch": 6.001597444089457, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3757 + }, + { + "epoch": 6.003194888178914, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3758 + }, + { + "epoch": 6.00479233226837, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3759 + }, + { + "epoch": 6.006389776357827, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 3760 + }, + { + "epoch": 6.007987220447284, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3761 + }, + { + "epoch": 6.0095846645367414, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3762 + }, + { + "epoch": 6.0111821086261985, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3763 + }, + { + "epoch": 6.012779552715655, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3764 + }, + { + "epoch": 6.014376996805112, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3765 + }, + { + "epoch": 6.015974440894569, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3766 + }, + { + "epoch": 6.017571884984026, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3767 + }, + { + "epoch": 6.019169329073482, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3768 + }, + { + "epoch": 6.020766773162939, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3769 + }, + { + "epoch": 6.022364217252396, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 3770 + }, + { + "epoch": 6.023961661341853, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3771 + }, + { + "epoch": 6.02555910543131, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3772 + }, + { + "epoch": 6.027156549520766, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3773 + }, + { + "epoch": 6.0287539936102235, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3774 + }, + { + "epoch": 6.0303514376996805, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3775 + }, + { + "epoch": 6.031948881789138, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3776 + }, + { + "epoch": 6.033546325878595, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3777 + }, + { + "epoch": 6.035143769968051, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3778 + }, + { + "epoch": 6.036741214057508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3779 + }, + { + "epoch": 6.038338658146965, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3780 + }, + { + "epoch": 6.039936102236422, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3781 + }, + { + "epoch": 6.041533546325879, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 3782 + }, + { + "epoch": 6.043130990415335, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 3783 + }, + { + "epoch": 6.044728434504792, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3784 + }, + { + "epoch": 6.046325878594249, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3785 + }, + { + "epoch": 6.047923322683706, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3786 + }, + { + "epoch": 6.0495207667731625, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3787 + }, + { + "epoch": 6.05111821086262, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3788 + }, + { + "epoch": 6.052715654952077, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3789 + }, + { + "epoch": 6.054313099041534, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3790 + }, + { + "epoch": 6.055910543130991, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3791 + }, + { + "epoch": 6.057507987220447, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 3792 + }, + { + "epoch": 6.059105431309904, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3793 + }, + { + "epoch": 6.060702875399361, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3794 + }, + { + "epoch": 6.062300319488818, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 3795 + }, + { + "epoch": 6.063897763578275, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3796 + }, + { + "epoch": 6.065495207667731, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3797 + }, + { + "epoch": 6.067092651757188, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3798 + }, + { + "epoch": 6.068690095846645, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3799 + }, + { + "epoch": 6.0702875399361025, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3800 + }, + { + "epoch": 6.0718849840255595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 3801 + }, + { + "epoch": 6.073482428115016, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3802 + }, + { + "epoch": 6.075079872204473, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3803 + }, + { + "epoch": 6.07667731629393, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3804 + }, + { + "epoch": 6.078274760383387, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3805 + }, + { + "epoch": 6.079872204472843, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3806 + }, + { + "epoch": 6.0814696485623, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 3807 + }, + { + "epoch": 6.083067092651757, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3808 + }, + { + "epoch": 6.084664536741214, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3809 + }, + { + "epoch": 6.086261980830671, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3810 + }, + { + "epoch": 6.087859424920127, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3811 + }, + { + "epoch": 6.0894568690095845, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3812 + }, + { + "epoch": 6.0910543130990416, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3813 + }, + { + "epoch": 6.092651757188499, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3814 + }, + { + "epoch": 6.094249201277956, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3815 + }, + { + "epoch": 6.095846645367412, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3816 + }, + { + "epoch": 6.097444089456869, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3817 + }, + { + "epoch": 6.099041533546326, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3818 + }, + { + "epoch": 6.100638977635783, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 3819 + }, + { + "epoch": 6.102236421725239, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 3820 + }, + { + "epoch": 6.103833865814696, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3821 + }, + { + "epoch": 6.105431309904153, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3822 + }, + { + "epoch": 6.10702875399361, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3823 + }, + { + "epoch": 6.108626198083067, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 3824 + }, + { + "epoch": 6.110223642172524, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3825 + }, + { + "epoch": 6.111821086261981, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3826 + }, + { + "epoch": 6.113418530351438, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3827 + }, + { + "epoch": 6.115015974440895, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3828 + }, + { + "epoch": 6.116613418530352, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3829 + }, + { + "epoch": 6.118210862619808, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3830 + }, + { + "epoch": 6.119808306709265, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3831 + }, + { + "epoch": 6.121405750798722, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3832 + }, + { + "epoch": 6.123003194888179, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3833 + }, + { + "epoch": 6.124600638977636, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3834 + }, + { + "epoch": 6.126198083067092, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3835 + }, + { + "epoch": 6.127795527156549, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3836 + }, + { + "epoch": 6.1293929712460065, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3837 + }, + { + "epoch": 6.1309904153354635, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3838 + }, + { + "epoch": 6.13258785942492, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3839 + }, + { + "epoch": 6.134185303514377, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3840 + }, + { + "epoch": 6.135782747603834, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3841 + }, + { + "epoch": 6.137380191693291, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3842 + }, + { + "epoch": 6.138977635782748, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3843 + }, + { + "epoch": 6.140575079872204, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3844 + }, + { + "epoch": 6.142172523961661, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 3845 + }, + { + "epoch": 6.143769968051118, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3846 + }, + { + "epoch": 6.145367412140575, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3847 + }, + { + "epoch": 6.146964856230032, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3848 + }, + { + "epoch": 6.1485623003194885, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3849 + }, + { + "epoch": 6.1501597444089455, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 3850 + }, + { + "epoch": 6.151757188498403, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3851 + }, + { + "epoch": 6.15335463258786, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 3852 + }, + { + "epoch": 6.154952076677317, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 3853 + }, + { + "epoch": 6.156549520766773, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3854 + }, + { + "epoch": 6.15814696485623, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3855 + }, + { + "epoch": 6.159744408945687, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3856 + }, + { + "epoch": 6.161341853035144, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3857 + }, + { + "epoch": 6.1629392971246, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3858 + }, + { + "epoch": 6.164536741214057, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 3859 + }, + { + "epoch": 6.166134185303514, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3860 + }, + { + "epoch": 6.167731629392971, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3861 + }, + { + "epoch": 6.169329073482428, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 3862 + }, + { + "epoch": 6.170926517571885, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3863 + }, + { + "epoch": 6.172523961661342, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3864 + }, + { + "epoch": 6.174121405750799, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3865 + }, + { + "epoch": 6.175718849840256, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3866 + }, + { + "epoch": 6.177316293929713, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3867 + }, + { + "epoch": 6.178913738019169, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 3868 + }, + { + "epoch": 6.180511182108626, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3869 + }, + { + "epoch": 6.182108626198083, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 3870 + }, + { + "epoch": 6.18370607028754, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3871 + }, + { + "epoch": 6.185303514376997, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 3872 + }, + { + "epoch": 6.186900958466453, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3873 + }, + { + "epoch": 6.18849840255591, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3874 + }, + { + "epoch": 6.1900958466453675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 3875 + }, + { + "epoch": 6.1916932907348246, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3876 + }, + { + "epoch": 6.193290734824281, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3877 + }, + { + "epoch": 6.194888178913738, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3878 + }, + { + "epoch": 6.196485623003195, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3879 + }, + { + "epoch": 6.198083067092652, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3880 + }, + { + "epoch": 6.199680511182109, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3881 + }, + { + "epoch": 6.201277955271565, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3882 + }, + { + "epoch": 6.202875399361022, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 3883 + }, + { + "epoch": 6.204472843450479, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3884 + }, + { + "epoch": 6.206070287539936, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3885 + }, + { + "epoch": 6.207667731629393, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 3886 + }, + { + "epoch": 6.2092651757188495, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3887 + }, + { + "epoch": 6.210862619808307, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3888 + }, + { + "epoch": 6.212460063897764, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3889 + }, + { + "epoch": 6.214057507987221, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3890 + }, + { + "epoch": 6.215654952076678, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 3891 + }, + { + "epoch": 6.217252396166134, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3892 + }, + { + "epoch": 6.218849840255591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3893 + }, + { + "epoch": 6.220447284345048, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3894 + }, + { + "epoch": 6.222044728434505, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3895 + }, + { + "epoch": 6.223642172523961, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3896 + }, + { + "epoch": 6.225239616613418, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3897 + }, + { + "epoch": 6.226837060702875, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3898 + }, + { + "epoch": 6.228434504792332, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3899 + }, + { + "epoch": 6.2300319488817895, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3900 + }, + { + "epoch": 6.231629392971246, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3901 + }, + { + "epoch": 6.233226837060703, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3902 + }, + { + "epoch": 6.23482428115016, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 3903 + }, + { + "epoch": 6.236421725239617, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3904 + }, + { + "epoch": 6.238019169329074, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 3905 + }, + { + "epoch": 6.23961661341853, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 3906 + }, + { + "epoch": 6.241214057507987, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3907 + }, + { + "epoch": 6.242811501597444, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3908 + }, + { + "epoch": 6.244408945686901, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3909 + }, + { + "epoch": 6.246006389776358, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3910 + }, + { + "epoch": 6.247603833865814, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 3911 + }, + { + "epoch": 6.2492012779552715, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3912 + }, + { + "epoch": 6.2507987220447285, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3913 + }, + { + "epoch": 6.252396166134186, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 3914 + }, + { + "epoch": 6.253993610223642, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3915 + }, + { + "epoch": 6.255591054313099, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3916 + }, + { + "epoch": 6.257188498402556, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3917 + }, + { + "epoch": 6.258785942492013, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 3918 + }, + { + "epoch": 6.26038338658147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3919 + }, + { + "epoch": 6.261980830670926, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3920 + }, + { + "epoch": 6.263578274760383, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3921 + }, + { + "epoch": 6.26517571884984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3922 + }, + { + "epoch": 6.266773162939297, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3923 + }, + { + "epoch": 6.268370607028754, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 3924 + }, + { + "epoch": 6.2699680511182105, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3925 + }, + { + "epoch": 6.271565495207668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3926 + }, + { + "epoch": 6.273162939297125, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3927 + }, + { + "epoch": 6.274760383386582, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3928 + }, + { + "epoch": 6.276357827476039, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3929 + }, + { + "epoch": 6.277955271565495, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 3930 + }, + { + "epoch": 6.279552715654952, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3931 + }, + { + "epoch": 6.281150159744409, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 3932 + }, + { + "epoch": 6.282747603833866, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3933 + }, + { + "epoch": 6.284345047923322, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3934 + }, + { + "epoch": 6.285942492012779, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3935 + }, + { + "epoch": 6.287539936102236, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3936 + }, + { + "epoch": 6.289137380191693, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3937 + }, + { + "epoch": 6.2907348242811505, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3938 + }, + { + "epoch": 6.292332268370607, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3939 + }, + { + "epoch": 6.293929712460064, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3940 + }, + { + "epoch": 6.295527156549521, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3941 + }, + { + "epoch": 6.297124600638978, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3942 + }, + { + "epoch": 6.298722044728435, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3943 + }, + { + "epoch": 6.300319488817891, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3944 + }, + { + "epoch": 6.301916932907348, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3945 + }, + { + "epoch": 6.303514376996805, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3946 + }, + { + "epoch": 6.305111821086262, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3947 + }, + { + "epoch": 6.306709265175719, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3948 + }, + { + "epoch": 6.3083067092651754, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3949 + }, + { + "epoch": 6.3099041533546325, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3950 + }, + { + "epoch": 6.31150159744409, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 3951 + }, + { + "epoch": 6.313099041533547, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3952 + }, + { + "epoch": 6.314696485623003, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3953 + }, + { + "epoch": 6.31629392971246, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3954 + }, + { + "epoch": 6.317891373801917, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3955 + }, + { + "epoch": 6.319488817891374, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3956 + }, + { + "epoch": 6.321086261980831, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3957 + }, + { + "epoch": 6.322683706070287, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3958 + }, + { + "epoch": 6.324281150159744, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 3959 + }, + { + "epoch": 6.325878594249201, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3960 + }, + { + "epoch": 6.327476038338658, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3961 + }, + { + "epoch": 6.329073482428115, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3962 + }, + { + "epoch": 6.330670926517572, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3963 + }, + { + "epoch": 6.332268370607029, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 3964 + }, + { + "epoch": 6.333865814696486, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3965 + }, + { + "epoch": 6.335463258785943, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3966 + }, + { + "epoch": 6.3370607028754, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 3967 + }, + { + "epoch": 6.338658146964856, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3968 + }, + { + "epoch": 6.340255591054313, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3969 + }, + { + "epoch": 6.34185303514377, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3970 + }, + { + "epoch": 6.343450479233227, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3971 + }, + { + "epoch": 6.345047923322683, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 3972 + }, + { + "epoch": 6.34664536741214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3973 + }, + { + "epoch": 6.348242811501597, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 3974 + }, + { + "epoch": 6.3498402555910545, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3975 + }, + { + "epoch": 6.3514376996805115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3976 + }, + { + "epoch": 6.353035143769968, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3977 + }, + { + "epoch": 6.354632587859425, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3978 + }, + { + "epoch": 6.356230031948882, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3979 + }, + { + "epoch": 6.357827476038339, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3980 + }, + { + "epoch": 6.359424920127796, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3981 + }, + { + "epoch": 6.361022364217252, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3982 + }, + { + "epoch": 6.362619808306709, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3983 + }, + { + "epoch": 6.364217252396166, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3984 + }, + { + "epoch": 6.365814696485623, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3985 + }, + { + "epoch": 6.36741214057508, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 3986 + }, + { + "epoch": 6.3690095846645365, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 3987 + }, + { + "epoch": 6.3706070287539935, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3988 + }, + { + "epoch": 6.372204472843451, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 3989 + }, + { + "epoch": 6.373801916932908, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3990 + }, + { + "epoch": 6.375399361022364, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3991 + }, + { + "epoch": 6.376996805111821, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3992 + }, + { + "epoch": 6.378594249201278, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3993 + }, + { + "epoch": 6.380191693290735, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3994 + }, + { + "epoch": 6.381789137380192, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 3995 + }, + { + "epoch": 6.383386581469648, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 3996 + }, + { + "epoch": 6.384984025559105, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 3997 + }, + { + "epoch": 6.386581469648562, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3998 + }, + { + "epoch": 6.388178913738019, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 3999 + }, + { + "epoch": 6.389776357827476, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4000 + }, + { + "epoch": 6.391373801916933, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4001 + }, + { + "epoch": 6.39297124600639, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4002 + }, + { + "epoch": 6.394568690095847, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4003 + }, + { + "epoch": 6.396166134185304, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 4004 + }, + { + "epoch": 6.397763578274761, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4005 + }, + { + "epoch": 6.399361022364217, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4006 + }, + { + "epoch": 6.400958466453674, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4007 + }, + { + "epoch": 6.402555910543131, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4008 + }, + { + "epoch": 6.404153354632588, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4009 + }, + { + "epoch": 6.405750798722044, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4010 + }, + { + "epoch": 6.407348242811501, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 4011 + }, + { + "epoch": 6.4089456869009584, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4012 + }, + { + "epoch": 6.4105431309904155, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4013 + }, + { + "epoch": 6.412140575079873, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 4014 + }, + { + "epoch": 6.413738019169329, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4015 + }, + { + "epoch": 6.415335463258786, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4016 + }, + { + "epoch": 6.416932907348243, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4017 + }, + { + "epoch": 6.4185303514377, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 4018 + }, + { + "epoch": 6.420127795527157, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4019 + }, + { + "epoch": 6.421725239616613, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4020 + }, + { + "epoch": 6.42332268370607, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4021 + }, + { + "epoch": 6.424920127795527, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 4022 + }, + { + "epoch": 6.426517571884984, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4023 + }, + { + "epoch": 6.428115015974441, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4024 + }, + { + "epoch": 6.4297124600638975, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4025 + }, + { + "epoch": 6.431309904153355, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4026 + }, + { + "epoch": 6.432907348242812, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4027 + }, + { + "epoch": 6.434504792332269, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4028 + }, + { + "epoch": 6.436102236421725, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4029 + }, + { + "epoch": 6.437699680511182, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4030 + }, + { + "epoch": 6.439297124600639, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4031 + }, + { + "epoch": 6.440894568690096, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4032 + }, + { + "epoch": 6.442492012779553, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4033 + }, + { + "epoch": 6.444089456869009, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4034 + }, + { + "epoch": 6.445686900958466, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 4035 + }, + { + "epoch": 6.447284345047923, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4036 + }, + { + "epoch": 6.44888178913738, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4037 + }, + { + "epoch": 6.4504792332268375, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4038 + }, + { + "epoch": 6.452076677316294, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4039 + }, + { + "epoch": 6.453674121405751, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 4040 + }, + { + "epoch": 6.455271565495208, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4041 + }, + { + "epoch": 6.456869009584665, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4042 + }, + { + "epoch": 6.458466453674122, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4043 + }, + { + "epoch": 6.460063897763578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4044 + }, + { + "epoch": 6.461661341853035, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4045 + }, + { + "epoch": 6.463258785942492, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4046 + }, + { + "epoch": 6.464856230031949, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4047 + }, + { + "epoch": 6.466453674121405, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4048 + }, + { + "epoch": 6.468051118210862, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4049 + }, + { + "epoch": 6.4696485623003195, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4050 + }, + { + "epoch": 6.4712460063897765, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4051 + }, + { + "epoch": 6.472843450479234, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4052 + }, + { + "epoch": 6.47444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4053 + }, + { + "epoch": 6.476038338658147, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 4054 + }, + { + "epoch": 6.477635782747604, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4055 + }, + { + "epoch": 6.479233226837061, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4056 + }, + { + "epoch": 6.480830670926517, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4057 + }, + { + "epoch": 6.482428115015974, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4058 + }, + { + "epoch": 6.484025559105431, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4059 + }, + { + "epoch": 6.485623003194888, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4060 + }, + { + "epoch": 6.487220447284345, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4061 + }, + { + "epoch": 6.488817891373802, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4062 + }, + { + "epoch": 6.4904153354632586, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4063 + }, + { + "epoch": 6.492012779552716, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4064 + }, + { + "epoch": 6.493610223642173, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4065 + }, + { + "epoch": 6.49520766773163, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4066 + }, + { + "epoch": 6.496805111821086, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4067 + }, + { + "epoch": 6.498402555910543, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4068 + }, + { + "epoch": 6.5, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 4069 + }, + { + "epoch": 6.501597444089457, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4070 + }, + { + "epoch": 6.503194888178914, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4071 + }, + { + "epoch": 6.50479233226837, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4072 + }, + { + "epoch": 6.506389776357827, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 4073 + }, + { + "epoch": 6.507987220447284, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4074 + }, + { + "epoch": 6.5095846645367414, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4075 + }, + { + "epoch": 6.511182108626198, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4076 + }, + { + "epoch": 6.512779552715655, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4077 + }, + { + "epoch": 6.514376996805112, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 4078 + }, + { + "epoch": 6.515974440894569, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4079 + }, + { + "epoch": 6.517571884984026, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4080 + }, + { + "epoch": 6.519169329073483, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4081 + }, + { + "epoch": 6.520766773162939, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4082 + }, + { + "epoch": 6.522364217252396, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4083 + }, + { + "epoch": 6.523961661341853, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4084 + }, + { + "epoch": 6.52555910543131, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4085 + }, + { + "epoch": 6.527156549520766, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4086 + }, + { + "epoch": 6.5287539936102235, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4087 + }, + { + "epoch": 6.5303514376996805, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4088 + }, + { + "epoch": 6.531948881789138, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4089 + }, + { + "epoch": 6.533546325878595, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 4090 + }, + { + "epoch": 6.535143769968051, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4091 + }, + { + "epoch": 6.536741214057508, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4092 + }, + { + "epoch": 6.538338658146965, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4093 + }, + { + "epoch": 6.539936102236422, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4094 + }, + { + "epoch": 6.541533546325878, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4095 + }, + { + "epoch": 6.543130990415335, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4096 + }, + { + "epoch": 6.544728434504792, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4097 + }, + { + "epoch": 6.546325878594249, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4098 + }, + { + "epoch": 6.547923322683706, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4099 + }, + { + "epoch": 6.549520766773163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4100 + }, + { + "epoch": 6.55111821086262, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4101 + }, + { + "epoch": 6.552715654952077, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4102 + }, + { + "epoch": 6.554313099041534, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4103 + }, + { + "epoch": 6.555910543130991, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 4104 + }, + { + "epoch": 6.557507987220447, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 4105 + }, + { + "epoch": 6.559105431309904, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4106 + }, + { + "epoch": 6.560702875399361, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4107 + }, + { + "epoch": 6.562300319488818, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4108 + }, + { + "epoch": 6.563897763578275, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4109 + }, + { + "epoch": 6.565495207667731, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4110 + }, + { + "epoch": 6.567092651757188, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4111 + }, + { + "epoch": 6.568690095846645, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4112 + }, + { + "epoch": 6.5702875399361025, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4113 + }, + { + "epoch": 6.571884984025559, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4114 + }, + { + "epoch": 6.573482428115016, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4115 + }, + { + "epoch": 6.575079872204473, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4116 + }, + { + "epoch": 6.57667731629393, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4117 + }, + { + "epoch": 6.578274760383387, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4118 + }, + { + "epoch": 6.579872204472844, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4119 + }, + { + "epoch": 6.5814696485623, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4120 + }, + { + "epoch": 6.583067092651757, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4121 + }, + { + "epoch": 6.584664536741214, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4122 + }, + { + "epoch": 6.586261980830671, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4123 + }, + { + "epoch": 6.587859424920127, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4124 + }, + { + "epoch": 6.5894568690095845, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4125 + }, + { + "epoch": 6.5910543130990416, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4126 + }, + { + "epoch": 6.592651757188499, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 4127 + }, + { + "epoch": 6.594249201277956, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4128 + }, + { + "epoch": 6.595846645367412, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4129 + }, + { + "epoch": 6.597444089456869, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 4130 + }, + { + "epoch": 6.599041533546326, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 4131 + }, + { + "epoch": 6.600638977635783, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4132 + }, + { + "epoch": 6.602236421725239, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4133 + }, + { + "epoch": 6.603833865814696, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4134 + }, + { + "epoch": 6.605431309904153, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4135 + }, + { + "epoch": 6.60702875399361, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 4136 + }, + { + "epoch": 6.608626198083067, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4137 + }, + { + "epoch": 6.6102236421725244, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4138 + }, + { + "epoch": 6.611821086261981, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 4139 + }, + { + "epoch": 6.613418530351438, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 4140 + }, + { + "epoch": 6.615015974440895, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4141 + }, + { + "epoch": 6.616613418530352, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4142 + }, + { + "epoch": 6.618210862619808, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4143 + }, + { + "epoch": 6.619808306709265, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4144 + }, + { + "epoch": 6.621405750798722, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4145 + }, + { + "epoch": 6.623003194888179, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4146 + }, + { + "epoch": 6.624600638977636, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4147 + }, + { + "epoch": 6.626198083067092, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4148 + }, + { + "epoch": 6.627795527156549, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4149 + }, + { + "epoch": 6.6293929712460065, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4150 + }, + { + "epoch": 6.6309904153354635, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4151 + }, + { + "epoch": 6.63258785942492, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 4152 + }, + { + "epoch": 6.634185303514377, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4153 + }, + { + "epoch": 6.635782747603834, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4154 + }, + { + "epoch": 6.637380191693291, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4155 + }, + { + "epoch": 6.638977635782748, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 4156 + }, + { + "epoch": 6.640575079872205, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4157 + }, + { + "epoch": 6.642172523961661, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4158 + }, + { + "epoch": 6.643769968051118, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4159 + }, + { + "epoch": 6.645367412140575, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4160 + }, + { + "epoch": 6.646964856230032, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4161 + }, + { + "epoch": 6.6485623003194885, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4162 + }, + { + "epoch": 6.6501597444089455, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4163 + }, + { + "epoch": 6.651757188498403, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4164 + }, + { + "epoch": 6.65335463258786, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4165 + }, + { + "epoch": 6.654952076677317, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4166 + }, + { + "epoch": 6.656549520766773, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4167 + }, + { + "epoch": 6.65814696485623, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4168 + }, + { + "epoch": 6.659744408945687, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4169 + }, + { + "epoch": 6.661341853035144, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4170 + }, + { + "epoch": 6.6629392971246, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4171 + }, + { + "epoch": 6.664536741214057, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 4172 + }, + { + "epoch": 6.666134185303514, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4173 + }, + { + "epoch": 6.667731629392971, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4174 + }, + { + "epoch": 6.669329073482428, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4175 + }, + { + "epoch": 6.6709265175718855, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4176 + }, + { + "epoch": 6.672523961661342, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4177 + }, + { + "epoch": 6.674121405750799, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4178 + }, + { + "epoch": 6.675718849840256, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4179 + }, + { + "epoch": 6.677316293929713, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4180 + }, + { + "epoch": 6.678913738019169, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4181 + }, + { + "epoch": 6.680511182108626, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4182 + }, + { + "epoch": 6.682108626198083, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4183 + }, + { + "epoch": 6.68370607028754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 4184 + }, + { + "epoch": 6.685303514376997, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 4185 + }, + { + "epoch": 6.686900958466453, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4186 + }, + { + "epoch": 6.68849840255591, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4187 + }, + { + "epoch": 6.6900958466453675, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4188 + }, + { + "epoch": 6.6916932907348246, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4189 + }, + { + "epoch": 6.693290734824281, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4190 + }, + { + "epoch": 6.694888178913738, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4191 + }, + { + "epoch": 6.696485623003195, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4192 + }, + { + "epoch": 6.698083067092652, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4193 + }, + { + "epoch": 6.699680511182109, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4194 + }, + { + "epoch": 6.701277955271565, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4195 + }, + { + "epoch": 6.702875399361022, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 4196 + }, + { + "epoch": 6.704472843450479, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4197 + }, + { + "epoch": 6.706070287539936, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4198 + }, + { + "epoch": 6.707667731629393, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4199 + }, + { + "epoch": 6.7092651757188495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4200 + }, + { + "epoch": 6.710862619808307, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 4201 + }, + { + "epoch": 6.712460063897764, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4202 + }, + { + "epoch": 6.714057507987221, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4203 + }, + { + "epoch": 6.715654952076678, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4204 + }, + { + "epoch": 6.717252396166134, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4205 + }, + { + "epoch": 6.718849840255591, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4206 + }, + { + "epoch": 6.720447284345048, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4207 + }, + { + "epoch": 6.722044728434505, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4208 + }, + { + "epoch": 6.723642172523961, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4209 + }, + { + "epoch": 6.725239616613418, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4210 + }, + { + "epoch": 6.726837060702875, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4211 + }, + { + "epoch": 6.728434504792332, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4212 + }, + { + "epoch": 6.7300319488817895, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4213 + }, + { + "epoch": 6.731629392971246, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4214 + }, + { + "epoch": 6.733226837060703, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4215 + }, + { + "epoch": 6.73482428115016, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4216 + }, + { + "epoch": 6.736421725239617, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4217 + }, + { + "epoch": 6.738019169329074, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 4218 + }, + { + "epoch": 6.73961661341853, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4219 + }, + { + "epoch": 6.741214057507987, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4220 + }, + { + "epoch": 6.742811501597444, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4221 + }, + { + "epoch": 6.744408945686901, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4222 + }, + { + "epoch": 6.746006389776358, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 4223 + }, + { + "epoch": 6.747603833865814, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4224 + }, + { + "epoch": 6.7492012779552715, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4225 + }, + { + "epoch": 6.7507987220447285, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4226 + }, + { + "epoch": 6.752396166134186, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 4227 + }, + { + "epoch": 6.753993610223642, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 4228 + }, + { + "epoch": 6.755591054313099, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 4229 + }, + { + "epoch": 6.757188498402556, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4230 + }, + { + "epoch": 6.758785942492013, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4231 + }, + { + "epoch": 6.76038338658147, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4232 + }, + { + "epoch": 6.761980830670926, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4233 + }, + { + "epoch": 6.763578274760383, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4234 + }, + { + "epoch": 6.76517571884984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4235 + }, + { + "epoch": 6.766773162939297, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4236 + }, + { + "epoch": 6.768370607028754, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 4237 + }, + { + "epoch": 6.7699680511182105, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4238 + }, + { + "epoch": 6.771565495207668, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4239 + }, + { + "epoch": 6.773162939297125, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4240 + }, + { + "epoch": 6.774760383386582, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 4241 + }, + { + "epoch": 6.776357827476039, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4242 + }, + { + "epoch": 6.777955271565495, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4243 + }, + { + "epoch": 6.779552715654952, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4244 + }, + { + "epoch": 6.781150159744409, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4245 + }, + { + "epoch": 6.782747603833866, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4246 + }, + { + "epoch": 6.784345047923322, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4247 + }, + { + "epoch": 6.785942492012779, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4248 + }, + { + "epoch": 6.787539936102236, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4249 + }, + { + "epoch": 6.789137380191693, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4250 + }, + { + "epoch": 6.7907348242811505, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4251 + }, + { + "epoch": 6.792332268370607, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4252 + }, + { + "epoch": 6.793929712460064, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4253 + }, + { + "epoch": 6.795527156549521, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4254 + }, + { + "epoch": 6.797124600638978, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4255 + }, + { + "epoch": 6.798722044728435, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4256 + }, + { + "epoch": 6.800319488817891, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4257 + }, + { + "epoch": 6.801916932907348, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4258 + }, + { + "epoch": 6.803514376996805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4259 + }, + { + "epoch": 6.805111821086262, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4260 + }, + { + "epoch": 6.806709265175719, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 4261 + }, + { + "epoch": 6.8083067092651754, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4262 + }, + { + "epoch": 6.8099041533546325, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4263 + }, + { + "epoch": 6.81150159744409, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4264 + }, + { + "epoch": 6.813099041533547, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4265 + }, + { + "epoch": 6.814696485623003, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4266 + }, + { + "epoch": 6.81629392971246, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4267 + }, + { + "epoch": 6.817891373801917, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4268 + }, + { + "epoch": 6.819488817891374, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4269 + }, + { + "epoch": 6.821086261980831, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 4270 + }, + { + "epoch": 6.822683706070287, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4271 + }, + { + "epoch": 6.824281150159744, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4272 + }, + { + "epoch": 6.825878594249201, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4273 + }, + { + "epoch": 6.827476038338658, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4274 + }, + { + "epoch": 6.8290734824281145, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4275 + }, + { + "epoch": 6.830670926517572, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4276 + }, + { + "epoch": 6.832268370607029, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4277 + }, + { + "epoch": 6.833865814696486, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4278 + }, + { + "epoch": 6.835463258785943, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4279 + }, + { + "epoch": 6.8370607028754, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4280 + }, + { + "epoch": 6.838658146964856, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4281 + }, + { + "epoch": 6.840255591054313, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4282 + }, + { + "epoch": 6.84185303514377, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4283 + }, + { + "epoch": 6.843450479233227, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4284 + }, + { + "epoch": 6.845047923322683, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4285 + }, + { + "epoch": 6.84664536741214, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4286 + }, + { + "epoch": 6.848242811501597, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4287 + }, + { + "epoch": 6.8498402555910545, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4288 + }, + { + "epoch": 6.8514376996805115, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4289 + }, + { + "epoch": 6.853035143769968, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4290 + }, + { + "epoch": 6.854632587859425, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4291 + }, + { + "epoch": 6.856230031948882, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4292 + }, + { + "epoch": 6.857827476038339, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4293 + }, + { + "epoch": 6.859424920127795, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4294 + }, + { + "epoch": 6.861022364217252, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4295 + }, + { + "epoch": 6.862619808306709, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4296 + }, + { + "epoch": 6.864217252396166, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4297 + }, + { + "epoch": 6.865814696485623, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4298 + }, + { + "epoch": 6.86741214057508, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4299 + }, + { + "epoch": 6.8690095846645365, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4300 + }, + { + "epoch": 6.8706070287539935, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4301 + }, + { + "epoch": 6.872204472843451, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 4302 + }, + { + "epoch": 6.873801916932908, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 4303 + }, + { + "epoch": 6.875399361022364, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4304 + }, + { + "epoch": 6.876996805111821, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4305 + }, + { + "epoch": 6.878594249201278, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4306 + }, + { + "epoch": 6.880191693290735, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4307 + }, + { + "epoch": 6.881789137380192, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4308 + }, + { + "epoch": 6.883386581469648, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4309 + }, + { + "epoch": 6.884984025559105, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4310 + }, + { + "epoch": 6.886581469648562, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4311 + }, + { + "epoch": 6.888178913738019, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4312 + }, + { + "epoch": 6.8897763578274756, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4313 + }, + { + "epoch": 6.891373801916933, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4314 + }, + { + "epoch": 6.89297124600639, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4315 + }, + { + "epoch": 6.894568690095847, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4316 + }, + { + "epoch": 6.896166134185304, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4317 + }, + { + "epoch": 6.897763578274761, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4318 + }, + { + "epoch": 6.899361022364217, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4319 + }, + { + "epoch": 6.900958466453674, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4320 + }, + { + "epoch": 6.902555910543131, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4321 + }, + { + "epoch": 6.904153354632588, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4322 + }, + { + "epoch": 6.905750798722044, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4323 + }, + { + "epoch": 6.907348242811501, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4324 + }, + { + "epoch": 6.9089456869009584, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4325 + }, + { + "epoch": 6.9105431309904155, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4326 + }, + { + "epoch": 6.912140575079873, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4327 + }, + { + "epoch": 6.913738019169329, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 4328 + }, + { + "epoch": 6.915335463258786, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4329 + }, + { + "epoch": 6.916932907348243, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4330 + }, + { + "epoch": 6.9185303514377, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4331 + }, + { + "epoch": 6.920127795527156, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4332 + }, + { + "epoch": 6.921725239616613, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4333 + }, + { + "epoch": 6.92332268370607, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4334 + }, + { + "epoch": 6.924920127795527, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 4335 + }, + { + "epoch": 6.926517571884984, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4336 + }, + { + "epoch": 6.928115015974441, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4337 + }, + { + "epoch": 6.9297124600638975, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4338 + }, + { + "epoch": 6.931309904153355, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4339 + }, + { + "epoch": 6.932907348242812, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4340 + }, + { + "epoch": 6.934504792332269, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4341 + }, + { + "epoch": 6.936102236421725, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4342 + }, + { + "epoch": 6.937699680511182, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4343 + }, + { + "epoch": 6.939297124600639, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4344 + }, + { + "epoch": 6.940894568690096, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4345 + }, + { + "epoch": 6.942492012779553, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 4346 + }, + { + "epoch": 6.944089456869009, + "grad_norm": 0.34375, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4347 + }, + { + "epoch": 6.945686900958466, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4348 + }, + { + "epoch": 6.947284345047923, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4349 + }, + { + "epoch": 6.94888178913738, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4350 + }, + { + "epoch": 6.950479233226837, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4351 + }, + { + "epoch": 6.952076677316294, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 4352 + }, + { + "epoch": 6.953674121405751, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4353 + }, + { + "epoch": 6.955271565495208, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4354 + }, + { + "epoch": 6.956869009584665, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4355 + }, + { + "epoch": 6.958466453674122, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4356 + }, + { + "epoch": 6.960063897763578, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 4357 + }, + { + "epoch": 6.961661341853035, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4358 + }, + { + "epoch": 6.963258785942492, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4359 + }, + { + "epoch": 6.964856230031949, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4360 + }, + { + "epoch": 6.966453674121405, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4361 + }, + { + "epoch": 6.968051118210862, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4362 + }, + { + "epoch": 6.9696485623003195, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4363 + }, + { + "epoch": 6.9712460063897765, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4364 + }, + { + "epoch": 6.972843450479234, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4365 + }, + { + "epoch": 6.97444089456869, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4366 + }, + { + "epoch": 6.976038338658147, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4367 + }, + { + "epoch": 6.977635782747604, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4368 + }, + { + "epoch": 6.979233226837061, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4369 + }, + { + "epoch": 6.980830670926517, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4370 + }, + { + "epoch": 6.982428115015974, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4371 + }, + { + "epoch": 6.984025559105431, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4372 + }, + { + "epoch": 6.985623003194888, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4373 + }, + { + "epoch": 6.987220447284345, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4374 + }, + { + "epoch": 6.988817891373802, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4375 + }, + { + "epoch": 6.9904153354632586, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4376 + }, + { + "epoch": 6.992012779552716, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4377 + }, + { + "epoch": 6.993610223642173, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4378 + }, + { + "epoch": 6.99520766773163, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4379 + }, + { + "epoch": 6.996805111821086, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4380 + }, + { + "epoch": 6.998402555910543, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4381 + }, + { + "epoch": 7.0, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4382 + }, + { + "epoch": 7.001597444089457, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4383 + }, + { + "epoch": 7.003194888178914, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 4384 + }, + { + "epoch": 7.00479233226837, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4385 + }, + { + "epoch": 7.006389776357827, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4386 + }, + { + "epoch": 7.007987220447284, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4387 + }, + { + "epoch": 7.0095846645367414, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4388 + }, + { + "epoch": 7.0111821086261985, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4389 + }, + { + "epoch": 7.012779552715655, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4390 + }, + { + "epoch": 7.014376996805112, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4391 + }, + { + "epoch": 7.015974440894569, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4392 + }, + { + "epoch": 7.017571884984026, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4393 + }, + { + "epoch": 7.019169329073482, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4394 + }, + { + "epoch": 7.020766773162939, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4395 + }, + { + "epoch": 7.022364217252396, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 4396 + }, + { + "epoch": 7.023961661341853, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4397 + }, + { + "epoch": 7.02555910543131, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4398 + }, + { + "epoch": 7.027156549520766, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4399 + }, + { + "epoch": 7.0287539936102235, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4400 + }, + { + "epoch": 7.0303514376996805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4401 + }, + { + "epoch": 7.031948881789138, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4402 + }, + { + "epoch": 7.033546325878595, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4403 + }, + { + "epoch": 7.035143769968051, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4404 + }, + { + "epoch": 7.036741214057508, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4405 + }, + { + "epoch": 7.038338658146965, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4406 + }, + { + "epoch": 7.039936102236422, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4407 + }, + { + "epoch": 7.041533546325879, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 4408 + }, + { + "epoch": 7.043130990415335, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4409 + }, + { + "epoch": 7.044728434504792, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4410 + }, + { + "epoch": 7.046325878594249, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4411 + }, + { + "epoch": 7.047923322683706, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4412 + }, + { + "epoch": 7.0495207667731625, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4413 + }, + { + "epoch": 7.05111821086262, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4414 + }, + { + "epoch": 7.052715654952077, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4415 + }, + { + "epoch": 7.054313099041534, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4416 + }, + { + "epoch": 7.055910543130991, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4417 + }, + { + "epoch": 7.057507987220447, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4418 + }, + { + "epoch": 7.059105431309904, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4419 + }, + { + "epoch": 7.060702875399361, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4420 + }, + { + "epoch": 7.062300319488818, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4421 + }, + { + "epoch": 7.063897763578275, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4422 + }, + { + "epoch": 7.065495207667731, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4423 + }, + { + "epoch": 7.067092651757188, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4424 + }, + { + "epoch": 7.068690095846645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 4425 + }, + { + "epoch": 7.0702875399361025, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4426 + }, + { + "epoch": 7.0718849840255595, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4427 + }, + { + "epoch": 7.073482428115016, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 4428 + }, + { + "epoch": 7.075079872204473, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4429 + }, + { + "epoch": 7.07667731629393, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4430 + }, + { + "epoch": 7.078274760383387, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4431 + }, + { + "epoch": 7.079872204472843, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4432 + }, + { + "epoch": 7.0814696485623, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4433 + }, + { + "epoch": 7.083067092651757, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4434 + }, + { + "epoch": 7.084664536741214, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 4435 + }, + { + "epoch": 7.086261980830671, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4436 + }, + { + "epoch": 7.087859424920127, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4437 + }, + { + "epoch": 7.0894568690095845, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4438 + }, + { + "epoch": 7.0910543130990416, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4439 + }, + { + "epoch": 7.092651757188499, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 4440 + }, + { + "epoch": 7.094249201277956, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4441 + }, + { + "epoch": 7.095846645367412, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4442 + }, + { + "epoch": 7.097444089456869, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4443 + }, + { + "epoch": 7.099041533546326, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4444 + }, + { + "epoch": 7.100638977635783, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4445 + }, + { + "epoch": 7.102236421725239, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4446 + }, + { + "epoch": 7.103833865814696, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4447 + }, + { + "epoch": 7.105431309904153, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4448 + }, + { + "epoch": 7.10702875399361, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4449 + }, + { + "epoch": 7.108626198083067, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4450 + }, + { + "epoch": 7.110223642172524, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4451 + }, + { + "epoch": 7.111821086261981, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4452 + }, + { + "epoch": 7.113418530351438, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4453 + }, + { + "epoch": 7.115015974440895, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4454 + }, + { + "epoch": 7.116613418530352, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4455 + }, + { + "epoch": 7.118210862619808, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4456 + }, + { + "epoch": 7.119808306709265, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4457 + }, + { + "epoch": 7.121405750798722, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4458 + }, + { + "epoch": 7.123003194888179, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4459 + }, + { + "epoch": 7.124600638977636, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4460 + }, + { + "epoch": 7.126198083067092, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4461 + }, + { + "epoch": 7.127795527156549, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4462 + }, + { + "epoch": 7.1293929712460065, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4463 + }, + { + "epoch": 7.1309904153354635, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4464 + }, + { + "epoch": 7.13258785942492, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4465 + }, + { + "epoch": 7.134185303514377, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4466 + }, + { + "epoch": 7.135782747603834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4467 + }, + { + "epoch": 7.137380191693291, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4468 + }, + { + "epoch": 7.138977635782748, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4469 + }, + { + "epoch": 7.140575079872204, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 4470 + }, + { + "epoch": 7.142172523961661, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4471 + }, + { + "epoch": 7.143769968051118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4472 + }, + { + "epoch": 7.145367412140575, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4473 + }, + { + "epoch": 7.146964856230032, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4474 + }, + { + "epoch": 7.1485623003194885, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4475 + }, + { + "epoch": 7.1501597444089455, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4476 + }, + { + "epoch": 7.151757188498403, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4477 + }, + { + "epoch": 7.15335463258786, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 4478 + }, + { + "epoch": 7.154952076677317, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4479 + }, + { + "epoch": 7.156549520766773, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 4480 + }, + { + "epoch": 7.15814696485623, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4481 + }, + { + "epoch": 7.159744408945687, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4482 + }, + { + "epoch": 7.161341853035144, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4483 + }, + { + "epoch": 7.1629392971246, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4484 + }, + { + "epoch": 7.164536741214057, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4485 + }, + { + "epoch": 7.166134185303514, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4486 + }, + { + "epoch": 7.167731629392971, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4487 + }, + { + "epoch": 7.169329073482428, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4488 + }, + { + "epoch": 7.170926517571885, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4489 + }, + { + "epoch": 7.172523961661342, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4490 + }, + { + "epoch": 7.174121405750799, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4491 + }, + { + "epoch": 7.175718849840256, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 4492 + }, + { + "epoch": 7.177316293929713, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4493 + }, + { + "epoch": 7.178913738019169, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4494 + }, + { + "epoch": 7.180511182108626, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4495 + }, + { + "epoch": 7.182108626198083, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 4496 + }, + { + "epoch": 7.18370607028754, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4497 + }, + { + "epoch": 7.185303514376997, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4498 + }, + { + "epoch": 7.186900958466453, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4499 + }, + { + "epoch": 7.18849840255591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4500 + }, + { + "epoch": 7.1900958466453675, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4501 + }, + { + "epoch": 7.1916932907348246, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 4502 + }, + { + "epoch": 7.193290734824281, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4503 + }, + { + "epoch": 7.194888178913738, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 4504 + }, + { + "epoch": 7.196485623003195, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4505 + }, + { + "epoch": 7.198083067092652, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4506 + }, + { + "epoch": 7.199680511182109, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4507 + }, + { + "epoch": 7.201277955271565, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4508 + }, + { + "epoch": 7.202875399361022, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4509 + }, + { + "epoch": 7.204472843450479, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4510 + }, + { + "epoch": 7.206070287539936, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4511 + }, + { + "epoch": 7.207667731629393, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4512 + }, + { + "epoch": 7.2092651757188495, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 4513 + }, + { + "epoch": 7.210862619808307, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4514 + }, + { + "epoch": 7.212460063897764, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4515 + }, + { + "epoch": 7.214057507987221, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4516 + }, + { + "epoch": 7.215654952076678, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4517 + }, + { + "epoch": 7.217252396166134, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4518 + }, + { + "epoch": 7.218849840255591, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4519 + }, + { + "epoch": 7.220447284345048, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4520 + }, + { + "epoch": 7.222044728434505, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4521 + }, + { + "epoch": 7.223642172523961, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4522 + }, + { + "epoch": 7.225239616613418, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 4523 + }, + { + "epoch": 7.226837060702875, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4524 + }, + { + "epoch": 7.228434504792332, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 4525 + }, + { + "epoch": 7.2300319488817895, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4526 + }, + { + "epoch": 7.231629392971246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4527 + }, + { + "epoch": 7.233226837060703, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4528 + }, + { + "epoch": 7.23482428115016, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4529 + }, + { + "epoch": 7.236421725239617, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4530 + }, + { + "epoch": 7.238019169329074, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4531 + }, + { + "epoch": 7.23961661341853, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4532 + }, + { + "epoch": 7.241214057507987, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4533 + }, + { + "epoch": 7.242811501597444, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4534 + }, + { + "epoch": 7.244408945686901, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4535 + }, + { + "epoch": 7.246006389776358, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4536 + }, + { + "epoch": 7.247603833865814, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4537 + }, + { + "epoch": 7.2492012779552715, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4538 + }, + { + "epoch": 7.2507987220447285, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4539 + }, + { + "epoch": 7.252396166134186, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4540 + }, + { + "epoch": 7.253993610223642, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4541 + }, + { + "epoch": 7.255591054313099, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4542 + }, + { + "epoch": 7.257188498402556, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 4543 + }, + { + "epoch": 7.258785942492013, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 4544 + }, + { + "epoch": 7.26038338658147, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 4545 + }, + { + "epoch": 7.261980830670926, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4546 + }, + { + "epoch": 7.263578274760383, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4547 + }, + { + "epoch": 7.26517571884984, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4548 + }, + { + "epoch": 7.266773162939297, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4549 + }, + { + "epoch": 7.268370607028754, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4550 + }, + { + "epoch": 7.2699680511182105, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4551 + }, + { + "epoch": 7.271565495207668, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4552 + }, + { + "epoch": 7.273162939297125, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4553 + }, + { + "epoch": 7.274760383386582, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4554 + }, + { + "epoch": 7.276357827476039, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4555 + }, + { + "epoch": 7.277955271565495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4556 + }, + { + "epoch": 7.279552715654952, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4557 + }, + { + "epoch": 7.281150159744409, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4558 + }, + { + "epoch": 7.282747603833866, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4559 + }, + { + "epoch": 7.284345047923322, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4560 + }, + { + "epoch": 7.285942492012779, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 4561 + }, + { + "epoch": 7.287539936102236, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4562 + }, + { + "epoch": 7.289137380191693, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 4563 + }, + { + "epoch": 7.2907348242811505, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 4564 + }, + { + "epoch": 7.292332268370607, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4565 + }, + { + "epoch": 7.293929712460064, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4566 + }, + { + "epoch": 7.295527156549521, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 4567 + }, + { + "epoch": 7.297124600638978, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4568 + }, + { + "epoch": 7.298722044728435, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4569 + }, + { + "epoch": 7.300319488817891, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4570 + }, + { + "epoch": 7.301916932907348, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4571 + }, + { + "epoch": 7.303514376996805, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4572 + }, + { + "epoch": 7.305111821086262, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 4573 + }, + { + "epoch": 7.306709265175719, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4574 + }, + { + "epoch": 7.3083067092651754, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4575 + }, + { + "epoch": 7.3099041533546325, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4576 + }, + { + "epoch": 7.31150159744409, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 4577 + }, + { + "epoch": 7.313099041533547, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4578 + }, + { + "epoch": 7.314696485623003, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 4579 + }, + { + "epoch": 7.31629392971246, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4580 + }, + { + "epoch": 7.317891373801917, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4581 + }, + { + "epoch": 7.319488817891374, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4582 + }, + { + "epoch": 7.321086261980831, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4583 + }, + { + "epoch": 7.322683706070287, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4584 + }, + { + "epoch": 7.324281150159744, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4585 + }, + { + "epoch": 7.325878594249201, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4586 + }, + { + "epoch": 7.327476038338658, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 4587 + }, + { + "epoch": 7.329073482428115, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4588 + }, + { + "epoch": 7.330670926517572, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4589 + }, + { + "epoch": 7.332268370607029, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4590 + }, + { + "epoch": 7.333865814696486, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4591 + }, + { + "epoch": 7.335463258785943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 4592 + }, + { + "epoch": 7.3370607028754, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4593 + }, + { + "epoch": 7.338658146964856, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4594 + }, + { + "epoch": 7.340255591054313, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 4595 + }, + { + "epoch": 7.34185303514377, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 4596 + }, + { + "epoch": 7.343450479233227, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4597 + }, + { + "epoch": 7.345047923322683, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4598 + }, + { + "epoch": 7.34664536741214, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4599 + }, + { + "epoch": 7.348242811501597, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4600 + }, + { + "epoch": 7.3498402555910545, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4601 + }, + { + "epoch": 7.3514376996805115, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4602 + }, + { + "epoch": 7.353035143769968, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4603 + }, + { + "epoch": 7.354632587859425, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4604 + }, + { + "epoch": 7.356230031948882, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4605 + }, + { + "epoch": 7.357827476038339, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4606 + }, + { + "epoch": 7.359424920127796, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4607 + }, + { + "epoch": 7.361022364217252, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4608 + }, + { + "epoch": 7.362619808306709, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 4609 + }, + { + "epoch": 7.364217252396166, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 4610 + }, + { + "epoch": 7.365814696485623, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4611 + }, + { + "epoch": 7.36741214057508, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4612 + }, + { + "epoch": 7.3690095846645365, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4613 + }, + { + "epoch": 7.3706070287539935, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4614 + }, + { + "epoch": 7.372204472843451, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4615 + }, + { + "epoch": 7.373801916932908, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4616 + }, + { + "epoch": 7.375399361022364, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4617 + }, + { + "epoch": 7.376996805111821, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4618 + }, + { + "epoch": 7.378594249201278, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4619 + }, + { + "epoch": 7.380191693290735, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4620 + }, + { + "epoch": 7.381789137380192, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4621 + }, + { + "epoch": 7.383386581469648, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4622 + }, + { + "epoch": 7.384984025559105, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 4623 + }, + { + "epoch": 7.386581469648562, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4624 + }, + { + "epoch": 7.388178913738019, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4625 + }, + { + "epoch": 7.389776357827476, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4626 + }, + { + "epoch": 7.391373801916933, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4627 + }, + { + "epoch": 7.39297124600639, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4628 + }, + { + "epoch": 7.394568690095847, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4629 + }, + { + "epoch": 7.396166134185304, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4630 + }, + { + "epoch": 7.397763578274761, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4631 + }, + { + "epoch": 7.399361022364217, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4632 + }, + { + "epoch": 7.400958466453674, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4633 + }, + { + "epoch": 7.402555910543131, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4634 + }, + { + "epoch": 7.404153354632588, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4635 + }, + { + "epoch": 7.405750798722044, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 4636 + }, + { + "epoch": 7.407348242811501, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4637 + }, + { + "epoch": 7.4089456869009584, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4638 + }, + { + "epoch": 7.4105431309904155, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4639 + }, + { + "epoch": 7.412140575079873, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4640 + }, + { + "epoch": 7.413738019169329, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4641 + }, + { + "epoch": 7.415335463258786, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4642 + }, + { + "epoch": 7.416932907348243, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4643 + }, + { + "epoch": 7.4185303514377, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4644 + }, + { + "epoch": 7.420127795527157, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4645 + }, + { + "epoch": 7.421725239616613, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4646 + }, + { + "epoch": 7.42332268370607, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4647 + }, + { + "epoch": 7.424920127795527, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4648 + }, + { + "epoch": 7.426517571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4649 + }, + { + "epoch": 7.428115015974441, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4650 + }, + { + "epoch": 7.4297124600638975, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4651 + }, + { + "epoch": 7.431309904153355, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4652 + }, + { + "epoch": 7.432907348242812, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 4653 + }, + { + "epoch": 7.434504792332269, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4654 + }, + { + "epoch": 7.436102236421725, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 4655 + }, + { + "epoch": 7.437699680511182, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4656 + }, + { + "epoch": 7.439297124600639, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4657 + }, + { + "epoch": 7.440894568690096, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4658 + }, + { + "epoch": 7.442492012779553, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4659 + }, + { + "epoch": 7.444089456869009, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4660 + }, + { + "epoch": 7.445686900958466, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4661 + }, + { + "epoch": 7.447284345047923, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4662 + }, + { + "epoch": 7.44888178913738, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4663 + }, + { + "epoch": 7.4504792332268375, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4664 + }, + { + "epoch": 7.452076677316294, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4665 + }, + { + "epoch": 7.453674121405751, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4666 + }, + { + "epoch": 7.455271565495208, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4667 + }, + { + "epoch": 7.456869009584665, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 4668 + }, + { + "epoch": 7.458466453674122, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4669 + }, + { + "epoch": 7.460063897763578, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 4670 + }, + { + "epoch": 7.461661341853035, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 4671 + }, + { + "epoch": 7.463258785942492, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4672 + }, + { + "epoch": 7.464856230031949, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4673 + }, + { + "epoch": 7.466453674121405, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4674 + }, + { + "epoch": 7.468051118210862, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4675 + }, + { + "epoch": 7.4696485623003195, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4676 + }, + { + "epoch": 7.4712460063897765, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4677 + }, + { + "epoch": 7.472843450479234, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4678 + }, + { + "epoch": 7.47444089456869, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4679 + }, + { + "epoch": 7.476038338658147, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4680 + }, + { + "epoch": 7.477635782747604, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4681 + }, + { + "epoch": 7.479233226837061, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4682 + }, + { + "epoch": 7.480830670926517, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4683 + }, + { + "epoch": 7.482428115015974, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4684 + }, + { + "epoch": 7.484025559105431, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4685 + }, + { + "epoch": 7.485623003194888, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4686 + }, + { + "epoch": 7.487220447284345, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4687 + }, + { + "epoch": 7.488817891373802, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4688 + }, + { + "epoch": 7.4904153354632586, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4689 + }, + { + "epoch": 7.492012779552716, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4690 + }, + { + "epoch": 7.493610223642173, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4691 + }, + { + "epoch": 7.49520766773163, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4692 + }, + { + "epoch": 7.496805111821086, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 4693 + }, + { + "epoch": 7.498402555910543, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4694 + }, + { + "epoch": 7.5, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4695 + }, + { + "epoch": 7.501597444089457, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4696 + }, + { + "epoch": 7.503194888178914, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4697 + }, + { + "epoch": 7.50479233226837, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4698 + }, + { + "epoch": 7.506389776357827, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4699 + }, + { + "epoch": 7.507987220447284, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4700 + }, + { + "epoch": 7.5095846645367414, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4701 + }, + { + "epoch": 7.511182108626198, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4702 + }, + { + "epoch": 7.512779552715655, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4703 + }, + { + "epoch": 7.514376996805112, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4704 + }, + { + "epoch": 7.515974440894569, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4705 + }, + { + "epoch": 7.517571884984026, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4706 + }, + { + "epoch": 7.519169329073483, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4707 + }, + { + "epoch": 7.520766773162939, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4708 + }, + { + "epoch": 7.522364217252396, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4709 + }, + { + "epoch": 7.523961661341853, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4710 + }, + { + "epoch": 7.52555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 4711 + }, + { + "epoch": 7.527156549520766, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4712 + }, + { + "epoch": 7.5287539936102235, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4713 + }, + { + "epoch": 7.5303514376996805, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4714 + }, + { + "epoch": 7.531948881789138, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4715 + }, + { + "epoch": 7.533546325878595, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4716 + }, + { + "epoch": 7.535143769968051, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4717 + }, + { + "epoch": 7.536741214057508, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4718 + }, + { + "epoch": 7.538338658146965, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4719 + }, + { + "epoch": 7.539936102236422, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4720 + }, + { + "epoch": 7.541533546325878, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4721 + }, + { + "epoch": 7.543130990415335, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 4722 + }, + { + "epoch": 7.544728434504792, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4723 + }, + { + "epoch": 7.546325878594249, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4724 + }, + { + "epoch": 7.547923322683706, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4725 + }, + { + "epoch": 7.549520766773163, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4726 + }, + { + "epoch": 7.55111821086262, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4727 + }, + { + "epoch": 7.552715654952077, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4728 + }, + { + "epoch": 7.554313099041534, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4729 + }, + { + "epoch": 7.555910543130991, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4730 + }, + { + "epoch": 7.557507987220447, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4731 + }, + { + "epoch": 7.559105431309904, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4732 + }, + { + "epoch": 7.560702875399361, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4733 + }, + { + "epoch": 7.562300319488818, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4734 + }, + { + "epoch": 7.563897763578275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4735 + }, + { + "epoch": 7.565495207667731, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4736 + }, + { + "epoch": 7.567092651757188, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4737 + }, + { + "epoch": 7.568690095846645, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 4738 + }, + { + "epoch": 7.5702875399361025, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 4739 + }, + { + "epoch": 7.571884984025559, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4740 + }, + { + "epoch": 7.573482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 4741 + }, + { + "epoch": 7.575079872204473, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4742 + }, + { + "epoch": 7.57667731629393, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4743 + }, + { + "epoch": 7.578274760383387, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4744 + }, + { + "epoch": 7.579872204472844, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4745 + }, + { + "epoch": 7.5814696485623, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 4746 + }, + { + "epoch": 7.583067092651757, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4747 + }, + { + "epoch": 7.584664536741214, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4748 + }, + { + "epoch": 7.586261980830671, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4749 + }, + { + "epoch": 7.587859424920127, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4750 + }, + { + "epoch": 7.5894568690095845, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4751 + }, + { + "epoch": 7.5910543130990416, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4752 + }, + { + "epoch": 7.592651757188499, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4753 + }, + { + "epoch": 7.594249201277956, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4754 + }, + { + "epoch": 7.595846645367412, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4755 + }, + { + "epoch": 7.597444089456869, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4756 + }, + { + "epoch": 7.599041533546326, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4757 + }, + { + "epoch": 7.600638977635783, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4758 + }, + { + "epoch": 7.602236421725239, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4759 + }, + { + "epoch": 7.603833865814696, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4760 + }, + { + "epoch": 7.605431309904153, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4761 + }, + { + "epoch": 7.60702875399361, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4762 + }, + { + "epoch": 7.608626198083067, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4763 + }, + { + "epoch": 7.6102236421725244, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4764 + }, + { + "epoch": 7.611821086261981, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4765 + }, + { + "epoch": 7.613418530351438, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4766 + }, + { + "epoch": 7.615015974440895, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4767 + }, + { + "epoch": 7.616613418530352, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4768 + }, + { + "epoch": 7.618210862619808, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4769 + }, + { + "epoch": 7.619808306709265, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4770 + }, + { + "epoch": 7.621405750798722, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4771 + }, + { + "epoch": 7.623003194888179, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4772 + }, + { + "epoch": 7.624600638977636, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4773 + }, + { + "epoch": 7.626198083067092, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4774 + }, + { + "epoch": 7.627795527156549, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4775 + }, + { + "epoch": 7.6293929712460065, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4776 + }, + { + "epoch": 7.6309904153354635, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4777 + }, + { + "epoch": 7.63258785942492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4778 + }, + { + "epoch": 7.634185303514377, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4779 + }, + { + "epoch": 7.635782747603834, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 4780 + }, + { + "epoch": 7.637380191693291, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4781 + }, + { + "epoch": 7.638977635782748, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4782 + }, + { + "epoch": 7.640575079872205, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4783 + }, + { + "epoch": 7.642172523961661, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4784 + }, + { + "epoch": 7.643769968051118, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4785 + }, + { + "epoch": 7.645367412140575, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4786 + }, + { + "epoch": 7.646964856230032, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 4787 + }, + { + "epoch": 7.6485623003194885, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 4788 + }, + { + "epoch": 7.6501597444089455, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4789 + }, + { + "epoch": 7.651757188498403, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4790 + }, + { + "epoch": 7.65335463258786, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 4791 + }, + { + "epoch": 7.654952076677317, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4792 + }, + { + "epoch": 7.656549520766773, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4793 + }, + { + "epoch": 7.65814696485623, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4794 + }, + { + "epoch": 7.659744408945687, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4795 + }, + { + "epoch": 7.661341853035144, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4796 + }, + { + "epoch": 7.6629392971246, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4797 + }, + { + "epoch": 7.664536741214057, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4798 + }, + { + "epoch": 7.666134185303514, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4799 + }, + { + "epoch": 7.667731629392971, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4800 + }, + { + "epoch": 7.669329073482428, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4801 + }, + { + "epoch": 7.6709265175718855, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4802 + }, + { + "epoch": 7.672523961661342, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 4803 + }, + { + "epoch": 7.674121405750799, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 4804 + }, + { + "epoch": 7.675718849840256, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4805 + }, + { + "epoch": 7.677316293929713, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 4806 + }, + { + "epoch": 7.678913738019169, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4807 + }, + { + "epoch": 7.680511182108626, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 4808 + }, + { + "epoch": 7.682108626198083, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4809 + }, + { + "epoch": 7.68370607028754, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4810 + }, + { + "epoch": 7.685303514376997, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 4811 + }, + { + "epoch": 7.686900958466453, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4812 + }, + { + "epoch": 7.68849840255591, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4813 + }, + { + "epoch": 7.6900958466453675, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4814 + }, + { + "epoch": 7.6916932907348246, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4815 + }, + { + "epoch": 7.693290734824281, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 4816 + }, + { + "epoch": 7.694888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4817 + }, + { + "epoch": 7.696485623003195, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 4818 + }, + { + "epoch": 7.698083067092652, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4819 + }, + { + "epoch": 7.699680511182109, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4820 + }, + { + "epoch": 7.701277955271565, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4821 + }, + { + "epoch": 7.702875399361022, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4822 + }, + { + "epoch": 7.704472843450479, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4823 + }, + { + "epoch": 7.706070287539936, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4824 + }, + { + "epoch": 7.707667731629393, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4825 + }, + { + "epoch": 7.7092651757188495, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4826 + }, + { + "epoch": 7.710862619808307, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4827 + }, + { + "epoch": 7.712460063897764, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4828 + }, + { + "epoch": 7.714057507987221, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 4829 + }, + { + "epoch": 7.715654952076678, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4830 + }, + { + "epoch": 7.717252396166134, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 4831 + }, + { + "epoch": 7.718849840255591, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4832 + }, + { + "epoch": 7.720447284345048, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4833 + }, + { + "epoch": 7.722044728434505, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4834 + }, + { + "epoch": 7.723642172523961, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4835 + }, + { + "epoch": 7.725239616613418, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4836 + }, + { + "epoch": 7.726837060702875, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4837 + }, + { + "epoch": 7.728434504792332, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4838 + }, + { + "epoch": 7.7300319488817895, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4839 + }, + { + "epoch": 7.731629392971246, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 4840 + }, + { + "epoch": 7.733226837060703, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4841 + }, + { + "epoch": 7.73482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4842 + }, + { + "epoch": 7.736421725239617, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4843 + }, + { + "epoch": 7.738019169329074, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4844 + }, + { + "epoch": 7.73961661341853, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4845 + }, + { + "epoch": 7.741214057507987, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4846 + }, + { + "epoch": 7.742811501597444, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4847 + }, + { + "epoch": 7.744408945686901, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4848 + }, + { + "epoch": 7.746006389776358, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 4849 + }, + { + "epoch": 7.747603833865814, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4850 + }, + { + "epoch": 7.7492012779552715, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4851 + }, + { + "epoch": 7.7507987220447285, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4852 + }, + { + "epoch": 7.752396166134186, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4853 + }, + { + "epoch": 7.753993610223642, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4854 + }, + { + "epoch": 7.755591054313099, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4855 + }, + { + "epoch": 7.757188498402556, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4856 + }, + { + "epoch": 7.758785942492013, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4857 + }, + { + "epoch": 7.76038338658147, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4858 + }, + { + "epoch": 7.761980830670926, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4859 + }, + { + "epoch": 7.763578274760383, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4860 + }, + { + "epoch": 7.76517571884984, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4861 + }, + { + "epoch": 7.766773162939297, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4862 + }, + { + "epoch": 7.768370607028754, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4863 + }, + { + "epoch": 7.7699680511182105, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4864 + }, + { + "epoch": 7.771565495207668, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4865 + }, + { + "epoch": 7.773162939297125, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4866 + }, + { + "epoch": 7.774760383386582, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4867 + }, + { + "epoch": 7.776357827476039, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4868 + }, + { + "epoch": 7.777955271565495, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4869 + }, + { + "epoch": 7.779552715654952, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 4870 + }, + { + "epoch": 7.781150159744409, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 4871 + }, + { + "epoch": 7.782747603833866, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4872 + }, + { + "epoch": 7.784345047923322, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4873 + }, + { + "epoch": 7.785942492012779, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4874 + }, + { + "epoch": 7.787539936102236, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4875 + }, + { + "epoch": 7.789137380191693, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 4876 + }, + { + "epoch": 7.7907348242811505, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4877 + }, + { + "epoch": 7.792332268370607, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4878 + }, + { + "epoch": 7.793929712460064, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4879 + }, + { + "epoch": 7.795527156549521, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4880 + }, + { + "epoch": 7.797124600638978, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4881 + }, + { + "epoch": 7.798722044728435, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4882 + }, + { + "epoch": 7.800319488817891, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4883 + }, + { + "epoch": 7.801916932907348, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4884 + }, + { + "epoch": 7.803514376996805, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4885 + }, + { + "epoch": 7.805111821086262, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 4886 + }, + { + "epoch": 7.806709265175719, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4887 + }, + { + "epoch": 7.8083067092651754, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 4888 + }, + { + "epoch": 7.8099041533546325, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4889 + }, + { + "epoch": 7.81150159744409, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4890 + }, + { + "epoch": 7.813099041533547, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4891 + }, + { + "epoch": 7.814696485623003, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4892 + }, + { + "epoch": 7.81629392971246, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4893 + }, + { + "epoch": 7.817891373801917, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4894 + }, + { + "epoch": 7.819488817891374, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 4895 + }, + { + "epoch": 7.821086261980831, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4896 + }, + { + "epoch": 7.822683706070287, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4897 + }, + { + "epoch": 7.824281150159744, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4898 + }, + { + "epoch": 7.825878594249201, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4899 + }, + { + "epoch": 7.827476038338658, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4900 + }, + { + "epoch": 7.8290734824281145, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4901 + }, + { + "epoch": 7.830670926517572, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4902 + }, + { + "epoch": 7.832268370607029, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 4903 + }, + { + "epoch": 7.833865814696486, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4904 + }, + { + "epoch": 7.835463258785943, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4905 + }, + { + "epoch": 7.8370607028754, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4906 + }, + { + "epoch": 7.838658146964856, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 4907 + }, + { + "epoch": 7.840255591054313, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4908 + }, + { + "epoch": 7.84185303514377, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4909 + }, + { + "epoch": 7.843450479233227, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4910 + }, + { + "epoch": 7.845047923322683, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 4911 + }, + { + "epoch": 7.84664536741214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4912 + }, + { + "epoch": 7.848242811501597, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 4913 + }, + { + "epoch": 7.8498402555910545, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4914 + }, + { + "epoch": 7.8514376996805115, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4915 + }, + { + "epoch": 7.853035143769968, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 4916 + }, + { + "epoch": 7.854632587859425, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4917 + }, + { + "epoch": 7.856230031948882, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4918 + }, + { + "epoch": 7.857827476038339, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 4919 + }, + { + "epoch": 7.859424920127795, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 4920 + }, + { + "epoch": 7.861022364217252, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4921 + }, + { + "epoch": 7.862619808306709, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4922 + }, + { + "epoch": 7.864217252396166, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4923 + }, + { + "epoch": 7.865814696485623, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4924 + }, + { + "epoch": 7.86741214057508, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 4925 + }, + { + "epoch": 7.8690095846645365, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4926 + }, + { + "epoch": 7.8706070287539935, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4927 + }, + { + "epoch": 7.872204472843451, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 4928 + }, + { + "epoch": 7.873801916932908, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4929 + }, + { + "epoch": 7.875399361022364, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4930 + }, + { + "epoch": 7.876996805111821, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4931 + }, + { + "epoch": 7.878594249201278, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4932 + }, + { + "epoch": 7.880191693290735, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4933 + }, + { + "epoch": 7.881789137380192, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4934 + }, + { + "epoch": 7.883386581469648, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4935 + }, + { + "epoch": 7.884984025559105, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4936 + }, + { + "epoch": 7.886581469648562, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 4937 + }, + { + "epoch": 7.888178913738019, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4938 + }, + { + "epoch": 7.8897763578274756, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4939 + }, + { + "epoch": 7.891373801916933, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4940 + }, + { + "epoch": 7.89297124600639, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4941 + }, + { + "epoch": 7.894568690095847, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 4942 + }, + { + "epoch": 7.896166134185304, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4943 + }, + { + "epoch": 7.897763578274761, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4944 + }, + { + "epoch": 7.899361022364217, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4945 + }, + { + "epoch": 7.900958466453674, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4946 + }, + { + "epoch": 7.902555910543131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4947 + }, + { + "epoch": 7.904153354632588, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4948 + }, + { + "epoch": 7.905750798722044, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4949 + }, + { + "epoch": 7.907348242811501, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4950 + }, + { + "epoch": 7.9089456869009584, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 4951 + }, + { + "epoch": 7.9105431309904155, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4952 + }, + { + "epoch": 7.912140575079873, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 4953 + }, + { + "epoch": 7.913738019169329, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4954 + }, + { + "epoch": 7.915335463258786, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4955 + }, + { + "epoch": 7.916932907348243, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4956 + }, + { + "epoch": 7.9185303514377, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4957 + }, + { + "epoch": 7.920127795527156, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4958 + }, + { + "epoch": 7.921725239616613, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4959 + }, + { + "epoch": 7.92332268370607, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4960 + }, + { + "epoch": 7.924920127795527, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4961 + }, + { + "epoch": 7.926517571884984, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4962 + }, + { + "epoch": 7.928115015974441, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 4963 + }, + { + "epoch": 7.9297124600638975, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4964 + }, + { + "epoch": 7.931309904153355, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 4965 + }, + { + "epoch": 7.932907348242812, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 4966 + }, + { + "epoch": 7.934504792332269, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4967 + }, + { + "epoch": 7.936102236421725, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4968 + }, + { + "epoch": 7.937699680511182, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 4969 + }, + { + "epoch": 7.939297124600639, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4970 + }, + { + "epoch": 7.940894568690096, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 4971 + }, + { + "epoch": 7.942492012779553, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 4972 + }, + { + "epoch": 7.944089456869009, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4973 + }, + { + "epoch": 7.945686900958466, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4974 + }, + { + "epoch": 7.947284345047923, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 4975 + }, + { + "epoch": 7.94888178913738, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4976 + }, + { + "epoch": 7.950479233226837, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4977 + }, + { + "epoch": 7.952076677316294, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4978 + }, + { + "epoch": 7.953674121405751, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4979 + }, + { + "epoch": 7.955271565495208, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4980 + }, + { + "epoch": 7.956869009584665, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 4981 + }, + { + "epoch": 7.958466453674122, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4982 + }, + { + "epoch": 7.960063897763578, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4983 + }, + { + "epoch": 7.961661341853035, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4984 + }, + { + "epoch": 7.963258785942492, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4985 + }, + { + "epoch": 7.964856230031949, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4986 + }, + { + "epoch": 7.966453674121405, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4987 + }, + { + "epoch": 7.968051118210862, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4988 + }, + { + "epoch": 7.9696485623003195, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4989 + }, + { + "epoch": 7.9712460063897765, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 4990 + }, + { + "epoch": 7.972843450479234, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4991 + }, + { + "epoch": 7.97444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4992 + }, + { + "epoch": 7.976038338658147, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4993 + }, + { + "epoch": 7.977635782747604, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4994 + }, + { + "epoch": 7.979233226837061, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4995 + }, + { + "epoch": 7.980830670926517, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4996 + }, + { + "epoch": 7.982428115015974, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4997 + }, + { + "epoch": 7.984025559105431, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4998 + }, + { + "epoch": 7.985623003194888, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4999 + }, + { + "epoch": 7.987220447284345, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5000 + }, + { + "epoch": 7.988817891373802, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5001 + }, + { + "epoch": 7.9904153354632586, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5002 + }, + { + "epoch": 7.992012779552716, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5003 + }, + { + "epoch": 7.993610223642173, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5004 + }, + { + "epoch": 7.99520766773163, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5005 + }, + { + "epoch": 7.996805111821086, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 5006 + }, + { + "epoch": 7.998402555910543, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 5007 + }, + { + "epoch": 8.0, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5008 + }, + { + "epoch": 8.001597444089457, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5009 + }, + { + "epoch": 8.003194888178914, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5010 + }, + { + "epoch": 8.004792332268371, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5011 + }, + { + "epoch": 8.006389776357828, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5012 + }, + { + "epoch": 8.007987220447284, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5013 + }, + { + "epoch": 8.00958466453674, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5014 + }, + { + "epoch": 8.011182108626198, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5015 + }, + { + "epoch": 8.012779552715655, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 5016 + }, + { + "epoch": 8.014376996805112, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 5017 + }, + { + "epoch": 8.015974440894569, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5018 + }, + { + "epoch": 8.017571884984026, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 5019 + }, + { + "epoch": 8.019169329073483, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5020 + }, + { + "epoch": 8.02076677316294, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5021 + }, + { + "epoch": 8.022364217252397, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5022 + }, + { + "epoch": 8.023961661341852, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5023 + }, + { + "epoch": 8.02555910543131, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5024 + }, + { + "epoch": 8.027156549520766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 5025 + }, + { + "epoch": 8.028753993610223, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5026 + }, + { + "epoch": 8.03035143769968, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5027 + }, + { + "epoch": 8.031948881789138, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5028 + }, + { + "epoch": 8.033546325878595, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5029 + }, + { + "epoch": 8.035143769968052, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5030 + }, + { + "epoch": 8.036741214057509, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5031 + }, + { + "epoch": 8.038338658146964, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5032 + }, + { + "epoch": 8.039936102236421, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 5033 + }, + { + "epoch": 8.041533546325878, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5034 + }, + { + "epoch": 8.043130990415335, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5035 + }, + { + "epoch": 8.044728434504792, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5036 + }, + { + "epoch": 8.04632587859425, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 5037 + }, + { + "epoch": 8.047923322683706, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 5038 + }, + { + "epoch": 8.049520766773163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5039 + }, + { + "epoch": 8.05111821086262, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5040 + }, + { + "epoch": 8.052715654952078, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5041 + }, + { + "epoch": 8.054313099041533, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5042 + }, + { + "epoch": 8.05591054313099, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5043 + }, + { + "epoch": 8.057507987220447, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5044 + }, + { + "epoch": 8.059105431309904, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5045 + }, + { + "epoch": 8.060702875399361, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5046 + }, + { + "epoch": 8.062300319488818, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5047 + }, + { + "epoch": 8.063897763578275, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5048 + }, + { + "epoch": 8.065495207667732, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5049 + }, + { + "epoch": 8.06709265175719, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5050 + }, + { + "epoch": 8.068690095846645, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 5051 + }, + { + "epoch": 8.070287539936102, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5052 + }, + { + "epoch": 8.071884984025559, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5053 + }, + { + "epoch": 8.073482428115016, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5054 + }, + { + "epoch": 8.075079872204473, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5055 + }, + { + "epoch": 8.07667731629393, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5056 + }, + { + "epoch": 8.078274760383387, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 5057 + }, + { + "epoch": 8.079872204472844, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5058 + }, + { + "epoch": 8.081469648562301, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5059 + }, + { + "epoch": 8.083067092651758, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5060 + }, + { + "epoch": 8.084664536741213, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5061 + }, + { + "epoch": 8.08626198083067, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 5062 + }, + { + "epoch": 8.087859424920127, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5063 + }, + { + "epoch": 8.089456869009584, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5064 + }, + { + "epoch": 8.091054313099042, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5065 + }, + { + "epoch": 8.092651757188499, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5066 + }, + { + "epoch": 8.094249201277956, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5067 + }, + { + "epoch": 8.095846645367413, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5068 + }, + { + "epoch": 8.09744408945687, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5069 + }, + { + "epoch": 8.099041533546325, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5070 + }, + { + "epoch": 8.100638977635782, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5071 + }, + { + "epoch": 8.10223642172524, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5072 + }, + { + "epoch": 8.103833865814696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5073 + }, + { + "epoch": 8.105431309904153, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5074 + }, + { + "epoch": 8.10702875399361, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5075 + }, + { + "epoch": 8.108626198083067, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5076 + }, + { + "epoch": 8.110223642172524, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5077 + }, + { + "epoch": 8.111821086261982, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 5078 + }, + { + "epoch": 8.113418530351439, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5079 + }, + { + "epoch": 8.115015974440894, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5080 + }, + { + "epoch": 8.116613418530351, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5081 + }, + { + "epoch": 8.118210862619808, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 5082 + }, + { + "epoch": 8.119808306709265, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5083 + }, + { + "epoch": 8.121405750798722, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5084 + }, + { + "epoch": 8.12300319488818, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5085 + }, + { + "epoch": 8.124600638977636, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 5086 + }, + { + "epoch": 8.126198083067093, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5087 + }, + { + "epoch": 8.12779552715655, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 5088 + }, + { + "epoch": 8.129392971246006, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5089 + }, + { + "epoch": 8.130990415335463, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5090 + }, + { + "epoch": 8.13258785942492, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5091 + }, + { + "epoch": 8.134185303514377, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 5092 + }, + { + "epoch": 8.135782747603834, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5093 + }, + { + "epoch": 8.13738019169329, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5094 + }, + { + "epoch": 8.138977635782748, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5095 + }, + { + "epoch": 8.140575079872205, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5096 + }, + { + "epoch": 8.142172523961662, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5097 + }, + { + "epoch": 8.143769968051119, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5098 + }, + { + "epoch": 8.145367412140574, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5099 + }, + { + "epoch": 8.146964856230031, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5100 + }, + { + "epoch": 8.148562300319488, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5101 + }, + { + "epoch": 8.150159744408946, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 5102 + }, + { + "epoch": 8.151757188498403, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5103 + }, + { + "epoch": 8.15335463258786, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5104 + }, + { + "epoch": 8.154952076677317, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5105 + }, + { + "epoch": 8.156549520766774, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5106 + }, + { + "epoch": 8.15814696485623, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5107 + }, + { + "epoch": 8.159744408945686, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5108 + }, + { + "epoch": 8.161341853035143, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5109 + }, + { + "epoch": 8.1629392971246, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5110 + }, + { + "epoch": 8.164536741214057, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5111 + }, + { + "epoch": 8.166134185303514, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5112 + }, + { + "epoch": 8.167731629392971, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5113 + }, + { + "epoch": 8.169329073482428, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5114 + }, + { + "epoch": 8.170926517571885, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5115 + }, + { + "epoch": 8.172523961661343, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5116 + }, + { + "epoch": 8.1741214057508, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5117 + }, + { + "epoch": 8.175718849840255, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5118 + }, + { + "epoch": 8.177316293929712, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5119 + }, + { + "epoch": 8.178913738019169, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5120 + }, + { + "epoch": 8.180511182108626, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5121 + }, + { + "epoch": 8.182108626198083, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5122 + }, + { + "epoch": 8.18370607028754, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5123 + }, + { + "epoch": 8.185303514376997, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 5124 + }, + { + "epoch": 8.186900958466454, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5125 + }, + { + "epoch": 8.188498402555911, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5126 + }, + { + "epoch": 8.190095846645367, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5127 + }, + { + "epoch": 8.191693290734824, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5128 + }, + { + "epoch": 8.19329073482428, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5129 + }, + { + "epoch": 8.194888178913738, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5130 + }, + { + "epoch": 8.196485623003195, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5131 + }, + { + "epoch": 8.198083067092652, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5132 + }, + { + "epoch": 8.199680511182109, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5133 + }, + { + "epoch": 8.201277955271566, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5134 + }, + { + "epoch": 8.202875399361023, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5135 + }, + { + "epoch": 8.204472843450478, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5136 + }, + { + "epoch": 8.206070287539935, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5137 + }, + { + "epoch": 8.207667731629392, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5138 + }, + { + "epoch": 8.20926517571885, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 5139 + }, + { + "epoch": 8.210862619808307, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5140 + }, + { + "epoch": 8.212460063897764, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5141 + }, + { + "epoch": 8.21405750798722, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5142 + }, + { + "epoch": 8.215654952076678, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5143 + }, + { + "epoch": 8.217252396166135, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5144 + }, + { + "epoch": 8.218849840255592, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 5145 + }, + { + "epoch": 8.220447284345047, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 5146 + }, + { + "epoch": 8.222044728434504, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5147 + }, + { + "epoch": 8.223642172523961, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5148 + }, + { + "epoch": 8.225239616613418, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5149 + }, + { + "epoch": 8.226837060702875, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5150 + }, + { + "epoch": 8.228434504792332, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5151 + }, + { + "epoch": 8.23003194888179, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5152 + }, + { + "epoch": 8.231629392971247, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5153 + }, + { + "epoch": 8.233226837060704, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 5154 + }, + { + "epoch": 8.23482428115016, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5155 + }, + { + "epoch": 8.236421725239616, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5156 + }, + { + "epoch": 8.238019169329073, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5157 + }, + { + "epoch": 8.23961661341853, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5158 + }, + { + "epoch": 8.241214057507987, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5159 + }, + { + "epoch": 8.242811501597444, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5160 + }, + { + "epoch": 8.244408945686901, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5161 + }, + { + "epoch": 8.246006389776358, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5162 + }, + { + "epoch": 8.247603833865815, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5163 + }, + { + "epoch": 8.249201277955272, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5164 + }, + { + "epoch": 8.250798722044728, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5165 + }, + { + "epoch": 8.252396166134185, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5166 + }, + { + "epoch": 8.253993610223642, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5167 + }, + { + "epoch": 8.255591054313099, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 5168 + }, + { + "epoch": 8.257188498402556, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5169 + }, + { + "epoch": 8.258785942492013, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5170 + }, + { + "epoch": 8.26038338658147, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5171 + }, + { + "epoch": 8.261980830670927, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 5172 + }, + { + "epoch": 8.263578274760384, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5173 + }, + { + "epoch": 8.26517571884984, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5174 + }, + { + "epoch": 8.266773162939296, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5175 + }, + { + "epoch": 8.268370607028753, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5176 + }, + { + "epoch": 8.26996805111821, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5177 + }, + { + "epoch": 8.271565495207668, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5178 + }, + { + "epoch": 8.273162939297125, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5179 + }, + { + "epoch": 8.274760383386582, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5180 + }, + { + "epoch": 8.276357827476039, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5181 + }, + { + "epoch": 8.277955271565496, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5182 + }, + { + "epoch": 8.279552715654953, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5183 + }, + { + "epoch": 8.281150159744408, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5184 + }, + { + "epoch": 8.282747603833865, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5185 + }, + { + "epoch": 8.284345047923322, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5186 + }, + { + "epoch": 8.28594249201278, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5187 + }, + { + "epoch": 8.287539936102236, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5188 + }, + { + "epoch": 8.289137380191693, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5189 + }, + { + "epoch": 8.29073482428115, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 5190 + }, + { + "epoch": 8.292332268370608, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 5191 + }, + { + "epoch": 8.293929712460065, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5192 + }, + { + "epoch": 8.295527156549522, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5193 + }, + { + "epoch": 8.297124600638977, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5194 + }, + { + "epoch": 8.298722044728434, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 5195 + }, + { + "epoch": 8.300319488817891, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5196 + }, + { + "epoch": 8.301916932907348, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5197 + }, + { + "epoch": 8.303514376996805, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5198 + }, + { + "epoch": 8.305111821086262, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5199 + }, + { + "epoch": 8.30670926517572, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5200 + }, + { + "epoch": 8.308306709265176, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 5201 + }, + { + "epoch": 8.309904153354633, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5202 + }, + { + "epoch": 8.311501597444089, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5203 + }, + { + "epoch": 8.313099041533546, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5204 + }, + { + "epoch": 8.314696485623003, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5205 + }, + { + "epoch": 8.31629392971246, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5206 + }, + { + "epoch": 8.317891373801917, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 5207 + }, + { + "epoch": 8.319488817891374, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5208 + }, + { + "epoch": 8.321086261980831, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5209 + }, + { + "epoch": 8.322683706070288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5210 + }, + { + "epoch": 8.324281150159745, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5211 + }, + { + "epoch": 8.3258785942492, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5212 + }, + { + "epoch": 8.327476038338657, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5213 + }, + { + "epoch": 8.329073482428115, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 5214 + }, + { + "epoch": 8.330670926517572, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5215 + }, + { + "epoch": 8.332268370607029, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5216 + }, + { + "epoch": 8.333865814696486, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5217 + }, + { + "epoch": 8.335463258785943, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5218 + }, + { + "epoch": 8.3370607028754, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5219 + }, + { + "epoch": 8.338658146964857, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5220 + }, + { + "epoch": 8.340255591054314, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 5221 + }, + { + "epoch": 8.34185303514377, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5222 + }, + { + "epoch": 8.343450479233226, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5223 + }, + { + "epoch": 8.345047923322683, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5224 + }, + { + "epoch": 8.34664536741214, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5225 + }, + { + "epoch": 8.348242811501597, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5226 + }, + { + "epoch": 8.349840255591054, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5227 + }, + { + "epoch": 8.351437699680512, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 5228 + }, + { + "epoch": 8.353035143769969, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5229 + }, + { + "epoch": 8.354632587859426, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5230 + }, + { + "epoch": 8.356230031948883, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 5231 + }, + { + "epoch": 8.357827476038338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 5232 + }, + { + "epoch": 8.359424920127795, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5233 + }, + { + "epoch": 8.361022364217252, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5234 + }, + { + "epoch": 8.36261980830671, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5235 + }, + { + "epoch": 8.364217252396166, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5236 + }, + { + "epoch": 8.365814696485623, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5237 + }, + { + "epoch": 8.36741214057508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5238 + }, + { + "epoch": 8.369009584664537, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5239 + }, + { + "epoch": 8.370607028753994, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5240 + }, + { + "epoch": 8.37220447284345, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5241 + }, + { + "epoch": 8.373801916932907, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5242 + }, + { + "epoch": 8.375399361022364, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5243 + }, + { + "epoch": 8.37699680511182, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5244 + }, + { + "epoch": 8.378594249201278, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 5245 + }, + { + "epoch": 8.380191693290735, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5246 + }, + { + "epoch": 8.381789137380192, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5247 + }, + { + "epoch": 8.383386581469649, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5248 + }, + { + "epoch": 8.384984025559106, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5249 + }, + { + "epoch": 8.386581469648561, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5250 + }, + { + "epoch": 8.388178913738018, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5251 + }, + { + "epoch": 8.389776357827476, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5252 + }, + { + "epoch": 8.391373801916933, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5253 + }, + { + "epoch": 8.39297124600639, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5254 + }, + { + "epoch": 8.394568690095847, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5255 + }, + { + "epoch": 8.396166134185304, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5256 + }, + { + "epoch": 8.39776357827476, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5257 + }, + { + "epoch": 8.399361022364218, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5258 + }, + { + "epoch": 8.400958466453675, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5259 + }, + { + "epoch": 8.40255591054313, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5260 + }, + { + "epoch": 8.404153354632587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5261 + }, + { + "epoch": 8.405750798722044, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 5262 + }, + { + "epoch": 8.407348242811501, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 5263 + }, + { + "epoch": 8.408945686900958, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5264 + }, + { + "epoch": 8.410543130990416, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5265 + }, + { + "epoch": 8.412140575079873, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5266 + }, + { + "epoch": 8.41373801916933, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5267 + }, + { + "epoch": 8.415335463258787, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5268 + }, + { + "epoch": 8.416932907348242, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5269 + }, + { + "epoch": 8.418530351437699, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5270 + }, + { + "epoch": 8.420127795527156, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5271 + }, + { + "epoch": 8.421725239616613, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5272 + }, + { + "epoch": 8.42332268370607, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5273 + }, + { + "epoch": 8.424920127795527, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5274 + }, + { + "epoch": 8.426517571884984, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5275 + }, + { + "epoch": 8.428115015974441, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5276 + }, + { + "epoch": 8.429712460063898, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5277 + }, + { + "epoch": 8.431309904153355, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5278 + }, + { + "epoch": 8.43290734824281, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5279 + }, + { + "epoch": 8.434504792332268, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5280 + }, + { + "epoch": 8.436102236421725, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5281 + }, + { + "epoch": 8.437699680511182, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 5282 + }, + { + "epoch": 8.439297124600639, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5283 + }, + { + "epoch": 8.440894568690096, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5284 + }, + { + "epoch": 8.442492012779553, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5285 + }, + { + "epoch": 8.44408945686901, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5286 + }, + { + "epoch": 8.445686900958467, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5287 + }, + { + "epoch": 8.447284345047922, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5288 + }, + { + "epoch": 8.44888178913738, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 5289 + }, + { + "epoch": 8.450479233226837, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 5290 + }, + { + "epoch": 8.452076677316294, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5291 + }, + { + "epoch": 8.45367412140575, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5292 + }, + { + "epoch": 8.455271565495208, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5293 + }, + { + "epoch": 8.456869009584665, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5294 + }, + { + "epoch": 8.458466453674122, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5295 + }, + { + "epoch": 8.460063897763579, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5296 + }, + { + "epoch": 8.461661341853036, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5297 + }, + { + "epoch": 8.463258785942491, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5298 + }, + { + "epoch": 8.464856230031948, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5299 + }, + { + "epoch": 8.466453674121405, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5300 + }, + { + "epoch": 8.468051118210862, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 5301 + }, + { + "epoch": 8.46964856230032, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5302 + }, + { + "epoch": 8.471246006389777, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5303 + }, + { + "epoch": 8.472843450479234, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5304 + }, + { + "epoch": 8.47444089456869, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 5305 + }, + { + "epoch": 8.476038338658148, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5306 + }, + { + "epoch": 8.477635782747603, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5307 + }, + { + "epoch": 8.47923322683706, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5308 + }, + { + "epoch": 8.480830670926517, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 5309 + }, + { + "epoch": 8.482428115015974, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5310 + }, + { + "epoch": 8.484025559105431, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5311 + }, + { + "epoch": 8.485623003194888, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 5312 + }, + { + "epoch": 8.487220447284345, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5313 + }, + { + "epoch": 8.488817891373802, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 5314 + }, + { + "epoch": 8.49041533546326, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 5315 + }, + { + "epoch": 8.492012779552716, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5316 + }, + { + "epoch": 8.493610223642172, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5317 + }, + { + "epoch": 8.495207667731629, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5318 + }, + { + "epoch": 8.496805111821086, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5319 + }, + { + "epoch": 8.498402555910543, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5320 + }, + { + "epoch": 8.5, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 5321 + }, + { + "epoch": 8.501597444089457, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5322 + }, + { + "epoch": 8.503194888178914, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5323 + }, + { + "epoch": 8.504792332268371, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5324 + }, + { + "epoch": 8.506389776357828, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5325 + }, + { + "epoch": 8.507987220447284, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5326 + }, + { + "epoch": 8.50958466453674, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5327 + }, + { + "epoch": 8.511182108626198, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5328 + }, + { + "epoch": 8.512779552715655, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5329 + }, + { + "epoch": 8.514376996805112, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5330 + }, + { + "epoch": 8.515974440894569, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5331 + }, + { + "epoch": 8.517571884984026, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5332 + }, + { + "epoch": 8.519169329073483, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5333 + }, + { + "epoch": 8.52076677316294, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5334 + }, + { + "epoch": 8.522364217252395, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5335 + }, + { + "epoch": 8.523961661341852, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5336 + }, + { + "epoch": 8.52555910543131, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5337 + }, + { + "epoch": 8.527156549520766, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5338 + }, + { + "epoch": 8.528753993610223, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5339 + }, + { + "epoch": 8.53035143769968, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 5340 + }, + { + "epoch": 8.531948881789138, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5341 + }, + { + "epoch": 8.533546325878595, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5342 + }, + { + "epoch": 8.535143769968052, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5343 + }, + { + "epoch": 8.536741214057509, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5344 + }, + { + "epoch": 8.538338658146966, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5345 + }, + { + "epoch": 8.539936102236421, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 5346 + }, + { + "epoch": 8.541533546325878, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5347 + }, + { + "epoch": 8.543130990415335, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5348 + }, + { + "epoch": 8.544728434504792, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 5349 + }, + { + "epoch": 8.54632587859425, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5350 + }, + { + "epoch": 8.547923322683706, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5351 + }, + { + "epoch": 8.549520766773163, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 5352 + }, + { + "epoch": 8.55111821086262, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5353 + }, + { + "epoch": 8.552715654952078, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 5354 + }, + { + "epoch": 8.554313099041533, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 5355 + }, + { + "epoch": 8.55591054313099, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 5356 + }, + { + "epoch": 8.557507987220447, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5357 + }, + { + "epoch": 8.559105431309904, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5358 + }, + { + "epoch": 8.560702875399361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 5359 + }, + { + "epoch": 8.562300319488818, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5360 + }, + { + "epoch": 8.563897763578275, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5361 + }, + { + "epoch": 8.565495207667732, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 5362 + }, + { + "epoch": 8.56709265175719, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 5363 + }, + { + "epoch": 8.568690095846645, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5364 + }, + { + "epoch": 8.570287539936102, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5365 + }, + { + "epoch": 8.571884984025559, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5366 + }, + { + "epoch": 8.573482428115016, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5367 + }, + { + "epoch": 8.575079872204473, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5368 + }, + { + "epoch": 8.57667731629393, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5369 + }, + { + "epoch": 8.578274760383387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5370 + }, + { + "epoch": 8.579872204472844, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 5371 + }, + { + "epoch": 8.581469648562301, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5372 + }, + { + "epoch": 8.583067092651756, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5373 + }, + { + "epoch": 8.584664536741213, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5374 + }, + { + "epoch": 8.58626198083067, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5375 + }, + { + "epoch": 8.587859424920127, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5376 + }, + { + "epoch": 8.589456869009584, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5377 + }, + { + "epoch": 8.591054313099042, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5378 + }, + { + "epoch": 8.592651757188499, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5379 + }, + { + "epoch": 8.594249201277956, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5380 + }, + { + "epoch": 8.595846645367413, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5381 + }, + { + "epoch": 8.59744408945687, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5382 + }, + { + "epoch": 8.599041533546325, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5383 + }, + { + "epoch": 8.600638977635782, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5384 + }, + { + "epoch": 8.60223642172524, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5385 + }, + { + "epoch": 8.603833865814696, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5386 + }, + { + "epoch": 8.605431309904153, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5387 + }, + { + "epoch": 8.60702875399361, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5388 + }, + { + "epoch": 8.608626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 5389 + }, + { + "epoch": 8.610223642172524, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5390 + }, + { + "epoch": 8.611821086261982, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 5391 + }, + { + "epoch": 8.613418530351439, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5392 + }, + { + "epoch": 8.615015974440894, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5393 + }, + { + "epoch": 8.616613418530351, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 5394 + }, + { + "epoch": 8.618210862619808, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5395 + }, + { + "epoch": 8.619808306709265, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5396 + }, + { + "epoch": 8.621405750798722, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5397 + }, + { + "epoch": 8.62300319488818, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5398 + }, + { + "epoch": 8.624600638977636, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5399 + }, + { + "epoch": 8.626198083067093, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5400 + }, + { + "epoch": 8.62779552715655, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5401 + }, + { + "epoch": 8.629392971246006, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5402 + }, + { + "epoch": 8.630990415335463, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5403 + }, + { + "epoch": 8.63258785942492, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5404 + }, + { + "epoch": 8.634185303514377, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5405 + }, + { + "epoch": 8.635782747603834, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5406 + }, + { + "epoch": 8.63738019169329, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5407 + }, + { + "epoch": 8.638977635782748, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5408 + }, + { + "epoch": 8.640575079872205, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5409 + }, + { + "epoch": 8.642172523961662, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5410 + }, + { + "epoch": 8.643769968051117, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5411 + }, + { + "epoch": 8.645367412140574, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5412 + }, + { + "epoch": 8.646964856230031, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 5413 + }, + { + "epoch": 8.648562300319488, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5414 + }, + { + "epoch": 8.650159744408946, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5415 + }, + { + "epoch": 8.651757188498403, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5416 + }, + { + "epoch": 8.65335463258786, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5417 + }, + { + "epoch": 8.654952076677317, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5418 + }, + { + "epoch": 8.656549520766774, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5419 + }, + { + "epoch": 8.65814696485623, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5420 + }, + { + "epoch": 8.659744408945686, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 5421 + }, + { + "epoch": 8.661341853035143, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5422 + }, + { + "epoch": 8.6629392971246, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 5423 + }, + { + "epoch": 8.664536741214057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 5424 + }, + { + "epoch": 8.666134185303514, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5425 + }, + { + "epoch": 8.667731629392971, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5426 + }, + { + "epoch": 8.669329073482428, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5427 + }, + { + "epoch": 8.670926517571885, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5428 + }, + { + "epoch": 8.672523961661343, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5429 + }, + { + "epoch": 8.6741214057508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5430 + }, + { + "epoch": 8.675718849840255, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5431 + }, + { + "epoch": 8.677316293929712, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5432 + }, + { + "epoch": 8.678913738019169, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5433 + }, + { + "epoch": 8.680511182108626, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5434 + }, + { + "epoch": 8.682108626198083, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 5435 + }, + { + "epoch": 8.68370607028754, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5436 + }, + { + "epoch": 8.685303514376997, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5437 + }, + { + "epoch": 8.686900958466454, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5438 + }, + { + "epoch": 8.688498402555911, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5439 + }, + { + "epoch": 8.690095846645367, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5440 + }, + { + "epoch": 8.691693290734824, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5441 + }, + { + "epoch": 8.69329073482428, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5442 + }, + { + "epoch": 8.694888178913738, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5443 + }, + { + "epoch": 8.696485623003195, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5444 + }, + { + "epoch": 8.698083067092652, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5445 + }, + { + "epoch": 8.699680511182109, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5446 + }, + { + "epoch": 8.701277955271566, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5447 + }, + { + "epoch": 8.702875399361023, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5448 + }, + { + "epoch": 8.704472843450478, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5449 + }, + { + "epoch": 8.706070287539935, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5450 + }, + { + "epoch": 8.707667731629392, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5451 + }, + { + "epoch": 8.70926517571885, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5452 + }, + { + "epoch": 8.710862619808307, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5453 + }, + { + "epoch": 8.712460063897764, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5454 + }, + { + "epoch": 8.71405750798722, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5455 + }, + { + "epoch": 8.715654952076678, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5456 + }, + { + "epoch": 8.717252396166135, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5457 + }, + { + "epoch": 8.718849840255592, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5458 + }, + { + "epoch": 8.720447284345047, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5459 + }, + { + "epoch": 8.722044728434504, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 5460 + }, + { + "epoch": 8.723642172523961, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5461 + }, + { + "epoch": 8.725239616613418, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5462 + }, + { + "epoch": 8.726837060702875, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5463 + }, + { + "epoch": 8.728434504792332, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5464 + }, + { + "epoch": 8.73003194888179, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5465 + }, + { + "epoch": 8.731629392971247, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5466 + }, + { + "epoch": 8.733226837060704, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5467 + }, + { + "epoch": 8.73482428115016, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5468 + }, + { + "epoch": 8.736421725239616, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5469 + }, + { + "epoch": 8.738019169329073, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5470 + }, + { + "epoch": 8.73961661341853, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5471 + }, + { + "epoch": 8.741214057507987, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5472 + }, + { + "epoch": 8.742811501597444, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 5473 + }, + { + "epoch": 8.744408945686901, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5474 + }, + { + "epoch": 8.746006389776358, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5475 + }, + { + "epoch": 8.747603833865815, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 5476 + }, + { + "epoch": 8.749201277955272, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5477 + }, + { + "epoch": 8.750798722044728, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5478 + }, + { + "epoch": 8.752396166134185, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5479 + }, + { + "epoch": 8.753993610223642, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5480 + }, + { + "epoch": 8.755591054313099, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5481 + }, + { + "epoch": 8.757188498402556, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5482 + }, + { + "epoch": 8.758785942492013, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5483 + }, + { + "epoch": 8.76038338658147, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5484 + }, + { + "epoch": 8.761980830670927, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5485 + }, + { + "epoch": 8.763578274760384, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5486 + }, + { + "epoch": 8.76517571884984, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5487 + }, + { + "epoch": 8.766773162939296, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5488 + }, + { + "epoch": 8.768370607028753, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5489 + }, + { + "epoch": 8.76996805111821, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5490 + }, + { + "epoch": 8.771565495207668, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5491 + }, + { + "epoch": 8.773162939297125, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5492 + }, + { + "epoch": 8.774760383386582, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5493 + }, + { + "epoch": 8.776357827476039, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5494 + }, + { + "epoch": 8.777955271565496, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5495 + }, + { + "epoch": 8.779552715654953, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 5496 + }, + { + "epoch": 8.781150159744408, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5497 + }, + { + "epoch": 8.782747603833865, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5498 + }, + { + "epoch": 8.784345047923322, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5499 + }, + { + "epoch": 8.78594249201278, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5500 + }, + { + "epoch": 8.787539936102236, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 5501 + }, + { + "epoch": 8.789137380191693, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 5502 + }, + { + "epoch": 8.79073482428115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5503 + }, + { + "epoch": 8.792332268370608, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5504 + }, + { + "epoch": 8.793929712460065, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5505 + }, + { + "epoch": 8.795527156549522, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5506 + }, + { + "epoch": 8.797124600638977, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5507 + }, + { + "epoch": 8.798722044728434, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5508 + }, + { + "epoch": 8.800319488817891, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5509 + }, + { + "epoch": 8.801916932907348, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5510 + }, + { + "epoch": 8.803514376996805, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5511 + }, + { + "epoch": 8.805111821086262, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5512 + }, + { + "epoch": 8.80670926517572, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 5513 + }, + { + "epoch": 8.808306709265176, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5514 + }, + { + "epoch": 8.809904153354633, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5515 + }, + { + "epoch": 8.811501597444089, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5516 + }, + { + "epoch": 8.813099041533546, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5517 + }, + { + "epoch": 8.814696485623003, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5518 + }, + { + "epoch": 8.81629392971246, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5519 + }, + { + "epoch": 8.817891373801917, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5520 + }, + { + "epoch": 8.819488817891374, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5521 + }, + { + "epoch": 8.821086261980831, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5522 + }, + { + "epoch": 8.822683706070288, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 5523 + }, + { + "epoch": 8.824281150159745, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 5524 + }, + { + "epoch": 8.8258785942492, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5525 + }, + { + "epoch": 8.827476038338657, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5526 + }, + { + "epoch": 8.829073482428115, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5527 + }, + { + "epoch": 8.830670926517572, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5528 + }, + { + "epoch": 8.832268370607029, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5529 + }, + { + "epoch": 8.833865814696486, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5530 + }, + { + "epoch": 8.835463258785943, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5531 + }, + { + "epoch": 8.8370607028754, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 5532 + }, + { + "epoch": 8.838658146964857, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5533 + }, + { + "epoch": 8.840255591054314, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5534 + }, + { + "epoch": 8.84185303514377, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5535 + }, + { + "epoch": 8.843450479233226, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5536 + }, + { + "epoch": 8.845047923322683, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5537 + }, + { + "epoch": 8.84664536741214, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5538 + }, + { + "epoch": 8.848242811501597, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5539 + }, + { + "epoch": 8.849840255591054, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5540 + }, + { + "epoch": 8.851437699680512, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5541 + }, + { + "epoch": 8.853035143769969, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5542 + }, + { + "epoch": 8.854632587859426, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5543 + }, + { + "epoch": 8.856230031948883, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5544 + }, + { + "epoch": 8.857827476038338, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5545 + }, + { + "epoch": 8.859424920127795, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5546 + }, + { + "epoch": 8.861022364217252, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 5547 + }, + { + "epoch": 8.86261980830671, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5548 + }, + { + "epoch": 8.864217252396166, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5549 + }, + { + "epoch": 8.865814696485623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5550 + }, + { + "epoch": 8.86741214057508, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5551 + }, + { + "epoch": 8.869009584664537, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5552 + }, + { + "epoch": 8.870607028753994, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5553 + }, + { + "epoch": 8.87220447284345, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5554 + }, + { + "epoch": 8.873801916932907, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5555 + }, + { + "epoch": 8.875399361022364, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5556 + }, + { + "epoch": 8.87699680511182, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5557 + }, + { + "epoch": 8.878594249201278, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5558 + }, + { + "epoch": 8.880191693290735, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5559 + }, + { + "epoch": 8.881789137380192, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 5560 + }, + { + "epoch": 8.883386581469649, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5561 + }, + { + "epoch": 8.884984025559106, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5562 + }, + { + "epoch": 8.886581469648561, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5563 + }, + { + "epoch": 8.888178913738018, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5564 + }, + { + "epoch": 8.889776357827476, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5565 + }, + { + "epoch": 8.891373801916933, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5566 + }, + { + "epoch": 8.89297124600639, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5567 + }, + { + "epoch": 8.894568690095847, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5568 + }, + { + "epoch": 8.896166134185304, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5569 + }, + { + "epoch": 8.89776357827476, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5570 + }, + { + "epoch": 8.899361022364218, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5571 + }, + { + "epoch": 8.900958466453675, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5572 + }, + { + "epoch": 8.90255591054313, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5573 + }, + { + "epoch": 8.904153354632587, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5574 + }, + { + "epoch": 8.905750798722044, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5575 + }, + { + "epoch": 8.907348242811501, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5576 + }, + { + "epoch": 8.908945686900958, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5577 + }, + { + "epoch": 8.910543130990416, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5578 + }, + { + "epoch": 8.912140575079873, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5579 + }, + { + "epoch": 8.91373801916933, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5580 + }, + { + "epoch": 8.915335463258787, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5581 + }, + { + "epoch": 8.916932907348244, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5582 + }, + { + "epoch": 8.918530351437699, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5583 + }, + { + "epoch": 8.920127795527156, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 5584 + }, + { + "epoch": 8.921725239616613, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5585 + }, + { + "epoch": 8.92332268370607, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5586 + }, + { + "epoch": 8.924920127795527, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5587 + }, + { + "epoch": 8.926517571884984, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5588 + }, + { + "epoch": 8.928115015974441, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5589 + }, + { + "epoch": 8.929712460063898, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5590 + }, + { + "epoch": 8.931309904153355, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5591 + }, + { + "epoch": 8.93290734824281, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 5592 + }, + { + "epoch": 8.934504792332268, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5593 + }, + { + "epoch": 8.936102236421725, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5594 + }, + { + "epoch": 8.937699680511182, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5595 + }, + { + "epoch": 8.939297124600639, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5596 + }, + { + "epoch": 8.940894568690096, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5597 + }, + { + "epoch": 8.942492012779553, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5598 + }, + { + "epoch": 8.94408945686901, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 5599 + }, + { + "epoch": 8.945686900958467, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5600 + }, + { + "epoch": 8.947284345047922, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5601 + }, + { + "epoch": 8.94888178913738, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5602 + }, + { + "epoch": 8.950479233226837, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5603 + }, + { + "epoch": 8.952076677316294, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5604 + }, + { + "epoch": 8.95367412140575, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5605 + }, + { + "epoch": 8.955271565495208, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5606 + }, + { + "epoch": 8.956869009584665, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 5607 + }, + { + "epoch": 8.958466453674122, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5608 + }, + { + "epoch": 8.960063897763579, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5609 + }, + { + "epoch": 8.961661341853034, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5610 + }, + { + "epoch": 8.963258785942491, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5611 + }, + { + "epoch": 8.964856230031948, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5612 + }, + { + "epoch": 8.966453674121405, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5613 + }, + { + "epoch": 8.968051118210862, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5614 + }, + { + "epoch": 8.96964856230032, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5615 + }, + { + "epoch": 8.971246006389777, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5616 + }, + { + "epoch": 8.972843450479234, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5617 + }, + { + "epoch": 8.97444089456869, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5618 + }, + { + "epoch": 8.976038338658148, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5619 + }, + { + "epoch": 8.977635782747605, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5620 + }, + { + "epoch": 8.97923322683706, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5621 + }, + { + "epoch": 8.980830670926517, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5622 + }, + { + "epoch": 8.982428115015974, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5623 + }, + { + "epoch": 8.984025559105431, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5624 + }, + { + "epoch": 8.985623003194888, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5625 + }, + { + "epoch": 8.987220447284345, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5626 + }, + { + "epoch": 8.988817891373802, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5627 + }, + { + "epoch": 8.99041533546326, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 5628 + }, + { + "epoch": 8.992012779552716, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5629 + }, + { + "epoch": 8.993610223642172, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5630 + }, + { + "epoch": 8.995207667731629, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 5631 + }, + { + "epoch": 8.996805111821086, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 5632 + }, + { + "epoch": 8.998402555910543, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5633 + }, + { + "epoch": 9.0, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 5634 + }, + { + "epoch": 9.001597444089457, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5635 + }, + { + "epoch": 9.003194888178914, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5636 + }, + { + "epoch": 9.004792332268371, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5637 + }, + { + "epoch": 9.006389776357828, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5638 + }, + { + "epoch": 9.007987220447284, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5639 + }, + { + "epoch": 9.00958466453674, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5640 + }, + { + "epoch": 9.011182108626198, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5641 + }, + { + "epoch": 9.012779552715655, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5642 + }, + { + "epoch": 9.014376996805112, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 5643 + }, + { + "epoch": 9.015974440894569, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 5644 + }, + { + "epoch": 9.017571884984026, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 5645 + }, + { + "epoch": 9.019169329073483, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5646 + }, + { + "epoch": 9.02076677316294, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5647 + }, + { + "epoch": 9.022364217252397, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5648 + }, + { + "epoch": 9.023961661341852, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5649 + }, + { + "epoch": 9.02555910543131, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5650 + }, + { + "epoch": 9.027156549520766, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 5651 + }, + { + "epoch": 9.028753993610223, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5652 + }, + { + "epoch": 9.03035143769968, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5653 + }, + { + "epoch": 9.031948881789138, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5654 + }, + { + "epoch": 9.033546325878595, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5655 + }, + { + "epoch": 9.035143769968052, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5656 + }, + { + "epoch": 9.036741214057509, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5657 + }, + { + "epoch": 9.038338658146964, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 5658 + }, + { + "epoch": 9.039936102236421, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5659 + }, + { + "epoch": 9.041533546325878, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5660 + }, + { + "epoch": 9.043130990415335, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5661 + }, + { + "epoch": 9.044728434504792, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 5662 + }, + { + "epoch": 9.04632587859425, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5663 + }, + { + "epoch": 9.047923322683706, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5664 + }, + { + "epoch": 9.049520766773163, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5665 + }, + { + "epoch": 9.05111821086262, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5666 + }, + { + "epoch": 9.052715654952078, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5667 + }, + { + "epoch": 9.054313099041533, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5668 + }, + { + "epoch": 9.05591054313099, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5669 + }, + { + "epoch": 9.057507987220447, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5670 + }, + { + "epoch": 9.059105431309904, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5671 + }, + { + "epoch": 9.060702875399361, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 5672 + }, + { + "epoch": 9.062300319488818, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5673 + }, + { + "epoch": 9.063897763578275, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5674 + }, + { + "epoch": 9.065495207667732, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 5675 + }, + { + "epoch": 9.06709265175719, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5676 + }, + { + "epoch": 9.068690095846645, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5677 + }, + { + "epoch": 9.070287539936102, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5678 + }, + { + "epoch": 9.071884984025559, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5679 + }, + { + "epoch": 9.073482428115016, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5680 + }, + { + "epoch": 9.075079872204473, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5681 + }, + { + "epoch": 9.07667731629393, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5682 + }, + { + "epoch": 9.078274760383387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5683 + }, + { + "epoch": 9.079872204472844, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5684 + }, + { + "epoch": 9.081469648562301, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5685 + }, + { + "epoch": 9.083067092651758, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5686 + }, + { + "epoch": 9.084664536741213, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5687 + }, + { + "epoch": 9.08626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5688 + }, + { + "epoch": 9.087859424920127, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5689 + }, + { + "epoch": 9.089456869009584, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5690 + }, + { + "epoch": 9.091054313099042, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5691 + }, + { + "epoch": 9.092651757188499, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5692 + }, + { + "epoch": 9.094249201277956, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5693 + }, + { + "epoch": 9.095846645367413, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5694 + }, + { + "epoch": 9.09744408945687, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5695 + }, + { + "epoch": 9.099041533546325, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 5696 + }, + { + "epoch": 9.100638977635782, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 5697 + }, + { + "epoch": 9.10223642172524, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5698 + }, + { + "epoch": 9.103833865814696, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5699 + }, + { + "epoch": 9.105431309904153, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5700 + }, + { + "epoch": 9.10702875399361, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5701 + }, + { + "epoch": 9.108626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5702 + }, + { + "epoch": 9.110223642172524, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5703 + }, + { + "epoch": 9.111821086261982, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 5704 + }, + { + "epoch": 9.113418530351439, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 5705 + }, + { + "epoch": 9.115015974440894, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5706 + }, + { + "epoch": 9.116613418530351, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5707 + }, + { + "epoch": 9.118210862619808, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 5708 + }, + { + "epoch": 9.119808306709265, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5709 + }, + { + "epoch": 9.121405750798722, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5710 + }, + { + "epoch": 9.12300319488818, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5711 + }, + { + "epoch": 9.124600638977636, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5712 + }, + { + "epoch": 9.126198083067093, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5713 + }, + { + "epoch": 9.12779552715655, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5714 + }, + { + "epoch": 9.129392971246006, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5715 + }, + { + "epoch": 9.130990415335463, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5716 + }, + { + "epoch": 9.13258785942492, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5717 + }, + { + "epoch": 9.134185303514377, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5718 + }, + { + "epoch": 9.135782747603834, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5719 + }, + { + "epoch": 9.13738019169329, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5720 + }, + { + "epoch": 9.138977635782748, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 5721 + }, + { + "epoch": 9.140575079872205, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5722 + }, + { + "epoch": 9.142172523961662, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5723 + }, + { + "epoch": 9.143769968051119, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5724 + }, + { + "epoch": 9.145367412140574, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5725 + }, + { + "epoch": 9.146964856230031, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5726 + }, + { + "epoch": 9.148562300319488, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5727 + }, + { + "epoch": 9.150159744408946, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5728 + }, + { + "epoch": 9.151757188498403, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5729 + }, + { + "epoch": 9.15335463258786, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 5730 + }, + { + "epoch": 9.154952076677317, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5731 + }, + { + "epoch": 9.156549520766774, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5732 + }, + { + "epoch": 9.15814696485623, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5733 + }, + { + "epoch": 9.159744408945686, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5734 + }, + { + "epoch": 9.161341853035143, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5735 + }, + { + "epoch": 9.1629392971246, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5736 + }, + { + "epoch": 9.164536741214057, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5737 + }, + { + "epoch": 9.166134185303514, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5738 + }, + { + "epoch": 9.167731629392971, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5739 + }, + { + "epoch": 9.169329073482428, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5740 + }, + { + "epoch": 9.170926517571885, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 5741 + }, + { + "epoch": 9.172523961661343, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5742 + }, + { + "epoch": 9.1741214057508, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5743 + }, + { + "epoch": 9.175718849840255, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5744 + }, + { + "epoch": 9.177316293929712, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5745 + }, + { + "epoch": 9.178913738019169, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5746 + }, + { + "epoch": 9.180511182108626, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5747 + }, + { + "epoch": 9.182108626198083, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5748 + }, + { + "epoch": 9.18370607028754, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 5749 + }, + { + "epoch": 9.185303514376997, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5750 + }, + { + "epoch": 9.186900958466454, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5751 + }, + { + "epoch": 9.188498402555911, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5752 + }, + { + "epoch": 9.190095846645367, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5753 + }, + { + "epoch": 9.191693290734824, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5754 + }, + { + "epoch": 9.19329073482428, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5755 + }, + { + "epoch": 9.194888178913738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5756 + }, + { + "epoch": 9.196485623003195, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 5757 + }, + { + "epoch": 9.198083067092652, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5758 + }, + { + "epoch": 9.199680511182109, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5759 + }, + { + "epoch": 9.201277955271566, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 5760 + }, + { + "epoch": 9.202875399361023, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5761 + }, + { + "epoch": 9.204472843450478, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5762 + }, + { + "epoch": 9.206070287539935, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5763 + }, + { + "epoch": 9.207667731629392, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 5764 + }, + { + "epoch": 9.20926517571885, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5765 + }, + { + "epoch": 9.210862619808307, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 5766 + }, + { + "epoch": 9.212460063897764, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5767 + }, + { + "epoch": 9.21405750798722, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 5768 + }, + { + "epoch": 9.215654952076678, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5769 + }, + { + "epoch": 9.217252396166135, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5770 + }, + { + "epoch": 9.218849840255592, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5771 + }, + { + "epoch": 9.220447284345047, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5772 + }, + { + "epoch": 9.222044728434504, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5773 + }, + { + "epoch": 9.223642172523961, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5774 + }, + { + "epoch": 9.225239616613418, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5775 + }, + { + "epoch": 9.226837060702875, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5776 + }, + { + "epoch": 9.228434504792332, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5777 + }, + { + "epoch": 9.23003194888179, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5778 + }, + { + "epoch": 9.231629392971247, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5779 + }, + { + "epoch": 9.233226837060704, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5780 + }, + { + "epoch": 9.23482428115016, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5781 + }, + { + "epoch": 9.236421725239616, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5782 + }, + { + "epoch": 9.238019169329073, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 5783 + }, + { + "epoch": 9.23961661341853, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5784 + }, + { + "epoch": 9.241214057507987, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5785 + }, + { + "epoch": 9.242811501597444, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5786 + }, + { + "epoch": 9.244408945686901, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5787 + }, + { + "epoch": 9.246006389776358, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5788 + }, + { + "epoch": 9.247603833865815, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5789 + }, + { + "epoch": 9.249201277955272, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5790 + }, + { + "epoch": 9.250798722044728, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5791 + }, + { + "epoch": 9.252396166134185, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5792 + }, + { + "epoch": 9.253993610223642, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5793 + }, + { + "epoch": 9.255591054313099, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5794 + }, + { + "epoch": 9.257188498402556, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5795 + }, + { + "epoch": 9.258785942492013, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5796 + }, + { + "epoch": 9.26038338658147, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5797 + }, + { + "epoch": 9.261980830670927, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 5798 + }, + { + "epoch": 9.263578274760384, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5799 + }, + { + "epoch": 9.26517571884984, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5800 + }, + { + "epoch": 9.266773162939296, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5801 + }, + { + "epoch": 9.268370607028753, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5802 + }, + { + "epoch": 9.26996805111821, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5803 + }, + { + "epoch": 9.271565495207668, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5804 + }, + { + "epoch": 9.273162939297125, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5805 + }, + { + "epoch": 9.274760383386582, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5806 + }, + { + "epoch": 9.276357827476039, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5807 + }, + { + "epoch": 9.277955271565496, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5808 + }, + { + "epoch": 9.279552715654953, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5809 + }, + { + "epoch": 9.281150159744408, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5810 + }, + { + "epoch": 9.282747603833865, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5811 + }, + { + "epoch": 9.284345047923322, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5812 + }, + { + "epoch": 9.28594249201278, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5813 + }, + { + "epoch": 9.287539936102236, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5814 + }, + { + "epoch": 9.289137380191693, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5815 + }, + { + "epoch": 9.29073482428115, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5816 + }, + { + "epoch": 9.292332268370608, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 5817 + }, + { + "epoch": 9.293929712460065, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5818 + }, + { + "epoch": 9.295527156549522, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5819 + }, + { + "epoch": 9.297124600638977, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5820 + }, + { + "epoch": 9.298722044728434, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5821 + }, + { + "epoch": 9.300319488817891, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5822 + }, + { + "epoch": 9.301916932907348, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5823 + }, + { + "epoch": 9.303514376996805, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 5824 + }, + { + "epoch": 9.305111821086262, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5825 + }, + { + "epoch": 9.30670926517572, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5826 + }, + { + "epoch": 9.308306709265176, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5827 + }, + { + "epoch": 9.309904153354633, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5828 + }, + { + "epoch": 9.311501597444089, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5829 + }, + { + "epoch": 9.313099041533546, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 5830 + }, + { + "epoch": 9.314696485623003, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5831 + }, + { + "epoch": 9.31629392971246, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5832 + }, + { + "epoch": 9.317891373801917, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5833 + }, + { + "epoch": 9.319488817891374, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5834 + }, + { + "epoch": 9.321086261980831, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5835 + }, + { + "epoch": 9.322683706070288, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5836 + }, + { + "epoch": 9.324281150159745, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5837 + }, + { + "epoch": 9.3258785942492, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5838 + }, + { + "epoch": 9.327476038338657, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5839 + }, + { + "epoch": 9.329073482428115, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5840 + }, + { + "epoch": 9.330670926517572, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5841 + }, + { + "epoch": 9.332268370607029, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5842 + }, + { + "epoch": 9.333865814696486, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5843 + }, + { + "epoch": 9.335463258785943, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5844 + }, + { + "epoch": 9.3370607028754, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5845 + }, + { + "epoch": 9.338658146964857, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5846 + }, + { + "epoch": 9.340255591054314, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5847 + }, + { + "epoch": 9.34185303514377, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5848 + }, + { + "epoch": 9.343450479233226, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5849 + }, + { + "epoch": 9.345047923322683, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5850 + }, + { + "epoch": 9.34664536741214, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5851 + }, + { + "epoch": 9.348242811501597, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5852 + }, + { + "epoch": 9.349840255591054, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5853 + }, + { + "epoch": 9.351437699680512, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5854 + }, + { + "epoch": 9.353035143769969, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5855 + }, + { + "epoch": 9.354632587859426, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5856 + }, + { + "epoch": 9.356230031948883, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5857 + }, + { + "epoch": 9.357827476038338, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5858 + }, + { + "epoch": 9.359424920127795, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5859 + }, + { + "epoch": 9.361022364217252, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5860 + }, + { + "epoch": 9.36261980830671, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 5861 + }, + { + "epoch": 9.364217252396166, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5862 + }, + { + "epoch": 9.365814696485623, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5863 + }, + { + "epoch": 9.36741214057508, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5864 + }, + { + "epoch": 9.369009584664537, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5865 + }, + { + "epoch": 9.370607028753994, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5866 + }, + { + "epoch": 9.37220447284345, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5867 + }, + { + "epoch": 9.373801916932907, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5868 + }, + { + "epoch": 9.375399361022364, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5869 + }, + { + "epoch": 9.37699680511182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5870 + }, + { + "epoch": 9.378594249201278, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5871 + }, + { + "epoch": 9.380191693290735, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 5872 + }, + { + "epoch": 9.381789137380192, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5873 + }, + { + "epoch": 9.383386581469649, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5874 + }, + { + "epoch": 9.384984025559106, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5875 + }, + { + "epoch": 9.386581469648561, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5876 + }, + { + "epoch": 9.388178913738018, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5877 + }, + { + "epoch": 9.389776357827476, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5878 + }, + { + "epoch": 9.391373801916933, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5879 + }, + { + "epoch": 9.39297124600639, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5880 + }, + { + "epoch": 9.394568690095847, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5881 + }, + { + "epoch": 9.396166134185304, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5882 + }, + { + "epoch": 9.39776357827476, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5883 + }, + { + "epoch": 9.399361022364218, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5884 + }, + { + "epoch": 9.400958466453675, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 5885 + }, + { + "epoch": 9.40255591054313, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5886 + }, + { + "epoch": 9.404153354632587, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5887 + }, + { + "epoch": 9.405750798722044, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 5888 + }, + { + "epoch": 9.407348242811501, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 5889 + }, + { + "epoch": 9.408945686900958, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5890 + }, + { + "epoch": 9.410543130990416, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5891 + }, + { + "epoch": 9.412140575079873, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5892 + }, + { + "epoch": 9.41373801916933, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5893 + }, + { + "epoch": 9.415335463258787, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5894 + }, + { + "epoch": 9.416932907348242, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5895 + }, + { + "epoch": 9.418530351437699, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5896 + }, + { + "epoch": 9.420127795527156, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5897 + }, + { + "epoch": 9.421725239616613, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5898 + }, + { + "epoch": 9.42332268370607, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5899 + }, + { + "epoch": 9.424920127795527, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5900 + }, + { + "epoch": 9.426517571884984, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5901 + }, + { + "epoch": 9.428115015974441, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5902 + }, + { + "epoch": 9.429712460063898, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5903 + }, + { + "epoch": 9.431309904153355, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5904 + }, + { + "epoch": 9.43290734824281, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 5905 + }, + { + "epoch": 9.434504792332268, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5906 + }, + { + "epoch": 9.436102236421725, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5907 + }, + { + "epoch": 9.437699680511182, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5908 + }, + { + "epoch": 9.439297124600639, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5909 + }, + { + "epoch": 9.440894568690096, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5910 + }, + { + "epoch": 9.442492012779553, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5911 + }, + { + "epoch": 9.44408945686901, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5912 + }, + { + "epoch": 9.445686900958467, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5913 + }, + { + "epoch": 9.447284345047922, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5914 + }, + { + "epoch": 9.44888178913738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5915 + }, + { + "epoch": 9.450479233226837, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5916 + }, + { + "epoch": 9.452076677316294, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 5917 + }, + { + "epoch": 9.45367412140575, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5918 + }, + { + "epoch": 9.455271565495208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 5919 + }, + { + "epoch": 9.456869009584665, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5920 + }, + { + "epoch": 9.458466453674122, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5921 + }, + { + "epoch": 9.460063897763579, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5922 + }, + { + "epoch": 9.461661341853036, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5923 + }, + { + "epoch": 9.463258785942491, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5924 + }, + { + "epoch": 9.464856230031948, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5925 + }, + { + "epoch": 9.466453674121405, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5926 + }, + { + "epoch": 9.468051118210862, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5927 + }, + { + "epoch": 9.46964856230032, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5928 + }, + { + "epoch": 9.471246006389777, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5929 + }, + { + "epoch": 9.472843450479234, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 5930 + }, + { + "epoch": 9.47444089456869, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5931 + }, + { + "epoch": 9.476038338658148, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5932 + }, + { + "epoch": 9.477635782747603, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5933 + }, + { + "epoch": 9.47923322683706, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5934 + }, + { + "epoch": 9.480830670926517, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5935 + }, + { + "epoch": 9.482428115015974, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5936 + }, + { + "epoch": 9.484025559105431, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5937 + }, + { + "epoch": 9.485623003194888, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5938 + }, + { + "epoch": 9.487220447284345, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5939 + }, + { + "epoch": 9.488817891373802, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5940 + }, + { + "epoch": 9.49041533546326, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 5941 + }, + { + "epoch": 9.492012779552716, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5942 + }, + { + "epoch": 9.493610223642172, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5943 + }, + { + "epoch": 9.495207667731629, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5944 + }, + { + "epoch": 9.496805111821086, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 5945 + }, + { + "epoch": 9.498402555910543, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5946 + }, + { + "epoch": 9.5, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5947 + }, + { + "epoch": 9.501597444089457, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5948 + }, + { + "epoch": 9.503194888178914, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5949 + }, + { + "epoch": 9.504792332268371, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5950 + }, + { + "epoch": 9.506389776357828, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5951 + }, + { + "epoch": 9.507987220447284, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5952 + }, + { + "epoch": 9.50958466453674, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5953 + }, + { + "epoch": 9.511182108626198, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5954 + }, + { + "epoch": 9.512779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5955 + }, + { + "epoch": 9.514376996805112, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5956 + }, + { + "epoch": 9.515974440894569, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5957 + }, + { + "epoch": 9.517571884984026, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5958 + }, + { + "epoch": 9.519169329073483, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5959 + }, + { + "epoch": 9.52076677316294, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5960 + }, + { + "epoch": 9.522364217252395, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5961 + }, + { + "epoch": 9.523961661341852, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5962 + }, + { + "epoch": 9.52555910543131, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5963 + }, + { + "epoch": 9.527156549520766, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5964 + }, + { + "epoch": 9.528753993610223, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5965 + }, + { + "epoch": 9.53035143769968, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5966 + }, + { + "epoch": 9.531948881789138, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 5967 + }, + { + "epoch": 9.533546325878595, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5968 + }, + { + "epoch": 9.535143769968052, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5969 + }, + { + "epoch": 9.536741214057509, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5970 + }, + { + "epoch": 9.538338658146966, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5971 + }, + { + "epoch": 9.539936102236421, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5972 + }, + { + "epoch": 9.541533546325878, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5973 + }, + { + "epoch": 9.543130990415335, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5974 + }, + { + "epoch": 9.544728434504792, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5975 + }, + { + "epoch": 9.54632587859425, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5976 + }, + { + "epoch": 9.547923322683706, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 5977 + }, + { + "epoch": 9.549520766773163, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5978 + }, + { + "epoch": 9.55111821086262, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5979 + }, + { + "epoch": 9.552715654952078, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5980 + }, + { + "epoch": 9.554313099041533, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5981 + }, + { + "epoch": 9.55591054313099, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5982 + }, + { + "epoch": 9.557507987220447, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5983 + }, + { + "epoch": 9.559105431309904, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5984 + }, + { + "epoch": 9.560702875399361, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5985 + }, + { + "epoch": 9.562300319488818, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5986 + }, + { + "epoch": 9.563897763578275, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5987 + }, + { + "epoch": 9.565495207667732, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5988 + }, + { + "epoch": 9.56709265175719, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5989 + }, + { + "epoch": 9.568690095846645, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5990 + }, + { + "epoch": 9.570287539936102, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5991 + }, + { + "epoch": 9.571884984025559, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5992 + }, + { + "epoch": 9.573482428115016, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 5993 + }, + { + "epoch": 9.575079872204473, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5994 + }, + { + "epoch": 9.57667731629393, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5995 + }, + { + "epoch": 9.578274760383387, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5996 + }, + { + "epoch": 9.579872204472844, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5997 + }, + { + "epoch": 9.581469648562301, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5998 + }, + { + "epoch": 9.583067092651756, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5999 + }, + { + "epoch": 9.584664536741213, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6000 + }, + { + "epoch": 9.58626198083067, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 6001 + }, + { + "epoch": 9.587859424920127, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6002 + }, + { + "epoch": 9.589456869009584, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 6003 + }, + { + "epoch": 9.591054313099042, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6004 + }, + { + "epoch": 9.592651757188499, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6005 + }, + { + "epoch": 9.594249201277956, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6006 + }, + { + "epoch": 9.595846645367413, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6007 + }, + { + "epoch": 9.59744408945687, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6008 + }, + { + "epoch": 9.599041533546325, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6009 + }, + { + "epoch": 9.600638977635782, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6010 + }, + { + "epoch": 9.60223642172524, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6011 + }, + { + "epoch": 9.603833865814696, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6012 + }, + { + "epoch": 9.605431309904153, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6013 + }, + { + "epoch": 9.60702875399361, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6014 + }, + { + "epoch": 9.608626198083067, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6015 + }, + { + "epoch": 9.610223642172524, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 6016 + }, + { + "epoch": 9.611821086261982, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 6017 + }, + { + "epoch": 9.613418530351439, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6018 + }, + { + "epoch": 9.615015974440894, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6019 + }, + { + "epoch": 9.616613418530351, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6020 + }, + { + "epoch": 9.618210862619808, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6021 + }, + { + "epoch": 9.619808306709265, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 6022 + }, + { + "epoch": 9.621405750798722, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6023 + }, + { + "epoch": 9.62300319488818, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6024 + }, + { + "epoch": 9.624600638977636, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6025 + }, + { + "epoch": 9.626198083067093, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6026 + }, + { + "epoch": 9.62779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6027 + }, + { + "epoch": 9.629392971246006, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6028 + }, + { + "epoch": 9.630990415335463, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6029 + }, + { + "epoch": 9.63258785942492, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6030 + }, + { + "epoch": 9.634185303514377, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6031 + }, + { + "epoch": 9.635782747603834, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6032 + }, + { + "epoch": 9.63738019169329, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6033 + }, + { + "epoch": 9.638977635782748, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6034 + }, + { + "epoch": 9.640575079872205, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6035 + }, + { + "epoch": 9.642172523961662, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6036 + }, + { + "epoch": 9.643769968051117, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6037 + }, + { + "epoch": 9.645367412140574, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6038 + }, + { + "epoch": 9.646964856230031, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6039 + }, + { + "epoch": 9.648562300319488, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 6040 + }, + { + "epoch": 9.650159744408946, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 6041 + }, + { + "epoch": 9.651757188498403, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6042 + }, + { + "epoch": 9.65335463258786, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6043 + }, + { + "epoch": 9.654952076677317, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6044 + }, + { + "epoch": 9.656549520766774, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6045 + }, + { + "epoch": 9.65814696485623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 6046 + }, + { + "epoch": 9.659744408945686, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6047 + }, + { + "epoch": 9.661341853035143, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6048 + }, + { + "epoch": 9.6629392971246, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6049 + }, + { + "epoch": 9.664536741214057, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6050 + }, + { + "epoch": 9.666134185303514, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6051 + }, + { + "epoch": 9.667731629392971, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6052 + }, + { + "epoch": 9.669329073482428, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6053 + }, + { + "epoch": 9.670926517571885, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6054 + }, + { + "epoch": 9.672523961661343, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 6055 + }, + { + "epoch": 9.6741214057508, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6056 + }, + { + "epoch": 9.675718849840255, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6057 + }, + { + "epoch": 9.677316293929712, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 6058 + }, + { + "epoch": 9.678913738019169, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6059 + }, + { + "epoch": 9.680511182108626, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6060 + }, + { + "epoch": 9.682108626198083, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6061 + }, + { + "epoch": 9.68370607028754, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6062 + }, + { + "epoch": 9.685303514376997, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6063 + }, + { + "epoch": 9.686900958466454, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 6064 + }, + { + "epoch": 9.688498402555911, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6065 + }, + { + "epoch": 9.690095846645367, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6066 + }, + { + "epoch": 9.691693290734824, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6067 + }, + { + "epoch": 9.69329073482428, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 6068 + }, + { + "epoch": 9.694888178913738, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6069 + }, + { + "epoch": 9.696485623003195, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6070 + }, + { + "epoch": 9.698083067092652, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6071 + }, + { + "epoch": 9.699680511182109, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6072 + }, + { + "epoch": 9.701277955271566, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6073 + }, + { + "epoch": 9.702875399361023, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6074 + }, + { + "epoch": 9.704472843450478, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6075 + }, + { + "epoch": 9.706070287539935, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6076 + }, + { + "epoch": 9.707667731629392, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6077 + }, + { + "epoch": 9.70926517571885, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6078 + }, + { + "epoch": 9.710862619808307, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6079 + }, + { + "epoch": 9.712460063897764, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6080 + }, + { + "epoch": 9.71405750798722, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6081 + }, + { + "epoch": 9.715654952076678, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6082 + }, + { + "epoch": 9.717252396166135, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6083 + }, + { + "epoch": 9.718849840255592, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 6084 + }, + { + "epoch": 9.720447284345047, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 6085 + }, + { + "epoch": 9.722044728434504, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 6086 + }, + { + "epoch": 9.723642172523961, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6087 + }, + { + "epoch": 9.725239616613418, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6088 + }, + { + "epoch": 9.726837060702875, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6089 + }, + { + "epoch": 9.728434504792332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 6090 + }, + { + "epoch": 9.73003194888179, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6091 + }, + { + "epoch": 9.731629392971247, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6092 + }, + { + "epoch": 9.733226837060704, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6093 + }, + { + "epoch": 9.73482428115016, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6094 + }, + { + "epoch": 9.736421725239616, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6095 + }, + { + "epoch": 9.738019169329073, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6096 + }, + { + "epoch": 9.73961661341853, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6097 + }, + { + "epoch": 9.741214057507987, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 6098 + }, + { + "epoch": 9.742811501597444, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6099 + }, + { + "epoch": 9.744408945686901, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6100 + }, + { + "epoch": 9.746006389776358, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6101 + }, + { + "epoch": 9.747603833865815, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6102 + }, + { + "epoch": 9.749201277955272, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6103 + }, + { + "epoch": 9.750798722044728, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6104 + }, + { + "epoch": 9.752396166134185, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6105 + }, + { + "epoch": 9.753993610223642, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6106 + }, + { + "epoch": 9.755591054313099, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 6107 + }, + { + "epoch": 9.757188498402556, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6108 + }, + { + "epoch": 9.758785942492013, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6109 + }, + { + "epoch": 9.76038338658147, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6110 + }, + { + "epoch": 9.761980830670927, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6111 + }, + { + "epoch": 9.763578274760384, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6112 + }, + { + "epoch": 9.76517571884984, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6113 + }, + { + "epoch": 9.766773162939296, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6114 + }, + { + "epoch": 9.768370607028753, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6115 + }, + { + "epoch": 9.76996805111821, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6116 + }, + { + "epoch": 9.771565495207668, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6117 + }, + { + "epoch": 9.773162939297125, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6118 + }, + { + "epoch": 9.774760383386582, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6119 + }, + { + "epoch": 9.776357827476039, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6120 + }, + { + "epoch": 9.777955271565496, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6121 + }, + { + "epoch": 9.779552715654953, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6122 + }, + { + "epoch": 9.781150159744408, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6123 + }, + { + "epoch": 9.782747603833865, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6124 + }, + { + "epoch": 9.784345047923322, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6125 + }, + { + "epoch": 9.78594249201278, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6126 + }, + { + "epoch": 9.787539936102236, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6127 + }, + { + "epoch": 9.789137380191693, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6128 + }, + { + "epoch": 9.79073482428115, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6129 + }, + { + "epoch": 9.792332268370608, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6130 + }, + { + "epoch": 9.793929712460065, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6131 + }, + { + "epoch": 9.795527156549522, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6132 + }, + { + "epoch": 9.797124600638977, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6133 + }, + { + "epoch": 9.798722044728434, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6134 + }, + { + "epoch": 9.800319488817891, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6135 + }, + { + "epoch": 9.801916932907348, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6136 + }, + { + "epoch": 9.803514376996805, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6137 + }, + { + "epoch": 9.805111821086262, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6138 + }, + { + "epoch": 9.80670926517572, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6139 + }, + { + "epoch": 9.808306709265176, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6140 + }, + { + "epoch": 9.809904153354633, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6141 + }, + { + "epoch": 9.811501597444089, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6142 + }, + { + "epoch": 9.813099041533546, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6143 + }, + { + "epoch": 9.814696485623003, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 6144 + }, + { + "epoch": 9.81629392971246, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 6145 + }, + { + "epoch": 9.817891373801917, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6146 + }, + { + "epoch": 9.819488817891374, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6147 + }, + { + "epoch": 9.821086261980831, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6148 + }, + { + "epoch": 9.822683706070288, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 6149 + }, + { + "epoch": 9.824281150159745, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6150 + }, + { + "epoch": 9.8258785942492, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6151 + }, + { + "epoch": 9.827476038338657, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6152 + }, + { + "epoch": 9.829073482428115, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6153 + }, + { + "epoch": 9.830670926517572, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6154 + }, + { + "epoch": 9.832268370607029, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6155 + }, + { + "epoch": 9.833865814696486, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6156 + }, + { + "epoch": 9.835463258785943, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6157 + }, + { + "epoch": 9.8370607028754, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6158 + }, + { + "epoch": 9.838658146964857, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6159 + }, + { + "epoch": 9.840255591054314, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6160 + }, + { + "epoch": 9.84185303514377, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6161 + }, + { + "epoch": 9.843450479233226, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6162 + }, + { + "epoch": 9.845047923322683, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6163 + }, + { + "epoch": 9.84664536741214, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6164 + }, + { + "epoch": 9.848242811501597, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 6165 + }, + { + "epoch": 9.849840255591054, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 6166 + }, + { + "epoch": 9.851437699680512, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6167 + }, + { + "epoch": 9.853035143769969, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6168 + }, + { + "epoch": 9.854632587859426, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6169 + }, + { + "epoch": 9.856230031948883, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6170 + }, + { + "epoch": 9.857827476038338, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6171 + }, + { + "epoch": 9.859424920127795, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6172 + }, + { + "epoch": 9.861022364217252, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6173 + }, + { + "epoch": 9.86261980830671, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6174 + }, + { + "epoch": 9.864217252396166, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 6175 + }, + { + "epoch": 9.865814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6176 + }, + { + "epoch": 9.86741214057508, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6177 + }, + { + "epoch": 9.869009584664537, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6178 + }, + { + "epoch": 9.870607028753994, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6179 + }, + { + "epoch": 9.87220447284345, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6180 + }, + { + "epoch": 9.873801916932907, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6181 + }, + { + "epoch": 9.875399361022364, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6182 + }, + { + "epoch": 9.87699680511182, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6183 + }, + { + "epoch": 9.878594249201278, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 6184 + }, + { + "epoch": 9.880191693290735, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6185 + }, + { + "epoch": 9.881789137380192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6186 + }, + { + "epoch": 9.883386581469649, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6187 + }, + { + "epoch": 9.884984025559106, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6188 + }, + { + "epoch": 9.886581469648561, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6189 + }, + { + "epoch": 9.888178913738018, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6190 + }, + { + "epoch": 9.889776357827476, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6191 + }, + { + "epoch": 9.891373801916933, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6192 + }, + { + "epoch": 9.89297124600639, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6193 + }, + { + "epoch": 9.894568690095847, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 6194 + }, + { + "epoch": 9.896166134185304, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6195 + }, + { + "epoch": 9.89776357827476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6196 + }, + { + "epoch": 9.899361022364218, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6197 + }, + { + "epoch": 9.900958466453675, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6198 + }, + { + "epoch": 9.90255591054313, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6199 + }, + { + "epoch": 9.904153354632587, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6200 + }, + { + "epoch": 9.905750798722044, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6201 + }, + { + "epoch": 9.907348242811501, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6202 + }, + { + "epoch": 9.908945686900958, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6203 + }, + { + "epoch": 9.910543130990416, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6204 + }, + { + "epoch": 9.912140575079873, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 6205 + }, + { + "epoch": 9.91373801916933, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6206 + }, + { + "epoch": 9.915335463258787, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6207 + }, + { + "epoch": 9.916932907348244, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6208 + }, + { + "epoch": 9.918530351437699, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 6209 + }, + { + "epoch": 9.920127795527156, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6210 + }, + { + "epoch": 9.921725239616613, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 6211 + }, + { + "epoch": 9.92332268370607, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6212 + }, + { + "epoch": 9.924920127795527, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6213 + }, + { + "epoch": 9.926517571884984, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6214 + }, + { + "epoch": 9.928115015974441, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6215 + }, + { + "epoch": 9.929712460063898, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6216 + }, + { + "epoch": 9.931309904153355, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6217 + }, + { + "epoch": 9.93290734824281, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6218 + }, + { + "epoch": 9.934504792332268, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6219 + }, + { + "epoch": 9.936102236421725, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6220 + }, + { + "epoch": 9.937699680511182, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6221 + }, + { + "epoch": 9.939297124600639, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6222 + }, + { + "epoch": 9.940894568690096, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6223 + }, + { + "epoch": 9.942492012779553, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6224 + }, + { + "epoch": 9.94408945686901, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6225 + }, + { + "epoch": 9.945686900958467, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6226 + }, + { + "epoch": 9.947284345047922, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6227 + }, + { + "epoch": 9.94888178913738, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6228 + }, + { + "epoch": 9.950479233226837, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6229 + }, + { + "epoch": 9.952076677316294, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6230 + }, + { + "epoch": 9.95367412140575, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6231 + }, + { + "epoch": 9.955271565495208, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6232 + }, + { + "epoch": 9.956869009584665, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6233 + }, + { + "epoch": 9.958466453674122, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6234 + }, + { + "epoch": 9.960063897763579, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 6235 + }, + { + "epoch": 9.961661341853034, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 6236 + }, + { + "epoch": 9.963258785942491, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6237 + }, + { + "epoch": 9.964856230031948, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6238 + }, + { + "epoch": 9.966453674121405, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6239 + }, + { + "epoch": 9.968051118210862, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6240 + }, + { + "epoch": 9.96964856230032, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6241 + }, + { + "epoch": 9.971246006389777, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6242 + }, + { + "epoch": 9.972843450479234, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6243 + }, + { + "epoch": 9.97444089456869, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6244 + }, + { + "epoch": 9.976038338658148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6245 + }, + { + "epoch": 9.977635782747605, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6246 + }, + { + "epoch": 9.97923322683706, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6247 + }, + { + "epoch": 9.980830670926517, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6248 + }, + { + "epoch": 9.982428115015974, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6249 + }, + { + "epoch": 9.984025559105431, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6250 + }, + { + "epoch": 9.985623003194888, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 6251 + }, + { + "epoch": 9.987220447284345, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6252 + }, + { + "epoch": 9.988817891373802, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6253 + }, + { + "epoch": 9.99041533546326, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6254 + }, + { + "epoch": 9.992012779552716, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 6255 + }, + { + "epoch": 9.993610223642172, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6256 + }, + { + "epoch": 9.995207667731629, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 6257 + }, + { + "epoch": 9.996805111821086, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6258 + }, + { + "epoch": 9.998402555910543, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6259 + }, + { + "epoch": 10.0, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6260 + }, + { + "epoch": 10.001597444089457, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6261 + }, + { + "epoch": 10.003194888178914, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6262 + }, + { + "epoch": 10.004792332268371, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6263 + }, + { + "epoch": 10.006389776357828, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6264 + }, + { + "epoch": 10.007987220447284, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6265 + }, + { + "epoch": 10.00958466453674, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6266 + }, + { + "epoch": 10.011182108626198, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6267 + }, + { + "epoch": 10.012779552715655, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6268 + }, + { + "epoch": 10.014376996805112, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6269 + }, + { + "epoch": 10.015974440894569, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6270 + }, + { + "epoch": 10.017571884984026, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6271 + }, + { + "epoch": 10.019169329073483, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6272 + }, + { + "epoch": 10.02076677316294, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6273 + }, + { + "epoch": 10.022364217252397, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6274 + }, + { + "epoch": 10.023961661341852, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6275 + }, + { + "epoch": 10.02555910543131, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6276 + }, + { + "epoch": 10.027156549520766, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6277 + }, + { + "epoch": 10.028753993610223, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6278 + }, + { + "epoch": 10.03035143769968, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6279 + }, + { + "epoch": 10.031948881789138, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6280 + }, + { + "epoch": 10.033546325878595, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 6281 + }, + { + "epoch": 10.035143769968052, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6282 + }, + { + "epoch": 10.036741214057509, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6283 + }, + { + "epoch": 10.038338658146964, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6284 + }, + { + "epoch": 10.039936102236421, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6285 + }, + { + "epoch": 10.041533546325878, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6286 + }, + { + "epoch": 10.043130990415335, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 6287 + }, + { + "epoch": 10.044728434504792, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6288 + }, + { + "epoch": 10.04632587859425, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6289 + }, + { + "epoch": 10.047923322683706, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 6290 + }, + { + "epoch": 10.049520766773163, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6291 + }, + { + "epoch": 10.05111821086262, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6292 + }, + { + "epoch": 10.052715654952078, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 6293 + }, + { + "epoch": 10.054313099041533, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6294 + }, + { + "epoch": 10.05591054313099, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 6295 + }, + { + "epoch": 10.057507987220447, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6296 + }, + { + "epoch": 10.059105431309904, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6297 + }, + { + "epoch": 10.060702875399361, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 6298 + }, + { + "epoch": 10.062300319488818, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6299 + }, + { + "epoch": 10.063897763578275, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6300 + }, + { + "epoch": 10.065495207667732, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6301 + }, + { + "epoch": 10.06709265175719, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6302 + }, + { + "epoch": 10.068690095846645, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6303 + }, + { + "epoch": 10.070287539936102, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6304 + }, + { + "epoch": 10.071884984025559, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 6305 + }, + { + "epoch": 10.073482428115016, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6306 + }, + { + "epoch": 10.075079872204473, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6307 + }, + { + "epoch": 10.07667731629393, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6308 + }, + { + "epoch": 10.078274760383387, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6309 + }, + { + "epoch": 10.079872204472844, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6310 + }, + { + "epoch": 10.081469648562301, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6311 + }, + { + "epoch": 10.083067092651758, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6312 + }, + { + "epoch": 10.084664536741213, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6313 + }, + { + "epoch": 10.08626198083067, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6314 + }, + { + "epoch": 10.087859424920127, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6315 + }, + { + "epoch": 10.089456869009584, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6316 + }, + { + "epoch": 10.091054313099042, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6317 + }, + { + "epoch": 10.092651757188499, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6318 + }, + { + "epoch": 10.094249201277956, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 6319 + }, + { + "epoch": 10.095846645367413, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6320 + }, + { + "epoch": 10.09744408945687, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6321 + }, + { + "epoch": 10.099041533546325, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6322 + }, + { + "epoch": 10.100638977635782, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6323 + }, + { + "epoch": 10.10223642172524, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6324 + }, + { + "epoch": 10.103833865814696, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6325 + }, + { + "epoch": 10.105431309904153, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6326 + }, + { + "epoch": 10.10702875399361, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 6327 + }, + { + "epoch": 10.108626198083067, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6328 + }, + { + "epoch": 10.110223642172524, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6329 + }, + { + "epoch": 10.111821086261982, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6330 + }, + { + "epoch": 10.113418530351439, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6331 + }, + { + "epoch": 10.115015974440894, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6332 + }, + { + "epoch": 10.116613418530351, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 6333 + }, + { + "epoch": 10.118210862619808, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6334 + }, + { + "epoch": 10.119808306709265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6335 + }, + { + "epoch": 10.121405750798722, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6336 + }, + { + "epoch": 10.12300319488818, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6337 + }, + { + "epoch": 10.124600638977636, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6338 + }, + { + "epoch": 10.126198083067093, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6339 + }, + { + "epoch": 10.12779552715655, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6340 + }, + { + "epoch": 10.129392971246006, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 6341 + }, + { + "epoch": 10.130990415335463, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6342 + }, + { + "epoch": 10.13258785942492, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6343 + }, + { + "epoch": 10.134185303514377, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6344 + }, + { + "epoch": 10.135782747603834, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6345 + }, + { + "epoch": 10.13738019169329, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6346 + }, + { + "epoch": 10.138977635782748, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6347 + }, + { + "epoch": 10.140575079872205, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6348 + }, + { + "epoch": 10.142172523961662, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6349 + }, + { + "epoch": 10.143769968051119, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6350 + }, + { + "epoch": 10.145367412140574, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6351 + }, + { + "epoch": 10.146964856230031, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6352 + }, + { + "epoch": 10.148562300319488, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6353 + }, + { + "epoch": 10.150159744408946, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6354 + }, + { + "epoch": 10.151757188498403, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6355 + }, + { + "epoch": 10.15335463258786, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6356 + }, + { + "epoch": 10.154952076677317, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6357 + }, + { + "epoch": 10.156549520766774, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6358 + }, + { + "epoch": 10.15814696485623, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6359 + }, + { + "epoch": 10.159744408945686, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6360 + }, + { + "epoch": 10.161341853035143, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6361 + }, + { + "epoch": 10.1629392971246, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6362 + }, + { + "epoch": 10.164536741214057, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6363 + }, + { + "epoch": 10.166134185303514, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6364 + }, + { + "epoch": 10.167731629392971, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6365 + }, + { + "epoch": 10.169329073482428, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6366 + }, + { + "epoch": 10.170926517571885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6367 + }, + { + "epoch": 10.172523961661343, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6368 + }, + { + "epoch": 10.1741214057508, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6369 + }, + { + "epoch": 10.175718849840255, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6370 + }, + { + "epoch": 10.177316293929712, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6371 + }, + { + "epoch": 10.178913738019169, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6372 + }, + { + "epoch": 10.180511182108626, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6373 + }, + { + "epoch": 10.182108626198083, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6374 + }, + { + "epoch": 10.18370607028754, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6375 + }, + { + "epoch": 10.185303514376997, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6376 + }, + { + "epoch": 10.186900958466454, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6377 + }, + { + "epoch": 10.188498402555911, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6378 + }, + { + "epoch": 10.190095846645367, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6379 + }, + { + "epoch": 10.191693290734824, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6380 + }, + { + "epoch": 10.19329073482428, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6381 + }, + { + "epoch": 10.194888178913738, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 6382 + }, + { + "epoch": 10.196485623003195, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6383 + }, + { + "epoch": 10.198083067092652, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6384 + }, + { + "epoch": 10.199680511182109, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6385 + }, + { + "epoch": 10.201277955271566, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6386 + }, + { + "epoch": 10.202875399361023, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6387 + }, + { + "epoch": 10.204472843450478, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6388 + }, + { + "epoch": 10.206070287539935, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6389 + }, + { + "epoch": 10.207667731629392, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6390 + }, + { + "epoch": 10.20926517571885, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6391 + }, + { + "epoch": 10.210862619808307, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 6392 + }, + { + "epoch": 10.212460063897764, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 6393 + }, + { + "epoch": 10.21405750798722, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6394 + }, + { + "epoch": 10.215654952076678, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6395 + }, + { + "epoch": 10.217252396166135, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6396 + }, + { + "epoch": 10.218849840255592, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6397 + }, + { + "epoch": 10.220447284345047, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 6398 + }, + { + "epoch": 10.222044728434504, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6399 + }, + { + "epoch": 10.223642172523961, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6400 + }, + { + "epoch": 10.225239616613418, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6401 + }, + { + "epoch": 10.226837060702875, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6402 + }, + { + "epoch": 10.228434504792332, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6403 + }, + { + "epoch": 10.23003194888179, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6404 + }, + { + "epoch": 10.231629392971247, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6405 + }, + { + "epoch": 10.233226837060704, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6406 + }, + { + "epoch": 10.23482428115016, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6407 + }, + { + "epoch": 10.236421725239616, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6408 + }, + { + "epoch": 10.238019169329073, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6409 + }, + { + "epoch": 10.23961661341853, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6410 + }, + { + "epoch": 10.241214057507987, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6411 + }, + { + "epoch": 10.242811501597444, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6412 + }, + { + "epoch": 10.244408945686901, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 6413 + }, + { + "epoch": 10.246006389776358, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6414 + }, + { + "epoch": 10.247603833865815, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6415 + }, + { + "epoch": 10.249201277955272, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6416 + }, + { + "epoch": 10.250798722044728, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6417 + }, + { + "epoch": 10.252396166134185, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6418 + }, + { + "epoch": 10.253993610223642, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6419 + }, + { + "epoch": 10.255591054313099, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6420 + }, + { + "epoch": 10.257188498402556, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6421 + }, + { + "epoch": 10.258785942492013, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6422 + }, + { + "epoch": 10.26038338658147, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6423 + }, + { + "epoch": 10.261980830670927, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6424 + }, + { + "epoch": 10.263578274760384, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6425 + }, + { + "epoch": 10.26517571884984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6426 + }, + { + "epoch": 10.266773162939296, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6427 + }, + { + "epoch": 10.268370607028753, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6428 + }, + { + "epoch": 10.26996805111821, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 6429 + }, + { + "epoch": 10.271565495207668, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6430 + }, + { + "epoch": 10.273162939297125, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6431 + }, + { + "epoch": 10.274760383386582, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6432 + }, + { + "epoch": 10.276357827476039, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 6433 + }, + { + "epoch": 10.277955271565496, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6434 + }, + { + "epoch": 10.279552715654953, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6435 + }, + { + "epoch": 10.281150159744408, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6436 + }, + { + "epoch": 10.282747603833865, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6437 + }, + { + "epoch": 10.284345047923322, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6438 + }, + { + "epoch": 10.28594249201278, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6439 + }, + { + "epoch": 10.287539936102236, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6440 + }, + { + "epoch": 10.289137380191693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6441 + }, + { + "epoch": 10.29073482428115, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6442 + }, + { + "epoch": 10.292332268370608, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6443 + }, + { + "epoch": 10.293929712460065, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6444 + }, + { + "epoch": 10.295527156549522, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6445 + }, + { + "epoch": 10.297124600638977, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 6446 + }, + { + "epoch": 10.298722044728434, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6447 + }, + { + "epoch": 10.300319488817891, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6448 + }, + { + "epoch": 10.301916932907348, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 6449 + }, + { + "epoch": 10.303514376996805, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6450 + }, + { + "epoch": 10.305111821086262, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6451 + }, + { + "epoch": 10.30670926517572, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6452 + }, + { + "epoch": 10.308306709265176, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6453 + }, + { + "epoch": 10.309904153354633, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6454 + }, + { + "epoch": 10.311501597444089, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6455 + }, + { + "epoch": 10.313099041533546, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6456 + }, + { + "epoch": 10.314696485623003, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6457 + }, + { + "epoch": 10.31629392971246, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6458 + }, + { + "epoch": 10.317891373801917, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6459 + }, + { + "epoch": 10.319488817891374, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6460 + }, + { + "epoch": 10.321086261980831, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6461 + }, + { + "epoch": 10.322683706070288, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6462 + }, + { + "epoch": 10.324281150159745, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6463 + }, + { + "epoch": 10.3258785942492, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 6464 + }, + { + "epoch": 10.327476038338657, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6465 + }, + { + "epoch": 10.329073482428115, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6466 + }, + { + "epoch": 10.330670926517572, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6467 + }, + { + "epoch": 10.332268370607029, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6468 + }, + { + "epoch": 10.333865814696486, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6469 + }, + { + "epoch": 10.335463258785943, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 6470 + }, + { + "epoch": 10.3370607028754, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6471 + }, + { + "epoch": 10.338658146964857, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6472 + }, + { + "epoch": 10.340255591054314, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6473 + }, + { + "epoch": 10.34185303514377, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6474 + }, + { + "epoch": 10.343450479233226, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6475 + }, + { + "epoch": 10.345047923322683, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6476 + }, + { + "epoch": 10.34664536741214, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6477 + }, + { + "epoch": 10.348242811501597, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6478 + }, + { + "epoch": 10.349840255591054, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6479 + }, + { + "epoch": 10.351437699680512, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6480 + }, + { + "epoch": 10.353035143769969, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6481 + }, + { + "epoch": 10.354632587859426, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 6482 + }, + { + "epoch": 10.356230031948883, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6483 + }, + { + "epoch": 10.357827476038338, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6484 + }, + { + "epoch": 10.359424920127795, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6485 + }, + { + "epoch": 10.361022364217252, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6486 + }, + { + "epoch": 10.36261980830671, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6487 + }, + { + "epoch": 10.364217252396166, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6488 + }, + { + "epoch": 10.365814696485623, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6489 + }, + { + "epoch": 10.36741214057508, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6490 + }, + { + "epoch": 10.369009584664537, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 6491 + }, + { + "epoch": 10.370607028753994, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6492 + }, + { + "epoch": 10.37220447284345, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6493 + }, + { + "epoch": 10.373801916932907, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 6494 + }, + { + "epoch": 10.375399361022364, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 6495 + }, + { + "epoch": 10.37699680511182, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6496 + }, + { + "epoch": 10.378594249201278, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 6497 + }, + { + "epoch": 10.380191693290735, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6498 + }, + { + "epoch": 10.381789137380192, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6499 + }, + { + "epoch": 10.383386581469649, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 6500 + }, + { + "epoch": 10.384984025559106, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6501 + }, + { + "epoch": 10.386581469648561, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6502 + }, + { + "epoch": 10.388178913738018, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6503 + }, + { + "epoch": 10.389776357827476, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6504 + }, + { + "epoch": 10.391373801916933, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6505 + }, + { + "epoch": 10.39297124600639, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6506 + }, + { + "epoch": 10.394568690095847, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6507 + }, + { + "epoch": 10.396166134185304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6508 + }, + { + "epoch": 10.39776357827476, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6509 + }, + { + "epoch": 10.399361022364218, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6510 + }, + { + "epoch": 10.400958466453675, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 6511 + }, + { + "epoch": 10.40255591054313, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6512 + }, + { + "epoch": 10.404153354632587, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6513 + }, + { + "epoch": 10.405750798722044, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6514 + }, + { + "epoch": 10.407348242811501, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6515 + }, + { + "epoch": 10.408945686900958, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6516 + }, + { + "epoch": 10.410543130990416, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 6517 + }, + { + "epoch": 10.412140575079873, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6518 + }, + { + "epoch": 10.41373801916933, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6519 + }, + { + "epoch": 10.415335463258787, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6520 + }, + { + "epoch": 10.416932907348242, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6521 + }, + { + "epoch": 10.418530351437699, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6522 + }, + { + "epoch": 10.420127795527156, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6523 + }, + { + "epoch": 10.421725239616613, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6524 + }, + { + "epoch": 10.42332268370607, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6525 + }, + { + "epoch": 10.424920127795527, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 6526 + }, + { + "epoch": 10.426517571884984, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6527 + }, + { + "epoch": 10.428115015974441, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6528 + }, + { + "epoch": 10.429712460063898, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6529 + }, + { + "epoch": 10.431309904153355, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6530 + }, + { + "epoch": 10.43290734824281, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6531 + }, + { + "epoch": 10.434504792332268, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 6532 + }, + { + "epoch": 10.436102236421725, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6533 + }, + { + "epoch": 10.437699680511182, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6534 + }, + { + "epoch": 10.439297124600639, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6535 + }, + { + "epoch": 10.440894568690096, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6536 + }, + { + "epoch": 10.442492012779553, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6537 + }, + { + "epoch": 10.44408945686901, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6538 + }, + { + "epoch": 10.445686900958467, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6539 + }, + { + "epoch": 10.447284345047922, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6540 + }, + { + "epoch": 10.44888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6541 + }, + { + "epoch": 10.450479233226837, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6542 + }, + { + "epoch": 10.452076677316294, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6543 + }, + { + "epoch": 10.45367412140575, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6544 + }, + { + "epoch": 10.455271565495208, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 6545 + }, + { + "epoch": 10.456869009584665, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6546 + }, + { + "epoch": 10.458466453674122, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 6547 + }, + { + "epoch": 10.460063897763579, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6548 + }, + { + "epoch": 10.461661341853036, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6549 + }, + { + "epoch": 10.463258785942491, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6550 + }, + { + "epoch": 10.464856230031948, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6551 + }, + { + "epoch": 10.466453674121405, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6552 + }, + { + "epoch": 10.468051118210862, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 6553 + }, + { + "epoch": 10.46964856230032, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6554 + }, + { + "epoch": 10.471246006389777, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6555 + }, + { + "epoch": 10.472843450479234, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6556 + }, + { + "epoch": 10.47444089456869, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6557 + }, + { + "epoch": 10.476038338658148, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6558 + }, + { + "epoch": 10.477635782747603, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6559 + }, + { + "epoch": 10.47923322683706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6560 + }, + { + "epoch": 10.480830670926517, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6561 + }, + { + "epoch": 10.482428115015974, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6562 + }, + { + "epoch": 10.484025559105431, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 6563 + }, + { + "epoch": 10.485623003194888, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6564 + }, + { + "epoch": 10.487220447284345, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6565 + }, + { + "epoch": 10.488817891373802, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6566 + }, + { + "epoch": 10.49041533546326, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6567 + }, + { + "epoch": 10.492012779552716, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6568 + }, + { + "epoch": 10.493610223642172, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6569 + }, + { + "epoch": 10.495207667731629, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6570 + }, + { + "epoch": 10.496805111821086, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6571 + }, + { + "epoch": 10.498402555910543, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6572 + }, + { + "epoch": 10.5, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6573 + }, + { + "epoch": 10.501597444089457, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 6574 + }, + { + "epoch": 10.503194888178914, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6575 + }, + { + "epoch": 10.504792332268371, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6576 + }, + { + "epoch": 10.506389776357828, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6577 + }, + { + "epoch": 10.507987220447284, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6578 + }, + { + "epoch": 10.50958466453674, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6579 + }, + { + "epoch": 10.511182108626198, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6580 + }, + { + "epoch": 10.512779552715655, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6581 + }, + { + "epoch": 10.514376996805112, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6582 + }, + { + "epoch": 10.515974440894569, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 6583 + }, + { + "epoch": 10.517571884984026, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6584 + }, + { + "epoch": 10.519169329073483, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6585 + }, + { + "epoch": 10.52076677316294, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6586 + }, + { + "epoch": 10.522364217252395, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6587 + }, + { + "epoch": 10.523961661341852, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6588 + }, + { + "epoch": 10.52555910543131, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 6589 + }, + { + "epoch": 10.527156549520766, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6590 + }, + { + "epoch": 10.528753993610223, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6591 + }, + { + "epoch": 10.53035143769968, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6592 + }, + { + "epoch": 10.531948881789138, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6593 + }, + { + "epoch": 10.533546325878595, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6594 + }, + { + "epoch": 10.535143769968052, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6595 + }, + { + "epoch": 10.536741214057509, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6596 + }, + { + "epoch": 10.538338658146966, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6597 + }, + { + "epoch": 10.539936102236421, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6598 + }, + { + "epoch": 10.541533546325878, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6599 + }, + { + "epoch": 10.543130990415335, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6600 + }, + { + "epoch": 10.544728434504792, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6601 + }, + { + "epoch": 10.54632587859425, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6602 + }, + { + "epoch": 10.547923322683706, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6603 + }, + { + "epoch": 10.549520766773163, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 6604 + }, + { + "epoch": 10.55111821086262, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6605 + }, + { + "epoch": 10.552715654952078, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 6606 + }, + { + "epoch": 10.554313099041533, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6607 + }, + { + "epoch": 10.55591054313099, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6608 + }, + { + "epoch": 10.557507987220447, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 6609 + }, + { + "epoch": 10.559105431309904, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6610 + }, + { + "epoch": 10.560702875399361, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6611 + }, + { + "epoch": 10.562300319488818, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6612 + }, + { + "epoch": 10.563897763578275, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6613 + }, + { + "epoch": 10.565495207667732, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6614 + }, + { + "epoch": 10.56709265175719, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6615 + }, + { + "epoch": 10.568690095846645, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 6616 + }, + { + "epoch": 10.570287539936102, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6617 + }, + { + "epoch": 10.571884984025559, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 6618 + }, + { + "epoch": 10.573482428115016, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6619 + }, + { + "epoch": 10.575079872204473, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6620 + }, + { + "epoch": 10.57667731629393, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6621 + }, + { + "epoch": 10.578274760383387, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 6622 + }, + { + "epoch": 10.579872204472844, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 6623 + }, + { + "epoch": 10.581469648562301, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6624 + }, + { + "epoch": 10.583067092651756, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6625 + }, + { + "epoch": 10.584664536741213, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6626 + }, + { + "epoch": 10.58626198083067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6627 + }, + { + "epoch": 10.587859424920127, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6628 + }, + { + "epoch": 10.589456869009584, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6629 + }, + { + "epoch": 10.591054313099042, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6630 + }, + { + "epoch": 10.592651757188499, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6631 + }, + { + "epoch": 10.594249201277956, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 6632 + }, + { + "epoch": 10.595846645367413, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6633 + }, + { + "epoch": 10.59744408945687, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6634 + }, + { + "epoch": 10.599041533546325, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6635 + }, + { + "epoch": 10.600638977635782, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6636 + }, + { + "epoch": 10.60223642172524, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6637 + }, + { + "epoch": 10.603833865814696, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6638 + }, + { + "epoch": 10.605431309904153, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6639 + }, + { + "epoch": 10.60702875399361, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6640 + }, + { + "epoch": 10.608626198083067, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6641 + }, + { + "epoch": 10.610223642172524, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6642 + }, + { + "epoch": 10.611821086261982, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6643 + }, + { + "epoch": 10.613418530351439, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6644 + }, + { + "epoch": 10.615015974440894, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6645 + }, + { + "epoch": 10.616613418530351, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6646 + }, + { + "epoch": 10.618210862619808, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6647 + }, + { + "epoch": 10.619808306709265, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6648 + }, + { + "epoch": 10.621405750798722, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6649 + }, + { + "epoch": 10.62300319488818, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6650 + }, + { + "epoch": 10.624600638977636, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 6651 + }, + { + "epoch": 10.626198083067093, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6652 + }, + { + "epoch": 10.62779552715655, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6653 + }, + { + "epoch": 10.629392971246006, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6654 + }, + { + "epoch": 10.630990415335463, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6655 + }, + { + "epoch": 10.63258785942492, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6656 + }, + { + "epoch": 10.634185303514377, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6657 + }, + { + "epoch": 10.635782747603834, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6658 + }, + { + "epoch": 10.63738019169329, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 6659 + }, + { + "epoch": 10.638977635782748, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6660 + }, + { + "epoch": 10.640575079872205, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6661 + }, + { + "epoch": 10.642172523961662, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 6662 + }, + { + "epoch": 10.643769968051117, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6663 + }, + { + "epoch": 10.645367412140574, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6664 + }, + { + "epoch": 10.646964856230031, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6665 + }, + { + "epoch": 10.648562300319488, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6666 + }, + { + "epoch": 10.650159744408946, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6667 + }, + { + "epoch": 10.651757188498403, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6668 + }, + { + "epoch": 10.65335463258786, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6669 + }, + { + "epoch": 10.654952076677317, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6670 + }, + { + "epoch": 10.656549520766774, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6671 + }, + { + "epoch": 10.65814696485623, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6672 + }, + { + "epoch": 10.659744408945686, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6673 + }, + { + "epoch": 10.661341853035143, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6674 + }, + { + "epoch": 10.6629392971246, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6675 + }, + { + "epoch": 10.664536741214057, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6676 + }, + { + "epoch": 10.666134185303514, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6677 + }, + { + "epoch": 10.667731629392971, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6678 + }, + { + "epoch": 10.669329073482428, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6679 + }, + { + "epoch": 10.670926517571885, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6680 + }, + { + "epoch": 10.672523961661343, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 6681 + }, + { + "epoch": 10.6741214057508, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6682 + }, + { + "epoch": 10.675718849840255, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6683 + }, + { + "epoch": 10.677316293929712, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6684 + }, + { + "epoch": 10.678913738019169, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6685 + }, + { + "epoch": 10.680511182108626, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6686 + }, + { + "epoch": 10.682108626198083, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6687 + }, + { + "epoch": 10.68370607028754, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6688 + }, + { + "epoch": 10.685303514376997, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6689 + }, + { + "epoch": 10.686900958466454, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6690 + }, + { + "epoch": 10.688498402555911, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6691 + }, + { + "epoch": 10.690095846645367, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6692 + }, + { + "epoch": 10.691693290734824, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6693 + }, + { + "epoch": 10.69329073482428, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6694 + }, + { + "epoch": 10.694888178913738, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6695 + }, + { + "epoch": 10.696485623003195, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6696 + }, + { + "epoch": 10.698083067092652, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6697 + }, + { + "epoch": 10.699680511182109, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6698 + }, + { + "epoch": 10.701277955271566, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 6699 + }, + { + "epoch": 10.702875399361023, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6700 + }, + { + "epoch": 10.704472843450478, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6701 + }, + { + "epoch": 10.706070287539935, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6702 + }, + { + "epoch": 10.707667731629392, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6703 + }, + { + "epoch": 10.70926517571885, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6704 + }, + { + "epoch": 10.710862619808307, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6705 + }, + { + "epoch": 10.712460063897764, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6706 + }, + { + "epoch": 10.71405750798722, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6707 + }, + { + "epoch": 10.715654952076678, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6708 + }, + { + "epoch": 10.717252396166135, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6709 + }, + { + "epoch": 10.718849840255592, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6710 + }, + { + "epoch": 10.720447284345047, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6711 + }, + { + "epoch": 10.722044728434504, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6712 + }, + { + "epoch": 10.723642172523961, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6713 + }, + { + "epoch": 10.725239616613418, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6714 + }, + { + "epoch": 10.726837060702875, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6715 + }, + { + "epoch": 10.728434504792332, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6716 + }, + { + "epoch": 10.73003194888179, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 6717 + }, + { + "epoch": 10.731629392971247, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6718 + }, + { + "epoch": 10.733226837060704, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6719 + }, + { + "epoch": 10.73482428115016, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6720 + }, + { + "epoch": 10.736421725239616, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6721 + }, + { + "epoch": 10.738019169329073, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6722 + }, + { + "epoch": 10.73961661341853, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 6723 + }, + { + "epoch": 10.741214057507987, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6724 + }, + { + "epoch": 10.742811501597444, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6725 + }, + { + "epoch": 10.744408945686901, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6726 + }, + { + "epoch": 10.746006389776358, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6727 + }, + { + "epoch": 10.747603833865815, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6728 + }, + { + "epoch": 10.749201277955272, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6729 + }, + { + "epoch": 10.750798722044728, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6730 + }, + { + "epoch": 10.752396166134185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6731 + }, + { + "epoch": 10.753993610223642, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6732 + }, + { + "epoch": 10.755591054313099, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 6733 + }, + { + "epoch": 10.757188498402556, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6734 + }, + { + "epoch": 10.758785942492013, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6735 + }, + { + "epoch": 10.76038338658147, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6736 + }, + { + "epoch": 10.761980830670927, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6737 + }, + { + "epoch": 10.763578274760384, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6738 + }, + { + "epoch": 10.76517571884984, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6739 + }, + { + "epoch": 10.766773162939296, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6740 + }, + { + "epoch": 10.768370607028753, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6741 + }, + { + "epoch": 10.76996805111821, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6742 + }, + { + "epoch": 10.771565495207668, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6743 + }, + { + "epoch": 10.773162939297125, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6744 + }, + { + "epoch": 10.774760383386582, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 6745 + }, + { + "epoch": 10.776357827476039, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6746 + }, + { + "epoch": 10.777955271565496, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6747 + }, + { + "epoch": 10.779552715654953, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6748 + }, + { + "epoch": 10.781150159744408, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6749 + }, + { + "epoch": 10.782747603833865, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6750 + }, + { + "epoch": 10.784345047923322, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6751 + }, + { + "epoch": 10.78594249201278, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 6752 + }, + { + "epoch": 10.787539936102236, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6753 + }, + { + "epoch": 10.789137380191693, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6754 + }, + { + "epoch": 10.79073482428115, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6755 + }, + { + "epoch": 10.792332268370608, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6756 + }, + { + "epoch": 10.793929712460065, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 6757 + }, + { + "epoch": 10.795527156549522, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6758 + }, + { + "epoch": 10.797124600638977, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6759 + }, + { + "epoch": 10.798722044728434, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6760 + }, + { + "epoch": 10.800319488817891, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6761 + }, + { + "epoch": 10.801916932907348, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6762 + }, + { + "epoch": 10.803514376996805, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6763 + }, + { + "epoch": 10.805111821086262, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6764 + }, + { + "epoch": 10.80670926517572, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6765 + }, + { + "epoch": 10.808306709265176, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6766 + }, + { + "epoch": 10.809904153354633, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6767 + }, + { + "epoch": 10.811501597444089, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6768 + }, + { + "epoch": 10.813099041533546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6769 + }, + { + "epoch": 10.814696485623003, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6770 + }, + { + "epoch": 10.81629392971246, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 6771 + }, + { + "epoch": 10.817891373801917, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6772 + }, + { + "epoch": 10.819488817891374, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6773 + }, + { + "epoch": 10.821086261980831, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6774 + }, + { + "epoch": 10.822683706070288, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6775 + }, + { + "epoch": 10.824281150159745, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6776 + }, + { + "epoch": 10.8258785942492, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6777 + }, + { + "epoch": 10.827476038338657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6778 + }, + { + "epoch": 10.829073482428115, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6779 + }, + { + "epoch": 10.830670926517572, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6780 + }, + { + "epoch": 10.832268370607029, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6781 + }, + { + "epoch": 10.833865814696486, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6782 + }, + { + "epoch": 10.835463258785943, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6783 + }, + { + "epoch": 10.8370607028754, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6784 + }, + { + "epoch": 10.838658146964857, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6785 + }, + { + "epoch": 10.840255591054314, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6786 + }, + { + "epoch": 10.84185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6787 + }, + { + "epoch": 10.843450479233226, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6788 + }, + { + "epoch": 10.845047923322683, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6789 + }, + { + "epoch": 10.84664536741214, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 6790 + }, + { + "epoch": 10.848242811501597, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6791 + }, + { + "epoch": 10.849840255591054, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 6792 + }, + { + "epoch": 10.851437699680512, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6793 + }, + { + "epoch": 10.853035143769969, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6794 + }, + { + "epoch": 10.854632587859426, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6795 + }, + { + "epoch": 10.856230031948883, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6796 + }, + { + "epoch": 10.857827476038338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6797 + }, + { + "epoch": 10.859424920127795, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6798 + }, + { + "epoch": 10.861022364217252, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6799 + }, + { + "epoch": 10.86261980830671, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 6800 + }, + { + "epoch": 10.864217252396166, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6801 + }, + { + "epoch": 10.865814696485623, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6802 + }, + { + "epoch": 10.86741214057508, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 6803 + }, + { + "epoch": 10.869009584664537, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6804 + }, + { + "epoch": 10.870607028753994, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6805 + }, + { + "epoch": 10.87220447284345, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6806 + }, + { + "epoch": 10.873801916932907, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6807 + }, + { + "epoch": 10.875399361022364, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6808 + }, + { + "epoch": 10.87699680511182, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 6809 + }, + { + "epoch": 10.878594249201278, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6810 + }, + { + "epoch": 10.880191693290735, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6811 + }, + { + "epoch": 10.881789137380192, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6812 + }, + { + "epoch": 10.883386581469649, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6813 + }, + { + "epoch": 10.884984025559106, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6814 + }, + { + "epoch": 10.886581469648561, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6815 + }, + { + "epoch": 10.888178913738018, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6816 + }, + { + "epoch": 10.889776357827476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6817 + }, + { + "epoch": 10.891373801916933, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6818 + }, + { + "epoch": 10.89297124600639, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6819 + }, + { + "epoch": 10.894568690095847, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 6820 + }, + { + "epoch": 10.896166134185304, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6821 + }, + { + "epoch": 10.89776357827476, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 6822 + }, + { + "epoch": 10.899361022364218, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6823 + }, + { + "epoch": 10.900958466453675, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6824 + }, + { + "epoch": 10.90255591054313, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6825 + }, + { + "epoch": 10.904153354632587, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6826 + }, + { + "epoch": 10.905750798722044, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6827 + }, + { + "epoch": 10.907348242811501, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6828 + }, + { + "epoch": 10.908945686900958, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6829 + }, + { + "epoch": 10.910543130990416, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6830 + }, + { + "epoch": 10.912140575079873, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6831 + }, + { + "epoch": 10.91373801916933, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6832 + }, + { + "epoch": 10.915335463258787, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6833 + }, + { + "epoch": 10.916932907348244, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6834 + }, + { + "epoch": 10.918530351437699, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6835 + }, + { + "epoch": 10.920127795527156, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6836 + }, + { + "epoch": 10.921725239616613, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 6837 + }, + { + "epoch": 10.92332268370607, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6838 + }, + { + "epoch": 10.924920127795527, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6839 + }, + { + "epoch": 10.926517571884984, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6840 + }, + { + "epoch": 10.928115015974441, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6841 + }, + { + "epoch": 10.929712460063898, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6842 + }, + { + "epoch": 10.931309904153355, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6843 + }, + { + "epoch": 10.93290734824281, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6844 + }, + { + "epoch": 10.934504792332268, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6845 + }, + { + "epoch": 10.936102236421725, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6846 + }, + { + "epoch": 10.937699680511182, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6847 + }, + { + "epoch": 10.939297124600639, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6848 + }, + { + "epoch": 10.940894568690096, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 6849 + }, + { + "epoch": 10.942492012779553, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6850 + }, + { + "epoch": 10.94408945686901, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6851 + }, + { + "epoch": 10.945686900958467, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6852 + }, + { + "epoch": 10.947284345047922, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6853 + }, + { + "epoch": 10.94888178913738, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6854 + }, + { + "epoch": 10.950479233226837, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6855 + }, + { + "epoch": 10.952076677316294, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6856 + }, + { + "epoch": 10.95367412140575, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6857 + }, + { + "epoch": 10.955271565495208, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6858 + }, + { + "epoch": 10.956869009584665, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6859 + }, + { + "epoch": 10.958466453674122, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6860 + }, + { + "epoch": 10.960063897763579, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6861 + }, + { + "epoch": 10.961661341853034, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6862 + }, + { + "epoch": 10.963258785942491, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6863 + }, + { + "epoch": 10.964856230031948, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6864 + }, + { + "epoch": 10.966453674121405, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6865 + }, + { + "epoch": 10.968051118210862, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6866 + }, + { + "epoch": 10.96964856230032, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 6867 + }, + { + "epoch": 10.971246006389777, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6868 + }, + { + "epoch": 10.972843450479234, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 6869 + }, + { + "epoch": 10.97444089456869, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6870 + }, + { + "epoch": 10.976038338658148, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6871 + }, + { + "epoch": 10.977635782747605, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6872 + }, + { + "epoch": 10.97923322683706, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6873 + }, + { + "epoch": 10.980830670926517, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 6874 + }, + { + "epoch": 10.982428115015974, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 6875 + }, + { + "epoch": 10.984025559105431, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6876 + }, + { + "epoch": 10.985623003194888, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6877 + }, + { + "epoch": 10.987220447284345, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 6878 + }, + { + "epoch": 10.988817891373802, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6879 + }, + { + "epoch": 10.99041533546326, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6880 + }, + { + "epoch": 10.992012779552716, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6881 + }, + { + "epoch": 10.993610223642172, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6882 + }, + { + "epoch": 10.995207667731629, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6883 + }, + { + "epoch": 10.996805111821086, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6884 + }, + { + "epoch": 10.998402555910543, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6885 + }, + { + "epoch": 11.0, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6886 + }, + { + "epoch": 11.001597444089457, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6887 + }, + { + "epoch": 11.003194888178914, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6888 + }, + { + "epoch": 11.004792332268371, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 6889 + }, + { + "epoch": 11.006389776357828, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6890 + }, + { + "epoch": 11.007987220447284, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6891 + }, + { + "epoch": 11.00958466453674, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6892 + }, + { + "epoch": 11.011182108626198, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6893 + }, + { + "epoch": 11.012779552715655, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6894 + }, + { + "epoch": 11.014376996805112, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6895 + }, + { + "epoch": 11.015974440894569, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 6896 + }, + { + "epoch": 11.017571884984026, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6897 + }, + { + "epoch": 11.019169329073483, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6898 + }, + { + "epoch": 11.02076677316294, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6899 + }, + { + "epoch": 11.022364217252397, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6900 + }, + { + "epoch": 11.023961661341852, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6901 + }, + { + "epoch": 11.02555910543131, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6902 + }, + { + "epoch": 11.027156549520766, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6903 + }, + { + "epoch": 11.028753993610223, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6904 + }, + { + "epoch": 11.03035143769968, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6905 + }, + { + "epoch": 11.031948881789138, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6906 + }, + { + "epoch": 11.033546325878595, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6907 + }, + { + "epoch": 11.035143769968052, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6908 + }, + { + "epoch": 11.036741214057509, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 6909 + }, + { + "epoch": 11.038338658146964, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6910 + }, + { + "epoch": 11.039936102236421, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 6911 + }, + { + "epoch": 11.041533546325878, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6912 + }, + { + "epoch": 11.043130990415335, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6913 + }, + { + "epoch": 11.044728434504792, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6914 + }, + { + "epoch": 11.04632587859425, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 6915 + }, + { + "epoch": 11.047923322683706, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6916 + }, + { + "epoch": 11.049520766773163, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6917 + }, + { + "epoch": 11.05111821086262, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 6918 + }, + { + "epoch": 11.052715654952078, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6919 + }, + { + "epoch": 11.054313099041533, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6920 + }, + { + "epoch": 11.05591054313099, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6921 + }, + { + "epoch": 11.057507987220447, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6922 + }, + { + "epoch": 11.059105431309904, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6923 + }, + { + "epoch": 11.060702875399361, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6924 + }, + { + "epoch": 11.062300319488818, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6925 + }, + { + "epoch": 11.063897763578275, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6926 + }, + { + "epoch": 11.065495207667732, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6927 + }, + { + "epoch": 11.06709265175719, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6928 + }, + { + "epoch": 11.068690095846645, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6929 + }, + { + "epoch": 11.070287539936102, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6930 + }, + { + "epoch": 11.071884984025559, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6931 + }, + { + "epoch": 11.073482428115016, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6932 + }, + { + "epoch": 11.075079872204473, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6933 + }, + { + "epoch": 11.07667731629393, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6934 + }, + { + "epoch": 11.078274760383387, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6935 + }, + { + "epoch": 11.079872204472844, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6936 + }, + { + "epoch": 11.081469648562301, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6937 + }, + { + "epoch": 11.083067092651758, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6938 + }, + { + "epoch": 11.084664536741213, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6939 + }, + { + "epoch": 11.08626198083067, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6940 + }, + { + "epoch": 11.087859424920127, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6941 + }, + { + "epoch": 11.089456869009584, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6942 + }, + { + "epoch": 11.091054313099042, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6943 + }, + { + "epoch": 11.092651757188499, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 6944 + }, + { + "epoch": 11.094249201277956, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6945 + }, + { + "epoch": 11.095846645367413, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6946 + }, + { + "epoch": 11.09744408945687, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6947 + }, + { + "epoch": 11.099041533546325, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6948 + }, + { + "epoch": 11.100638977635782, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6949 + }, + { + "epoch": 11.10223642172524, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 6950 + }, + { + "epoch": 11.103833865814696, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6951 + }, + { + "epoch": 11.105431309904153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6952 + }, + { + "epoch": 11.10702875399361, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6953 + }, + { + "epoch": 11.108626198083067, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6954 + }, + { + "epoch": 11.110223642172524, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 6955 + }, + { + "epoch": 11.111821086261982, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6956 + }, + { + "epoch": 11.113418530351439, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6957 + }, + { + "epoch": 11.115015974440894, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6958 + }, + { + "epoch": 11.116613418530351, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6959 + }, + { + "epoch": 11.118210862619808, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6960 + }, + { + "epoch": 11.119808306709265, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6961 + }, + { + "epoch": 11.121405750798722, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6962 + }, + { + "epoch": 11.12300319488818, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6963 + }, + { + "epoch": 11.124600638977636, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 6964 + }, + { + "epoch": 11.126198083067093, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6965 + }, + { + "epoch": 11.12779552715655, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6966 + }, + { + "epoch": 11.129392971246006, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6967 + }, + { + "epoch": 11.130990415335463, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6968 + }, + { + "epoch": 11.13258785942492, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6969 + }, + { + "epoch": 11.134185303514377, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6970 + }, + { + "epoch": 11.135782747603834, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 6971 + }, + { + "epoch": 11.13738019169329, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6972 + }, + { + "epoch": 11.138977635782748, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6973 + }, + { + "epoch": 11.140575079872205, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6974 + }, + { + "epoch": 11.142172523961662, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6975 + }, + { + "epoch": 11.143769968051119, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6976 + }, + { + "epoch": 11.145367412140574, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6977 + }, + { + "epoch": 11.146964856230031, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6978 + }, + { + "epoch": 11.148562300319488, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6979 + }, + { + "epoch": 11.150159744408946, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6980 + }, + { + "epoch": 11.151757188498403, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6981 + }, + { + "epoch": 11.15335463258786, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6982 + }, + { + "epoch": 11.154952076677317, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6983 + }, + { + "epoch": 11.156549520766774, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6984 + }, + { + "epoch": 11.15814696485623, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6985 + }, + { + "epoch": 11.159744408945686, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6986 + }, + { + "epoch": 11.161341853035143, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6987 + }, + { + "epoch": 11.1629392971246, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6988 + }, + { + "epoch": 11.164536741214057, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6989 + }, + { + "epoch": 11.166134185303514, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 6990 + }, + { + "epoch": 11.167731629392971, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6991 + }, + { + "epoch": 11.169329073482428, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6992 + }, + { + "epoch": 11.170926517571885, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6993 + }, + { + "epoch": 11.172523961661343, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6994 + }, + { + "epoch": 11.1741214057508, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6995 + }, + { + "epoch": 11.175718849840255, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6996 + }, + { + "epoch": 11.177316293929712, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6997 + }, + { + "epoch": 11.178913738019169, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6998 + }, + { + "epoch": 11.180511182108626, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6999 + }, + { + "epoch": 11.182108626198083, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7000 + }, + { + "epoch": 11.18370607028754, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7001 + }, + { + "epoch": 11.185303514376997, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7002 + }, + { + "epoch": 11.186900958466454, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7003 + }, + { + "epoch": 11.188498402555911, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7004 + }, + { + "epoch": 11.190095846645367, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7005 + }, + { + "epoch": 11.191693290734824, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7006 + }, + { + "epoch": 11.19329073482428, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7007 + }, + { + "epoch": 11.194888178913738, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7008 + }, + { + "epoch": 11.196485623003195, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7009 + }, + { + "epoch": 11.198083067092652, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7010 + }, + { + "epoch": 11.199680511182109, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7011 + }, + { + "epoch": 11.201277955271566, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7012 + }, + { + "epoch": 11.202875399361023, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7013 + }, + { + "epoch": 11.204472843450478, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7014 + }, + { + "epoch": 11.206070287539935, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7015 + }, + { + "epoch": 11.207667731629392, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7016 + }, + { + "epoch": 11.20926517571885, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7017 + }, + { + "epoch": 11.210862619808307, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7018 + }, + { + "epoch": 11.212460063897764, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7019 + }, + { + "epoch": 11.21405750798722, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7020 + }, + { + "epoch": 11.215654952076678, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7021 + }, + { + "epoch": 11.217252396166135, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7022 + }, + { + "epoch": 11.218849840255592, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7023 + }, + { + "epoch": 11.220447284345047, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 7024 + }, + { + "epoch": 11.222044728434504, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7025 + }, + { + "epoch": 11.223642172523961, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7026 + }, + { + "epoch": 11.225239616613418, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7027 + }, + { + "epoch": 11.226837060702875, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7028 + }, + { + "epoch": 11.228434504792332, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7029 + }, + { + "epoch": 11.23003194888179, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7030 + }, + { + "epoch": 11.231629392971247, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7031 + }, + { + "epoch": 11.233226837060704, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7032 + }, + { + "epoch": 11.23482428115016, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7033 + }, + { + "epoch": 11.236421725239616, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7034 + }, + { + "epoch": 11.238019169329073, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7035 + }, + { + "epoch": 11.23961661341853, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 7036 + }, + { + "epoch": 11.241214057507987, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7037 + }, + { + "epoch": 11.242811501597444, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7038 + }, + { + "epoch": 11.244408945686901, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7039 + }, + { + "epoch": 11.246006389776358, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7040 + }, + { + "epoch": 11.247603833865815, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7041 + }, + { + "epoch": 11.249201277955272, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 7042 + }, + { + "epoch": 11.250798722044728, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 7043 + }, + { + "epoch": 11.252396166134185, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7044 + }, + { + "epoch": 11.253993610223642, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7045 + }, + { + "epoch": 11.255591054313099, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7046 + }, + { + "epoch": 11.257188498402556, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7047 + }, + { + "epoch": 11.258785942492013, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7048 + }, + { + "epoch": 11.26038338658147, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7049 + }, + { + "epoch": 11.261980830670927, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7050 + }, + { + "epoch": 11.263578274760384, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7051 + }, + { + "epoch": 11.26517571884984, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7052 + }, + { + "epoch": 11.266773162939296, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7053 + }, + { + "epoch": 11.268370607028753, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7054 + }, + { + "epoch": 11.26996805111821, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7055 + }, + { + "epoch": 11.271565495207668, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7056 + }, + { + "epoch": 11.273162939297125, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 7057 + }, + { + "epoch": 11.274760383386582, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 7058 + }, + { + "epoch": 11.276357827476039, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7059 + }, + { + "epoch": 11.277955271565496, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7060 + }, + { + "epoch": 11.279552715654953, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7061 + }, + { + "epoch": 11.281150159744408, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7062 + }, + { + "epoch": 11.282747603833865, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7063 + }, + { + "epoch": 11.284345047923322, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7064 + }, + { + "epoch": 11.28594249201278, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7065 + }, + { + "epoch": 11.287539936102236, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 7066 + }, + { + "epoch": 11.289137380191693, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7067 + }, + { + "epoch": 11.29073482428115, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7068 + }, + { + "epoch": 11.292332268370608, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 7069 + }, + { + "epoch": 11.293929712460065, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 7070 + }, + { + "epoch": 11.295527156549522, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7071 + }, + { + "epoch": 11.297124600638977, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 7072 + }, + { + "epoch": 11.298722044728434, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7073 + }, + { + "epoch": 11.300319488817891, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7074 + }, + { + "epoch": 11.301916932907348, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7075 + }, + { + "epoch": 11.303514376996805, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7076 + }, + { + "epoch": 11.305111821086262, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7077 + }, + { + "epoch": 11.30670926517572, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7078 + }, + { + "epoch": 11.308306709265176, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7079 + }, + { + "epoch": 11.309904153354633, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7080 + }, + { + "epoch": 11.311501597444089, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7081 + }, + { + "epoch": 11.313099041533546, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 7082 + }, + { + "epoch": 11.314696485623003, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7083 + }, + { + "epoch": 11.31629392971246, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7084 + }, + { + "epoch": 11.317891373801917, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7085 + }, + { + "epoch": 11.319488817891374, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7086 + }, + { + "epoch": 11.321086261980831, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7087 + }, + { + "epoch": 11.322683706070288, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 7088 + }, + { + "epoch": 11.324281150159745, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7089 + }, + { + "epoch": 11.3258785942492, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7090 + }, + { + "epoch": 11.327476038338657, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7091 + }, + { + "epoch": 11.329073482428115, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7092 + }, + { + "epoch": 11.330670926517572, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7093 + }, + { + "epoch": 11.332268370607029, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7094 + }, + { + "epoch": 11.333865814696486, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7095 + }, + { + "epoch": 11.335463258785943, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7096 + }, + { + "epoch": 11.3370607028754, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7097 + }, + { + "epoch": 11.338658146964857, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7098 + }, + { + "epoch": 11.340255591054314, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7099 + }, + { + "epoch": 11.34185303514377, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7100 + }, + { + "epoch": 11.343450479233226, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7101 + }, + { + "epoch": 11.345047923322683, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7102 + }, + { + "epoch": 11.34664536741214, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7103 + }, + { + "epoch": 11.348242811501597, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7104 + }, + { + "epoch": 11.349840255591054, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7105 + }, + { + "epoch": 11.351437699680512, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7106 + }, + { + "epoch": 11.353035143769969, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7107 + }, + { + "epoch": 11.354632587859426, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7108 + }, + { + "epoch": 11.356230031948883, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7109 + }, + { + "epoch": 11.357827476038338, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7110 + }, + { + "epoch": 11.359424920127795, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7111 + }, + { + "epoch": 11.361022364217252, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7112 + }, + { + "epoch": 11.36261980830671, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7113 + }, + { + "epoch": 11.364217252396166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7114 + }, + { + "epoch": 11.365814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7115 + }, + { + "epoch": 11.36741214057508, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7116 + }, + { + "epoch": 11.369009584664537, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 7117 + }, + { + "epoch": 11.370607028753994, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7118 + }, + { + "epoch": 11.37220447284345, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7119 + }, + { + "epoch": 11.373801916932907, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7120 + }, + { + "epoch": 11.375399361022364, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 7121 + }, + { + "epoch": 11.37699680511182, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7122 + }, + { + "epoch": 11.378594249201278, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7123 + }, + { + "epoch": 11.380191693290735, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7124 + }, + { + "epoch": 11.381789137380192, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7125 + }, + { + "epoch": 11.383386581469649, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7126 + }, + { + "epoch": 11.384984025559106, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7127 + }, + { + "epoch": 11.386581469648561, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7128 + }, + { + "epoch": 11.388178913738018, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7129 + }, + { + "epoch": 11.389776357827476, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7130 + }, + { + "epoch": 11.391373801916933, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7131 + }, + { + "epoch": 11.39297124600639, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7132 + }, + { + "epoch": 11.394568690095847, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 7133 + }, + { + "epoch": 11.396166134185304, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7134 + }, + { + "epoch": 11.39776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7135 + }, + { + "epoch": 11.399361022364218, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7136 + }, + { + "epoch": 11.400958466453675, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7137 + }, + { + "epoch": 11.40255591054313, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7138 + }, + { + "epoch": 11.404153354632587, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7139 + }, + { + "epoch": 11.405750798722044, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7140 + }, + { + "epoch": 11.407348242811501, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7141 + }, + { + "epoch": 11.408945686900958, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7142 + }, + { + "epoch": 11.410543130990416, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7143 + }, + { + "epoch": 11.412140575079873, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7144 + }, + { + "epoch": 11.41373801916933, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7145 + }, + { + "epoch": 11.415335463258787, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 7146 + }, + { + "epoch": 11.416932907348242, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7147 + }, + { + "epoch": 11.418530351437699, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 7148 + }, + { + "epoch": 11.420127795527156, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7149 + }, + { + "epoch": 11.421725239616613, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7150 + }, + { + "epoch": 11.42332268370607, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7151 + }, + { + "epoch": 11.424920127795527, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7152 + }, + { + "epoch": 11.426517571884984, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7153 + }, + { + "epoch": 11.428115015974441, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7154 + }, + { + "epoch": 11.429712460063898, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7155 + }, + { + "epoch": 11.431309904153355, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7156 + }, + { + "epoch": 11.43290734824281, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7157 + }, + { + "epoch": 11.434504792332268, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7158 + }, + { + "epoch": 11.436102236421725, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7159 + }, + { + "epoch": 11.437699680511182, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7160 + }, + { + "epoch": 11.439297124600639, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7161 + }, + { + "epoch": 11.440894568690096, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7162 + }, + { + "epoch": 11.442492012779553, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7163 + }, + { + "epoch": 11.44408945686901, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7164 + }, + { + "epoch": 11.445686900958467, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7165 + }, + { + "epoch": 11.447284345047922, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 7166 + }, + { + "epoch": 11.44888178913738, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7167 + }, + { + "epoch": 11.450479233226837, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7168 + }, + { + "epoch": 11.452076677316294, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7169 + }, + { + "epoch": 11.45367412140575, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7170 + }, + { + "epoch": 11.455271565495208, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7171 + }, + { + "epoch": 11.456869009584665, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7172 + }, + { + "epoch": 11.458466453674122, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7173 + }, + { + "epoch": 11.460063897763579, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7174 + }, + { + "epoch": 11.461661341853036, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7175 + }, + { + "epoch": 11.463258785942491, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7176 + }, + { + "epoch": 11.464856230031948, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7177 + }, + { + "epoch": 11.466453674121405, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7178 + }, + { + "epoch": 11.468051118210862, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7179 + }, + { + "epoch": 11.46964856230032, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7180 + }, + { + "epoch": 11.471246006389777, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 7181 + }, + { + "epoch": 11.472843450479234, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 7182 + }, + { + "epoch": 11.47444089456869, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7183 + }, + { + "epoch": 11.476038338658148, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7184 + }, + { + "epoch": 11.477635782747603, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 7185 + }, + { + "epoch": 11.47923322683706, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7186 + }, + { + "epoch": 11.480830670926517, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7187 + }, + { + "epoch": 11.482428115015974, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7188 + }, + { + "epoch": 11.484025559105431, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7189 + }, + { + "epoch": 11.485623003194888, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7190 + }, + { + "epoch": 11.487220447284345, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 7191 + }, + { + "epoch": 11.488817891373802, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7192 + }, + { + "epoch": 11.49041533546326, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7193 + }, + { + "epoch": 11.492012779552716, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7194 + }, + { + "epoch": 11.493610223642172, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7195 + }, + { + "epoch": 11.495207667731629, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7196 + }, + { + "epoch": 11.496805111821086, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 7197 + }, + { + "epoch": 11.498402555910543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7198 + }, + { + "epoch": 11.5, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7199 + }, + { + "epoch": 11.501597444089457, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7200 + }, + { + "epoch": 11.503194888178914, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 7201 + }, + { + "epoch": 11.504792332268371, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7202 + }, + { + "epoch": 11.506389776357828, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7203 + }, + { + "epoch": 11.507987220447284, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7204 + }, + { + "epoch": 11.50958466453674, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7205 + }, + { + "epoch": 11.511182108626198, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7206 + }, + { + "epoch": 11.512779552715655, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 7207 + }, + { + "epoch": 11.514376996805112, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7208 + }, + { + "epoch": 11.515974440894569, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7209 + }, + { + "epoch": 11.517571884984026, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7210 + }, + { + "epoch": 11.519169329073483, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 7211 + }, + { + "epoch": 11.52076677316294, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7212 + }, + { + "epoch": 11.522364217252395, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7213 + }, + { + "epoch": 11.523961661341852, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 7214 + }, + { + "epoch": 11.52555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7215 + }, + { + "epoch": 11.527156549520766, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7216 + }, + { + "epoch": 11.528753993610223, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7217 + }, + { + "epoch": 11.53035143769968, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7218 + }, + { + "epoch": 11.531948881789138, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7219 + }, + { + "epoch": 11.533546325878595, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7220 + }, + { + "epoch": 11.535143769968052, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7221 + }, + { + "epoch": 11.536741214057509, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7222 + }, + { + "epoch": 11.538338658146966, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7223 + }, + { + "epoch": 11.539936102236421, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7224 + }, + { + "epoch": 11.541533546325878, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 7225 + }, + { + "epoch": 11.543130990415335, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7226 + }, + { + "epoch": 11.544728434504792, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7227 + }, + { + "epoch": 11.54632587859425, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7228 + }, + { + "epoch": 11.547923322683706, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7229 + }, + { + "epoch": 11.549520766773163, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7230 + }, + { + "epoch": 11.55111821086262, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7231 + }, + { + "epoch": 11.552715654952078, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7232 + }, + { + "epoch": 11.554313099041533, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7233 + }, + { + "epoch": 11.55591054313099, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7234 + }, + { + "epoch": 11.557507987220447, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7235 + }, + { + "epoch": 11.559105431309904, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7236 + }, + { + "epoch": 11.560702875399361, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7237 + }, + { + "epoch": 11.562300319488818, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7238 + }, + { + "epoch": 11.563897763578275, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7239 + }, + { + "epoch": 11.565495207667732, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7240 + }, + { + "epoch": 11.56709265175719, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 7241 + }, + { + "epoch": 11.568690095846645, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7242 + }, + { + "epoch": 11.570287539936102, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7243 + }, + { + "epoch": 11.571884984025559, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7244 + }, + { + "epoch": 11.573482428115016, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7245 + }, + { + "epoch": 11.575079872204473, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7246 + }, + { + "epoch": 11.57667731629393, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7247 + }, + { + "epoch": 11.578274760383387, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 7248 + }, + { + "epoch": 11.579872204472844, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7249 + }, + { + "epoch": 11.581469648562301, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7250 + }, + { + "epoch": 11.583067092651756, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7251 + }, + { + "epoch": 11.584664536741213, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 7252 + }, + { + "epoch": 11.58626198083067, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7253 + }, + { + "epoch": 11.587859424920127, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7254 + }, + { + "epoch": 11.589456869009584, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7255 + }, + { + "epoch": 11.591054313099042, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7256 + }, + { + "epoch": 11.592651757188499, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7257 + }, + { + "epoch": 11.594249201277956, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7258 + }, + { + "epoch": 11.595846645367413, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7259 + }, + { + "epoch": 11.59744408945687, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7260 + }, + { + "epoch": 11.599041533546325, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7261 + }, + { + "epoch": 11.600638977635782, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 7262 + }, + { + "epoch": 11.60223642172524, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7263 + }, + { + "epoch": 11.603833865814696, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7264 + }, + { + "epoch": 11.605431309904153, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 7265 + }, + { + "epoch": 11.60702875399361, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7266 + }, + { + "epoch": 11.608626198083067, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7267 + }, + { + "epoch": 11.610223642172524, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7268 + }, + { + "epoch": 11.611821086261982, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7269 + }, + { + "epoch": 11.613418530351439, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7270 + }, + { + "epoch": 11.615015974440894, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7271 + }, + { + "epoch": 11.616613418530351, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7272 + }, + { + "epoch": 11.618210862619808, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7273 + }, + { + "epoch": 11.619808306709265, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7274 + }, + { + "epoch": 11.621405750798722, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7275 + }, + { + "epoch": 11.62300319488818, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7276 + }, + { + "epoch": 11.624600638977636, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7277 + }, + { + "epoch": 11.626198083067093, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 7278 + }, + { + "epoch": 11.62779552715655, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7279 + }, + { + "epoch": 11.629392971246006, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7280 + }, + { + "epoch": 11.630990415335463, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7281 + }, + { + "epoch": 11.63258785942492, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7282 + }, + { + "epoch": 11.634185303514377, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7283 + }, + { + "epoch": 11.635782747603834, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7284 + }, + { + "epoch": 11.63738019169329, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7285 + }, + { + "epoch": 11.638977635782748, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7286 + }, + { + "epoch": 11.640575079872205, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7287 + }, + { + "epoch": 11.642172523961662, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7288 + }, + { + "epoch": 11.643769968051117, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7289 + }, + { + "epoch": 11.645367412140574, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7290 + }, + { + "epoch": 11.646964856230031, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7291 + }, + { + "epoch": 11.648562300319488, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 7292 + }, + { + "epoch": 11.650159744408946, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7293 + }, + { + "epoch": 11.651757188498403, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7294 + }, + { + "epoch": 11.65335463258786, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 7295 + }, + { + "epoch": 11.654952076677317, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7296 + }, + { + "epoch": 11.656549520766774, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7297 + }, + { + "epoch": 11.65814696485623, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7298 + }, + { + "epoch": 11.659744408945686, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7299 + }, + { + "epoch": 11.661341853035143, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 7300 + }, + { + "epoch": 11.6629392971246, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7301 + }, + { + "epoch": 11.664536741214057, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7302 + }, + { + "epoch": 11.666134185303514, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7303 + }, + { + "epoch": 11.667731629392971, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7304 + }, + { + "epoch": 11.669329073482428, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7305 + }, + { + "epoch": 11.670926517571885, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 7306 + }, + { + "epoch": 11.672523961661343, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7307 + }, + { + "epoch": 11.6741214057508, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7308 + }, + { + "epoch": 11.675718849840255, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7309 + }, + { + "epoch": 11.677316293929712, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7310 + }, + { + "epoch": 11.678913738019169, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7311 + }, + { + "epoch": 11.680511182108626, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7312 + }, + { + "epoch": 11.682108626198083, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7313 + }, + { + "epoch": 11.68370607028754, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7314 + }, + { + "epoch": 11.685303514376997, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7315 + }, + { + "epoch": 11.686900958466454, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7316 + }, + { + "epoch": 11.688498402555911, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7317 + }, + { + "epoch": 11.690095846645367, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7318 + }, + { + "epoch": 11.691693290734824, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7319 + }, + { + "epoch": 11.69329073482428, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7320 + }, + { + "epoch": 11.694888178913738, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 7321 + }, + { + "epoch": 11.696485623003195, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7322 + }, + { + "epoch": 11.698083067092652, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7323 + }, + { + "epoch": 11.699680511182109, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7324 + }, + { + "epoch": 11.701277955271566, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7325 + }, + { + "epoch": 11.702875399361023, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 7326 + }, + { + "epoch": 11.704472843450478, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7327 + }, + { + "epoch": 11.706070287539935, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7328 + }, + { + "epoch": 11.707667731629392, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7329 + }, + { + "epoch": 11.70926517571885, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7330 + }, + { + "epoch": 11.710862619808307, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7331 + }, + { + "epoch": 11.712460063897764, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7332 + }, + { + "epoch": 11.71405750798722, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7333 + }, + { + "epoch": 11.715654952076678, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7334 + }, + { + "epoch": 11.717252396166135, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7335 + }, + { + "epoch": 11.718849840255592, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7336 + }, + { + "epoch": 11.720447284345047, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7337 + }, + { + "epoch": 11.722044728434504, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7338 + }, + { + "epoch": 11.723642172523961, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7339 + }, + { + "epoch": 11.725239616613418, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7340 + }, + { + "epoch": 11.726837060702875, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7341 + }, + { + "epoch": 11.728434504792332, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7342 + }, + { + "epoch": 11.73003194888179, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7343 + }, + { + "epoch": 11.731629392971247, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7344 + }, + { + "epoch": 11.733226837060704, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7345 + }, + { + "epoch": 11.73482428115016, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7346 + }, + { + "epoch": 11.736421725239616, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7347 + }, + { + "epoch": 11.738019169329073, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7348 + }, + { + "epoch": 11.73961661341853, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7349 + }, + { + "epoch": 11.741214057507987, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7350 + }, + { + "epoch": 11.742811501597444, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7351 + }, + { + "epoch": 11.744408945686901, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7352 + }, + { + "epoch": 11.746006389776358, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7353 + }, + { + "epoch": 11.747603833865815, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7354 + }, + { + "epoch": 11.749201277955272, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7355 + }, + { + "epoch": 11.750798722044728, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7356 + }, + { + "epoch": 11.752396166134185, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7357 + }, + { + "epoch": 11.753993610223642, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7358 + }, + { + "epoch": 11.755591054313099, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7359 + }, + { + "epoch": 11.757188498402556, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7360 + }, + { + "epoch": 11.758785942492013, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7361 + }, + { + "epoch": 11.76038338658147, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7362 + }, + { + "epoch": 11.761980830670927, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7363 + }, + { + "epoch": 11.763578274760384, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7364 + }, + { + "epoch": 11.76517571884984, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7365 + }, + { + "epoch": 11.766773162939296, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7366 + }, + { + "epoch": 11.768370607028753, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7367 + }, + { + "epoch": 11.76996805111821, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7368 + }, + { + "epoch": 11.771565495207668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7369 + }, + { + "epoch": 11.773162939297125, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7370 + }, + { + "epoch": 11.774760383386582, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7371 + }, + { + "epoch": 11.776357827476039, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 7372 + }, + { + "epoch": 11.777955271565496, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7373 + }, + { + "epoch": 11.779552715654953, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7374 + }, + { + "epoch": 11.781150159744408, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7375 + }, + { + "epoch": 11.782747603833865, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7376 + }, + { + "epoch": 11.784345047923322, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7377 + }, + { + "epoch": 11.78594249201278, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7378 + }, + { + "epoch": 11.787539936102236, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7379 + }, + { + "epoch": 11.789137380191693, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7380 + }, + { + "epoch": 11.79073482428115, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7381 + }, + { + "epoch": 11.792332268370608, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7382 + }, + { + "epoch": 11.793929712460065, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7383 + }, + { + "epoch": 11.795527156549522, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7384 + }, + { + "epoch": 11.797124600638977, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7385 + }, + { + "epoch": 11.798722044728434, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7386 + }, + { + "epoch": 11.800319488817891, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7387 + }, + { + "epoch": 11.801916932907348, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7388 + }, + { + "epoch": 11.803514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7389 + }, + { + "epoch": 11.805111821086262, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7390 + }, + { + "epoch": 11.80670926517572, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7391 + }, + { + "epoch": 11.808306709265176, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7392 + }, + { + "epoch": 11.809904153354633, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7393 + }, + { + "epoch": 11.811501597444089, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7394 + }, + { + "epoch": 11.813099041533546, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7395 + }, + { + "epoch": 11.814696485623003, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7396 + }, + { + "epoch": 11.81629392971246, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 7397 + }, + { + "epoch": 11.817891373801917, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7398 + }, + { + "epoch": 11.819488817891374, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7399 + }, + { + "epoch": 11.821086261980831, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7400 + }, + { + "epoch": 11.822683706070288, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7401 + }, + { + "epoch": 11.824281150159745, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7402 + }, + { + "epoch": 11.8258785942492, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7403 + }, + { + "epoch": 11.827476038338657, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7404 + }, + { + "epoch": 11.829073482428115, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7405 + }, + { + "epoch": 11.830670926517572, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7406 + }, + { + "epoch": 11.832268370607029, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7407 + }, + { + "epoch": 11.833865814696486, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 7408 + }, + { + "epoch": 11.835463258785943, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7409 + }, + { + "epoch": 11.8370607028754, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7410 + }, + { + "epoch": 11.838658146964857, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7411 + }, + { + "epoch": 11.840255591054314, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7412 + }, + { + "epoch": 11.84185303514377, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7413 + }, + { + "epoch": 11.843450479233226, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 7414 + }, + { + "epoch": 11.845047923322683, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7415 + }, + { + "epoch": 11.84664536741214, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7416 + }, + { + "epoch": 11.848242811501597, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7417 + }, + { + "epoch": 11.849840255591054, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7418 + }, + { + "epoch": 11.851437699680512, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7419 + }, + { + "epoch": 11.853035143769969, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7420 + }, + { + "epoch": 11.854632587859426, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7421 + }, + { + "epoch": 11.856230031948883, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7422 + }, + { + "epoch": 11.857827476038338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7423 + }, + { + "epoch": 11.859424920127795, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7424 + }, + { + "epoch": 11.861022364217252, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7425 + }, + { + "epoch": 11.86261980830671, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7426 + }, + { + "epoch": 11.864217252396166, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7427 + }, + { + "epoch": 11.865814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7428 + }, + { + "epoch": 11.86741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 7429 + }, + { + "epoch": 11.869009584664537, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7430 + }, + { + "epoch": 11.870607028753994, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7431 + }, + { + "epoch": 11.87220447284345, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7432 + }, + { + "epoch": 11.873801916932907, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7433 + }, + { + "epoch": 11.875399361022364, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7434 + }, + { + "epoch": 11.87699680511182, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7435 + }, + { + "epoch": 11.878594249201278, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7436 + }, + { + "epoch": 11.880191693290735, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7437 + }, + { + "epoch": 11.881789137380192, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7438 + }, + { + "epoch": 11.883386581469649, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7439 + }, + { + "epoch": 11.884984025559106, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7440 + }, + { + "epoch": 11.886581469648561, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7441 + }, + { + "epoch": 11.888178913738018, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7442 + }, + { + "epoch": 11.889776357827476, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7443 + }, + { + "epoch": 11.891373801916933, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7444 + }, + { + "epoch": 11.89297124600639, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7445 + }, + { + "epoch": 11.894568690095847, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7446 + }, + { + "epoch": 11.896166134185304, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7447 + }, + { + "epoch": 11.89776357827476, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7448 + }, + { + "epoch": 11.899361022364218, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7449 + }, + { + "epoch": 11.900958466453675, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7450 + }, + { + "epoch": 11.90255591054313, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7451 + }, + { + "epoch": 11.904153354632587, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7452 + }, + { + "epoch": 11.905750798722044, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7453 + }, + { + "epoch": 11.907348242811501, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7454 + }, + { + "epoch": 11.908945686900958, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7455 + }, + { + "epoch": 11.910543130990416, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7456 + }, + { + "epoch": 11.912140575079873, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7457 + }, + { + "epoch": 11.91373801916933, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 7458 + }, + { + "epoch": 11.915335463258787, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7459 + }, + { + "epoch": 11.916932907348244, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7460 + }, + { + "epoch": 11.918530351437699, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7461 + }, + { + "epoch": 11.920127795527156, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7462 + }, + { + "epoch": 11.921725239616613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7463 + }, + { + "epoch": 11.92332268370607, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7464 + }, + { + "epoch": 11.924920127795527, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7465 + }, + { + "epoch": 11.926517571884984, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7466 + }, + { + "epoch": 11.928115015974441, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7467 + }, + { + "epoch": 11.929712460063898, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7468 + }, + { + "epoch": 11.931309904153355, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7469 + }, + { + "epoch": 11.93290734824281, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7470 + }, + { + "epoch": 11.934504792332268, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 7471 + }, + { + "epoch": 11.936102236421725, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7472 + }, + { + "epoch": 11.937699680511182, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 7473 + }, + { + "epoch": 11.939297124600639, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7474 + }, + { + "epoch": 11.940894568690096, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7475 + }, + { + "epoch": 11.942492012779553, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7476 + }, + { + "epoch": 11.94408945686901, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7477 + }, + { + "epoch": 11.945686900958467, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 7478 + }, + { + "epoch": 11.947284345047922, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7479 + }, + { + "epoch": 11.94888178913738, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7480 + }, + { + "epoch": 11.950479233226837, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7481 + }, + { + "epoch": 11.952076677316294, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7482 + }, + { + "epoch": 11.95367412140575, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7483 + }, + { + "epoch": 11.955271565495208, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7484 + }, + { + "epoch": 11.956869009584665, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7485 + }, + { + "epoch": 11.958466453674122, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7486 + }, + { + "epoch": 11.960063897763579, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7487 + }, + { + "epoch": 11.961661341853034, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7488 + }, + { + "epoch": 11.963258785942491, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7489 + }, + { + "epoch": 11.964856230031948, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 7490 + }, + { + "epoch": 11.966453674121405, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 7491 + }, + { + "epoch": 11.968051118210862, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7492 + }, + { + "epoch": 11.96964856230032, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7493 + }, + { + "epoch": 11.971246006389777, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7494 + }, + { + "epoch": 11.972843450479234, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7495 + }, + { + "epoch": 11.97444089456869, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7496 + }, + { + "epoch": 11.976038338658148, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7497 + }, + { + "epoch": 11.977635782747605, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7498 + }, + { + "epoch": 11.97923322683706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7499 + }, + { + "epoch": 11.980830670926517, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7500 + }, + { + "epoch": 11.982428115015974, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7501 + }, + { + "epoch": 11.984025559105431, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7502 + }, + { + "epoch": 11.985623003194888, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7503 + }, + { + "epoch": 11.987220447284345, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7504 + }, + { + "epoch": 11.988817891373802, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7505 + }, + { + "epoch": 11.99041533546326, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7506 + }, + { + "epoch": 11.992012779552716, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7507 + }, + { + "epoch": 11.993610223642172, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7508 + }, + { + "epoch": 11.995207667731629, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7509 + }, + { + "epoch": 11.996805111821086, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7510 + }, + { + "epoch": 11.998402555910543, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7511 + }, + { + "epoch": 12.0, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7512 + }, + { + "epoch": 12.001597444089457, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7513 + }, + { + "epoch": 12.003194888178914, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7514 + }, + { + "epoch": 12.004792332268371, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7515 + }, + { + "epoch": 12.006389776357828, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7516 + }, + { + "epoch": 12.007987220447284, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7517 + }, + { + "epoch": 12.00958466453674, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7518 + }, + { + "epoch": 12.011182108626198, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7519 + }, + { + "epoch": 12.012779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 7520 + }, + { + "epoch": 12.014376996805112, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7521 + }, + { + "epoch": 12.015974440894569, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7522 + }, + { + "epoch": 12.017571884984026, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 7523 + }, + { + "epoch": 12.019169329073483, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7524 + }, + { + "epoch": 12.02076677316294, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7525 + }, + { + "epoch": 12.022364217252397, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7526 + }, + { + "epoch": 12.023961661341852, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7527 + }, + { + "epoch": 12.02555910543131, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7528 + }, + { + "epoch": 12.027156549520766, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7529 + }, + { + "epoch": 12.028753993610223, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7530 + }, + { + "epoch": 12.03035143769968, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7531 + }, + { + "epoch": 12.031948881789138, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7532 + }, + { + "epoch": 12.033546325878595, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 7533 + }, + { + "epoch": 12.035143769968052, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7534 + }, + { + "epoch": 12.036741214057509, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7535 + }, + { + "epoch": 12.038338658146964, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7536 + }, + { + "epoch": 12.039936102236421, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7537 + }, + { + "epoch": 12.041533546325878, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7538 + }, + { + "epoch": 12.043130990415335, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7539 + }, + { + "epoch": 12.044728434504792, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 7540 + }, + { + "epoch": 12.04632587859425, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 7541 + }, + { + "epoch": 12.047923322683706, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7542 + }, + { + "epoch": 12.049520766773163, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7543 + }, + { + "epoch": 12.05111821086262, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7544 + }, + { + "epoch": 12.052715654952078, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7545 + }, + { + "epoch": 12.054313099041533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7546 + }, + { + "epoch": 12.05591054313099, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7547 + }, + { + "epoch": 12.057507987220447, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7548 + }, + { + "epoch": 12.059105431309904, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7549 + }, + { + "epoch": 12.060702875399361, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7550 + }, + { + "epoch": 12.062300319488818, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7551 + }, + { + "epoch": 12.063897763578275, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7552 + }, + { + "epoch": 12.065495207667732, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7553 + }, + { + "epoch": 12.06709265175719, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7554 + }, + { + "epoch": 12.068690095846645, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7555 + }, + { + "epoch": 12.070287539936102, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7556 + }, + { + "epoch": 12.071884984025559, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7557 + }, + { + "epoch": 12.073482428115016, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7558 + }, + { + "epoch": 12.075079872204473, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 7559 + }, + { + "epoch": 12.07667731629393, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7560 + }, + { + "epoch": 12.078274760383387, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7561 + }, + { + "epoch": 12.079872204472844, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 7562 + }, + { + "epoch": 12.081469648562301, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7563 + }, + { + "epoch": 12.083067092651758, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7564 + }, + { + "epoch": 12.084664536741213, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7565 + }, + { + "epoch": 12.08626198083067, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7566 + }, + { + "epoch": 12.087859424920127, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 7567 + }, + { + "epoch": 12.089456869009584, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7568 + }, + { + "epoch": 12.091054313099042, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7569 + }, + { + "epoch": 12.092651757188499, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7570 + }, + { + "epoch": 12.094249201277956, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7571 + }, + { + "epoch": 12.095846645367413, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 7572 + }, + { + "epoch": 12.09744408945687, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7573 + }, + { + "epoch": 12.099041533546325, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7574 + }, + { + "epoch": 12.100638977635782, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7575 + }, + { + "epoch": 12.10223642172524, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7576 + }, + { + "epoch": 12.103833865814696, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7577 + }, + { + "epoch": 12.105431309904153, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7578 + }, + { + "epoch": 12.10702875399361, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7579 + }, + { + "epoch": 12.108626198083067, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7580 + }, + { + "epoch": 12.110223642172524, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7581 + }, + { + "epoch": 12.111821086261982, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7582 + }, + { + "epoch": 12.113418530351439, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7583 + }, + { + "epoch": 12.115015974440894, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7584 + }, + { + "epoch": 12.116613418530351, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 7585 + }, + { + "epoch": 12.118210862619808, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7586 + }, + { + "epoch": 12.119808306709265, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7587 + }, + { + "epoch": 12.121405750798722, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7588 + }, + { + "epoch": 12.12300319488818, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7589 + }, + { + "epoch": 12.124600638977636, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7590 + }, + { + "epoch": 12.126198083067093, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7591 + }, + { + "epoch": 12.12779552715655, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7592 + }, + { + "epoch": 12.129392971246006, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7593 + }, + { + "epoch": 12.130990415335463, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7594 + }, + { + "epoch": 12.13258785942492, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7595 + }, + { + "epoch": 12.134185303514377, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7596 + }, + { + "epoch": 12.135782747603834, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7597 + }, + { + "epoch": 12.13738019169329, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7598 + }, + { + "epoch": 12.138977635782748, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7599 + }, + { + "epoch": 12.140575079872205, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7600 + }, + { + "epoch": 12.142172523961662, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7601 + }, + { + "epoch": 12.143769968051119, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7602 + }, + { + "epoch": 12.145367412140574, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7603 + }, + { + "epoch": 12.146964856230031, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7604 + }, + { + "epoch": 12.148562300319488, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7605 + }, + { + "epoch": 12.150159744408946, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7606 + }, + { + "epoch": 12.151757188498403, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7607 + }, + { + "epoch": 12.15335463258786, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7608 + }, + { + "epoch": 12.154952076677317, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7609 + }, + { + "epoch": 12.156549520766774, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7610 + }, + { + "epoch": 12.15814696485623, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 7611 + }, + { + "epoch": 12.159744408945686, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7612 + }, + { + "epoch": 12.161341853035143, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7613 + }, + { + "epoch": 12.1629392971246, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7614 + }, + { + "epoch": 12.164536741214057, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7615 + }, + { + "epoch": 12.166134185303514, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7616 + }, + { + "epoch": 12.167731629392971, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7617 + }, + { + "epoch": 12.169329073482428, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7618 + }, + { + "epoch": 12.170926517571885, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7619 + }, + { + "epoch": 12.172523961661343, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7620 + }, + { + "epoch": 12.1741214057508, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7621 + }, + { + "epoch": 12.175718849840255, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7622 + }, + { + "epoch": 12.177316293929712, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7623 + }, + { + "epoch": 12.178913738019169, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7624 + }, + { + "epoch": 12.180511182108626, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7625 + }, + { + "epoch": 12.182108626198083, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 7626 + }, + { + "epoch": 12.18370607028754, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7627 + }, + { + "epoch": 12.185303514376997, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7628 + }, + { + "epoch": 12.186900958466454, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7629 + }, + { + "epoch": 12.188498402555911, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7630 + }, + { + "epoch": 12.190095846645367, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7631 + }, + { + "epoch": 12.191693290734824, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7632 + }, + { + "epoch": 12.19329073482428, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7633 + }, + { + "epoch": 12.194888178913738, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7634 + }, + { + "epoch": 12.196485623003195, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7635 + }, + { + "epoch": 12.198083067092652, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7636 + }, + { + "epoch": 12.199680511182109, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7637 + }, + { + "epoch": 12.201277955271566, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7638 + }, + { + "epoch": 12.202875399361023, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7639 + }, + { + "epoch": 12.204472843450478, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7640 + }, + { + "epoch": 12.206070287539935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7641 + }, + { + "epoch": 12.207667731629392, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7642 + }, + { + "epoch": 12.20926517571885, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7643 + }, + { + "epoch": 12.210862619808307, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7644 + }, + { + "epoch": 12.212460063897764, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 7645 + }, + { + "epoch": 12.21405750798722, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7646 + }, + { + "epoch": 12.215654952076678, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7647 + }, + { + "epoch": 12.217252396166135, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7648 + }, + { + "epoch": 12.218849840255592, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 7649 + }, + { + "epoch": 12.220447284345047, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7650 + }, + { + "epoch": 12.222044728434504, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7651 + }, + { + "epoch": 12.223642172523961, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7652 + }, + { + "epoch": 12.225239616613418, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7653 + }, + { + "epoch": 12.226837060702875, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7654 + }, + { + "epoch": 12.228434504792332, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7655 + }, + { + "epoch": 12.23003194888179, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7656 + }, + { + "epoch": 12.231629392971247, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7657 + }, + { + "epoch": 12.233226837060704, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7658 + }, + { + "epoch": 12.23482428115016, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 7659 + }, + { + "epoch": 12.236421725239616, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7660 + }, + { + "epoch": 12.238019169329073, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7661 + }, + { + "epoch": 12.23961661341853, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7662 + }, + { + "epoch": 12.241214057507987, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7663 + }, + { + "epoch": 12.242811501597444, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 7664 + }, + { + "epoch": 12.244408945686901, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7665 + }, + { + "epoch": 12.246006389776358, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7666 + }, + { + "epoch": 12.247603833865815, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7667 + }, + { + "epoch": 12.249201277955272, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7668 + }, + { + "epoch": 12.250798722044728, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7669 + }, + { + "epoch": 12.252396166134185, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7670 + }, + { + "epoch": 12.253993610223642, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7671 + }, + { + "epoch": 12.255591054313099, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7672 + }, + { + "epoch": 12.257188498402556, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 7673 + }, + { + "epoch": 12.258785942492013, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7674 + }, + { + "epoch": 12.26038338658147, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7675 + }, + { + "epoch": 12.261980830670927, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7676 + }, + { + "epoch": 12.263578274760384, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 7677 + }, + { + "epoch": 12.26517571884984, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7678 + }, + { + "epoch": 12.266773162939296, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7679 + }, + { + "epoch": 12.268370607028753, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7680 + }, + { + "epoch": 12.26996805111821, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7681 + }, + { + "epoch": 12.271565495207668, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7682 + }, + { + "epoch": 12.273162939297125, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7683 + }, + { + "epoch": 12.274760383386582, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7684 + }, + { + "epoch": 12.276357827476039, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7685 + }, + { + "epoch": 12.277955271565496, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7686 + }, + { + "epoch": 12.279552715654953, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7687 + }, + { + "epoch": 12.281150159744408, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7688 + }, + { + "epoch": 12.282747603833865, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7689 + }, + { + "epoch": 12.284345047923322, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7690 + }, + { + "epoch": 12.28594249201278, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7691 + }, + { + "epoch": 12.287539936102236, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7692 + }, + { + "epoch": 12.289137380191693, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7693 + }, + { + "epoch": 12.29073482428115, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7694 + }, + { + "epoch": 12.292332268370608, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7695 + }, + { + "epoch": 12.293929712460065, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7696 + }, + { + "epoch": 12.295527156549522, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7697 + }, + { + "epoch": 12.297124600638977, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7698 + }, + { + "epoch": 12.298722044728434, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7699 + }, + { + "epoch": 12.300319488817891, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7700 + }, + { + "epoch": 12.301916932907348, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7701 + }, + { + "epoch": 12.303514376996805, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 7702 + }, + { + "epoch": 12.305111821086262, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7703 + }, + { + "epoch": 12.30670926517572, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7704 + }, + { + "epoch": 12.308306709265176, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7705 + }, + { + "epoch": 12.309904153354633, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7706 + }, + { + "epoch": 12.311501597444089, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7707 + }, + { + "epoch": 12.313099041533546, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7708 + }, + { + "epoch": 12.314696485623003, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7709 + }, + { + "epoch": 12.31629392971246, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7710 + }, + { + "epoch": 12.317891373801917, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7711 + }, + { + "epoch": 12.319488817891374, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7712 + }, + { + "epoch": 12.321086261980831, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7713 + }, + { + "epoch": 12.322683706070288, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7714 + }, + { + "epoch": 12.324281150159745, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7715 + }, + { + "epoch": 12.3258785942492, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7716 + }, + { + "epoch": 12.327476038338657, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7717 + }, + { + "epoch": 12.329073482428115, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7718 + }, + { + "epoch": 12.330670926517572, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7719 + }, + { + "epoch": 12.332268370607029, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7720 + }, + { + "epoch": 12.333865814696486, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7721 + }, + { + "epoch": 12.335463258785943, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7722 + }, + { + "epoch": 12.3370607028754, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7723 + }, + { + "epoch": 12.338658146964857, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7724 + }, + { + "epoch": 12.340255591054314, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7725 + }, + { + "epoch": 12.34185303514377, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7726 + }, + { + "epoch": 12.343450479233226, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7727 + }, + { + "epoch": 12.345047923322683, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 7728 + }, + { + "epoch": 12.34664536741214, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 7729 + }, + { + "epoch": 12.348242811501597, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7730 + }, + { + "epoch": 12.349840255591054, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7731 + }, + { + "epoch": 12.351437699680512, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 7732 + }, + { + "epoch": 12.353035143769969, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7733 + }, + { + "epoch": 12.354632587859426, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7734 + }, + { + "epoch": 12.356230031948883, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7735 + }, + { + "epoch": 12.357827476038338, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7736 + }, + { + "epoch": 12.359424920127795, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7737 + }, + { + "epoch": 12.361022364217252, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7738 + }, + { + "epoch": 12.36261980830671, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 7739 + }, + { + "epoch": 12.364217252396166, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7740 + }, + { + "epoch": 12.365814696485623, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7741 + }, + { + "epoch": 12.36741214057508, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7742 + }, + { + "epoch": 12.369009584664537, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7743 + }, + { + "epoch": 12.370607028753994, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7744 + }, + { + "epoch": 12.37220447284345, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 7745 + }, + { + "epoch": 12.373801916932907, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7746 + }, + { + "epoch": 12.375399361022364, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 7747 + }, + { + "epoch": 12.37699680511182, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7748 + }, + { + "epoch": 12.378594249201278, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7749 + }, + { + "epoch": 12.380191693290735, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7750 + }, + { + "epoch": 12.381789137380192, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7751 + }, + { + "epoch": 12.383386581469649, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7752 + }, + { + "epoch": 12.384984025559106, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7753 + }, + { + "epoch": 12.386581469648561, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7754 + }, + { + "epoch": 12.388178913738018, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7755 + }, + { + "epoch": 12.389776357827476, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7756 + }, + { + "epoch": 12.391373801916933, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7757 + }, + { + "epoch": 12.39297124600639, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 7758 + }, + { + "epoch": 12.394568690095847, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7759 + }, + { + "epoch": 12.396166134185304, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7760 + }, + { + "epoch": 12.39776357827476, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7761 + }, + { + "epoch": 12.399361022364218, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7762 + }, + { + "epoch": 12.400958466453675, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7763 + }, + { + "epoch": 12.40255591054313, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7764 + }, + { + "epoch": 12.404153354632587, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7765 + }, + { + "epoch": 12.405750798722044, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7766 + }, + { + "epoch": 12.407348242811501, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7767 + }, + { + "epoch": 12.408945686900958, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7768 + }, + { + "epoch": 12.410543130990416, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 7769 + }, + { + "epoch": 12.412140575079873, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 7770 + }, + { + "epoch": 12.41373801916933, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 7771 + }, + { + "epoch": 12.415335463258787, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7772 + }, + { + "epoch": 12.416932907348242, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7773 + }, + { + "epoch": 12.418530351437699, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7774 + }, + { + "epoch": 12.420127795527156, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7775 + }, + { + "epoch": 12.421725239616613, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7776 + }, + { + "epoch": 12.42332268370607, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7777 + }, + { + "epoch": 12.424920127795527, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7778 + }, + { + "epoch": 12.426517571884984, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7779 + }, + { + "epoch": 12.428115015974441, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7780 + }, + { + "epoch": 12.429712460063898, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7781 + }, + { + "epoch": 12.431309904153355, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7782 + }, + { + "epoch": 12.43290734824281, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7783 + }, + { + "epoch": 12.434504792332268, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7784 + }, + { + "epoch": 12.436102236421725, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7785 + }, + { + "epoch": 12.437699680511182, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7786 + }, + { + "epoch": 12.439297124600639, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7787 + }, + { + "epoch": 12.440894568690096, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7788 + }, + { + "epoch": 12.442492012779553, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7789 + }, + { + "epoch": 12.44408945686901, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7790 + }, + { + "epoch": 12.445686900958467, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7791 + }, + { + "epoch": 12.447284345047922, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7792 + }, + { + "epoch": 12.44888178913738, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7793 + }, + { + "epoch": 12.450479233226837, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 7794 + }, + { + "epoch": 12.452076677316294, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7795 + }, + { + "epoch": 12.45367412140575, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7796 + }, + { + "epoch": 12.455271565495208, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7797 + }, + { + "epoch": 12.456869009584665, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7798 + }, + { + "epoch": 12.458466453674122, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 7799 + }, + { + "epoch": 12.460063897763579, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7800 + }, + { + "epoch": 12.461661341853036, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7801 + }, + { + "epoch": 12.463258785942491, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7802 + }, + { + "epoch": 12.464856230031948, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7803 + }, + { + "epoch": 12.466453674121405, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7804 + }, + { + "epoch": 12.468051118210862, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7805 + }, + { + "epoch": 12.46964856230032, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7806 + }, + { + "epoch": 12.471246006389777, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7807 + }, + { + "epoch": 12.472843450479234, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 7808 + }, + { + "epoch": 12.47444089456869, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7809 + }, + { + "epoch": 12.476038338658148, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7810 + }, + { + "epoch": 12.477635782747603, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7811 + }, + { + "epoch": 12.47923322683706, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7812 + }, + { + "epoch": 12.480830670926517, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7813 + }, + { + "epoch": 12.482428115015974, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7814 + }, + { + "epoch": 12.484025559105431, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7815 + }, + { + "epoch": 12.485623003194888, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7816 + }, + { + "epoch": 12.487220447284345, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7817 + }, + { + "epoch": 12.488817891373802, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7818 + }, + { + "epoch": 12.49041533546326, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7819 + }, + { + "epoch": 12.492012779552716, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7820 + }, + { + "epoch": 12.493610223642172, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7821 + }, + { + "epoch": 12.495207667731629, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7822 + }, + { + "epoch": 12.496805111821086, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7823 + }, + { + "epoch": 12.498402555910543, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7824 + }, + { + "epoch": 12.5, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7825 + }, + { + "epoch": 12.501597444089457, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7826 + }, + { + "epoch": 12.503194888178914, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7827 + }, + { + "epoch": 12.504792332268371, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7828 + }, + { + "epoch": 12.506389776357828, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7829 + }, + { + "epoch": 12.507987220447284, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7830 + }, + { + "epoch": 12.50958466453674, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7831 + }, + { + "epoch": 12.511182108626198, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7832 + }, + { + "epoch": 12.512779552715655, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7833 + }, + { + "epoch": 12.514376996805112, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7834 + }, + { + "epoch": 12.515974440894569, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 7835 + }, + { + "epoch": 12.517571884984026, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7836 + }, + { + "epoch": 12.519169329073483, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7837 + }, + { + "epoch": 12.52076677316294, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7838 + }, + { + "epoch": 12.522364217252395, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7839 + }, + { + "epoch": 12.523961661341852, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 7840 + }, + { + "epoch": 12.52555910543131, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7841 + }, + { + "epoch": 12.527156549520766, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 7842 + }, + { + "epoch": 12.528753993610223, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7843 + }, + { + "epoch": 12.53035143769968, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7844 + }, + { + "epoch": 12.531948881789138, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7845 + }, + { + "epoch": 12.533546325878595, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7846 + }, + { + "epoch": 12.535143769968052, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7847 + }, + { + "epoch": 12.536741214057509, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7848 + }, + { + "epoch": 12.538338658146966, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7849 + }, + { + "epoch": 12.539936102236421, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7850 + }, + { + "epoch": 12.541533546325878, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7851 + }, + { + "epoch": 12.543130990415335, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7852 + }, + { + "epoch": 12.544728434504792, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7853 + }, + { + "epoch": 12.54632587859425, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7854 + }, + { + "epoch": 12.547923322683706, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7855 + }, + { + "epoch": 12.549520766773163, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7856 + }, + { + "epoch": 12.55111821086262, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7857 + }, + { + "epoch": 12.552715654952078, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7858 + }, + { + "epoch": 12.554313099041533, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7859 + }, + { + "epoch": 12.55591054313099, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7860 + }, + { + "epoch": 12.557507987220447, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7861 + }, + { + "epoch": 12.559105431309904, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7862 + }, + { + "epoch": 12.560702875399361, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7863 + }, + { + "epoch": 12.562300319488818, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7864 + }, + { + "epoch": 12.563897763578275, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7865 + }, + { + "epoch": 12.565495207667732, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7866 + }, + { + "epoch": 12.56709265175719, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7867 + }, + { + "epoch": 12.568690095846645, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7868 + }, + { + "epoch": 12.570287539936102, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7869 + }, + { + "epoch": 12.571884984025559, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7870 + }, + { + "epoch": 12.573482428115016, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7871 + }, + { + "epoch": 12.575079872204473, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7872 + }, + { + "epoch": 12.57667731629393, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7873 + }, + { + "epoch": 12.578274760383387, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7874 + }, + { + "epoch": 12.579872204472844, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7875 + }, + { + "epoch": 12.581469648562301, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7876 + }, + { + "epoch": 12.583067092651756, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7877 + }, + { + "epoch": 12.584664536741213, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7878 + }, + { + "epoch": 12.58626198083067, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7879 + }, + { + "epoch": 12.587859424920127, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 7880 + }, + { + "epoch": 12.589456869009584, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7881 + }, + { + "epoch": 12.591054313099042, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7882 + }, + { + "epoch": 12.592651757188499, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7883 + }, + { + "epoch": 12.594249201277956, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7884 + }, + { + "epoch": 12.595846645367413, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7885 + }, + { + "epoch": 12.59744408945687, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7886 + }, + { + "epoch": 12.599041533546325, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7887 + }, + { + "epoch": 12.600638977635782, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7888 + }, + { + "epoch": 12.60223642172524, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7889 + }, + { + "epoch": 12.603833865814696, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7890 + }, + { + "epoch": 12.605431309904153, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7891 + }, + { + "epoch": 12.60702875399361, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7892 + }, + { + "epoch": 12.608626198083067, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7893 + }, + { + "epoch": 12.610223642172524, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7894 + }, + { + "epoch": 12.611821086261982, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7895 + }, + { + "epoch": 12.613418530351439, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7896 + }, + { + "epoch": 12.615015974440894, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7897 + }, + { + "epoch": 12.616613418530351, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7898 + }, + { + "epoch": 12.618210862619808, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7899 + }, + { + "epoch": 12.619808306709265, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7900 + }, + { + "epoch": 12.621405750798722, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7901 + }, + { + "epoch": 12.62300319488818, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7902 + }, + { + "epoch": 12.624600638977636, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7903 + }, + { + "epoch": 12.626198083067093, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7904 + }, + { + "epoch": 12.62779552715655, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 7905 + }, + { + "epoch": 12.629392971246006, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 7906 + }, + { + "epoch": 12.630990415335463, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 7907 + }, + { + "epoch": 12.63258785942492, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7908 + }, + { + "epoch": 12.634185303514377, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7909 + }, + { + "epoch": 12.635782747603834, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7910 + }, + { + "epoch": 12.63738019169329, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 7911 + }, + { + "epoch": 12.638977635782748, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7912 + }, + { + "epoch": 12.640575079872205, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7913 + }, + { + "epoch": 12.642172523961662, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7914 + }, + { + "epoch": 12.643769968051117, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7915 + }, + { + "epoch": 12.645367412140574, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7916 + }, + { + "epoch": 12.646964856230031, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7917 + }, + { + "epoch": 12.648562300319488, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7918 + }, + { + "epoch": 12.650159744408946, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7919 + }, + { + "epoch": 12.651757188498403, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7920 + }, + { + "epoch": 12.65335463258786, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7921 + }, + { + "epoch": 12.654952076677317, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7922 + }, + { + "epoch": 12.656549520766774, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7923 + }, + { + "epoch": 12.65814696485623, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7924 + }, + { + "epoch": 12.659744408945686, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7925 + }, + { + "epoch": 12.661341853035143, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7926 + }, + { + "epoch": 12.6629392971246, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7927 + }, + { + "epoch": 12.664536741214057, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7928 + }, + { + "epoch": 12.666134185303514, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7929 + }, + { + "epoch": 12.667731629392971, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7930 + }, + { + "epoch": 12.669329073482428, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7931 + }, + { + "epoch": 12.670926517571885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 7932 + }, + { + "epoch": 12.672523961661343, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7933 + }, + { + "epoch": 12.6741214057508, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7934 + }, + { + "epoch": 12.675718849840255, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7935 + }, + { + "epoch": 12.677316293929712, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7936 + }, + { + "epoch": 12.678913738019169, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7937 + }, + { + "epoch": 12.680511182108626, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7938 + }, + { + "epoch": 12.682108626198083, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7939 + }, + { + "epoch": 12.68370607028754, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7940 + }, + { + "epoch": 12.685303514376997, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7941 + }, + { + "epoch": 12.686900958466454, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7942 + }, + { + "epoch": 12.688498402555911, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7943 + }, + { + "epoch": 12.690095846645367, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 7944 + }, + { + "epoch": 12.691693290734824, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 7945 + }, + { + "epoch": 12.69329073482428, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7946 + }, + { + "epoch": 12.694888178913738, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 7947 + }, + { + "epoch": 12.696485623003195, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7948 + }, + { + "epoch": 12.698083067092652, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7949 + }, + { + "epoch": 12.699680511182109, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7950 + }, + { + "epoch": 12.701277955271566, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7951 + }, + { + "epoch": 12.702875399361023, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7952 + }, + { + "epoch": 12.704472843450478, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7953 + }, + { + "epoch": 12.706070287539935, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7954 + }, + { + "epoch": 12.707667731629392, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7955 + }, + { + "epoch": 12.70926517571885, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7956 + }, + { + "epoch": 12.710862619808307, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7957 + }, + { + "epoch": 12.712460063897764, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7958 + }, + { + "epoch": 12.71405750798722, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7959 + }, + { + "epoch": 12.715654952076678, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7960 + }, + { + "epoch": 12.717252396166135, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7961 + }, + { + "epoch": 12.718849840255592, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7962 + }, + { + "epoch": 12.720447284345047, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7963 + }, + { + "epoch": 12.722044728434504, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7964 + }, + { + "epoch": 12.723642172523961, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7965 + }, + { + "epoch": 12.725239616613418, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7966 + }, + { + "epoch": 12.726837060702875, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7967 + }, + { + "epoch": 12.728434504792332, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7968 + }, + { + "epoch": 12.73003194888179, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7969 + }, + { + "epoch": 12.731629392971247, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7970 + }, + { + "epoch": 12.733226837060704, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7971 + }, + { + "epoch": 12.73482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7972 + }, + { + "epoch": 12.736421725239616, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7973 + }, + { + "epoch": 12.738019169329073, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7974 + }, + { + "epoch": 12.73961661341853, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7975 + }, + { + "epoch": 12.741214057507987, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7976 + }, + { + "epoch": 12.742811501597444, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 7977 + }, + { + "epoch": 12.744408945686901, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7978 + }, + { + "epoch": 12.746006389776358, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7979 + }, + { + "epoch": 12.747603833865815, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7980 + }, + { + "epoch": 12.749201277955272, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7981 + }, + { + "epoch": 12.750798722044728, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7982 + }, + { + "epoch": 12.752396166134185, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7983 + }, + { + "epoch": 12.753993610223642, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7984 + }, + { + "epoch": 12.755591054313099, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7985 + }, + { + "epoch": 12.757188498402556, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7986 + }, + { + "epoch": 12.758785942492013, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7987 + }, + { + "epoch": 12.76038338658147, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7988 + }, + { + "epoch": 12.761980830670927, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 7989 + }, + { + "epoch": 12.763578274760384, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7990 + }, + { + "epoch": 12.76517571884984, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7991 + }, + { + "epoch": 12.766773162939296, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7992 + }, + { + "epoch": 12.768370607028753, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7993 + }, + { + "epoch": 12.76996805111821, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7994 + }, + { + "epoch": 12.771565495207668, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7995 + }, + { + "epoch": 12.773162939297125, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7996 + }, + { + "epoch": 12.774760383386582, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7997 + }, + { + "epoch": 12.776357827476039, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7998 + }, + { + "epoch": 12.777955271565496, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 7999 + }, + { + "epoch": 12.779552715654953, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8000 + }, + { + "epoch": 12.781150159744408, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8001 + }, + { + "epoch": 12.782747603833865, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8002 + }, + { + "epoch": 12.784345047923322, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 8003 + }, + { + "epoch": 12.78594249201278, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 8004 + }, + { + "epoch": 12.787539936102236, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8005 + }, + { + "epoch": 12.789137380191693, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 8006 + }, + { + "epoch": 12.79073482428115, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8007 + }, + { + "epoch": 12.792332268370608, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8008 + }, + { + "epoch": 12.793929712460065, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8009 + }, + { + "epoch": 12.795527156549522, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 8010 + }, + { + "epoch": 12.797124600638977, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8011 + }, + { + "epoch": 12.798722044728434, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 8012 + }, + { + "epoch": 12.800319488817891, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8013 + }, + { + "epoch": 12.801916932907348, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8014 + }, + { + "epoch": 12.803514376996805, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8015 + }, + { + "epoch": 12.805111821086262, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8016 + }, + { + "epoch": 12.80670926517572, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8017 + }, + { + "epoch": 12.808306709265176, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8018 + }, + { + "epoch": 12.809904153354633, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8019 + }, + { + "epoch": 12.811501597444089, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8020 + }, + { + "epoch": 12.813099041533546, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 8021 + }, + { + "epoch": 12.814696485623003, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8022 + }, + { + "epoch": 12.81629392971246, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8023 + }, + { + "epoch": 12.817891373801917, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8024 + }, + { + "epoch": 12.819488817891374, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8025 + }, + { + "epoch": 12.821086261980831, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 8026 + }, + { + "epoch": 12.822683706070288, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 8027 + }, + { + "epoch": 12.824281150159745, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8028 + }, + { + "epoch": 12.8258785942492, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8029 + }, + { + "epoch": 12.827476038338657, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8030 + }, + { + "epoch": 12.829073482428115, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8031 + }, + { + "epoch": 12.830670926517572, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8032 + }, + { + "epoch": 12.832268370607029, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8033 + }, + { + "epoch": 12.833865814696486, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 8034 + }, + { + "epoch": 12.835463258785943, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 8035 + }, + { + "epoch": 12.8370607028754, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8036 + }, + { + "epoch": 12.838658146964857, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8037 + }, + { + "epoch": 12.840255591054314, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8038 + }, + { + "epoch": 12.84185303514377, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8039 + }, + { + "epoch": 12.843450479233226, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8040 + }, + { + "epoch": 12.845047923322683, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8041 + }, + { + "epoch": 12.84664536741214, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8042 + }, + { + "epoch": 12.848242811501597, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8043 + }, + { + "epoch": 12.849840255591054, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 8044 + }, + { + "epoch": 12.851437699680512, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 8045 + }, + { + "epoch": 12.853035143769969, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8046 + }, + { + "epoch": 12.854632587859426, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8047 + }, + { + "epoch": 12.856230031948883, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8048 + }, + { + "epoch": 12.857827476038338, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8049 + }, + { + "epoch": 12.859424920127795, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8050 + }, + { + "epoch": 12.861022364217252, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8051 + }, + { + "epoch": 12.86261980830671, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8052 + }, + { + "epoch": 12.864217252396166, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8053 + }, + { + "epoch": 12.865814696485623, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8054 + }, + { + "epoch": 12.86741214057508, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8055 + }, + { + "epoch": 12.869009584664537, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8056 + }, + { + "epoch": 12.870607028753994, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8057 + }, + { + "epoch": 12.87220447284345, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8058 + }, + { + "epoch": 12.873801916932907, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8059 + }, + { + "epoch": 12.875399361022364, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8060 + }, + { + "epoch": 12.87699680511182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8061 + }, + { + "epoch": 12.878594249201278, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8062 + }, + { + "epoch": 12.880191693290735, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8063 + }, + { + "epoch": 12.881789137380192, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8064 + }, + { + "epoch": 12.883386581469649, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8065 + }, + { + "epoch": 12.884984025559106, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8066 + }, + { + "epoch": 12.886581469648561, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8067 + }, + { + "epoch": 12.888178913738018, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8068 + }, + { + "epoch": 12.889776357827476, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 8069 + }, + { + "epoch": 12.891373801916933, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8070 + }, + { + "epoch": 12.89297124600639, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8071 + }, + { + "epoch": 12.894568690095847, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8072 + }, + { + "epoch": 12.896166134185304, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8073 + }, + { + "epoch": 12.89776357827476, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8074 + }, + { + "epoch": 12.899361022364218, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8075 + }, + { + "epoch": 12.900958466453675, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8076 + }, + { + "epoch": 12.90255591054313, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8077 + }, + { + "epoch": 12.904153354632587, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 8078 + }, + { + "epoch": 12.905750798722044, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8079 + }, + { + "epoch": 12.907348242811501, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8080 + }, + { + "epoch": 12.908945686900958, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8081 + }, + { + "epoch": 12.910543130990416, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8082 + }, + { + "epoch": 12.912140575079873, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8083 + }, + { + "epoch": 12.91373801916933, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8084 + }, + { + "epoch": 12.915335463258787, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 8085 + }, + { + "epoch": 12.916932907348244, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8086 + }, + { + "epoch": 12.918530351437699, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8087 + }, + { + "epoch": 12.920127795527156, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 8088 + }, + { + "epoch": 12.921725239616613, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8089 + }, + { + "epoch": 12.92332268370607, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8090 + }, + { + "epoch": 12.924920127795527, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8091 + }, + { + "epoch": 12.926517571884984, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8092 + }, + { + "epoch": 12.928115015974441, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8093 + }, + { + "epoch": 12.929712460063898, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8094 + }, + { + "epoch": 12.931309904153355, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8095 + }, + { + "epoch": 12.93290734824281, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8096 + }, + { + "epoch": 12.934504792332268, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 8097 + }, + { + "epoch": 12.936102236421725, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8098 + }, + { + "epoch": 12.937699680511182, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8099 + }, + { + "epoch": 12.939297124600639, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8100 + }, + { + "epoch": 12.940894568690096, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8101 + }, + { + "epoch": 12.942492012779553, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8102 + }, + { + "epoch": 12.94408945686901, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8103 + }, + { + "epoch": 12.945686900958467, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8104 + }, + { + "epoch": 12.947284345047922, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8105 + }, + { + "epoch": 12.94888178913738, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8106 + }, + { + "epoch": 12.950479233226837, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8107 + }, + { + "epoch": 12.952076677316294, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8108 + }, + { + "epoch": 12.95367412140575, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8109 + }, + { + "epoch": 12.955271565495208, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 8110 + }, + { + "epoch": 12.956869009584665, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8111 + }, + { + "epoch": 12.958466453674122, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8112 + }, + { + "epoch": 12.960063897763579, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8113 + }, + { + "epoch": 12.961661341853034, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8114 + }, + { + "epoch": 12.963258785942491, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8115 + }, + { + "epoch": 12.964856230031948, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 8116 + }, + { + "epoch": 12.966453674121405, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8117 + }, + { + "epoch": 12.968051118210862, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8118 + }, + { + "epoch": 12.96964856230032, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8119 + }, + { + "epoch": 12.971246006389777, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 8120 + }, + { + "epoch": 12.972843450479234, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 8121 + }, + { + "epoch": 12.97444089456869, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 8122 + }, + { + "epoch": 12.976038338658148, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8123 + }, + { + "epoch": 12.977635782747605, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 8124 + }, + { + "epoch": 12.97923322683706, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8125 + }, + { + "epoch": 12.980830670926517, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8126 + }, + { + "epoch": 12.982428115015974, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8127 + }, + { + "epoch": 12.984025559105431, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8128 + }, + { + "epoch": 12.985623003194888, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8129 + }, + { + "epoch": 12.987220447284345, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8130 + }, + { + "epoch": 12.988817891373802, + "grad_norm": 0.51171875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8131 + }, + { + "epoch": 12.99041533546326, + "grad_norm": 1.25, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 8132 + }, + { + "epoch": 12.992012779552716, + "grad_norm": 1.515625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 8133 + }, + { + "epoch": 12.993610223642172, + "grad_norm": 0.41015625, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 8134 + }, + { + "epoch": 12.995207667731629, + "grad_norm": 1.1640625, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 8135 + }, + { + "epoch": 12.996805111821086, + "grad_norm": 0.71875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 8136 + }, + { + "epoch": 12.998402555910543, + "grad_norm": 0.8046875, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 8137 + }, + { + "epoch": 13.0, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 8138 + }, + { + "epoch": 13.001597444089457, + "grad_norm": 0.73046875, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 8139 + }, + { + "epoch": 13.003194888178914, + "grad_norm": 1.0546875, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 8140 + }, + { + "epoch": 13.004792332268371, + "grad_norm": 0.87109375, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 8141 + }, + { + "epoch": 13.006389776357828, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 8142 + }, + { + "epoch": 13.007987220447284, + "grad_norm": 0.6015625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 8143 + }, + { + "epoch": 13.00958466453674, + "grad_norm": 0.5, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 8144 + }, + { + "epoch": 13.011182108626198, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 8145 + }, + { + "epoch": 13.012779552715655, + "grad_norm": 0.38671875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 8146 + }, + { + "epoch": 13.014376996805112, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 8147 + }, + { + "epoch": 13.015974440894569, + "grad_norm": 0.33203125, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 8148 + }, + { + "epoch": 13.017571884984026, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 8149 + }, + { + "epoch": 13.019169329073483, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 8150 + }, + { + "epoch": 13.02076677316294, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8151 + }, + { + "epoch": 13.022364217252397, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 8152 + }, + { + "epoch": 13.023961661341852, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 8153 + }, + { + "epoch": 13.02555910543131, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8154 + }, + { + "epoch": 13.027156549520766, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8155 + }, + { + "epoch": 13.028753993610223, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 8156 + }, + { + "epoch": 13.03035143769968, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8157 + }, + { + "epoch": 13.031948881789138, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 8158 + }, + { + "epoch": 13.033546325878595, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 8159 + }, + { + "epoch": 13.035143769968052, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8160 + }, + { + "epoch": 13.036741214057509, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 8161 + }, + { + "epoch": 13.038338658146964, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8162 + }, + { + "epoch": 13.039936102236421, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8163 + }, + { + "epoch": 13.041533546325878, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8164 + }, + { + "epoch": 13.043130990415335, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8165 + }, + { + "epoch": 13.044728434504792, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 8166 + }, + { + "epoch": 13.04632587859425, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8167 + }, + { + "epoch": 13.047923322683706, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8168 + }, + { + "epoch": 13.049520766773163, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8169 + }, + { + "epoch": 13.05111821086262, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8170 + }, + { + "epoch": 13.052715654952078, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8171 + }, + { + "epoch": 13.054313099041533, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8172 + }, + { + "epoch": 13.05591054313099, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8173 + }, + { + "epoch": 13.057507987220447, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8174 + }, + { + "epoch": 13.059105431309904, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8175 + }, + { + "epoch": 13.060702875399361, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 8176 + }, + { + "epoch": 13.062300319488818, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8177 + }, + { + "epoch": 13.063897763578275, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 8178 + }, + { + "epoch": 13.065495207667732, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8179 + }, + { + "epoch": 13.06709265175719, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8180 + }, + { + "epoch": 13.068690095846645, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8181 + }, + { + "epoch": 13.070287539936102, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8182 + }, + { + "epoch": 13.071884984025559, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8183 + }, + { + "epoch": 13.073482428115016, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8184 + }, + { + "epoch": 13.075079872204473, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8185 + }, + { + "epoch": 13.07667731629393, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8186 + }, + { + "epoch": 13.078274760383387, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8187 + }, + { + "epoch": 13.079872204472844, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8188 + }, + { + "epoch": 13.081469648562301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8189 + }, + { + "epoch": 13.083067092651758, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8190 + }, + { + "epoch": 13.084664536741213, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8191 + }, + { + "epoch": 13.08626198083067, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8192 + }, + { + "epoch": 13.087859424920127, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8193 + }, + { + "epoch": 13.089456869009584, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8194 + }, + { + "epoch": 13.091054313099042, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8195 + }, + { + "epoch": 13.092651757188499, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8196 + }, + { + "epoch": 13.094249201277956, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8197 + }, + { + "epoch": 13.095846645367413, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8198 + }, + { + "epoch": 13.09744408945687, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8199 + }, + { + "epoch": 13.099041533546325, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8200 + }, + { + "epoch": 13.100638977635782, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8201 + }, + { + "epoch": 13.10223642172524, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8202 + }, + { + "epoch": 13.103833865814696, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 8203 + }, + { + "epoch": 13.105431309904153, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8204 + }, + { + "epoch": 13.10702875399361, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8205 + }, + { + "epoch": 13.108626198083067, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8206 + }, + { + "epoch": 13.110223642172524, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8207 + }, + { + "epoch": 13.111821086261982, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8208 + }, + { + "epoch": 13.113418530351439, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8209 + }, + { + "epoch": 13.115015974440894, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 8210 + }, + { + "epoch": 13.116613418530351, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8211 + }, + { + "epoch": 13.118210862619808, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8212 + }, + { + "epoch": 13.119808306709265, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8213 + }, + { + "epoch": 13.121405750798722, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8214 + }, + { + "epoch": 13.12300319488818, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8215 + }, + { + "epoch": 13.124600638977636, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 8216 + }, + { + "epoch": 13.126198083067093, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8217 + }, + { + "epoch": 13.12779552715655, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8218 + }, + { + "epoch": 13.129392971246006, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8219 + }, + { + "epoch": 13.130990415335463, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8220 + }, + { + "epoch": 13.13258785942492, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8221 + }, + { + "epoch": 13.134185303514377, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8222 + }, + { + "epoch": 13.135782747603834, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8223 + }, + { + "epoch": 13.13738019169329, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8224 + }, + { + "epoch": 13.138977635782748, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8225 + }, + { + "epoch": 13.140575079872205, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8226 + }, + { + "epoch": 13.142172523961662, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8227 + }, + { + "epoch": 13.143769968051119, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8228 + }, + { + "epoch": 13.145367412140574, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8229 + }, + { + "epoch": 13.146964856230031, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 8230 + }, + { + "epoch": 13.148562300319488, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 8231 + }, + { + "epoch": 13.150159744408946, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8232 + }, + { + "epoch": 13.151757188498403, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8233 + }, + { + "epoch": 13.15335463258786, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8234 + }, + { + "epoch": 13.154952076677317, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8235 + }, + { + "epoch": 13.156549520766774, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8236 + }, + { + "epoch": 13.15814696485623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8237 + }, + { + "epoch": 13.159744408945686, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8238 + }, + { + "epoch": 13.161341853035143, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8239 + }, + { + "epoch": 13.1629392971246, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 8240 + }, + { + "epoch": 13.164536741214057, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 8241 + }, + { + "epoch": 13.166134185303514, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 8242 + }, + { + "epoch": 13.167731629392971, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8243 + }, + { + "epoch": 13.169329073482428, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8244 + }, + { + "epoch": 13.170926517571885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8245 + }, + { + "epoch": 13.172523961661343, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8246 + }, + { + "epoch": 13.1741214057508, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8247 + }, + { + "epoch": 13.175718849840255, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8248 + }, + { + "epoch": 13.177316293929712, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8249 + }, + { + "epoch": 13.178913738019169, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8250 + }, + { + "epoch": 13.180511182108626, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8251 + }, + { + "epoch": 13.182108626198083, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8252 + }, + { + "epoch": 13.18370607028754, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8253 + }, + { + "epoch": 13.185303514376997, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8254 + }, + { + "epoch": 13.186900958466454, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8255 + }, + { + "epoch": 13.188498402555911, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8256 + }, + { + "epoch": 13.190095846645367, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8257 + }, + { + "epoch": 13.191693290734824, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8258 + }, + { + "epoch": 13.19329073482428, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8259 + }, + { + "epoch": 13.194888178913738, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8260 + }, + { + "epoch": 13.196485623003195, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8261 + }, + { + "epoch": 13.198083067092652, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 8262 + }, + { + "epoch": 13.199680511182109, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8263 + }, + { + "epoch": 13.201277955271566, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8264 + }, + { + "epoch": 13.202875399361023, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8265 + }, + { + "epoch": 13.204472843450478, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 8266 + }, + { + "epoch": 13.206070287539935, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8267 + }, + { + "epoch": 13.207667731629392, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8268 + }, + { + "epoch": 13.20926517571885, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 8269 + }, + { + "epoch": 13.210862619808307, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8270 + }, + { + "epoch": 13.212460063897764, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8271 + }, + { + "epoch": 13.21405750798722, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8272 + }, + { + "epoch": 13.215654952076678, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 8273 + }, + { + "epoch": 13.217252396166135, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 8274 + }, + { + "epoch": 13.218849840255592, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8275 + }, + { + "epoch": 13.220447284345047, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8276 + }, + { + "epoch": 13.222044728434504, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8277 + }, + { + "epoch": 13.223642172523961, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 8278 + }, + { + "epoch": 13.225239616613418, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8279 + }, + { + "epoch": 13.226837060702875, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8280 + }, + { + "epoch": 13.228434504792332, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 8281 + }, + { + "epoch": 13.23003194888179, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8282 + }, + { + "epoch": 13.231629392971247, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8283 + }, + { + "epoch": 13.233226837060704, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8284 + }, + { + "epoch": 13.23482428115016, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8285 + }, + { + "epoch": 13.236421725239616, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8286 + }, + { + "epoch": 13.238019169329073, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8287 + }, + { + "epoch": 13.23961661341853, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8288 + }, + { + "epoch": 13.241214057507987, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8289 + }, + { + "epoch": 13.242811501597444, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8290 + }, + { + "epoch": 13.244408945686901, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8291 + }, + { + "epoch": 13.246006389776358, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8292 + }, + { + "epoch": 13.247603833865815, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8293 + }, + { + "epoch": 13.249201277955272, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8294 + }, + { + "epoch": 13.250798722044728, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8295 + }, + { + "epoch": 13.252396166134185, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8296 + }, + { + "epoch": 13.253993610223642, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8297 + }, + { + "epoch": 13.255591054313099, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8298 + }, + { + "epoch": 13.257188498402556, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8299 + }, + { + "epoch": 13.258785942492013, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8300 + }, + { + "epoch": 13.26038338658147, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8301 + }, + { + "epoch": 13.261980830670927, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8302 + }, + { + "epoch": 13.263578274760384, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8303 + }, + { + "epoch": 13.26517571884984, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8304 + }, + { + "epoch": 13.266773162939296, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8305 + }, + { + "epoch": 13.268370607028753, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8306 + }, + { + "epoch": 13.26996805111821, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8307 + }, + { + "epoch": 13.271565495207668, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8308 + }, + { + "epoch": 13.273162939297125, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8309 + }, + { + "epoch": 13.274760383386582, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8310 + }, + { + "epoch": 13.276357827476039, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8311 + }, + { + "epoch": 13.277955271565496, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8312 + }, + { + "epoch": 13.279552715654953, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8313 + }, + { + "epoch": 13.281150159744408, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8314 + }, + { + "epoch": 13.282747603833865, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 8315 + }, + { + "epoch": 13.284345047923322, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8316 + }, + { + "epoch": 13.28594249201278, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8317 + }, + { + "epoch": 13.287539936102236, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8318 + }, + { + "epoch": 13.289137380191693, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8319 + }, + { + "epoch": 13.29073482428115, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8320 + }, + { + "epoch": 13.292332268370608, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8321 + }, + { + "epoch": 13.293929712460065, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8322 + }, + { + "epoch": 13.295527156549522, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8323 + }, + { + "epoch": 13.297124600638977, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8324 + }, + { + "epoch": 13.298722044728434, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8325 + }, + { + "epoch": 13.300319488817891, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8326 + }, + { + "epoch": 13.301916932907348, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8327 + }, + { + "epoch": 13.303514376996805, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8328 + }, + { + "epoch": 13.305111821086262, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 8329 + }, + { + "epoch": 13.30670926517572, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8330 + }, + { + "epoch": 13.308306709265176, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8331 + }, + { + "epoch": 13.309904153354633, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8332 + }, + { + "epoch": 13.311501597444089, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8333 + }, + { + "epoch": 13.313099041533546, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 8334 + }, + { + "epoch": 13.314696485623003, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8335 + }, + { + "epoch": 13.31629392971246, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 8336 + }, + { + "epoch": 13.317891373801917, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8337 + }, + { + "epoch": 13.319488817891374, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8338 + }, + { + "epoch": 13.321086261980831, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8339 + }, + { + "epoch": 13.322683706070288, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 8340 + }, + { + "epoch": 13.324281150159745, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8341 + }, + { + "epoch": 13.3258785942492, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8342 + }, + { + "epoch": 13.327476038338657, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8343 + }, + { + "epoch": 13.329073482428115, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8344 + }, + { + "epoch": 13.330670926517572, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8345 + }, + { + "epoch": 13.332268370607029, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8346 + }, + { + "epoch": 13.333865814696486, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8347 + }, + { + "epoch": 13.335463258785943, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8348 + }, + { + "epoch": 13.3370607028754, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8349 + }, + { + "epoch": 13.338658146964857, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 8350 + }, + { + "epoch": 13.340255591054314, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8351 + }, + { + "epoch": 13.34185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8352 + }, + { + "epoch": 13.343450479233226, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8353 + }, + { + "epoch": 13.345047923322683, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8354 + }, + { + "epoch": 13.34664536741214, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8355 + }, + { + "epoch": 13.348242811501597, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8356 + }, + { + "epoch": 13.349840255591054, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8357 + }, + { + "epoch": 13.351437699680512, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8358 + }, + { + "epoch": 13.353035143769969, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8359 + }, + { + "epoch": 13.354632587859426, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8360 + }, + { + "epoch": 13.356230031948883, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8361 + }, + { + "epoch": 13.357827476038338, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8362 + }, + { + "epoch": 13.359424920127795, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8363 + }, + { + "epoch": 13.361022364217252, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8364 + }, + { + "epoch": 13.36261980830671, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8365 + }, + { + "epoch": 13.364217252396166, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8366 + }, + { + "epoch": 13.365814696485623, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8367 + }, + { + "epoch": 13.36741214057508, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8368 + }, + { + "epoch": 13.369009584664537, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 8369 + }, + { + "epoch": 13.370607028753994, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8370 + }, + { + "epoch": 13.37220447284345, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8371 + }, + { + "epoch": 13.373801916932907, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8372 + }, + { + "epoch": 13.375399361022364, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8373 + }, + { + "epoch": 13.37699680511182, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8374 + }, + { + "epoch": 13.378594249201278, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8375 + }, + { + "epoch": 13.380191693290735, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8376 + }, + { + "epoch": 13.381789137380192, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8377 + }, + { + "epoch": 13.383386581469649, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8378 + }, + { + "epoch": 13.384984025559106, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8379 + }, + { + "epoch": 13.386581469648561, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8380 + }, + { + "epoch": 13.388178913738018, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 8381 + }, + { + "epoch": 13.389776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8382 + }, + { + "epoch": 13.391373801916933, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8383 + }, + { + "epoch": 13.39297124600639, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8384 + }, + { + "epoch": 13.394568690095847, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8385 + }, + { + "epoch": 13.396166134185304, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 8386 + }, + { + "epoch": 13.39776357827476, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8387 + }, + { + "epoch": 13.399361022364218, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 8388 + }, + { + "epoch": 13.400958466453675, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8389 + }, + { + "epoch": 13.40255591054313, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8390 + }, + { + "epoch": 13.404153354632587, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8391 + }, + { + "epoch": 13.405750798722044, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8392 + }, + { + "epoch": 13.407348242811501, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8393 + }, + { + "epoch": 13.408945686900958, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 8394 + }, + { + "epoch": 13.410543130990416, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8395 + }, + { + "epoch": 13.412140575079873, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8396 + }, + { + "epoch": 13.41373801916933, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8397 + }, + { + "epoch": 13.415335463258787, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8398 + }, + { + "epoch": 13.416932907348242, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8399 + }, + { + "epoch": 13.418530351437699, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8400 + }, + { + "epoch": 13.420127795527156, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8401 + }, + { + "epoch": 13.421725239616613, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8402 + }, + { + "epoch": 13.42332268370607, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8403 + }, + { + "epoch": 13.424920127795527, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8404 + }, + { + "epoch": 13.426517571884984, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8405 + }, + { + "epoch": 13.428115015974441, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8406 + }, + { + "epoch": 13.429712460063898, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8407 + }, + { + "epoch": 13.431309904153355, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8408 + }, + { + "epoch": 13.43290734824281, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8409 + }, + { + "epoch": 13.434504792332268, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8410 + }, + { + "epoch": 13.436102236421725, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8411 + }, + { + "epoch": 13.437699680511182, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 8412 + }, + { + "epoch": 13.439297124600639, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8413 + }, + { + "epoch": 13.440894568690096, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8414 + }, + { + "epoch": 13.442492012779553, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8415 + }, + { + "epoch": 13.44408945686901, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8416 + }, + { + "epoch": 13.445686900958467, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8417 + }, + { + "epoch": 13.447284345047922, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8418 + }, + { + "epoch": 13.44888178913738, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8419 + }, + { + "epoch": 13.450479233226837, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8420 + }, + { + "epoch": 13.452076677316294, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8421 + }, + { + "epoch": 13.45367412140575, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 8422 + }, + { + "epoch": 13.455271565495208, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8423 + }, + { + "epoch": 13.456869009584665, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8424 + }, + { + "epoch": 13.458466453674122, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8425 + }, + { + "epoch": 13.460063897763579, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8426 + }, + { + "epoch": 13.461661341853036, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8427 + }, + { + "epoch": 13.463258785942491, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8428 + }, + { + "epoch": 13.464856230031948, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8429 + }, + { + "epoch": 13.466453674121405, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8430 + }, + { + "epoch": 13.468051118210862, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8431 + }, + { + "epoch": 13.46964856230032, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8432 + }, + { + "epoch": 13.471246006389777, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8433 + }, + { + "epoch": 13.472843450479234, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8434 + }, + { + "epoch": 13.47444089456869, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8435 + }, + { + "epoch": 13.476038338658148, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8436 + }, + { + "epoch": 13.477635782747603, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 8437 + }, + { + "epoch": 13.47923322683706, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8438 + }, + { + "epoch": 13.480830670926517, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8439 + }, + { + "epoch": 13.482428115015974, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8440 + }, + { + "epoch": 13.484025559105431, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8441 + }, + { + "epoch": 13.485623003194888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8442 + }, + { + "epoch": 13.487220447284345, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 8443 + }, + { + "epoch": 13.488817891373802, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8444 + }, + { + "epoch": 13.49041533546326, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8445 + }, + { + "epoch": 13.492012779552716, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8446 + }, + { + "epoch": 13.493610223642172, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8447 + }, + { + "epoch": 13.495207667731629, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 8448 + }, + { + "epoch": 13.496805111821086, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8449 + }, + { + "epoch": 13.498402555910543, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8450 + }, + { + "epoch": 13.5, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8451 + }, + { + "epoch": 13.501597444089457, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 8452 + }, + { + "epoch": 13.503194888178914, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8453 + }, + { + "epoch": 13.504792332268371, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8454 + }, + { + "epoch": 13.506389776357828, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8455 + }, + { + "epoch": 13.507987220447284, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8456 + }, + { + "epoch": 13.50958466453674, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8457 + }, + { + "epoch": 13.511182108626198, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8458 + }, + { + "epoch": 13.512779552715655, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8459 + }, + { + "epoch": 13.514376996805112, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 8460 + }, + { + "epoch": 13.515974440894569, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8461 + }, + { + "epoch": 13.517571884984026, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8462 + }, + { + "epoch": 13.519169329073483, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8463 + }, + { + "epoch": 13.52076677316294, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8464 + }, + { + "epoch": 13.522364217252395, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8465 + }, + { + "epoch": 13.523961661341852, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8466 + }, + { + "epoch": 13.52555910543131, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 8467 + }, + { + "epoch": 13.527156549520766, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8468 + }, + { + "epoch": 13.528753993610223, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 8469 + }, + { + "epoch": 13.53035143769968, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8470 + }, + { + "epoch": 13.531948881789138, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8471 + }, + { + "epoch": 13.533546325878595, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8472 + }, + { + "epoch": 13.535143769968052, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8473 + }, + { + "epoch": 13.536741214057509, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8474 + }, + { + "epoch": 13.538338658146966, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8475 + }, + { + "epoch": 13.539936102236421, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8476 + }, + { + "epoch": 13.541533546325878, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8477 + }, + { + "epoch": 13.543130990415335, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 8478 + }, + { + "epoch": 13.544728434504792, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8479 + }, + { + "epoch": 13.54632587859425, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8480 + }, + { + "epoch": 13.547923322683706, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8481 + }, + { + "epoch": 13.549520766773163, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8482 + }, + { + "epoch": 13.55111821086262, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8483 + }, + { + "epoch": 13.552715654952078, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 8484 + }, + { + "epoch": 13.554313099041533, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8485 + }, + { + "epoch": 13.55591054313099, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8486 + }, + { + "epoch": 13.557507987220447, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8487 + }, + { + "epoch": 13.559105431309904, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8488 + }, + { + "epoch": 13.560702875399361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8489 + }, + { + "epoch": 13.562300319488818, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8490 + }, + { + "epoch": 13.563897763578275, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8491 + }, + { + "epoch": 13.565495207667732, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8492 + }, + { + "epoch": 13.56709265175719, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8493 + }, + { + "epoch": 13.568690095846645, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8494 + }, + { + "epoch": 13.570287539936102, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8495 + }, + { + "epoch": 13.571884984025559, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8496 + }, + { + "epoch": 13.573482428115016, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8497 + }, + { + "epoch": 13.575079872204473, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8498 + }, + { + "epoch": 13.57667731629393, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8499 + }, + { + "epoch": 13.578274760383387, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8500 + }, + { + "epoch": 13.579872204472844, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 8501 + }, + { + "epoch": 13.581469648562301, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8502 + }, + { + "epoch": 13.583067092651756, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8503 + }, + { + "epoch": 13.584664536741213, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8504 + }, + { + "epoch": 13.58626198083067, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8505 + }, + { + "epoch": 13.587859424920127, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8506 + }, + { + "epoch": 13.589456869009584, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8507 + }, + { + "epoch": 13.591054313099042, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8508 + }, + { + "epoch": 13.592651757188499, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8509 + }, + { + "epoch": 13.594249201277956, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8510 + }, + { + "epoch": 13.595846645367413, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8511 + }, + { + "epoch": 13.59744408945687, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8512 + }, + { + "epoch": 13.599041533546325, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8513 + }, + { + "epoch": 13.600638977635782, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8514 + }, + { + "epoch": 13.60223642172524, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8515 + }, + { + "epoch": 13.603833865814696, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8516 + }, + { + "epoch": 13.605431309904153, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8517 + }, + { + "epoch": 13.60702875399361, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8518 + }, + { + "epoch": 13.608626198083067, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8519 + }, + { + "epoch": 13.610223642172524, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8520 + }, + { + "epoch": 13.611821086261982, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8521 + }, + { + "epoch": 13.613418530351439, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8522 + }, + { + "epoch": 13.615015974440894, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8523 + }, + { + "epoch": 13.616613418530351, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8524 + }, + { + "epoch": 13.618210862619808, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8525 + }, + { + "epoch": 13.619808306709265, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 8526 + }, + { + "epoch": 13.621405750798722, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8527 + }, + { + "epoch": 13.62300319488818, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8528 + }, + { + "epoch": 13.624600638977636, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8529 + }, + { + "epoch": 13.626198083067093, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8530 + }, + { + "epoch": 13.62779552715655, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8531 + }, + { + "epoch": 13.629392971246006, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 8532 + }, + { + "epoch": 13.630990415335463, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8533 + }, + { + "epoch": 13.63258785942492, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8534 + }, + { + "epoch": 13.634185303514377, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8535 + }, + { + "epoch": 13.635782747603834, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8536 + }, + { + "epoch": 13.63738019169329, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8537 + }, + { + "epoch": 13.638977635782748, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8538 + }, + { + "epoch": 13.640575079872205, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8539 + }, + { + "epoch": 13.642172523961662, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8540 + }, + { + "epoch": 13.643769968051117, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8541 + }, + { + "epoch": 13.645367412140574, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8542 + }, + { + "epoch": 13.646964856230031, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8543 + }, + { + "epoch": 13.648562300319488, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8544 + }, + { + "epoch": 13.650159744408946, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8545 + }, + { + "epoch": 13.651757188498403, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8546 + }, + { + "epoch": 13.65335463258786, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8547 + }, + { + "epoch": 13.654952076677317, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 8548 + }, + { + "epoch": 13.656549520766774, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8549 + }, + { + "epoch": 13.65814696485623, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8550 + }, + { + "epoch": 13.659744408945686, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 8551 + }, + { + "epoch": 13.661341853035143, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8552 + }, + { + "epoch": 13.6629392971246, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8553 + }, + { + "epoch": 13.664536741214057, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8554 + }, + { + "epoch": 13.666134185303514, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8555 + }, + { + "epoch": 13.667731629392971, + "grad_norm": 0.03173828125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8556 + }, + { + "epoch": 13.669329073482428, + "grad_norm": 0.031494140625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8557 + }, + { + "epoch": 13.670926517571885, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8558 + }, + { + "epoch": 13.672523961661343, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8559 + }, + { + "epoch": 13.6741214057508, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8560 + }, + { + "epoch": 13.675718849840255, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8561 + }, + { + "epoch": 13.677316293929712, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8562 + }, + { + "epoch": 13.678913738019169, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8563 + }, + { + "epoch": 13.680511182108626, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8564 + }, + { + "epoch": 13.682108626198083, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 8565 + }, + { + "epoch": 13.68370607028754, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8566 + }, + { + "epoch": 13.685303514376997, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8567 + }, + { + "epoch": 13.686900958466454, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8568 + }, + { + "epoch": 13.688498402555911, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8569 + }, + { + "epoch": 13.690095846645367, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8570 + }, + { + "epoch": 13.691693290734824, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8571 + }, + { + "epoch": 13.69329073482428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8572 + }, + { + "epoch": 13.694888178913738, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8573 + }, + { + "epoch": 13.696485623003195, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8574 + }, + { + "epoch": 13.698083067092652, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8575 + }, + { + "epoch": 13.699680511182109, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8576 + }, + { + "epoch": 13.701277955271566, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8577 + }, + { + "epoch": 13.702875399361023, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8578 + }, + { + "epoch": 13.704472843450478, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8579 + }, + { + "epoch": 13.706070287539935, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8580 + }, + { + "epoch": 13.707667731629392, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8581 + }, + { + "epoch": 13.70926517571885, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8582 + }, + { + "epoch": 13.710862619808307, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8583 + }, + { + "epoch": 13.712460063897764, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8584 + }, + { + "epoch": 13.71405750798722, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8585 + }, + { + "epoch": 13.715654952076678, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8586 + }, + { + "epoch": 13.717252396166135, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 8587 + }, + { + "epoch": 13.718849840255592, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8588 + }, + { + "epoch": 13.720447284345047, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 8589 + }, + { + "epoch": 13.722044728434504, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8590 + }, + { + "epoch": 13.723642172523961, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8591 + }, + { + "epoch": 13.725239616613418, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8592 + }, + { + "epoch": 13.726837060702875, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8593 + }, + { + "epoch": 13.728434504792332, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8594 + }, + { + "epoch": 13.73003194888179, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8595 + }, + { + "epoch": 13.731629392971247, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8596 + }, + { + "epoch": 13.733226837060704, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8597 + }, + { + "epoch": 13.73482428115016, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 8598 + }, + { + "epoch": 13.736421725239616, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8599 + }, + { + "epoch": 13.738019169329073, + "grad_norm": 0.031494140625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8600 + }, + { + "epoch": 13.73961661341853, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8601 + }, + { + "epoch": 13.741214057507987, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8602 + }, + { + "epoch": 13.742811501597444, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8603 + }, + { + "epoch": 13.744408945686901, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8604 + }, + { + "epoch": 13.746006389776358, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8605 + }, + { + "epoch": 13.747603833865815, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 8606 + }, + { + "epoch": 13.749201277955272, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8607 + }, + { + "epoch": 13.750798722044728, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8608 + }, + { + "epoch": 13.752396166134185, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8609 + }, + { + "epoch": 13.753993610223642, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8610 + }, + { + "epoch": 13.755591054313099, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8611 + }, + { + "epoch": 13.757188498402556, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8612 + }, + { + "epoch": 13.758785942492013, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8613 + }, + { + "epoch": 13.76038338658147, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8614 + }, + { + "epoch": 13.761980830670927, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8615 + }, + { + "epoch": 13.763578274760384, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8616 + }, + { + "epoch": 13.76517571884984, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8617 + }, + { + "epoch": 13.766773162939296, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8618 + }, + { + "epoch": 13.768370607028753, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8619 + }, + { + "epoch": 13.76996805111821, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8620 + }, + { + "epoch": 13.771565495207668, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8621 + }, + { + "epoch": 13.773162939297125, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8622 + }, + { + "epoch": 13.774760383386582, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 8623 + }, + { + "epoch": 13.776357827476039, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8624 + }, + { + "epoch": 13.777955271565496, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8625 + }, + { + "epoch": 13.779552715654953, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 8626 + }, + { + "epoch": 13.781150159744408, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 8627 + }, + { + "epoch": 13.782747603833865, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8628 + }, + { + "epoch": 13.784345047923322, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8629 + }, + { + "epoch": 13.78594249201278, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8630 + }, + { + "epoch": 13.787539936102236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8631 + }, + { + "epoch": 13.789137380191693, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8632 + }, + { + "epoch": 13.79073482428115, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8633 + }, + { + "epoch": 13.792332268370608, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8634 + }, + { + "epoch": 13.793929712460065, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8635 + }, + { + "epoch": 13.795527156549522, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 8636 + }, + { + "epoch": 13.797124600638977, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8637 + }, + { + "epoch": 13.798722044728434, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8638 + }, + { + "epoch": 13.800319488817891, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8639 + }, + { + "epoch": 13.801916932907348, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 8640 + }, + { + "epoch": 13.803514376996805, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8641 + }, + { + "epoch": 13.805111821086262, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8642 + }, + { + "epoch": 13.80670926517572, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8643 + }, + { + "epoch": 13.808306709265176, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8644 + }, + { + "epoch": 13.809904153354633, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8645 + }, + { + "epoch": 13.811501597444089, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8646 + }, + { + "epoch": 13.813099041533546, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8647 + }, + { + "epoch": 13.814696485623003, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8648 + }, + { + "epoch": 13.81629392971246, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8649 + }, + { + "epoch": 13.817891373801917, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8650 + }, + { + "epoch": 13.819488817891374, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8651 + }, + { + "epoch": 13.821086261980831, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8652 + }, + { + "epoch": 13.822683706070288, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8653 + }, + { + "epoch": 13.824281150159745, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8654 + }, + { + "epoch": 13.8258785942492, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8655 + }, + { + "epoch": 13.827476038338657, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8656 + }, + { + "epoch": 13.829073482428115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8657 + }, + { + "epoch": 13.830670926517572, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8658 + }, + { + "epoch": 13.832268370607029, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8659 + }, + { + "epoch": 13.833865814696486, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8660 + }, + { + "epoch": 13.835463258785943, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8661 + }, + { + "epoch": 13.8370607028754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8662 + }, + { + "epoch": 13.838658146964857, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8663 + }, + { + "epoch": 13.840255591054314, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8664 + }, + { + "epoch": 13.84185303514377, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8665 + }, + { + "epoch": 13.843450479233226, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8666 + }, + { + "epoch": 13.845047923322683, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8667 + }, + { + "epoch": 13.84664536741214, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8668 + }, + { + "epoch": 13.848242811501597, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8669 + }, + { + "epoch": 13.849840255591054, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8670 + }, + { + "epoch": 13.851437699680512, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8671 + }, + { + "epoch": 13.853035143769969, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8672 + }, + { + "epoch": 13.854632587859426, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8673 + }, + { + "epoch": 13.856230031948883, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8674 + }, + { + "epoch": 13.857827476038338, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8675 + }, + { + "epoch": 13.859424920127795, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8676 + }, + { + "epoch": 13.861022364217252, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8677 + }, + { + "epoch": 13.86261980830671, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8678 + }, + { + "epoch": 13.864217252396166, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8679 + }, + { + "epoch": 13.865814696485623, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8680 + }, + { + "epoch": 13.86741214057508, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8681 + }, + { + "epoch": 13.869009584664537, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8682 + }, + { + "epoch": 13.870607028753994, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8683 + }, + { + "epoch": 13.87220447284345, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8684 + }, + { + "epoch": 13.873801916932907, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 8685 + }, + { + "epoch": 13.875399361022364, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8686 + }, + { + "epoch": 13.87699680511182, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8687 + }, + { + "epoch": 13.878594249201278, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8688 + }, + { + "epoch": 13.880191693290735, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8689 + }, + { + "epoch": 13.881789137380192, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 8690 + }, + { + "epoch": 13.883386581469649, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8691 + }, + { + "epoch": 13.884984025559106, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8692 + }, + { + "epoch": 13.886581469648561, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8693 + }, + { + "epoch": 13.888178913738018, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8694 + }, + { + "epoch": 13.889776357827476, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8695 + }, + { + "epoch": 13.891373801916933, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8696 + }, + { + "epoch": 13.89297124600639, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8697 + }, + { + "epoch": 13.894568690095847, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8698 + }, + { + "epoch": 13.896166134185304, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8699 + }, + { + "epoch": 13.89776357827476, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8700 + }, + { + "epoch": 13.899361022364218, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8701 + }, + { + "epoch": 13.900958466453675, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8702 + }, + { + "epoch": 13.90255591054313, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8703 + }, + { + "epoch": 13.904153354632587, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8704 + }, + { + "epoch": 13.905750798722044, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8705 + }, + { + "epoch": 13.907348242811501, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8706 + }, + { + "epoch": 13.908945686900958, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8707 + }, + { + "epoch": 13.910543130990416, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8708 + }, + { + "epoch": 13.912140575079873, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8709 + }, + { + "epoch": 13.91373801916933, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8710 + }, + { + "epoch": 13.915335463258787, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8711 + }, + { + "epoch": 13.916932907348244, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8712 + }, + { + "epoch": 13.918530351437699, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 8713 + }, + { + "epoch": 13.920127795527156, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8714 + }, + { + "epoch": 13.921725239616613, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8715 + }, + { + "epoch": 13.92332268370607, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8716 + }, + { + "epoch": 13.924920127795527, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8717 + }, + { + "epoch": 13.926517571884984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 8718 + }, + { + "epoch": 13.928115015974441, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8719 + }, + { + "epoch": 13.929712460063898, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8720 + }, + { + "epoch": 13.931309904153355, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8721 + }, + { + "epoch": 13.93290734824281, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8722 + }, + { + "epoch": 13.934504792332268, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8723 + }, + { + "epoch": 13.936102236421725, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8724 + }, + { + "epoch": 13.937699680511182, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8725 + }, + { + "epoch": 13.939297124600639, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8726 + }, + { + "epoch": 13.940894568690096, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8727 + }, + { + "epoch": 13.942492012779553, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8728 + }, + { + "epoch": 13.94408945686901, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8729 + }, + { + "epoch": 13.945686900958467, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8730 + }, + { + "epoch": 13.947284345047922, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8731 + }, + { + "epoch": 13.94888178913738, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8732 + }, + { + "epoch": 13.950479233226837, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 8733 + }, + { + "epoch": 13.952076677316294, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8734 + }, + { + "epoch": 13.95367412140575, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8735 + }, + { + "epoch": 13.955271565495208, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8736 + }, + { + "epoch": 13.956869009584665, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8737 + }, + { + "epoch": 13.958466453674122, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8738 + }, + { + "epoch": 13.960063897763579, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8739 + }, + { + "epoch": 13.961661341853034, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8740 + }, + { + "epoch": 13.963258785942491, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8741 + }, + { + "epoch": 13.964856230031948, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8742 + }, + { + "epoch": 13.966453674121405, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8743 + }, + { + "epoch": 13.968051118210862, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8744 + }, + { + "epoch": 13.96964856230032, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8745 + }, + { + "epoch": 13.971246006389777, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8746 + }, + { + "epoch": 13.972843450479234, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 8747 + }, + { + "epoch": 13.97444089456869, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8748 + }, + { + "epoch": 13.976038338658148, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8749 + }, + { + "epoch": 13.977635782747605, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8750 + }, + { + "epoch": 13.97923322683706, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8751 + }, + { + "epoch": 13.980830670926517, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8752 + }, + { + "epoch": 13.982428115015974, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8753 + }, + { + "epoch": 13.984025559105431, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8754 + }, + { + "epoch": 13.985623003194888, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 8755 + }, + { + "epoch": 13.987220447284345, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8756 + }, + { + "epoch": 13.988817891373802, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8757 + }, + { + "epoch": 13.99041533546326, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8758 + }, + { + "epoch": 13.992012779552716, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8759 + }, + { + "epoch": 13.993610223642172, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8760 + }, + { + "epoch": 13.995207667731629, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 8761 + }, + { + "epoch": 13.996805111821086, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8762 + }, + { + "epoch": 13.998402555910543, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8763 + }, + { + "epoch": 14.0, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8764 + }, + { + "epoch": 14.001597444089457, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8765 + }, + { + "epoch": 14.003194888178914, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8766 + }, + { + "epoch": 14.004792332268371, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8767 + }, + { + "epoch": 14.006389776357828, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8768 + }, + { + "epoch": 14.007987220447284, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8769 + }, + { + "epoch": 14.00958466453674, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 8770 + }, + { + "epoch": 14.011182108626198, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8771 + }, + { + "epoch": 14.012779552715655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8772 + }, + { + "epoch": 14.014376996805112, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8773 + }, + { + "epoch": 14.015974440894569, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8774 + }, + { + "epoch": 14.017571884984026, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8775 + }, + { + "epoch": 14.019169329073483, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8776 + }, + { + "epoch": 14.02076677316294, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8777 + }, + { + "epoch": 14.022364217252397, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8778 + }, + { + "epoch": 14.023961661341852, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8779 + }, + { + "epoch": 14.02555910543131, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8780 + }, + { + "epoch": 14.027156549520766, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8781 + }, + { + "epoch": 14.028753993610223, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8782 + }, + { + "epoch": 14.03035143769968, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8783 + }, + { + "epoch": 14.031948881789138, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8784 + }, + { + "epoch": 14.033546325878595, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8785 + }, + { + "epoch": 14.035143769968052, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8786 + }, + { + "epoch": 14.036741214057509, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8787 + }, + { + "epoch": 14.038338658146964, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 8788 + }, + { + "epoch": 14.039936102236421, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8789 + }, + { + "epoch": 14.041533546325878, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8790 + }, + { + "epoch": 14.043130990415335, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8791 + }, + { + "epoch": 14.044728434504792, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8792 + }, + { + "epoch": 14.04632587859425, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8793 + }, + { + "epoch": 14.047923322683706, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8794 + }, + { + "epoch": 14.049520766773163, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8795 + }, + { + "epoch": 14.05111821086262, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8796 + }, + { + "epoch": 14.052715654952078, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8797 + }, + { + "epoch": 14.054313099041533, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8798 + }, + { + "epoch": 14.05591054313099, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 8799 + }, + { + "epoch": 14.057507987220447, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8800 + }, + { + "epoch": 14.059105431309904, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8801 + }, + { + "epoch": 14.060702875399361, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8802 + }, + { + "epoch": 14.062300319488818, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8803 + }, + { + "epoch": 14.063897763578275, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8804 + }, + { + "epoch": 14.065495207667732, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8805 + }, + { + "epoch": 14.06709265175719, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8806 + }, + { + "epoch": 14.068690095846645, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8807 + }, + { + "epoch": 14.070287539936102, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8808 + }, + { + "epoch": 14.071884984025559, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8809 + }, + { + "epoch": 14.073482428115016, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8810 + }, + { + "epoch": 14.075079872204473, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8811 + }, + { + "epoch": 14.07667731629393, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8812 + }, + { + "epoch": 14.078274760383387, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8813 + }, + { + "epoch": 14.079872204472844, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8814 + }, + { + "epoch": 14.081469648562301, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8815 + }, + { + "epoch": 14.083067092651758, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8816 + }, + { + "epoch": 14.084664536741213, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8817 + }, + { + "epoch": 14.08626198083067, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8818 + }, + { + "epoch": 14.087859424920127, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8819 + }, + { + "epoch": 14.089456869009584, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8820 + }, + { + "epoch": 14.091054313099042, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8821 + }, + { + "epoch": 14.092651757188499, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8822 + }, + { + "epoch": 14.094249201277956, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8823 + }, + { + "epoch": 14.095846645367413, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8824 + }, + { + "epoch": 14.09744408945687, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 8825 + }, + { + "epoch": 14.099041533546325, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8826 + }, + { + "epoch": 14.100638977635782, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8827 + }, + { + "epoch": 14.10223642172524, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8828 + }, + { + "epoch": 14.103833865814696, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8829 + }, + { + "epoch": 14.105431309904153, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8830 + }, + { + "epoch": 14.10702875399361, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8831 + }, + { + "epoch": 14.108626198083067, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8832 + }, + { + "epoch": 14.110223642172524, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8833 + }, + { + "epoch": 14.111821086261982, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8834 + }, + { + "epoch": 14.113418530351439, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 8835 + }, + { + "epoch": 14.115015974440894, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8836 + }, + { + "epoch": 14.116613418530351, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8837 + }, + { + "epoch": 14.118210862619808, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8838 + }, + { + "epoch": 14.119808306709265, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8839 + }, + { + "epoch": 14.121405750798722, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8840 + }, + { + "epoch": 14.12300319488818, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8841 + }, + { + "epoch": 14.124600638977636, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8842 + }, + { + "epoch": 14.126198083067093, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8843 + }, + { + "epoch": 14.12779552715655, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8844 + }, + { + "epoch": 14.129392971246006, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8845 + }, + { + "epoch": 14.130990415335463, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 8846 + }, + { + "epoch": 14.13258785942492, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8847 + }, + { + "epoch": 14.134185303514377, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8848 + }, + { + "epoch": 14.135782747603834, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8849 + }, + { + "epoch": 14.13738019169329, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8850 + }, + { + "epoch": 14.138977635782748, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8851 + }, + { + "epoch": 14.140575079872205, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8852 + }, + { + "epoch": 14.142172523961662, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8853 + }, + { + "epoch": 14.143769968051119, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8854 + }, + { + "epoch": 14.145367412140574, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8855 + }, + { + "epoch": 14.146964856230031, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8856 + }, + { + "epoch": 14.148562300319488, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8857 + }, + { + "epoch": 14.150159744408946, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8858 + }, + { + "epoch": 14.151757188498403, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8859 + }, + { + "epoch": 14.15335463258786, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8860 + }, + { + "epoch": 14.154952076677317, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8861 + }, + { + "epoch": 14.156549520766774, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8862 + }, + { + "epoch": 14.15814696485623, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8863 + }, + { + "epoch": 14.159744408945686, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 8864 + }, + { + "epoch": 14.161341853035143, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8865 + }, + { + "epoch": 14.1629392971246, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8866 + }, + { + "epoch": 14.164536741214057, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8867 + }, + { + "epoch": 14.166134185303514, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8868 + }, + { + "epoch": 14.167731629392971, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8869 + }, + { + "epoch": 14.169329073482428, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 8870 + }, + { + "epoch": 14.170926517571885, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8871 + }, + { + "epoch": 14.172523961661343, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8872 + }, + { + "epoch": 14.1741214057508, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8873 + }, + { + "epoch": 14.175718849840255, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8874 + }, + { + "epoch": 14.177316293929712, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8875 + }, + { + "epoch": 14.178913738019169, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8876 + }, + { + "epoch": 14.180511182108626, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8877 + }, + { + "epoch": 14.182108626198083, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8878 + }, + { + "epoch": 14.18370607028754, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8879 + }, + { + "epoch": 14.185303514376997, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 8880 + }, + { + "epoch": 14.186900958466454, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8881 + }, + { + "epoch": 14.188498402555911, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8882 + }, + { + "epoch": 14.190095846645367, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8883 + }, + { + "epoch": 14.191693290734824, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8884 + }, + { + "epoch": 14.19329073482428, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8885 + }, + { + "epoch": 14.194888178913738, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8886 + }, + { + "epoch": 14.196485623003195, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8887 + }, + { + "epoch": 14.198083067092652, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 8888 + }, + { + "epoch": 14.199680511182109, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 8889 + }, + { + "epoch": 14.201277955271566, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8890 + }, + { + "epoch": 14.202875399361023, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8891 + }, + { + "epoch": 14.204472843450478, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8892 + }, + { + "epoch": 14.206070287539935, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8893 + }, + { + "epoch": 14.207667731629392, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8894 + }, + { + "epoch": 14.20926517571885, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8895 + }, + { + "epoch": 14.210862619808307, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8896 + }, + { + "epoch": 14.212460063897764, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8897 + }, + { + "epoch": 14.21405750798722, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8898 + }, + { + "epoch": 14.215654952076678, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8899 + }, + { + "epoch": 14.217252396166135, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8900 + }, + { + "epoch": 14.218849840255592, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8901 + }, + { + "epoch": 14.220447284345047, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8902 + }, + { + "epoch": 14.222044728434504, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8903 + }, + { + "epoch": 14.223642172523961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8904 + }, + { + "epoch": 14.225239616613418, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8905 + }, + { + "epoch": 14.226837060702875, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8906 + }, + { + "epoch": 14.228434504792332, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8907 + }, + { + "epoch": 14.23003194888179, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8908 + }, + { + "epoch": 14.231629392971247, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8909 + }, + { + "epoch": 14.233226837060704, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8910 + }, + { + "epoch": 14.23482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8911 + }, + { + "epoch": 14.236421725239616, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8912 + }, + { + "epoch": 14.238019169329073, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8913 + }, + { + "epoch": 14.23961661341853, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 8914 + }, + { + "epoch": 14.241214057507987, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 8915 + }, + { + "epoch": 14.242811501597444, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8916 + }, + { + "epoch": 14.244408945686901, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8917 + }, + { + "epoch": 14.246006389776358, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8918 + }, + { + "epoch": 14.247603833865815, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8919 + }, + { + "epoch": 14.249201277955272, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8920 + }, + { + "epoch": 14.250798722044728, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8921 + }, + { + "epoch": 14.252396166134185, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8922 + }, + { + "epoch": 14.253993610223642, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8923 + }, + { + "epoch": 14.255591054313099, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8924 + }, + { + "epoch": 14.257188498402556, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8925 + }, + { + "epoch": 14.258785942492013, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8926 + }, + { + "epoch": 14.26038338658147, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 8927 + }, + { + "epoch": 14.261980830670927, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8928 + }, + { + "epoch": 14.263578274760384, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8929 + }, + { + "epoch": 14.26517571884984, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8930 + }, + { + "epoch": 14.266773162939296, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8931 + }, + { + "epoch": 14.268370607028753, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8932 + }, + { + "epoch": 14.26996805111821, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8933 + }, + { + "epoch": 14.271565495207668, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8934 + }, + { + "epoch": 14.273162939297125, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8935 + }, + { + "epoch": 14.274760383386582, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8936 + }, + { + "epoch": 14.276357827476039, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8937 + }, + { + "epoch": 14.277955271565496, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8938 + }, + { + "epoch": 14.279552715654953, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8939 + }, + { + "epoch": 14.281150159744408, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8940 + }, + { + "epoch": 14.282747603833865, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8941 + }, + { + "epoch": 14.284345047923322, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 8942 + }, + { + "epoch": 14.28594249201278, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 8943 + }, + { + "epoch": 14.287539936102236, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8944 + }, + { + "epoch": 14.289137380191693, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8945 + }, + { + "epoch": 14.29073482428115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8946 + }, + { + "epoch": 14.292332268370608, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8947 + }, + { + "epoch": 14.293929712460065, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8948 + }, + { + "epoch": 14.295527156549522, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 8949 + }, + { + "epoch": 14.297124600638977, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8950 + }, + { + "epoch": 14.298722044728434, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8951 + }, + { + "epoch": 14.300319488817891, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8952 + }, + { + "epoch": 14.301916932907348, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8953 + }, + { + "epoch": 14.303514376996805, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8954 + }, + { + "epoch": 14.305111821086262, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8955 + }, + { + "epoch": 14.30670926517572, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8956 + }, + { + "epoch": 14.308306709265176, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8957 + }, + { + "epoch": 14.309904153354633, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8958 + }, + { + "epoch": 14.311501597444089, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 8959 + }, + { + "epoch": 14.313099041533546, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8960 + }, + { + "epoch": 14.314696485623003, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8961 + }, + { + "epoch": 14.31629392971246, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 8962 + }, + { + "epoch": 14.317891373801917, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 8963 + }, + { + "epoch": 14.319488817891374, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8964 + }, + { + "epoch": 14.321086261980831, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8965 + }, + { + "epoch": 14.322683706070288, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8966 + }, + { + "epoch": 14.324281150159745, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 8967 + }, + { + "epoch": 14.3258785942492, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8968 + }, + { + "epoch": 14.327476038338657, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 8969 + }, + { + "epoch": 14.329073482428115, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8970 + }, + { + "epoch": 14.330670926517572, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8971 + }, + { + "epoch": 14.332268370607029, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8972 + }, + { + "epoch": 14.333865814696486, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8973 + }, + { + "epoch": 14.335463258785943, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 8974 + }, + { + "epoch": 14.3370607028754, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8975 + }, + { + "epoch": 14.338658146964857, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8976 + }, + { + "epoch": 14.340255591054314, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8977 + }, + { + "epoch": 14.34185303514377, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8978 + }, + { + "epoch": 14.343450479233226, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8979 + }, + { + "epoch": 14.345047923322683, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8980 + }, + { + "epoch": 14.34664536741214, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8981 + }, + { + "epoch": 14.348242811501597, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8982 + }, + { + "epoch": 14.349840255591054, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8983 + }, + { + "epoch": 14.351437699680512, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8984 + }, + { + "epoch": 14.353035143769969, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 8985 + }, + { + "epoch": 14.354632587859426, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8986 + }, + { + "epoch": 14.356230031948883, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8987 + }, + { + "epoch": 14.357827476038338, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8988 + }, + { + "epoch": 14.359424920127795, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8989 + }, + { + "epoch": 14.361022364217252, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8990 + }, + { + "epoch": 14.36261980830671, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 8991 + }, + { + "epoch": 14.364217252396166, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8992 + }, + { + "epoch": 14.365814696485623, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8993 + }, + { + "epoch": 14.36741214057508, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8994 + }, + { + "epoch": 14.369009584664537, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8995 + }, + { + "epoch": 14.370607028753994, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8996 + }, + { + "epoch": 14.37220447284345, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8997 + }, + { + "epoch": 14.373801916932907, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8998 + }, + { + "epoch": 14.375399361022364, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8999 + }, + { + "epoch": 14.37699680511182, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9000 + }, + { + "epoch": 14.378594249201278, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9001 + }, + { + "epoch": 14.380191693290735, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9002 + }, + { + "epoch": 14.381789137380192, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9003 + }, + { + "epoch": 14.383386581469649, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9004 + }, + { + "epoch": 14.384984025559106, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9005 + }, + { + "epoch": 14.386581469648561, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9006 + }, + { + "epoch": 14.388178913738018, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 9007 + }, + { + "epoch": 14.389776357827476, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9008 + }, + { + "epoch": 14.391373801916933, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9009 + }, + { + "epoch": 14.39297124600639, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 9010 + }, + { + "epoch": 14.394568690095847, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9011 + }, + { + "epoch": 14.396166134185304, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9012 + }, + { + "epoch": 14.39776357827476, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9013 + }, + { + "epoch": 14.399361022364218, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9014 + }, + { + "epoch": 14.400958466453675, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9015 + }, + { + "epoch": 14.40255591054313, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9016 + }, + { + "epoch": 14.404153354632587, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9017 + }, + { + "epoch": 14.405750798722044, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9018 + }, + { + "epoch": 14.407348242811501, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9019 + }, + { + "epoch": 14.408945686900958, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9020 + }, + { + "epoch": 14.410543130990416, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9021 + }, + { + "epoch": 14.412140575079873, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9022 + }, + { + "epoch": 14.41373801916933, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9023 + }, + { + "epoch": 14.415335463258787, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9024 + }, + { + "epoch": 14.416932907348242, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9025 + }, + { + "epoch": 14.418530351437699, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9026 + }, + { + "epoch": 14.420127795527156, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9027 + }, + { + "epoch": 14.421725239616613, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9028 + }, + { + "epoch": 14.42332268370607, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9029 + }, + { + "epoch": 14.424920127795527, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9030 + }, + { + "epoch": 14.426517571884984, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9031 + }, + { + "epoch": 14.428115015974441, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9032 + }, + { + "epoch": 14.429712460063898, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9033 + }, + { + "epoch": 14.431309904153355, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9034 + }, + { + "epoch": 14.43290734824281, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9035 + }, + { + "epoch": 14.434504792332268, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9036 + }, + { + "epoch": 14.436102236421725, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9037 + }, + { + "epoch": 14.437699680511182, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9038 + }, + { + "epoch": 14.439297124600639, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 9039 + }, + { + "epoch": 14.440894568690096, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9040 + }, + { + "epoch": 14.442492012779553, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9041 + }, + { + "epoch": 14.44408945686901, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9042 + }, + { + "epoch": 14.445686900958467, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9043 + }, + { + "epoch": 14.447284345047922, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 9044 + }, + { + "epoch": 14.44888178913738, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9045 + }, + { + "epoch": 14.450479233226837, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9046 + }, + { + "epoch": 14.452076677316294, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9047 + }, + { + "epoch": 14.45367412140575, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 9048 + }, + { + "epoch": 14.455271565495208, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9049 + }, + { + "epoch": 14.456869009584665, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9050 + }, + { + "epoch": 14.458466453674122, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9051 + }, + { + "epoch": 14.460063897763579, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9052 + }, + { + "epoch": 14.461661341853036, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9053 + }, + { + "epoch": 14.463258785942491, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9054 + }, + { + "epoch": 14.464856230031948, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9055 + }, + { + "epoch": 14.466453674121405, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9056 + }, + { + "epoch": 14.468051118210862, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9057 + }, + { + "epoch": 14.46964856230032, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9058 + }, + { + "epoch": 14.471246006389777, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9059 + }, + { + "epoch": 14.472843450479234, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9060 + }, + { + "epoch": 14.47444089456869, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9061 + }, + { + "epoch": 14.476038338658148, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9062 + }, + { + "epoch": 14.477635782747603, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9063 + }, + { + "epoch": 14.47923322683706, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9064 + }, + { + "epoch": 14.480830670926517, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9065 + }, + { + "epoch": 14.482428115015974, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9066 + }, + { + "epoch": 14.484025559105431, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9067 + }, + { + "epoch": 14.485623003194888, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9068 + }, + { + "epoch": 14.487220447284345, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9069 + }, + { + "epoch": 14.488817891373802, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9070 + }, + { + "epoch": 14.49041533546326, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 9071 + }, + { + "epoch": 14.492012779552716, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9072 + }, + { + "epoch": 14.493610223642172, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9073 + }, + { + "epoch": 14.495207667731629, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9074 + }, + { + "epoch": 14.496805111821086, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9075 + }, + { + "epoch": 14.498402555910543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 9076 + }, + { + "epoch": 14.5, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9077 + }, + { + "epoch": 14.501597444089457, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9078 + }, + { + "epoch": 14.503194888178914, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9079 + }, + { + "epoch": 14.504792332268371, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9080 + }, + { + "epoch": 14.506389776357828, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9081 + }, + { + "epoch": 14.507987220447284, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9082 + }, + { + "epoch": 14.50958466453674, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9083 + }, + { + "epoch": 14.511182108626198, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9084 + }, + { + "epoch": 14.512779552715655, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9085 + }, + { + "epoch": 14.514376996805112, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9086 + }, + { + "epoch": 14.515974440894569, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9087 + }, + { + "epoch": 14.517571884984026, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9088 + }, + { + "epoch": 14.519169329073483, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9089 + }, + { + "epoch": 14.52076677316294, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9090 + }, + { + "epoch": 14.522364217252395, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9091 + }, + { + "epoch": 14.523961661341852, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9092 + }, + { + "epoch": 14.52555910543131, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9093 + }, + { + "epoch": 14.527156549520766, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9094 + }, + { + "epoch": 14.528753993610223, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9095 + }, + { + "epoch": 14.53035143769968, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9096 + }, + { + "epoch": 14.531948881789138, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9097 + }, + { + "epoch": 14.533546325878595, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9098 + }, + { + "epoch": 14.535143769968052, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9099 + }, + { + "epoch": 14.536741214057509, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9100 + }, + { + "epoch": 14.538338658146966, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9101 + }, + { + "epoch": 14.539936102236421, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9102 + }, + { + "epoch": 14.541533546325878, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9103 + }, + { + "epoch": 14.543130990415335, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9104 + }, + { + "epoch": 14.544728434504792, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9105 + }, + { + "epoch": 14.54632587859425, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9106 + }, + { + "epoch": 14.547923322683706, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9107 + }, + { + "epoch": 14.549520766773163, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9108 + }, + { + "epoch": 14.55111821086262, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9109 + }, + { + "epoch": 14.552715654952078, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9110 + }, + { + "epoch": 14.554313099041533, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9111 + }, + { + "epoch": 14.55591054313099, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9112 + }, + { + "epoch": 14.557507987220447, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9113 + }, + { + "epoch": 14.559105431309904, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 9114 + }, + { + "epoch": 14.560702875399361, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9115 + }, + { + "epoch": 14.562300319488818, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9116 + }, + { + "epoch": 14.563897763578275, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9117 + }, + { + "epoch": 14.565495207667732, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9118 + }, + { + "epoch": 14.56709265175719, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9119 + }, + { + "epoch": 14.568690095846645, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9120 + }, + { + "epoch": 14.570287539936102, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9121 + }, + { + "epoch": 14.571884984025559, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 9122 + }, + { + "epoch": 14.573482428115016, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9123 + }, + { + "epoch": 14.575079872204473, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9124 + }, + { + "epoch": 14.57667731629393, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9125 + }, + { + "epoch": 14.578274760383387, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9126 + }, + { + "epoch": 14.579872204472844, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 9127 + }, + { + "epoch": 14.581469648562301, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9128 + }, + { + "epoch": 14.583067092651756, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 9129 + }, + { + "epoch": 14.584664536741213, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9130 + }, + { + "epoch": 14.58626198083067, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9131 + }, + { + "epoch": 14.587859424920127, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9132 + }, + { + "epoch": 14.589456869009584, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9133 + }, + { + "epoch": 14.591054313099042, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 9134 + }, + { + "epoch": 14.592651757188499, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9135 + }, + { + "epoch": 14.594249201277956, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9136 + }, + { + "epoch": 14.595846645367413, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9137 + }, + { + "epoch": 14.59744408945687, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9138 + }, + { + "epoch": 14.599041533546325, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9139 + }, + { + "epoch": 14.600638977635782, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9140 + }, + { + "epoch": 14.60223642172524, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 9141 + }, + { + "epoch": 14.603833865814696, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9142 + }, + { + "epoch": 14.605431309904153, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9143 + }, + { + "epoch": 14.60702875399361, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9144 + }, + { + "epoch": 14.608626198083067, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9145 + }, + { + "epoch": 14.610223642172524, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9146 + }, + { + "epoch": 14.611821086261982, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9147 + }, + { + "epoch": 14.613418530351439, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9148 + }, + { + "epoch": 14.615015974440894, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 9149 + }, + { + "epoch": 14.616613418530351, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9150 + }, + { + "epoch": 14.618210862619808, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9151 + }, + { + "epoch": 14.619808306709265, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9152 + }, + { + "epoch": 14.621405750798722, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9153 + }, + { + "epoch": 14.62300319488818, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9154 + }, + { + "epoch": 14.624600638977636, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9155 + }, + { + "epoch": 14.626198083067093, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9156 + }, + { + "epoch": 14.62779552715655, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9157 + }, + { + "epoch": 14.629392971246006, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9158 + }, + { + "epoch": 14.630990415335463, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9159 + }, + { + "epoch": 14.63258785942492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9160 + }, + { + "epoch": 14.634185303514377, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9161 + }, + { + "epoch": 14.635782747603834, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9162 + }, + { + "epoch": 14.63738019169329, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9163 + }, + { + "epoch": 14.638977635782748, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9164 + }, + { + "epoch": 14.640575079872205, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 9165 + }, + { + "epoch": 14.642172523961662, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9166 + }, + { + "epoch": 14.643769968051117, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9167 + }, + { + "epoch": 14.645367412140574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9168 + }, + { + "epoch": 14.646964856230031, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9169 + }, + { + "epoch": 14.648562300319488, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9170 + }, + { + "epoch": 14.650159744408946, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9171 + }, + { + "epoch": 14.651757188498403, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9172 + }, + { + "epoch": 14.65335463258786, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 9173 + }, + { + "epoch": 14.654952076677317, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9174 + }, + { + "epoch": 14.656549520766774, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9175 + }, + { + "epoch": 14.65814696485623, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9176 + }, + { + "epoch": 14.659744408945686, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9177 + }, + { + "epoch": 14.661341853035143, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9178 + }, + { + "epoch": 14.6629392971246, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9179 + }, + { + "epoch": 14.664536741214057, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9180 + }, + { + "epoch": 14.666134185303514, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9181 + }, + { + "epoch": 14.667731629392971, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9182 + }, + { + "epoch": 14.669329073482428, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9183 + }, + { + "epoch": 14.670926517571885, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9184 + }, + { + "epoch": 14.672523961661343, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9185 + }, + { + "epoch": 14.6741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9186 + }, + { + "epoch": 14.675718849840255, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9187 + }, + { + "epoch": 14.677316293929712, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9188 + }, + { + "epoch": 14.678913738019169, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9189 + }, + { + "epoch": 14.680511182108626, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9190 + }, + { + "epoch": 14.682108626198083, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9191 + }, + { + "epoch": 14.68370607028754, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9192 + }, + { + "epoch": 14.685303514376997, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9193 + }, + { + "epoch": 14.686900958466454, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 9194 + }, + { + "epoch": 14.688498402555911, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9195 + }, + { + "epoch": 14.690095846645367, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9196 + }, + { + "epoch": 14.691693290734824, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9197 + }, + { + "epoch": 14.69329073482428, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9198 + }, + { + "epoch": 14.694888178913738, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9199 + }, + { + "epoch": 14.696485623003195, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 9200 + }, + { + "epoch": 14.698083067092652, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9201 + }, + { + "epoch": 14.699680511182109, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9202 + }, + { + "epoch": 14.701277955271566, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9203 + }, + { + "epoch": 14.702875399361023, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9204 + }, + { + "epoch": 14.704472843450478, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 9205 + }, + { + "epoch": 14.706070287539935, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9206 + }, + { + "epoch": 14.707667731629392, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9207 + }, + { + "epoch": 14.70926517571885, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 9208 + }, + { + "epoch": 14.710862619808307, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9209 + }, + { + "epoch": 14.712460063897764, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9210 + }, + { + "epoch": 14.71405750798722, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9211 + }, + { + "epoch": 14.715654952076678, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9212 + }, + { + "epoch": 14.717252396166135, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9213 + }, + { + "epoch": 14.718849840255592, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 9214 + }, + { + "epoch": 14.720447284345047, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9215 + }, + { + "epoch": 14.722044728434504, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9216 + }, + { + "epoch": 14.723642172523961, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9217 + }, + { + "epoch": 14.725239616613418, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9218 + }, + { + "epoch": 14.726837060702875, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9219 + }, + { + "epoch": 14.728434504792332, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9220 + }, + { + "epoch": 14.73003194888179, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9221 + }, + { + "epoch": 14.731629392971247, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9222 + }, + { + "epoch": 14.733226837060704, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9223 + }, + { + "epoch": 14.73482428115016, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9224 + }, + { + "epoch": 14.736421725239616, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9225 + }, + { + "epoch": 14.738019169329073, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9226 + }, + { + "epoch": 14.73961661341853, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 9227 + }, + { + "epoch": 14.741214057507987, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 9228 + }, + { + "epoch": 14.742811501597444, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 9229 + }, + { + "epoch": 14.744408945686901, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9230 + }, + { + "epoch": 14.746006389776358, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9231 + }, + { + "epoch": 14.747603833865815, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9232 + }, + { + "epoch": 14.749201277955272, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9233 + }, + { + "epoch": 14.750798722044728, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9234 + }, + { + "epoch": 14.752396166134185, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9235 + }, + { + "epoch": 14.753993610223642, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9236 + }, + { + "epoch": 14.755591054313099, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9237 + }, + { + "epoch": 14.757188498402556, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 9238 + }, + { + "epoch": 14.758785942492013, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9239 + }, + { + "epoch": 14.76038338658147, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 9240 + }, + { + "epoch": 14.761980830670927, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9241 + }, + { + "epoch": 14.763578274760384, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9242 + }, + { + "epoch": 14.76517571884984, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9243 + }, + { + "epoch": 14.766773162939296, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9244 + }, + { + "epoch": 14.768370607028753, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9245 + }, + { + "epoch": 14.76996805111821, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9246 + }, + { + "epoch": 14.771565495207668, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 9247 + }, + { + "epoch": 14.773162939297125, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9248 + }, + { + "epoch": 14.774760383386582, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9249 + }, + { + "epoch": 14.776357827476039, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9250 + }, + { + "epoch": 14.777955271565496, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9251 + }, + { + "epoch": 14.779552715654953, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9252 + }, + { + "epoch": 14.781150159744408, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9253 + }, + { + "epoch": 14.782747603833865, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9254 + }, + { + "epoch": 14.784345047923322, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9255 + }, + { + "epoch": 14.78594249201278, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9256 + }, + { + "epoch": 14.787539936102236, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9257 + }, + { + "epoch": 14.789137380191693, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9258 + }, + { + "epoch": 14.79073482428115, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9259 + }, + { + "epoch": 14.792332268370608, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9260 + }, + { + "epoch": 14.793929712460065, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9261 + }, + { + "epoch": 14.795527156549522, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9262 + }, + { + "epoch": 14.797124600638977, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9263 + }, + { + "epoch": 14.798722044728434, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 9264 + }, + { + "epoch": 14.800319488817891, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9265 + }, + { + "epoch": 14.801916932907348, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9266 + }, + { + "epoch": 14.803514376996805, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9267 + }, + { + "epoch": 14.805111821086262, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9268 + }, + { + "epoch": 14.80670926517572, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9269 + }, + { + "epoch": 14.808306709265176, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9270 + }, + { + "epoch": 14.809904153354633, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9271 + }, + { + "epoch": 14.811501597444089, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9272 + }, + { + "epoch": 14.813099041533546, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9273 + }, + { + "epoch": 14.814696485623003, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9274 + }, + { + "epoch": 14.81629392971246, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9275 + }, + { + "epoch": 14.817891373801917, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9276 + }, + { + "epoch": 14.819488817891374, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9277 + }, + { + "epoch": 14.821086261980831, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9278 + }, + { + "epoch": 14.822683706070288, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9279 + }, + { + "epoch": 14.824281150159745, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 9280 + }, + { + "epoch": 14.8258785942492, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9281 + }, + { + "epoch": 14.827476038338657, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9282 + }, + { + "epoch": 14.829073482428115, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9283 + }, + { + "epoch": 14.830670926517572, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9284 + }, + { + "epoch": 14.832268370607029, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9285 + }, + { + "epoch": 14.833865814696486, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9286 + }, + { + "epoch": 14.835463258785943, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9287 + }, + { + "epoch": 14.8370607028754, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 9288 + }, + { + "epoch": 14.838658146964857, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9289 + }, + { + "epoch": 14.840255591054314, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9290 + }, + { + "epoch": 14.84185303514377, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9291 + }, + { + "epoch": 14.843450479233226, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9292 + }, + { + "epoch": 14.845047923322683, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9293 + }, + { + "epoch": 14.84664536741214, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9294 + }, + { + "epoch": 14.848242811501597, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9295 + }, + { + "epoch": 14.849840255591054, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9296 + }, + { + "epoch": 14.851437699680512, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9297 + }, + { + "epoch": 14.853035143769969, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9298 + }, + { + "epoch": 14.854632587859426, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9299 + }, + { + "epoch": 14.856230031948883, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9300 + }, + { + "epoch": 14.857827476038338, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9301 + }, + { + "epoch": 14.859424920127795, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9302 + }, + { + "epoch": 14.861022364217252, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9303 + }, + { + "epoch": 14.86261980830671, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9304 + }, + { + "epoch": 14.864217252396166, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9305 + }, + { + "epoch": 14.865814696485623, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9306 + }, + { + "epoch": 14.86741214057508, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9307 + }, + { + "epoch": 14.869009584664537, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9308 + }, + { + "epoch": 14.870607028753994, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9309 + }, + { + "epoch": 14.87220447284345, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9310 + }, + { + "epoch": 14.873801916932907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9311 + }, + { + "epoch": 14.875399361022364, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9312 + }, + { + "epoch": 14.87699680511182, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9313 + }, + { + "epoch": 14.878594249201278, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9314 + }, + { + "epoch": 14.880191693290735, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9315 + }, + { + "epoch": 14.881789137380192, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9316 + }, + { + "epoch": 14.883386581469649, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9317 + }, + { + "epoch": 14.884984025559106, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9318 + }, + { + "epoch": 14.886581469648561, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9319 + }, + { + "epoch": 14.888178913738018, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9320 + }, + { + "epoch": 14.889776357827476, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9321 + }, + { + "epoch": 14.891373801916933, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 9322 + }, + { + "epoch": 14.89297124600639, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9323 + }, + { + "epoch": 14.894568690095847, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9324 + }, + { + "epoch": 14.896166134185304, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9325 + }, + { + "epoch": 14.89776357827476, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 9326 + }, + { + "epoch": 14.899361022364218, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9327 + }, + { + "epoch": 14.900958466453675, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9328 + }, + { + "epoch": 14.90255591054313, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 9329 + }, + { + "epoch": 14.904153354632587, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9330 + }, + { + "epoch": 14.905750798722044, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9331 + }, + { + "epoch": 14.907348242811501, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9332 + }, + { + "epoch": 14.908945686900958, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9333 + }, + { + "epoch": 14.910543130990416, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9334 + }, + { + "epoch": 14.912140575079873, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9335 + }, + { + "epoch": 14.91373801916933, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9336 + }, + { + "epoch": 14.915335463258787, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9337 + }, + { + "epoch": 14.916932907348244, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9338 + }, + { + "epoch": 14.918530351437699, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9339 + }, + { + "epoch": 14.920127795527156, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9340 + }, + { + "epoch": 14.921725239616613, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9341 + }, + { + "epoch": 14.92332268370607, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9342 + }, + { + "epoch": 14.924920127795527, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9343 + }, + { + "epoch": 14.926517571884984, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9344 + }, + { + "epoch": 14.928115015974441, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9345 + }, + { + "epoch": 14.929712460063898, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9346 + }, + { + "epoch": 14.931309904153355, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9347 + }, + { + "epoch": 14.93290734824281, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 9348 + }, + { + "epoch": 14.934504792332268, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9349 + }, + { + "epoch": 14.936102236421725, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9350 + }, + { + "epoch": 14.937699680511182, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9351 + }, + { + "epoch": 14.939297124600639, + "grad_norm": 0.031982421875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9352 + }, + { + "epoch": 14.940894568690096, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9353 + }, + { + "epoch": 14.942492012779553, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 9354 + }, + { + "epoch": 14.94408945686901, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9355 + }, + { + "epoch": 14.945686900958467, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9356 + }, + { + "epoch": 14.947284345047922, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9357 + }, + { + "epoch": 14.94888178913738, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9358 + }, + { + "epoch": 14.950479233226837, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9359 + }, + { + "epoch": 14.952076677316294, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9360 + }, + { + "epoch": 14.95367412140575, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9361 + }, + { + "epoch": 14.955271565495208, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9362 + }, + { + "epoch": 14.956869009584665, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9363 + }, + { + "epoch": 14.958466453674122, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 9364 + }, + { + "epoch": 14.960063897763579, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9365 + }, + { + "epoch": 14.961661341853034, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9366 + }, + { + "epoch": 14.963258785942491, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9367 + }, + { + "epoch": 14.964856230031948, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9368 + }, + { + "epoch": 14.966453674121405, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9369 + }, + { + "epoch": 14.968051118210862, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9370 + }, + { + "epoch": 14.96964856230032, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9371 + }, + { + "epoch": 14.971246006389777, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9372 + }, + { + "epoch": 14.972843450479234, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9373 + }, + { + "epoch": 14.97444089456869, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9374 + }, + { + "epoch": 14.976038338658148, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9375 + }, + { + "epoch": 14.977635782747605, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9376 + }, + { + "epoch": 14.97923322683706, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9377 + }, + { + "epoch": 14.980830670926517, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9378 + }, + { + "epoch": 14.982428115015974, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9379 + }, + { + "epoch": 14.984025559105431, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9380 + }, + { + "epoch": 14.985623003194888, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9381 + }, + { + "epoch": 14.987220447284345, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9382 + }, + { + "epoch": 14.988817891373802, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9383 + }, + { + "epoch": 14.99041533546326, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9384 + }, + { + "epoch": 14.992012779552716, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9385 + }, + { + "epoch": 14.993610223642172, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9386 + }, + { + "epoch": 14.995207667731629, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9387 + }, + { + "epoch": 14.996805111821086, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9388 + }, + { + "epoch": 14.998402555910543, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9389 + }, + { + "epoch": 15.0, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9390 + }, + { + "epoch": 15.001597444089457, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 9391 + }, + { + "epoch": 15.003194888178914, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9392 + }, + { + "epoch": 15.004792332268371, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9393 + }, + { + "epoch": 15.006389776357828, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9394 + }, + { + "epoch": 15.007987220447284, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9395 + }, + { + "epoch": 15.00958466453674, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9396 + }, + { + "epoch": 15.011182108626198, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9397 + }, + { + "epoch": 15.012779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9398 + }, + { + "epoch": 15.014376996805112, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9399 + }, + { + "epoch": 15.015974440894569, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9400 + }, + { + "epoch": 15.017571884984026, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9401 + }, + { + "epoch": 15.019169329073483, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 9402 + }, + { + "epoch": 15.02076677316294, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9403 + }, + { + "epoch": 15.022364217252397, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9404 + }, + { + "epoch": 15.023961661341852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9405 + }, + { + "epoch": 15.02555910543131, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 9406 + }, + { + "epoch": 15.027156549520766, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9407 + }, + { + "epoch": 15.028753993610223, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9408 + }, + { + "epoch": 15.03035143769968, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9409 + }, + { + "epoch": 15.031948881789138, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9410 + }, + { + "epoch": 15.033546325878595, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9411 + }, + { + "epoch": 15.035143769968052, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9412 + }, + { + "epoch": 15.036741214057509, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9413 + }, + { + "epoch": 15.038338658146964, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9414 + }, + { + "epoch": 15.039936102236421, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 9415 + }, + { + "epoch": 15.041533546325878, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9416 + }, + { + "epoch": 15.043130990415335, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9417 + }, + { + "epoch": 15.044728434504792, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9418 + }, + { + "epoch": 15.04632587859425, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9419 + }, + { + "epoch": 15.047923322683706, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9420 + }, + { + "epoch": 15.049520766773163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9421 + }, + { + "epoch": 15.05111821086262, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9422 + }, + { + "epoch": 15.052715654952078, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9423 + }, + { + "epoch": 15.054313099041533, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9424 + }, + { + "epoch": 15.05591054313099, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9425 + }, + { + "epoch": 15.057507987220447, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9426 + }, + { + "epoch": 15.059105431309904, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9427 + }, + { + "epoch": 15.060702875399361, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9428 + }, + { + "epoch": 15.062300319488818, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 9429 + }, + { + "epoch": 15.063897763578275, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9430 + }, + { + "epoch": 15.065495207667732, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 9431 + }, + { + "epoch": 15.06709265175719, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9432 + }, + { + "epoch": 15.068690095846645, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9433 + }, + { + "epoch": 15.070287539936102, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9434 + }, + { + "epoch": 15.071884984025559, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9435 + }, + { + "epoch": 15.073482428115016, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9436 + }, + { + "epoch": 15.075079872204473, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9437 + }, + { + "epoch": 15.07667731629393, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9438 + }, + { + "epoch": 15.078274760383387, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 9439 + }, + { + "epoch": 15.079872204472844, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9440 + }, + { + "epoch": 15.081469648562301, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9441 + }, + { + "epoch": 15.083067092651758, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9442 + }, + { + "epoch": 15.084664536741213, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9443 + }, + { + "epoch": 15.08626198083067, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9444 + }, + { + "epoch": 15.087859424920127, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9445 + }, + { + "epoch": 15.089456869009584, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9446 + }, + { + "epoch": 15.091054313099042, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 9447 + }, + { + "epoch": 15.092651757188499, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9448 + }, + { + "epoch": 15.094249201277956, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9449 + }, + { + "epoch": 15.095846645367413, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9450 + }, + { + "epoch": 15.09744408945687, + "grad_norm": 0.031982421875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9451 + }, + { + "epoch": 15.099041533546325, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9452 + }, + { + "epoch": 15.100638977635782, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9453 + }, + { + "epoch": 15.10223642172524, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9454 + }, + { + "epoch": 15.103833865814696, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9455 + }, + { + "epoch": 15.105431309904153, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9456 + }, + { + "epoch": 15.10702875399361, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9457 + }, + { + "epoch": 15.108626198083067, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9458 + }, + { + "epoch": 15.110223642172524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9459 + }, + { + "epoch": 15.111821086261982, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9460 + }, + { + "epoch": 15.113418530351439, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9461 + }, + { + "epoch": 15.115015974440894, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9462 + }, + { + "epoch": 15.116613418530351, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9463 + }, + { + "epoch": 15.118210862619808, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9464 + }, + { + "epoch": 15.119808306709265, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9465 + }, + { + "epoch": 15.121405750798722, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9466 + }, + { + "epoch": 15.12300319488818, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9467 + }, + { + "epoch": 15.124600638977636, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9468 + }, + { + "epoch": 15.126198083067093, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9469 + }, + { + "epoch": 15.12779552715655, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 9470 + }, + { + "epoch": 15.129392971246006, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9471 + }, + { + "epoch": 15.130990415335463, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9472 + }, + { + "epoch": 15.13258785942492, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9473 + }, + { + "epoch": 15.134185303514377, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9474 + }, + { + "epoch": 15.135782747603834, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9475 + }, + { + "epoch": 15.13738019169329, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9476 + }, + { + "epoch": 15.138977635782748, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9477 + }, + { + "epoch": 15.140575079872205, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9478 + }, + { + "epoch": 15.142172523961662, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9479 + }, + { + "epoch": 15.143769968051119, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9480 + }, + { + "epoch": 15.145367412140574, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9481 + }, + { + "epoch": 15.146964856230031, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9482 + }, + { + "epoch": 15.148562300319488, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9483 + }, + { + "epoch": 15.150159744408946, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9484 + }, + { + "epoch": 15.151757188498403, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9485 + }, + { + "epoch": 15.15335463258786, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9486 + }, + { + "epoch": 15.154952076677317, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9487 + }, + { + "epoch": 15.156549520766774, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9488 + }, + { + "epoch": 15.15814696485623, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9489 + }, + { + "epoch": 15.159744408945686, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 9490 + }, + { + "epoch": 15.161341853035143, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9491 + }, + { + "epoch": 15.1629392971246, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9492 + }, + { + "epoch": 15.164536741214057, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9493 + }, + { + "epoch": 15.166134185303514, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9494 + }, + { + "epoch": 15.167731629392971, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9495 + }, + { + "epoch": 15.169329073482428, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9496 + }, + { + "epoch": 15.170926517571885, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9497 + }, + { + "epoch": 15.172523961661343, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9498 + }, + { + "epoch": 15.1741214057508, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9499 + }, + { + "epoch": 15.175718849840255, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9500 + }, + { + "epoch": 15.177316293929712, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9501 + }, + { + "epoch": 15.178913738019169, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9502 + }, + { + "epoch": 15.180511182108626, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9503 + }, + { + "epoch": 15.182108626198083, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9504 + }, + { + "epoch": 15.18370607028754, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9505 + }, + { + "epoch": 15.185303514376997, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9506 + }, + { + "epoch": 15.186900958466454, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9507 + }, + { + "epoch": 15.188498402555911, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9508 + }, + { + "epoch": 15.190095846645367, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9509 + }, + { + "epoch": 15.191693290734824, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9510 + }, + { + "epoch": 15.19329073482428, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9511 + }, + { + "epoch": 15.194888178913738, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9512 + }, + { + "epoch": 15.196485623003195, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9513 + }, + { + "epoch": 15.198083067092652, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9514 + }, + { + "epoch": 15.199680511182109, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 9515 + }, + { + "epoch": 15.201277955271566, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9516 + }, + { + "epoch": 15.202875399361023, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9517 + }, + { + "epoch": 15.204472843450478, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9518 + }, + { + "epoch": 15.206070287539935, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9519 + }, + { + "epoch": 15.207667731629392, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9520 + }, + { + "epoch": 15.20926517571885, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 9521 + }, + { + "epoch": 15.210862619808307, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9522 + }, + { + "epoch": 15.212460063897764, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9523 + }, + { + "epoch": 15.21405750798722, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9524 + }, + { + "epoch": 15.215654952076678, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9525 + }, + { + "epoch": 15.217252396166135, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9526 + }, + { + "epoch": 15.218849840255592, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9527 + }, + { + "epoch": 15.220447284345047, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9528 + }, + { + "epoch": 15.222044728434504, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9529 + }, + { + "epoch": 15.223642172523961, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9530 + }, + { + "epoch": 15.225239616613418, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 9531 + }, + { + "epoch": 15.226837060702875, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9532 + }, + { + "epoch": 15.228434504792332, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9533 + }, + { + "epoch": 15.23003194888179, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9534 + }, + { + "epoch": 15.231629392971247, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9535 + }, + { + "epoch": 15.233226837060704, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9536 + }, + { + "epoch": 15.23482428115016, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9537 + }, + { + "epoch": 15.236421725239616, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9538 + }, + { + "epoch": 15.238019169329073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9539 + }, + { + "epoch": 15.23961661341853, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9540 + }, + { + "epoch": 15.241214057507987, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9541 + }, + { + "epoch": 15.242811501597444, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9542 + }, + { + "epoch": 15.244408945686901, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9543 + }, + { + "epoch": 15.246006389776358, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9544 + }, + { + "epoch": 15.247603833865815, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 9545 + }, + { + "epoch": 15.249201277955272, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9546 + }, + { + "epoch": 15.250798722044728, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9547 + }, + { + "epoch": 15.252396166134185, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9548 + }, + { + "epoch": 15.253993610223642, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9549 + }, + { + "epoch": 15.255591054313099, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9550 + }, + { + "epoch": 15.257188498402556, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9551 + }, + { + "epoch": 15.258785942492013, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 9552 + }, + { + "epoch": 15.26038338658147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9553 + }, + { + "epoch": 15.261980830670927, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9554 + }, + { + "epoch": 15.263578274760384, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9555 + }, + { + "epoch": 15.26517571884984, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 9556 + }, + { + "epoch": 15.266773162939296, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9557 + }, + { + "epoch": 15.268370607028753, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9558 + }, + { + "epoch": 15.26996805111821, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9559 + }, + { + "epoch": 15.271565495207668, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9560 + }, + { + "epoch": 15.273162939297125, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9561 + }, + { + "epoch": 15.274760383386582, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9562 + }, + { + "epoch": 15.276357827476039, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9563 + }, + { + "epoch": 15.277955271565496, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9564 + }, + { + "epoch": 15.279552715654953, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9565 + }, + { + "epoch": 15.281150159744408, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9566 + }, + { + "epoch": 15.282747603833865, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9567 + }, + { + "epoch": 15.284345047923322, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9568 + }, + { + "epoch": 15.28594249201278, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9569 + }, + { + "epoch": 15.287539936102236, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9570 + }, + { + "epoch": 15.289137380191693, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9571 + }, + { + "epoch": 15.29073482428115, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 9572 + }, + { + "epoch": 15.292332268370608, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9573 + }, + { + "epoch": 15.293929712460065, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9574 + }, + { + "epoch": 15.295527156549522, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9575 + }, + { + "epoch": 15.297124600638977, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9576 + }, + { + "epoch": 15.298722044728434, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9577 + }, + { + "epoch": 15.300319488817891, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 9578 + }, + { + "epoch": 15.301916932907348, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9579 + }, + { + "epoch": 15.303514376996805, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9580 + }, + { + "epoch": 15.305111821086262, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9581 + }, + { + "epoch": 15.30670926517572, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9582 + }, + { + "epoch": 15.308306709265176, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9583 + }, + { + "epoch": 15.309904153354633, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9584 + }, + { + "epoch": 15.311501597444089, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 9585 + }, + { + "epoch": 15.313099041533546, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9586 + }, + { + "epoch": 15.314696485623003, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 9587 + }, + { + "epoch": 15.31629392971246, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9588 + }, + { + "epoch": 15.317891373801917, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9589 + }, + { + "epoch": 15.319488817891374, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9590 + }, + { + "epoch": 15.321086261980831, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9591 + }, + { + "epoch": 15.322683706070288, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9592 + }, + { + "epoch": 15.324281150159745, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9593 + }, + { + "epoch": 15.3258785942492, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9594 + }, + { + "epoch": 15.327476038338657, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9595 + }, + { + "epoch": 15.329073482428115, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9596 + }, + { + "epoch": 15.330670926517572, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9597 + }, + { + "epoch": 15.332268370607029, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9598 + }, + { + "epoch": 15.333865814696486, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 9599 + }, + { + "epoch": 15.335463258785943, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9600 + }, + { + "epoch": 15.3370607028754, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9601 + }, + { + "epoch": 15.338658146964857, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9602 + }, + { + "epoch": 15.340255591054314, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9603 + }, + { + "epoch": 15.34185303514377, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 9604 + }, + { + "epoch": 15.343450479233226, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9605 + }, + { + "epoch": 15.345047923322683, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9606 + }, + { + "epoch": 15.34664536741214, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9607 + }, + { + "epoch": 15.348242811501597, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9608 + }, + { + "epoch": 15.349840255591054, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 9609 + }, + { + "epoch": 15.351437699680512, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9610 + }, + { + "epoch": 15.353035143769969, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9611 + }, + { + "epoch": 15.354632587859426, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9612 + }, + { + "epoch": 15.356230031948883, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9613 + }, + { + "epoch": 15.357827476038338, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9614 + }, + { + "epoch": 15.359424920127795, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9615 + }, + { + "epoch": 15.361022364217252, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9616 + }, + { + "epoch": 15.36261980830671, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9617 + }, + { + "epoch": 15.364217252396166, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9618 + }, + { + "epoch": 15.365814696485623, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9619 + }, + { + "epoch": 15.36741214057508, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9620 + }, + { + "epoch": 15.369009584664537, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9621 + }, + { + "epoch": 15.370607028753994, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9622 + }, + { + "epoch": 15.37220447284345, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9623 + }, + { + "epoch": 15.373801916932907, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9624 + }, + { + "epoch": 15.375399361022364, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9625 + }, + { + "epoch": 15.37699680511182, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9626 + }, + { + "epoch": 15.378594249201278, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9627 + }, + { + "epoch": 15.380191693290735, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9628 + }, + { + "epoch": 15.381789137380192, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9629 + }, + { + "epoch": 15.383386581469649, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9630 + }, + { + "epoch": 15.384984025559106, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 9631 + }, + { + "epoch": 15.386581469648561, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9632 + }, + { + "epoch": 15.388178913738018, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9633 + }, + { + "epoch": 15.389776357827476, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9634 + }, + { + "epoch": 15.391373801916933, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9635 + }, + { + "epoch": 15.39297124600639, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9636 + }, + { + "epoch": 15.394568690095847, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9637 + }, + { + "epoch": 15.396166134185304, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9638 + }, + { + "epoch": 15.39776357827476, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9639 + }, + { + "epoch": 15.399361022364218, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9640 + }, + { + "epoch": 15.400958466453675, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9641 + }, + { + "epoch": 15.40255591054313, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9642 + }, + { + "epoch": 15.404153354632587, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9643 + }, + { + "epoch": 15.405750798722044, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9644 + }, + { + "epoch": 15.407348242811501, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9645 + }, + { + "epoch": 15.408945686900958, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9646 + }, + { + "epoch": 15.410543130990416, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9647 + }, + { + "epoch": 15.412140575079873, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 9648 + }, + { + "epoch": 15.41373801916933, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9649 + }, + { + "epoch": 15.415335463258787, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9650 + }, + { + "epoch": 15.416932907348242, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9651 + }, + { + "epoch": 15.418530351437699, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9652 + }, + { + "epoch": 15.420127795527156, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9653 + }, + { + "epoch": 15.421725239616613, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9654 + }, + { + "epoch": 15.42332268370607, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9655 + }, + { + "epoch": 15.424920127795527, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9656 + }, + { + "epoch": 15.426517571884984, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9657 + }, + { + "epoch": 15.428115015974441, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9658 + }, + { + "epoch": 15.429712460063898, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9659 + }, + { + "epoch": 15.431309904153355, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9660 + }, + { + "epoch": 15.43290734824281, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9661 + }, + { + "epoch": 15.434504792332268, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9662 + }, + { + "epoch": 15.436102236421725, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9663 + }, + { + "epoch": 15.437699680511182, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9664 + }, + { + "epoch": 15.439297124600639, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9665 + }, + { + "epoch": 15.440894568690096, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9666 + }, + { + "epoch": 15.442492012779553, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9667 + }, + { + "epoch": 15.44408945686901, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9668 + }, + { + "epoch": 15.445686900958467, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9669 + }, + { + "epoch": 15.447284345047922, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9670 + }, + { + "epoch": 15.44888178913738, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 9671 + }, + { + "epoch": 15.450479233226837, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9672 + }, + { + "epoch": 15.452076677316294, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9673 + }, + { + "epoch": 15.45367412140575, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9674 + }, + { + "epoch": 15.455271565495208, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9675 + }, + { + "epoch": 15.456869009584665, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9676 + }, + { + "epoch": 15.458466453674122, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9677 + }, + { + "epoch": 15.460063897763579, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9678 + }, + { + "epoch": 15.461661341853036, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9679 + }, + { + "epoch": 15.463258785942491, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9680 + }, + { + "epoch": 15.464856230031948, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9681 + }, + { + "epoch": 15.466453674121405, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9682 + }, + { + "epoch": 15.468051118210862, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9683 + }, + { + "epoch": 15.46964856230032, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9684 + }, + { + "epoch": 15.471246006389777, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9685 + }, + { + "epoch": 15.472843450479234, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9686 + }, + { + "epoch": 15.47444089456869, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9687 + }, + { + "epoch": 15.476038338658148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9688 + }, + { + "epoch": 15.477635782747603, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9689 + }, + { + "epoch": 15.47923322683706, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9690 + }, + { + "epoch": 15.480830670926517, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9691 + }, + { + "epoch": 15.482428115015974, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9692 + }, + { + "epoch": 15.484025559105431, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9693 + }, + { + "epoch": 15.485623003194888, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 9694 + }, + { + "epoch": 15.487220447284345, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9695 + }, + { + "epoch": 15.488817891373802, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9696 + }, + { + "epoch": 15.49041533546326, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9697 + }, + { + "epoch": 15.492012779552716, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9698 + }, + { + "epoch": 15.493610223642172, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9699 + }, + { + "epoch": 15.495207667731629, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9700 + }, + { + "epoch": 15.496805111821086, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9701 + }, + { + "epoch": 15.498402555910543, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9702 + }, + { + "epoch": 15.5, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9703 + }, + { + "epoch": 15.501597444089457, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9704 + }, + { + "epoch": 15.503194888178914, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9705 + }, + { + "epoch": 15.504792332268371, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 9706 + }, + { + "epoch": 15.506389776357828, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9707 + }, + { + "epoch": 15.507987220447284, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 9708 + }, + { + "epoch": 15.50958466453674, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9709 + }, + { + "epoch": 15.511182108626198, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 9710 + }, + { + "epoch": 15.512779552715655, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9711 + }, + { + "epoch": 15.514376996805112, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9712 + }, + { + "epoch": 15.515974440894569, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9713 + }, + { + "epoch": 15.517571884984026, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9714 + }, + { + "epoch": 15.519169329073483, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9715 + }, + { + "epoch": 15.52076677316294, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9716 + }, + { + "epoch": 15.522364217252395, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9717 + }, + { + "epoch": 15.523961661341852, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9718 + }, + { + "epoch": 15.52555910543131, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 9719 + }, + { + "epoch": 15.527156549520766, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9720 + }, + { + "epoch": 15.528753993610223, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9721 + }, + { + "epoch": 15.53035143769968, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9722 + }, + { + "epoch": 15.531948881789138, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9723 + }, + { + "epoch": 15.533546325878595, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9724 + }, + { + "epoch": 15.535143769968052, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9725 + }, + { + "epoch": 15.536741214057509, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9726 + }, + { + "epoch": 15.538338658146966, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9727 + }, + { + "epoch": 15.539936102236421, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9728 + }, + { + "epoch": 15.541533546325878, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9729 + }, + { + "epoch": 15.543130990415335, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9730 + }, + { + "epoch": 15.544728434504792, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9731 + }, + { + "epoch": 15.54632587859425, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 9732 + }, + { + "epoch": 15.547923322683706, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9733 + }, + { + "epoch": 15.549520766773163, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9734 + }, + { + "epoch": 15.55111821086262, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 9735 + }, + { + "epoch": 15.552715654952078, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9736 + }, + { + "epoch": 15.554313099041533, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 9737 + }, + { + "epoch": 15.55591054313099, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9738 + }, + { + "epoch": 15.557507987220447, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9739 + }, + { + "epoch": 15.559105431309904, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9740 + }, + { + "epoch": 15.560702875399361, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9741 + }, + { + "epoch": 15.562300319488818, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9742 + }, + { + "epoch": 15.563897763578275, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9743 + }, + { + "epoch": 15.565495207667732, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9744 + }, + { + "epoch": 15.56709265175719, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 9745 + }, + { + "epoch": 15.568690095846645, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 9746 + }, + { + "epoch": 15.570287539936102, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9747 + }, + { + "epoch": 15.571884984025559, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9748 + }, + { + "epoch": 15.573482428115016, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9749 + }, + { + "epoch": 15.575079872204473, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9750 + }, + { + "epoch": 15.57667731629393, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9751 + }, + { + "epoch": 15.578274760383387, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9752 + }, + { + "epoch": 15.579872204472844, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9753 + }, + { + "epoch": 15.581469648562301, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9754 + }, + { + "epoch": 15.583067092651756, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9755 + }, + { + "epoch": 15.584664536741213, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9756 + }, + { + "epoch": 15.58626198083067, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9757 + }, + { + "epoch": 15.587859424920127, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 9758 + }, + { + "epoch": 15.589456869009584, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9759 + }, + { + "epoch": 15.591054313099042, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9760 + }, + { + "epoch": 15.592651757188499, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9761 + }, + { + "epoch": 15.594249201277956, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9762 + }, + { + "epoch": 15.595846645367413, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9763 + }, + { + "epoch": 15.59744408945687, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9764 + }, + { + "epoch": 15.599041533546325, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9765 + }, + { + "epoch": 15.600638977635782, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9766 + }, + { + "epoch": 15.60223642172524, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9767 + }, + { + "epoch": 15.603833865814696, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9768 + }, + { + "epoch": 15.605431309904153, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9769 + }, + { + "epoch": 15.60702875399361, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9770 + }, + { + "epoch": 15.608626198083067, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9771 + }, + { + "epoch": 15.610223642172524, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9772 + }, + { + "epoch": 15.611821086261982, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9773 + }, + { + "epoch": 15.613418530351439, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9774 + }, + { + "epoch": 15.615015974440894, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9775 + }, + { + "epoch": 15.616613418530351, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 9776 + }, + { + "epoch": 15.618210862619808, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9777 + }, + { + "epoch": 15.619808306709265, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9778 + }, + { + "epoch": 15.621405750798722, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9779 + }, + { + "epoch": 15.62300319488818, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9780 + }, + { + "epoch": 15.624600638977636, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 9781 + }, + { + "epoch": 15.626198083067093, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 9782 + }, + { + "epoch": 15.62779552715655, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9783 + }, + { + "epoch": 15.629392971246006, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9784 + }, + { + "epoch": 15.630990415335463, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 9785 + }, + { + "epoch": 15.63258785942492, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9786 + }, + { + "epoch": 15.634185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9787 + }, + { + "epoch": 15.635782747603834, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9788 + }, + { + "epoch": 15.63738019169329, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9789 + }, + { + "epoch": 15.638977635782748, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9790 + }, + { + "epoch": 15.640575079872205, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9791 + }, + { + "epoch": 15.642172523961662, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9792 + }, + { + "epoch": 15.643769968051117, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9793 + }, + { + "epoch": 15.645367412140574, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9794 + }, + { + "epoch": 15.646964856230031, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 9795 + }, + { + "epoch": 15.648562300319488, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9796 + }, + { + "epoch": 15.650159744408946, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9797 + }, + { + "epoch": 15.651757188498403, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 9798 + }, + { + "epoch": 15.65335463258786, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9799 + }, + { + "epoch": 15.654952076677317, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9800 + }, + { + "epoch": 15.656549520766774, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9801 + }, + { + "epoch": 15.65814696485623, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9802 + }, + { + "epoch": 15.659744408945686, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 9803 + }, + { + "epoch": 15.661341853035143, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9804 + }, + { + "epoch": 15.6629392971246, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9805 + }, + { + "epoch": 15.664536741214057, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9806 + }, + { + "epoch": 15.666134185303514, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9807 + }, + { + "epoch": 15.667731629392971, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 9808 + }, + { + "epoch": 15.669329073482428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9809 + }, + { + "epoch": 15.670926517571885, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9810 + }, + { + "epoch": 15.672523961661343, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9811 + }, + { + "epoch": 15.6741214057508, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9812 + }, + { + "epoch": 15.675718849840255, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9813 + }, + { + "epoch": 15.677316293929712, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9814 + }, + { + "epoch": 15.678913738019169, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9815 + }, + { + "epoch": 15.680511182108626, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9816 + }, + { + "epoch": 15.682108626198083, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9817 + }, + { + "epoch": 15.68370607028754, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9818 + }, + { + "epoch": 15.685303514376997, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9819 + }, + { + "epoch": 15.686900958466454, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9820 + }, + { + "epoch": 15.688498402555911, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9821 + }, + { + "epoch": 15.690095846645367, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 9822 + }, + { + "epoch": 15.691693290734824, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9823 + }, + { + "epoch": 15.69329073482428, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9824 + }, + { + "epoch": 15.694888178913738, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9825 + }, + { + "epoch": 15.696485623003195, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9826 + }, + { + "epoch": 15.698083067092652, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 9827 + }, + { + "epoch": 15.699680511182109, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9828 + }, + { + "epoch": 15.701277955271566, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9829 + }, + { + "epoch": 15.702875399361023, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9830 + }, + { + "epoch": 15.704472843450478, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9831 + }, + { + "epoch": 15.706070287539935, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9832 + }, + { + "epoch": 15.707667731629392, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9833 + }, + { + "epoch": 15.70926517571885, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 9834 + }, + { + "epoch": 15.710862619808307, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9835 + }, + { + "epoch": 15.712460063897764, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9836 + }, + { + "epoch": 15.71405750798722, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9837 + }, + { + "epoch": 15.715654952076678, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 9838 + }, + { + "epoch": 15.717252396166135, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9839 + }, + { + "epoch": 15.718849840255592, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9840 + }, + { + "epoch": 15.720447284345047, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 9841 + }, + { + "epoch": 15.722044728434504, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9842 + }, + { + "epoch": 15.723642172523961, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9843 + }, + { + "epoch": 15.725239616613418, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9844 + }, + { + "epoch": 15.726837060702875, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9845 + }, + { + "epoch": 15.728434504792332, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9846 + }, + { + "epoch": 15.73003194888179, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 9847 + }, + { + "epoch": 15.731629392971247, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9848 + }, + { + "epoch": 15.733226837060704, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9849 + }, + { + "epoch": 15.73482428115016, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9850 + }, + { + "epoch": 15.736421725239616, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9851 + }, + { + "epoch": 15.738019169329073, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9852 + }, + { + "epoch": 15.73961661341853, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9853 + }, + { + "epoch": 15.741214057507987, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9854 + }, + { + "epoch": 15.742811501597444, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9855 + }, + { + "epoch": 15.744408945686901, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9856 + }, + { + "epoch": 15.746006389776358, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9857 + }, + { + "epoch": 15.747603833865815, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 9858 + }, + { + "epoch": 15.749201277955272, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9859 + }, + { + "epoch": 15.750798722044728, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9860 + }, + { + "epoch": 15.752396166134185, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9861 + }, + { + "epoch": 15.753993610223642, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9862 + }, + { + "epoch": 15.755591054313099, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 9863 + }, + { + "epoch": 15.757188498402556, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9864 + }, + { + "epoch": 15.758785942492013, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9865 + }, + { + "epoch": 15.76038338658147, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9866 + }, + { + "epoch": 15.761980830670927, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9867 + }, + { + "epoch": 15.763578274760384, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9868 + }, + { + "epoch": 15.76517571884984, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9869 + }, + { + "epoch": 15.766773162939296, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9870 + }, + { + "epoch": 15.768370607028753, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9871 + }, + { + "epoch": 15.76996805111821, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 9872 + }, + { + "epoch": 15.771565495207668, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9873 + }, + { + "epoch": 15.773162939297125, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9874 + }, + { + "epoch": 15.774760383386582, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9875 + }, + { + "epoch": 15.776357827476039, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9876 + }, + { + "epoch": 15.777955271565496, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9877 + }, + { + "epoch": 15.779552715654953, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9878 + }, + { + "epoch": 15.781150159744408, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9879 + }, + { + "epoch": 15.782747603833865, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 9880 + }, + { + "epoch": 15.784345047923322, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9881 + }, + { + "epoch": 15.78594249201278, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9882 + }, + { + "epoch": 15.787539936102236, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 9883 + }, + { + "epoch": 15.789137380191693, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9884 + }, + { + "epoch": 15.79073482428115, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9885 + }, + { + "epoch": 15.792332268370608, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9886 + }, + { + "epoch": 15.793929712460065, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9887 + }, + { + "epoch": 15.795527156549522, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9888 + }, + { + "epoch": 15.797124600638977, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9889 + }, + { + "epoch": 15.798722044728434, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9890 + }, + { + "epoch": 15.800319488817891, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9891 + }, + { + "epoch": 15.801916932907348, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9892 + }, + { + "epoch": 15.803514376996805, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9893 + }, + { + "epoch": 15.805111821086262, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9894 + }, + { + "epoch": 15.80670926517572, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9895 + }, + { + "epoch": 15.808306709265176, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9896 + }, + { + "epoch": 15.809904153354633, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9897 + }, + { + "epoch": 15.811501597444089, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9898 + }, + { + "epoch": 15.813099041533546, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9899 + }, + { + "epoch": 15.814696485623003, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9900 + }, + { + "epoch": 15.81629392971246, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 9901 + }, + { + "epoch": 15.817891373801917, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9902 + }, + { + "epoch": 15.819488817891374, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9903 + }, + { + "epoch": 15.821086261980831, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9904 + }, + { + "epoch": 15.822683706070288, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9905 + }, + { + "epoch": 15.824281150159745, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9906 + }, + { + "epoch": 15.8258785942492, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 9907 + }, + { + "epoch": 15.827476038338657, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 9908 + }, + { + "epoch": 15.829073482428115, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9909 + }, + { + "epoch": 15.830670926517572, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9910 + }, + { + "epoch": 15.832268370607029, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9911 + }, + { + "epoch": 15.833865814696486, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9912 + }, + { + "epoch": 15.835463258785943, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9913 + }, + { + "epoch": 15.8370607028754, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9914 + }, + { + "epoch": 15.838658146964857, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9915 + }, + { + "epoch": 15.840255591054314, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9916 + }, + { + "epoch": 15.84185303514377, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9917 + }, + { + "epoch": 15.843450479233226, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9918 + }, + { + "epoch": 15.845047923322683, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9919 + }, + { + "epoch": 15.84664536741214, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9920 + }, + { + "epoch": 15.848242811501597, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9921 + }, + { + "epoch": 15.849840255591054, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 9922 + }, + { + "epoch": 15.851437699680512, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9923 + }, + { + "epoch": 15.853035143769969, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 9924 + }, + { + "epoch": 15.854632587859426, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9925 + }, + { + "epoch": 15.856230031948883, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9926 + }, + { + "epoch": 15.857827476038338, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 9927 + }, + { + "epoch": 15.859424920127795, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9928 + }, + { + "epoch": 15.861022364217252, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9929 + }, + { + "epoch": 15.86261980830671, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9930 + }, + { + "epoch": 15.864217252396166, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9931 + }, + { + "epoch": 15.865814696485623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9932 + }, + { + "epoch": 15.86741214057508, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9933 + }, + { + "epoch": 15.869009584664537, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9934 + }, + { + "epoch": 15.870607028753994, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9935 + }, + { + "epoch": 15.87220447284345, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9936 + }, + { + "epoch": 15.873801916932907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9937 + }, + { + "epoch": 15.875399361022364, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9938 + }, + { + "epoch": 15.87699680511182, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9939 + }, + { + "epoch": 15.878594249201278, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9940 + }, + { + "epoch": 15.880191693290735, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9941 + }, + { + "epoch": 15.881789137380192, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9942 + }, + { + "epoch": 15.883386581469649, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 9943 + }, + { + "epoch": 15.884984025559106, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9944 + }, + { + "epoch": 15.886581469648561, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9945 + }, + { + "epoch": 15.888178913738018, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9946 + }, + { + "epoch": 15.889776357827476, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 9947 + }, + { + "epoch": 15.891373801916933, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9948 + }, + { + "epoch": 15.89297124600639, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9949 + }, + { + "epoch": 15.894568690095847, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9950 + }, + { + "epoch": 15.896166134185304, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9951 + }, + { + "epoch": 15.89776357827476, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9952 + }, + { + "epoch": 15.899361022364218, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9953 + }, + { + "epoch": 15.900958466453675, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9954 + }, + { + "epoch": 15.90255591054313, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9955 + }, + { + "epoch": 15.904153354632587, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9956 + }, + { + "epoch": 15.905750798722044, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9957 + }, + { + "epoch": 15.907348242811501, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9958 + }, + { + "epoch": 15.908945686900958, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9959 + }, + { + "epoch": 15.910543130990416, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9960 + }, + { + "epoch": 15.912140575079873, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9961 + }, + { + "epoch": 15.91373801916933, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9962 + }, + { + "epoch": 15.915335463258787, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9963 + }, + { + "epoch": 15.916932907348244, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9964 + }, + { + "epoch": 15.918530351437699, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 9965 + }, + { + "epoch": 15.920127795527156, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9966 + }, + { + "epoch": 15.921725239616613, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9967 + }, + { + "epoch": 15.92332268370607, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9968 + }, + { + "epoch": 15.924920127795527, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9969 + }, + { + "epoch": 15.926517571884984, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9970 + }, + { + "epoch": 15.928115015974441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9971 + }, + { + "epoch": 15.929712460063898, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9972 + }, + { + "epoch": 15.931309904153355, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9973 + }, + { + "epoch": 15.93290734824281, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 9974 + }, + { + "epoch": 15.934504792332268, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9975 + }, + { + "epoch": 15.936102236421725, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9976 + }, + { + "epoch": 15.937699680511182, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9977 + }, + { + "epoch": 15.939297124600639, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 9978 + }, + { + "epoch": 15.940894568690096, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9979 + }, + { + "epoch": 15.942492012779553, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9980 + }, + { + "epoch": 15.94408945686901, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9981 + }, + { + "epoch": 15.945686900958467, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9982 + }, + { + "epoch": 15.947284345047922, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9983 + }, + { + "epoch": 15.94888178913738, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9984 + }, + { + "epoch": 15.950479233226837, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9985 + }, + { + "epoch": 15.952076677316294, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9986 + }, + { + "epoch": 15.95367412140575, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9987 + }, + { + "epoch": 15.955271565495208, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9988 + }, + { + "epoch": 15.956869009584665, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9989 + }, + { + "epoch": 15.958466453674122, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 9990 + }, + { + "epoch": 15.960063897763579, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9991 + }, + { + "epoch": 15.961661341853034, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9992 + }, + { + "epoch": 15.963258785942491, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9993 + }, + { + "epoch": 15.964856230031948, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9994 + }, + { + "epoch": 15.966453674121405, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9995 + }, + { + "epoch": 15.968051118210862, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9996 + }, + { + "epoch": 15.96964856230032, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9997 + }, + { + "epoch": 15.971246006389777, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9998 + }, + { + "epoch": 15.972843450479234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9999 + }, + { + "epoch": 15.97444089456869, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10000 + }, + { + "epoch": 15.976038338658148, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10001 + }, + { + "epoch": 15.977635782747605, + "grad_norm": 0.48828125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10002 + }, + { + "epoch": 15.97923322683706, + "grad_norm": 0.56640625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 10003 + }, + { + "epoch": 15.980830670926517, + "grad_norm": 1.171875, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 10004 + }, + { + "epoch": 15.982428115015974, + "grad_norm": 1.703125, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 10005 + }, + { + "epoch": 15.984025559105431, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 10006 + }, + { + "epoch": 15.985623003194888, + "grad_norm": 0.9609375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 10007 + }, + { + "epoch": 15.987220447284345, + "grad_norm": 1.1015625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 10008 + }, + { + "epoch": 15.988817891373802, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 10009 + }, + { + "epoch": 15.99041533546326, + "grad_norm": 0.8125, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 10010 + }, + { + "epoch": 15.992012779552716, + "grad_norm": 0.5703125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 10011 + }, + { + "epoch": 15.993610223642172, + "grad_norm": 0.333984375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 10012 + }, + { + "epoch": 15.995207667731629, + "grad_norm": 0.54296875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 10013 + }, + { + "epoch": 15.996805111821086, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 10014 + }, + { + "epoch": 15.998402555910543, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 10015 + }, + { + "epoch": 16.0, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 10016 + }, + { + "epoch": 16.001597444089455, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 10017 + }, + { + "epoch": 16.003194888178914, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 10018 + }, + { + "epoch": 16.00479233226837, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 10019 + }, + { + "epoch": 16.00638977635783, + "grad_norm": 0.333984375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10020 + }, + { + "epoch": 16.007987220447284, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 10021 + }, + { + "epoch": 16.009584664536742, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 10022 + }, + { + "epoch": 16.011182108626198, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10023 + }, + { + "epoch": 16.012779552715656, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 10024 + }, + { + "epoch": 16.01437699680511, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 10025 + }, + { + "epoch": 16.015974440894567, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10026 + }, + { + "epoch": 16.017571884984026, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 10027 + }, + { + "epoch": 16.01916932907348, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10028 + }, + { + "epoch": 16.02076677316294, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 10029 + }, + { + "epoch": 16.022364217252395, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10030 + }, + { + "epoch": 16.023961661341854, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10031 + }, + { + "epoch": 16.02555910543131, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10032 + }, + { + "epoch": 16.027156549520768, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10033 + }, + { + "epoch": 16.028753993610223, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 10034 + }, + { + "epoch": 16.03035143769968, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10035 + }, + { + "epoch": 16.031948881789138, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10036 + }, + { + "epoch": 16.033546325878593, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10037 + }, + { + "epoch": 16.03514376996805, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 10038 + }, + { + "epoch": 16.036741214057507, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10039 + }, + { + "epoch": 16.038338658146966, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10040 + }, + { + "epoch": 16.03993610223642, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10041 + }, + { + "epoch": 16.04153354632588, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 10042 + }, + { + "epoch": 16.043130990415335, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10043 + }, + { + "epoch": 16.044728434504794, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10044 + }, + { + "epoch": 16.04632587859425, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10045 + }, + { + "epoch": 16.047923322683705, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10046 + }, + { + "epoch": 16.049520766773163, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 10047 + }, + { + "epoch": 16.05111821086262, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10048 + }, + { + "epoch": 16.052715654952078, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10049 + }, + { + "epoch": 16.054313099041533, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10050 + }, + { + "epoch": 16.05591054313099, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10051 + }, + { + "epoch": 16.057507987220447, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10052 + }, + { + "epoch": 16.059105431309906, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 10053 + }, + { + "epoch": 16.06070287539936, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10054 + }, + { + "epoch": 16.062300319488816, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10055 + }, + { + "epoch": 16.063897763578275, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10056 + }, + { + "epoch": 16.06549520766773, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10057 + }, + { + "epoch": 16.06709265175719, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10058 + }, + { + "epoch": 16.068690095846645, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10059 + }, + { + "epoch": 16.070287539936103, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10060 + }, + { + "epoch": 16.07188498402556, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 10061 + }, + { + "epoch": 16.073482428115017, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10062 + }, + { + "epoch": 16.075079872204473, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10063 + }, + { + "epoch": 16.076677316293928, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10064 + }, + { + "epoch": 16.078274760383387, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10065 + }, + { + "epoch": 16.079872204472842, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10066 + }, + { + "epoch": 16.0814696485623, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10067 + }, + { + "epoch": 16.083067092651756, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10068 + }, + { + "epoch": 16.084664536741215, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10069 + }, + { + "epoch": 16.08626198083067, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 10070 + }, + { + "epoch": 16.08785942492013, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10071 + }, + { + "epoch": 16.089456869009584, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10072 + }, + { + "epoch": 16.09105431309904, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10073 + }, + { + "epoch": 16.0926517571885, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10074 + }, + { + "epoch": 16.094249201277954, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10075 + }, + { + "epoch": 16.095846645367413, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10076 + }, + { + "epoch": 16.097444089456868, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10077 + }, + { + "epoch": 16.099041533546327, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10078 + }, + { + "epoch": 16.100638977635782, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10079 + }, + { + "epoch": 16.10223642172524, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10080 + }, + { + "epoch": 16.103833865814696, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10081 + }, + { + "epoch": 16.105431309904155, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10082 + }, + { + "epoch": 16.10702875399361, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10083 + }, + { + "epoch": 16.108626198083066, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10084 + }, + { + "epoch": 16.110223642172524, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10085 + }, + { + "epoch": 16.11182108626198, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10086 + }, + { + "epoch": 16.11341853035144, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10087 + }, + { + "epoch": 16.115015974440894, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10088 + }, + { + "epoch": 16.116613418530353, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10089 + }, + { + "epoch": 16.118210862619808, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10090 + }, + { + "epoch": 16.119808306709267, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10091 + }, + { + "epoch": 16.121405750798722, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10092 + }, + { + "epoch": 16.123003194888177, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10093 + }, + { + "epoch": 16.124600638977636, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10094 + }, + { + "epoch": 16.12619808306709, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10095 + }, + { + "epoch": 16.12779552715655, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10096 + }, + { + "epoch": 16.129392971246006, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10097 + }, + { + "epoch": 16.130990415335464, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10098 + }, + { + "epoch": 16.13258785942492, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10099 + }, + { + "epoch": 16.13418530351438, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10100 + }, + { + "epoch": 16.135782747603834, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10101 + }, + { + "epoch": 16.13738019169329, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10102 + }, + { + "epoch": 16.138977635782748, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10103 + }, + { + "epoch": 16.140575079872203, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10104 + }, + { + "epoch": 16.142172523961662, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 10105 + }, + { + "epoch": 16.143769968051117, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10106 + }, + { + "epoch": 16.145367412140576, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10107 + }, + { + "epoch": 16.14696485623003, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10108 + }, + { + "epoch": 16.14856230031949, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10109 + }, + { + "epoch": 16.150159744408946, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10110 + }, + { + "epoch": 16.1517571884984, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10111 + }, + { + "epoch": 16.15335463258786, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10112 + }, + { + "epoch": 16.154952076677315, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10113 + }, + { + "epoch": 16.156549520766774, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10114 + }, + { + "epoch": 16.15814696485623, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10115 + }, + { + "epoch": 16.159744408945688, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10116 + }, + { + "epoch": 16.161341853035143, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10117 + }, + { + "epoch": 16.162939297124602, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10118 + }, + { + "epoch": 16.164536741214057, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 10119 + }, + { + "epoch": 16.166134185303516, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10120 + }, + { + "epoch": 16.16773162939297, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10121 + }, + { + "epoch": 16.169329073482427, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10122 + }, + { + "epoch": 16.170926517571885, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10123 + }, + { + "epoch": 16.17252396166134, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10124 + }, + { + "epoch": 16.1741214057508, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10125 + }, + { + "epoch": 16.175718849840255, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10126 + }, + { + "epoch": 16.177316293929714, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10127 + }, + { + "epoch": 16.17891373801917, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10128 + }, + { + "epoch": 16.180511182108628, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10129 + }, + { + "epoch": 16.182108626198083, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10130 + }, + { + "epoch": 16.18370607028754, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10131 + }, + { + "epoch": 16.185303514376997, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10132 + }, + { + "epoch": 16.186900958466452, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10133 + }, + { + "epoch": 16.18849840255591, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 10134 + }, + { + "epoch": 16.190095846645367, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10135 + }, + { + "epoch": 16.191693290734825, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 10136 + }, + { + "epoch": 16.19329073482428, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 10137 + }, + { + "epoch": 16.19488817891374, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10138 + }, + { + "epoch": 16.196485623003195, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10139 + }, + { + "epoch": 16.19808306709265, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10140 + }, + { + "epoch": 16.19968051118211, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10141 + }, + { + "epoch": 16.201277955271564, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10142 + }, + { + "epoch": 16.202875399361023, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10143 + }, + { + "epoch": 16.20447284345048, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10144 + }, + { + "epoch": 16.206070287539937, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10145 + }, + { + "epoch": 16.207667731629392, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10146 + }, + { + "epoch": 16.20926517571885, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 10147 + }, + { + "epoch": 16.210862619808307, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10148 + }, + { + "epoch": 16.212460063897762, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10149 + }, + { + "epoch": 16.21405750798722, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 10150 + }, + { + "epoch": 16.215654952076676, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10151 + }, + { + "epoch": 16.217252396166135, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10152 + }, + { + "epoch": 16.21884984025559, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10153 + }, + { + "epoch": 16.22044728434505, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10154 + }, + { + "epoch": 16.222044728434504, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 10155 + }, + { + "epoch": 16.223642172523963, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10156 + }, + { + "epoch": 16.22523961661342, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10157 + }, + { + "epoch": 16.226837060702877, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10158 + }, + { + "epoch": 16.228434504792332, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 10159 + }, + { + "epoch": 16.230031948881788, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10160 + }, + { + "epoch": 16.231629392971247, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10161 + }, + { + "epoch": 16.233226837060702, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10162 + }, + { + "epoch": 16.23482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10163 + }, + { + "epoch": 16.236421725239616, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10164 + }, + { + "epoch": 16.238019169329075, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10165 + }, + { + "epoch": 16.23961661341853, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10166 + }, + { + "epoch": 16.24121405750799, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10167 + }, + { + "epoch": 16.242811501597444, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10168 + }, + { + "epoch": 16.2444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10169 + }, + { + "epoch": 16.24600638977636, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10170 + }, + { + "epoch": 16.247603833865814, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10171 + }, + { + "epoch": 16.249201277955272, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10172 + }, + { + "epoch": 16.250798722044728, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10173 + }, + { + "epoch": 16.252396166134186, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 10174 + }, + { + "epoch": 16.25399361022364, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10175 + }, + { + "epoch": 16.2555910543131, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 10176 + }, + { + "epoch": 16.257188498402556, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10177 + }, + { + "epoch": 16.25878594249201, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10178 + }, + { + "epoch": 16.26038338658147, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10179 + }, + { + "epoch": 16.261980830670925, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10180 + }, + { + "epoch": 16.263578274760384, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10181 + }, + { + "epoch": 16.26517571884984, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10182 + }, + { + "epoch": 16.266773162939298, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10183 + }, + { + "epoch": 16.268370607028753, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10184 + }, + { + "epoch": 16.269968051118212, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10185 + }, + { + "epoch": 16.271565495207668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10186 + }, + { + "epoch": 16.273162939297123, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10187 + }, + { + "epoch": 16.27476038338658, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10188 + }, + { + "epoch": 16.276357827476037, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10189 + }, + { + "epoch": 16.277955271565496, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10190 + }, + { + "epoch": 16.27955271565495, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10191 + }, + { + "epoch": 16.28115015974441, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10192 + }, + { + "epoch": 16.282747603833865, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10193 + }, + { + "epoch": 16.284345047923324, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10194 + }, + { + "epoch": 16.28594249201278, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10195 + }, + { + "epoch": 16.287539936102238, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10196 + }, + { + "epoch": 16.289137380191693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10197 + }, + { + "epoch": 16.29073482428115, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10198 + }, + { + "epoch": 16.292332268370608, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10199 + }, + { + "epoch": 16.293929712460063, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10200 + }, + { + "epoch": 16.29552715654952, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 10201 + }, + { + "epoch": 16.297124600638977, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10202 + }, + { + "epoch": 16.298722044728436, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10203 + }, + { + "epoch": 16.30031948881789, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10204 + }, + { + "epoch": 16.30191693290735, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10205 + }, + { + "epoch": 16.303514376996805, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 10206 + }, + { + "epoch": 16.30511182108626, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10207 + }, + { + "epoch": 16.30670926517572, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10208 + }, + { + "epoch": 16.308306709265175, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10209 + }, + { + "epoch": 16.309904153354633, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10210 + }, + { + "epoch": 16.31150159744409, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10211 + }, + { + "epoch": 16.313099041533548, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10212 + }, + { + "epoch": 16.314696485623003, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10213 + }, + { + "epoch": 16.31629392971246, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10214 + }, + { + "epoch": 16.317891373801917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 10215 + }, + { + "epoch": 16.319488817891372, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10216 + }, + { + "epoch": 16.32108626198083, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10217 + }, + { + "epoch": 16.322683706070286, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10218 + }, + { + "epoch": 16.324281150159745, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10219 + }, + { + "epoch": 16.3258785942492, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10220 + }, + { + "epoch": 16.32747603833866, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10221 + }, + { + "epoch": 16.329073482428115, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10222 + }, + { + "epoch": 16.330670926517573, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10223 + }, + { + "epoch": 16.33226837060703, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10224 + }, + { + "epoch": 16.333865814696484, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10225 + }, + { + "epoch": 16.335463258785943, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10226 + }, + { + "epoch": 16.337060702875398, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 10227 + }, + { + "epoch": 16.338658146964857, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10228 + }, + { + "epoch": 16.340255591054312, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10229 + }, + { + "epoch": 16.34185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10230 + }, + { + "epoch": 16.343450479233226, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10231 + }, + { + "epoch": 16.345047923322685, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 10232 + }, + { + "epoch": 16.34664536741214, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10233 + }, + { + "epoch": 16.3482428115016, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10234 + }, + { + "epoch": 16.349840255591054, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10235 + }, + { + "epoch": 16.35143769968051, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10236 + }, + { + "epoch": 16.35303514376997, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10237 + }, + { + "epoch": 16.354632587859424, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 10238 + }, + { + "epoch": 16.356230031948883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10239 + }, + { + "epoch": 16.357827476038338, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10240 + }, + { + "epoch": 16.359424920127797, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 10241 + }, + { + "epoch": 16.361022364217252, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 10242 + }, + { + "epoch": 16.36261980830671, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10243 + }, + { + "epoch": 16.364217252396166, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10244 + }, + { + "epoch": 16.36581469648562, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10245 + }, + { + "epoch": 16.36741214057508, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10246 + }, + { + "epoch": 16.369009584664536, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10247 + }, + { + "epoch": 16.370607028753994, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10248 + }, + { + "epoch": 16.37220447284345, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10249 + }, + { + "epoch": 16.37380191693291, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10250 + }, + { + "epoch": 16.375399361022364, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10251 + }, + { + "epoch": 16.376996805111823, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10252 + }, + { + "epoch": 16.378594249201278, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10253 + }, + { + "epoch": 16.380191693290733, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 10254 + }, + { + "epoch": 16.381789137380192, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 10255 + }, + { + "epoch": 16.383386581469647, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10256 + }, + { + "epoch": 16.384984025559106, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10257 + }, + { + "epoch": 16.38658146964856, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10258 + }, + { + "epoch": 16.38817891373802, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10259 + }, + { + "epoch": 16.389776357827476, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10260 + }, + { + "epoch": 16.391373801916934, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10261 + }, + { + "epoch": 16.39297124600639, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10262 + }, + { + "epoch": 16.394568690095845, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10263 + }, + { + "epoch": 16.396166134185304, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10264 + }, + { + "epoch": 16.39776357827476, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10265 + }, + { + "epoch": 16.399361022364218, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10266 + }, + { + "epoch": 16.400958466453673, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10267 + }, + { + "epoch": 16.402555910543132, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10268 + }, + { + "epoch": 16.404153354632587, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10269 + }, + { + "epoch": 16.405750798722046, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10270 + }, + { + "epoch": 16.4073482428115, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10271 + }, + { + "epoch": 16.408945686900957, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10272 + }, + { + "epoch": 16.410543130990416, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10273 + }, + { + "epoch": 16.41214057507987, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10274 + }, + { + "epoch": 16.41373801916933, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10275 + }, + { + "epoch": 16.415335463258785, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10276 + }, + { + "epoch": 16.416932907348244, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10277 + }, + { + "epoch": 16.4185303514377, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10278 + }, + { + "epoch": 16.420127795527158, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10279 + }, + { + "epoch": 16.421725239616613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 10280 + }, + { + "epoch": 16.423322683706072, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10281 + }, + { + "epoch": 16.424920127795527, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10282 + }, + { + "epoch": 16.426517571884983, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10283 + }, + { + "epoch": 16.42811501597444, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10284 + }, + { + "epoch": 16.429712460063897, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10285 + }, + { + "epoch": 16.431309904153355, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10286 + }, + { + "epoch": 16.43290734824281, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10287 + }, + { + "epoch": 16.43450479233227, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10288 + }, + { + "epoch": 16.436102236421725, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10289 + }, + { + "epoch": 16.437699680511184, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 10290 + }, + { + "epoch": 16.43929712460064, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10291 + }, + { + "epoch": 16.440894568690094, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10292 + }, + { + "epoch": 16.442492012779553, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 10293 + }, + { + "epoch": 16.44408945686901, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10294 + }, + { + "epoch": 16.445686900958467, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10295 + }, + { + "epoch": 16.447284345047922, + "grad_norm": 0.031494140625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10296 + }, + { + "epoch": 16.44888178913738, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10297 + }, + { + "epoch": 16.450479233226837, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 10298 + }, + { + "epoch": 16.452076677316295, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 10299 + }, + { + "epoch": 16.45367412140575, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10300 + }, + { + "epoch": 16.455271565495206, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 10301 + }, + { + "epoch": 16.456869009584665, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 10302 + }, + { + "epoch": 16.45846645367412, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10303 + }, + { + "epoch": 16.46006389776358, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10304 + }, + { + "epoch": 16.461661341853034, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10305 + }, + { + "epoch": 16.463258785942493, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10306 + }, + { + "epoch": 16.46485623003195, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10307 + }, + { + "epoch": 16.466453674121407, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10308 + }, + { + "epoch": 16.468051118210862, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10309 + }, + { + "epoch": 16.46964856230032, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10310 + }, + { + "epoch": 16.471246006389777, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 10311 + }, + { + "epoch": 16.472843450479232, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10312 + }, + { + "epoch": 16.47444089456869, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10313 + }, + { + "epoch": 16.476038338658146, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10314 + }, + { + "epoch": 16.477635782747605, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10315 + }, + { + "epoch": 16.47923322683706, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10316 + }, + { + "epoch": 16.48083067092652, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10317 + }, + { + "epoch": 16.482428115015974, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10318 + }, + { + "epoch": 16.484025559105433, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10319 + }, + { + "epoch": 16.48562300319489, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 10320 + }, + { + "epoch": 16.487220447284344, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 10321 + }, + { + "epoch": 16.488817891373802, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10322 + }, + { + "epoch": 16.490415335463258, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 10323 + }, + { + "epoch": 16.492012779552716, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10324 + }, + { + "epoch": 16.49361022364217, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10325 + }, + { + "epoch": 16.49520766773163, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10326 + }, + { + "epoch": 16.496805111821086, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10327 + }, + { + "epoch": 16.498402555910545, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 10328 + }, + { + "epoch": 16.5, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10329 + }, + { + "epoch": 16.501597444089455, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10330 + }, + { + "epoch": 16.503194888178914, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10331 + }, + { + "epoch": 16.50479233226837, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10332 + }, + { + "epoch": 16.50638977635783, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10333 + }, + { + "epoch": 16.507987220447284, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10334 + }, + { + "epoch": 16.509584664536742, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10335 + }, + { + "epoch": 16.511182108626198, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10336 + }, + { + "epoch": 16.512779552715656, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10337 + }, + { + "epoch": 16.51437699680511, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 10338 + }, + { + "epoch": 16.515974440894567, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10339 + }, + { + "epoch": 16.517571884984026, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10340 + }, + { + "epoch": 16.51916932907348, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 10341 + }, + { + "epoch": 16.52076677316294, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10342 + }, + { + "epoch": 16.522364217252395, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10343 + }, + { + "epoch": 16.523961661341854, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 10344 + }, + { + "epoch": 16.52555910543131, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10345 + }, + { + "epoch": 16.527156549520768, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 10346 + }, + { + "epoch": 16.528753993610223, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 10347 + }, + { + "epoch": 16.53035143769968, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 10348 + }, + { + "epoch": 16.531948881789138, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10349 + }, + { + "epoch": 16.533546325878593, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10350 + }, + { + "epoch": 16.53514376996805, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10351 + }, + { + "epoch": 16.536741214057507, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10352 + }, + { + "epoch": 16.538338658146966, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10353 + }, + { + "epoch": 16.53993610223642, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10354 + }, + { + "epoch": 16.54153354632588, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10355 + }, + { + "epoch": 16.543130990415335, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10356 + }, + { + "epoch": 16.544728434504794, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10357 + }, + { + "epoch": 16.54632587859425, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10358 + }, + { + "epoch": 16.547923322683705, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10359 + }, + { + "epoch": 16.549520766773163, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10360 + }, + { + "epoch": 16.55111821086262, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10361 + }, + { + "epoch": 16.552715654952078, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10362 + }, + { + "epoch": 16.554313099041533, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 10363 + }, + { + "epoch": 16.55591054313099, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10364 + }, + { + "epoch": 16.557507987220447, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10365 + }, + { + "epoch": 16.559105431309906, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10366 + }, + { + "epoch": 16.56070287539936, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10367 + }, + { + "epoch": 16.562300319488816, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10368 + }, + { + "epoch": 16.563897763578275, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10369 + }, + { + "epoch": 16.56549520766773, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 10370 + }, + { + "epoch": 16.56709265175719, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10371 + }, + { + "epoch": 16.568690095846645, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10372 + }, + { + "epoch": 16.570287539936103, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 10373 + }, + { + "epoch": 16.57188498402556, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10374 + }, + { + "epoch": 16.573482428115017, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10375 + }, + { + "epoch": 16.575079872204473, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10376 + }, + { + "epoch": 16.576677316293928, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10377 + }, + { + "epoch": 16.578274760383387, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10378 + }, + { + "epoch": 16.579872204472842, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10379 + }, + { + "epoch": 16.5814696485623, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 10380 + }, + { + "epoch": 16.583067092651756, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10381 + }, + { + "epoch": 16.584664536741215, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10382 + }, + { + "epoch": 16.58626198083067, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10383 + }, + { + "epoch": 16.58785942492013, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10384 + }, + { + "epoch": 16.589456869009584, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10385 + }, + { + "epoch": 16.591054313099043, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10386 + }, + { + "epoch": 16.5926517571885, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10387 + }, + { + "epoch": 16.594249201277954, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10388 + }, + { + "epoch": 16.595846645367413, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10389 + }, + { + "epoch": 16.597444089456868, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10390 + }, + { + "epoch": 16.599041533546327, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 10391 + }, + { + "epoch": 16.600638977635782, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10392 + }, + { + "epoch": 16.60223642172524, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10393 + }, + { + "epoch": 16.603833865814696, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10394 + }, + { + "epoch": 16.605431309904155, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10395 + }, + { + "epoch": 16.60702875399361, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10396 + }, + { + "epoch": 16.608626198083066, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10397 + }, + { + "epoch": 16.610223642172524, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10398 + }, + { + "epoch": 16.61182108626198, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10399 + }, + { + "epoch": 16.61341853035144, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10400 + }, + { + "epoch": 16.615015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 10401 + }, + { + "epoch": 16.616613418530353, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10402 + }, + { + "epoch": 16.618210862619808, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10403 + }, + { + "epoch": 16.619808306709267, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10404 + }, + { + "epoch": 16.621405750798722, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10405 + }, + { + "epoch": 16.623003194888177, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10406 + }, + { + "epoch": 16.624600638977636, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10407 + }, + { + "epoch": 16.62619808306709, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10408 + }, + { + "epoch": 16.62779552715655, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10409 + }, + { + "epoch": 16.629392971246006, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10410 + }, + { + "epoch": 16.630990415335464, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10411 + }, + { + "epoch": 16.63258785942492, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10412 + }, + { + "epoch": 16.63418530351438, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10413 + }, + { + "epoch": 16.635782747603834, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 10414 + }, + { + "epoch": 16.63738019169329, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10415 + }, + { + "epoch": 16.638977635782748, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 10416 + }, + { + "epoch": 16.640575079872203, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 10417 + }, + { + "epoch": 16.642172523961662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10418 + }, + { + "epoch": 16.643769968051117, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10419 + }, + { + "epoch": 16.645367412140576, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10420 + }, + { + "epoch": 16.64696485623003, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10421 + }, + { + "epoch": 16.64856230031949, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10422 + }, + { + "epoch": 16.650159744408946, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10423 + }, + { + "epoch": 16.6517571884984, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10424 + }, + { + "epoch": 16.65335463258786, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10425 + }, + { + "epoch": 16.654952076677315, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10426 + }, + { + "epoch": 16.656549520766774, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10427 + }, + { + "epoch": 16.65814696485623, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 10428 + }, + { + "epoch": 16.659744408945688, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10429 + }, + { + "epoch": 16.661341853035143, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10430 + }, + { + "epoch": 16.662939297124602, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10431 + }, + { + "epoch": 16.664536741214057, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 10432 + }, + { + "epoch": 16.666134185303516, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10433 + }, + { + "epoch": 16.66773162939297, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10434 + }, + { + "epoch": 16.669329073482427, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10435 + }, + { + "epoch": 16.670926517571885, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10436 + }, + { + "epoch": 16.67252396166134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10437 + }, + { + "epoch": 16.6741214057508, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10438 + }, + { + "epoch": 16.675718849840255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10439 + }, + { + "epoch": 16.677316293929714, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10440 + }, + { + "epoch": 16.67891373801917, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10441 + }, + { + "epoch": 16.680511182108628, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10442 + }, + { + "epoch": 16.682108626198083, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10443 + }, + { + "epoch": 16.68370607028754, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10444 + }, + { + "epoch": 16.685303514376997, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 10445 + }, + { + "epoch": 16.686900958466452, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10446 + }, + { + "epoch": 16.68849840255591, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10447 + }, + { + "epoch": 16.690095846645367, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10448 + }, + { + "epoch": 16.691693290734825, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 10449 + }, + { + "epoch": 16.69329073482428, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10450 + }, + { + "epoch": 16.69488817891374, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10451 + }, + { + "epoch": 16.696485623003195, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10452 + }, + { + "epoch": 16.69808306709265, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10453 + }, + { + "epoch": 16.69968051118211, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10454 + }, + { + "epoch": 16.701277955271564, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10455 + }, + { + "epoch": 16.702875399361023, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10456 + }, + { + "epoch": 16.70447284345048, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10457 + }, + { + "epoch": 16.706070287539937, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10458 + }, + { + "epoch": 16.707667731629392, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 10459 + }, + { + "epoch": 16.70926517571885, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10460 + }, + { + "epoch": 16.710862619808307, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10461 + }, + { + "epoch": 16.712460063897765, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10462 + }, + { + "epoch": 16.71405750798722, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10463 + }, + { + "epoch": 16.715654952076676, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10464 + }, + { + "epoch": 16.717252396166135, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10465 + }, + { + "epoch": 16.71884984025559, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10466 + }, + { + "epoch": 16.72044728434505, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 10467 + }, + { + "epoch": 16.722044728434504, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10468 + }, + { + "epoch": 16.723642172523963, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10469 + }, + { + "epoch": 16.72523961661342, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10470 + }, + { + "epoch": 16.726837060702877, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10471 + }, + { + "epoch": 16.728434504792332, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10472 + }, + { + "epoch": 16.730031948881788, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10473 + }, + { + "epoch": 16.731629392971247, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10474 + }, + { + "epoch": 16.733226837060702, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10475 + }, + { + "epoch": 16.73482428115016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10476 + }, + { + "epoch": 16.736421725239616, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10477 + }, + { + "epoch": 16.738019169329075, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10478 + }, + { + "epoch": 16.73961661341853, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10479 + }, + { + "epoch": 16.74121405750799, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10480 + }, + { + "epoch": 16.742811501597444, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10481 + }, + { + "epoch": 16.7444089456869, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10482 + }, + { + "epoch": 16.74600638977636, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 10483 + }, + { + "epoch": 16.747603833865814, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10484 + }, + { + "epoch": 16.749201277955272, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10485 + }, + { + "epoch": 16.750798722044728, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10486 + }, + { + "epoch": 16.752396166134186, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10487 + }, + { + "epoch": 16.75399361022364, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 10488 + }, + { + "epoch": 16.7555910543131, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10489 + }, + { + "epoch": 16.757188498402556, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10490 + }, + { + "epoch": 16.75878594249201, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10491 + }, + { + "epoch": 16.76038338658147, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10492 + }, + { + "epoch": 16.761980830670925, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10493 + }, + { + "epoch": 16.763578274760384, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10494 + }, + { + "epoch": 16.76517571884984, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10495 + }, + { + "epoch": 16.766773162939298, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10496 + }, + { + "epoch": 16.768370607028753, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10497 + }, + { + "epoch": 16.769968051118212, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10498 + }, + { + "epoch": 16.771565495207668, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10499 + }, + { + "epoch": 16.773162939297123, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10500 + }, + { + "epoch": 16.77476038338658, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10501 + }, + { + "epoch": 16.776357827476037, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10502 + }, + { + "epoch": 16.777955271565496, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10503 + }, + { + "epoch": 16.77955271565495, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10504 + }, + { + "epoch": 16.78115015974441, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10505 + }, + { + "epoch": 16.782747603833865, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10506 + }, + { + "epoch": 16.784345047923324, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10507 + }, + { + "epoch": 16.78594249201278, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10508 + }, + { + "epoch": 16.787539936102235, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10509 + }, + { + "epoch": 16.789137380191693, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10510 + }, + { + "epoch": 16.79073482428115, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10511 + }, + { + "epoch": 16.792332268370608, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10512 + }, + { + "epoch": 16.793929712460063, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 10513 + }, + { + "epoch": 16.79552715654952, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10514 + }, + { + "epoch": 16.797124600638977, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10515 + }, + { + "epoch": 16.798722044728436, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 10516 + }, + { + "epoch": 16.80031948881789, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10517 + }, + { + "epoch": 16.80191693290735, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 10518 + }, + { + "epoch": 16.803514376996805, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10519 + }, + { + "epoch": 16.80511182108626, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10520 + }, + { + "epoch": 16.80670926517572, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10521 + }, + { + "epoch": 16.808306709265175, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10522 + }, + { + "epoch": 16.809904153354633, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10523 + }, + { + "epoch": 16.81150159744409, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10524 + }, + { + "epoch": 16.813099041533548, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10525 + }, + { + "epoch": 16.814696485623003, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10526 + }, + { + "epoch": 16.81629392971246, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10527 + }, + { + "epoch": 16.817891373801917, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10528 + }, + { + "epoch": 16.819488817891372, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10529 + }, + { + "epoch": 16.82108626198083, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10530 + }, + { + "epoch": 16.822683706070286, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10531 + }, + { + "epoch": 16.824281150159745, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 10532 + }, + { + "epoch": 16.8258785942492, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10533 + }, + { + "epoch": 16.82747603833866, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10534 + }, + { + "epoch": 16.829073482428115, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10535 + }, + { + "epoch": 16.830670926517573, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10536 + }, + { + "epoch": 16.83226837060703, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10537 + }, + { + "epoch": 16.833865814696484, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10538 + }, + { + "epoch": 16.835463258785943, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10539 + }, + { + "epoch": 16.837060702875398, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10540 + }, + { + "epoch": 16.838658146964857, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10541 + }, + { + "epoch": 16.840255591054312, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 10542 + }, + { + "epoch": 16.84185303514377, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10543 + }, + { + "epoch": 16.843450479233226, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10544 + }, + { + "epoch": 16.845047923322685, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10545 + }, + { + "epoch": 16.84664536741214, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10546 + }, + { + "epoch": 16.8482428115016, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10547 + }, + { + "epoch": 16.849840255591054, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10548 + }, + { + "epoch": 16.85143769968051, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10549 + }, + { + "epoch": 16.85303514376997, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10550 + }, + { + "epoch": 16.854632587859424, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10551 + }, + { + "epoch": 16.856230031948883, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10552 + }, + { + "epoch": 16.857827476038338, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10553 + }, + { + "epoch": 16.859424920127797, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10554 + }, + { + "epoch": 16.861022364217252, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10555 + }, + { + "epoch": 16.86261980830671, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10556 + }, + { + "epoch": 16.864217252396166, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 10557 + }, + { + "epoch": 16.86581469648562, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10558 + }, + { + "epoch": 16.86741214057508, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10559 + }, + { + "epoch": 16.869009584664536, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10560 + }, + { + "epoch": 16.870607028753994, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 10561 + }, + { + "epoch": 16.87220447284345, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 10562 + }, + { + "epoch": 16.87380191693291, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10563 + }, + { + "epoch": 16.875399361022364, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 10564 + }, + { + "epoch": 16.876996805111823, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10565 + }, + { + "epoch": 16.878594249201278, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10566 + }, + { + "epoch": 16.880191693290733, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10567 + }, + { + "epoch": 16.881789137380192, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 10568 + }, + { + "epoch": 16.883386581469647, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10569 + }, + { + "epoch": 16.884984025559106, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10570 + }, + { + "epoch": 16.88658146964856, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10571 + }, + { + "epoch": 16.88817891373802, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10572 + }, + { + "epoch": 16.889776357827476, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10573 + }, + { + "epoch": 16.891373801916934, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10574 + }, + { + "epoch": 16.89297124600639, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10575 + }, + { + "epoch": 16.894568690095845, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 10576 + }, + { + "epoch": 16.896166134185304, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10577 + }, + { + "epoch": 16.89776357827476, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 10578 + }, + { + "epoch": 16.899361022364218, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10579 + }, + { + "epoch": 16.900958466453673, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10580 + }, + { + "epoch": 16.902555910543132, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10581 + }, + { + "epoch": 16.904153354632587, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10582 + }, + { + "epoch": 16.905750798722046, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10583 + }, + { + "epoch": 16.9073482428115, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 10584 + }, + { + "epoch": 16.908945686900957, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 10585 + }, + { + "epoch": 16.910543130990416, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10586 + }, + { + "epoch": 16.91214057507987, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10587 + }, + { + "epoch": 16.91373801916933, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10588 + }, + { + "epoch": 16.915335463258785, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10589 + }, + { + "epoch": 16.916932907348244, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 10590 + }, + { + "epoch": 16.9185303514377, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10591 + }, + { + "epoch": 16.920127795527158, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10592 + }, + { + "epoch": 16.921725239616613, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10593 + }, + { + "epoch": 16.923322683706072, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10594 + }, + { + "epoch": 16.924920127795527, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10595 + }, + { + "epoch": 16.926517571884983, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10596 + }, + { + "epoch": 16.92811501597444, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10597 + }, + { + "epoch": 16.929712460063897, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10598 + }, + { + "epoch": 16.931309904153355, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10599 + }, + { + "epoch": 16.93290734824281, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10600 + }, + { + "epoch": 16.93450479233227, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10601 + }, + { + "epoch": 16.936102236421725, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10602 + }, + { + "epoch": 16.937699680511184, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 10603 + }, + { + "epoch": 16.93929712460064, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10604 + }, + { + "epoch": 16.940894568690094, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10605 + }, + { + "epoch": 16.942492012779553, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10606 + }, + { + "epoch": 16.94408945686901, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10607 + }, + { + "epoch": 16.945686900958467, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 10608 + }, + { + "epoch": 16.947284345047922, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10609 + }, + { + "epoch": 16.94888178913738, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10610 + }, + { + "epoch": 16.950479233226837, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10611 + }, + { + "epoch": 16.952076677316295, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10612 + }, + { + "epoch": 16.95367412140575, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10613 + }, + { + "epoch": 16.955271565495206, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10614 + }, + { + "epoch": 16.956869009584665, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10615 + }, + { + "epoch": 16.95846645367412, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10616 + }, + { + "epoch": 16.96006389776358, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10617 + }, + { + "epoch": 16.961661341853034, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10618 + }, + { + "epoch": 16.963258785942493, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 10619 + }, + { + "epoch": 16.96485623003195, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10620 + }, + { + "epoch": 16.966453674121407, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10621 + }, + { + "epoch": 16.968051118210862, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10622 + }, + { + "epoch": 16.96964856230032, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10623 + }, + { + "epoch": 16.971246006389777, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10624 + }, + { + "epoch": 16.972843450479232, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10625 + }, + { + "epoch": 16.97444089456869, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10626 + }, + { + "epoch": 16.976038338658146, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10627 + }, + { + "epoch": 16.977635782747605, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10628 + }, + { + "epoch": 16.97923322683706, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10629 + }, + { + "epoch": 16.98083067092652, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10630 + }, + { + "epoch": 16.982428115015974, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10631 + }, + { + "epoch": 16.984025559105433, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10632 + }, + { + "epoch": 16.98562300319489, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10633 + }, + { + "epoch": 16.987220447284344, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10634 + }, + { + "epoch": 16.988817891373802, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 10635 + }, + { + "epoch": 16.990415335463258, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10636 + }, + { + "epoch": 16.992012779552716, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10637 + }, + { + "epoch": 16.99361022364217, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 10638 + }, + { + "epoch": 16.99520766773163, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10639 + }, + { + "epoch": 16.996805111821086, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10640 + }, + { + "epoch": 16.998402555910545, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10641 + }, + { + "epoch": 17.0, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10642 + }, + { + "epoch": 17.001597444089455, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10643 + }, + { + "epoch": 17.003194888178914, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10644 + }, + { + "epoch": 17.00479233226837, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10645 + }, + { + "epoch": 17.00638977635783, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10646 + }, + { + "epoch": 17.007987220447284, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10647 + }, + { + "epoch": 17.009584664536742, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10648 + }, + { + "epoch": 17.011182108626198, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10649 + }, + { + "epoch": 17.012779552715656, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10650 + }, + { + "epoch": 17.01437699680511, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10651 + }, + { + "epoch": 17.015974440894567, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 10652 + }, + { + "epoch": 17.017571884984026, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10653 + }, + { + "epoch": 17.01916932907348, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10654 + }, + { + "epoch": 17.02076677316294, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10655 + }, + { + "epoch": 17.022364217252395, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 10656 + }, + { + "epoch": 17.023961661341854, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10657 + }, + { + "epoch": 17.02555910543131, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10658 + }, + { + "epoch": 17.027156549520768, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10659 + }, + { + "epoch": 17.028753993610223, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10660 + }, + { + "epoch": 17.03035143769968, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10661 + }, + { + "epoch": 17.031948881789138, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10662 + }, + { + "epoch": 17.033546325878593, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10663 + }, + { + "epoch": 17.03514376996805, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10664 + }, + { + "epoch": 17.036741214057507, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 10665 + }, + { + "epoch": 17.038338658146966, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10666 + }, + { + "epoch": 17.03993610223642, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10667 + }, + { + "epoch": 17.04153354632588, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10668 + }, + { + "epoch": 17.043130990415335, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10669 + }, + { + "epoch": 17.044728434504794, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10670 + }, + { + "epoch": 17.04632587859425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10671 + }, + { + "epoch": 17.047923322683705, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10672 + }, + { + "epoch": 17.049520766773163, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10673 + }, + { + "epoch": 17.05111821086262, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10674 + }, + { + "epoch": 17.052715654952078, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10675 + }, + { + "epoch": 17.054313099041533, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10676 + }, + { + "epoch": 17.05591054313099, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10677 + }, + { + "epoch": 17.057507987220447, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10678 + }, + { + "epoch": 17.059105431309906, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10679 + }, + { + "epoch": 17.06070287539936, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10680 + }, + { + "epoch": 17.062300319488816, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10681 + }, + { + "epoch": 17.063897763578275, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10682 + }, + { + "epoch": 17.06549520766773, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10683 + }, + { + "epoch": 17.06709265175719, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10684 + }, + { + "epoch": 17.068690095846645, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10685 + }, + { + "epoch": 17.070287539936103, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10686 + }, + { + "epoch": 17.07188498402556, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 10687 + }, + { + "epoch": 17.073482428115017, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10688 + }, + { + "epoch": 17.075079872204473, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10689 + }, + { + "epoch": 17.076677316293928, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10690 + }, + { + "epoch": 17.078274760383387, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10691 + }, + { + "epoch": 17.079872204472842, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10692 + }, + { + "epoch": 17.0814696485623, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 10693 + }, + { + "epoch": 17.083067092651756, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 10694 + }, + { + "epoch": 17.084664536741215, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10695 + }, + { + "epoch": 17.08626198083067, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10696 + }, + { + "epoch": 17.08785942492013, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10697 + }, + { + "epoch": 17.089456869009584, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 10698 + }, + { + "epoch": 17.09105431309904, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10699 + }, + { + "epoch": 17.0926517571885, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10700 + }, + { + "epoch": 17.094249201277954, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 10701 + }, + { + "epoch": 17.095846645367413, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 10702 + }, + { + "epoch": 17.097444089456868, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10703 + }, + { + "epoch": 17.099041533546327, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10704 + }, + { + "epoch": 17.100638977635782, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10705 + }, + { + "epoch": 17.10223642172524, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10706 + }, + { + "epoch": 17.103833865814696, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10707 + }, + { + "epoch": 17.105431309904155, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10708 + }, + { + "epoch": 17.10702875399361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10709 + }, + { + "epoch": 17.108626198083066, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10710 + }, + { + "epoch": 17.110223642172524, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10711 + }, + { + "epoch": 17.11182108626198, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10712 + }, + { + "epoch": 17.11341853035144, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10713 + }, + { + "epoch": 17.115015974440894, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10714 + }, + { + "epoch": 17.116613418530353, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10715 + }, + { + "epoch": 17.118210862619808, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10716 + }, + { + "epoch": 17.119808306709267, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10717 + }, + { + "epoch": 17.121405750798722, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10718 + }, + { + "epoch": 17.123003194888177, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10719 + }, + { + "epoch": 17.124600638977636, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10720 + }, + { + "epoch": 17.12619808306709, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10721 + }, + { + "epoch": 17.12779552715655, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10722 + }, + { + "epoch": 17.129392971246006, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10723 + }, + { + "epoch": 17.130990415335464, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10724 + }, + { + "epoch": 17.13258785942492, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10725 + }, + { + "epoch": 17.13418530351438, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 10726 + }, + { + "epoch": 17.135782747603834, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10727 + }, + { + "epoch": 17.13738019169329, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10728 + }, + { + "epoch": 17.138977635782748, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10729 + }, + { + "epoch": 17.140575079872203, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 10730 + }, + { + "epoch": 17.142172523961662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10731 + }, + { + "epoch": 17.143769968051117, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10732 + }, + { + "epoch": 17.145367412140576, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10733 + }, + { + "epoch": 17.14696485623003, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10734 + }, + { + "epoch": 17.14856230031949, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10735 + }, + { + "epoch": 17.150159744408946, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10736 + }, + { + "epoch": 17.1517571884984, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 10737 + }, + { + "epoch": 17.15335463258786, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10738 + }, + { + "epoch": 17.154952076677315, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10739 + }, + { + "epoch": 17.156549520766774, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 10740 + }, + { + "epoch": 17.15814696485623, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10741 + }, + { + "epoch": 17.159744408945688, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 10742 + }, + { + "epoch": 17.161341853035143, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10743 + }, + { + "epoch": 17.162939297124602, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10744 + }, + { + "epoch": 17.164536741214057, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10745 + }, + { + "epoch": 17.166134185303516, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 10746 + }, + { + "epoch": 17.16773162939297, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10747 + }, + { + "epoch": 17.169329073482427, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10748 + }, + { + "epoch": 17.170926517571885, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10749 + }, + { + "epoch": 17.17252396166134, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10750 + }, + { + "epoch": 17.1741214057508, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10751 + }, + { + "epoch": 17.175718849840255, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10752 + }, + { + "epoch": 17.177316293929714, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10753 + }, + { + "epoch": 17.17891373801917, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10754 + }, + { + "epoch": 17.180511182108628, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10755 + }, + { + "epoch": 17.182108626198083, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 10756 + }, + { + "epoch": 17.18370607028754, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10757 + }, + { + "epoch": 17.185303514376997, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10758 + }, + { + "epoch": 17.186900958466452, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10759 + }, + { + "epoch": 17.18849840255591, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10760 + }, + { + "epoch": 17.190095846645367, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10761 + }, + { + "epoch": 17.191693290734825, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10762 + }, + { + "epoch": 17.19329073482428, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10763 + }, + { + "epoch": 17.19488817891374, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10764 + }, + { + "epoch": 17.196485623003195, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10765 + }, + { + "epoch": 17.19808306709265, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 10766 + }, + { + "epoch": 17.19968051118211, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10767 + }, + { + "epoch": 17.201277955271564, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10768 + }, + { + "epoch": 17.202875399361023, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10769 + }, + { + "epoch": 17.20447284345048, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10770 + }, + { + "epoch": 17.206070287539937, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10771 + }, + { + "epoch": 17.207667731629392, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10772 + }, + { + "epoch": 17.20926517571885, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10773 + }, + { + "epoch": 17.210862619808307, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10774 + }, + { + "epoch": 17.212460063897762, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10775 + }, + { + "epoch": 17.21405750798722, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10776 + }, + { + "epoch": 17.215654952076676, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10777 + }, + { + "epoch": 17.217252396166135, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10778 + }, + { + "epoch": 17.21884984025559, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10779 + }, + { + "epoch": 17.22044728434505, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10780 + }, + { + "epoch": 17.222044728434504, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10781 + }, + { + "epoch": 17.223642172523963, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 10782 + }, + { + "epoch": 17.22523961661342, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10783 + }, + { + "epoch": 17.226837060702877, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 10784 + }, + { + "epoch": 17.228434504792332, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10785 + }, + { + "epoch": 17.230031948881788, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 10786 + }, + { + "epoch": 17.231629392971247, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 10787 + }, + { + "epoch": 17.233226837060702, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10788 + }, + { + "epoch": 17.23482428115016, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 10789 + }, + { + "epoch": 17.236421725239616, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10790 + }, + { + "epoch": 17.238019169329075, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10791 + }, + { + "epoch": 17.23961661341853, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10792 + }, + { + "epoch": 17.24121405750799, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 10793 + }, + { + "epoch": 17.242811501597444, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10794 + }, + { + "epoch": 17.2444089456869, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10795 + }, + { + "epoch": 17.24600638977636, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10796 + }, + { + "epoch": 17.247603833865814, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10797 + }, + { + "epoch": 17.249201277955272, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10798 + }, + { + "epoch": 17.250798722044728, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10799 + }, + { + "epoch": 17.252396166134186, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10800 + }, + { + "epoch": 17.25399361022364, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10801 + }, + { + "epoch": 17.2555910543131, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10802 + }, + { + "epoch": 17.257188498402556, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10803 + }, + { + "epoch": 17.25878594249201, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10804 + }, + { + "epoch": 17.26038338658147, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10805 + }, + { + "epoch": 17.261980830670925, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10806 + }, + { + "epoch": 17.263578274760384, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10807 + }, + { + "epoch": 17.26517571884984, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10808 + }, + { + "epoch": 17.266773162939298, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10809 + }, + { + "epoch": 17.268370607028753, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10810 + }, + { + "epoch": 17.269968051118212, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10811 + }, + { + "epoch": 17.271565495207668, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 10812 + }, + { + "epoch": 17.273162939297123, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10813 + }, + { + "epoch": 17.27476038338658, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10814 + }, + { + "epoch": 17.276357827476037, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10815 + }, + { + "epoch": 17.277955271565496, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10816 + }, + { + "epoch": 17.27955271565495, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10817 + }, + { + "epoch": 17.28115015974441, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10818 + }, + { + "epoch": 17.282747603833865, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10819 + }, + { + "epoch": 17.284345047923324, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10820 + }, + { + "epoch": 17.28594249201278, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10821 + }, + { + "epoch": 17.287539936102238, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10822 + }, + { + "epoch": 17.289137380191693, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10823 + }, + { + "epoch": 17.29073482428115, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 10824 + }, + { + "epoch": 17.292332268370608, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 10825 + }, + { + "epoch": 17.293929712460063, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10826 + }, + { + "epoch": 17.29552715654952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 10827 + }, + { + "epoch": 17.297124600638977, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 10828 + }, + { + "epoch": 17.298722044728436, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 10829 + }, + { + "epoch": 17.30031948881789, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 10830 + }, + { + "epoch": 17.30191693290735, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10831 + }, + { + "epoch": 17.303514376996805, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10832 + }, + { + "epoch": 17.30511182108626, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10833 + }, + { + "epoch": 17.30670926517572, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10834 + }, + { + "epoch": 17.308306709265175, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10835 + }, + { + "epoch": 17.309904153354633, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10836 + }, + { + "epoch": 17.31150159744409, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10837 + }, + { + "epoch": 17.313099041533548, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10838 + }, + { + "epoch": 17.314696485623003, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 10839 + }, + { + "epoch": 17.31629392971246, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10840 + }, + { + "epoch": 17.317891373801917, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 10841 + }, + { + "epoch": 17.319488817891372, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10842 + }, + { + "epoch": 17.32108626198083, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10843 + }, + { + "epoch": 17.322683706070286, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10844 + }, + { + "epoch": 17.324281150159745, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10845 + }, + { + "epoch": 17.3258785942492, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10846 + }, + { + "epoch": 17.32747603833866, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10847 + }, + { + "epoch": 17.329073482428115, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10848 + }, + { + "epoch": 17.330670926517573, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10849 + }, + { + "epoch": 17.33226837060703, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10850 + }, + { + "epoch": 17.333865814696484, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10851 + }, + { + "epoch": 17.335463258785943, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10852 + }, + { + "epoch": 17.337060702875398, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10853 + }, + { + "epoch": 17.338658146964857, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10854 + }, + { + "epoch": 17.340255591054312, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 10855 + }, + { + "epoch": 17.34185303514377, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10856 + }, + { + "epoch": 17.343450479233226, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10857 + }, + { + "epoch": 17.345047923322685, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 10858 + }, + { + "epoch": 17.34664536741214, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10859 + }, + { + "epoch": 17.3482428115016, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 10860 + }, + { + "epoch": 17.349840255591054, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10861 + }, + { + "epoch": 17.35143769968051, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10862 + }, + { + "epoch": 17.35303514376997, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10863 + }, + { + "epoch": 17.354632587859424, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10864 + }, + { + "epoch": 17.356230031948883, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10865 + }, + { + "epoch": 17.357827476038338, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10866 + }, + { + "epoch": 17.359424920127797, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10867 + }, + { + "epoch": 17.361022364217252, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10868 + }, + { + "epoch": 17.36261980830671, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10869 + }, + { + "epoch": 17.364217252396166, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10870 + }, + { + "epoch": 17.36581469648562, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 10871 + }, + { + "epoch": 17.36741214057508, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10872 + }, + { + "epoch": 17.369009584664536, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10873 + }, + { + "epoch": 17.370607028753994, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10874 + }, + { + "epoch": 17.37220447284345, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10875 + }, + { + "epoch": 17.37380191693291, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10876 + }, + { + "epoch": 17.375399361022364, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10877 + }, + { + "epoch": 17.376996805111823, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10878 + }, + { + "epoch": 17.378594249201278, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10879 + }, + { + "epoch": 17.380191693290733, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10880 + }, + { + "epoch": 17.381789137380192, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10881 + }, + { + "epoch": 17.383386581469647, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 10882 + }, + { + "epoch": 17.384984025559106, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10883 + }, + { + "epoch": 17.38658146964856, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10884 + }, + { + "epoch": 17.38817891373802, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10885 + }, + { + "epoch": 17.389776357827476, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10886 + }, + { + "epoch": 17.391373801916934, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10887 + }, + { + "epoch": 17.39297124600639, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10888 + }, + { + "epoch": 17.394568690095845, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10889 + }, + { + "epoch": 17.396166134185304, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10890 + }, + { + "epoch": 17.39776357827476, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 10891 + }, + { + "epoch": 17.399361022364218, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 10892 + }, + { + "epoch": 17.400958466453673, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10893 + }, + { + "epoch": 17.402555910543132, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 10894 + }, + { + "epoch": 17.404153354632587, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10895 + }, + { + "epoch": 17.405750798722046, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10896 + }, + { + "epoch": 17.4073482428115, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10897 + }, + { + "epoch": 17.408945686900957, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 10898 + }, + { + "epoch": 17.410543130990416, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10899 + }, + { + "epoch": 17.41214057507987, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10900 + }, + { + "epoch": 17.41373801916933, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 10901 + }, + { + "epoch": 17.415335463258785, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10902 + }, + { + "epoch": 17.416932907348244, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 10903 + }, + { + "epoch": 17.4185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10904 + }, + { + "epoch": 17.420127795527158, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10905 + }, + { + "epoch": 17.421725239616613, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 10906 + }, + { + "epoch": 17.423322683706072, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10907 + }, + { + "epoch": 17.424920127795527, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10908 + }, + { + "epoch": 17.426517571884983, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10909 + }, + { + "epoch": 17.42811501597444, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10910 + }, + { + "epoch": 17.429712460063897, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10911 + }, + { + "epoch": 17.431309904153355, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10912 + }, + { + "epoch": 17.43290734824281, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10913 + }, + { + "epoch": 17.43450479233227, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 10914 + }, + { + "epoch": 17.436102236421725, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10915 + }, + { + "epoch": 17.437699680511184, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10916 + }, + { + "epoch": 17.43929712460064, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10917 + }, + { + "epoch": 17.440894568690094, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10918 + }, + { + "epoch": 17.442492012779553, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10919 + }, + { + "epoch": 17.44408945686901, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10920 + }, + { + "epoch": 17.445686900958467, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10921 + }, + { + "epoch": 17.447284345047922, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10922 + }, + { + "epoch": 17.44888178913738, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10923 + }, + { + "epoch": 17.450479233226837, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10924 + }, + { + "epoch": 17.452076677316295, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10925 + }, + { + "epoch": 17.45367412140575, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10926 + }, + { + "epoch": 17.455271565495206, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10927 + }, + { + "epoch": 17.456869009584665, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10928 + }, + { + "epoch": 17.45846645367412, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10929 + }, + { + "epoch": 17.46006389776358, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10930 + }, + { + "epoch": 17.461661341853034, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 10931 + }, + { + "epoch": 17.463258785942493, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10932 + }, + { + "epoch": 17.46485623003195, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10933 + }, + { + "epoch": 17.466453674121407, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10934 + }, + { + "epoch": 17.468051118210862, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10935 + }, + { + "epoch": 17.46964856230032, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10936 + }, + { + "epoch": 17.471246006389777, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 10937 + }, + { + "epoch": 17.472843450479232, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10938 + }, + { + "epoch": 17.47444089456869, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10939 + }, + { + "epoch": 17.476038338658146, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10940 + }, + { + "epoch": 17.477635782747605, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 10941 + }, + { + "epoch": 17.47923322683706, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10942 + }, + { + "epoch": 17.48083067092652, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10943 + }, + { + "epoch": 17.482428115015974, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10944 + }, + { + "epoch": 17.484025559105433, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10945 + }, + { + "epoch": 17.48562300319489, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 10946 + }, + { + "epoch": 17.487220447284344, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 10947 + }, + { + "epoch": 17.488817891373802, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10948 + }, + { + "epoch": 17.490415335463258, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10949 + }, + { + "epoch": 17.492012779552716, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10950 + }, + { + "epoch": 17.49361022364217, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10951 + }, + { + "epoch": 17.49520766773163, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10952 + }, + { + "epoch": 17.496805111821086, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10953 + }, + { + "epoch": 17.498402555910545, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10954 + }, + { + "epoch": 17.5, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10955 + }, + { + "epoch": 17.501597444089455, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10956 + }, + { + "epoch": 17.503194888178914, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10957 + }, + { + "epoch": 17.50479233226837, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10958 + }, + { + "epoch": 17.50638977635783, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10959 + }, + { + "epoch": 17.507987220447284, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10960 + }, + { + "epoch": 17.509584664536742, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10961 + }, + { + "epoch": 17.511182108626198, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10962 + }, + { + "epoch": 17.512779552715656, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10963 + }, + { + "epoch": 17.51437699680511, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10964 + }, + { + "epoch": 17.515974440894567, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10965 + }, + { + "epoch": 17.517571884984026, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 10966 + }, + { + "epoch": 17.51916932907348, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10967 + }, + { + "epoch": 17.52076677316294, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10968 + }, + { + "epoch": 17.522364217252395, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 10969 + }, + { + "epoch": 17.523961661341854, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10970 + }, + { + "epoch": 17.52555910543131, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 10971 + }, + { + "epoch": 17.527156549520768, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 10972 + }, + { + "epoch": 17.528753993610223, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10973 + }, + { + "epoch": 17.53035143769968, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10974 + }, + { + "epoch": 17.531948881789138, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10975 + }, + { + "epoch": 17.533546325878593, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10976 + }, + { + "epoch": 17.53514376996805, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10977 + }, + { + "epoch": 17.536741214057507, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10978 + }, + { + "epoch": 17.538338658146966, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10979 + }, + { + "epoch": 17.53993610223642, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 10980 + }, + { + "epoch": 17.54153354632588, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10981 + }, + { + "epoch": 17.543130990415335, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10982 + }, + { + "epoch": 17.544728434504794, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10983 + }, + { + "epoch": 17.54632587859425, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10984 + }, + { + "epoch": 17.547923322683705, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10985 + }, + { + "epoch": 17.549520766773163, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10986 + }, + { + "epoch": 17.55111821086262, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 10987 + }, + { + "epoch": 17.552715654952078, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10988 + }, + { + "epoch": 17.554313099041533, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10989 + }, + { + "epoch": 17.55591054313099, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 10990 + }, + { + "epoch": 17.557507987220447, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10991 + }, + { + "epoch": 17.559105431309906, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10992 + }, + { + "epoch": 17.56070287539936, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10993 + }, + { + "epoch": 17.562300319488816, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 10994 + }, + { + "epoch": 17.563897763578275, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10995 + }, + { + "epoch": 17.56549520766773, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10996 + }, + { + "epoch": 17.56709265175719, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10997 + }, + { + "epoch": 17.568690095846645, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10998 + }, + { + "epoch": 17.570287539936103, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10999 + }, + { + "epoch": 17.57188498402556, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 11000 + }, + { + "epoch": 17.573482428115017, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11001 + }, + { + "epoch": 17.575079872204473, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11002 + }, + { + "epoch": 17.576677316293928, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11003 + }, + { + "epoch": 17.578274760383387, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11004 + }, + { + "epoch": 17.579872204472842, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11005 + }, + { + "epoch": 17.5814696485623, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11006 + }, + { + "epoch": 17.583067092651756, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11007 + }, + { + "epoch": 17.584664536741215, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11008 + }, + { + "epoch": 17.58626198083067, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 11009 + }, + { + "epoch": 17.58785942492013, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11010 + }, + { + "epoch": 17.589456869009584, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11011 + }, + { + "epoch": 17.591054313099043, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11012 + }, + { + "epoch": 17.5926517571885, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11013 + }, + { + "epoch": 17.594249201277954, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11014 + }, + { + "epoch": 17.595846645367413, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11015 + }, + { + "epoch": 17.597444089456868, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11016 + }, + { + "epoch": 17.599041533546327, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 11017 + }, + { + "epoch": 17.600638977635782, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 11018 + }, + { + "epoch": 17.60223642172524, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11019 + }, + { + "epoch": 17.603833865814696, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11020 + }, + { + "epoch": 17.605431309904155, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11021 + }, + { + "epoch": 17.60702875399361, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11022 + }, + { + "epoch": 17.608626198083066, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11023 + }, + { + "epoch": 17.610223642172524, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11024 + }, + { + "epoch": 17.61182108626198, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11025 + }, + { + "epoch": 17.61341853035144, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11026 + }, + { + "epoch": 17.615015974440894, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11027 + }, + { + "epoch": 17.616613418530353, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11028 + }, + { + "epoch": 17.618210862619808, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11029 + }, + { + "epoch": 17.619808306709267, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11030 + }, + { + "epoch": 17.621405750798722, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11031 + }, + { + "epoch": 17.623003194888177, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11032 + }, + { + "epoch": 17.624600638977636, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11033 + }, + { + "epoch": 17.62619808306709, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11034 + }, + { + "epoch": 17.62779552715655, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11035 + }, + { + "epoch": 17.629392971246006, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11036 + }, + { + "epoch": 17.630990415335464, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 11037 + }, + { + "epoch": 17.63258785942492, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11038 + }, + { + "epoch": 17.63418530351438, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11039 + }, + { + "epoch": 17.635782747603834, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 11040 + }, + { + "epoch": 17.63738019169329, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11041 + }, + { + "epoch": 17.638977635782748, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11042 + }, + { + "epoch": 17.640575079872203, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 11043 + }, + { + "epoch": 17.642172523961662, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11044 + }, + { + "epoch": 17.643769968051117, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11045 + }, + { + "epoch": 17.645367412140576, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11046 + }, + { + "epoch": 17.64696485623003, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11047 + }, + { + "epoch": 17.64856230031949, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11048 + }, + { + "epoch": 17.650159744408946, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11049 + }, + { + "epoch": 17.6517571884984, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11050 + }, + { + "epoch": 17.65335463258786, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11051 + }, + { + "epoch": 17.654952076677315, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11052 + }, + { + "epoch": 17.656549520766774, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11053 + }, + { + "epoch": 17.65814696485623, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11054 + }, + { + "epoch": 17.659744408945688, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11055 + }, + { + "epoch": 17.661341853035143, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11056 + }, + { + "epoch": 17.662939297124602, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11057 + }, + { + "epoch": 17.664536741214057, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11058 + }, + { + "epoch": 17.666134185303516, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11059 + }, + { + "epoch": 17.66773162939297, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11060 + }, + { + "epoch": 17.669329073482427, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11061 + }, + { + "epoch": 17.670926517571885, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11062 + }, + { + "epoch": 17.67252396166134, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11063 + }, + { + "epoch": 17.6741214057508, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 11064 + }, + { + "epoch": 17.675718849840255, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11065 + }, + { + "epoch": 17.677316293929714, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11066 + }, + { + "epoch": 17.67891373801917, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11067 + }, + { + "epoch": 17.680511182108628, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11068 + }, + { + "epoch": 17.682108626198083, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11069 + }, + { + "epoch": 17.68370607028754, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11070 + }, + { + "epoch": 17.685303514376997, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11071 + }, + { + "epoch": 17.686900958466452, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11072 + }, + { + "epoch": 17.68849840255591, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11073 + }, + { + "epoch": 17.690095846645367, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11074 + }, + { + "epoch": 17.691693290734825, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11075 + }, + { + "epoch": 17.69329073482428, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11076 + }, + { + "epoch": 17.69488817891374, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11077 + }, + { + "epoch": 17.696485623003195, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11078 + }, + { + "epoch": 17.69808306709265, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 11079 + }, + { + "epoch": 17.69968051118211, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 11080 + }, + { + "epoch": 17.701277955271564, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11081 + }, + { + "epoch": 17.702875399361023, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11082 + }, + { + "epoch": 17.70447284345048, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11083 + }, + { + "epoch": 17.706070287539937, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 11084 + }, + { + "epoch": 17.707667731629392, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11085 + }, + { + "epoch": 17.70926517571885, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11086 + }, + { + "epoch": 17.710862619808307, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 11087 + }, + { + "epoch": 17.712460063897765, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11088 + }, + { + "epoch": 17.71405750798722, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 11089 + }, + { + "epoch": 17.715654952076676, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11090 + }, + { + "epoch": 17.717252396166135, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11091 + }, + { + "epoch": 17.71884984025559, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11092 + }, + { + "epoch": 17.72044728434505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11093 + }, + { + "epoch": 17.722044728434504, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11094 + }, + { + "epoch": 17.723642172523963, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 11095 + }, + { + "epoch": 17.72523961661342, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11096 + }, + { + "epoch": 17.726837060702877, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11097 + }, + { + "epoch": 17.728434504792332, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11098 + }, + { + "epoch": 17.730031948881788, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11099 + }, + { + "epoch": 17.731629392971247, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11100 + }, + { + "epoch": 17.733226837060702, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11101 + }, + { + "epoch": 17.73482428115016, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11102 + }, + { + "epoch": 17.736421725239616, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11103 + }, + { + "epoch": 17.738019169329075, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11104 + }, + { + "epoch": 17.73961661341853, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11105 + }, + { + "epoch": 17.74121405750799, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11106 + }, + { + "epoch": 17.742811501597444, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11107 + }, + { + "epoch": 17.7444089456869, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11108 + }, + { + "epoch": 17.74600638977636, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11109 + }, + { + "epoch": 17.747603833865814, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11110 + }, + { + "epoch": 17.749201277955272, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11111 + }, + { + "epoch": 17.750798722044728, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11112 + }, + { + "epoch": 17.752396166134186, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11113 + }, + { + "epoch": 17.75399361022364, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11114 + }, + { + "epoch": 17.7555910543131, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11115 + }, + { + "epoch": 17.757188498402556, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11116 + }, + { + "epoch": 17.75878594249201, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11117 + }, + { + "epoch": 17.76038338658147, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11118 + }, + { + "epoch": 17.761980830670925, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11119 + }, + { + "epoch": 17.763578274760384, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11120 + }, + { + "epoch": 17.76517571884984, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11121 + }, + { + "epoch": 17.766773162939298, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11122 + }, + { + "epoch": 17.768370607028753, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11123 + }, + { + "epoch": 17.769968051118212, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 11124 + }, + { + "epoch": 17.771565495207668, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11125 + }, + { + "epoch": 17.773162939297123, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 11126 + }, + { + "epoch": 17.77476038338658, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11127 + }, + { + "epoch": 17.776357827476037, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 11128 + }, + { + "epoch": 17.777955271565496, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11129 + }, + { + "epoch": 17.77955271565495, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11130 + }, + { + "epoch": 17.78115015974441, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11131 + }, + { + "epoch": 17.782747603833865, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11132 + }, + { + "epoch": 17.784345047923324, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 11133 + }, + { + "epoch": 17.78594249201278, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11134 + }, + { + "epoch": 17.787539936102235, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11135 + }, + { + "epoch": 17.789137380191693, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11136 + }, + { + "epoch": 17.79073482428115, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11137 + }, + { + "epoch": 17.792332268370608, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11138 + }, + { + "epoch": 17.793929712460063, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11139 + }, + { + "epoch": 17.79552715654952, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11140 + }, + { + "epoch": 17.797124600638977, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11141 + }, + { + "epoch": 17.798722044728436, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11142 + }, + { + "epoch": 17.80031948881789, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11143 + }, + { + "epoch": 17.80191693290735, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11144 + }, + { + "epoch": 17.803514376996805, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11145 + }, + { + "epoch": 17.80511182108626, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11146 + }, + { + "epoch": 17.80670926517572, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11147 + }, + { + "epoch": 17.808306709265175, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 11148 + }, + { + "epoch": 17.809904153354633, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11149 + }, + { + "epoch": 17.81150159744409, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11150 + }, + { + "epoch": 17.813099041533548, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 11151 + }, + { + "epoch": 17.814696485623003, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 11152 + }, + { + "epoch": 17.81629392971246, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11153 + }, + { + "epoch": 17.817891373801917, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11154 + }, + { + "epoch": 17.819488817891372, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11155 + }, + { + "epoch": 17.82108626198083, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11156 + }, + { + "epoch": 17.822683706070286, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11157 + }, + { + "epoch": 17.824281150159745, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11158 + }, + { + "epoch": 17.8258785942492, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11159 + }, + { + "epoch": 17.82747603833866, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11160 + }, + { + "epoch": 17.829073482428115, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11161 + }, + { + "epoch": 17.830670926517573, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 11162 + }, + { + "epoch": 17.83226837060703, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11163 + }, + { + "epoch": 17.833865814696484, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11164 + }, + { + "epoch": 17.835463258785943, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 11165 + }, + { + "epoch": 17.837060702875398, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11166 + }, + { + "epoch": 17.838658146964857, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11167 + }, + { + "epoch": 17.840255591054312, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11168 + }, + { + "epoch": 17.84185303514377, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11169 + }, + { + "epoch": 17.843450479233226, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11170 + }, + { + "epoch": 17.845047923322685, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11171 + }, + { + "epoch": 17.84664536741214, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11172 + }, + { + "epoch": 17.8482428115016, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11173 + }, + { + "epoch": 17.849840255591054, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11174 + }, + { + "epoch": 17.85143769968051, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 11175 + }, + { + "epoch": 17.85303514376997, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11176 + }, + { + "epoch": 17.854632587859424, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11177 + }, + { + "epoch": 17.856230031948883, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11178 + }, + { + "epoch": 17.857827476038338, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11179 + }, + { + "epoch": 17.859424920127797, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11180 + }, + { + "epoch": 17.861022364217252, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11181 + }, + { + "epoch": 17.86261980830671, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11182 + }, + { + "epoch": 17.864217252396166, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11183 + }, + { + "epoch": 17.86581469648562, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 11184 + }, + { + "epoch": 17.86741214057508, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11185 + }, + { + "epoch": 17.869009584664536, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11186 + }, + { + "epoch": 17.870607028753994, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11187 + }, + { + "epoch": 17.87220447284345, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11188 + }, + { + "epoch": 17.87380191693291, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11189 + }, + { + "epoch": 17.875399361022364, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11190 + }, + { + "epoch": 17.876996805111823, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11191 + }, + { + "epoch": 17.878594249201278, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11192 + }, + { + "epoch": 17.880191693290733, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11193 + }, + { + "epoch": 17.881789137380192, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11194 + }, + { + "epoch": 17.883386581469647, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11195 + }, + { + "epoch": 17.884984025559106, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 11196 + }, + { + "epoch": 17.88658146964856, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11197 + }, + { + "epoch": 17.88817891373802, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11198 + }, + { + "epoch": 17.889776357827476, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11199 + }, + { + "epoch": 17.891373801916934, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 11200 + }, + { + "epoch": 17.89297124600639, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11201 + }, + { + "epoch": 17.894568690095845, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11202 + }, + { + "epoch": 17.896166134185304, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11203 + }, + { + "epoch": 17.89776357827476, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11204 + }, + { + "epoch": 17.899361022364218, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11205 + }, + { + "epoch": 17.900958466453673, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11206 + }, + { + "epoch": 17.902555910543132, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11207 + }, + { + "epoch": 17.904153354632587, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11208 + }, + { + "epoch": 17.905750798722046, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11209 + }, + { + "epoch": 17.9073482428115, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11210 + }, + { + "epoch": 17.908945686900957, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11211 + }, + { + "epoch": 17.910543130990416, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11212 + }, + { + "epoch": 17.91214057507987, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11213 + }, + { + "epoch": 17.91373801916933, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11214 + }, + { + "epoch": 17.915335463258785, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11215 + }, + { + "epoch": 17.916932907348244, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11216 + }, + { + "epoch": 17.9185303514377, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11217 + }, + { + "epoch": 17.920127795527158, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11218 + }, + { + "epoch": 17.921725239616613, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11219 + }, + { + "epoch": 17.923322683706072, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 11220 + }, + { + "epoch": 17.924920127795527, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 11221 + }, + { + "epoch": 17.926517571884983, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11222 + }, + { + "epoch": 17.92811501597444, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11223 + }, + { + "epoch": 17.929712460063897, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 11224 + }, + { + "epoch": 17.931309904153355, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11225 + }, + { + "epoch": 17.93290734824281, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11226 + }, + { + "epoch": 17.93450479233227, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11227 + }, + { + "epoch": 17.936102236421725, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11228 + }, + { + "epoch": 17.937699680511184, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11229 + }, + { + "epoch": 17.93929712460064, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11230 + }, + { + "epoch": 17.940894568690094, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11231 + }, + { + "epoch": 17.942492012779553, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 11232 + }, + { + "epoch": 17.94408945686901, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11233 + }, + { + "epoch": 17.945686900958467, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11234 + }, + { + "epoch": 17.947284345047922, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11235 + }, + { + "epoch": 17.94888178913738, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11236 + }, + { + "epoch": 17.950479233226837, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11237 + }, + { + "epoch": 17.952076677316295, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11238 + }, + { + "epoch": 17.95367412140575, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 11239 + }, + { + "epoch": 17.955271565495206, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11240 + }, + { + "epoch": 17.956869009584665, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11241 + }, + { + "epoch": 17.95846645367412, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 11242 + }, + { + "epoch": 17.96006389776358, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11243 + }, + { + "epoch": 17.961661341853034, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11244 + }, + { + "epoch": 17.963258785942493, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11245 + }, + { + "epoch": 17.96485623003195, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11246 + }, + { + "epoch": 17.966453674121407, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 11247 + }, + { + "epoch": 17.968051118210862, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 11248 + }, + { + "epoch": 17.96964856230032, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11249 + }, + { + "epoch": 17.971246006389777, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11250 + }, + { + "epoch": 17.972843450479232, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11251 + }, + { + "epoch": 17.97444089456869, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11252 + }, + { + "epoch": 17.976038338658146, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11253 + }, + { + "epoch": 17.977635782747605, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11254 + }, + { + "epoch": 17.97923322683706, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11255 + }, + { + "epoch": 17.98083067092652, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11256 + }, + { + "epoch": 17.982428115015974, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11257 + }, + { + "epoch": 17.984025559105433, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11258 + }, + { + "epoch": 17.98562300319489, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11259 + }, + { + "epoch": 17.987220447284344, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11260 + }, + { + "epoch": 17.988817891373802, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11261 + }, + { + "epoch": 17.990415335463258, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11262 + }, + { + "epoch": 17.992012779552716, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11263 + }, + { + "epoch": 17.99361022364217, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11264 + }, + { + "epoch": 17.99520766773163, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11265 + }, + { + "epoch": 17.996805111821086, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 11266 + }, + { + "epoch": 17.998402555910545, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11267 + }, + { + "epoch": 18.0, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11268 + }, + { + "epoch": 18.001597444089455, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11269 + }, + { + "epoch": 18.003194888178914, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11270 + }, + { + "epoch": 18.00479233226837, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11271 + }, + { + "epoch": 18.00638977635783, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11272 + }, + { + "epoch": 18.007987220447284, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11273 + }, + { + "epoch": 18.009584664536742, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11274 + }, + { + "epoch": 18.011182108626198, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11275 + }, + { + "epoch": 18.012779552715656, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11276 + }, + { + "epoch": 18.01437699680511, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 11277 + }, + { + "epoch": 18.015974440894567, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11278 + }, + { + "epoch": 18.017571884984026, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11279 + }, + { + "epoch": 18.01916932907348, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11280 + }, + { + "epoch": 18.02076677316294, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11281 + }, + { + "epoch": 18.022364217252395, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11282 + }, + { + "epoch": 18.023961661341854, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11283 + }, + { + "epoch": 18.02555910543131, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11284 + }, + { + "epoch": 18.027156549520768, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11285 + }, + { + "epoch": 18.028753993610223, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11286 + }, + { + "epoch": 18.03035143769968, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11287 + }, + { + "epoch": 18.031948881789138, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11288 + }, + { + "epoch": 18.033546325878593, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11289 + }, + { + "epoch": 18.03514376996805, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11290 + }, + { + "epoch": 18.036741214057507, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11291 + }, + { + "epoch": 18.038338658146966, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11292 + }, + { + "epoch": 18.03993610223642, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11293 + }, + { + "epoch": 18.04153354632588, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11294 + }, + { + "epoch": 18.043130990415335, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11295 + }, + { + "epoch": 18.044728434504794, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11296 + }, + { + "epoch": 18.04632587859425, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11297 + }, + { + "epoch": 18.047923322683705, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11298 + }, + { + "epoch": 18.049520766773163, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11299 + }, + { + "epoch": 18.05111821086262, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11300 + }, + { + "epoch": 18.052715654952078, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11301 + }, + { + "epoch": 18.054313099041533, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11302 + }, + { + "epoch": 18.05591054313099, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11303 + }, + { + "epoch": 18.057507987220447, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11304 + }, + { + "epoch": 18.059105431309906, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11305 + }, + { + "epoch": 18.06070287539936, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11306 + }, + { + "epoch": 18.062300319488816, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11307 + }, + { + "epoch": 18.063897763578275, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11308 + }, + { + "epoch": 18.06549520766773, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11309 + }, + { + "epoch": 18.06709265175719, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11310 + }, + { + "epoch": 18.068690095846645, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11311 + }, + { + "epoch": 18.070287539936103, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11312 + }, + { + "epoch": 18.07188498402556, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11313 + }, + { + "epoch": 18.073482428115017, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11314 + }, + { + "epoch": 18.075079872204473, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11315 + }, + { + "epoch": 18.076677316293928, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11316 + }, + { + "epoch": 18.078274760383387, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11317 + }, + { + "epoch": 18.079872204472842, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11318 + }, + { + "epoch": 18.0814696485623, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11319 + }, + { + "epoch": 18.083067092651756, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11320 + }, + { + "epoch": 18.084664536741215, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11321 + }, + { + "epoch": 18.08626198083067, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11322 + }, + { + "epoch": 18.08785942492013, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11323 + }, + { + "epoch": 18.089456869009584, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11324 + }, + { + "epoch": 18.09105431309904, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11325 + }, + { + "epoch": 18.0926517571885, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11326 + }, + { + "epoch": 18.094249201277954, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11327 + }, + { + "epoch": 18.095846645367413, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11328 + }, + { + "epoch": 18.097444089456868, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11329 + }, + { + "epoch": 18.099041533546327, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11330 + }, + { + "epoch": 18.100638977635782, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11331 + }, + { + "epoch": 18.10223642172524, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11332 + }, + { + "epoch": 18.103833865814696, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11333 + }, + { + "epoch": 18.105431309904155, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11334 + }, + { + "epoch": 18.10702875399361, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11335 + }, + { + "epoch": 18.108626198083066, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11336 + }, + { + "epoch": 18.110223642172524, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11337 + }, + { + "epoch": 18.11182108626198, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11338 + }, + { + "epoch": 18.11341853035144, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11339 + }, + { + "epoch": 18.115015974440894, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11340 + }, + { + "epoch": 18.116613418530353, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11341 + }, + { + "epoch": 18.118210862619808, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11342 + }, + { + "epoch": 18.119808306709267, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11343 + }, + { + "epoch": 18.121405750798722, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11344 + }, + { + "epoch": 18.123003194888177, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 11345 + }, + { + "epoch": 18.124600638977636, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11346 + }, + { + "epoch": 18.12619808306709, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11347 + }, + { + "epoch": 18.12779552715655, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11348 + }, + { + "epoch": 18.129392971246006, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11349 + }, + { + "epoch": 18.130990415335464, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11350 + }, + { + "epoch": 18.13258785942492, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11351 + }, + { + "epoch": 18.13418530351438, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11352 + }, + { + "epoch": 18.135782747603834, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11353 + }, + { + "epoch": 18.13738019169329, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11354 + }, + { + "epoch": 18.138977635782748, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 11355 + }, + { + "epoch": 18.140575079872203, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11356 + }, + { + "epoch": 18.142172523961662, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11357 + }, + { + "epoch": 18.143769968051117, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 11358 + }, + { + "epoch": 18.145367412140576, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11359 + }, + { + "epoch": 18.14696485623003, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11360 + }, + { + "epoch": 18.14856230031949, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11361 + }, + { + "epoch": 18.150159744408946, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11362 + }, + { + "epoch": 18.1517571884984, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11363 + }, + { + "epoch": 18.15335463258786, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11364 + }, + { + "epoch": 18.154952076677315, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 11365 + }, + { + "epoch": 18.156549520766774, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11366 + }, + { + "epoch": 18.15814696485623, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11367 + }, + { + "epoch": 18.159744408945688, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11368 + }, + { + "epoch": 18.161341853035143, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11369 + }, + { + "epoch": 18.162939297124602, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 11370 + }, + { + "epoch": 18.164536741214057, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11371 + }, + { + "epoch": 18.166134185303516, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11372 + }, + { + "epoch": 18.16773162939297, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11373 + }, + { + "epoch": 18.169329073482427, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11374 + }, + { + "epoch": 18.170926517571885, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11375 + }, + { + "epoch": 18.17252396166134, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11376 + }, + { + "epoch": 18.1741214057508, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11377 + }, + { + "epoch": 18.175718849840255, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11378 + }, + { + "epoch": 18.177316293929714, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11379 + }, + { + "epoch": 18.17891373801917, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11380 + }, + { + "epoch": 18.180511182108628, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11381 + }, + { + "epoch": 18.182108626198083, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11382 + }, + { + "epoch": 18.18370607028754, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11383 + }, + { + "epoch": 18.185303514376997, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11384 + }, + { + "epoch": 18.186900958466452, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11385 + }, + { + "epoch": 18.18849840255591, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11386 + }, + { + "epoch": 18.190095846645367, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 11387 + }, + { + "epoch": 18.191693290734825, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11388 + }, + { + "epoch": 18.19329073482428, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11389 + }, + { + "epoch": 18.19488817891374, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11390 + }, + { + "epoch": 18.196485623003195, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11391 + }, + { + "epoch": 18.19808306709265, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11392 + }, + { + "epoch": 18.19968051118211, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11393 + }, + { + "epoch": 18.201277955271564, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11394 + }, + { + "epoch": 18.202875399361023, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11395 + }, + { + "epoch": 18.20447284345048, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 11396 + }, + { + "epoch": 18.206070287539937, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11397 + }, + { + "epoch": 18.207667731629392, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11398 + }, + { + "epoch": 18.20926517571885, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11399 + }, + { + "epoch": 18.210862619808307, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11400 + }, + { + "epoch": 18.212460063897762, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11401 + }, + { + "epoch": 18.21405750798722, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11402 + }, + { + "epoch": 18.215654952076676, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 11403 + }, + { + "epoch": 18.217252396166135, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11404 + }, + { + "epoch": 18.21884984025559, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11405 + }, + { + "epoch": 18.22044728434505, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11406 + }, + { + "epoch": 18.222044728434504, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11407 + }, + { + "epoch": 18.223642172523963, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11408 + }, + { + "epoch": 18.22523961661342, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11409 + }, + { + "epoch": 18.226837060702877, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 11410 + }, + { + "epoch": 18.228434504792332, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11411 + }, + { + "epoch": 18.230031948881788, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11412 + }, + { + "epoch": 18.231629392971247, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11413 + }, + { + "epoch": 18.233226837060702, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11414 + }, + { + "epoch": 18.23482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11415 + }, + { + "epoch": 18.236421725239616, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11416 + }, + { + "epoch": 18.238019169329075, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11417 + }, + { + "epoch": 18.23961661341853, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11418 + }, + { + "epoch": 18.24121405750799, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11419 + }, + { + "epoch": 18.242811501597444, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11420 + }, + { + "epoch": 18.2444089456869, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11421 + }, + { + "epoch": 18.24600638977636, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11422 + }, + { + "epoch": 18.247603833865814, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11423 + }, + { + "epoch": 18.249201277955272, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11424 + }, + { + "epoch": 18.250798722044728, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11425 + }, + { + "epoch": 18.252396166134186, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 11426 + }, + { + "epoch": 18.25399361022364, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11427 + }, + { + "epoch": 18.2555910543131, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11428 + }, + { + "epoch": 18.257188498402556, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11429 + }, + { + "epoch": 18.25878594249201, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11430 + }, + { + "epoch": 18.26038338658147, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11431 + }, + { + "epoch": 18.261980830670925, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11432 + }, + { + "epoch": 18.263578274760384, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11433 + }, + { + "epoch": 18.26517571884984, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11434 + }, + { + "epoch": 18.266773162939298, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11435 + }, + { + "epoch": 18.268370607028753, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11436 + }, + { + "epoch": 18.269968051118212, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11437 + }, + { + "epoch": 18.271565495207668, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11438 + }, + { + "epoch": 18.273162939297123, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11439 + }, + { + "epoch": 18.27476038338658, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11440 + }, + { + "epoch": 18.276357827476037, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11441 + }, + { + "epoch": 18.277955271565496, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11442 + }, + { + "epoch": 18.27955271565495, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11443 + }, + { + "epoch": 18.28115015974441, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 11444 + }, + { + "epoch": 18.282747603833865, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11445 + }, + { + "epoch": 18.284345047923324, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11446 + }, + { + "epoch": 18.28594249201278, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11447 + }, + { + "epoch": 18.287539936102238, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11448 + }, + { + "epoch": 18.289137380191693, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11449 + }, + { + "epoch": 18.29073482428115, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11450 + }, + { + "epoch": 18.292332268370608, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11451 + }, + { + "epoch": 18.293929712460063, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11452 + }, + { + "epoch": 18.29552715654952, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11453 + }, + { + "epoch": 18.297124600638977, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11454 + }, + { + "epoch": 18.298722044728436, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11455 + }, + { + "epoch": 18.30031948881789, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 11456 + }, + { + "epoch": 18.30191693290735, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11457 + }, + { + "epoch": 18.303514376996805, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11458 + }, + { + "epoch": 18.30511182108626, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11459 + }, + { + "epoch": 18.30670926517572, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11460 + }, + { + "epoch": 18.308306709265175, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11461 + }, + { + "epoch": 18.309904153354633, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 11462 + }, + { + "epoch": 18.31150159744409, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11463 + }, + { + "epoch": 18.313099041533548, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11464 + }, + { + "epoch": 18.314696485623003, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11465 + }, + { + "epoch": 18.31629392971246, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 11466 + }, + { + "epoch": 18.317891373801917, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 11467 + }, + { + "epoch": 18.319488817891372, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11468 + }, + { + "epoch": 18.32108626198083, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11469 + }, + { + "epoch": 18.322683706070286, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11470 + }, + { + "epoch": 18.324281150159745, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11471 + }, + { + "epoch": 18.3258785942492, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11472 + }, + { + "epoch": 18.32747603833866, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11473 + }, + { + "epoch": 18.329073482428115, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11474 + }, + { + "epoch": 18.330670926517573, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11475 + }, + { + "epoch": 18.33226837060703, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11476 + }, + { + "epoch": 18.333865814696484, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11477 + }, + { + "epoch": 18.335463258785943, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 11478 + }, + { + "epoch": 18.337060702875398, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11479 + }, + { + "epoch": 18.338658146964857, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11480 + }, + { + "epoch": 18.340255591054312, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11481 + }, + { + "epoch": 18.34185303514377, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11482 + }, + { + "epoch": 18.343450479233226, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11483 + }, + { + "epoch": 18.345047923322685, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11484 + }, + { + "epoch": 18.34664536741214, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11485 + }, + { + "epoch": 18.3482428115016, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11486 + }, + { + "epoch": 18.349840255591054, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 11487 + }, + { + "epoch": 18.35143769968051, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11488 + }, + { + "epoch": 18.35303514376997, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11489 + }, + { + "epoch": 18.354632587859424, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11490 + }, + { + "epoch": 18.356230031948883, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11491 + }, + { + "epoch": 18.357827476038338, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11492 + }, + { + "epoch": 18.359424920127797, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11493 + }, + { + "epoch": 18.361022364217252, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11494 + }, + { + "epoch": 18.36261980830671, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11495 + }, + { + "epoch": 18.364217252396166, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11496 + }, + { + "epoch": 18.36581469648562, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11497 + }, + { + "epoch": 18.36741214057508, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11498 + }, + { + "epoch": 18.369009584664536, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 11499 + }, + { + "epoch": 18.370607028753994, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11500 + }, + { + "epoch": 18.37220447284345, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11501 + }, + { + "epoch": 18.37380191693291, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 11502 + }, + { + "epoch": 18.375399361022364, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 11503 + }, + { + "epoch": 18.376996805111823, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11504 + }, + { + "epoch": 18.378594249201278, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11505 + }, + { + "epoch": 18.380191693290733, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11506 + }, + { + "epoch": 18.381789137380192, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11507 + }, + { + "epoch": 18.383386581469647, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11508 + }, + { + "epoch": 18.384984025559106, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11509 + }, + { + "epoch": 18.38658146964856, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11510 + }, + { + "epoch": 18.38817891373802, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11511 + }, + { + "epoch": 18.389776357827476, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 11512 + }, + { + "epoch": 18.391373801916934, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11513 + }, + { + "epoch": 18.39297124600639, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11514 + }, + { + "epoch": 18.394568690095845, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11515 + }, + { + "epoch": 18.396166134185304, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11516 + }, + { + "epoch": 18.39776357827476, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11517 + }, + { + "epoch": 18.399361022364218, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11518 + }, + { + "epoch": 18.400958466453673, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11519 + }, + { + "epoch": 18.402555910543132, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 11520 + }, + { + "epoch": 18.404153354632587, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11521 + }, + { + "epoch": 18.405750798722046, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 11522 + }, + { + "epoch": 18.4073482428115, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11523 + }, + { + "epoch": 18.408945686900957, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11524 + }, + { + "epoch": 18.410543130990416, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11525 + }, + { + "epoch": 18.41214057507987, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11526 + }, + { + "epoch": 18.41373801916933, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11527 + }, + { + "epoch": 18.415335463258785, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11528 + }, + { + "epoch": 18.416932907348244, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 11529 + }, + { + "epoch": 18.4185303514377, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11530 + }, + { + "epoch": 18.420127795527158, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11531 + }, + { + "epoch": 18.421725239616613, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 11532 + }, + { + "epoch": 18.423322683706072, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11533 + }, + { + "epoch": 18.424920127795527, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11534 + }, + { + "epoch": 18.426517571884983, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 11535 + }, + { + "epoch": 18.42811501597444, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11536 + }, + { + "epoch": 18.429712460063897, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11537 + }, + { + "epoch": 18.431309904153355, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11538 + }, + { + "epoch": 18.43290734824281, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11539 + }, + { + "epoch": 18.43450479233227, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11540 + }, + { + "epoch": 18.436102236421725, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11541 + }, + { + "epoch": 18.437699680511184, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11542 + }, + { + "epoch": 18.43929712460064, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11543 + }, + { + "epoch": 18.440894568690094, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11544 + }, + { + "epoch": 18.442492012779553, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11545 + }, + { + "epoch": 18.44408945686901, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 11546 + }, + { + "epoch": 18.445686900958467, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 11547 + }, + { + "epoch": 18.447284345047922, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11548 + }, + { + "epoch": 18.44888178913738, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 11549 + }, + { + "epoch": 18.450479233226837, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11550 + }, + { + "epoch": 18.452076677316295, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11551 + }, + { + "epoch": 18.45367412140575, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11552 + }, + { + "epoch": 18.455271565495206, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11553 + }, + { + "epoch": 18.456869009584665, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11554 + }, + { + "epoch": 18.45846645367412, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11555 + }, + { + "epoch": 18.46006389776358, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11556 + }, + { + "epoch": 18.461661341853034, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 11557 + }, + { + "epoch": 18.463258785942493, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11558 + }, + { + "epoch": 18.46485623003195, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11559 + }, + { + "epoch": 18.466453674121407, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11560 + }, + { + "epoch": 18.468051118210862, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11561 + }, + { + "epoch": 18.46964856230032, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11562 + }, + { + "epoch": 18.471246006389777, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11563 + }, + { + "epoch": 18.472843450479232, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11564 + }, + { + "epoch": 18.47444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11565 + }, + { + "epoch": 18.476038338658146, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 11566 + }, + { + "epoch": 18.477635782747605, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11567 + }, + { + "epoch": 18.47923322683706, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11568 + }, + { + "epoch": 18.48083067092652, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11569 + }, + { + "epoch": 18.482428115015974, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 11570 + }, + { + "epoch": 18.484025559105433, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11571 + }, + { + "epoch": 18.48562300319489, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11572 + }, + { + "epoch": 18.487220447284344, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11573 + }, + { + "epoch": 18.488817891373802, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11574 + }, + { + "epoch": 18.490415335463258, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11575 + }, + { + "epoch": 18.492012779552716, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11576 + }, + { + "epoch": 18.49361022364217, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11577 + }, + { + "epoch": 18.49520766773163, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11578 + }, + { + "epoch": 18.496805111821086, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11579 + }, + { + "epoch": 18.498402555910545, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11580 + }, + { + "epoch": 18.5, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11581 + }, + { + "epoch": 18.501597444089455, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11582 + }, + { + "epoch": 18.503194888178914, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11583 + }, + { + "epoch": 18.50479233226837, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11584 + }, + { + "epoch": 18.50638977635783, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11585 + }, + { + "epoch": 18.507987220447284, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11586 + }, + { + "epoch": 18.509584664536742, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11587 + }, + { + "epoch": 18.511182108626198, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11588 + }, + { + "epoch": 18.512779552715656, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11589 + }, + { + "epoch": 18.51437699680511, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11590 + }, + { + "epoch": 18.515974440894567, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11591 + }, + { + "epoch": 18.517571884984026, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11592 + }, + { + "epoch": 18.51916932907348, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11593 + }, + { + "epoch": 18.52076677316294, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11594 + }, + { + "epoch": 18.522364217252395, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11595 + }, + { + "epoch": 18.523961661341854, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11596 + }, + { + "epoch": 18.52555910543131, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11597 + }, + { + "epoch": 18.527156549520768, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11598 + }, + { + "epoch": 18.528753993610223, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11599 + }, + { + "epoch": 18.53035143769968, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11600 + }, + { + "epoch": 18.531948881789138, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11601 + }, + { + "epoch": 18.533546325878593, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11602 + }, + { + "epoch": 18.53514376996805, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11603 + }, + { + "epoch": 18.536741214057507, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11604 + }, + { + "epoch": 18.538338658146966, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11605 + }, + { + "epoch": 18.53993610223642, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 11606 + }, + { + "epoch": 18.54153354632588, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11607 + }, + { + "epoch": 18.543130990415335, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11608 + }, + { + "epoch": 18.544728434504794, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11609 + }, + { + "epoch": 18.54632587859425, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11610 + }, + { + "epoch": 18.547923322683705, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 11611 + }, + { + "epoch": 18.549520766773163, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11612 + }, + { + "epoch": 18.55111821086262, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 11613 + }, + { + "epoch": 18.552715654952078, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11614 + }, + { + "epoch": 18.554313099041533, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11615 + }, + { + "epoch": 18.55591054313099, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11616 + }, + { + "epoch": 18.557507987220447, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11617 + }, + { + "epoch": 18.559105431309906, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11618 + }, + { + "epoch": 18.56070287539936, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11619 + }, + { + "epoch": 18.562300319488816, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11620 + }, + { + "epoch": 18.563897763578275, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 11621 + }, + { + "epoch": 18.56549520766773, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 11622 + }, + { + "epoch": 18.56709265175719, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11623 + }, + { + "epoch": 18.568690095846645, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11624 + }, + { + "epoch": 18.570287539936103, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11625 + }, + { + "epoch": 18.57188498402556, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11626 + }, + { + "epoch": 18.573482428115017, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11627 + }, + { + "epoch": 18.575079872204473, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11628 + }, + { + "epoch": 18.576677316293928, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11629 + }, + { + "epoch": 18.578274760383387, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11630 + }, + { + "epoch": 18.579872204472842, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11631 + }, + { + "epoch": 18.5814696485623, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11632 + }, + { + "epoch": 18.583067092651756, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11633 + }, + { + "epoch": 18.584664536741215, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11634 + }, + { + "epoch": 18.58626198083067, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11635 + }, + { + "epoch": 18.58785942492013, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 11636 + }, + { + "epoch": 18.589456869009584, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11637 + }, + { + "epoch": 18.591054313099043, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 11638 + }, + { + "epoch": 18.5926517571885, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11639 + }, + { + "epoch": 18.594249201277954, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11640 + }, + { + "epoch": 18.595846645367413, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11641 + }, + { + "epoch": 18.597444089456868, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11642 + }, + { + "epoch": 18.599041533546327, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11643 + }, + { + "epoch": 18.600638977635782, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 11644 + }, + { + "epoch": 18.60223642172524, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11645 + }, + { + "epoch": 18.603833865814696, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11646 + }, + { + "epoch": 18.605431309904155, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11647 + }, + { + "epoch": 18.60702875399361, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11648 + }, + { + "epoch": 18.608626198083066, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11649 + }, + { + "epoch": 18.610223642172524, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11650 + }, + { + "epoch": 18.61182108626198, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 11651 + }, + { + "epoch": 18.61341853035144, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11652 + }, + { + "epoch": 18.615015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11653 + }, + { + "epoch": 18.616613418530353, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11654 + }, + { + "epoch": 18.618210862619808, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11655 + }, + { + "epoch": 18.619808306709267, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11656 + }, + { + "epoch": 18.621405750798722, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11657 + }, + { + "epoch": 18.623003194888177, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11658 + }, + { + "epoch": 18.624600638977636, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 11659 + }, + { + "epoch": 18.62619808306709, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11660 + }, + { + "epoch": 18.62779552715655, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11661 + }, + { + "epoch": 18.629392971246006, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11662 + }, + { + "epoch": 18.630990415335464, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11663 + }, + { + "epoch": 18.63258785942492, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 11664 + }, + { + "epoch": 18.63418530351438, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11665 + }, + { + "epoch": 18.635782747603834, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11666 + }, + { + "epoch": 18.63738019169329, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11667 + }, + { + "epoch": 18.638977635782748, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11668 + }, + { + "epoch": 18.640575079872203, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 11669 + }, + { + "epoch": 18.642172523961662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11670 + }, + { + "epoch": 18.643769968051117, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11671 + }, + { + "epoch": 18.645367412140576, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11672 + }, + { + "epoch": 18.64696485623003, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11673 + }, + { + "epoch": 18.64856230031949, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11674 + }, + { + "epoch": 18.650159744408946, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11675 + }, + { + "epoch": 18.6517571884984, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11676 + }, + { + "epoch": 18.65335463258786, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11677 + }, + { + "epoch": 18.654952076677315, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 11678 + }, + { + "epoch": 18.656549520766774, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11679 + }, + { + "epoch": 18.65814696485623, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11680 + }, + { + "epoch": 18.659744408945688, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11681 + }, + { + "epoch": 18.661341853035143, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11682 + }, + { + "epoch": 18.662939297124602, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 11683 + }, + { + "epoch": 18.664536741214057, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11684 + }, + { + "epoch": 18.666134185303516, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11685 + }, + { + "epoch": 18.66773162939297, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11686 + }, + { + "epoch": 18.669329073482427, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11687 + }, + { + "epoch": 18.670926517571885, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11688 + }, + { + "epoch": 18.67252396166134, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11689 + }, + { + "epoch": 18.6741214057508, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11690 + }, + { + "epoch": 18.675718849840255, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11691 + }, + { + "epoch": 18.677316293929714, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11692 + }, + { + "epoch": 18.67891373801917, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11693 + }, + { + "epoch": 18.680511182108628, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11694 + }, + { + "epoch": 18.682108626198083, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11695 + }, + { + "epoch": 18.68370607028754, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11696 + }, + { + "epoch": 18.685303514376997, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11697 + }, + { + "epoch": 18.686900958466452, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11698 + }, + { + "epoch": 18.68849840255591, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11699 + }, + { + "epoch": 18.690095846645367, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11700 + }, + { + "epoch": 18.691693290734825, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11701 + }, + { + "epoch": 18.69329073482428, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11702 + }, + { + "epoch": 18.69488817891374, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 11703 + }, + { + "epoch": 18.696485623003195, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11704 + }, + { + "epoch": 18.69808306709265, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11705 + }, + { + "epoch": 18.69968051118211, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11706 + }, + { + "epoch": 18.701277955271564, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11707 + }, + { + "epoch": 18.702875399361023, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 11708 + }, + { + "epoch": 18.70447284345048, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 11709 + }, + { + "epoch": 18.706070287539937, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11710 + }, + { + "epoch": 18.707667731629392, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11711 + }, + { + "epoch": 18.70926517571885, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11712 + }, + { + "epoch": 18.710862619808307, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11713 + }, + { + "epoch": 18.712460063897765, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 11714 + }, + { + "epoch": 18.71405750798722, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11715 + }, + { + "epoch": 18.715654952076676, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11716 + }, + { + "epoch": 18.717252396166135, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11717 + }, + { + "epoch": 18.71884984025559, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11718 + }, + { + "epoch": 18.72044728434505, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11719 + }, + { + "epoch": 18.722044728434504, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 11720 + }, + { + "epoch": 18.723642172523963, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11721 + }, + { + "epoch": 18.72523961661342, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11722 + }, + { + "epoch": 18.726837060702877, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11723 + }, + { + "epoch": 18.728434504792332, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11724 + }, + { + "epoch": 18.730031948881788, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11725 + }, + { + "epoch": 18.731629392971247, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11726 + }, + { + "epoch": 18.733226837060702, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11727 + }, + { + "epoch": 18.73482428115016, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11728 + }, + { + "epoch": 18.736421725239616, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11729 + }, + { + "epoch": 18.738019169329075, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 11730 + }, + { + "epoch": 18.73961661341853, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11731 + }, + { + "epoch": 18.74121405750799, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11732 + }, + { + "epoch": 18.742811501597444, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11733 + }, + { + "epoch": 18.7444089456869, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11734 + }, + { + "epoch": 18.74600638977636, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11735 + }, + { + "epoch": 18.747603833865814, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11736 + }, + { + "epoch": 18.749201277955272, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11737 + }, + { + "epoch": 18.750798722044728, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11738 + }, + { + "epoch": 18.752396166134186, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11739 + }, + { + "epoch": 18.75399361022364, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11740 + }, + { + "epoch": 18.7555910543131, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11741 + }, + { + "epoch": 18.757188498402556, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11742 + }, + { + "epoch": 18.75878594249201, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11743 + }, + { + "epoch": 18.76038338658147, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11744 + }, + { + "epoch": 18.761980830670925, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11745 + }, + { + "epoch": 18.763578274760384, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11746 + }, + { + "epoch": 18.76517571884984, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11747 + }, + { + "epoch": 18.766773162939298, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11748 + }, + { + "epoch": 18.768370607028753, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11749 + }, + { + "epoch": 18.769968051118212, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11750 + }, + { + "epoch": 18.771565495207668, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11751 + }, + { + "epoch": 18.773162939297123, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11752 + }, + { + "epoch": 18.77476038338658, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11753 + }, + { + "epoch": 18.776357827476037, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11754 + }, + { + "epoch": 18.777955271565496, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11755 + }, + { + "epoch": 18.77955271565495, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11756 + }, + { + "epoch": 18.78115015974441, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11757 + }, + { + "epoch": 18.782747603833865, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 11758 + }, + { + "epoch": 18.784345047923324, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11759 + }, + { + "epoch": 18.78594249201278, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11760 + }, + { + "epoch": 18.787539936102235, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 11761 + }, + { + "epoch": 18.789137380191693, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11762 + }, + { + "epoch": 18.79073482428115, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11763 + }, + { + "epoch": 18.792332268370608, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11764 + }, + { + "epoch": 18.793929712460063, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11765 + }, + { + "epoch": 18.79552715654952, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11766 + }, + { + "epoch": 18.797124600638977, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11767 + }, + { + "epoch": 18.798722044728436, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11768 + }, + { + "epoch": 18.80031948881789, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11769 + }, + { + "epoch": 18.80191693290735, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11770 + }, + { + "epoch": 18.803514376996805, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11771 + }, + { + "epoch": 18.80511182108626, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11772 + }, + { + "epoch": 18.80670926517572, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 11773 + }, + { + "epoch": 18.808306709265175, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11774 + }, + { + "epoch": 18.809904153354633, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 11775 + }, + { + "epoch": 18.81150159744409, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 11776 + }, + { + "epoch": 18.813099041533548, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11777 + }, + { + "epoch": 18.814696485623003, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11778 + }, + { + "epoch": 18.81629392971246, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11779 + }, + { + "epoch": 18.817891373801917, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11780 + }, + { + "epoch": 18.819488817891372, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11781 + }, + { + "epoch": 18.82108626198083, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11782 + }, + { + "epoch": 18.822683706070286, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11783 + }, + { + "epoch": 18.824281150159745, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 11784 + }, + { + "epoch": 18.8258785942492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11785 + }, + { + "epoch": 18.82747603833866, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11786 + }, + { + "epoch": 18.829073482428115, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11787 + }, + { + "epoch": 18.830670926517573, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11788 + }, + { + "epoch": 18.83226837060703, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11789 + }, + { + "epoch": 18.833865814696484, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11790 + }, + { + "epoch": 18.835463258785943, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11791 + }, + { + "epoch": 18.837060702875398, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11792 + }, + { + "epoch": 18.838658146964857, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11793 + }, + { + "epoch": 18.840255591054312, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11794 + }, + { + "epoch": 18.84185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11795 + }, + { + "epoch": 18.843450479233226, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 11796 + }, + { + "epoch": 18.845047923322685, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11797 + }, + { + "epoch": 18.84664536741214, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11798 + }, + { + "epoch": 18.8482428115016, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11799 + }, + { + "epoch": 18.849840255591054, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 11800 + }, + { + "epoch": 18.85143769968051, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11801 + }, + { + "epoch": 18.85303514376997, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 11802 + }, + { + "epoch": 18.854632587859424, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11803 + }, + { + "epoch": 18.856230031948883, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11804 + }, + { + "epoch": 18.857827476038338, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11805 + }, + { + "epoch": 18.859424920127797, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11806 + }, + { + "epoch": 18.861022364217252, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11807 + }, + { + "epoch": 18.86261980830671, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11808 + }, + { + "epoch": 18.864217252396166, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 11809 + }, + { + "epoch": 18.86581469648562, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11810 + }, + { + "epoch": 18.86741214057508, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11811 + }, + { + "epoch": 18.869009584664536, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11812 + }, + { + "epoch": 18.870607028753994, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11813 + }, + { + "epoch": 18.87220447284345, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11814 + }, + { + "epoch": 18.87380191693291, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11815 + }, + { + "epoch": 18.875399361022364, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11816 + }, + { + "epoch": 18.876996805111823, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11817 + }, + { + "epoch": 18.878594249201278, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11818 + }, + { + "epoch": 18.880191693290733, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11819 + }, + { + "epoch": 18.881789137380192, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11820 + }, + { + "epoch": 18.883386581469647, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11821 + }, + { + "epoch": 18.884984025559106, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11822 + }, + { + "epoch": 18.88658146964856, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11823 + }, + { + "epoch": 18.88817891373802, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11824 + }, + { + "epoch": 18.889776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11825 + }, + { + "epoch": 18.891373801916934, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11826 + }, + { + "epoch": 18.89297124600639, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11827 + }, + { + "epoch": 18.894568690095845, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11828 + }, + { + "epoch": 18.896166134185304, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11829 + }, + { + "epoch": 18.89776357827476, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11830 + }, + { + "epoch": 18.899361022364218, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11831 + }, + { + "epoch": 18.900958466453673, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11832 + }, + { + "epoch": 18.902555910543132, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11833 + }, + { + "epoch": 18.904153354632587, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11834 + }, + { + "epoch": 18.905750798722046, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11835 + }, + { + "epoch": 18.9073482428115, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11836 + }, + { + "epoch": 18.908945686900957, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 11837 + }, + { + "epoch": 18.910543130990416, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11838 + }, + { + "epoch": 18.91214057507987, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 11839 + }, + { + "epoch": 18.91373801916933, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11840 + }, + { + "epoch": 18.915335463258785, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11841 + }, + { + "epoch": 18.916932907348244, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11842 + }, + { + "epoch": 18.9185303514377, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11843 + }, + { + "epoch": 18.920127795527158, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11844 + }, + { + "epoch": 18.921725239616613, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11845 + }, + { + "epoch": 18.923322683706072, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11846 + }, + { + "epoch": 18.924920127795527, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11847 + }, + { + "epoch": 18.926517571884983, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11848 + }, + { + "epoch": 18.92811501597444, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11849 + }, + { + "epoch": 18.929712460063897, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11850 + }, + { + "epoch": 18.931309904153355, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11851 + }, + { + "epoch": 18.93290734824281, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 11852 + }, + { + "epoch": 18.93450479233227, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11853 + }, + { + "epoch": 18.936102236421725, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11854 + }, + { + "epoch": 18.937699680511184, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11855 + }, + { + "epoch": 18.93929712460064, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11856 + }, + { + "epoch": 18.940894568690094, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11857 + }, + { + "epoch": 18.942492012779553, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11858 + }, + { + "epoch": 18.94408945686901, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11859 + }, + { + "epoch": 18.945686900958467, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11860 + }, + { + "epoch": 18.947284345047922, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11861 + }, + { + "epoch": 18.94888178913738, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11862 + }, + { + "epoch": 18.950479233226837, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 11863 + }, + { + "epoch": 18.952076677316295, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11864 + }, + { + "epoch": 18.95367412140575, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11865 + }, + { + "epoch": 18.955271565495206, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11866 + }, + { + "epoch": 18.956869009584665, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11867 + }, + { + "epoch": 18.95846645367412, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11868 + }, + { + "epoch": 18.96006389776358, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11869 + }, + { + "epoch": 18.961661341853034, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11870 + }, + { + "epoch": 18.963258785942493, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 11871 + }, + { + "epoch": 18.96485623003195, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11872 + }, + { + "epoch": 18.966453674121407, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11873 + }, + { + "epoch": 18.968051118210862, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11874 + }, + { + "epoch": 18.96964856230032, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11875 + }, + { + "epoch": 18.971246006389777, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11876 + }, + { + "epoch": 18.972843450479232, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11877 + }, + { + "epoch": 18.97444089456869, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11878 + }, + { + "epoch": 18.976038338658146, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11879 + }, + { + "epoch": 18.977635782747605, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11880 + }, + { + "epoch": 18.97923322683706, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11881 + }, + { + "epoch": 18.98083067092652, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 11882 + }, + { + "epoch": 18.982428115015974, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11883 + }, + { + "epoch": 18.984025559105433, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11884 + }, + { + "epoch": 18.98562300319489, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11885 + }, + { + "epoch": 18.987220447284344, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11886 + }, + { + "epoch": 18.988817891373802, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11887 + }, + { + "epoch": 18.990415335463258, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11888 + }, + { + "epoch": 18.992012779552716, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11889 + }, + { + "epoch": 18.99361022364217, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11890 + }, + { + "epoch": 18.99520766773163, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11891 + }, + { + "epoch": 18.996805111821086, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11892 + }, + { + "epoch": 18.998402555910545, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11893 + }, + { + "epoch": 19.0, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11894 + }, + { + "epoch": 19.001597444089455, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11895 + }, + { + "epoch": 19.003194888178914, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11896 + }, + { + "epoch": 19.00479233226837, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11897 + }, + { + "epoch": 19.00638977635783, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11898 + }, + { + "epoch": 19.007987220447284, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 11899 + }, + { + "epoch": 19.009584664536742, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11900 + }, + { + "epoch": 19.011182108626198, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11901 + }, + { + "epoch": 19.012779552715656, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11902 + }, + { + "epoch": 19.01437699680511, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 11903 + }, + { + "epoch": 19.015974440894567, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11904 + }, + { + "epoch": 19.017571884984026, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11905 + }, + { + "epoch": 19.01916932907348, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11906 + }, + { + "epoch": 19.02076677316294, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 11907 + }, + { + "epoch": 19.022364217252395, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11908 + }, + { + "epoch": 19.023961661341854, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11909 + }, + { + "epoch": 19.02555910543131, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11910 + }, + { + "epoch": 19.027156549520768, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11911 + }, + { + "epoch": 19.028753993610223, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 11912 + }, + { + "epoch": 19.03035143769968, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11913 + }, + { + "epoch": 19.031948881789138, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11914 + }, + { + "epoch": 19.033546325878593, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11915 + }, + { + "epoch": 19.03514376996805, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 11916 + }, + { + "epoch": 19.036741214057507, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11917 + }, + { + "epoch": 19.038338658146966, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11918 + }, + { + "epoch": 19.03993610223642, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11919 + }, + { + "epoch": 19.04153354632588, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11920 + }, + { + "epoch": 19.043130990415335, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 11921 + }, + { + "epoch": 19.044728434504794, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11922 + }, + { + "epoch": 19.04632587859425, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11923 + }, + { + "epoch": 19.047923322683705, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11924 + }, + { + "epoch": 19.049520766773163, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11925 + }, + { + "epoch": 19.05111821086262, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11926 + }, + { + "epoch": 19.052715654952078, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11927 + }, + { + "epoch": 19.054313099041533, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11928 + }, + { + "epoch": 19.05591054313099, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11929 + }, + { + "epoch": 19.057507987220447, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11930 + }, + { + "epoch": 19.059105431309906, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11931 + }, + { + "epoch": 19.06070287539936, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11932 + }, + { + "epoch": 19.062300319488816, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11933 + }, + { + "epoch": 19.063897763578275, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 11934 + }, + { + "epoch": 19.06549520766773, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11935 + }, + { + "epoch": 19.06709265175719, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11936 + }, + { + "epoch": 19.068690095846645, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11937 + }, + { + "epoch": 19.070287539936103, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 11938 + }, + { + "epoch": 19.07188498402556, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 11939 + }, + { + "epoch": 19.073482428115017, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 11940 + }, + { + "epoch": 19.075079872204473, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11941 + }, + { + "epoch": 19.076677316293928, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11942 + }, + { + "epoch": 19.078274760383387, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11943 + }, + { + "epoch": 19.079872204472842, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11944 + }, + { + "epoch": 19.0814696485623, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11945 + }, + { + "epoch": 19.083067092651756, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11946 + }, + { + "epoch": 19.084664536741215, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11947 + }, + { + "epoch": 19.08626198083067, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11948 + }, + { + "epoch": 19.08785942492013, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11949 + }, + { + "epoch": 19.089456869009584, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11950 + }, + { + "epoch": 19.09105431309904, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11951 + }, + { + "epoch": 19.0926517571885, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11952 + }, + { + "epoch": 19.094249201277954, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11953 + }, + { + "epoch": 19.095846645367413, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11954 + }, + { + "epoch": 19.097444089456868, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 11955 + }, + { + "epoch": 19.099041533546327, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11956 + }, + { + "epoch": 19.100638977635782, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11957 + }, + { + "epoch": 19.10223642172524, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11958 + }, + { + "epoch": 19.103833865814696, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11959 + }, + { + "epoch": 19.105431309904155, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11960 + }, + { + "epoch": 19.10702875399361, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 11961 + }, + { + "epoch": 19.108626198083066, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11962 + }, + { + "epoch": 19.110223642172524, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11963 + }, + { + "epoch": 19.11182108626198, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 11964 + }, + { + "epoch": 19.11341853035144, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11965 + }, + { + "epoch": 19.115015974440894, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 11966 + }, + { + "epoch": 19.116613418530353, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11967 + }, + { + "epoch": 19.118210862619808, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 11968 + }, + { + "epoch": 19.119808306709267, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11969 + }, + { + "epoch": 19.121405750798722, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11970 + }, + { + "epoch": 19.123003194888177, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11971 + }, + { + "epoch": 19.124600638977636, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11972 + }, + { + "epoch": 19.12619808306709, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11973 + }, + { + "epoch": 19.12779552715655, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11974 + }, + { + "epoch": 19.129392971246006, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 11975 + }, + { + "epoch": 19.130990415335464, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11976 + }, + { + "epoch": 19.13258785942492, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11977 + }, + { + "epoch": 19.13418530351438, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11978 + }, + { + "epoch": 19.135782747603834, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11979 + }, + { + "epoch": 19.13738019169329, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11980 + }, + { + "epoch": 19.138977635782748, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 11981 + }, + { + "epoch": 19.140575079872203, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11982 + }, + { + "epoch": 19.142172523961662, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 11983 + }, + { + "epoch": 19.143769968051117, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11984 + }, + { + "epoch": 19.145367412140576, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11985 + }, + { + "epoch": 19.14696485623003, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11986 + }, + { + "epoch": 19.14856230031949, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11987 + }, + { + "epoch": 19.150159744408946, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11988 + }, + { + "epoch": 19.1517571884984, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11989 + }, + { + "epoch": 19.15335463258786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11990 + }, + { + "epoch": 19.154952076677315, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11991 + }, + { + "epoch": 19.156549520766774, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11992 + }, + { + "epoch": 19.15814696485623, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11993 + }, + { + "epoch": 19.159744408945688, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11994 + }, + { + "epoch": 19.161341853035143, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11995 + }, + { + "epoch": 19.162939297124602, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11996 + }, + { + "epoch": 19.164536741214057, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 11997 + }, + { + "epoch": 19.166134185303516, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11998 + }, + { + "epoch": 19.16773162939297, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11999 + }, + { + "epoch": 19.169329073482427, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12000 + }, + { + "epoch": 19.170926517571885, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12001 + }, + { + "epoch": 19.17252396166134, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12002 + }, + { + "epoch": 19.1741214057508, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12003 + }, + { + "epoch": 19.175718849840255, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12004 + }, + { + "epoch": 19.177316293929714, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12005 + }, + { + "epoch": 19.17891373801917, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12006 + }, + { + "epoch": 19.180511182108628, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 12007 + }, + { + "epoch": 19.182108626198083, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12008 + }, + { + "epoch": 19.18370607028754, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12009 + }, + { + "epoch": 19.185303514376997, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12010 + }, + { + "epoch": 19.186900958466452, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 12011 + }, + { + "epoch": 19.18849840255591, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 12012 + }, + { + "epoch": 19.190095846645367, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12013 + }, + { + "epoch": 19.191693290734825, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 12014 + }, + { + "epoch": 19.19329073482428, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12015 + }, + { + "epoch": 19.19488817891374, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 12016 + }, + { + "epoch": 19.196485623003195, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12017 + }, + { + "epoch": 19.19808306709265, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12018 + }, + { + "epoch": 19.19968051118211, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12019 + }, + { + "epoch": 19.201277955271564, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12020 + }, + { + "epoch": 19.202875399361023, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12021 + }, + { + "epoch": 19.20447284345048, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12022 + }, + { + "epoch": 19.206070287539937, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12023 + }, + { + "epoch": 19.207667731629392, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 12024 + }, + { + "epoch": 19.20926517571885, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12025 + }, + { + "epoch": 19.210862619808307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12026 + }, + { + "epoch": 19.212460063897762, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12027 + }, + { + "epoch": 19.21405750798722, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12028 + }, + { + "epoch": 19.215654952076676, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12029 + }, + { + "epoch": 19.217252396166135, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12030 + }, + { + "epoch": 19.21884984025559, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12031 + }, + { + "epoch": 19.22044728434505, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12032 + }, + { + "epoch": 19.222044728434504, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12033 + }, + { + "epoch": 19.223642172523963, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12034 + }, + { + "epoch": 19.22523961661342, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12035 + }, + { + "epoch": 19.226837060702877, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12036 + }, + { + "epoch": 19.228434504792332, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12037 + }, + { + "epoch": 19.230031948881788, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 12038 + }, + { + "epoch": 19.231629392971247, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12039 + }, + { + "epoch": 19.233226837060702, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12040 + }, + { + "epoch": 19.23482428115016, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12041 + }, + { + "epoch": 19.236421725239616, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12042 + }, + { + "epoch": 19.238019169329075, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12043 + }, + { + "epoch": 19.23961661341853, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12044 + }, + { + "epoch": 19.24121405750799, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12045 + }, + { + "epoch": 19.242811501597444, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12046 + }, + { + "epoch": 19.2444089456869, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12047 + }, + { + "epoch": 19.24600638977636, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12048 + }, + { + "epoch": 19.247603833865814, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 12049 + }, + { + "epoch": 19.249201277955272, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12050 + }, + { + "epoch": 19.250798722044728, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12051 + }, + { + "epoch": 19.252396166134186, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12052 + }, + { + "epoch": 19.25399361022364, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12053 + }, + { + "epoch": 19.2555910543131, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12054 + }, + { + "epoch": 19.257188498402556, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12055 + }, + { + "epoch": 19.25878594249201, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12056 + }, + { + "epoch": 19.26038338658147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12057 + }, + { + "epoch": 19.261980830670925, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12058 + }, + { + "epoch": 19.263578274760384, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12059 + }, + { + "epoch": 19.26517571884984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12060 + }, + { + "epoch": 19.266773162939298, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 12061 + }, + { + "epoch": 19.268370607028753, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12062 + }, + { + "epoch": 19.269968051118212, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12063 + }, + { + "epoch": 19.271565495207668, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12064 + }, + { + "epoch": 19.273162939297123, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12065 + }, + { + "epoch": 19.27476038338658, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12066 + }, + { + "epoch": 19.276357827476037, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12067 + }, + { + "epoch": 19.277955271565496, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12068 + }, + { + "epoch": 19.27955271565495, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12069 + }, + { + "epoch": 19.28115015974441, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12070 + }, + { + "epoch": 19.282747603833865, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12071 + }, + { + "epoch": 19.284345047923324, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12072 + }, + { + "epoch": 19.28594249201278, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12073 + }, + { + "epoch": 19.287539936102238, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12074 + }, + { + "epoch": 19.289137380191693, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12075 + }, + { + "epoch": 19.29073482428115, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12076 + }, + { + "epoch": 19.292332268370608, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12077 + }, + { + "epoch": 19.293929712460063, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12078 + }, + { + "epoch": 19.29552715654952, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12079 + }, + { + "epoch": 19.297124600638977, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 12080 + }, + { + "epoch": 19.298722044728436, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12081 + }, + { + "epoch": 19.30031948881789, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12082 + }, + { + "epoch": 19.30191693290735, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 12083 + }, + { + "epoch": 19.303514376996805, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12084 + }, + { + "epoch": 19.30511182108626, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12085 + }, + { + "epoch": 19.30670926517572, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12086 + }, + { + "epoch": 19.308306709265175, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 12087 + }, + { + "epoch": 19.309904153354633, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12088 + }, + { + "epoch": 19.31150159744409, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12089 + }, + { + "epoch": 19.313099041533548, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12090 + }, + { + "epoch": 19.314696485623003, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12091 + }, + { + "epoch": 19.31629392971246, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12092 + }, + { + "epoch": 19.317891373801917, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12093 + }, + { + "epoch": 19.319488817891372, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12094 + }, + { + "epoch": 19.32108626198083, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12095 + }, + { + "epoch": 19.322683706070286, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12096 + }, + { + "epoch": 19.324281150159745, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12097 + }, + { + "epoch": 19.3258785942492, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12098 + }, + { + "epoch": 19.32747603833866, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 12099 + }, + { + "epoch": 19.329073482428115, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12100 + }, + { + "epoch": 19.330670926517573, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 12101 + }, + { + "epoch": 19.33226837060703, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12102 + }, + { + "epoch": 19.333865814696484, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12103 + }, + { + "epoch": 19.335463258785943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12104 + }, + { + "epoch": 19.337060702875398, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12105 + }, + { + "epoch": 19.338658146964857, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12106 + }, + { + "epoch": 19.340255591054312, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12107 + }, + { + "epoch": 19.34185303514377, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12108 + }, + { + "epoch": 19.343450479233226, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12109 + }, + { + "epoch": 19.345047923322685, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12110 + }, + { + "epoch": 19.34664536741214, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12111 + }, + { + "epoch": 19.3482428115016, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12112 + }, + { + "epoch": 19.349840255591054, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12113 + }, + { + "epoch": 19.35143769968051, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12114 + }, + { + "epoch": 19.35303514376997, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12115 + }, + { + "epoch": 19.354632587859424, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12116 + }, + { + "epoch": 19.356230031948883, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12117 + }, + { + "epoch": 19.357827476038338, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12118 + }, + { + "epoch": 19.359424920127797, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12119 + }, + { + "epoch": 19.361022364217252, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12120 + }, + { + "epoch": 19.36261980830671, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12121 + }, + { + "epoch": 19.364217252396166, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12122 + }, + { + "epoch": 19.36581469648562, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12123 + }, + { + "epoch": 19.36741214057508, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12124 + }, + { + "epoch": 19.369009584664536, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 12125 + }, + { + "epoch": 19.370607028753994, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 12126 + }, + { + "epoch": 19.37220447284345, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12127 + }, + { + "epoch": 19.37380191693291, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 12128 + }, + { + "epoch": 19.375399361022364, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12129 + }, + { + "epoch": 19.376996805111823, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12130 + }, + { + "epoch": 19.378594249201278, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12131 + }, + { + "epoch": 19.380191693290733, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12132 + }, + { + "epoch": 19.381789137380192, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12133 + }, + { + "epoch": 19.383386581469647, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 12134 + }, + { + "epoch": 19.384984025559106, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12135 + }, + { + "epoch": 19.38658146964856, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12136 + }, + { + "epoch": 19.38817891373802, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12137 + }, + { + "epoch": 19.389776357827476, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12138 + }, + { + "epoch": 19.391373801916934, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12139 + }, + { + "epoch": 19.39297124600639, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12140 + }, + { + "epoch": 19.394568690095845, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 12141 + }, + { + "epoch": 19.396166134185304, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 12142 + }, + { + "epoch": 19.39776357827476, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 12143 + }, + { + "epoch": 19.399361022364218, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12144 + }, + { + "epoch": 19.400958466453673, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12145 + }, + { + "epoch": 19.402555910543132, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12146 + }, + { + "epoch": 19.404153354632587, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12147 + }, + { + "epoch": 19.405750798722046, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12148 + }, + { + "epoch": 19.4073482428115, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12149 + }, + { + "epoch": 19.408945686900957, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 12150 + }, + { + "epoch": 19.410543130990416, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12151 + }, + { + "epoch": 19.41214057507987, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12152 + }, + { + "epoch": 19.41373801916933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12153 + }, + { + "epoch": 19.415335463258785, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12154 + }, + { + "epoch": 19.416932907348244, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12155 + }, + { + "epoch": 19.4185303514377, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12156 + }, + { + "epoch": 19.420127795527158, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12157 + }, + { + "epoch": 19.421725239616613, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12158 + }, + { + "epoch": 19.423322683706072, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12159 + }, + { + "epoch": 19.424920127795527, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12160 + }, + { + "epoch": 19.426517571884983, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12161 + }, + { + "epoch": 19.42811501597444, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 12162 + }, + { + "epoch": 19.429712460063897, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12163 + }, + { + "epoch": 19.431309904153355, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 12164 + }, + { + "epoch": 19.43290734824281, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12165 + }, + { + "epoch": 19.43450479233227, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12166 + }, + { + "epoch": 19.436102236421725, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 12167 + }, + { + "epoch": 19.437699680511184, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12168 + }, + { + "epoch": 19.43929712460064, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12169 + }, + { + "epoch": 19.440894568690094, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12170 + }, + { + "epoch": 19.442492012779553, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12171 + }, + { + "epoch": 19.44408945686901, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12172 + }, + { + "epoch": 19.445686900958467, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12173 + }, + { + "epoch": 19.447284345047922, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12174 + }, + { + "epoch": 19.44888178913738, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 12175 + }, + { + "epoch": 19.450479233226837, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12176 + }, + { + "epoch": 19.452076677316295, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12177 + }, + { + "epoch": 19.45367412140575, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12178 + }, + { + "epoch": 19.455271565495206, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12179 + }, + { + "epoch": 19.456869009584665, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12180 + }, + { + "epoch": 19.45846645367412, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12181 + }, + { + "epoch": 19.46006389776358, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12182 + }, + { + "epoch": 19.461661341853034, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12183 + }, + { + "epoch": 19.463258785942493, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 12184 + }, + { + "epoch": 19.46485623003195, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12185 + }, + { + "epoch": 19.466453674121407, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12186 + }, + { + "epoch": 19.468051118210862, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12187 + }, + { + "epoch": 19.46964856230032, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12188 + }, + { + "epoch": 19.471246006389777, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12189 + }, + { + "epoch": 19.472843450479232, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12190 + }, + { + "epoch": 19.47444089456869, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12191 + }, + { + "epoch": 19.476038338658146, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12192 + }, + { + "epoch": 19.477635782747605, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12193 + }, + { + "epoch": 19.47923322683706, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12194 + }, + { + "epoch": 19.48083067092652, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12195 + }, + { + "epoch": 19.482428115015974, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12196 + }, + { + "epoch": 19.484025559105433, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12197 + }, + { + "epoch": 19.48562300319489, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12198 + }, + { + "epoch": 19.487220447284344, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12199 + }, + { + "epoch": 19.488817891373802, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12200 + }, + { + "epoch": 19.490415335463258, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12201 + }, + { + "epoch": 19.492012779552716, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 12202 + }, + { + "epoch": 19.49361022364217, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12203 + }, + { + "epoch": 19.49520766773163, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12204 + }, + { + "epoch": 19.496805111821086, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12205 + }, + { + "epoch": 19.498402555910545, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12206 + }, + { + "epoch": 19.5, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 12207 + }, + { + "epoch": 19.501597444089455, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12208 + }, + { + "epoch": 19.503194888178914, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12209 + }, + { + "epoch": 19.50479233226837, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12210 + }, + { + "epoch": 19.50638977635783, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12211 + }, + { + "epoch": 19.507987220447284, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 12212 + }, + { + "epoch": 19.509584664536742, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 12213 + }, + { + "epoch": 19.511182108626198, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12214 + }, + { + "epoch": 19.512779552715656, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12215 + }, + { + "epoch": 19.51437699680511, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12216 + }, + { + "epoch": 19.515974440894567, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12217 + }, + { + "epoch": 19.517571884984026, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12218 + }, + { + "epoch": 19.51916932907348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12219 + }, + { + "epoch": 19.52076677316294, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12220 + }, + { + "epoch": 19.522364217252395, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12221 + }, + { + "epoch": 19.523961661341854, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12222 + }, + { + "epoch": 19.52555910543131, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 12223 + }, + { + "epoch": 19.527156549520768, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12224 + }, + { + "epoch": 19.528753993610223, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 12225 + }, + { + "epoch": 19.53035143769968, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12226 + }, + { + "epoch": 19.531948881789138, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12227 + }, + { + "epoch": 19.533546325878593, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12228 + }, + { + "epoch": 19.53514376996805, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12229 + }, + { + "epoch": 19.536741214057507, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12230 + }, + { + "epoch": 19.538338658146966, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12231 + }, + { + "epoch": 19.53993610223642, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 12232 + }, + { + "epoch": 19.54153354632588, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12233 + }, + { + "epoch": 19.543130990415335, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12234 + }, + { + "epoch": 19.544728434504794, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12235 + }, + { + "epoch": 19.54632587859425, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12236 + }, + { + "epoch": 19.547923322683705, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12237 + }, + { + "epoch": 19.549520766773163, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12238 + }, + { + "epoch": 19.55111821086262, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12239 + }, + { + "epoch": 19.552715654952078, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 12240 + }, + { + "epoch": 19.554313099041533, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12241 + }, + { + "epoch": 19.55591054313099, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12242 + }, + { + "epoch": 19.557507987220447, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12243 + }, + { + "epoch": 19.559105431309906, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12244 + }, + { + "epoch": 19.56070287539936, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 12245 + }, + { + "epoch": 19.562300319488816, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12246 + }, + { + "epoch": 19.563897763578275, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12247 + }, + { + "epoch": 19.56549520766773, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12248 + }, + { + "epoch": 19.56709265175719, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12249 + }, + { + "epoch": 19.568690095846645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 12250 + }, + { + "epoch": 19.570287539936103, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12251 + }, + { + "epoch": 19.57188498402556, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12252 + }, + { + "epoch": 19.573482428115017, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 12253 + }, + { + "epoch": 19.575079872204473, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12254 + }, + { + "epoch": 19.576677316293928, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12255 + }, + { + "epoch": 19.578274760383387, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12256 + }, + { + "epoch": 19.579872204472842, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12257 + }, + { + "epoch": 19.5814696485623, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12258 + }, + { + "epoch": 19.583067092651756, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12259 + }, + { + "epoch": 19.584664536741215, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12260 + }, + { + "epoch": 19.58626198083067, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12261 + }, + { + "epoch": 19.58785942492013, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12262 + }, + { + "epoch": 19.589456869009584, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12263 + }, + { + "epoch": 19.591054313099043, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12264 + }, + { + "epoch": 19.5926517571885, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12265 + }, + { + "epoch": 19.594249201277954, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12266 + }, + { + "epoch": 19.595846645367413, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12267 + }, + { + "epoch": 19.597444089456868, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12268 + }, + { + "epoch": 19.599041533546327, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12269 + }, + { + "epoch": 19.600638977635782, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12270 + }, + { + "epoch": 19.60223642172524, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12271 + }, + { + "epoch": 19.603833865814696, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12272 + }, + { + "epoch": 19.605431309904155, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12273 + }, + { + "epoch": 19.60702875399361, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 12274 + }, + { + "epoch": 19.608626198083066, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12275 + }, + { + "epoch": 19.610223642172524, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12276 + }, + { + "epoch": 19.61182108626198, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12277 + }, + { + "epoch": 19.61341853035144, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12278 + }, + { + "epoch": 19.615015974440894, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12279 + }, + { + "epoch": 19.616613418530353, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12280 + }, + { + "epoch": 19.618210862619808, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 12281 + }, + { + "epoch": 19.619808306709267, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12282 + }, + { + "epoch": 19.621405750798722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12283 + }, + { + "epoch": 19.623003194888177, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12284 + }, + { + "epoch": 19.624600638977636, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12285 + }, + { + "epoch": 19.62619808306709, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12286 + }, + { + "epoch": 19.62779552715655, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12287 + }, + { + "epoch": 19.629392971246006, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 12288 + }, + { + "epoch": 19.630990415335464, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 12289 + }, + { + "epoch": 19.63258785942492, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12290 + }, + { + "epoch": 19.63418530351438, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12291 + }, + { + "epoch": 19.635782747603834, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12292 + }, + { + "epoch": 19.63738019169329, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12293 + }, + { + "epoch": 19.638977635782748, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12294 + }, + { + "epoch": 19.640575079872203, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12295 + }, + { + "epoch": 19.642172523961662, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 12296 + }, + { + "epoch": 19.643769968051117, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12297 + }, + { + "epoch": 19.645367412140576, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12298 + }, + { + "epoch": 19.64696485623003, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12299 + }, + { + "epoch": 19.64856230031949, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12300 + }, + { + "epoch": 19.650159744408946, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12301 + }, + { + "epoch": 19.6517571884984, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12302 + }, + { + "epoch": 19.65335463258786, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12303 + }, + { + "epoch": 19.654952076677315, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12304 + }, + { + "epoch": 19.656549520766774, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12305 + }, + { + "epoch": 19.65814696485623, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12306 + }, + { + "epoch": 19.659744408945688, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12307 + }, + { + "epoch": 19.661341853035143, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12308 + }, + { + "epoch": 19.662939297124602, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12309 + }, + { + "epoch": 19.664536741214057, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12310 + }, + { + "epoch": 19.666134185303516, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12311 + }, + { + "epoch": 19.66773162939297, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12312 + }, + { + "epoch": 19.669329073482427, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12313 + }, + { + "epoch": 19.670926517571885, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12314 + }, + { + "epoch": 19.67252396166134, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12315 + }, + { + "epoch": 19.6741214057508, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12316 + }, + { + "epoch": 19.675718849840255, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12317 + }, + { + "epoch": 19.677316293929714, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12318 + }, + { + "epoch": 19.67891373801917, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12319 + }, + { + "epoch": 19.680511182108628, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 12320 + }, + { + "epoch": 19.682108626198083, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12321 + }, + { + "epoch": 19.68370607028754, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12322 + }, + { + "epoch": 19.685303514376997, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12323 + }, + { + "epoch": 19.686900958466452, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 12324 + }, + { + "epoch": 19.68849840255591, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12325 + }, + { + "epoch": 19.690095846645367, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12326 + }, + { + "epoch": 19.691693290734825, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12327 + }, + { + "epoch": 19.69329073482428, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12328 + }, + { + "epoch": 19.69488817891374, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12329 + }, + { + "epoch": 19.696485623003195, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12330 + }, + { + "epoch": 19.69808306709265, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 12331 + }, + { + "epoch": 19.69968051118211, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12332 + }, + { + "epoch": 19.701277955271564, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12333 + }, + { + "epoch": 19.702875399361023, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12334 + }, + { + "epoch": 19.70447284345048, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12335 + }, + { + "epoch": 19.706070287539937, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12336 + }, + { + "epoch": 19.707667731629392, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 12337 + }, + { + "epoch": 19.70926517571885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 12338 + }, + { + "epoch": 19.710862619808307, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12339 + }, + { + "epoch": 19.712460063897765, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12340 + }, + { + "epoch": 19.71405750798722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 12341 + }, + { + "epoch": 19.715654952076676, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12342 + }, + { + "epoch": 19.717252396166135, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12343 + }, + { + "epoch": 19.71884984025559, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12344 + }, + { + "epoch": 19.72044728434505, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 12345 + }, + { + "epoch": 19.722044728434504, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12346 + }, + { + "epoch": 19.723642172523963, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12347 + }, + { + "epoch": 19.72523961661342, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12348 + }, + { + "epoch": 19.726837060702877, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12349 + }, + { + "epoch": 19.728434504792332, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12350 + }, + { + "epoch": 19.730031948881788, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12351 + }, + { + "epoch": 19.731629392971247, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12352 + }, + { + "epoch": 19.733226837060702, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12353 + }, + { + "epoch": 19.73482428115016, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 12354 + }, + { + "epoch": 19.736421725239616, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12355 + }, + { + "epoch": 19.738019169329075, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12356 + }, + { + "epoch": 19.73961661341853, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12357 + }, + { + "epoch": 19.74121405750799, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12358 + }, + { + "epoch": 19.742811501597444, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12359 + }, + { + "epoch": 19.7444089456869, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12360 + }, + { + "epoch": 19.74600638977636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12361 + }, + { + "epoch": 19.747603833865814, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12362 + }, + { + "epoch": 19.749201277955272, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12363 + }, + { + "epoch": 19.750798722044728, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12364 + }, + { + "epoch": 19.752396166134186, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 12365 + }, + { + "epoch": 19.75399361022364, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12366 + }, + { + "epoch": 19.7555910543131, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12367 + }, + { + "epoch": 19.757188498402556, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12368 + }, + { + "epoch": 19.75878594249201, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12369 + }, + { + "epoch": 19.76038338658147, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 12370 + }, + { + "epoch": 19.761980830670925, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12371 + }, + { + "epoch": 19.763578274760384, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12372 + }, + { + "epoch": 19.76517571884984, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12373 + }, + { + "epoch": 19.766773162939298, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12374 + }, + { + "epoch": 19.768370607028753, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12375 + }, + { + "epoch": 19.769968051118212, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 12376 + }, + { + "epoch": 19.771565495207668, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12377 + }, + { + "epoch": 19.773162939297123, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12378 + }, + { + "epoch": 19.77476038338658, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12379 + }, + { + "epoch": 19.776357827476037, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12380 + }, + { + "epoch": 19.777955271565496, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12381 + }, + { + "epoch": 19.77955271565495, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12382 + }, + { + "epoch": 19.78115015974441, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12383 + }, + { + "epoch": 19.782747603833865, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12384 + }, + { + "epoch": 19.784345047923324, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12385 + }, + { + "epoch": 19.78594249201278, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12386 + }, + { + "epoch": 19.787539936102235, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12387 + }, + { + "epoch": 19.789137380191693, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12388 + }, + { + "epoch": 19.79073482428115, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12389 + }, + { + "epoch": 19.792332268370608, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12390 + }, + { + "epoch": 19.793929712460063, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12391 + }, + { + "epoch": 19.79552715654952, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12392 + }, + { + "epoch": 19.797124600638977, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 12393 + }, + { + "epoch": 19.798722044728436, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12394 + }, + { + "epoch": 19.80031948881789, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12395 + }, + { + "epoch": 19.80191693290735, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12396 + }, + { + "epoch": 19.803514376996805, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12397 + }, + { + "epoch": 19.80511182108626, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12398 + }, + { + "epoch": 19.80670926517572, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12399 + }, + { + "epoch": 19.808306709265175, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12400 + }, + { + "epoch": 19.809904153354633, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12401 + }, + { + "epoch": 19.81150159744409, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12402 + }, + { + "epoch": 19.813099041533548, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12403 + }, + { + "epoch": 19.814696485623003, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12404 + }, + { + "epoch": 19.81629392971246, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12405 + }, + { + "epoch": 19.817891373801917, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12406 + }, + { + "epoch": 19.819488817891372, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 12407 + }, + { + "epoch": 19.82108626198083, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12408 + }, + { + "epoch": 19.822683706070286, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12409 + }, + { + "epoch": 19.824281150159745, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12410 + }, + { + "epoch": 19.8258785942492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12411 + }, + { + "epoch": 19.82747603833866, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12412 + }, + { + "epoch": 19.829073482428115, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 12413 + }, + { + "epoch": 19.830670926517573, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12414 + }, + { + "epoch": 19.83226837060703, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12415 + }, + { + "epoch": 19.833865814696484, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12416 + }, + { + "epoch": 19.835463258785943, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12417 + }, + { + "epoch": 19.837060702875398, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12418 + }, + { + "epoch": 19.838658146964857, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12419 + }, + { + "epoch": 19.840255591054312, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12420 + }, + { + "epoch": 19.84185303514377, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12421 + }, + { + "epoch": 19.843450479233226, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12422 + }, + { + "epoch": 19.845047923322685, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12423 + }, + { + "epoch": 19.84664536741214, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12424 + }, + { + "epoch": 19.8482428115016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12425 + }, + { + "epoch": 19.849840255591054, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12426 + }, + { + "epoch": 19.85143769968051, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12427 + }, + { + "epoch": 19.85303514376997, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12428 + }, + { + "epoch": 19.854632587859424, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12429 + }, + { + "epoch": 19.856230031948883, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 12430 + }, + { + "epoch": 19.857827476038338, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12431 + }, + { + "epoch": 19.859424920127797, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12432 + }, + { + "epoch": 19.861022364217252, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12433 + }, + { + "epoch": 19.86261980830671, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12434 + }, + { + "epoch": 19.864217252396166, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12435 + }, + { + "epoch": 19.86581469648562, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12436 + }, + { + "epoch": 19.86741214057508, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12437 + }, + { + "epoch": 19.869009584664536, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12438 + }, + { + "epoch": 19.870607028753994, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12439 + }, + { + "epoch": 19.87220447284345, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12440 + }, + { + "epoch": 19.87380191693291, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12441 + }, + { + "epoch": 19.875399361022364, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12442 + }, + { + "epoch": 19.876996805111823, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12443 + }, + { + "epoch": 19.878594249201278, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12444 + }, + { + "epoch": 19.880191693290733, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12445 + }, + { + "epoch": 19.881789137380192, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 12446 + }, + { + "epoch": 19.883386581469647, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12447 + }, + { + "epoch": 19.884984025559106, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12448 + }, + { + "epoch": 19.88658146964856, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 12449 + }, + { + "epoch": 19.88817891373802, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12450 + }, + { + "epoch": 19.889776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12451 + }, + { + "epoch": 19.891373801916934, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12452 + }, + { + "epoch": 19.89297124600639, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12453 + }, + { + "epoch": 19.894568690095845, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12454 + }, + { + "epoch": 19.896166134185304, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12455 + }, + { + "epoch": 19.89776357827476, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12456 + }, + { + "epoch": 19.899361022364218, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 12457 + }, + { + "epoch": 19.900958466453673, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12458 + }, + { + "epoch": 19.902555910543132, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12459 + }, + { + "epoch": 19.904153354632587, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12460 + }, + { + "epoch": 19.905750798722046, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12461 + }, + { + "epoch": 19.9073482428115, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12462 + }, + { + "epoch": 19.908945686900957, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 12463 + }, + { + "epoch": 19.910543130990416, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12464 + }, + { + "epoch": 19.91214057507987, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12465 + }, + { + "epoch": 19.91373801916933, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12466 + }, + { + "epoch": 19.915335463258785, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12467 + }, + { + "epoch": 19.916932907348244, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12468 + }, + { + "epoch": 19.9185303514377, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12469 + }, + { + "epoch": 19.920127795527158, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12470 + }, + { + "epoch": 19.921725239616613, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12471 + }, + { + "epoch": 19.923322683706072, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 12472 + }, + { + "epoch": 19.924920127795527, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12473 + }, + { + "epoch": 19.926517571884983, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 12474 + }, + { + "epoch": 19.92811501597444, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12475 + }, + { + "epoch": 19.929712460063897, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12476 + }, + { + "epoch": 19.931309904153355, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12477 + }, + { + "epoch": 19.93290734824281, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12478 + }, + { + "epoch": 19.93450479233227, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12479 + }, + { + "epoch": 19.936102236421725, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12480 + }, + { + "epoch": 19.937699680511184, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12481 + }, + { + "epoch": 19.93929712460064, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12482 + }, + { + "epoch": 19.940894568690094, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12483 + }, + { + "epoch": 19.942492012779553, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12484 + }, + { + "epoch": 19.94408945686901, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12485 + }, + { + "epoch": 19.945686900958467, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12486 + }, + { + "epoch": 19.947284345047922, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12487 + }, + { + "epoch": 19.94888178913738, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12488 + }, + { + "epoch": 19.950479233226837, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12489 + }, + { + "epoch": 19.952076677316295, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12490 + }, + { + "epoch": 19.95367412140575, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12491 + }, + { + "epoch": 19.955271565495206, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12492 + }, + { + "epoch": 19.956869009584665, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12493 + }, + { + "epoch": 19.95846645367412, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12494 + }, + { + "epoch": 19.96006389776358, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12495 + }, + { + "epoch": 19.961661341853034, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12496 + }, + { + "epoch": 19.963258785942493, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12497 + }, + { + "epoch": 19.96485623003195, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12498 + }, + { + "epoch": 19.966453674121407, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12499 + }, + { + "epoch": 19.968051118210862, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12500 + }, + { + "epoch": 19.96964856230032, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12501 + }, + { + "epoch": 19.971246006389777, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12502 + }, + { + "epoch": 19.972843450479232, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12503 + }, + { + "epoch": 19.97444089456869, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12504 + }, + { + "epoch": 19.976038338658146, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12505 + }, + { + "epoch": 19.977635782747605, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12506 + }, + { + "epoch": 19.97923322683706, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 12507 + }, + { + "epoch": 19.98083067092652, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12508 + }, + { + "epoch": 19.982428115015974, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12509 + }, + { + "epoch": 19.984025559105433, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12510 + }, + { + "epoch": 19.98562300319489, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 12511 + }, + { + "epoch": 19.987220447284344, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 12512 + }, + { + "epoch": 19.988817891373802, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12513 + }, + { + "epoch": 19.990415335463258, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12514 + }, + { + "epoch": 19.992012779552716, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12515 + }, + { + "epoch": 19.99361022364217, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12516 + }, + { + "epoch": 19.99520766773163, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12517 + }, + { + "epoch": 19.996805111821086, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 12518 + }, + { + "epoch": 19.998402555910545, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12519 + }, + { + "epoch": 20.0, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12520 + }, + { + "epoch": 20.001597444089455, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12521 + }, + { + "epoch": 20.003194888178914, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12522 + }, + { + "epoch": 20.00479233226837, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 12523 + }, + { + "epoch": 20.00638977635783, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12524 + }, + { + "epoch": 20.007987220447284, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12525 + }, + { + "epoch": 20.009584664536742, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12526 + }, + { + "epoch": 20.011182108626198, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12527 + }, + { + "epoch": 20.012779552715656, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12528 + }, + { + "epoch": 20.01437699680511, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12529 + }, + { + "epoch": 20.015974440894567, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12530 + }, + { + "epoch": 20.017571884984026, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12531 + }, + { + "epoch": 20.01916932907348, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12532 + }, + { + "epoch": 20.02076677316294, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12533 + }, + { + "epoch": 20.022364217252395, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 12534 + }, + { + "epoch": 20.023961661341854, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12535 + }, + { + "epoch": 20.02555910543131, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12536 + }, + { + "epoch": 20.027156549520768, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 12537 + }, + { + "epoch": 20.028753993610223, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12538 + }, + { + "epoch": 20.03035143769968, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12539 + }, + { + "epoch": 20.031948881789138, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 12540 + }, + { + "epoch": 20.033546325878593, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 12541 + }, + { + "epoch": 20.03514376996805, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12542 + }, + { + "epoch": 20.036741214057507, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12543 + }, + { + "epoch": 20.038338658146966, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12544 + }, + { + "epoch": 20.03993610223642, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12545 + }, + { + "epoch": 20.04153354632588, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 12546 + }, + { + "epoch": 20.043130990415335, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12547 + }, + { + "epoch": 20.044728434504794, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12548 + }, + { + "epoch": 20.04632587859425, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12549 + }, + { + "epoch": 20.047923322683705, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12550 + }, + { + "epoch": 20.049520766773163, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12551 + }, + { + "epoch": 20.05111821086262, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12552 + }, + { + "epoch": 20.052715654952078, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12553 + }, + { + "epoch": 20.054313099041533, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12554 + }, + { + "epoch": 20.05591054313099, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12555 + }, + { + "epoch": 20.057507987220447, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12556 + }, + { + "epoch": 20.059105431309906, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12557 + }, + { + "epoch": 20.06070287539936, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12558 + }, + { + "epoch": 20.062300319488816, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12559 + }, + { + "epoch": 20.063897763578275, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12560 + }, + { + "epoch": 20.06549520766773, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12561 + }, + { + "epoch": 20.06709265175719, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12562 + }, + { + "epoch": 20.068690095846645, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12563 + }, + { + "epoch": 20.070287539936103, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12564 + }, + { + "epoch": 20.07188498402556, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12565 + }, + { + "epoch": 20.073482428115017, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 12566 + }, + { + "epoch": 20.075079872204473, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 12567 + }, + { + "epoch": 20.076677316293928, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12568 + }, + { + "epoch": 20.078274760383387, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12569 + }, + { + "epoch": 20.079872204472842, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12570 + }, + { + "epoch": 20.0814696485623, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12571 + }, + { + "epoch": 20.083067092651756, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12572 + }, + { + "epoch": 20.084664536741215, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12573 + }, + { + "epoch": 20.08626198083067, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12574 + }, + { + "epoch": 20.08785942492013, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12575 + }, + { + "epoch": 20.089456869009584, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12576 + }, + { + "epoch": 20.09105431309904, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12577 + }, + { + "epoch": 20.0926517571885, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12578 + }, + { + "epoch": 20.094249201277954, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12579 + }, + { + "epoch": 20.095846645367413, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12580 + }, + { + "epoch": 20.097444089456868, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12581 + }, + { + "epoch": 20.099041533546327, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12582 + }, + { + "epoch": 20.100638977635782, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 12583 + }, + { + "epoch": 20.10223642172524, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12584 + }, + { + "epoch": 20.103833865814696, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12585 + }, + { + "epoch": 20.105431309904155, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12586 + }, + { + "epoch": 20.10702875399361, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12587 + }, + { + "epoch": 20.108626198083066, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12588 + }, + { + "epoch": 20.110223642172524, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12589 + }, + { + "epoch": 20.11182108626198, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 12590 + }, + { + "epoch": 20.11341853035144, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12591 + }, + { + "epoch": 20.115015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12592 + }, + { + "epoch": 20.116613418530353, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12593 + }, + { + "epoch": 20.118210862619808, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12594 + }, + { + "epoch": 20.119808306709267, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12595 + }, + { + "epoch": 20.121405750798722, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12596 + }, + { + "epoch": 20.123003194888177, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12597 + }, + { + "epoch": 20.124600638977636, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12598 + }, + { + "epoch": 20.12619808306709, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12599 + }, + { + "epoch": 20.12779552715655, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12600 + }, + { + "epoch": 20.129392971246006, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12601 + }, + { + "epoch": 20.130990415335464, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12602 + }, + { + "epoch": 20.13258785942492, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12603 + }, + { + "epoch": 20.13418530351438, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 12604 + }, + { + "epoch": 20.135782747603834, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12605 + }, + { + "epoch": 20.13738019169329, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12606 + }, + { + "epoch": 20.138977635782748, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12607 + }, + { + "epoch": 20.140575079872203, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 12608 + }, + { + "epoch": 20.142172523961662, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12609 + }, + { + "epoch": 20.143769968051117, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12610 + }, + { + "epoch": 20.145367412140576, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12611 + }, + { + "epoch": 20.14696485623003, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12612 + }, + { + "epoch": 20.14856230031949, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12613 + }, + { + "epoch": 20.150159744408946, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12614 + }, + { + "epoch": 20.1517571884984, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12615 + }, + { + "epoch": 20.15335463258786, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12616 + }, + { + "epoch": 20.154952076677315, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12617 + }, + { + "epoch": 20.156549520766774, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12618 + }, + { + "epoch": 20.15814696485623, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12619 + }, + { + "epoch": 20.159744408945688, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12620 + }, + { + "epoch": 20.161341853035143, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12621 + }, + { + "epoch": 20.162939297124602, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12622 + }, + { + "epoch": 20.164536741214057, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12623 + }, + { + "epoch": 20.166134185303516, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12624 + }, + { + "epoch": 20.16773162939297, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12625 + }, + { + "epoch": 20.169329073482427, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12626 + }, + { + "epoch": 20.170926517571885, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12627 + }, + { + "epoch": 20.17252396166134, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12628 + }, + { + "epoch": 20.1741214057508, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12629 + }, + { + "epoch": 20.175718849840255, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12630 + }, + { + "epoch": 20.177316293929714, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12631 + }, + { + "epoch": 20.17891373801917, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12632 + }, + { + "epoch": 20.180511182108628, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12633 + }, + { + "epoch": 20.182108626198083, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12634 + }, + { + "epoch": 20.18370607028754, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12635 + }, + { + "epoch": 20.185303514376997, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12636 + }, + { + "epoch": 20.186900958466452, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12637 + }, + { + "epoch": 20.18849840255591, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12638 + }, + { + "epoch": 20.190095846645367, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12639 + }, + { + "epoch": 20.191693290734825, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12640 + }, + { + "epoch": 20.19329073482428, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12641 + }, + { + "epoch": 20.19488817891374, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12642 + }, + { + "epoch": 20.196485623003195, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12643 + }, + { + "epoch": 20.19808306709265, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12644 + }, + { + "epoch": 20.19968051118211, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12645 + }, + { + "epoch": 20.201277955271564, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12646 + }, + { + "epoch": 20.202875399361023, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12647 + }, + { + "epoch": 20.20447284345048, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12648 + }, + { + "epoch": 20.206070287539937, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12649 + }, + { + "epoch": 20.207667731629392, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12650 + }, + { + "epoch": 20.20926517571885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12651 + }, + { + "epoch": 20.210862619808307, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 12652 + }, + { + "epoch": 20.212460063897762, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12653 + }, + { + "epoch": 20.21405750798722, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12654 + }, + { + "epoch": 20.215654952076676, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12655 + }, + { + "epoch": 20.217252396166135, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12656 + }, + { + "epoch": 20.21884984025559, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 12657 + }, + { + "epoch": 20.22044728434505, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12658 + }, + { + "epoch": 20.222044728434504, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12659 + }, + { + "epoch": 20.223642172523963, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12660 + }, + { + "epoch": 20.22523961661342, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12661 + }, + { + "epoch": 20.226837060702877, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12662 + }, + { + "epoch": 20.228434504792332, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12663 + }, + { + "epoch": 20.230031948881788, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12664 + }, + { + "epoch": 20.231629392971247, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12665 + }, + { + "epoch": 20.233226837060702, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12666 + }, + { + "epoch": 20.23482428115016, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12667 + }, + { + "epoch": 20.236421725239616, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12668 + }, + { + "epoch": 20.238019169329075, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12669 + }, + { + "epoch": 20.23961661341853, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12670 + }, + { + "epoch": 20.24121405750799, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 12671 + }, + { + "epoch": 20.242811501597444, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12672 + }, + { + "epoch": 20.2444089456869, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12673 + }, + { + "epoch": 20.24600638977636, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12674 + }, + { + "epoch": 20.247603833865814, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12675 + }, + { + "epoch": 20.249201277955272, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 12676 + }, + { + "epoch": 20.250798722044728, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12677 + }, + { + "epoch": 20.252396166134186, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12678 + }, + { + "epoch": 20.25399361022364, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12679 + }, + { + "epoch": 20.2555910543131, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12680 + }, + { + "epoch": 20.257188498402556, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12681 + }, + { + "epoch": 20.25878594249201, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12682 + }, + { + "epoch": 20.26038338658147, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12683 + }, + { + "epoch": 20.261980830670925, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12684 + }, + { + "epoch": 20.263578274760384, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12685 + }, + { + "epoch": 20.26517571884984, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12686 + }, + { + "epoch": 20.266773162939298, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12687 + }, + { + "epoch": 20.268370607028753, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 12688 + }, + { + "epoch": 20.269968051118212, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12689 + }, + { + "epoch": 20.271565495207668, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 12690 + }, + { + "epoch": 20.273162939297123, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12691 + }, + { + "epoch": 20.27476038338658, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12692 + }, + { + "epoch": 20.276357827476037, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12693 + }, + { + "epoch": 20.277955271565496, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12694 + }, + { + "epoch": 20.27955271565495, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12695 + }, + { + "epoch": 20.28115015974441, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12696 + }, + { + "epoch": 20.282747603833865, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12697 + }, + { + "epoch": 20.284345047923324, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12698 + }, + { + "epoch": 20.28594249201278, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 12699 + }, + { + "epoch": 20.287539936102238, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12700 + }, + { + "epoch": 20.289137380191693, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12701 + }, + { + "epoch": 20.29073482428115, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12702 + }, + { + "epoch": 20.292332268370608, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12703 + }, + { + "epoch": 20.293929712460063, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12704 + }, + { + "epoch": 20.29552715654952, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12705 + }, + { + "epoch": 20.297124600638977, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12706 + }, + { + "epoch": 20.298722044728436, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12707 + }, + { + "epoch": 20.30031948881789, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12708 + }, + { + "epoch": 20.30191693290735, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12709 + }, + { + "epoch": 20.303514376996805, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12710 + }, + { + "epoch": 20.30511182108626, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12711 + }, + { + "epoch": 20.30670926517572, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12712 + }, + { + "epoch": 20.308306709265175, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12713 + }, + { + "epoch": 20.309904153354633, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12714 + }, + { + "epoch": 20.31150159744409, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12715 + }, + { + "epoch": 20.313099041533548, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12716 + }, + { + "epoch": 20.314696485623003, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12717 + }, + { + "epoch": 20.31629392971246, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12718 + }, + { + "epoch": 20.317891373801917, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12719 + }, + { + "epoch": 20.319488817891372, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12720 + }, + { + "epoch": 20.32108626198083, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12721 + }, + { + "epoch": 20.322683706070286, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12722 + }, + { + "epoch": 20.324281150159745, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12723 + }, + { + "epoch": 20.3258785942492, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12724 + }, + { + "epoch": 20.32747603833866, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12725 + }, + { + "epoch": 20.329073482428115, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12726 + }, + { + "epoch": 20.330670926517573, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 12727 + }, + { + "epoch": 20.33226837060703, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12728 + }, + { + "epoch": 20.333865814696484, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12729 + }, + { + "epoch": 20.335463258785943, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12730 + }, + { + "epoch": 20.337060702875398, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12731 + }, + { + "epoch": 20.338658146964857, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12732 + }, + { + "epoch": 20.340255591054312, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12733 + }, + { + "epoch": 20.34185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12734 + }, + { + "epoch": 20.343450479233226, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12735 + }, + { + "epoch": 20.345047923322685, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12736 + }, + { + "epoch": 20.34664536741214, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12737 + }, + { + "epoch": 20.3482428115016, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12738 + }, + { + "epoch": 20.349840255591054, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12739 + }, + { + "epoch": 20.35143769968051, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12740 + }, + { + "epoch": 20.35303514376997, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12741 + }, + { + "epoch": 20.354632587859424, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12742 + }, + { + "epoch": 20.356230031948883, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12743 + }, + { + "epoch": 20.357827476038338, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12744 + }, + { + "epoch": 20.359424920127797, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12745 + }, + { + "epoch": 20.361022364217252, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12746 + }, + { + "epoch": 20.36261980830671, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12747 + }, + { + "epoch": 20.364217252396166, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12748 + }, + { + "epoch": 20.36581469648562, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12749 + }, + { + "epoch": 20.36741214057508, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 12750 + }, + { + "epoch": 20.369009584664536, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12751 + }, + { + "epoch": 20.370607028753994, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12752 + }, + { + "epoch": 20.37220447284345, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12753 + }, + { + "epoch": 20.37380191693291, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12754 + }, + { + "epoch": 20.375399361022364, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12755 + }, + { + "epoch": 20.376996805111823, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12756 + }, + { + "epoch": 20.378594249201278, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12757 + }, + { + "epoch": 20.380191693290733, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12758 + }, + { + "epoch": 20.381789137380192, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12759 + }, + { + "epoch": 20.383386581469647, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12760 + }, + { + "epoch": 20.384984025559106, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12761 + }, + { + "epoch": 20.38658146964856, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12762 + }, + { + "epoch": 20.38817891373802, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12763 + }, + { + "epoch": 20.389776357827476, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12764 + }, + { + "epoch": 20.391373801916934, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12765 + }, + { + "epoch": 20.39297124600639, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12766 + }, + { + "epoch": 20.394568690095845, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12767 + }, + { + "epoch": 20.396166134185304, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12768 + }, + { + "epoch": 20.39776357827476, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12769 + }, + { + "epoch": 20.399361022364218, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 12770 + }, + { + "epoch": 20.400958466453673, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12771 + }, + { + "epoch": 20.402555910543132, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12772 + }, + { + "epoch": 20.404153354632587, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12773 + }, + { + "epoch": 20.405750798722046, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12774 + }, + { + "epoch": 20.4073482428115, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12775 + }, + { + "epoch": 20.408945686900957, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12776 + }, + { + "epoch": 20.410543130990416, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 12777 + }, + { + "epoch": 20.41214057507987, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 12778 + }, + { + "epoch": 20.41373801916933, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 12779 + }, + { + "epoch": 20.415335463258785, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12780 + }, + { + "epoch": 20.416932907348244, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 12781 + }, + { + "epoch": 20.4185303514377, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12782 + }, + { + "epoch": 20.420127795527158, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12783 + }, + { + "epoch": 20.421725239616613, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12784 + }, + { + "epoch": 20.423322683706072, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12785 + }, + { + "epoch": 20.424920127795527, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12786 + }, + { + "epoch": 20.426517571884983, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12787 + }, + { + "epoch": 20.42811501597444, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12788 + }, + { + "epoch": 20.429712460063897, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 12789 + }, + { + "epoch": 20.431309904153355, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12790 + }, + { + "epoch": 20.43290734824281, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12791 + }, + { + "epoch": 20.43450479233227, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12792 + }, + { + "epoch": 20.436102236421725, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12793 + }, + { + "epoch": 20.437699680511184, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12794 + }, + { + "epoch": 20.43929712460064, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12795 + }, + { + "epoch": 20.440894568690094, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12796 + }, + { + "epoch": 20.442492012779553, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 12797 + }, + { + "epoch": 20.44408945686901, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12798 + }, + { + "epoch": 20.445686900958467, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12799 + }, + { + "epoch": 20.447284345047922, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12800 + }, + { + "epoch": 20.44888178913738, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12801 + }, + { + "epoch": 20.450479233226837, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12802 + }, + { + "epoch": 20.452076677316295, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12803 + }, + { + "epoch": 20.45367412140575, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12804 + }, + { + "epoch": 20.455271565495206, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 12805 + }, + { + "epoch": 20.456869009584665, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12806 + }, + { + "epoch": 20.45846645367412, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12807 + }, + { + "epoch": 20.46006389776358, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12808 + }, + { + "epoch": 20.461661341853034, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 12809 + }, + { + "epoch": 20.463258785942493, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 12810 + }, + { + "epoch": 20.46485623003195, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12811 + }, + { + "epoch": 20.466453674121407, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12812 + }, + { + "epoch": 20.468051118210862, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12813 + }, + { + "epoch": 20.46964856230032, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12814 + }, + { + "epoch": 20.471246006389777, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12815 + }, + { + "epoch": 20.472843450479232, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12816 + }, + { + "epoch": 20.47444089456869, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12817 + }, + { + "epoch": 20.476038338658146, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12818 + }, + { + "epoch": 20.477635782747605, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12819 + }, + { + "epoch": 20.47923322683706, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12820 + }, + { + "epoch": 20.48083067092652, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12821 + }, + { + "epoch": 20.482428115015974, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12822 + }, + { + "epoch": 20.484025559105433, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12823 + }, + { + "epoch": 20.48562300319489, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12824 + }, + { + "epoch": 20.487220447284344, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12825 + }, + { + "epoch": 20.488817891373802, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 12826 + }, + { + "epoch": 20.490415335463258, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 12827 + }, + { + "epoch": 20.492012779552716, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12828 + }, + { + "epoch": 20.49361022364217, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12829 + }, + { + "epoch": 20.49520766773163, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 12830 + }, + { + "epoch": 20.496805111821086, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12831 + }, + { + "epoch": 20.498402555910545, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12832 + }, + { + "epoch": 20.5, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12833 + }, + { + "epoch": 20.501597444089455, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12834 + }, + { + "epoch": 20.503194888178914, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12835 + }, + { + "epoch": 20.50479233226837, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12836 + }, + { + "epoch": 20.50638977635783, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12837 + }, + { + "epoch": 20.507987220447284, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12838 + }, + { + "epoch": 20.509584664536742, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12839 + }, + { + "epoch": 20.511182108626198, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12840 + }, + { + "epoch": 20.512779552715656, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12841 + }, + { + "epoch": 20.51437699680511, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12842 + }, + { + "epoch": 20.515974440894567, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12843 + }, + { + "epoch": 20.517571884984026, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12844 + }, + { + "epoch": 20.51916932907348, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 12845 + }, + { + "epoch": 20.52076677316294, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12846 + }, + { + "epoch": 20.522364217252395, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 12847 + }, + { + "epoch": 20.523961661341854, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12848 + }, + { + "epoch": 20.52555910543131, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12849 + }, + { + "epoch": 20.527156549520768, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12850 + }, + { + "epoch": 20.528753993610223, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12851 + }, + { + "epoch": 20.53035143769968, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12852 + }, + { + "epoch": 20.531948881789138, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12853 + }, + { + "epoch": 20.533546325878593, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12854 + }, + { + "epoch": 20.53514376996805, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12855 + }, + { + "epoch": 20.536741214057507, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12856 + }, + { + "epoch": 20.538338658146966, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12857 + }, + { + "epoch": 20.53993610223642, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12858 + }, + { + "epoch": 20.54153354632588, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12859 + }, + { + "epoch": 20.543130990415335, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 12860 + }, + { + "epoch": 20.544728434504794, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12861 + }, + { + "epoch": 20.54632587859425, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12862 + }, + { + "epoch": 20.547923322683705, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12863 + }, + { + "epoch": 20.549520766773163, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12864 + }, + { + "epoch": 20.55111821086262, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12865 + }, + { + "epoch": 20.552715654952078, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12866 + }, + { + "epoch": 20.554313099041533, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12867 + }, + { + "epoch": 20.55591054313099, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12868 + }, + { + "epoch": 20.557507987220447, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12869 + }, + { + "epoch": 20.559105431309906, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12870 + }, + { + "epoch": 20.56070287539936, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12871 + }, + { + "epoch": 20.562300319488816, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12872 + }, + { + "epoch": 20.563897763578275, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 12873 + }, + { + "epoch": 20.56549520766773, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12874 + }, + { + "epoch": 20.56709265175719, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12875 + }, + { + "epoch": 20.568690095846645, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12876 + }, + { + "epoch": 20.570287539936103, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12877 + }, + { + "epoch": 20.57188498402556, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12878 + }, + { + "epoch": 20.573482428115017, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12879 + }, + { + "epoch": 20.575079872204473, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12880 + }, + { + "epoch": 20.576677316293928, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12881 + }, + { + "epoch": 20.578274760383387, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12882 + }, + { + "epoch": 20.579872204472842, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12883 + }, + { + "epoch": 20.5814696485623, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12884 + }, + { + "epoch": 20.583067092651756, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12885 + }, + { + "epoch": 20.584664536741215, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12886 + }, + { + "epoch": 20.58626198083067, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 12887 + }, + { + "epoch": 20.58785942492013, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12888 + }, + { + "epoch": 20.589456869009584, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12889 + }, + { + "epoch": 20.591054313099043, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12890 + }, + { + "epoch": 20.5926517571885, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12891 + }, + { + "epoch": 20.594249201277954, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12892 + }, + { + "epoch": 20.595846645367413, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12893 + }, + { + "epoch": 20.597444089456868, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12894 + }, + { + "epoch": 20.599041533546327, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12895 + }, + { + "epoch": 20.600638977635782, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12896 + }, + { + "epoch": 20.60223642172524, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 12897 + }, + { + "epoch": 20.603833865814696, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12898 + }, + { + "epoch": 20.605431309904155, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12899 + }, + { + "epoch": 20.60702875399361, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12900 + }, + { + "epoch": 20.608626198083066, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12901 + }, + { + "epoch": 20.610223642172524, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12902 + }, + { + "epoch": 20.61182108626198, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12903 + }, + { + "epoch": 20.61341853035144, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12904 + }, + { + "epoch": 20.615015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12905 + }, + { + "epoch": 20.616613418530353, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 12906 + }, + { + "epoch": 20.618210862619808, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12907 + }, + { + "epoch": 20.619808306709267, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12908 + }, + { + "epoch": 20.621405750798722, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12909 + }, + { + "epoch": 20.623003194888177, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12910 + }, + { + "epoch": 20.624600638977636, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12911 + }, + { + "epoch": 20.62619808306709, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12912 + }, + { + "epoch": 20.62779552715655, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 12913 + }, + { + "epoch": 20.629392971246006, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12914 + }, + { + "epoch": 20.630990415335464, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12915 + }, + { + "epoch": 20.63258785942492, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 12916 + }, + { + "epoch": 20.63418530351438, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12917 + }, + { + "epoch": 20.635782747603834, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 12918 + }, + { + "epoch": 20.63738019169329, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12919 + }, + { + "epoch": 20.638977635782748, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12920 + }, + { + "epoch": 20.640575079872203, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12921 + }, + { + "epoch": 20.642172523961662, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12922 + }, + { + "epoch": 20.643769968051117, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12923 + }, + { + "epoch": 20.645367412140576, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12924 + }, + { + "epoch": 20.64696485623003, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12925 + }, + { + "epoch": 20.64856230031949, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12926 + }, + { + "epoch": 20.650159744408946, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12927 + }, + { + "epoch": 20.6517571884984, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12928 + }, + { + "epoch": 20.65335463258786, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12929 + }, + { + "epoch": 20.654952076677315, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12930 + }, + { + "epoch": 20.656549520766774, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12931 + }, + { + "epoch": 20.65814696485623, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12932 + }, + { + "epoch": 20.659744408945688, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 12933 + }, + { + "epoch": 20.661341853035143, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12934 + }, + { + "epoch": 20.662939297124602, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12935 + }, + { + "epoch": 20.664536741214057, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12936 + }, + { + "epoch": 20.666134185303516, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12937 + }, + { + "epoch": 20.66773162939297, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12938 + }, + { + "epoch": 20.669329073482427, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12939 + }, + { + "epoch": 20.670926517571885, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 12940 + }, + { + "epoch": 20.67252396166134, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12941 + }, + { + "epoch": 20.6741214057508, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 12942 + }, + { + "epoch": 20.675718849840255, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12943 + }, + { + "epoch": 20.677316293929714, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12944 + }, + { + "epoch": 20.67891373801917, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12945 + }, + { + "epoch": 20.680511182108628, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12946 + }, + { + "epoch": 20.682108626198083, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12947 + }, + { + "epoch": 20.68370607028754, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12948 + }, + { + "epoch": 20.685303514376997, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12949 + }, + { + "epoch": 20.686900958466452, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12950 + }, + { + "epoch": 20.68849840255591, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12951 + }, + { + "epoch": 20.690095846645367, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12952 + }, + { + "epoch": 20.691693290734825, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12953 + }, + { + "epoch": 20.69329073482428, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12954 + }, + { + "epoch": 20.69488817891374, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12955 + }, + { + "epoch": 20.696485623003195, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12956 + }, + { + "epoch": 20.69808306709265, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12957 + }, + { + "epoch": 20.69968051118211, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12958 + }, + { + "epoch": 20.701277955271564, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12959 + }, + { + "epoch": 20.702875399361023, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12960 + }, + { + "epoch": 20.70447284345048, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12961 + }, + { + "epoch": 20.706070287539937, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12962 + }, + { + "epoch": 20.707667731629392, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12963 + }, + { + "epoch": 20.70926517571885, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12964 + }, + { + "epoch": 20.710862619808307, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12965 + }, + { + "epoch": 20.712460063897765, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12966 + }, + { + "epoch": 20.71405750798722, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12967 + }, + { + "epoch": 20.715654952076676, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12968 + }, + { + "epoch": 20.717252396166135, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12969 + }, + { + "epoch": 20.71884984025559, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12970 + }, + { + "epoch": 20.72044728434505, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12971 + }, + { + "epoch": 20.722044728434504, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12972 + }, + { + "epoch": 20.723642172523963, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12973 + }, + { + "epoch": 20.72523961661342, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 12974 + }, + { + "epoch": 20.726837060702877, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12975 + }, + { + "epoch": 20.728434504792332, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12976 + }, + { + "epoch": 20.730031948881788, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12977 + }, + { + "epoch": 20.731629392971247, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12978 + }, + { + "epoch": 20.733226837060702, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12979 + }, + { + "epoch": 20.73482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12980 + }, + { + "epoch": 20.736421725239616, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12981 + }, + { + "epoch": 20.738019169329075, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 12982 + }, + { + "epoch": 20.73961661341853, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12983 + }, + { + "epoch": 20.74121405750799, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 12984 + }, + { + "epoch": 20.742811501597444, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12985 + }, + { + "epoch": 20.7444089456869, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12986 + }, + { + "epoch": 20.74600638977636, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12987 + }, + { + "epoch": 20.747603833865814, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12988 + }, + { + "epoch": 20.749201277955272, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12989 + }, + { + "epoch": 20.750798722044728, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12990 + }, + { + "epoch": 20.752396166134186, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12991 + }, + { + "epoch": 20.75399361022364, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12992 + }, + { + "epoch": 20.7555910543131, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12993 + }, + { + "epoch": 20.757188498402556, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12994 + }, + { + "epoch": 20.75878594249201, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12995 + }, + { + "epoch": 20.76038338658147, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12996 + }, + { + "epoch": 20.761980830670925, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12997 + }, + { + "epoch": 20.763578274760384, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12998 + }, + { + "epoch": 20.76517571884984, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12999 + }, + { + "epoch": 20.766773162939298, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 13000 + } + ], + "logging_steps": 1.0, + "max_steps": 40064, + "num_input_tokens_seen": 0, + "num_train_epochs": 64, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 8.983976018746278e+18, + "train_batch_size": 256, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-13000/training_args.bin b/checkpoint-13000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..f0fb42436d77a51b765ce5b2ee9e58f75d8bebed --- /dev/null +++ b/checkpoint-13000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3892ab69705b96d80c03f1e7acdafde89ecdd24967bc603de8e73f84852bf5df +size 6008 diff --git a/checkpoint-13500/config.json b/checkpoint-13500/config.json new file mode 100644 index 0000000000000000000000000000000000000000..60a257df017ccff177cc7be1151b89b4bd290c54 --- /dev/null +++ b/checkpoint-13500/config.json @@ -0,0 +1,79 @@ +{ + "ar_steps": 1, + "architectures": [ + "DiffVLMDiffusion" + ], + "attention_dropout": 0.0, + "bos_token_id": 151643, + "condition_layer": -1, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "image_token_id": 151655, + "img_cross_attention_dim": 2048, + "img_diffuser_depth": 2, + "img_ffn_dim_multiplier": null, + "img_hidden_size": 1536, + "img_multiple_of": 256, + "img_norm_eps": 1e-05, + "img_num_attention_heads": 12, + "img_num_kv_heads": 12, + "img_qk_norm": true, + "in_channels": 32, + "initializer_range": 0.02, + "inject_img_diffuser": false, + "input_size": 32, + "intermediate_size": 8960, + "layer_group_size": 7, + "layerwise_start_idx": 0, + "lora_alpha": 128, + "lora_bias": "none", + "lora_dropout": 0.05, + "lora_enable": false, + "lora_r": 64, + "max_position_embeddings": 32768, + "max_window_layers": 28, + "model_type": "qwen2_vl", + "non_linearity": 1, + "norm_elementwise_affine": true, + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "patch_size": 2, + "repa_coeff": 0.1, + "repa_layers": "2", + "repa_shared": false, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sample_size": 128, + "sampling_steps": 28, + "sliding_window": null, + "tie_word_embeddings": true, + "torch_dtype": "bfloat16", + "transformers_version": "4.47.0", + "use_cache": true, + "use_repa": false, + "use_residual_attn": true, + "use_sliding_window": false, + "vae_path": "mit-han-lab/dc-ae-f32c32-sana-1.1-diffusers", + "video_token_id": 151656, + "vision_config": { + "hidden_size": 1536, + "in_chans": 3, + "model_type": "qwen2_vl", + "spatial_patch_size": 14 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 151936 +} diff --git a/checkpoint-13500/generation_config.json b/checkpoint-13500/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0e3d465da90cb0ab59d8ba7babdefb0e88fbfa6b --- /dev/null +++ b/checkpoint-13500/generation_config.json @@ -0,0 +1,6 @@ +{ + "_from_model_config": true, + "bos_token_id": 151643, + "eos_token_id": 151645, + "transformers_version": "4.47.0" +} diff --git a/checkpoint-13500/model.safetensors b/checkpoint-13500/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a23d66a0b3274c3aa659cb6fba523d59b84f5c1a --- /dev/null +++ b/checkpoint-13500/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84eda8d8258151d7397c29e73d3a2b0f65cd871868a45f081d4d86755aaf8deb +size 4539487822 diff --git a/checkpoint-13500/optimizer.pt b/checkpoint-13500/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..238191fca9b22a3268539ae4906d0e32edb9864f --- /dev/null +++ b/checkpoint-13500/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6860c80bbc792f87367353a20a4d4a2275c6d8fce3d78ec9a51f9b4435453151 +size 6632345482 diff --git a/checkpoint-13500/rng_state_0.pth b/checkpoint-13500/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..793dbe2508e42a752963ff6c6cecd42e09513f9d --- /dev/null +++ b/checkpoint-13500/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8cc3f5bf46ecddf7fe837e6ff32f08d757e26eb5230bc8b381583d319368fc5d +size 15984 diff --git a/checkpoint-13500/rng_state_1.pth b/checkpoint-13500/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..75bf45120cc4174ddd3b3bba104a14b1dd59f0e3 --- /dev/null +++ b/checkpoint-13500/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d2f4d9a79c44d532f0807d53056a1f498578b6df59cfd8859bb91d64220f3ea +size 15984 diff --git a/checkpoint-13500/rng_state_2.pth b/checkpoint-13500/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..76f85dcae41eb45182bb56ab1923390925b6f9bd --- /dev/null +++ b/checkpoint-13500/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c06b83f46f2c476a53fa4a0d6e4bcbe197cfd550df4522e24490542c783ffbdf +size 15984 diff --git a/checkpoint-13500/rng_state_3.pth b/checkpoint-13500/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..d91d7503924282b064a76737bf2e039fbadb18db --- /dev/null +++ b/checkpoint-13500/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23ce89feb84d5cb6a1d198bf6fc003f8ea96cac4cad897e88535de5b531e5715 +size 15984 diff --git a/checkpoint-13500/rng_state_4.pth b/checkpoint-13500/rng_state_4.pth new file mode 100644 index 0000000000000000000000000000000000000000..35e6995e07e7c0cce347c1fe3dbad8602e11f8f3 --- /dev/null +++ b/checkpoint-13500/rng_state_4.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de5fe31f526ebea59a78cb7abd830295f2ef0d46c538c91bce0b253ce6e8f1d6 +size 15984 diff --git a/checkpoint-13500/rng_state_5.pth b/checkpoint-13500/rng_state_5.pth new file mode 100644 index 0000000000000000000000000000000000000000..7ab822ef98c007de5214399dccaf11dcf3d47c34 --- /dev/null +++ b/checkpoint-13500/rng_state_5.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca4863cccbffc4e615799076aaa77605d155a32ea9b447c79393f52ec22935f2 +size 15984 diff --git a/checkpoint-13500/rng_state_6.pth b/checkpoint-13500/rng_state_6.pth new file mode 100644 index 0000000000000000000000000000000000000000..9176148d499aa7b69cf7f1b6359dde51bae652c4 --- /dev/null +++ b/checkpoint-13500/rng_state_6.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e778ea24c8551ab774dbf961e35a61cd8c3a249c86c02c5001020dc102218fb8 +size 15984 diff --git a/checkpoint-13500/rng_state_7.pth b/checkpoint-13500/rng_state_7.pth new file mode 100644 index 0000000000000000000000000000000000000000..a80410c4e1b76d79ba9fe70c431d2572aba85d33 --- /dev/null +++ b/checkpoint-13500/rng_state_7.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ce028eba857001ef259849e32f0b4fb8f29a63a96aa0b53edb89ba441b0e15f9 +size 15984 diff --git a/checkpoint-13500/scheduler.pt b/checkpoint-13500/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..664d9d3512ea60661a622606ca4c8e7ffb3127c8 --- /dev/null +++ b/checkpoint-13500/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d3bd3d33f03c30839702ba58b03206b7185403e472a45db995e8ab0f1188ba0c +size 1064 diff --git a/checkpoint-13500/trainer_state.json b/checkpoint-13500/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..fe9a365b6d66a5edefe8d3a42cad27aec7f4ea14 --- /dev/null +++ b/checkpoint-13500/trainer_state.json @@ -0,0 +1,94533 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 21.56549520766773, + "eval_steps": 500, + "global_step": 13500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001597444089456869, + "grad_norm": 3.390625, + "learning_rate": 0.0005, + "loss": 3.6303, + "step": 1 + }, + { + "epoch": 0.003194888178913738, + "grad_norm": 17.5, + "learning_rate": 0.0005, + "loss": 5.5153, + "step": 2 + }, + { + "epoch": 0.004792332268370607, + "grad_norm": 7.84375, + "learning_rate": 0.0005, + "loss": 2.7811, + "step": 3 + }, + { + "epoch": 0.006389776357827476, + "grad_norm": 7.65625, + "learning_rate": 0.0005, + "loss": 2.8695, + "step": 4 + }, + { + "epoch": 0.007987220447284345, + "grad_norm": 3.734375, + "learning_rate": 0.0005, + "loss": 2.3246, + "step": 5 + }, + { + "epoch": 0.009584664536741214, + "grad_norm": 3.640625, + "learning_rate": 0.0005, + "loss": 2.0843, + "step": 6 + }, + { + "epoch": 0.011182108626198083, + "grad_norm": 1.015625, + "learning_rate": 0.0005, + "loss": 1.7376, + "step": 7 + }, + { + "epoch": 0.012779552715654952, + "grad_norm": 4.65625, + "learning_rate": 0.0005, + "loss": 1.863, + "step": 8 + }, + { + "epoch": 0.01437699680511182, + "grad_norm": 1.3828125, + "learning_rate": 0.0005, + "loss": 1.7027, + "step": 9 + }, + { + "epoch": 0.01597444089456869, + "grad_norm": 4.90625, + "learning_rate": 0.0005, + "loss": 1.8852, + "step": 10 + }, + { + "epoch": 0.01757188498402556, + "grad_norm": 4.46875, + "learning_rate": 0.0005, + "loss": 1.8597, + "step": 11 + }, + { + "epoch": 0.019169329073482427, + "grad_norm": 1.015625, + "learning_rate": 0.0005, + "loss": 1.6276, + "step": 12 + }, + { + "epoch": 0.020766773162939296, + "grad_norm": 2.640625, + "learning_rate": 0.0005, + "loss": 1.6508, + "step": 13 + }, + { + "epoch": 0.022364217252396165, + "grad_norm": 2.578125, + "learning_rate": 0.0005, + "loss": 1.6636, + "step": 14 + }, + { + "epoch": 0.023961661341853034, + "grad_norm": 0.703125, + "learning_rate": 0.0005, + "loss": 1.5623, + "step": 15 + }, + { + "epoch": 0.025559105431309903, + "grad_norm": 1.34375, + "learning_rate": 0.0005, + "loss": 1.5314, + "step": 16 + }, + { + "epoch": 0.027156549520766772, + "grad_norm": 1.0546875, + "learning_rate": 0.0005, + "loss": 1.4973, + "step": 17 + }, + { + "epoch": 0.02875399361022364, + "grad_norm": 0.412109375, + "learning_rate": 0.0005, + "loss": 1.4629, + "step": 18 + }, + { + "epoch": 0.03035143769968051, + "grad_norm": 0.640625, + "learning_rate": 0.0005, + "loss": 1.445, + "step": 19 + }, + { + "epoch": 0.03194888178913738, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.4098, + "step": 20 + }, + { + "epoch": 0.03354632587859425, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.3997, + "step": 21 + }, + { + "epoch": 0.03514376996805112, + "grad_norm": 0.376953125, + "learning_rate": 0.0005, + "loss": 1.401, + "step": 22 + }, + { + "epoch": 0.036741214057507986, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.378, + "step": 23 + }, + { + "epoch": 0.038338658146964855, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.3717, + "step": 24 + }, + { + "epoch": 0.039936102236421724, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.3598, + "step": 25 + }, + { + "epoch": 0.04153354632587859, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.3511, + "step": 26 + }, + { + "epoch": 0.04313099041533546, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.3562, + "step": 27 + }, + { + "epoch": 0.04472843450479233, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.3464, + "step": 28 + }, + { + "epoch": 0.0463258785942492, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.3385, + "step": 29 + }, + { + "epoch": 0.04792332268370607, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.3363, + "step": 30 + }, + { + "epoch": 0.04952076677316294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.3294, + "step": 31 + }, + { + "epoch": 0.051118210862619806, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.3254, + "step": 32 + }, + { + "epoch": 0.052715654952076675, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.315, + "step": 33 + }, + { + "epoch": 0.054313099041533544, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.3153, + "step": 34 + }, + { + "epoch": 0.05591054313099041, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.3213, + "step": 35 + }, + { + "epoch": 0.05750798722044728, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.309, + "step": 36 + }, + { + "epoch": 0.05910543130990415, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.3139, + "step": 37 + }, + { + "epoch": 0.06070287539936102, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.3085, + "step": 38 + }, + { + "epoch": 0.06230031948881789, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.2991, + "step": 39 + }, + { + "epoch": 0.06389776357827476, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.3113, + "step": 40 + }, + { + "epoch": 0.06549520766773163, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.3159, + "step": 41 + }, + { + "epoch": 0.0670926517571885, + "grad_norm": 0.2373046875, + "learning_rate": 0.0005, + "loss": 1.295, + "step": 42 + }, + { + "epoch": 0.06869009584664537, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.2948, + "step": 43 + }, + { + "epoch": 0.07028753993610223, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.303, + "step": 44 + }, + { + "epoch": 0.07188498402555911, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.3037, + "step": 45 + }, + { + "epoch": 0.07348242811501597, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.29, + "step": 46 + }, + { + "epoch": 0.07507987220447285, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.2882, + "step": 47 + }, + { + "epoch": 0.07667731629392971, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.2929, + "step": 48 + }, + { + "epoch": 0.07827476038338659, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.2836, + "step": 49 + }, + { + "epoch": 0.07987220447284345, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.2844, + "step": 50 + }, + { + "epoch": 0.08146964856230032, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.2863, + "step": 51 + }, + { + "epoch": 0.08306709265175719, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.2772, + "step": 52 + }, + { + "epoch": 0.08466453674121406, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.2777, + "step": 53 + }, + { + "epoch": 0.08626198083067092, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.2746, + "step": 54 + }, + { + "epoch": 0.0878594249201278, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.2789, + "step": 55 + }, + { + "epoch": 0.08945686900958466, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.2659, + "step": 56 + }, + { + "epoch": 0.09105431309904154, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.2669, + "step": 57 + }, + { + "epoch": 0.0926517571884984, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.2729, + "step": 58 + }, + { + "epoch": 0.09424920127795527, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.269, + "step": 59 + }, + { + "epoch": 0.09584664536741214, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.2707, + "step": 60 + }, + { + "epoch": 0.09744408945686901, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.2673, + "step": 61 + }, + { + "epoch": 0.09904153354632587, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.2676, + "step": 62 + }, + { + "epoch": 0.10063897763578275, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.2581, + "step": 63 + }, + { + "epoch": 0.10223642172523961, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.2608, + "step": 64 + }, + { + "epoch": 0.10383386581469649, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.2584, + "step": 65 + }, + { + "epoch": 0.10543130990415335, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.2597, + "step": 66 + }, + { + "epoch": 0.10702875399361023, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.2582, + "step": 67 + }, + { + "epoch": 0.10862619808306709, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.2544, + "step": 68 + }, + { + "epoch": 0.11022364217252396, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.2545, + "step": 69 + }, + { + "epoch": 0.11182108626198083, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.2556, + "step": 70 + }, + { + "epoch": 0.1134185303514377, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.2508, + "step": 71 + }, + { + "epoch": 0.11501597444089456, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.2514, + "step": 72 + }, + { + "epoch": 0.11661341853035144, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.2509, + "step": 73 + }, + { + "epoch": 0.1182108626198083, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.2508, + "step": 74 + }, + { + "epoch": 0.11980830670926518, + "grad_norm": 0.3125, + "learning_rate": 0.0005, + "loss": 1.2531, + "step": 75 + }, + { + "epoch": 0.12140575079872204, + "grad_norm": 0.5390625, + "learning_rate": 0.0005, + "loss": 1.2642, + "step": 76 + }, + { + "epoch": 0.12300319488817892, + "grad_norm": 0.9453125, + "learning_rate": 0.0005, + "loss": 1.2656, + "step": 77 + }, + { + "epoch": 0.12460063897763578, + "grad_norm": 0.9609375, + "learning_rate": 0.0005, + "loss": 1.3036, + "step": 78 + }, + { + "epoch": 0.12619808306709265, + "grad_norm": 0.498046875, + "learning_rate": 0.0005, + "loss": 1.2778, + "step": 79 + }, + { + "epoch": 0.12779552715654952, + "grad_norm": 1.03125, + "learning_rate": 0.0005, + "loss": 1.2695, + "step": 80 + }, + { + "epoch": 0.12939297124600638, + "grad_norm": 0.578125, + "learning_rate": 0.0005, + "loss": 1.2628, + "step": 81 + }, + { + "epoch": 0.13099041533546327, + "grad_norm": 0.39453125, + "learning_rate": 0.0005, + "loss": 1.2666, + "step": 82 + }, + { + "epoch": 0.13258785942492013, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.2624, + "step": 83 + }, + { + "epoch": 0.134185303514377, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.2575, + "step": 84 + }, + { + "epoch": 0.13578274760383385, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.253, + "step": 85 + }, + { + "epoch": 0.13738019169329074, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.2532, + "step": 86 + }, + { + "epoch": 0.1389776357827476, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.2481, + "step": 87 + }, + { + "epoch": 0.14057507987220447, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.2461, + "step": 88 + }, + { + "epoch": 0.14217252396166133, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.2395, + "step": 89 + }, + { + "epoch": 0.14376996805111822, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.2436, + "step": 90 + }, + { + "epoch": 0.14536741214057508, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.2458, + "step": 91 + }, + { + "epoch": 0.14696485623003194, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.2398, + "step": 92 + }, + { + "epoch": 0.1485623003194888, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.2416, + "step": 93 + }, + { + "epoch": 0.1501597444089457, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.2359, + "step": 94 + }, + { + "epoch": 0.15175718849840256, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.2355, + "step": 95 + }, + { + "epoch": 0.15335463258785942, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.2457, + "step": 96 + }, + { + "epoch": 0.15495207667731628, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.2436, + "step": 97 + }, + { + "epoch": 0.15654952076677317, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.2392, + "step": 98 + }, + { + "epoch": 0.15814696485623003, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.2442, + "step": 99 + }, + { + "epoch": 0.1597444089456869, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.2386, + "step": 100 + }, + { + "epoch": 0.16134185303514376, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.2429, + "step": 101 + }, + { + "epoch": 0.16293929712460065, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.2238, + "step": 102 + }, + { + "epoch": 0.1645367412140575, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.238, + "step": 103 + }, + { + "epoch": 0.16613418530351437, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.2289, + "step": 104 + }, + { + "epoch": 0.16773162939297126, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.2384, + "step": 105 + }, + { + "epoch": 0.16932907348242812, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.2244, + "step": 106 + }, + { + "epoch": 0.17092651757188498, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.2248, + "step": 107 + }, + { + "epoch": 0.17252396166134185, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.2196, + "step": 108 + }, + { + "epoch": 0.17412140575079874, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.2263, + "step": 109 + }, + { + "epoch": 0.1757188498402556, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.2254, + "step": 110 + }, + { + "epoch": 0.17731629392971246, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.2265, + "step": 111 + }, + { + "epoch": 0.17891373801916932, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.2184, + "step": 112 + }, + { + "epoch": 0.1805111821086262, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.224, + "step": 113 + }, + { + "epoch": 0.18210862619808307, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.2227, + "step": 114 + }, + { + "epoch": 0.18370607028753994, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.2152, + "step": 115 + }, + { + "epoch": 0.1853035143769968, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.2217, + "step": 116 + }, + { + "epoch": 0.1869009584664537, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.2197, + "step": 117 + }, + { + "epoch": 0.18849840255591055, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.215, + "step": 118 + }, + { + "epoch": 0.1900958466453674, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.2158, + "step": 119 + }, + { + "epoch": 0.19169329073482427, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.219, + "step": 120 + }, + { + "epoch": 0.19329073482428116, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.2184, + "step": 121 + }, + { + "epoch": 0.19488817891373802, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.2155, + "step": 122 + }, + { + "epoch": 0.1964856230031949, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.218, + "step": 123 + }, + { + "epoch": 0.19808306709265175, + "grad_norm": 0.5703125, + "learning_rate": 0.0005, + "loss": 1.2236, + "step": 124 + }, + { + "epoch": 0.19968051118210864, + "grad_norm": 1.2265625, + "learning_rate": 0.0005, + "loss": 1.2427, + "step": 125 + }, + { + "epoch": 0.2012779552715655, + "grad_norm": 0.83203125, + "learning_rate": 0.0005, + "loss": 1.2307, + "step": 126 + }, + { + "epoch": 0.20287539936102236, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.2214, + "step": 127 + }, + { + "epoch": 0.20447284345047922, + "grad_norm": 0.416015625, + "learning_rate": 0.0005, + "loss": 1.2235, + "step": 128 + }, + { + "epoch": 0.20607028753993611, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.2175, + "step": 129 + }, + { + "epoch": 0.20766773162939298, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.2159, + "step": 130 + }, + { + "epoch": 0.20926517571884984, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.2158, + "step": 131 + }, + { + "epoch": 0.2108626198083067, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.2192, + "step": 132 + }, + { + "epoch": 0.2124600638977636, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.2172, + "step": 133 + }, + { + "epoch": 0.21405750798722045, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.2174, + "step": 134 + }, + { + "epoch": 0.21565495207667731, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.214, + "step": 135 + }, + { + "epoch": 0.21725239616613418, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.2048, + "step": 136 + }, + { + "epoch": 0.21884984025559107, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.215, + "step": 137 + }, + { + "epoch": 0.22044728434504793, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.2119, + "step": 138 + }, + { + "epoch": 0.2220447284345048, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.2138, + "step": 139 + }, + { + "epoch": 0.22364217252396165, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.208, + "step": 140 + }, + { + "epoch": 0.22523961661341854, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.2094, + "step": 141 + }, + { + "epoch": 0.2268370607028754, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.1994, + "step": 142 + }, + { + "epoch": 0.22843450479233227, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.2043, + "step": 143 + }, + { + "epoch": 0.23003194888178913, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.2063, + "step": 144 + }, + { + "epoch": 0.23162939297124602, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.2027, + "step": 145 + }, + { + "epoch": 0.23322683706070288, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.2019, + "step": 146 + }, + { + "epoch": 0.23482428115015974, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.2065, + "step": 147 + }, + { + "epoch": 0.2364217252396166, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.2055, + "step": 148 + }, + { + "epoch": 0.2380191693290735, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.2036, + "step": 149 + }, + { + "epoch": 0.23961661341853036, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.2094, + "step": 150 + }, + { + "epoch": 0.24121405750798722, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.207, + "step": 151 + }, + { + "epoch": 0.24281150159744408, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.2027, + "step": 152 + }, + { + "epoch": 0.24440894568690097, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1961, + "step": 153 + }, + { + "epoch": 0.24600638977635783, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1957, + "step": 154 + }, + { + "epoch": 0.2476038338658147, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.2024, + "step": 155 + }, + { + "epoch": 0.24920127795527156, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.1964, + "step": 156 + }, + { + "epoch": 0.2507987220447284, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.2021, + "step": 157 + }, + { + "epoch": 0.2523961661341853, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1907, + "step": 158 + }, + { + "epoch": 0.2539936102236422, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.1935, + "step": 159 + }, + { + "epoch": 0.25559105431309903, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1975, + "step": 160 + }, + { + "epoch": 0.2571884984025559, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1983, + "step": 161 + }, + { + "epoch": 0.25878594249201275, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1941, + "step": 162 + }, + { + "epoch": 0.26038338658146964, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1872, + "step": 163 + }, + { + "epoch": 0.26198083067092653, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.2013, + "step": 164 + }, + { + "epoch": 0.26357827476038337, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1971, + "step": 165 + }, + { + "epoch": 0.26517571884984026, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.2061, + "step": 166 + }, + { + "epoch": 0.26677316293929715, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1886, + "step": 167 + }, + { + "epoch": 0.268370607028754, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1969, + "step": 168 + }, + { + "epoch": 0.26996805111821087, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.194, + "step": 169 + }, + { + "epoch": 0.2715654952076677, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1917, + "step": 170 + }, + { + "epoch": 0.2731629392971246, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1935, + "step": 171 + }, + { + "epoch": 0.2747603833865815, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1936, + "step": 172 + }, + { + "epoch": 0.2763578274760383, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1906, + "step": 173 + }, + { + "epoch": 0.2779552715654952, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1991, + "step": 174 + }, + { + "epoch": 0.2795527156549521, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.1884, + "step": 175 + }, + { + "epoch": 0.28115015974440893, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.1936, + "step": 176 + }, + { + "epoch": 0.2827476038338658, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1833, + "step": 177 + }, + { + "epoch": 0.28434504792332266, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.187, + "step": 178 + }, + { + "epoch": 0.28594249201277955, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.1916, + "step": 179 + }, + { + "epoch": 0.28753993610223644, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.1884, + "step": 180 + }, + { + "epoch": 0.28913738019169327, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1856, + "step": 181 + }, + { + "epoch": 0.29073482428115016, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1918, + "step": 182 + }, + { + "epoch": 0.29233226837060705, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1908, + "step": 183 + }, + { + "epoch": 0.2939297124600639, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1946, + "step": 184 + }, + { + "epoch": 0.2955271565495208, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1867, + "step": 185 + }, + { + "epoch": 0.2971246006389776, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1944, + "step": 186 + }, + { + "epoch": 0.2987220447284345, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1837, + "step": 187 + }, + { + "epoch": 0.3003194888178914, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1921, + "step": 188 + }, + { + "epoch": 0.3019169329073482, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1931, + "step": 189 + }, + { + "epoch": 0.3035143769968051, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1835, + "step": 190 + }, + { + "epoch": 0.305111821086262, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.1885, + "step": 191 + }, + { + "epoch": 0.30670926517571884, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.1809, + "step": 192 + }, + { + "epoch": 0.3083067092651757, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.1812, + "step": 193 + }, + { + "epoch": 0.30990415335463256, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.1909, + "step": 194 + }, + { + "epoch": 0.31150159744408945, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1774, + "step": 195 + }, + { + "epoch": 0.31309904153354634, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.1777, + "step": 196 + }, + { + "epoch": 0.3146964856230032, + "grad_norm": 0.578125, + "learning_rate": 0.0005, + "loss": 1.1873, + "step": 197 + }, + { + "epoch": 0.31629392971246006, + "grad_norm": 2.203125, + "learning_rate": 0.0005, + "loss": 1.2371, + "step": 198 + }, + { + "epoch": 0.31789137380191695, + "grad_norm": 0.828125, + "learning_rate": 0.0005, + "loss": 1.2053, + "step": 199 + }, + { + "epoch": 0.3194888178913738, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.1954, + "step": 200 + }, + { + "epoch": 0.3210862619808307, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.1942, + "step": 201 + }, + { + "epoch": 0.3226837060702875, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.192, + "step": 202 + }, + { + "epoch": 0.3242811501597444, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1883, + "step": 203 + }, + { + "epoch": 0.3258785942492013, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1916, + "step": 204 + }, + { + "epoch": 0.3274760383386581, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1866, + "step": 205 + }, + { + "epoch": 0.329073482428115, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1961, + "step": 206 + }, + { + "epoch": 0.3306709265175719, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1856, + "step": 207 + }, + { + "epoch": 0.33226837060702874, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.1813, + "step": 208 + }, + { + "epoch": 0.33386581469648563, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1809, + "step": 209 + }, + { + "epoch": 0.3354632587859425, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.18, + "step": 210 + }, + { + "epoch": 0.33706070287539935, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1808, + "step": 211 + }, + { + "epoch": 0.33865814696485624, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.1839, + "step": 212 + }, + { + "epoch": 0.3402555910543131, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.1841, + "step": 213 + }, + { + "epoch": 0.34185303514376997, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1847, + "step": 214 + }, + { + "epoch": 0.34345047923322686, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.1796, + "step": 215 + }, + { + "epoch": 0.3450479233226837, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1798, + "step": 216 + }, + { + "epoch": 0.3466453674121406, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1732, + "step": 217 + }, + { + "epoch": 0.34824281150159747, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1827, + "step": 218 + }, + { + "epoch": 0.3498402555910543, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1789, + "step": 219 + }, + { + "epoch": 0.3514376996805112, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1765, + "step": 220 + }, + { + "epoch": 0.35303514376996803, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1797, + "step": 221 + }, + { + "epoch": 0.3546325878594249, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1816, + "step": 222 + }, + { + "epoch": 0.3562300319488818, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1773, + "step": 223 + }, + { + "epoch": 0.35782747603833864, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.173, + "step": 224 + }, + { + "epoch": 0.35942492012779553, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1787, + "step": 225 + }, + { + "epoch": 0.3610223642172524, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.187, + "step": 226 + }, + { + "epoch": 0.36261980830670926, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.178, + "step": 227 + }, + { + "epoch": 0.36421725239616615, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1788, + "step": 228 + }, + { + "epoch": 0.365814696485623, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1764, + "step": 229 + }, + { + "epoch": 0.36741214057507987, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1771, + "step": 230 + }, + { + "epoch": 0.36900958466453676, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1738, + "step": 231 + }, + { + "epoch": 0.3706070287539936, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1827, + "step": 232 + }, + { + "epoch": 0.3722044728434505, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.1733, + "step": 233 + }, + { + "epoch": 0.3738019169329074, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1692, + "step": 234 + }, + { + "epoch": 0.3753993610223642, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.1743, + "step": 235 + }, + { + "epoch": 0.3769968051118211, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1739, + "step": 236 + }, + { + "epoch": 0.37859424920127793, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1723, + "step": 237 + }, + { + "epoch": 0.3801916932907348, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1715, + "step": 238 + }, + { + "epoch": 0.3817891373801917, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1725, + "step": 239 + }, + { + "epoch": 0.38338658146964855, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1695, + "step": 240 + }, + { + "epoch": 0.38498402555910544, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1688, + "step": 241 + }, + { + "epoch": 0.3865814696485623, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1654, + "step": 242 + }, + { + "epoch": 0.38817891373801916, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1729, + "step": 243 + }, + { + "epoch": 0.38977635782747605, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1694, + "step": 244 + }, + { + "epoch": 0.3913738019169329, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1674, + "step": 245 + }, + { + "epoch": 0.3929712460063898, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.169, + "step": 246 + }, + { + "epoch": 0.39456869009584666, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.177, + "step": 247 + }, + { + "epoch": 0.3961661341853035, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1646, + "step": 248 + }, + { + "epoch": 0.3977635782747604, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1635, + "step": 249 + }, + { + "epoch": 0.3993610223642173, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1784, + "step": 250 + }, + { + "epoch": 0.4009584664536741, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1592, + "step": 251 + }, + { + "epoch": 0.402555910543131, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1734, + "step": 252 + }, + { + "epoch": 0.40415335463258784, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1726, + "step": 253 + }, + { + "epoch": 0.4057507987220447, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1565, + "step": 254 + }, + { + "epoch": 0.4073482428115016, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.167, + "step": 255 + }, + { + "epoch": 0.40894568690095845, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1711, + "step": 256 + }, + { + "epoch": 0.41054313099041534, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1626, + "step": 257 + }, + { + "epoch": 0.41214057507987223, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.171, + "step": 258 + }, + { + "epoch": 0.41373801916932906, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.1686, + "step": 259 + }, + { + "epoch": 0.41533546325878595, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1719, + "step": 260 + }, + { + "epoch": 0.4169329073482428, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1659, + "step": 261 + }, + { + "epoch": 0.4185303514376997, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.1659, + "step": 262 + }, + { + "epoch": 0.42012779552715657, + "grad_norm": 0.427734375, + "learning_rate": 0.0005, + "loss": 1.1693, + "step": 263 + }, + { + "epoch": 0.4217252396166134, + "grad_norm": 1.4765625, + "learning_rate": 0.0005, + "loss": 1.1913, + "step": 264 + }, + { + "epoch": 0.4233226837060703, + "grad_norm": 2.140625, + "learning_rate": 0.0005, + "loss": 1.2367, + "step": 265 + }, + { + "epoch": 0.4249201277955272, + "grad_norm": 0.66015625, + "learning_rate": 0.0005, + "loss": 1.1874, + "step": 266 + }, + { + "epoch": 0.426517571884984, + "grad_norm": 2.4375, + "learning_rate": 0.0005, + "loss": 1.2668, + "step": 267 + }, + { + "epoch": 0.4281150159744409, + "grad_norm": 0.81640625, + "learning_rate": 0.0005, + "loss": 1.2247, + "step": 268 + }, + { + "epoch": 0.42971246006389774, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.2079, + "step": 269 + }, + { + "epoch": 0.43130990415335463, + "grad_norm": 0.37890625, + "learning_rate": 0.0005, + "loss": 1.2128, + "step": 270 + }, + { + "epoch": 0.4329073482428115, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.2106, + "step": 271 + }, + { + "epoch": 0.43450479233226835, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.2115, + "step": 272 + }, + { + "epoch": 0.43610223642172524, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.2015, + "step": 273 + }, + { + "epoch": 0.43769968051118213, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1962, + "step": 274 + }, + { + "epoch": 0.43929712460063897, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.194, + "step": 275 + }, + { + "epoch": 0.44089456869009586, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1941, + "step": 276 + }, + { + "epoch": 0.4424920127795527, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.1882, + "step": 277 + }, + { + "epoch": 0.4440894568690096, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1838, + "step": 278 + }, + { + "epoch": 0.44568690095846647, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1878, + "step": 279 + }, + { + "epoch": 0.4472843450479233, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.1924, + "step": 280 + }, + { + "epoch": 0.4488817891373802, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.1842, + "step": 281 + }, + { + "epoch": 0.4504792332268371, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1834, + "step": 282 + }, + { + "epoch": 0.4520766773162939, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.176, + "step": 283 + }, + { + "epoch": 0.4536741214057508, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1806, + "step": 284 + }, + { + "epoch": 0.45527156549520764, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1785, + "step": 285 + }, + { + "epoch": 0.45686900958466453, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.178, + "step": 286 + }, + { + "epoch": 0.4584664536741214, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1791, + "step": 287 + }, + { + "epoch": 0.46006389776357826, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.176, + "step": 288 + }, + { + "epoch": 0.46166134185303515, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1777, + "step": 289 + }, + { + "epoch": 0.46325878594249204, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1672, + "step": 290 + }, + { + "epoch": 0.46485623003194887, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1717, + "step": 291 + }, + { + "epoch": 0.46645367412140576, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1705, + "step": 292 + }, + { + "epoch": 0.4680511182108626, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1731, + "step": 293 + }, + { + "epoch": 0.4696485623003195, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1703, + "step": 294 + }, + { + "epoch": 0.4712460063897764, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1692, + "step": 295 + }, + { + "epoch": 0.4728434504792332, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1634, + "step": 296 + }, + { + "epoch": 0.4744408945686901, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1687, + "step": 297 + }, + { + "epoch": 0.476038338658147, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1705, + "step": 298 + }, + { + "epoch": 0.4776357827476038, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1623, + "step": 299 + }, + { + "epoch": 0.4792332268370607, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.1619, + "step": 300 + }, + { + "epoch": 0.48083067092651754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1682, + "step": 301 + }, + { + "epoch": 0.48242811501597443, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1659, + "step": 302 + }, + { + "epoch": 0.4840255591054313, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1653, + "step": 303 + }, + { + "epoch": 0.48562300319488816, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1651, + "step": 304 + }, + { + "epoch": 0.48722044728434505, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1712, + "step": 305 + }, + { + "epoch": 0.48881789137380194, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1699, + "step": 306 + }, + { + "epoch": 0.4904153354632588, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 307 + }, + { + "epoch": 0.49201277955271566, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1681, + "step": 308 + }, + { + "epoch": 0.4936102236421725, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.163, + "step": 309 + }, + { + "epoch": 0.4952076677316294, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1573, + "step": 310 + }, + { + "epoch": 0.4968051118210863, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1596, + "step": 311 + }, + { + "epoch": 0.4984025559105431, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1613, + "step": 312 + }, + { + "epoch": 0.5, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1677, + "step": 313 + }, + { + "epoch": 0.5015974440894568, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 314 + }, + { + "epoch": 0.5031948881789138, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.155, + "step": 315 + }, + { + "epoch": 0.5047923322683706, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.173, + "step": 316 + }, + { + "epoch": 0.5063897763578274, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1627, + "step": 317 + }, + { + "epoch": 0.5079872204472844, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 318 + }, + { + "epoch": 0.5095846645367412, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1573, + "step": 319 + }, + { + "epoch": 0.5111821086261981, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.16, + "step": 320 + }, + { + "epoch": 0.512779552715655, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1642, + "step": 321 + }, + { + "epoch": 0.5143769968051118, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1613, + "step": 322 + }, + { + "epoch": 0.5159744408945687, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1608, + "step": 323 + }, + { + "epoch": 0.5175718849840255, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1655, + "step": 324 + }, + { + "epoch": 0.5191693290734825, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1624, + "step": 325 + }, + { + "epoch": 0.5207667731629393, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1572, + "step": 326 + }, + { + "epoch": 0.5223642172523961, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1549, + "step": 327 + }, + { + "epoch": 0.5239616613418531, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1572, + "step": 328 + }, + { + "epoch": 0.5255591054313099, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1581, + "step": 329 + }, + { + "epoch": 0.5271565495207667, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.1651, + "step": 330 + }, + { + "epoch": 0.5287539936102237, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 331 + }, + { + "epoch": 0.5303514376996805, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1526, + "step": 332 + }, + { + "epoch": 0.5319488817891374, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1547, + "step": 333 + }, + { + "epoch": 0.5335463258785943, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1605, + "step": 334 + }, + { + "epoch": 0.5351437699680511, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1533, + "step": 335 + }, + { + "epoch": 0.536741214057508, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.1524, + "step": 336 + }, + { + "epoch": 0.5383386581469649, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.152, + "step": 337 + }, + { + "epoch": 0.5399361022364217, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1595, + "step": 338 + }, + { + "epoch": 0.5415335463258786, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1675, + "step": 339 + }, + { + "epoch": 0.5431309904153354, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.1577, + "step": 340 + }, + { + "epoch": 0.5447284345047924, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1598, + "step": 341 + }, + { + "epoch": 0.5463258785942492, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1569, + "step": 342 + }, + { + "epoch": 0.547923322683706, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1548, + "step": 343 + }, + { + "epoch": 0.549520766773163, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1543, + "step": 344 + }, + { + "epoch": 0.5511182108626198, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1538, + "step": 345 + }, + { + "epoch": 0.5527156549520766, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.1539, + "step": 346 + }, + { + "epoch": 0.5543130990415336, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1618, + "step": 347 + }, + { + "epoch": 0.5559105431309904, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1558, + "step": 348 + }, + { + "epoch": 0.5575079872204473, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1596, + "step": 349 + }, + { + "epoch": 0.5591054313099042, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.1602, + "step": 350 + }, + { + "epoch": 0.560702875399361, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1512, + "step": 351 + }, + { + "epoch": 0.5623003194888179, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.1569, + "step": 352 + }, + { + "epoch": 0.5638977635782748, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.1529, + "step": 353 + }, + { + "epoch": 0.5654952076677316, + "grad_norm": 0.419921875, + "learning_rate": 0.0005, + "loss": 1.1527, + "step": 354 + }, + { + "epoch": 0.5670926517571885, + "grad_norm": 1.3125, + "learning_rate": 0.0005, + "loss": 1.1742, + "step": 355 + }, + { + "epoch": 0.5686900958466453, + "grad_norm": 1.5625, + "learning_rate": 0.0005, + "loss": 1.2129, + "step": 356 + }, + { + "epoch": 0.5702875399361023, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.1702, + "step": 357 + }, + { + "epoch": 0.5718849840255591, + "grad_norm": 0.8203125, + "learning_rate": 0.0005, + "loss": 1.1896, + "step": 358 + }, + { + "epoch": 0.5734824281150159, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.1831, + "step": 359 + }, + { + "epoch": 0.5750798722044729, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.1721, + "step": 360 + }, + { + "epoch": 0.5766773162939297, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.1761, + "step": 361 + }, + { + "epoch": 0.5782747603833865, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.174, + "step": 362 + }, + { + "epoch": 0.5798722044728435, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.1736, + "step": 363 + }, + { + "epoch": 0.5814696485623003, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.1748, + "step": 364 + }, + { + "epoch": 0.5830670926517572, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.1559, + "step": 365 + }, + { + "epoch": 0.5846645367412141, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1733, + "step": 366 + }, + { + "epoch": 0.5862619808306709, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.1667, + "step": 367 + }, + { + "epoch": 0.5878594249201278, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1627, + "step": 368 + }, + { + "epoch": 0.5894568690095847, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1644, + "step": 369 + }, + { + "epoch": 0.5910543130990416, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1565, + "step": 370 + }, + { + "epoch": 0.5926517571884984, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1622, + "step": 371 + }, + { + "epoch": 0.5942492012779552, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1589, + "step": 372 + }, + { + "epoch": 0.5958466453674122, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1561, + "step": 373 + }, + { + "epoch": 0.597444089456869, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1561, + "step": 374 + }, + { + "epoch": 0.5990415335463258, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1579, + "step": 375 + }, + { + "epoch": 0.6006389776357828, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.161, + "step": 376 + }, + { + "epoch": 0.6022364217252396, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1574, + "step": 377 + }, + { + "epoch": 0.6038338658146964, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1564, + "step": 378 + }, + { + "epoch": 0.6054313099041534, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1547, + "step": 379 + }, + { + "epoch": 0.6070287539936102, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1603, + "step": 380 + }, + { + "epoch": 0.6086261980830671, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1552, + "step": 381 + }, + { + "epoch": 0.610223642172524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1622, + "step": 382 + }, + { + "epoch": 0.6118210862619808, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1624, + "step": 383 + }, + { + "epoch": 0.6134185303514377, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1534, + "step": 384 + }, + { + "epoch": 0.6150159744408946, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1592, + "step": 385 + }, + { + "epoch": 0.6166134185303515, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1529, + "step": 386 + }, + { + "epoch": 0.6182108626198083, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.1488, + "step": 387 + }, + { + "epoch": 0.6198083067092651, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1631, + "step": 388 + }, + { + "epoch": 0.6214057507987221, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1545, + "step": 389 + }, + { + "epoch": 0.6230031948881789, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1534, + "step": 390 + }, + { + "epoch": 0.6246006389776357, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1502, + "step": 391 + }, + { + "epoch": 0.6261980830670927, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.16, + "step": 392 + }, + { + "epoch": 0.6277955271565495, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1615, + "step": 393 + }, + { + "epoch": 0.6293929712460063, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.152, + "step": 394 + }, + { + "epoch": 0.6309904153354633, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1425, + "step": 395 + }, + { + "epoch": 0.6325878594249201, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1518, + "step": 396 + }, + { + "epoch": 0.634185303514377, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1514, + "step": 397 + }, + { + "epoch": 0.6357827476038339, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1516, + "step": 398 + }, + { + "epoch": 0.6373801916932907, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1415, + "step": 399 + }, + { + "epoch": 0.6389776357827476, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1537, + "step": 400 + }, + { + "epoch": 0.6405750798722045, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.154, + "step": 401 + }, + { + "epoch": 0.6421725239616614, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1529, + "step": 402 + }, + { + "epoch": 0.6437699680511182, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.144, + "step": 403 + }, + { + "epoch": 0.645367412140575, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1508, + "step": 404 + }, + { + "epoch": 0.646964856230032, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1469, + "step": 405 + }, + { + "epoch": 0.6485623003194888, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.154, + "step": 406 + }, + { + "epoch": 0.6501597444089456, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1507, + "step": 407 + }, + { + "epoch": 0.6517571884984026, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1458, + "step": 408 + }, + { + "epoch": 0.6533546325878594, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1458, + "step": 409 + }, + { + "epoch": 0.6549520766773163, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1489, + "step": 410 + }, + { + "epoch": 0.6565495207667732, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1467, + "step": 411 + }, + { + "epoch": 0.65814696485623, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1494, + "step": 412 + }, + { + "epoch": 0.6597444089456869, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1538, + "step": 413 + }, + { + "epoch": 0.6613418530351438, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1482, + "step": 414 + }, + { + "epoch": 0.6629392971246006, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1509, + "step": 415 + }, + { + "epoch": 0.6645367412140575, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1487, + "step": 416 + }, + { + "epoch": 0.6661341853035144, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1535, + "step": 417 + }, + { + "epoch": 0.6677316293929713, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1516, + "step": 418 + }, + { + "epoch": 0.6693290734824281, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1538, + "step": 419 + }, + { + "epoch": 0.670926517571885, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1428, + "step": 420 + }, + { + "epoch": 0.6725239616613419, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 421 + }, + { + "epoch": 0.6741214057507987, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1504, + "step": 422 + }, + { + "epoch": 0.6757188498402555, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1445, + "step": 423 + }, + { + "epoch": 0.6773162939297125, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 424 + }, + { + "epoch": 0.6789137380191693, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1373, + "step": 425 + }, + { + "epoch": 0.6805111821086262, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1476, + "step": 426 + }, + { + "epoch": 0.6821086261980831, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1445, + "step": 427 + }, + { + "epoch": 0.6837060702875399, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1463, + "step": 428 + }, + { + "epoch": 0.6853035143769968, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1443, + "step": 429 + }, + { + "epoch": 0.6869009584664537, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 430 + }, + { + "epoch": 0.6884984025559105, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 431 + }, + { + "epoch": 0.6900958466453674, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1411, + "step": 432 + }, + { + "epoch": 0.6916932907348243, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1499, + "step": 433 + }, + { + "epoch": 0.6932907348242812, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1478, + "step": 434 + }, + { + "epoch": 0.694888178913738, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 435 + }, + { + "epoch": 0.6964856230031949, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.145, + "step": 436 + }, + { + "epoch": 0.6980830670926518, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1501, + "step": 437 + }, + { + "epoch": 0.6996805111821086, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1392, + "step": 438 + }, + { + "epoch": 0.7012779552715654, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1467, + "step": 439 + }, + { + "epoch": 0.7028753993610224, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1399, + "step": 440 + }, + { + "epoch": 0.7044728434504792, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1474, + "step": 441 + }, + { + "epoch": 0.7060702875399361, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1481, + "step": 442 + }, + { + "epoch": 0.707667731629393, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1448, + "step": 443 + }, + { + "epoch": 0.7092651757188498, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1421, + "step": 444 + }, + { + "epoch": 0.7108626198083067, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1441, + "step": 445 + }, + { + "epoch": 0.7124600638977636, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1408, + "step": 446 + }, + { + "epoch": 0.7140575079872205, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1486, + "step": 447 + }, + { + "epoch": 0.7156549520766773, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1442, + "step": 448 + }, + { + "epoch": 0.7172523961661342, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.1455, + "step": 449 + }, + { + "epoch": 0.7188498402555911, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.1424, + "step": 450 + }, + { + "epoch": 0.7204472843450479, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1436, + "step": 451 + }, + { + "epoch": 0.7220447284345048, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1496, + "step": 452 + }, + { + "epoch": 0.7236421725239617, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.1472, + "step": 453 + }, + { + "epoch": 0.7252396166134185, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1452, + "step": 454 + }, + { + "epoch": 0.7268370607028753, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1463, + "step": 455 + }, + { + "epoch": 0.7284345047923323, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1481, + "step": 456 + }, + { + "epoch": 0.7300319488817891, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 457 + }, + { + "epoch": 0.731629392971246, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1487, + "step": 458 + }, + { + "epoch": 0.7332268370607029, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 459 + }, + { + "epoch": 0.7348242811501597, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1434, + "step": 460 + }, + { + "epoch": 0.7364217252396166, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1468, + "step": 461 + }, + { + "epoch": 0.7380191693290735, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 462 + }, + { + "epoch": 0.7396166134185304, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1406, + "step": 463 + }, + { + "epoch": 0.7412140575079872, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1433, + "step": 464 + }, + { + "epoch": 0.7428115015974441, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1493, + "step": 465 + }, + { + "epoch": 0.744408945686901, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1488, + "step": 466 + }, + { + "epoch": 0.7460063897763578, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1449, + "step": 467 + }, + { + "epoch": 0.7476038338658147, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1379, + "step": 468 + }, + { + "epoch": 0.7492012779552716, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1415, + "step": 469 + }, + { + "epoch": 0.7507987220447284, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.1439, + "step": 470 + }, + { + "epoch": 0.7523961661341853, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.1436, + "step": 471 + }, + { + "epoch": 0.7539936102236422, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1468, + "step": 472 + }, + { + "epoch": 0.755591054313099, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1449, + "step": 473 + }, + { + "epoch": 0.7571884984025559, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1393, + "step": 474 + }, + { + "epoch": 0.7587859424920128, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1366, + "step": 475 + }, + { + "epoch": 0.7603833865814696, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.1453, + "step": 476 + }, + { + "epoch": 0.7619808306709265, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1398, + "step": 477 + }, + { + "epoch": 0.7635782747603834, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 478 + }, + { + "epoch": 0.7651757188498403, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1498, + "step": 479 + }, + { + "epoch": 0.7667731629392971, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1414, + "step": 480 + }, + { + "epoch": 0.768370607028754, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.143, + "step": 481 + }, + { + "epoch": 0.7699680511182109, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 482 + }, + { + "epoch": 0.7715654952076677, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1431, + "step": 483 + }, + { + "epoch": 0.7731629392971247, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.1494, + "step": 484 + }, + { + "epoch": 0.7747603833865815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1452, + "step": 485 + }, + { + "epoch": 0.7763578274760383, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1342, + "step": 486 + }, + { + "epoch": 0.7779552715654952, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 487 + }, + { + "epoch": 0.7795527156549521, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1376, + "step": 488 + }, + { + "epoch": 0.7811501597444089, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1412, + "step": 489 + }, + { + "epoch": 0.7827476038338658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1347, + "step": 490 + }, + { + "epoch": 0.7843450479233227, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.1423, + "step": 491 + }, + { + "epoch": 0.7859424920127795, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1385, + "step": 492 + }, + { + "epoch": 0.7875399361022364, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.1343, + "step": 493 + }, + { + "epoch": 0.7891373801916933, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1404, + "step": 494 + }, + { + "epoch": 0.7907348242811502, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.1346, + "step": 495 + }, + { + "epoch": 0.792332268370607, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1386, + "step": 496 + }, + { + "epoch": 0.7939297124600639, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1481, + "step": 497 + }, + { + "epoch": 0.7955271565495208, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1382, + "step": 498 + }, + { + "epoch": 0.7971246006389776, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1337, + "step": 499 + }, + { + "epoch": 0.7987220447284346, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1437, + "step": 500 + }, + { + "epoch": 0.8003194888178914, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.139, + "step": 501 + }, + { + "epoch": 0.8019169329073482, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1396, + "step": 502 + }, + { + "epoch": 0.8035143769968051, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1437, + "step": 503 + }, + { + "epoch": 0.805111821086262, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1411, + "step": 504 + }, + { + "epoch": 0.8067092651757188, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.1381, + "step": 505 + }, + { + "epoch": 0.8083067092651757, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1341, + "step": 506 + }, + { + "epoch": 0.8099041533546326, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1376, + "step": 507 + }, + { + "epoch": 0.8115015974440895, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.1377, + "step": 508 + }, + { + "epoch": 0.8130990415335463, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1346, + "step": 509 + }, + { + "epoch": 0.8146964856230032, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 510 + }, + { + "epoch": 0.8162939297124601, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 511 + }, + { + "epoch": 0.8178913738019169, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1397, + "step": 512 + }, + { + "epoch": 0.8194888178913738, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1431, + "step": 513 + }, + { + "epoch": 0.8210862619808307, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1336, + "step": 514 + }, + { + "epoch": 0.8226837060702875, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 515 + }, + { + "epoch": 0.8242811501597445, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.1342, + "step": 516 + }, + { + "epoch": 0.8258785942492013, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1366, + "step": 517 + }, + { + "epoch": 0.8274760383386581, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1398, + "step": 518 + }, + { + "epoch": 0.829073482428115, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1421, + "step": 519 + }, + { + "epoch": 0.8306709265175719, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.1385, + "step": 520 + }, + { + "epoch": 0.8322683706070287, + "grad_norm": 0.2255859375, + "learning_rate": 0.0005, + "loss": 1.1344, + "step": 521 + }, + { + "epoch": 0.8338658146964856, + "grad_norm": 0.466796875, + "learning_rate": 0.0005, + "loss": 1.1325, + "step": 522 + }, + { + "epoch": 0.8354632587859425, + "grad_norm": 1.5859375, + "learning_rate": 0.0005, + "loss": 1.1728, + "step": 523 + }, + { + "epoch": 0.8370607028753994, + "grad_norm": 1.8125, + "learning_rate": 0.0005, + "loss": 1.1896, + "step": 524 + }, + { + "epoch": 0.8386581469648562, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.1537, + "step": 525 + }, + { + "epoch": 0.8402555910543131, + "grad_norm": 1.796875, + "learning_rate": 0.0005, + "loss": 1.2008, + "step": 526 + }, + { + "epoch": 0.84185303514377, + "grad_norm": 0.82421875, + "learning_rate": 0.0005, + "loss": 1.1803, + "step": 527 + }, + { + "epoch": 0.8434504792332268, + "grad_norm": 0.859375, + "learning_rate": 0.0005, + "loss": 1.1776, + "step": 528 + }, + { + "epoch": 0.8450479233226837, + "grad_norm": 0.408203125, + "learning_rate": 0.0005, + "loss": 1.1669, + "step": 529 + }, + { + "epoch": 0.8466453674121406, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.1631, + "step": 530 + }, + { + "epoch": 0.8482428115015974, + "grad_norm": 0.4921875, + "learning_rate": 0.0005, + "loss": 1.1665, + "step": 531 + }, + { + "epoch": 0.8498402555910544, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.149, + "step": 532 + }, + { + "epoch": 0.8514376996805112, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1572, + "step": 533 + }, + { + "epoch": 0.853035143769968, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.1569, + "step": 534 + }, + { + "epoch": 0.854632587859425, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1521, + "step": 535 + }, + { + "epoch": 0.8562300319488818, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1544, + "step": 536 + }, + { + "epoch": 0.8578274760383386, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.1528, + "step": 537 + }, + { + "epoch": 0.8594249201277955, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.1478, + "step": 538 + }, + { + "epoch": 0.8610223642172524, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.1565, + "step": 539 + }, + { + "epoch": 0.8626198083067093, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1381, + "step": 540 + }, + { + "epoch": 0.8642172523961661, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.14, + "step": 541 + }, + { + "epoch": 0.865814696485623, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1485, + "step": 542 + }, + { + "epoch": 0.8674121405750799, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1363, + "step": 543 + }, + { + "epoch": 0.8690095846645367, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.1419, + "step": 544 + }, + { + "epoch": 0.8706070287539937, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1422, + "step": 545 + }, + { + "epoch": 0.8722044728434505, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1422, + "step": 546 + }, + { + "epoch": 0.8738019169329073, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1393, + "step": 547 + }, + { + "epoch": 0.8753993610223643, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1409, + "step": 548 + }, + { + "epoch": 0.8769968051118211, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1389, + "step": 549 + }, + { + "epoch": 0.8785942492012779, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1445, + "step": 550 + }, + { + "epoch": 0.8801916932907349, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1483, + "step": 551 + }, + { + "epoch": 0.8817891373801917, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.136, + "step": 552 + }, + { + "epoch": 0.8833865814696485, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1369, + "step": 553 + }, + { + "epoch": 0.8849840255591054, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1393, + "step": 554 + }, + { + "epoch": 0.8865814696485623, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1409, + "step": 555 + }, + { + "epoch": 0.8881789137380192, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1412, + "step": 556 + }, + { + "epoch": 0.889776357827476, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1403, + "step": 557 + }, + { + "epoch": 0.8913738019169329, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1394, + "step": 558 + }, + { + "epoch": 0.8929712460063898, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.142, + "step": 559 + }, + { + "epoch": 0.8945686900958466, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1377, + "step": 560 + }, + { + "epoch": 0.8961661341853036, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1366, + "step": 561 + }, + { + "epoch": 0.8977635782747604, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.132, + "step": 562 + }, + { + "epoch": 0.8993610223642172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 563 + }, + { + "epoch": 0.9009584664536742, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1332, + "step": 564 + }, + { + "epoch": 0.902555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1398, + "step": 565 + }, + { + "epoch": 0.9041533546325878, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 566 + }, + { + "epoch": 0.9057507987220448, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1395, + "step": 567 + }, + { + "epoch": 0.9073482428115016, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1408, + "step": 568 + }, + { + "epoch": 0.9089456869009584, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1376, + "step": 569 + }, + { + "epoch": 0.9105431309904153, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1392, + "step": 570 + }, + { + "epoch": 0.9121405750798722, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1396, + "step": 571 + }, + { + "epoch": 0.9137380191693291, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1382, + "step": 572 + }, + { + "epoch": 0.9153354632587859, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 573 + }, + { + "epoch": 0.9169329073482428, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1336, + "step": 574 + }, + { + "epoch": 0.9185303514376997, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 575 + }, + { + "epoch": 0.9201277955271565, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1451, + "step": 576 + }, + { + "epoch": 0.9217252396166135, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1305, + "step": 577 + }, + { + "epoch": 0.9233226837060703, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 578 + }, + { + "epoch": 0.9249201277955271, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1411, + "step": 579 + }, + { + "epoch": 0.9265175718849841, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1365, + "step": 580 + }, + { + "epoch": 0.9281150159744409, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 581 + }, + { + "epoch": 0.9297124600638977, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1318, + "step": 582 + }, + { + "epoch": 0.9313099041533547, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1374, + "step": 583 + }, + { + "epoch": 0.9329073482428115, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.137, + "step": 584 + }, + { + "epoch": 0.9345047923322684, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1343, + "step": 585 + }, + { + "epoch": 0.9361022364217252, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1339, + "step": 586 + }, + { + "epoch": 0.9376996805111821, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1307, + "step": 587 + }, + { + "epoch": 0.939297124600639, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1364, + "step": 588 + }, + { + "epoch": 0.9408945686900958, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1364, + "step": 589 + }, + { + "epoch": 0.9424920127795527, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 590 + }, + { + "epoch": 0.9440894568690096, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1352, + "step": 591 + }, + { + "epoch": 0.9456869009584664, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1364, + "step": 592 + }, + { + "epoch": 0.9472843450479234, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1382, + "step": 593 + }, + { + "epoch": 0.9488817891373802, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1333, + "step": 594 + }, + { + "epoch": 0.950479233226837, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1314, + "step": 595 + }, + { + "epoch": 0.952076677316294, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 596 + }, + { + "epoch": 0.9536741214057508, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1303, + "step": 597 + }, + { + "epoch": 0.9552715654952076, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 598 + }, + { + "epoch": 0.9568690095846646, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1291, + "step": 599 + }, + { + "epoch": 0.9584664536741214, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1397, + "step": 600 + }, + { + "epoch": 0.9600638977635783, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.1302, + "step": 601 + }, + { + "epoch": 0.9616613418530351, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.129, + "step": 602 + }, + { + "epoch": 0.963258785942492, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1343, + "step": 603 + }, + { + "epoch": 0.9648562300319489, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 604 + }, + { + "epoch": 0.9664536741214057, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1325, + "step": 605 + }, + { + "epoch": 0.9680511182108626, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1328, + "step": 606 + }, + { + "epoch": 0.9696485623003195, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1349, + "step": 607 + }, + { + "epoch": 0.9712460063897763, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 608 + }, + { + "epoch": 0.9728434504792333, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 609 + }, + { + "epoch": 0.9744408945686901, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 610 + }, + { + "epoch": 0.9760383386581469, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1284, + "step": 611 + }, + { + "epoch": 0.9776357827476039, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1313, + "step": 612 + }, + { + "epoch": 0.9792332268370607, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 613 + }, + { + "epoch": 0.9808306709265175, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1303, + "step": 614 + }, + { + "epoch": 0.9824281150159745, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1289, + "step": 615 + }, + { + "epoch": 0.9840255591054313, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1326, + "step": 616 + }, + { + "epoch": 0.9856230031948882, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1346, + "step": 617 + }, + { + "epoch": 0.987220447284345, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 618 + }, + { + "epoch": 0.9888178913738019, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1292, + "step": 619 + }, + { + "epoch": 0.9904153354632588, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1349, + "step": 620 + }, + { + "epoch": 0.9920127795527156, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1301, + "step": 621 + }, + { + "epoch": 0.9936102236421726, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1256, + "step": 622 + }, + { + "epoch": 0.9952076677316294, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1375, + "step": 623 + }, + { + "epoch": 0.9968051118210862, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 624 + }, + { + "epoch": 0.9984025559105432, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 625 + }, + { + "epoch": 1.0, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 626 + }, + { + "epoch": 1.0015974440894568, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.131, + "step": 627 + }, + { + "epoch": 1.0031948881789137, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1355, + "step": 628 + }, + { + "epoch": 1.0047923322683705, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 629 + }, + { + "epoch": 1.0063897763578276, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1337, + "step": 630 + }, + { + "epoch": 1.0079872204472844, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1331, + "step": 631 + }, + { + "epoch": 1.0095846645367412, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1246, + "step": 632 + }, + { + "epoch": 1.011182108626198, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1276, + "step": 633 + }, + { + "epoch": 1.012779552715655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.1298, + "step": 634 + }, + { + "epoch": 1.0143769968051117, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1247, + "step": 635 + }, + { + "epoch": 1.0159744408945688, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1303, + "step": 636 + }, + { + "epoch": 1.0175718849840256, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 637 + }, + { + "epoch": 1.0191693290734825, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 638 + }, + { + "epoch": 1.0207667731629393, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 639 + }, + { + "epoch": 1.0223642172523961, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 640 + }, + { + "epoch": 1.023961661341853, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 641 + }, + { + "epoch": 1.0255591054313098, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1305, + "step": 642 + }, + { + "epoch": 1.0271565495207668, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 643 + }, + { + "epoch": 1.0287539936102237, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1328, + "step": 644 + }, + { + "epoch": 1.0303514376996805, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 645 + }, + { + "epoch": 1.0319488817891374, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 646 + }, + { + "epoch": 1.0335463258785942, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 647 + }, + { + "epoch": 1.035143769968051, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.1291, + "step": 648 + }, + { + "epoch": 1.036741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1218, + "step": 649 + }, + { + "epoch": 1.038338658146965, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 650 + }, + { + "epoch": 1.0399361022364217, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 651 + }, + { + "epoch": 1.0415335463258786, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 652 + }, + { + "epoch": 1.0431309904153354, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1283, + "step": 653 + }, + { + "epoch": 1.0447284345047922, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.132, + "step": 654 + }, + { + "epoch": 1.0463258785942493, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.128, + "step": 655 + }, + { + "epoch": 1.0479233226837061, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 656 + }, + { + "epoch": 1.049520766773163, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1214, + "step": 657 + }, + { + "epoch": 1.0511182108626198, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1275, + "step": 658 + }, + { + "epoch": 1.0527156549520766, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 659 + }, + { + "epoch": 1.0543130990415335, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1321, + "step": 660 + }, + { + "epoch": 1.0559105431309903, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 661 + }, + { + "epoch": 1.0575079872204474, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 662 + }, + { + "epoch": 1.0591054313099042, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1292, + "step": 663 + }, + { + "epoch": 1.060702875399361, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1309, + "step": 664 + }, + { + "epoch": 1.0623003194888179, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 665 + }, + { + "epoch": 1.0638977635782747, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1296, + "step": 666 + }, + { + "epoch": 1.0654952076677315, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1286, + "step": 667 + }, + { + "epoch": 1.0670926517571886, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 668 + }, + { + "epoch": 1.0686900958466454, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1271, + "step": 669 + }, + { + "epoch": 1.0702875399361023, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1238, + "step": 670 + }, + { + "epoch": 1.071884984025559, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1271, + "step": 671 + }, + { + "epoch": 1.073482428115016, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.1297, + "step": 672 + }, + { + "epoch": 1.0750798722044728, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.1168, + "step": 673 + }, + { + "epoch": 1.0766773162939298, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 674 + }, + { + "epoch": 1.0782747603833867, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 675 + }, + { + "epoch": 1.0798722044728435, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 676 + }, + { + "epoch": 1.0814696485623003, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1294, + "step": 677 + }, + { + "epoch": 1.0830670926517572, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1248, + "step": 678 + }, + { + "epoch": 1.084664536741214, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1313, + "step": 679 + }, + { + "epoch": 1.0862619808306708, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 680 + }, + { + "epoch": 1.0878594249201279, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 681 + }, + { + "epoch": 1.0894568690095847, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 682 + }, + { + "epoch": 1.0910543130990416, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1255, + "step": 683 + }, + { + "epoch": 1.0926517571884984, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 684 + }, + { + "epoch": 1.0942492012779552, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 685 + }, + { + "epoch": 1.095846645367412, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1322, + "step": 686 + }, + { + "epoch": 1.097444089456869, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 687 + }, + { + "epoch": 1.099041533546326, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1314, + "step": 688 + }, + { + "epoch": 1.1006389776357828, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.128, + "step": 689 + }, + { + "epoch": 1.1022364217252396, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 690 + }, + { + "epoch": 1.1038338658146964, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1262, + "step": 691 + }, + { + "epoch": 1.1054313099041533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 692 + }, + { + "epoch": 1.1070287539936103, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1208, + "step": 693 + }, + { + "epoch": 1.1086261980830672, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.123, + "step": 694 + }, + { + "epoch": 1.110223642172524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1287, + "step": 695 + }, + { + "epoch": 1.1118210862619808, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 696 + }, + { + "epoch": 1.1134185303514377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 697 + }, + { + "epoch": 1.1150159744408945, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1319, + "step": 698 + }, + { + "epoch": 1.1166134185303513, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 699 + }, + { + "epoch": 1.1182108626198084, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1214, + "step": 700 + }, + { + "epoch": 1.1198083067092652, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1261, + "step": 701 + }, + { + "epoch": 1.121405750798722, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1327, + "step": 702 + }, + { + "epoch": 1.123003194888179, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 703 + }, + { + "epoch": 1.1246006389776357, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 704 + }, + { + "epoch": 1.1261980830670926, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1237, + "step": 705 + }, + { + "epoch": 1.1277955271565494, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1306, + "step": 706 + }, + { + "epoch": 1.1293929712460065, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1246, + "step": 707 + }, + { + "epoch": 1.1309904153354633, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1263, + "step": 708 + }, + { + "epoch": 1.1325878594249201, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 709 + }, + { + "epoch": 1.134185303514377, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.1217, + "step": 710 + }, + { + "epoch": 1.1357827476038338, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1193, + "step": 711 + }, + { + "epoch": 1.1373801916932909, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1253, + "step": 712 + }, + { + "epoch": 1.1389776357827477, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1311, + "step": 713 + }, + { + "epoch": 1.1405750798722045, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 714 + }, + { + "epoch": 1.1421725239616614, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1302, + "step": 715 + }, + { + "epoch": 1.1437699680511182, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1235, + "step": 716 + }, + { + "epoch": 1.145367412140575, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 717 + }, + { + "epoch": 1.1469648562300319, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.1266, + "step": 718 + }, + { + "epoch": 1.148562300319489, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.125, + "step": 719 + }, + { + "epoch": 1.1501597444089458, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 720 + }, + { + "epoch": 1.1517571884984026, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1209, + "step": 721 + }, + { + "epoch": 1.1533546325878594, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.1327, + "step": 722 + }, + { + "epoch": 1.1549520766773163, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1228, + "step": 723 + }, + { + "epoch": 1.156549520766773, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.1277, + "step": 724 + }, + { + "epoch": 1.15814696485623, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 725 + }, + { + "epoch": 1.159744408945687, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 726 + }, + { + "epoch": 1.1613418530351438, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.1213, + "step": 727 + }, + { + "epoch": 1.1629392971246006, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 728 + }, + { + "epoch": 1.1645367412140575, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.1277, + "step": 729 + }, + { + "epoch": 1.1661341853035143, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.1318, + "step": 730 + }, + { + "epoch": 1.1677316293929714, + "grad_norm": 0.33203125, + "learning_rate": 0.0005, + "loss": 1.1302, + "step": 731 + }, + { + "epoch": 1.1693290734824282, + "grad_norm": 1.046875, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 732 + }, + { + "epoch": 1.170926517571885, + "grad_norm": 2.453125, + "learning_rate": 0.0005, + "loss": 1.234, + "step": 733 + }, + { + "epoch": 1.1725239616613419, + "grad_norm": 0.7578125, + "learning_rate": 0.0005, + "loss": 1.1564, + "step": 734 + }, + { + "epoch": 1.1741214057507987, + "grad_norm": 1.328125, + "learning_rate": 0.0005, + "loss": 1.1934, + "step": 735 + }, + { + "epoch": 1.1757188498402555, + "grad_norm": 0.71875, + "learning_rate": 0.0005, + "loss": 1.1747, + "step": 736 + }, + { + "epoch": 1.1773162939297124, + "grad_norm": 0.80078125, + "learning_rate": 0.0005, + "loss": 1.1646, + "step": 737 + }, + { + "epoch": 1.1789137380191694, + "grad_norm": 0.921875, + "learning_rate": 0.0005, + "loss": 1.2053, + "step": 738 + }, + { + "epoch": 1.1805111821086263, + "grad_norm": 0.93359375, + "learning_rate": 0.0005, + "loss": 1.1907, + "step": 739 + }, + { + "epoch": 1.182108626198083, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.1779, + "step": 740 + }, + { + "epoch": 1.18370607028754, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.168, + "step": 741 + }, + { + "epoch": 1.1853035143769968, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.1682, + "step": 742 + }, + { + "epoch": 1.1869009584664536, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1609, + "step": 743 + }, + { + "epoch": 1.1884984025559104, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.1537, + "step": 744 + }, + { + "epoch": 1.1900958466453675, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.1637, + "step": 745 + }, + { + "epoch": 1.1916932907348243, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.16, + "step": 746 + }, + { + "epoch": 1.1932907348242812, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.156, + "step": 747 + }, + { + "epoch": 1.194888178913738, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1511, + "step": 748 + }, + { + "epoch": 1.1964856230031948, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.1443, + "step": 749 + }, + { + "epoch": 1.1980830670926517, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1474, + "step": 750 + }, + { + "epoch": 1.1996805111821087, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.1467, + "step": 751 + }, + { + "epoch": 1.2012779552715656, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1418, + "step": 752 + }, + { + "epoch": 1.2028753993610224, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1434, + "step": 753 + }, + { + "epoch": 1.2044728434504792, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1402, + "step": 754 + }, + { + "epoch": 1.206070287539936, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1406, + "step": 755 + }, + { + "epoch": 1.207667731629393, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1351, + "step": 756 + }, + { + "epoch": 1.20926517571885, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1358, + "step": 757 + }, + { + "epoch": 1.2108626198083068, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1402, + "step": 758 + }, + { + "epoch": 1.2124600638977636, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1326, + "step": 759 + }, + { + "epoch": 1.2140575079872205, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1347, + "step": 760 + }, + { + "epoch": 1.2156549520766773, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1353, + "step": 761 + }, + { + "epoch": 1.2172523961661341, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.14, + "step": 762 + }, + { + "epoch": 1.218849840255591, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.1326, + "step": 763 + }, + { + "epoch": 1.220447284345048, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.136, + "step": 764 + }, + { + "epoch": 1.2220447284345048, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 765 + }, + { + "epoch": 1.2236421725239617, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1325, + "step": 766 + }, + { + "epoch": 1.2252396166134185, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1351, + "step": 767 + }, + { + "epoch": 1.2268370607028753, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.131, + "step": 768 + }, + { + "epoch": 1.2284345047923322, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1289, + "step": 769 + }, + { + "epoch": 1.230031948881789, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 770 + }, + { + "epoch": 1.231629392971246, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1219, + "step": 771 + }, + { + "epoch": 1.233226837060703, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1227, + "step": 772 + }, + { + "epoch": 1.2348242811501597, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 773 + }, + { + "epoch": 1.2364217252396166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1244, + "step": 774 + }, + { + "epoch": 1.2380191693290734, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 775 + }, + { + "epoch": 1.2396166134185305, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1308, + "step": 776 + }, + { + "epoch": 1.2412140575079873, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1356, + "step": 777 + }, + { + "epoch": 1.2428115015974441, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1287, + "step": 778 + }, + { + "epoch": 1.244408945686901, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 779 + }, + { + "epoch": 1.2460063897763578, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.1331, + "step": 780 + }, + { + "epoch": 1.2476038338658146, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.127, + "step": 781 + }, + { + "epoch": 1.2492012779552715, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 782 + }, + { + "epoch": 1.2507987220447285, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 783 + }, + { + "epoch": 1.2523961661341854, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1241, + "step": 784 + }, + { + "epoch": 1.2539936102236422, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1203, + "step": 785 + }, + { + "epoch": 1.255591054313099, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 786 + }, + { + "epoch": 1.2571884984025559, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 787 + }, + { + "epoch": 1.2587859424920127, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1204, + "step": 788 + }, + { + "epoch": 1.2603833865814695, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1212, + "step": 789 + }, + { + "epoch": 1.2619808306709266, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1245, + "step": 790 + }, + { + "epoch": 1.2635782747603834, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1229, + "step": 791 + }, + { + "epoch": 1.2651757188498403, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.122, + "step": 792 + }, + { + "epoch": 1.266773162939297, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 793 + }, + { + "epoch": 1.268370607028754, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.1196, + "step": 794 + }, + { + "epoch": 1.269968051118211, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1263, + "step": 795 + }, + { + "epoch": 1.2715654952076676, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1289, + "step": 796 + }, + { + "epoch": 1.2731629392971247, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1304, + "step": 797 + }, + { + "epoch": 1.2747603833865815, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1288, + "step": 798 + }, + { + "epoch": 1.2763578274760383, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1161, + "step": 799 + }, + { + "epoch": 1.2779552715654952, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1276, + "step": 800 + }, + { + "epoch": 1.279552715654952, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1221, + "step": 801 + }, + { + "epoch": 1.281150159744409, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1276, + "step": 802 + }, + { + "epoch": 1.2827476038338659, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 803 + }, + { + "epoch": 1.2843450479233227, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1208, + "step": 804 + }, + { + "epoch": 1.2859424920127795, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1251, + "step": 805 + }, + { + "epoch": 1.2875399361022364, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 806 + }, + { + "epoch": 1.2891373801916932, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1278, + "step": 807 + }, + { + "epoch": 1.29073482428115, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 808 + }, + { + "epoch": 1.292332268370607, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 809 + }, + { + "epoch": 1.293929712460064, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 810 + }, + { + "epoch": 1.2955271565495208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 811 + }, + { + "epoch": 1.2971246006389776, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1252, + "step": 812 + }, + { + "epoch": 1.2987220447284344, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.118, + "step": 813 + }, + { + "epoch": 1.3003194888178915, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.1198, + "step": 814 + }, + { + "epoch": 1.3019169329073481, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1241, + "step": 815 + }, + { + "epoch": 1.3035143769968052, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 816 + }, + { + "epoch": 1.305111821086262, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1231, + "step": 817 + }, + { + "epoch": 1.3067092651757188, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1227, + "step": 818 + }, + { + "epoch": 1.3083067092651757, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 819 + }, + { + "epoch": 1.3099041533546325, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.1242, + "step": 820 + }, + { + "epoch": 1.3115015974440896, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1278, + "step": 821 + }, + { + "epoch": 1.3130990415335464, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 822 + }, + { + "epoch": 1.3146964856230032, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 823 + }, + { + "epoch": 1.31629392971246, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1217, + "step": 824 + }, + { + "epoch": 1.317891373801917, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1279, + "step": 825 + }, + { + "epoch": 1.3194888178913737, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.1239, + "step": 826 + }, + { + "epoch": 1.3210862619808306, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 827 + }, + { + "epoch": 1.3226837060702876, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.1231, + "step": 828 + }, + { + "epoch": 1.3242811501597445, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 829 + }, + { + "epoch": 1.3258785942492013, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1159, + "step": 830 + }, + { + "epoch": 1.3274760383386581, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 831 + }, + { + "epoch": 1.329073482428115, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1199, + "step": 832 + }, + { + "epoch": 1.330670926517572, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 833 + }, + { + "epoch": 1.3322683706070286, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1239, + "step": 834 + }, + { + "epoch": 1.3338658146964857, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1207, + "step": 835 + }, + { + "epoch": 1.3354632587859425, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.123, + "step": 836 + }, + { + "epoch": 1.3370607028753994, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1115, + "step": 837 + }, + { + "epoch": 1.3386581469648562, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1162, + "step": 838 + }, + { + "epoch": 1.340255591054313, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 839 + }, + { + "epoch": 1.34185303514377, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 840 + }, + { + "epoch": 1.343450479233227, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 841 + }, + { + "epoch": 1.3450479233226837, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 842 + }, + { + "epoch": 1.3466453674121406, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 843 + }, + { + "epoch": 1.3482428115015974, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1272, + "step": 844 + }, + { + "epoch": 1.3498402555910542, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1135, + "step": 845 + }, + { + "epoch": 1.351437699680511, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 846 + }, + { + "epoch": 1.3530351437699681, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1167, + "step": 847 + }, + { + "epoch": 1.354632587859425, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1239, + "step": 848 + }, + { + "epoch": 1.3562300319488818, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1149, + "step": 849 + }, + { + "epoch": 1.3578274760383386, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 850 + }, + { + "epoch": 1.3594249201277955, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 851 + }, + { + "epoch": 1.3610223642172525, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1243, + "step": 852 + }, + { + "epoch": 1.3626198083067091, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.124, + "step": 853 + }, + { + "epoch": 1.3642172523961662, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1211, + "step": 854 + }, + { + "epoch": 1.365814696485623, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 855 + }, + { + "epoch": 1.3674121405750799, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 856 + }, + { + "epoch": 1.3690095846645367, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 857 + }, + { + "epoch": 1.3706070287539935, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 858 + }, + { + "epoch": 1.3722044728434506, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1202, + "step": 859 + }, + { + "epoch": 1.3738019169329074, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 860 + }, + { + "epoch": 1.3753993610223643, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1233, + "step": 861 + }, + { + "epoch": 1.376996805111821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 862 + }, + { + "epoch": 1.378594249201278, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1202, + "step": 863 + }, + { + "epoch": 1.3801916932907348, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1258, + "step": 864 + }, + { + "epoch": 1.3817891373801916, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.123, + "step": 865 + }, + { + "epoch": 1.3833865814696487, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1204, + "step": 866 + }, + { + "epoch": 1.3849840255591055, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1237, + "step": 867 + }, + { + "epoch": 1.3865814696485623, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 868 + }, + { + "epoch": 1.3881789137380192, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1146, + "step": 869 + }, + { + "epoch": 1.389776357827476, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1288, + "step": 870 + }, + { + "epoch": 1.3913738019169328, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.118, + "step": 871 + }, + { + "epoch": 1.3929712460063897, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 872 + }, + { + "epoch": 1.3945686900958467, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 873 + }, + { + "epoch": 1.3961661341853036, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1256, + "step": 874 + }, + { + "epoch": 1.3977635782747604, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 875 + }, + { + "epoch": 1.3993610223642172, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 876 + }, + { + "epoch": 1.400958466453674, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.12, + "step": 877 + }, + { + "epoch": 1.4025559105431311, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 878 + }, + { + "epoch": 1.4041533546325877, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 879 + }, + { + "epoch": 1.4057507987220448, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1214, + "step": 880 + }, + { + "epoch": 1.4073482428115016, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 881 + }, + { + "epoch": 1.4089456869009584, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 882 + }, + { + "epoch": 1.4105431309904153, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 883 + }, + { + "epoch": 1.4121405750798721, + "grad_norm": 0.53515625, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 884 + }, + { + "epoch": 1.4137380191693292, + "grad_norm": 1.3359375, + "learning_rate": 0.0005, + "loss": 1.1437, + "step": 885 + }, + { + "epoch": 1.415335463258786, + "grad_norm": 1.7578125, + "learning_rate": 0.0005, + "loss": 1.1625, + "step": 886 + }, + { + "epoch": 1.4169329073482428, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.1297, + "step": 887 + }, + { + "epoch": 1.4185303514376997, + "grad_norm": 1.09375, + "learning_rate": 0.0005, + "loss": 1.147, + "step": 888 + }, + { + "epoch": 1.4201277955271565, + "grad_norm": 0.6328125, + "learning_rate": 0.0005, + "loss": 1.1441, + "step": 889 + }, + { + "epoch": 1.4217252396166133, + "grad_norm": 0.443359375, + "learning_rate": 0.0005, + "loss": 1.1385, + "step": 890 + }, + { + "epoch": 1.4233226837060702, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 891 + }, + { + "epoch": 1.4249201277955272, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.133, + "step": 892 + }, + { + "epoch": 1.426517571884984, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.1278, + "step": 893 + }, + { + "epoch": 1.428115015974441, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.134, + "step": 894 + }, + { + "epoch": 1.4297124600638977, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.1251, + "step": 895 + }, + { + "epoch": 1.4313099041533546, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.1338, + "step": 896 + }, + { + "epoch": 1.4329073482428116, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 897 + }, + { + "epoch": 1.4345047923322682, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.132, + "step": 898 + }, + { + "epoch": 1.4361022364217253, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 899 + }, + { + "epoch": 1.4376996805111821, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 900 + }, + { + "epoch": 1.439297124600639, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 901 + }, + { + "epoch": 1.4408945686900958, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.1308, + "step": 902 + }, + { + "epoch": 1.4424920127795526, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1225, + "step": 903 + }, + { + "epoch": 1.4440894568690097, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 904 + }, + { + "epoch": 1.4456869009584665, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1258, + "step": 905 + }, + { + "epoch": 1.4472843450479234, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.1212, + "step": 906 + }, + { + "epoch": 1.4488817891373802, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 907 + }, + { + "epoch": 1.450479233226837, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.1262, + "step": 908 + }, + { + "epoch": 1.4520766773162939, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 909 + }, + { + "epoch": 1.4536741214057507, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.1293, + "step": 910 + }, + { + "epoch": 1.4552715654952078, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1197, + "step": 911 + }, + { + "epoch": 1.4568690095846646, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.1187, + "step": 912 + }, + { + "epoch": 1.4584664536741214, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.1165, + "step": 913 + }, + { + "epoch": 1.4600638977635783, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 914 + }, + { + "epoch": 1.461661341853035, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.1211, + "step": 915 + }, + { + "epoch": 1.4632587859424921, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.1258, + "step": 916 + }, + { + "epoch": 1.4648562300319488, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 917 + }, + { + "epoch": 1.4664536741214058, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1159, + "step": 918 + }, + { + "epoch": 1.4680511182108626, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 919 + }, + { + "epoch": 1.4696485623003195, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1264, + "step": 920 + }, + { + "epoch": 1.4712460063897763, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1217, + "step": 921 + }, + { + "epoch": 1.4728434504792332, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 922 + }, + { + "epoch": 1.4744408945686902, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 923 + }, + { + "epoch": 1.476038338658147, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.1215, + "step": 924 + }, + { + "epoch": 1.4776357827476039, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1255, + "step": 925 + }, + { + "epoch": 1.4792332268370607, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 926 + }, + { + "epoch": 1.4808306709265175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 927 + }, + { + "epoch": 1.4824281150159744, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1196, + "step": 928 + }, + { + "epoch": 1.4840255591054312, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 929 + }, + { + "epoch": 1.4856230031948883, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 930 + }, + { + "epoch": 1.487220447284345, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.12, + "step": 931 + }, + { + "epoch": 1.488817891373802, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 932 + }, + { + "epoch": 1.4904153354632588, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 933 + }, + { + "epoch": 1.4920127795527156, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 934 + }, + { + "epoch": 1.4936102236421724, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 935 + }, + { + "epoch": 1.4952076677316293, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 936 + }, + { + "epoch": 1.4968051118210863, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1183, + "step": 937 + }, + { + "epoch": 1.4984025559105432, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1204, + "step": 938 + }, + { + "epoch": 1.5, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1229, + "step": 939 + }, + { + "epoch": 1.5015974440894568, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 940 + }, + { + "epoch": 1.5031948881789137, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1152, + "step": 941 + }, + { + "epoch": 1.5047923322683707, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 942 + }, + { + "epoch": 1.5063897763578273, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1171, + "step": 943 + }, + { + "epoch": 1.5079872204472844, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1181, + "step": 944 + }, + { + "epoch": 1.5095846645367412, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 945 + }, + { + "epoch": 1.511182108626198, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 946 + }, + { + "epoch": 1.5127795527156551, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 947 + }, + { + "epoch": 1.5143769968051117, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1215, + "step": 948 + }, + { + "epoch": 1.5159744408945688, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1109, + "step": 949 + }, + { + "epoch": 1.5175718849840254, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1147, + "step": 950 + }, + { + "epoch": 1.5191693290734825, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 951 + }, + { + "epoch": 1.5207667731629393, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 952 + }, + { + "epoch": 1.5223642172523961, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1136, + "step": 953 + }, + { + "epoch": 1.5239616613418532, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1182, + "step": 954 + }, + { + "epoch": 1.5255591054313098, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1186, + "step": 955 + }, + { + "epoch": 1.5271565495207668, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1253, + "step": 956 + }, + { + "epoch": 1.5287539936102237, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1107, + "step": 957 + }, + { + "epoch": 1.5303514376996805, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 958 + }, + { + "epoch": 1.5319488817891374, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 959 + }, + { + "epoch": 1.5335463258785942, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 960 + }, + { + "epoch": 1.5351437699680512, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 961 + }, + { + "epoch": 1.5367412140575079, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1119, + "step": 962 + }, + { + "epoch": 1.538338658146965, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1248, + "step": 963 + }, + { + "epoch": 1.5399361022364217, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1137, + "step": 964 + }, + { + "epoch": 1.5415335463258786, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 965 + }, + { + "epoch": 1.5431309904153354, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 966 + }, + { + "epoch": 1.5447284345047922, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 967 + }, + { + "epoch": 1.5463258785942493, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 968 + }, + { + "epoch": 1.547923322683706, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 969 + }, + { + "epoch": 1.549520766773163, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1157, + "step": 970 + }, + { + "epoch": 1.5511182108626198, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.1186, + "step": 971 + }, + { + "epoch": 1.5527156549520766, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 972 + }, + { + "epoch": 1.5543130990415337, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.115, + "step": 973 + }, + { + "epoch": 1.5559105431309903, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 974 + }, + { + "epoch": 1.5575079872204474, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 975 + }, + { + "epoch": 1.5591054313099042, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1171, + "step": 976 + }, + { + "epoch": 1.560702875399361, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 977 + }, + { + "epoch": 1.5623003194888179, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1189, + "step": 978 + }, + { + "epoch": 1.5638977635782747, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1177, + "step": 979 + }, + { + "epoch": 1.5654952076677318, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1184, + "step": 980 + }, + { + "epoch": 1.5670926517571884, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 981 + }, + { + "epoch": 1.5686900958466454, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 982 + }, + { + "epoch": 1.5702875399361023, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 983 + }, + { + "epoch": 1.571884984025559, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 984 + }, + { + "epoch": 1.573482428115016, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.116, + "step": 985 + }, + { + "epoch": 1.5750798722044728, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1218, + "step": 986 + }, + { + "epoch": 1.5766773162939298, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.1195, + "step": 987 + }, + { + "epoch": 1.5782747603833864, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1125, + "step": 988 + }, + { + "epoch": 1.5798722044728435, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 989 + }, + { + "epoch": 1.5814696485623003, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 990 + }, + { + "epoch": 1.5830670926517572, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1174, + "step": 991 + }, + { + "epoch": 1.5846645367412142, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1094, + "step": 992 + }, + { + "epoch": 1.5862619808306708, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 993 + }, + { + "epoch": 1.5878594249201279, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.118, + "step": 994 + }, + { + "epoch": 1.5894568690095847, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.115, + "step": 995 + }, + { + "epoch": 1.5910543130990416, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1146, + "step": 996 + }, + { + "epoch": 1.5926517571884984, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 997 + }, + { + "epoch": 1.5942492012779552, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1151, + "step": 998 + }, + { + "epoch": 1.5958466453674123, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 999 + }, + { + "epoch": 1.5974440894568689, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 1000 + }, + { + "epoch": 1.599041533546326, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1128, + "step": 1001 + }, + { + "epoch": 1.6006389776357828, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 1002 + }, + { + "epoch": 1.6022364217252396, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1112, + "step": 1003 + }, + { + "epoch": 1.6038338658146964, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.121, + "step": 1004 + }, + { + "epoch": 1.6054313099041533, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1143, + "step": 1005 + }, + { + "epoch": 1.6070287539936103, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 1006 + }, + { + "epoch": 1.608626198083067, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 1007 + }, + { + "epoch": 1.610223642172524, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1008 + }, + { + "epoch": 1.6118210862619808, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 1009 + }, + { + "epoch": 1.6134185303514377, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 1010 + }, + { + "epoch": 1.6150159744408947, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1197, + "step": 1011 + }, + { + "epoch": 1.6166134185303513, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.1135, + "step": 1012 + }, + { + "epoch": 1.6182108626198084, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.1107, + "step": 1013 + }, + { + "epoch": 1.619808306709265, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 1014 + }, + { + "epoch": 1.621405750798722, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1171, + "step": 1015 + }, + { + "epoch": 1.623003194888179, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1016 + }, + { + "epoch": 1.6246006389776357, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1017 + }, + { + "epoch": 1.6261980830670928, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.119, + "step": 1018 + }, + { + "epoch": 1.6277955271565494, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 1019 + }, + { + "epoch": 1.6293929712460065, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1173, + "step": 1020 + }, + { + "epoch": 1.6309904153354633, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1021 + }, + { + "epoch": 1.6325878594249201, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1022 + }, + { + "epoch": 1.634185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1161, + "step": 1023 + }, + { + "epoch": 1.6357827476038338, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1173, + "step": 1024 + }, + { + "epoch": 1.6373801916932909, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1175, + "step": 1025 + }, + { + "epoch": 1.6389776357827475, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1026 + }, + { + "epoch": 1.6405750798722045, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 1027 + }, + { + "epoch": 1.6421725239616614, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1028 + }, + { + "epoch": 1.6437699680511182, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 1029 + }, + { + "epoch": 1.645367412140575, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1118, + "step": 1030 + }, + { + "epoch": 1.6469648562300319, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 1031 + }, + { + "epoch": 1.648562300319489, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 1032 + }, + { + "epoch": 1.6501597444089455, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 1033 + }, + { + "epoch": 1.6517571884984026, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1114, + "step": 1034 + }, + { + "epoch": 1.6533546325878594, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1136, + "step": 1035 + }, + { + "epoch": 1.6549520766773163, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 1036 + }, + { + "epoch": 1.6565495207667733, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1167, + "step": 1037 + }, + { + "epoch": 1.65814696485623, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1038 + }, + { + "epoch": 1.659744408945687, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 1039 + }, + { + "epoch": 1.6613418530351438, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1040 + }, + { + "epoch": 1.6629392971246006, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1041 + }, + { + "epoch": 1.6645367412140575, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 1042 + }, + { + "epoch": 1.6661341853035143, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.1077, + "step": 1043 + }, + { + "epoch": 1.6677316293929714, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1044 + }, + { + "epoch": 1.669329073482428, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1041, + "step": 1045 + }, + { + "epoch": 1.670926517571885, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 1046 + }, + { + "epoch": 1.6725239616613419, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1102, + "step": 1047 + }, + { + "epoch": 1.6741214057507987, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 1048 + }, + { + "epoch": 1.6757188498402555, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1049 + }, + { + "epoch": 1.6773162939297124, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1148, + "step": 1050 + }, + { + "epoch": 1.6789137380191694, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1125, + "step": 1051 + }, + { + "epoch": 1.680511182108626, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1147, + "step": 1052 + }, + { + "epoch": 1.682108626198083, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1053 + }, + { + "epoch": 1.68370607028754, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1157, + "step": 1054 + }, + { + "epoch": 1.6853035143769968, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 1055 + }, + { + "epoch": 1.6869009584664538, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1146, + "step": 1056 + }, + { + "epoch": 1.6884984025559104, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1154, + "step": 1057 + }, + { + "epoch": 1.6900958466453675, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1149, + "step": 1058 + }, + { + "epoch": 1.6916932907348243, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1059 + }, + { + "epoch": 1.6932907348242812, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1122, + "step": 1060 + }, + { + "epoch": 1.694888178913738, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1164, + "step": 1061 + }, + { + "epoch": 1.6964856230031948, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1062 + }, + { + "epoch": 1.6980830670926519, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 1063 + }, + { + "epoch": 1.6996805111821085, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1119, + "step": 1064 + }, + { + "epoch": 1.7012779552715656, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 1065 + }, + { + "epoch": 1.7028753993610224, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1165, + "step": 1066 + }, + { + "epoch": 1.7044728434504792, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1067 + }, + { + "epoch": 1.706070287539936, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.116, + "step": 1068 + }, + { + "epoch": 1.707667731629393, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1131, + "step": 1069 + }, + { + "epoch": 1.70926517571885, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.1163, + "step": 1070 + }, + { + "epoch": 1.7108626198083066, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1066, + "step": 1071 + }, + { + "epoch": 1.7124600638977636, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1114, + "step": 1072 + }, + { + "epoch": 1.7140575079872205, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.1142, + "step": 1073 + }, + { + "epoch": 1.7156549520766773, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.1177, + "step": 1074 + }, + { + "epoch": 1.7172523961661343, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1075 + }, + { + "epoch": 1.718849840255591, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1121, + "step": 1076 + }, + { + "epoch": 1.720447284345048, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1077 + }, + { + "epoch": 1.7220447284345048, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1112, + "step": 1078 + }, + { + "epoch": 1.7236421725239617, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1079 + }, + { + "epoch": 1.7252396166134185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 1080 + }, + { + "epoch": 1.7268370607028753, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1105, + "step": 1081 + }, + { + "epoch": 1.7284345047923324, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 1082 + }, + { + "epoch": 1.730031948881789, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 1083 + }, + { + "epoch": 1.731629392971246, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 1084 + }, + { + "epoch": 1.733226837060703, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 1085 + }, + { + "epoch": 1.7348242811501597, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 1086 + }, + { + "epoch": 1.7364217252396166, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1087 + }, + { + "epoch": 1.7380191693290734, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1088 + }, + { + "epoch": 1.7396166134185305, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 1089 + }, + { + "epoch": 1.741214057507987, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 1090 + }, + { + "epoch": 1.7428115015974441, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.1095, + "step": 1091 + }, + { + "epoch": 1.744408945686901, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1092 + }, + { + "epoch": 1.7460063897763578, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1111, + "step": 1093 + }, + { + "epoch": 1.7476038338658149, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.1113, + "step": 1094 + }, + { + "epoch": 1.7492012779552715, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1095 + }, + { + "epoch": 1.7507987220447285, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1096 + }, + { + "epoch": 1.7523961661341851, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1097 + }, + { + "epoch": 1.7539936102236422, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1098 + }, + { + "epoch": 1.755591054313099, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 1099 + }, + { + "epoch": 1.7571884984025559, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 1100 + }, + { + "epoch": 1.758785942492013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1176, + "step": 1101 + }, + { + "epoch": 1.7603833865814695, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1102 + }, + { + "epoch": 1.7619808306709266, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.1166, + "step": 1103 + }, + { + "epoch": 1.7635782747603834, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 1104 + }, + { + "epoch": 1.7651757188498403, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1105 + }, + { + "epoch": 1.766773162939297, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1106 + }, + { + "epoch": 1.768370607028754, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1107 + }, + { + "epoch": 1.769968051118211, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 1108 + }, + { + "epoch": 1.7715654952076676, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1109 + }, + { + "epoch": 1.7731629392971247, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1138, + "step": 1110 + }, + { + "epoch": 1.7747603833865815, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1111 + }, + { + "epoch": 1.7763578274760383, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 1112 + }, + { + "epoch": 1.7779552715654952, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1113 + }, + { + "epoch": 1.779552715654952, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1123, + "step": 1114 + }, + { + "epoch": 1.781150159744409, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1105, + "step": 1115 + }, + { + "epoch": 1.7827476038338657, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1116 + }, + { + "epoch": 1.7843450479233227, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 1117 + }, + { + "epoch": 1.7859424920127795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.1126, + "step": 1118 + }, + { + "epoch": 1.7875399361022364, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1119 + }, + { + "epoch": 1.7891373801916934, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 1120 + }, + { + "epoch": 1.79073482428115, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1207, + "step": 1121 + }, + { + "epoch": 1.792332268370607, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1122 + }, + { + "epoch": 1.793929712460064, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1123 + }, + { + "epoch": 1.7955271565495208, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 1124 + }, + { + "epoch": 1.7971246006389776, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1125 + }, + { + "epoch": 1.7987220447284344, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1126 + }, + { + "epoch": 1.8003194888178915, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1127 + }, + { + "epoch": 1.8019169329073481, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1128 + }, + { + "epoch": 1.8035143769968052, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1137, + "step": 1129 + }, + { + "epoch": 1.805111821086262, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1162, + "step": 1130 + }, + { + "epoch": 1.8067092651757188, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1131 + }, + { + "epoch": 1.8083067092651757, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1132 + }, + { + "epoch": 1.8099041533546325, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1133 + }, + { + "epoch": 1.8115015974440896, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 1134 + }, + { + "epoch": 1.8130990415335462, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 1135 + }, + { + "epoch": 1.8146964856230032, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1136 + }, + { + "epoch": 1.81629392971246, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1137 + }, + { + "epoch": 1.817891373801917, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1138 + }, + { + "epoch": 1.819488817891374, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1108, + "step": 1139 + }, + { + "epoch": 1.8210862619808306, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1115, + "step": 1140 + }, + { + "epoch": 1.8226837060702876, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1141 + }, + { + "epoch": 1.8242811501597445, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1135, + "step": 1142 + }, + { + "epoch": 1.8258785942492013, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1143 + }, + { + "epoch": 1.8274760383386581, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1144 + }, + { + "epoch": 1.829073482428115, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1048, + "step": 1145 + }, + { + "epoch": 1.830670926517572, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1102, + "step": 1146 + }, + { + "epoch": 1.8322683706070286, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1128, + "step": 1147 + }, + { + "epoch": 1.8338658146964857, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1148 + }, + { + "epoch": 1.8354632587859425, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1149 + }, + { + "epoch": 1.8370607028753994, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1150 + }, + { + "epoch": 1.8386581469648562, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1151 + }, + { + "epoch": 1.840255591054313, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 1152 + }, + { + "epoch": 1.84185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1153 + }, + { + "epoch": 1.8434504792332267, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 1154 + }, + { + "epoch": 1.8450479233226837, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1155 + }, + { + "epoch": 1.8466453674121406, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1156 + }, + { + "epoch": 1.8482428115015974, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1095, + "step": 1157 + }, + { + "epoch": 1.8498402555910545, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1158 + }, + { + "epoch": 1.851437699680511, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 1159 + }, + { + "epoch": 1.8530351437699681, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1160 + }, + { + "epoch": 1.854632587859425, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.1119, + "step": 1161 + }, + { + "epoch": 1.8562300319488818, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1162 + }, + { + "epoch": 1.8578274760383386, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1071, + "step": 1163 + }, + { + "epoch": 1.8594249201277955, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1164 + }, + { + "epoch": 1.8610223642172525, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 1165 + }, + { + "epoch": 1.8626198083067091, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1166 + }, + { + "epoch": 1.8642172523961662, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 1167 + }, + { + "epoch": 1.865814696485623, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1168 + }, + { + "epoch": 1.8674121405750799, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1169 + }, + { + "epoch": 1.8690095846645367, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 1170 + }, + { + "epoch": 1.8706070287539935, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1171 + }, + { + "epoch": 1.8722044728434506, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1076, + "step": 1172 + }, + { + "epoch": 1.8738019169329072, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 1173 + }, + { + "epoch": 1.8753993610223643, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1174 + }, + { + "epoch": 1.876996805111821, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1175 + }, + { + "epoch": 1.878594249201278, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1176 + }, + { + "epoch": 1.880191693290735, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1177 + }, + { + "epoch": 1.8817891373801916, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1178 + }, + { + "epoch": 1.8833865814696487, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1179 + }, + { + "epoch": 1.8849840255591053, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1159, + "step": 1180 + }, + { + "epoch": 1.8865814696485623, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1181 + }, + { + "epoch": 1.8881789137380192, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1182 + }, + { + "epoch": 1.889776357827476, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1183 + }, + { + "epoch": 1.891373801916933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1184 + }, + { + "epoch": 1.8929712460063897, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1185 + }, + { + "epoch": 1.8945686900958467, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1186 + }, + { + "epoch": 1.8961661341853036, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1187 + }, + { + "epoch": 1.8977635782747604, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1188 + }, + { + "epoch": 1.8993610223642172, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1189 + }, + { + "epoch": 1.900958466453674, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 1190 + }, + { + "epoch": 1.9025559105431311, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1191 + }, + { + "epoch": 1.9041533546325877, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1192 + }, + { + "epoch": 1.9057507987220448, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 1193 + }, + { + "epoch": 1.9073482428115016, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1194 + }, + { + "epoch": 1.9089456869009584, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 1195 + }, + { + "epoch": 1.9105431309904153, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 1196 + }, + { + "epoch": 1.9121405750798721, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1197 + }, + { + "epoch": 1.9137380191693292, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1198 + }, + { + "epoch": 1.9153354632587858, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1199 + }, + { + "epoch": 1.9169329073482428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1200 + }, + { + "epoch": 1.9185303514376997, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1201 + }, + { + "epoch": 1.9201277955271565, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 1202 + }, + { + "epoch": 1.9217252396166136, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1125, + "step": 1203 + }, + { + "epoch": 1.9233226837060702, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1152, + "step": 1204 + }, + { + "epoch": 1.9249201277955272, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1107, + "step": 1205 + }, + { + "epoch": 1.926517571884984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 1206 + }, + { + "epoch": 1.928115015974441, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1109, + "step": 1207 + }, + { + "epoch": 1.9297124600638977, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1208 + }, + { + "epoch": 1.9313099041533546, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1209 + }, + { + "epoch": 1.9329073482428116, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1210 + }, + { + "epoch": 1.9345047923322682, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1211 + }, + { + "epoch": 1.9361022364217253, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1086, + "step": 1212 + }, + { + "epoch": 1.9376996805111821, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1213 + }, + { + "epoch": 1.939297124600639, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 1214 + }, + { + "epoch": 1.9408945686900958, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1215 + }, + { + "epoch": 1.9424920127795526, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1143, + "step": 1216 + }, + { + "epoch": 1.9440894568690097, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1217 + }, + { + "epoch": 1.9456869009584663, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.114, + "step": 1218 + }, + { + "epoch": 1.9472843450479234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 1219 + }, + { + "epoch": 1.9488817891373802, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1220 + }, + { + "epoch": 1.950479233226837, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 1221 + }, + { + "epoch": 1.952076677316294, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1222 + }, + { + "epoch": 1.9536741214057507, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 1223 + }, + { + "epoch": 1.9552715654952078, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1224 + }, + { + "epoch": 1.9568690095846646, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1225 + }, + { + "epoch": 1.9584664536741214, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 1226 + }, + { + "epoch": 1.9600638977635783, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 1227 + }, + { + "epoch": 1.961661341853035, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1228 + }, + { + "epoch": 1.9632587859424921, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1066, + "step": 1229 + }, + { + "epoch": 1.9648562300319488, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.1033, + "step": 1230 + }, + { + "epoch": 1.9664536741214058, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0984, + "step": 1231 + }, + { + "epoch": 1.9680511182108626, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.1071, + "step": 1232 + }, + { + "epoch": 1.9696485623003195, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1233 + }, + { + "epoch": 1.9712460063897763, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1234 + }, + { + "epoch": 1.9728434504792332, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1235 + }, + { + "epoch": 1.9744408945686902, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1066, + "step": 1236 + }, + { + "epoch": 1.9760383386581468, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 1237 + }, + { + "epoch": 1.9776357827476039, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1238 + }, + { + "epoch": 1.9792332268370607, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1048, + "step": 1239 + }, + { + "epoch": 1.9808306709265175, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1240 + }, + { + "epoch": 1.9824281150159746, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1241 + }, + { + "epoch": 1.9840255591054312, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 1242 + }, + { + "epoch": 1.9856230031948883, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1052, + "step": 1243 + }, + { + "epoch": 1.9872204472843449, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1111, + "step": 1244 + }, + { + "epoch": 1.988817891373802, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1245 + }, + { + "epoch": 1.9904153354632588, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1246 + }, + { + "epoch": 1.9920127795527156, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 1247 + }, + { + "epoch": 1.9936102236421727, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1248 + }, + { + "epoch": 1.9952076677316293, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1043, + "step": 1249 + }, + { + "epoch": 1.9968051118210863, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1043, + "step": 1250 + }, + { + "epoch": 1.9984025559105432, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 1251 + }, + { + "epoch": 2.0, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1252 + }, + { + "epoch": 2.001597444089457, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1253 + }, + { + "epoch": 2.0031948881789137, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.1086, + "step": 1254 + }, + { + "epoch": 2.0047923322683707, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 1255 + }, + { + "epoch": 2.0063897763578273, + "grad_norm": 0.73828125, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1256 + }, + { + "epoch": 2.0079872204472844, + "grad_norm": 1.9609375, + "learning_rate": 0.0005, + "loss": 1.1489, + "step": 1257 + }, + { + "epoch": 2.009584664536741, + "grad_norm": 0.431640625, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 1258 + }, + { + "epoch": 2.011182108626198, + "grad_norm": 0.5, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 1259 + }, + { + "epoch": 2.012779552715655, + "grad_norm": 0.921875, + "learning_rate": 0.0005, + "loss": 1.1174, + "step": 1260 + }, + { + "epoch": 2.0143769968051117, + "grad_norm": 0.671875, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 1261 + }, + { + "epoch": 2.015974440894569, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.1147, + "step": 1262 + }, + { + "epoch": 2.0175718849840254, + "grad_norm": 0.8359375, + "learning_rate": 0.0005, + "loss": 1.1288, + "step": 1263 + }, + { + "epoch": 2.0191693290734825, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.1124, + "step": 1264 + }, + { + "epoch": 2.0207667731629395, + "grad_norm": 0.55859375, + "learning_rate": 0.0005, + "loss": 1.1187, + "step": 1265 + }, + { + "epoch": 2.022364217252396, + "grad_norm": 0.52734375, + "learning_rate": 0.0005, + "loss": 1.1175, + "step": 1266 + }, + { + "epoch": 2.023961661341853, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.1108, + "step": 1267 + }, + { + "epoch": 2.02555910543131, + "grad_norm": 0.75, + "learning_rate": 0.0005, + "loss": 1.1191, + "step": 1268 + }, + { + "epoch": 2.027156549520767, + "grad_norm": 0.6875, + "learning_rate": 0.0005, + "loss": 1.1229, + "step": 1269 + }, + { + "epoch": 2.0287539936102235, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.1143, + "step": 1270 + }, + { + "epoch": 2.0303514376996805, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 1271 + }, + { + "epoch": 2.0319488817891376, + "grad_norm": 0.326171875, + "learning_rate": 0.0005, + "loss": 1.1166, + "step": 1272 + }, + { + "epoch": 2.033546325878594, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1273 + }, + { + "epoch": 2.0351437699680512, + "grad_norm": 0.392578125, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 1274 + }, + { + "epoch": 2.036741214057508, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1111, + "step": 1275 + }, + { + "epoch": 2.038338658146965, + "grad_norm": 0.2216796875, + "learning_rate": 0.0005, + "loss": 1.1128, + "step": 1276 + }, + { + "epoch": 2.0399361022364215, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1277 + }, + { + "epoch": 2.0415335463258786, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.1175, + "step": 1278 + }, + { + "epoch": 2.0431309904153356, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 1279 + }, + { + "epoch": 2.0447284345047922, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1172, + "step": 1280 + }, + { + "epoch": 2.0463258785942493, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1281 + }, + { + "epoch": 2.047923322683706, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 1282 + }, + { + "epoch": 2.049520766773163, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1283 + }, + { + "epoch": 2.0511182108626196, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1284 + }, + { + "epoch": 2.0527156549520766, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1144, + "step": 1285 + }, + { + "epoch": 2.0543130990415337, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 1286 + }, + { + "epoch": 2.0559105431309903, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.1043, + "step": 1287 + }, + { + "epoch": 2.0575079872204474, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1288 + }, + { + "epoch": 2.059105431309904, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1289 + }, + { + "epoch": 2.060702875399361, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1290 + }, + { + "epoch": 2.062300319488818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1082, + "step": 1291 + }, + { + "epoch": 2.0638977635782747, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1076, + "step": 1292 + }, + { + "epoch": 2.0654952076677318, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1293 + }, + { + "epoch": 2.0670926517571884, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1294 + }, + { + "epoch": 2.0686900958466454, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1295 + }, + { + "epoch": 2.070287539936102, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1296 + }, + { + "epoch": 2.071884984025559, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1297 + }, + { + "epoch": 2.073482428115016, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.1095, + "step": 1298 + }, + { + "epoch": 2.0750798722044728, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1131, + "step": 1299 + }, + { + "epoch": 2.07667731629393, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1300 + }, + { + "epoch": 2.0782747603833864, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1301 + }, + { + "epoch": 2.0798722044728435, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1302 + }, + { + "epoch": 2.0814696485623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1041, + "step": 1303 + }, + { + "epoch": 2.083067092651757, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1304 + }, + { + "epoch": 2.084664536741214, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1305 + }, + { + "epoch": 2.086261980830671, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1306 + }, + { + "epoch": 2.087859424920128, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1307 + }, + { + "epoch": 2.0894568690095845, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1308 + }, + { + "epoch": 2.0910543130990416, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.1025, + "step": 1309 + }, + { + "epoch": 2.0926517571884986, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1310 + }, + { + "epoch": 2.094249201277955, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.1053, + "step": 1311 + }, + { + "epoch": 2.0958466453674123, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1312 + }, + { + "epoch": 2.097444089456869, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1313 + }, + { + "epoch": 2.099041533546326, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1079, + "step": 1314 + }, + { + "epoch": 2.1006389776357826, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 1315 + }, + { + "epoch": 2.1022364217252396, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1316 + }, + { + "epoch": 2.1038338658146967, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1317 + }, + { + "epoch": 2.1054313099041533, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1318 + }, + { + "epoch": 2.1070287539936103, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 1319 + }, + { + "epoch": 2.108626198083067, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1320 + }, + { + "epoch": 2.110223642172524, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1321 + }, + { + "epoch": 2.1118210862619806, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 1322 + }, + { + "epoch": 2.1134185303514377, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1323 + }, + { + "epoch": 2.1150159744408947, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1324 + }, + { + "epoch": 2.1166134185303513, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.102, + "step": 1325 + }, + { + "epoch": 2.1182108626198084, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1326 + }, + { + "epoch": 2.119808306709265, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1327 + }, + { + "epoch": 2.121405750798722, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1328 + }, + { + "epoch": 2.123003194888179, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1329 + }, + { + "epoch": 2.1246006389776357, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1330 + }, + { + "epoch": 2.126198083067093, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1331 + }, + { + "epoch": 2.1277955271565494, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1332 + }, + { + "epoch": 2.1293929712460065, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1333 + }, + { + "epoch": 2.130990415335463, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 1334 + }, + { + "epoch": 2.13258785942492, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 1335 + }, + { + "epoch": 2.134185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 1336 + }, + { + "epoch": 2.135782747603834, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1337 + }, + { + "epoch": 2.137380191693291, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1338 + }, + { + "epoch": 2.1389776357827475, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1036, + "step": 1339 + }, + { + "epoch": 2.1405750798722045, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1340 + }, + { + "epoch": 2.142172523961661, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1341 + }, + { + "epoch": 2.143769968051118, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.1077, + "step": 1342 + }, + { + "epoch": 2.1453674121405752, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1343 + }, + { + "epoch": 2.146964856230032, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1344 + }, + { + "epoch": 2.148562300319489, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1345 + }, + { + "epoch": 2.1501597444089455, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1346 + }, + { + "epoch": 2.1517571884984026, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 1347 + }, + { + "epoch": 2.1533546325878596, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1348 + }, + { + "epoch": 2.1549520766773163, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1349 + }, + { + "epoch": 2.1565495207667733, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.1033, + "step": 1350 + }, + { + "epoch": 2.15814696485623, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1351 + }, + { + "epoch": 2.159744408945687, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1352 + }, + { + "epoch": 2.1613418530351436, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1353 + }, + { + "epoch": 2.1629392971246006, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1052, + "step": 1354 + }, + { + "epoch": 2.1645367412140577, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1355 + }, + { + "epoch": 2.1661341853035143, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1356 + }, + { + "epoch": 2.1677316293929714, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1357 + }, + { + "epoch": 2.169329073482428, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.1053, + "step": 1358 + }, + { + "epoch": 2.170926517571885, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 1359 + }, + { + "epoch": 2.1725239616613417, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 1360 + }, + { + "epoch": 2.1741214057507987, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1361 + }, + { + "epoch": 2.1757188498402558, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1362 + }, + { + "epoch": 2.1773162939297124, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.103, + "step": 1363 + }, + { + "epoch": 2.1789137380191694, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1041, + "step": 1364 + }, + { + "epoch": 2.180511182108626, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 1365 + }, + { + "epoch": 2.182108626198083, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1366 + }, + { + "epoch": 2.18370607028754, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1367 + }, + { + "epoch": 2.1853035143769968, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1368 + }, + { + "epoch": 2.186900958466454, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1369 + }, + { + "epoch": 2.1884984025559104, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1370 + }, + { + "epoch": 2.1900958466453675, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1371 + }, + { + "epoch": 2.191693290734824, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1372 + }, + { + "epoch": 2.193290734824281, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1373 + }, + { + "epoch": 2.194888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1374 + }, + { + "epoch": 2.196485623003195, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1375 + }, + { + "epoch": 2.198083067092652, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 1376 + }, + { + "epoch": 2.1996805111821085, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1377 + }, + { + "epoch": 2.2012779552715656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.1082, + "step": 1378 + }, + { + "epoch": 2.202875399361022, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1379 + }, + { + "epoch": 2.2044728434504792, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 1380 + }, + { + "epoch": 2.2060702875399363, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1071, + "step": 1381 + }, + { + "epoch": 2.207667731629393, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1382 + }, + { + "epoch": 2.20926517571885, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.1006, + "step": 1383 + }, + { + "epoch": 2.2108626198083066, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1384 + }, + { + "epoch": 2.2124600638977636, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 1385 + }, + { + "epoch": 2.2140575079872207, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1386 + }, + { + "epoch": 2.2156549520766773, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.1021, + "step": 1387 + }, + { + "epoch": 2.2172523961661343, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1388 + }, + { + "epoch": 2.218849840255591, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 1389 + }, + { + "epoch": 2.220447284345048, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.103, + "step": 1390 + }, + { + "epoch": 2.2220447284345046, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 1391 + }, + { + "epoch": 2.2236421725239617, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1392 + }, + { + "epoch": 2.2252396166134187, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1393 + }, + { + "epoch": 2.2268370607028753, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1394 + }, + { + "epoch": 2.2284345047923324, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1085, + "step": 1395 + }, + { + "epoch": 2.230031948881789, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1396 + }, + { + "epoch": 2.231629392971246, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.1036, + "step": 1397 + }, + { + "epoch": 2.2332268370607027, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1398 + }, + { + "epoch": 2.2348242811501597, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 1399 + }, + { + "epoch": 2.236421725239617, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1400 + }, + { + "epoch": 2.2380191693290734, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1401 + }, + { + "epoch": 2.2396166134185305, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1402 + }, + { + "epoch": 2.241214057507987, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1403 + }, + { + "epoch": 2.242811501597444, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1404 + }, + { + "epoch": 2.244408945686901, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1405 + }, + { + "epoch": 2.246006389776358, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1406 + }, + { + "epoch": 2.247603833865815, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1407 + }, + { + "epoch": 2.2492012779552715, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1408 + }, + { + "epoch": 2.2507987220447285, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1409 + }, + { + "epoch": 2.252396166134185, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1410 + }, + { + "epoch": 2.253993610223642, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.1055, + "step": 1411 + }, + { + "epoch": 2.255591054313099, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1412 + }, + { + "epoch": 2.257188498402556, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 1413 + }, + { + "epoch": 2.258785942492013, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.1096, + "step": 1414 + }, + { + "epoch": 2.2603833865814695, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1415 + }, + { + "epoch": 2.2619808306709266, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1416 + }, + { + "epoch": 2.263578274760383, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1417 + }, + { + "epoch": 2.2651757188498403, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1418 + }, + { + "epoch": 2.2667731629392973, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1419 + }, + { + "epoch": 2.268370607028754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1420 + }, + { + "epoch": 2.269968051118211, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 1421 + }, + { + "epoch": 2.2715654952076676, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1422 + }, + { + "epoch": 2.2731629392971247, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1423 + }, + { + "epoch": 2.2747603833865817, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 1424 + }, + { + "epoch": 2.2763578274760383, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1425 + }, + { + "epoch": 2.2779552715654954, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.1138, + "step": 1426 + }, + { + "epoch": 2.279552715654952, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 1427 + }, + { + "epoch": 2.281150159744409, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1428 + }, + { + "epoch": 2.2827476038338657, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1429 + }, + { + "epoch": 2.2843450479233227, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 1430 + }, + { + "epoch": 2.2859424920127793, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1431 + }, + { + "epoch": 2.2875399361022364, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1432 + }, + { + "epoch": 2.2891373801916934, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1433 + }, + { + "epoch": 2.29073482428115, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 1434 + }, + { + "epoch": 2.292332268370607, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1435 + }, + { + "epoch": 2.2939297124600637, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1436 + }, + { + "epoch": 2.2955271565495208, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1437 + }, + { + "epoch": 2.297124600638978, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.103, + "step": 1438 + }, + { + "epoch": 2.2987220447284344, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1439 + }, + { + "epoch": 2.3003194888178915, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 1440 + }, + { + "epoch": 2.301916932907348, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1441 + }, + { + "epoch": 2.303514376996805, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1442 + }, + { + "epoch": 2.3051118210862622, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1443 + }, + { + "epoch": 2.306709265175719, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 1444 + }, + { + "epoch": 2.308306709265176, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1445 + }, + { + "epoch": 2.3099041533546325, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1446 + }, + { + "epoch": 2.3115015974440896, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1447 + }, + { + "epoch": 2.313099041533546, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 1448 + }, + { + "epoch": 2.3146964856230032, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1449 + }, + { + "epoch": 2.31629392971246, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1450 + }, + { + "epoch": 2.317891373801917, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1451 + }, + { + "epoch": 2.319488817891374, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 1452 + }, + { + "epoch": 2.3210862619808306, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1453 + }, + { + "epoch": 2.3226837060702876, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1454 + }, + { + "epoch": 2.3242811501597442, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 1455 + }, + { + "epoch": 2.3258785942492013, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1456 + }, + { + "epoch": 2.3274760383386583, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 1457 + }, + { + "epoch": 2.329073482428115, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1458 + }, + { + "epoch": 2.330670926517572, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1459 + }, + { + "epoch": 2.3322683706070286, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1460 + }, + { + "epoch": 2.3338658146964857, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 1461 + }, + { + "epoch": 2.3354632587859427, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1462 + }, + { + "epoch": 2.3370607028753994, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 1463 + }, + { + "epoch": 2.3386581469648564, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 1464 + }, + { + "epoch": 2.340255591054313, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1465 + }, + { + "epoch": 2.34185303514377, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1466 + }, + { + "epoch": 2.3434504792332267, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1467 + }, + { + "epoch": 2.3450479233226837, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.102, + "step": 1468 + }, + { + "epoch": 2.3466453674121404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 1469 + }, + { + "epoch": 2.3482428115015974, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1470 + }, + { + "epoch": 2.3498402555910545, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.104, + "step": 1471 + }, + { + "epoch": 2.351437699680511, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.1025, + "step": 1472 + }, + { + "epoch": 2.353035143769968, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1473 + }, + { + "epoch": 2.3546325878594248, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1474 + }, + { + "epoch": 2.356230031948882, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1475 + }, + { + "epoch": 2.357827476038339, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1476 + }, + { + "epoch": 2.3594249201277955, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1477 + }, + { + "epoch": 2.3610223642172525, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1478 + }, + { + "epoch": 2.362619808306709, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1479 + }, + { + "epoch": 2.364217252396166, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1480 + }, + { + "epoch": 2.365814696485623, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1481 + }, + { + "epoch": 2.36741214057508, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 1482 + }, + { + "epoch": 2.369009584664537, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1483 + }, + { + "epoch": 2.3706070287539935, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1484 + }, + { + "epoch": 2.3722044728434506, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1485 + }, + { + "epoch": 2.373801916932907, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1486 + }, + { + "epoch": 2.3753993610223643, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.107, + "step": 1487 + }, + { + "epoch": 2.376996805111821, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1488 + }, + { + "epoch": 2.378594249201278, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1489 + }, + { + "epoch": 2.380191693290735, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1490 + }, + { + "epoch": 2.3817891373801916, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1491 + }, + { + "epoch": 2.3833865814696487, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1492 + }, + { + "epoch": 2.3849840255591053, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 1493 + }, + { + "epoch": 2.3865814696485623, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1494 + }, + { + "epoch": 2.3881789137380194, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0984, + "step": 1495 + }, + { + "epoch": 2.389776357827476, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1496 + }, + { + "epoch": 2.391373801916933, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 1497 + }, + { + "epoch": 2.3929712460063897, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1498 + }, + { + "epoch": 2.3945686900958467, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1499 + }, + { + "epoch": 2.3961661341853033, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 1500 + }, + { + "epoch": 2.3977635782747604, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1501 + }, + { + "epoch": 2.3993610223642174, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 1502 + }, + { + "epoch": 2.400958466453674, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1503 + }, + { + "epoch": 2.402555910543131, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 1504 + }, + { + "epoch": 2.4041533546325877, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1505 + }, + { + "epoch": 2.405750798722045, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1506 + }, + { + "epoch": 2.4073482428115014, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1507 + }, + { + "epoch": 2.4089456869009584, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1044, + "step": 1508 + }, + { + "epoch": 2.4105431309904155, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 1509 + }, + { + "epoch": 2.412140575079872, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1510 + }, + { + "epoch": 2.413738019169329, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1511 + }, + { + "epoch": 2.415335463258786, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 1512 + }, + { + "epoch": 2.416932907348243, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1513 + }, + { + "epoch": 2.4185303514377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1514 + }, + { + "epoch": 2.4201277955271565, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1515 + }, + { + "epoch": 2.4217252396166136, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1516 + }, + { + "epoch": 2.42332268370607, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1517 + }, + { + "epoch": 2.4249201277955272, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 1518 + }, + { + "epoch": 2.426517571884984, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1519 + }, + { + "epoch": 2.428115015974441, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1520 + }, + { + "epoch": 2.4297124600638975, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1521 + }, + { + "epoch": 2.4313099041533546, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 1522 + }, + { + "epoch": 2.4329073482428116, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 1523 + }, + { + "epoch": 2.4345047923322682, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1524 + }, + { + "epoch": 2.4361022364217253, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 1525 + }, + { + "epoch": 2.437699680511182, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1526 + }, + { + "epoch": 2.439297124600639, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1527 + }, + { + "epoch": 2.440894568690096, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1528 + }, + { + "epoch": 2.4424920127795526, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.102, + "step": 1529 + }, + { + "epoch": 2.4440894568690097, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 1530 + }, + { + "epoch": 2.4456869009584663, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1531 + }, + { + "epoch": 2.4472843450479234, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 1532 + }, + { + "epoch": 2.4488817891373804, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1533 + }, + { + "epoch": 2.450479233226837, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1534 + }, + { + "epoch": 2.452076677316294, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1535 + }, + { + "epoch": 2.4536741214057507, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1536 + }, + { + "epoch": 2.4552715654952078, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1064, + "step": 1537 + }, + { + "epoch": 2.4568690095846644, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0997, + "step": 1538 + }, + { + "epoch": 2.4584664536741214, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1539 + }, + { + "epoch": 2.460063897763578, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1540 + }, + { + "epoch": 2.461661341853035, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1541 + }, + { + "epoch": 2.463258785942492, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1542 + }, + { + "epoch": 2.4648562300319488, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 1543 + }, + { + "epoch": 2.466453674121406, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1544 + }, + { + "epoch": 2.4680511182108624, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1545 + }, + { + "epoch": 2.4696485623003195, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1546 + }, + { + "epoch": 2.4712460063897765, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1547 + }, + { + "epoch": 2.472843450479233, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1548 + }, + { + "epoch": 2.47444089456869, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1549 + }, + { + "epoch": 2.476038338658147, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1550 + }, + { + "epoch": 2.477635782747604, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 1551 + }, + { + "epoch": 2.479233226837061, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1552 + }, + { + "epoch": 2.4808306709265175, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1553 + }, + { + "epoch": 2.4824281150159746, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1554 + }, + { + "epoch": 2.484025559105431, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1555 + }, + { + "epoch": 2.4856230031948883, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1556 + }, + { + "epoch": 2.487220447284345, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 1557 + }, + { + "epoch": 2.488817891373802, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1558 + }, + { + "epoch": 2.4904153354632586, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 1559 + }, + { + "epoch": 2.4920127795527156, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1560 + }, + { + "epoch": 2.4936102236421727, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1561 + }, + { + "epoch": 2.4952076677316293, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 1562 + }, + { + "epoch": 2.4968051118210863, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1563 + }, + { + "epoch": 2.498402555910543, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1564 + }, + { + "epoch": 2.5, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1565 + }, + { + "epoch": 2.501597444089457, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1566 + }, + { + "epoch": 2.5031948881789137, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1567 + }, + { + "epoch": 2.5047923322683707, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 1568 + }, + { + "epoch": 2.5063897763578273, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 1569 + }, + { + "epoch": 2.5079872204472844, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1570 + }, + { + "epoch": 2.5095846645367414, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1003, + "step": 1571 + }, + { + "epoch": 2.511182108626198, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1572 + }, + { + "epoch": 2.512779552715655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.096, + "step": 1573 + }, + { + "epoch": 2.5143769968051117, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1574 + }, + { + "epoch": 2.515974440894569, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.1051, + "step": 1575 + }, + { + "epoch": 2.5175718849840254, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1576 + }, + { + "epoch": 2.5191693290734825, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1577 + }, + { + "epoch": 2.520766773162939, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1578 + }, + { + "epoch": 2.522364217252396, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1579 + }, + { + "epoch": 2.523961661341853, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1003, + "step": 1580 + }, + { + "epoch": 2.52555910543131, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1581 + }, + { + "epoch": 2.527156549520767, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 1582 + }, + { + "epoch": 2.5287539936102235, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 1583 + }, + { + "epoch": 2.5303514376996805, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1584 + }, + { + "epoch": 2.5319488817891376, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1585 + }, + { + "epoch": 2.533546325878594, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1586 + }, + { + "epoch": 2.5351437699680512, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1587 + }, + { + "epoch": 2.536741214057508, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1588 + }, + { + "epoch": 2.538338658146965, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 1589 + }, + { + "epoch": 2.539936102236422, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1590 + }, + { + "epoch": 2.5415335463258786, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 1591 + }, + { + "epoch": 2.543130990415335, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1592 + }, + { + "epoch": 2.5447284345047922, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1593 + }, + { + "epoch": 2.5463258785942493, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1594 + }, + { + "epoch": 2.547923322683706, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1595 + }, + { + "epoch": 2.549520766773163, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1596 + }, + { + "epoch": 2.5511182108626196, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1597 + }, + { + "epoch": 2.5527156549520766, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1598 + }, + { + "epoch": 2.5543130990415337, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1599 + }, + { + "epoch": 2.5559105431309903, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 1600 + }, + { + "epoch": 2.5575079872204474, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1601 + }, + { + "epoch": 2.559105431309904, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1602 + }, + { + "epoch": 2.560702875399361, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1603 + }, + { + "epoch": 2.562300319488818, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.1003, + "step": 1604 + }, + { + "epoch": 2.5638977635782747, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 1605 + }, + { + "epoch": 2.5654952076677318, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1606 + }, + { + "epoch": 2.5670926517571884, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1607 + }, + { + "epoch": 2.5686900958466454, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.1006, + "step": 1608 + }, + { + "epoch": 2.5702875399361025, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.1053, + "step": 1609 + }, + { + "epoch": 2.571884984025559, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1610 + }, + { + "epoch": 2.5734824281150157, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1611 + }, + { + "epoch": 2.5750798722044728, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 1612 + }, + { + "epoch": 2.57667731629393, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1613 + }, + { + "epoch": 2.5782747603833864, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1614 + }, + { + "epoch": 2.5798722044728435, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 1615 + }, + { + "epoch": 2.5814696485623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1616 + }, + { + "epoch": 2.583067092651757, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1617 + }, + { + "epoch": 2.584664536741214, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1618 + }, + { + "epoch": 2.586261980830671, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1619 + }, + { + "epoch": 2.587859424920128, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 1620 + }, + { + "epoch": 2.5894568690095845, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1036, + "step": 1621 + }, + { + "epoch": 2.5910543130990416, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1622 + }, + { + "epoch": 2.5926517571884986, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1623 + }, + { + "epoch": 2.594249201277955, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1624 + }, + { + "epoch": 2.5958466453674123, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1625 + }, + { + "epoch": 2.597444089456869, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1025, + "step": 1626 + }, + { + "epoch": 2.599041533546326, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 1627 + }, + { + "epoch": 2.600638977635783, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0972, + "step": 1628 + }, + { + "epoch": 2.6022364217252396, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1629 + }, + { + "epoch": 2.6038338658146962, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1630 + }, + { + "epoch": 2.6054313099041533, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1631 + }, + { + "epoch": 2.6070287539936103, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 1632 + }, + { + "epoch": 2.608626198083067, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 1633 + }, + { + "epoch": 2.610223642172524, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1634 + }, + { + "epoch": 2.6118210862619806, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1635 + }, + { + "epoch": 2.6134185303514377, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 1636 + }, + { + "epoch": 2.6150159744408947, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1637 + }, + { + "epoch": 2.6166134185303513, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1638 + }, + { + "epoch": 2.6182108626198084, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1639 + }, + { + "epoch": 2.619808306709265, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1640 + }, + { + "epoch": 2.621405750798722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 1641 + }, + { + "epoch": 2.623003194888179, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1642 + }, + { + "epoch": 2.6246006389776357, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1643 + }, + { + "epoch": 2.626198083067093, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1644 + }, + { + "epoch": 2.6277955271565494, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1645 + }, + { + "epoch": 2.6293929712460065, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 1646 + }, + { + "epoch": 2.6309904153354635, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 1647 + }, + { + "epoch": 2.63258785942492, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1648 + }, + { + "epoch": 2.6341853035143767, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1649 + }, + { + "epoch": 2.635782747603834, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1650 + }, + { + "epoch": 2.637380191693291, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0997, + "step": 1651 + }, + { + "epoch": 2.6389776357827475, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1652 + }, + { + "epoch": 2.6405750798722045, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1653 + }, + { + "epoch": 2.642172523961661, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1654 + }, + { + "epoch": 2.643769968051118, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1655 + }, + { + "epoch": 2.6453674121405752, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1656 + }, + { + "epoch": 2.646964856230032, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1657 + }, + { + "epoch": 2.648562300319489, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0925, + "step": 1658 + }, + { + "epoch": 2.6501597444089455, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1659 + }, + { + "epoch": 2.6517571884984026, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1660 + }, + { + "epoch": 2.6533546325878596, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1661 + }, + { + "epoch": 2.6549520766773163, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1662 + }, + { + "epoch": 2.6565495207667733, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0992, + "step": 1663 + }, + { + "epoch": 2.65814696485623, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1664 + }, + { + "epoch": 2.659744408945687, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1665 + }, + { + "epoch": 2.661341853035144, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1666 + }, + { + "epoch": 2.6629392971246006, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1667 + }, + { + "epoch": 2.6645367412140573, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1668 + }, + { + "epoch": 2.6661341853035143, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1669 + }, + { + "epoch": 2.6677316293929714, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1670 + }, + { + "epoch": 2.669329073482428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1671 + }, + { + "epoch": 2.670926517571885, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1672 + }, + { + "epoch": 2.6725239616613417, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1673 + }, + { + "epoch": 2.6741214057507987, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1674 + }, + { + "epoch": 2.6757188498402558, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1675 + }, + { + "epoch": 2.6773162939297124, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1676 + }, + { + "epoch": 2.6789137380191694, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1677 + }, + { + "epoch": 2.680511182108626, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1678 + }, + { + "epoch": 2.682108626198083, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1679 + }, + { + "epoch": 2.68370607028754, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1680 + }, + { + "epoch": 2.6853035143769968, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1681 + }, + { + "epoch": 2.686900958466454, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 1682 + }, + { + "epoch": 2.6884984025559104, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 1683 + }, + { + "epoch": 2.6900958466453675, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 1684 + }, + { + "epoch": 2.6916932907348246, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1685 + }, + { + "epoch": 2.693290734824281, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1686 + }, + { + "epoch": 2.6948881789137378, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1687 + }, + { + "epoch": 2.696485623003195, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1688 + }, + { + "epoch": 2.698083067092652, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 1689 + }, + { + "epoch": 2.6996805111821085, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1690 + }, + { + "epoch": 2.7012779552715656, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 1691 + }, + { + "epoch": 2.702875399361022, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1692 + }, + { + "epoch": 2.7044728434504792, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1693 + }, + { + "epoch": 2.7060702875399363, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1694 + }, + { + "epoch": 2.707667731629393, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1695 + }, + { + "epoch": 2.70926517571885, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1696 + }, + { + "epoch": 2.7108626198083066, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.1024, + "step": 1697 + }, + { + "epoch": 2.7124600638977636, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1698 + }, + { + "epoch": 2.7140575079872207, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1699 + }, + { + "epoch": 2.7156549520766773, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1700 + }, + { + "epoch": 2.7172523961661343, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1701 + }, + { + "epoch": 2.718849840255591, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1702 + }, + { + "epoch": 2.720447284345048, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 1703 + }, + { + "epoch": 2.722044728434505, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0963, + "step": 1704 + }, + { + "epoch": 2.7236421725239617, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1705 + }, + { + "epoch": 2.7252396166134183, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 1706 + }, + { + "epoch": 2.7268370607028753, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0972, + "step": 1707 + }, + { + "epoch": 2.7284345047923324, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1708 + }, + { + "epoch": 2.730031948881789, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 1709 + }, + { + "epoch": 2.731629392971246, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1710 + }, + { + "epoch": 2.7332268370607027, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 1711 + }, + { + "epoch": 2.7348242811501597, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1712 + }, + { + "epoch": 2.736421725239617, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 1713 + }, + { + "epoch": 2.7380191693290734, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 1714 + }, + { + "epoch": 2.7396166134185305, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1715 + }, + { + "epoch": 2.741214057507987, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1716 + }, + { + "epoch": 2.742811501597444, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1717 + }, + { + "epoch": 2.744408945686901, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 1718 + }, + { + "epoch": 2.746006389776358, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 1719 + }, + { + "epoch": 2.747603833865815, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 1720 + }, + { + "epoch": 2.7492012779552715, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1721 + }, + { + "epoch": 2.7507987220447285, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0971, + "step": 1722 + }, + { + "epoch": 2.752396166134185, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 1723 + }, + { + "epoch": 2.753993610223642, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1724 + }, + { + "epoch": 2.755591054313099, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1725 + }, + { + "epoch": 2.757188498402556, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1726 + }, + { + "epoch": 2.758785942492013, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1727 + }, + { + "epoch": 2.7603833865814695, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1728 + }, + { + "epoch": 2.7619808306709266, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1729 + }, + { + "epoch": 2.763578274760383, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0963, + "step": 1730 + }, + { + "epoch": 2.7651757188498403, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1731 + }, + { + "epoch": 2.7667731629392973, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 1732 + }, + { + "epoch": 2.768370607028754, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1733 + }, + { + "epoch": 2.769968051118211, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 1734 + }, + { + "epoch": 2.7715654952076676, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1735 + }, + { + "epoch": 2.7731629392971247, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1736 + }, + { + "epoch": 2.7747603833865817, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1737 + }, + { + "epoch": 2.7763578274760383, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 1738 + }, + { + "epoch": 2.777955271565495, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 1739 + }, + { + "epoch": 2.779552715654952, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1740 + }, + { + "epoch": 2.781150159744409, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 1741 + }, + { + "epoch": 2.7827476038338657, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1742 + }, + { + "epoch": 2.7843450479233227, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1743 + }, + { + "epoch": 2.7859424920127793, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 1744 + }, + { + "epoch": 2.7875399361022364, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1745 + }, + { + "epoch": 2.7891373801916934, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1746 + }, + { + "epoch": 2.79073482428115, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1747 + }, + { + "epoch": 2.792332268370607, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1748 + }, + { + "epoch": 2.7939297124600637, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1749 + }, + { + "epoch": 2.7955271565495208, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 1750 + }, + { + "epoch": 2.797124600638978, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1751 + }, + { + "epoch": 2.7987220447284344, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1752 + }, + { + "epoch": 2.8003194888178915, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1753 + }, + { + "epoch": 2.801916932907348, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 1754 + }, + { + "epoch": 2.803514376996805, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 1755 + }, + { + "epoch": 2.8051118210862622, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 1756 + }, + { + "epoch": 2.806709265175719, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1757 + }, + { + "epoch": 2.8083067092651754, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1758 + }, + { + "epoch": 2.8099041533546325, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1759 + }, + { + "epoch": 2.8115015974440896, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1760 + }, + { + "epoch": 2.813099041533546, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 1761 + }, + { + "epoch": 2.8146964856230032, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1762 + }, + { + "epoch": 2.81629392971246, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.096, + "step": 1763 + }, + { + "epoch": 2.817891373801917, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1764 + }, + { + "epoch": 2.819488817891374, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 1765 + }, + { + "epoch": 2.8210862619808306, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1766 + }, + { + "epoch": 2.8226837060702876, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1767 + }, + { + "epoch": 2.8242811501597442, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1768 + }, + { + "epoch": 2.8258785942492013, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1769 + }, + { + "epoch": 2.8274760383386583, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 1770 + }, + { + "epoch": 2.829073482428115, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1771 + }, + { + "epoch": 2.830670926517572, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 1772 + }, + { + "epoch": 2.8322683706070286, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1773 + }, + { + "epoch": 2.8338658146964857, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1774 + }, + { + "epoch": 2.8354632587859427, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1775 + }, + { + "epoch": 2.8370607028753994, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 1776 + }, + { + "epoch": 2.838658146964856, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1777 + }, + { + "epoch": 2.840255591054313, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1778 + }, + { + "epoch": 2.84185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1779 + }, + { + "epoch": 2.8434504792332267, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1780 + }, + { + "epoch": 2.8450479233226837, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1781 + }, + { + "epoch": 2.8466453674121404, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1782 + }, + { + "epoch": 2.8482428115015974, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1783 + }, + { + "epoch": 2.8498402555910545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 1784 + }, + { + "epoch": 2.851437699680511, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1785 + }, + { + "epoch": 2.853035143769968, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1786 + }, + { + "epoch": 2.8546325878594248, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 1787 + }, + { + "epoch": 2.856230031948882, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 1788 + }, + { + "epoch": 2.857827476038339, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 1789 + }, + { + "epoch": 2.8594249201277955, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1790 + }, + { + "epoch": 2.8610223642172525, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 1791 + }, + { + "epoch": 2.862619808306709, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 1792 + }, + { + "epoch": 2.864217252396166, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 1793 + }, + { + "epoch": 2.8658146964856233, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1794 + }, + { + "epoch": 2.86741214057508, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1795 + }, + { + "epoch": 2.8690095846645365, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 1796 + }, + { + "epoch": 2.8706070287539935, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1797 + }, + { + "epoch": 2.8722044728434506, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 1798 + }, + { + "epoch": 2.873801916932907, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1799 + }, + { + "epoch": 2.8753993610223643, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1800 + }, + { + "epoch": 2.876996805111821, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 1801 + }, + { + "epoch": 2.878594249201278, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 1802 + }, + { + "epoch": 2.880191693290735, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0997, + "step": 1803 + }, + { + "epoch": 2.8817891373801916, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1804 + }, + { + "epoch": 2.8833865814696487, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 1805 + }, + { + "epoch": 2.8849840255591053, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 1806 + }, + { + "epoch": 2.8865814696485623, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 1807 + }, + { + "epoch": 2.8881789137380194, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1808 + }, + { + "epoch": 2.889776357827476, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1809 + }, + { + "epoch": 2.891373801916933, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1810 + }, + { + "epoch": 2.8929712460063897, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1811 + }, + { + "epoch": 2.8945686900958467, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0966, + "step": 1812 + }, + { + "epoch": 2.8961661341853038, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 1813 + }, + { + "epoch": 2.8977635782747604, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 1814 + }, + { + "epoch": 2.899361022364217, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 1815 + }, + { + "epoch": 2.900958466453674, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1816 + }, + { + "epoch": 2.902555910543131, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 1817 + }, + { + "epoch": 2.9041533546325877, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1818 + }, + { + "epoch": 2.905750798722045, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 1819 + }, + { + "epoch": 2.9073482428115014, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 1820 + }, + { + "epoch": 2.9089456869009584, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1821 + }, + { + "epoch": 2.9105431309904155, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 1822 + }, + { + "epoch": 2.912140575079872, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.1005, + "step": 1823 + }, + { + "epoch": 2.913738019169329, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 1824 + }, + { + "epoch": 2.915335463258786, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 1825 + }, + { + "epoch": 2.916932907348243, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1826 + }, + { + "epoch": 2.9185303514377, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 1827 + }, + { + "epoch": 2.9201277955271565, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1828 + }, + { + "epoch": 2.9217252396166136, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 1829 + }, + { + "epoch": 2.92332268370607, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 1830 + }, + { + "epoch": 2.9249201277955272, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1831 + }, + { + "epoch": 2.9265175718849843, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1832 + }, + { + "epoch": 2.928115015974441, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1833 + }, + { + "epoch": 2.9297124600638975, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 1834 + }, + { + "epoch": 2.9313099041533546, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1835 + }, + { + "epoch": 2.9329073482428116, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 1836 + }, + { + "epoch": 2.9345047923322682, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 1837 + }, + { + "epoch": 2.9361022364217253, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1838 + }, + { + "epoch": 2.937699680511182, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0966, + "step": 1839 + }, + { + "epoch": 2.939297124600639, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1840 + }, + { + "epoch": 2.940894568690096, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 1841 + }, + { + "epoch": 2.9424920127795526, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 1842 + }, + { + "epoch": 2.9440894568690097, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1843 + }, + { + "epoch": 2.9456869009584663, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1844 + }, + { + "epoch": 2.9472843450479234, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 1845 + }, + { + "epoch": 2.9488817891373804, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 1846 + }, + { + "epoch": 2.950479233226837, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 1847 + }, + { + "epoch": 2.952076677316294, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 1848 + }, + { + "epoch": 2.9536741214057507, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1849 + }, + { + "epoch": 2.9552715654952078, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 1850 + }, + { + "epoch": 2.956869009584665, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0925, + "step": 1851 + }, + { + "epoch": 2.9584664536741214, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1852 + }, + { + "epoch": 2.960063897763578, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1853 + }, + { + "epoch": 2.961661341853035, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 1854 + }, + { + "epoch": 2.963258785942492, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 1855 + }, + { + "epoch": 2.9648562300319488, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 1856 + }, + { + "epoch": 2.966453674121406, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1857 + }, + { + "epoch": 2.9680511182108624, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 1858 + }, + { + "epoch": 2.9696485623003195, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 1859 + }, + { + "epoch": 2.9712460063897765, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1860 + }, + { + "epoch": 2.972843450479233, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1861 + }, + { + "epoch": 2.97444089456869, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1862 + }, + { + "epoch": 2.976038338658147, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1863 + }, + { + "epoch": 2.977635782747604, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 1864 + }, + { + "epoch": 2.979233226837061, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 1865 + }, + { + "epoch": 2.9808306709265175, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1866 + }, + { + "epoch": 2.9824281150159746, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1867 + }, + { + "epoch": 2.984025559105431, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1868 + }, + { + "epoch": 2.9856230031948883, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 1869 + }, + { + "epoch": 2.987220447284345, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 1870 + }, + { + "epoch": 2.988817891373802, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 1871 + }, + { + "epoch": 2.9904153354632586, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 1872 + }, + { + "epoch": 2.9920127795527156, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 1873 + }, + { + "epoch": 2.9936102236421727, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1874 + }, + { + "epoch": 2.9952076677316293, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1875 + }, + { + "epoch": 2.9968051118210863, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1876 + }, + { + "epoch": 2.998402555910543, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1877 + }, + { + "epoch": 3.0, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1878 + }, + { + "epoch": 3.001597444089457, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 1879 + }, + { + "epoch": 3.0031948881789137, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0925, + "step": 1880 + }, + { + "epoch": 3.0047923322683707, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1881 + }, + { + "epoch": 3.0063897763578273, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 1882 + }, + { + "epoch": 3.0079872204472844, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1883 + }, + { + "epoch": 3.009584664536741, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 1884 + }, + { + "epoch": 3.011182108626198, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1885 + }, + { + "epoch": 3.012779552715655, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 1886 + }, + { + "epoch": 3.0143769968051117, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1887 + }, + { + "epoch": 3.015974440894569, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 1888 + }, + { + "epoch": 3.0175718849840254, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0919, + "step": 1889 + }, + { + "epoch": 3.0191693290734825, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1890 + }, + { + "epoch": 3.0207667731629395, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 1891 + }, + { + "epoch": 3.022364217252396, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1892 + }, + { + "epoch": 3.023961661341853, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1893 + }, + { + "epoch": 3.02555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 1894 + }, + { + "epoch": 3.027156549520767, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1895 + }, + { + "epoch": 3.0287539936102235, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1896 + }, + { + "epoch": 3.0303514376996805, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1897 + }, + { + "epoch": 3.0319488817891376, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 1898 + }, + { + "epoch": 3.033546325878594, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 1899 + }, + { + "epoch": 3.0351437699680512, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 1900 + }, + { + "epoch": 3.036741214057508, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 1901 + }, + { + "epoch": 3.038338658146965, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1902 + }, + { + "epoch": 3.0399361022364215, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1903 + }, + { + "epoch": 3.0415335463258786, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1904 + }, + { + "epoch": 3.0431309904153356, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1905 + }, + { + "epoch": 3.0447284345047922, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1906 + }, + { + "epoch": 3.0463258785942493, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 1907 + }, + { + "epoch": 3.047923322683706, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 1908 + }, + { + "epoch": 3.049520766773163, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 1909 + }, + { + "epoch": 3.0511182108626196, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 1910 + }, + { + "epoch": 3.0527156549520766, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1911 + }, + { + "epoch": 3.0543130990415337, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 1912 + }, + { + "epoch": 3.0559105431309903, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1913 + }, + { + "epoch": 3.0575079872204474, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 1914 + }, + { + "epoch": 3.059105431309904, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1915 + }, + { + "epoch": 3.060702875399361, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 1916 + }, + { + "epoch": 3.062300319488818, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0963, + "step": 1917 + }, + { + "epoch": 3.0638977635782747, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.096, + "step": 1918 + }, + { + "epoch": 3.0654952076677318, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1919 + }, + { + "epoch": 3.0670926517571884, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 1920 + }, + { + "epoch": 3.0686900958466454, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 1921 + }, + { + "epoch": 3.070287539936102, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 1922 + }, + { + "epoch": 3.071884984025559, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 1923 + }, + { + "epoch": 3.073482428115016, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1924 + }, + { + "epoch": 3.0750798722044728, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0929, + "step": 1925 + }, + { + "epoch": 3.07667731629393, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1926 + }, + { + "epoch": 3.0782747603833864, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1927 + }, + { + "epoch": 3.0798722044728435, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1928 + }, + { + "epoch": 3.0814696485623, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 1929 + }, + { + "epoch": 3.083067092651757, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 1930 + }, + { + "epoch": 3.084664536741214, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1931 + }, + { + "epoch": 3.086261980830671, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 1932 + }, + { + "epoch": 3.087859424920128, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 1933 + }, + { + "epoch": 3.0894568690095845, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1934 + }, + { + "epoch": 3.0910543130990416, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1935 + }, + { + "epoch": 3.0926517571884986, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 1936 + }, + { + "epoch": 3.094249201277955, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 1937 + }, + { + "epoch": 3.0958466453674123, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 1938 + }, + { + "epoch": 3.097444089456869, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1939 + }, + { + "epoch": 3.099041533546326, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 1940 + }, + { + "epoch": 3.1006389776357826, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 1941 + }, + { + "epoch": 3.1022364217252396, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 1942 + }, + { + "epoch": 3.1038338658146967, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 1943 + }, + { + "epoch": 3.1054313099041533, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1944 + }, + { + "epoch": 3.1070287539936103, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1945 + }, + { + "epoch": 3.108626198083067, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1946 + }, + { + "epoch": 3.110223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 1947 + }, + { + "epoch": 3.1118210862619806, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 1948 + }, + { + "epoch": 3.1134185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1949 + }, + { + "epoch": 3.1150159744408947, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 1950 + }, + { + "epoch": 3.1166134185303513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1951 + }, + { + "epoch": 3.1182108626198084, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1952 + }, + { + "epoch": 3.119808306709265, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 1953 + }, + { + "epoch": 3.121405750798722, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1954 + }, + { + "epoch": 3.123003194888179, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 1955 + }, + { + "epoch": 3.1246006389776357, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1956 + }, + { + "epoch": 3.126198083067093, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 1957 + }, + { + "epoch": 3.1277955271565494, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1958 + }, + { + "epoch": 3.1293929712460065, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 1959 + }, + { + "epoch": 3.130990415335463, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 1960 + }, + { + "epoch": 3.13258785942492, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 1961 + }, + { + "epoch": 3.134185303514377, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 1962 + }, + { + "epoch": 3.135782747603834, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1963 + }, + { + "epoch": 3.137380191693291, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 1964 + }, + { + "epoch": 3.1389776357827475, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1965 + }, + { + "epoch": 3.1405750798722045, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 1966 + }, + { + "epoch": 3.142172523961661, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 1967 + }, + { + "epoch": 3.143769968051118, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1968 + }, + { + "epoch": 3.1453674121405752, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 1969 + }, + { + "epoch": 3.146964856230032, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 1970 + }, + { + "epoch": 3.148562300319489, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1971 + }, + { + "epoch": 3.1501597444089455, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 1972 + }, + { + "epoch": 3.1517571884984026, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 1973 + }, + { + "epoch": 3.1533546325878596, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1974 + }, + { + "epoch": 3.1549520766773163, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 1975 + }, + { + "epoch": 3.1565495207667733, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1976 + }, + { + "epoch": 3.15814696485623, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 1977 + }, + { + "epoch": 3.159744408945687, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1978 + }, + { + "epoch": 3.1613418530351436, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 1979 + }, + { + "epoch": 3.1629392971246006, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1980 + }, + { + "epoch": 3.1645367412140577, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 1981 + }, + { + "epoch": 3.1661341853035143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 1982 + }, + { + "epoch": 3.1677316293929714, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 1983 + }, + { + "epoch": 3.169329073482428, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1984 + }, + { + "epoch": 3.170926517571885, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1985 + }, + { + "epoch": 3.1725239616613417, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 1986 + }, + { + "epoch": 3.1741214057507987, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1987 + }, + { + "epoch": 3.1757188498402558, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 1988 + }, + { + "epoch": 3.1773162939297124, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 1989 + }, + { + "epoch": 3.1789137380191694, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 1990 + }, + { + "epoch": 3.180511182108626, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 1991 + }, + { + "epoch": 3.182108626198083, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 1992 + }, + { + "epoch": 3.18370607028754, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 1993 + }, + { + "epoch": 3.1853035143769968, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 1994 + }, + { + "epoch": 3.186900958466454, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0929, + "step": 1995 + }, + { + "epoch": 3.1884984025559104, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 1996 + }, + { + "epoch": 3.1900958466453675, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1997 + }, + { + "epoch": 3.191693290734824, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 1998 + }, + { + "epoch": 3.193290734824281, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 1999 + }, + { + "epoch": 3.194888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2000 + }, + { + "epoch": 3.196485623003195, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 2001 + }, + { + "epoch": 3.198083067092652, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 2002 + }, + { + "epoch": 3.1996805111821085, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 2003 + }, + { + "epoch": 3.2012779552715656, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 2004 + }, + { + "epoch": 3.202875399361022, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2005 + }, + { + "epoch": 3.2044728434504792, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2006 + }, + { + "epoch": 3.2060702875399363, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 2007 + }, + { + "epoch": 3.207667731629393, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2008 + }, + { + "epoch": 3.20926517571885, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0975, + "step": 2009 + }, + { + "epoch": 3.2108626198083066, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2010 + }, + { + "epoch": 3.2124600638977636, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 2011 + }, + { + "epoch": 3.2140575079872207, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2012 + }, + { + "epoch": 3.2156549520766773, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2013 + }, + { + "epoch": 3.2172523961661343, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2014 + }, + { + "epoch": 3.218849840255591, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2015 + }, + { + "epoch": 3.220447284345048, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 2016 + }, + { + "epoch": 3.2220447284345046, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 2017 + }, + { + "epoch": 3.2236421725239617, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 2018 + }, + { + "epoch": 3.2252396166134187, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2019 + }, + { + "epoch": 3.2268370607028753, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2020 + }, + { + "epoch": 3.2284345047923324, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2021 + }, + { + "epoch": 3.230031948881789, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2022 + }, + { + "epoch": 3.231629392971246, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2023 + }, + { + "epoch": 3.2332268370607027, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2024 + }, + { + "epoch": 3.2348242811501597, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2025 + }, + { + "epoch": 3.236421725239617, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 2026 + }, + { + "epoch": 3.2380191693290734, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2027 + }, + { + "epoch": 3.2396166134185305, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2028 + }, + { + "epoch": 3.241214057507987, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2029 + }, + { + "epoch": 3.242811501597444, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2030 + }, + { + "epoch": 3.244408945686901, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 2031 + }, + { + "epoch": 3.246006389776358, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2032 + }, + { + "epoch": 3.247603833865815, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2033 + }, + { + "epoch": 3.2492012779552715, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2034 + }, + { + "epoch": 3.2507987220447285, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2035 + }, + { + "epoch": 3.252396166134185, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2036 + }, + { + "epoch": 3.253993610223642, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2037 + }, + { + "epoch": 3.255591054313099, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2038 + }, + { + "epoch": 3.257188498402556, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2039 + }, + { + "epoch": 3.258785942492013, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2040 + }, + { + "epoch": 3.2603833865814695, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2041 + }, + { + "epoch": 3.2619808306709266, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 2042 + }, + { + "epoch": 3.263578274760383, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2043 + }, + { + "epoch": 3.2651757188498403, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2044 + }, + { + "epoch": 3.2667731629392973, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2045 + }, + { + "epoch": 3.268370607028754, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2046 + }, + { + "epoch": 3.269968051118211, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 2047 + }, + { + "epoch": 3.2715654952076676, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2048 + }, + { + "epoch": 3.2731629392971247, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2049 + }, + { + "epoch": 3.2747603833865817, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 2050 + }, + { + "epoch": 3.2763578274760383, + "grad_norm": 0.404296875, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 2051 + }, + { + "epoch": 3.2779552715654954, + "grad_norm": 0.8984375, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 2052 + }, + { + "epoch": 3.279552715654952, + "grad_norm": 2.359375, + "learning_rate": 0.0005, + "loss": 1.1479, + "step": 2053 + }, + { + "epoch": 3.281150159744409, + "grad_norm": 0.58984375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 2054 + }, + { + "epoch": 3.2827476038338657, + "grad_norm": 2.0, + "learning_rate": 0.0005, + "loss": 1.1742, + "step": 2055 + }, + { + "epoch": 3.2843450479233227, + "grad_norm": 0.85546875, + "learning_rate": 0.0005, + "loss": 1.1357, + "step": 2056 + }, + { + "epoch": 3.2859424920127793, + "grad_norm": 0.55078125, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 2057 + }, + { + "epoch": 3.2875399361022364, + "grad_norm": 0.5546875, + "learning_rate": 0.0005, + "loss": 1.1345, + "step": 2058 + }, + { + "epoch": 3.2891373801916934, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 2059 + }, + { + "epoch": 3.29073482428115, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.1267, + "step": 2060 + }, + { + "epoch": 3.292332268370607, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.1142, + "step": 2061 + }, + { + "epoch": 3.2939297124600637, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.1185, + "step": 2062 + }, + { + "epoch": 3.2955271565495208, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.1149, + "step": 2063 + }, + { + "epoch": 3.297124600638978, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.1106, + "step": 2064 + }, + { + "epoch": 3.2987220447284344, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.113, + "step": 2065 + }, + { + "epoch": 3.3003194888178915, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1077, + "step": 2066 + }, + { + "epoch": 3.301916932907348, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 2067 + }, + { + "epoch": 3.303514376996805, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 2068 + }, + { + "epoch": 3.3051118210862622, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 2069 + }, + { + "epoch": 3.306709265175719, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.1052, + "step": 2070 + }, + { + "epoch": 3.308306709265176, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.1082, + "step": 2071 + }, + { + "epoch": 3.3099041533546325, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 2072 + }, + { + "epoch": 3.3115015974440896, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2073 + }, + { + "epoch": 3.313099041533546, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 2074 + }, + { + "epoch": 3.3146964856230032, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 2075 + }, + { + "epoch": 3.31629392971246, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 2076 + }, + { + "epoch": 3.317891373801917, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 2077 + }, + { + "epoch": 3.319488817891374, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 2078 + }, + { + "epoch": 3.3210862619808306, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 2079 + }, + { + "epoch": 3.3226837060702876, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 2080 + }, + { + "epoch": 3.3242811501597442, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 2081 + }, + { + "epoch": 3.3258785942492013, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 2082 + }, + { + "epoch": 3.3274760383386583, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 2083 + }, + { + "epoch": 3.329073482428115, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2084 + }, + { + "epoch": 3.330670926517572, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2085 + }, + { + "epoch": 3.3322683706070286, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 2086 + }, + { + "epoch": 3.3338658146964857, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 2087 + }, + { + "epoch": 3.3354632587859427, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 2088 + }, + { + "epoch": 3.3370607028753994, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 2089 + }, + { + "epoch": 3.3386581469648564, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2090 + }, + { + "epoch": 3.340255591054313, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2091 + }, + { + "epoch": 3.34185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2092 + }, + { + "epoch": 3.3434504792332267, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2093 + }, + { + "epoch": 3.3450479233226837, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 2094 + }, + { + "epoch": 3.3466453674121404, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2095 + }, + { + "epoch": 3.3482428115015974, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2096 + }, + { + "epoch": 3.3498402555910545, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 2097 + }, + { + "epoch": 3.351437699680511, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2098 + }, + { + "epoch": 3.353035143769968, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 2099 + }, + { + "epoch": 3.3546325878594248, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2100 + }, + { + "epoch": 3.356230031948882, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2101 + }, + { + "epoch": 3.357827476038339, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2102 + }, + { + "epoch": 3.3594249201277955, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 2103 + }, + { + "epoch": 3.3610223642172525, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2104 + }, + { + "epoch": 3.362619808306709, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 2105 + }, + { + "epoch": 3.364217252396166, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 2106 + }, + { + "epoch": 3.365814696485623, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2107 + }, + { + "epoch": 3.36741214057508, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2108 + }, + { + "epoch": 3.369009584664537, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2109 + }, + { + "epoch": 3.3706070287539935, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2110 + }, + { + "epoch": 3.3722044728434506, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2111 + }, + { + "epoch": 3.373801916932907, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2112 + }, + { + "epoch": 3.3753993610223643, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2113 + }, + { + "epoch": 3.376996805111821, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2114 + }, + { + "epoch": 3.378594249201278, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2115 + }, + { + "epoch": 3.380191693290735, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2116 + }, + { + "epoch": 3.3817891373801916, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 2117 + }, + { + "epoch": 3.3833865814696487, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2118 + }, + { + "epoch": 3.3849840255591053, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2119 + }, + { + "epoch": 3.3865814696485623, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2120 + }, + { + "epoch": 3.3881789137380194, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2121 + }, + { + "epoch": 3.389776357827476, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 2122 + }, + { + "epoch": 3.391373801916933, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2123 + }, + { + "epoch": 3.3929712460063897, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 2124 + }, + { + "epoch": 3.3945686900958467, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 2125 + }, + { + "epoch": 3.3961661341853033, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 2126 + }, + { + "epoch": 3.3977635782747604, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2127 + }, + { + "epoch": 3.3993610223642174, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 2128 + }, + { + "epoch": 3.400958466453674, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2129 + }, + { + "epoch": 3.402555910543131, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0937, + "step": 2130 + }, + { + "epoch": 3.4041533546325877, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2131 + }, + { + "epoch": 3.405750798722045, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2132 + }, + { + "epoch": 3.4073482428115014, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2133 + }, + { + "epoch": 3.4089456869009584, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0919, + "step": 2134 + }, + { + "epoch": 3.4105431309904155, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2135 + }, + { + "epoch": 3.412140575079872, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2136 + }, + { + "epoch": 3.413738019169329, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2137 + }, + { + "epoch": 3.415335463258786, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2138 + }, + { + "epoch": 3.416932907348243, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2139 + }, + { + "epoch": 3.4185303514377, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 2140 + }, + { + "epoch": 3.4201277955271565, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2141 + }, + { + "epoch": 3.4217252396166136, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2142 + }, + { + "epoch": 3.42332268370607, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2143 + }, + { + "epoch": 3.4249201277955272, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2144 + }, + { + "epoch": 3.426517571884984, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2145 + }, + { + "epoch": 3.428115015974441, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2146 + }, + { + "epoch": 3.4297124600638975, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2147 + }, + { + "epoch": 3.4313099041533546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2148 + }, + { + "epoch": 3.4329073482428116, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2149 + }, + { + "epoch": 3.4345047923322682, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2150 + }, + { + "epoch": 3.4361022364217253, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2151 + }, + { + "epoch": 3.437699680511182, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0987, + "step": 2152 + }, + { + "epoch": 3.439297124600639, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 2153 + }, + { + "epoch": 3.440894568690096, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2154 + }, + { + "epoch": 3.4424920127795526, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2155 + }, + { + "epoch": 3.4440894568690097, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2156 + }, + { + "epoch": 3.4456869009584663, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2157 + }, + { + "epoch": 3.4472843450479234, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2158 + }, + { + "epoch": 3.4488817891373804, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 2159 + }, + { + "epoch": 3.450479233226837, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 2160 + }, + { + "epoch": 3.452076677316294, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2161 + }, + { + "epoch": 3.4536741214057507, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2162 + }, + { + "epoch": 3.4552715654952078, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2163 + }, + { + "epoch": 3.4568690095846644, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 2164 + }, + { + "epoch": 3.4584664536741214, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 2165 + }, + { + "epoch": 3.460063897763578, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2166 + }, + { + "epoch": 3.461661341853035, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2167 + }, + { + "epoch": 3.463258785942492, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2168 + }, + { + "epoch": 3.4648562300319488, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2169 + }, + { + "epoch": 3.466453674121406, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2170 + }, + { + "epoch": 3.4680511182108624, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 2171 + }, + { + "epoch": 3.4696485623003195, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2172 + }, + { + "epoch": 3.4712460063897765, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2173 + }, + { + "epoch": 3.472843450479233, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2174 + }, + { + "epoch": 3.47444089456869, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2175 + }, + { + "epoch": 3.476038338658147, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0929, + "step": 2176 + }, + { + "epoch": 3.477635782747604, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2177 + }, + { + "epoch": 3.479233226837061, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2178 + }, + { + "epoch": 3.4808306709265175, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2179 + }, + { + "epoch": 3.4824281150159746, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2180 + }, + { + "epoch": 3.484025559105431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0946, + "step": 2181 + }, + { + "epoch": 3.4856230031948883, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2182 + }, + { + "epoch": 3.487220447284345, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 2183 + }, + { + "epoch": 3.488817891373802, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2184 + }, + { + "epoch": 3.4904153354632586, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 2185 + }, + { + "epoch": 3.4920127795527156, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2186 + }, + { + "epoch": 3.4936102236421727, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2187 + }, + { + "epoch": 3.4952076677316293, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 2188 + }, + { + "epoch": 3.4968051118210863, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2189 + }, + { + "epoch": 3.498402555910543, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2190 + }, + { + "epoch": 3.5, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2191 + }, + { + "epoch": 3.501597444089457, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 2192 + }, + { + "epoch": 3.5031948881789137, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0983, + "step": 2193 + }, + { + "epoch": 3.5047923322683707, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2194 + }, + { + "epoch": 3.5063897763578273, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 2195 + }, + { + "epoch": 3.5079872204472844, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 2196 + }, + { + "epoch": 3.5095846645367414, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2197 + }, + { + "epoch": 3.511182108626198, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2198 + }, + { + "epoch": 3.512779552715655, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2199 + }, + { + "epoch": 3.5143769968051117, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2200 + }, + { + "epoch": 3.515974440894569, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2201 + }, + { + "epoch": 3.5175718849840254, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2202 + }, + { + "epoch": 3.5191693290734825, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 2203 + }, + { + "epoch": 3.520766773162939, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2204 + }, + { + "epoch": 3.522364217252396, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2205 + }, + { + "epoch": 3.523961661341853, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2206 + }, + { + "epoch": 3.52555910543131, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.1006, + "step": 2207 + }, + { + "epoch": 3.527156549520767, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2208 + }, + { + "epoch": 3.5287539936102235, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2209 + }, + { + "epoch": 3.5303514376996805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2210 + }, + { + "epoch": 3.5319488817891376, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2211 + }, + { + "epoch": 3.533546325878594, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2212 + }, + { + "epoch": 3.5351437699680512, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2213 + }, + { + "epoch": 3.536741214057508, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2214 + }, + { + "epoch": 3.538338658146965, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2215 + }, + { + "epoch": 3.539936102236422, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2216 + }, + { + "epoch": 3.5415335463258786, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2217 + }, + { + "epoch": 3.543130990415335, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 2218 + }, + { + "epoch": 3.5447284345047922, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 2219 + }, + { + "epoch": 3.5463258785942493, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2220 + }, + { + "epoch": 3.547923322683706, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2221 + }, + { + "epoch": 3.549520766773163, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2222 + }, + { + "epoch": 3.5511182108626196, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2223 + }, + { + "epoch": 3.5527156549520766, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2224 + }, + { + "epoch": 3.5543130990415337, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2225 + }, + { + "epoch": 3.5559105431309903, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 2226 + }, + { + "epoch": 3.5575079872204474, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 2227 + }, + { + "epoch": 3.559105431309904, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2228 + }, + { + "epoch": 3.560702875399361, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2229 + }, + { + "epoch": 3.562300319488818, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 2230 + }, + { + "epoch": 3.5638977635782747, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 2231 + }, + { + "epoch": 3.5654952076677318, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2232 + }, + { + "epoch": 3.5670926517571884, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2233 + }, + { + "epoch": 3.5686900958466454, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2234 + }, + { + "epoch": 3.5702875399361025, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2235 + }, + { + "epoch": 3.571884984025559, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2236 + }, + { + "epoch": 3.5734824281150157, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2237 + }, + { + "epoch": 3.5750798722044728, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2238 + }, + { + "epoch": 3.57667731629393, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2239 + }, + { + "epoch": 3.5782747603833864, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2240 + }, + { + "epoch": 3.5798722044728435, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2241 + }, + { + "epoch": 3.5814696485623, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2242 + }, + { + "epoch": 3.583067092651757, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2243 + }, + { + "epoch": 3.584664536741214, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2244 + }, + { + "epoch": 3.586261980830671, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 2245 + }, + { + "epoch": 3.587859424920128, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2246 + }, + { + "epoch": 3.5894568690095845, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2247 + }, + { + "epoch": 3.5910543130990416, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 2248 + }, + { + "epoch": 3.5926517571884986, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2249 + }, + { + "epoch": 3.594249201277955, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2250 + }, + { + "epoch": 3.5958466453674123, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2251 + }, + { + "epoch": 3.597444089456869, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2252 + }, + { + "epoch": 3.599041533546326, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2253 + }, + { + "epoch": 3.600638977635783, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2254 + }, + { + "epoch": 3.6022364217252396, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2255 + }, + { + "epoch": 3.6038338658146962, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2256 + }, + { + "epoch": 3.6054313099041533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2257 + }, + { + "epoch": 3.6070287539936103, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2258 + }, + { + "epoch": 3.608626198083067, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 2259 + }, + { + "epoch": 3.610223642172524, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2260 + }, + { + "epoch": 3.6118210862619806, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2261 + }, + { + "epoch": 3.6134185303514377, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2262 + }, + { + "epoch": 3.6150159744408947, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 2263 + }, + { + "epoch": 3.6166134185303513, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2264 + }, + { + "epoch": 3.6182108626198084, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2265 + }, + { + "epoch": 3.619808306709265, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2266 + }, + { + "epoch": 3.621405750798722, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 2267 + }, + { + "epoch": 3.623003194888179, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2268 + }, + { + "epoch": 3.6246006389776357, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2269 + }, + { + "epoch": 3.626198083067093, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2270 + }, + { + "epoch": 3.6277955271565494, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2271 + }, + { + "epoch": 3.6293929712460065, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2272 + }, + { + "epoch": 3.6309904153354635, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2273 + }, + { + "epoch": 3.63258785942492, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 2274 + }, + { + "epoch": 3.6341853035143767, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2275 + }, + { + "epoch": 3.635782747603834, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2276 + }, + { + "epoch": 3.637380191693291, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2277 + }, + { + "epoch": 3.6389776357827475, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2278 + }, + { + "epoch": 3.6405750798722045, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2279 + }, + { + "epoch": 3.642172523961661, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2280 + }, + { + "epoch": 3.643769968051118, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2281 + }, + { + "epoch": 3.6453674121405752, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2282 + }, + { + "epoch": 3.646964856230032, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2283 + }, + { + "epoch": 3.648562300319489, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2284 + }, + { + "epoch": 3.6501597444089455, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2285 + }, + { + "epoch": 3.6517571884984026, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 2286 + }, + { + "epoch": 3.6533546325878596, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2287 + }, + { + "epoch": 3.6549520766773163, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2288 + }, + { + "epoch": 3.6565495207667733, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2289 + }, + { + "epoch": 3.65814696485623, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2290 + }, + { + "epoch": 3.659744408945687, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2291 + }, + { + "epoch": 3.661341853035144, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2292 + }, + { + "epoch": 3.6629392971246006, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2293 + }, + { + "epoch": 3.6645367412140573, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2294 + }, + { + "epoch": 3.6661341853035143, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2295 + }, + { + "epoch": 3.6677316293929714, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2296 + }, + { + "epoch": 3.669329073482428, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2297 + }, + { + "epoch": 3.670926517571885, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2298 + }, + { + "epoch": 3.6725239616613417, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2299 + }, + { + "epoch": 3.6741214057507987, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2300 + }, + { + "epoch": 3.6757188498402558, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2301 + }, + { + "epoch": 3.6773162939297124, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 2302 + }, + { + "epoch": 3.6789137380191694, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2303 + }, + { + "epoch": 3.680511182108626, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2304 + }, + { + "epoch": 3.682108626198083, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2305 + }, + { + "epoch": 3.68370607028754, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2306 + }, + { + "epoch": 3.6853035143769968, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2307 + }, + { + "epoch": 3.686900958466454, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2308 + }, + { + "epoch": 3.6884984025559104, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2309 + }, + { + "epoch": 3.6900958466453675, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 2310 + }, + { + "epoch": 3.6916932907348246, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2311 + }, + { + "epoch": 3.693290734824281, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2312 + }, + { + "epoch": 3.6948881789137378, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2313 + }, + { + "epoch": 3.696485623003195, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 2314 + }, + { + "epoch": 3.698083067092652, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2315 + }, + { + "epoch": 3.6996805111821085, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2316 + }, + { + "epoch": 3.7012779552715656, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2317 + }, + { + "epoch": 3.702875399361022, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2318 + }, + { + "epoch": 3.7044728434504792, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2319 + }, + { + "epoch": 3.7060702875399363, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2320 + }, + { + "epoch": 3.707667731629393, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2321 + }, + { + "epoch": 3.70926517571885, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2322 + }, + { + "epoch": 3.7108626198083066, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2323 + }, + { + "epoch": 3.7124600638977636, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 2324 + }, + { + "epoch": 3.7140575079872207, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2325 + }, + { + "epoch": 3.7156549520766773, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 2326 + }, + { + "epoch": 3.7172523961661343, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2327 + }, + { + "epoch": 3.718849840255591, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2328 + }, + { + "epoch": 3.720447284345048, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 2329 + }, + { + "epoch": 3.722044728434505, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 2330 + }, + { + "epoch": 3.7236421725239617, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2331 + }, + { + "epoch": 3.7252396166134183, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 2332 + }, + { + "epoch": 3.7268370607028753, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2333 + }, + { + "epoch": 3.7284345047923324, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 2334 + }, + { + "epoch": 3.730031948881789, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2335 + }, + { + "epoch": 3.731629392971246, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2336 + }, + { + "epoch": 3.7332268370607027, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2337 + }, + { + "epoch": 3.7348242811501597, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2338 + }, + { + "epoch": 3.736421725239617, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2339 + }, + { + "epoch": 3.7380191693290734, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2340 + }, + { + "epoch": 3.7396166134185305, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2341 + }, + { + "epoch": 3.741214057507987, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2342 + }, + { + "epoch": 3.742811501597444, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 2343 + }, + { + "epoch": 3.744408945686901, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2344 + }, + { + "epoch": 3.746006389776358, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2345 + }, + { + "epoch": 3.747603833865815, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2346 + }, + { + "epoch": 3.7492012779552715, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2347 + }, + { + "epoch": 3.7507987220447285, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 2348 + }, + { + "epoch": 3.752396166134185, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2349 + }, + { + "epoch": 3.753993610223642, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2350 + }, + { + "epoch": 3.755591054313099, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2351 + }, + { + "epoch": 3.757188498402556, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2352 + }, + { + "epoch": 3.758785942492013, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2353 + }, + { + "epoch": 3.7603833865814695, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2354 + }, + { + "epoch": 3.7619808306709266, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 2355 + }, + { + "epoch": 3.763578274760383, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2356 + }, + { + "epoch": 3.7651757188498403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 2357 + }, + { + "epoch": 3.7667731629392973, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0904, + "step": 2358 + }, + { + "epoch": 3.768370607028754, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 2359 + }, + { + "epoch": 3.769968051118211, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2360 + }, + { + "epoch": 3.7715654952076676, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2361 + }, + { + "epoch": 3.7731629392971247, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2362 + }, + { + "epoch": 3.7747603833865817, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2363 + }, + { + "epoch": 3.7763578274760383, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2364 + }, + { + "epoch": 3.777955271565495, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2365 + }, + { + "epoch": 3.779552715654952, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2366 + }, + { + "epoch": 3.781150159744409, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2367 + }, + { + "epoch": 3.7827476038338657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2368 + }, + { + "epoch": 3.7843450479233227, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2369 + }, + { + "epoch": 3.7859424920127793, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 2370 + }, + { + "epoch": 3.7875399361022364, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2371 + }, + { + "epoch": 3.7891373801916934, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2372 + }, + { + "epoch": 3.79073482428115, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2373 + }, + { + "epoch": 3.792332268370607, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2374 + }, + { + "epoch": 3.7939297124600637, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2375 + }, + { + "epoch": 3.7955271565495208, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 2376 + }, + { + "epoch": 3.797124600638978, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2377 + }, + { + "epoch": 3.7987220447284344, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2378 + }, + { + "epoch": 3.8003194888178915, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2379 + }, + { + "epoch": 3.801916932907348, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2380 + }, + { + "epoch": 3.803514376996805, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2381 + }, + { + "epoch": 3.8051118210862622, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2382 + }, + { + "epoch": 3.806709265175719, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2383 + }, + { + "epoch": 3.8083067092651754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2384 + }, + { + "epoch": 3.8099041533546325, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2385 + }, + { + "epoch": 3.8115015974440896, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2386 + }, + { + "epoch": 3.813099041533546, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2387 + }, + { + "epoch": 3.8146964856230032, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2388 + }, + { + "epoch": 3.81629392971246, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2389 + }, + { + "epoch": 3.817891373801917, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2390 + }, + { + "epoch": 3.819488817891374, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2391 + }, + { + "epoch": 3.8210862619808306, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2392 + }, + { + "epoch": 3.8226837060702876, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2393 + }, + { + "epoch": 3.8242811501597442, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2394 + }, + { + "epoch": 3.8258785942492013, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2395 + }, + { + "epoch": 3.8274760383386583, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2396 + }, + { + "epoch": 3.829073482428115, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2397 + }, + { + "epoch": 3.830670926517572, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2398 + }, + { + "epoch": 3.8322683706070286, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2399 + }, + { + "epoch": 3.8338658146964857, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2400 + }, + { + "epoch": 3.8354632587859427, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2401 + }, + { + "epoch": 3.8370607028753994, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2402 + }, + { + "epoch": 3.838658146964856, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2403 + }, + { + "epoch": 3.840255591054313, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2404 + }, + { + "epoch": 3.84185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2405 + }, + { + "epoch": 3.8434504792332267, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2406 + }, + { + "epoch": 3.8450479233226837, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2407 + }, + { + "epoch": 3.8466453674121404, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2408 + }, + { + "epoch": 3.8482428115015974, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 2409 + }, + { + "epoch": 3.8498402555910545, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2410 + }, + { + "epoch": 3.851437699680511, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2411 + }, + { + "epoch": 3.853035143769968, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2412 + }, + { + "epoch": 3.8546325878594248, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2413 + }, + { + "epoch": 3.856230031948882, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2414 + }, + { + "epoch": 3.857827476038339, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2415 + }, + { + "epoch": 3.8594249201277955, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2416 + }, + { + "epoch": 3.8610223642172525, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2417 + }, + { + "epoch": 3.862619808306709, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2418 + }, + { + "epoch": 3.864217252396166, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2419 + }, + { + "epoch": 3.8658146964856233, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2420 + }, + { + "epoch": 3.86741214057508, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2421 + }, + { + "epoch": 3.8690095846645365, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2422 + }, + { + "epoch": 3.8706070287539935, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2423 + }, + { + "epoch": 3.8722044728434506, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2424 + }, + { + "epoch": 3.873801916932907, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2425 + }, + { + "epoch": 3.8753993610223643, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2426 + }, + { + "epoch": 3.876996805111821, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2427 + }, + { + "epoch": 3.878594249201278, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2428 + }, + { + "epoch": 3.880191693290735, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2429 + }, + { + "epoch": 3.8817891373801916, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2430 + }, + { + "epoch": 3.8833865814696487, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2431 + }, + { + "epoch": 3.8849840255591053, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2432 + }, + { + "epoch": 3.8865814696485623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2433 + }, + { + "epoch": 3.8881789137380194, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2434 + }, + { + "epoch": 3.889776357827476, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2435 + }, + { + "epoch": 3.891373801916933, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2436 + }, + { + "epoch": 3.8929712460063897, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2437 + }, + { + "epoch": 3.8945686900958467, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2438 + }, + { + "epoch": 3.8961661341853038, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2439 + }, + { + "epoch": 3.8977635782747604, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2440 + }, + { + "epoch": 3.899361022364217, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 2441 + }, + { + "epoch": 3.900958466453674, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2442 + }, + { + "epoch": 3.902555910543131, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2443 + }, + { + "epoch": 3.9041533546325877, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2444 + }, + { + "epoch": 3.905750798722045, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2445 + }, + { + "epoch": 3.9073482428115014, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2446 + }, + { + "epoch": 3.9089456869009584, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2447 + }, + { + "epoch": 3.9105431309904155, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 2448 + }, + { + "epoch": 3.912140575079872, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2449 + }, + { + "epoch": 3.913738019169329, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 2450 + }, + { + "epoch": 3.915335463258786, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 2451 + }, + { + "epoch": 3.916932907348243, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2452 + }, + { + "epoch": 3.9185303514377, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2453 + }, + { + "epoch": 3.9201277955271565, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2454 + }, + { + "epoch": 3.9217252396166136, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2455 + }, + { + "epoch": 3.92332268370607, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 2456 + }, + { + "epoch": 3.9249201277955272, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2457 + }, + { + "epoch": 3.9265175718849843, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2458 + }, + { + "epoch": 3.928115015974441, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2459 + }, + { + "epoch": 3.9297124600638975, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 2460 + }, + { + "epoch": 3.9313099041533546, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2461 + }, + { + "epoch": 3.9329073482428116, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 2462 + }, + { + "epoch": 3.9345047923322682, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2463 + }, + { + "epoch": 3.9361022364217253, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2464 + }, + { + "epoch": 3.937699680511182, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 2465 + }, + { + "epoch": 3.939297124600639, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2466 + }, + { + "epoch": 3.940894568690096, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2467 + }, + { + "epoch": 3.9424920127795526, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2468 + }, + { + "epoch": 3.9440894568690097, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2469 + }, + { + "epoch": 3.9456869009584663, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2470 + }, + { + "epoch": 3.9472843450479234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2471 + }, + { + "epoch": 3.9488817891373804, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2472 + }, + { + "epoch": 3.950479233226837, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2473 + }, + { + "epoch": 3.952076677316294, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2474 + }, + { + "epoch": 3.9536741214057507, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 2475 + }, + { + "epoch": 3.9552715654952078, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2476 + }, + { + "epoch": 3.956869009584665, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2477 + }, + { + "epoch": 3.9584664536741214, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 2478 + }, + { + "epoch": 3.960063897763578, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2479 + }, + { + "epoch": 3.961661341853035, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2480 + }, + { + "epoch": 3.963258785942492, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2481 + }, + { + "epoch": 3.9648562300319488, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2482 + }, + { + "epoch": 3.966453674121406, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 2483 + }, + { + "epoch": 3.9680511182108624, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2484 + }, + { + "epoch": 3.9696485623003195, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 2485 + }, + { + "epoch": 3.9712460063897765, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2486 + }, + { + "epoch": 3.972843450479233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2487 + }, + { + "epoch": 3.97444089456869, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2488 + }, + { + "epoch": 3.976038338658147, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2489 + }, + { + "epoch": 3.977635782747604, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2490 + }, + { + "epoch": 3.979233226837061, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 2491 + }, + { + "epoch": 3.9808306709265175, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 2492 + }, + { + "epoch": 3.9824281150159746, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2493 + }, + { + "epoch": 3.984025559105431, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2494 + }, + { + "epoch": 3.9856230031948883, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2495 + }, + { + "epoch": 3.987220447284345, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2496 + }, + { + "epoch": 3.988817891373802, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2497 + }, + { + "epoch": 3.9904153354632586, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2498 + }, + { + "epoch": 3.9920127795527156, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2499 + }, + { + "epoch": 3.9936102236421727, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2500 + }, + { + "epoch": 3.9952076677316293, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2501 + }, + { + "epoch": 3.9968051118210863, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2502 + }, + { + "epoch": 3.998402555910543, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2503 + }, + { + "epoch": 4.0, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2504 + }, + { + "epoch": 4.001597444089457, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2505 + }, + { + "epoch": 4.003194888178914, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2506 + }, + { + "epoch": 4.00479233226837, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2507 + }, + { + "epoch": 4.006389776357827, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2508 + }, + { + "epoch": 4.007987220447284, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2509 + }, + { + "epoch": 4.0095846645367414, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2510 + }, + { + "epoch": 4.0111821086261985, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2511 + }, + { + "epoch": 4.012779552715655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2512 + }, + { + "epoch": 4.014376996805112, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2513 + }, + { + "epoch": 4.015974440894569, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2514 + }, + { + "epoch": 4.017571884984026, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2515 + }, + { + "epoch": 4.019169329073482, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2516 + }, + { + "epoch": 4.020766773162939, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2517 + }, + { + "epoch": 4.022364217252396, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2518 + }, + { + "epoch": 4.023961661341853, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 2519 + }, + { + "epoch": 4.02555910543131, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2520 + }, + { + "epoch": 4.027156549520766, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2521 + }, + { + "epoch": 4.0287539936102235, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2522 + }, + { + "epoch": 4.0303514376996805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2523 + }, + { + "epoch": 4.031948881789138, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 2524 + }, + { + "epoch": 4.033546325878595, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2525 + }, + { + "epoch": 4.035143769968051, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 2526 + }, + { + "epoch": 4.036741214057508, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2527 + }, + { + "epoch": 4.038338658146965, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2528 + }, + { + "epoch": 4.039936102236422, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2529 + }, + { + "epoch": 4.041533546325879, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2530 + }, + { + "epoch": 4.043130990415335, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 2531 + }, + { + "epoch": 4.044728434504792, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 2532 + }, + { + "epoch": 4.046325878594249, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2533 + }, + { + "epoch": 4.047923322683706, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2534 + }, + { + "epoch": 4.0495207667731625, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2535 + }, + { + "epoch": 4.05111821086262, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2536 + }, + { + "epoch": 4.052715654952077, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2537 + }, + { + "epoch": 4.054313099041534, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2538 + }, + { + "epoch": 4.055910543130991, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2539 + }, + { + "epoch": 4.057507987220447, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2540 + }, + { + "epoch": 4.059105431309904, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2541 + }, + { + "epoch": 4.060702875399361, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2542 + }, + { + "epoch": 4.062300319488818, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 2543 + }, + { + "epoch": 4.063897763578275, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2544 + }, + { + "epoch": 4.065495207667731, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 2545 + }, + { + "epoch": 4.067092651757188, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2546 + }, + { + "epoch": 4.068690095846645, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2547 + }, + { + "epoch": 4.0702875399361025, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2548 + }, + { + "epoch": 4.0718849840255595, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 2549 + }, + { + "epoch": 4.073482428115016, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2550 + }, + { + "epoch": 4.075079872204473, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2551 + }, + { + "epoch": 4.07667731629393, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2552 + }, + { + "epoch": 4.078274760383387, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2553 + }, + { + "epoch": 4.079872204472843, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 2554 + }, + { + "epoch": 4.0814696485623, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2555 + }, + { + "epoch": 4.083067092651757, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2556 + }, + { + "epoch": 4.084664536741214, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2557 + }, + { + "epoch": 4.086261980830671, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2558 + }, + { + "epoch": 4.087859424920127, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2559 + }, + { + "epoch": 4.0894568690095845, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2560 + }, + { + "epoch": 4.0910543130990416, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2561 + }, + { + "epoch": 4.092651757188499, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2562 + }, + { + "epoch": 4.094249201277956, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2563 + }, + { + "epoch": 4.095846645367412, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2564 + }, + { + "epoch": 4.097444089456869, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 2565 + }, + { + "epoch": 4.099041533546326, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2566 + }, + { + "epoch": 4.100638977635783, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2567 + }, + { + "epoch": 4.102236421725239, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 2568 + }, + { + "epoch": 4.103833865814696, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2569 + }, + { + "epoch": 4.105431309904153, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 2570 + }, + { + "epoch": 4.10702875399361, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2571 + }, + { + "epoch": 4.108626198083067, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2572 + }, + { + "epoch": 4.110223642172524, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2573 + }, + { + "epoch": 4.111821086261981, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2574 + }, + { + "epoch": 4.113418530351438, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2575 + }, + { + "epoch": 4.115015974440895, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2576 + }, + { + "epoch": 4.116613418530352, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 2577 + }, + { + "epoch": 4.118210862619808, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2578 + }, + { + "epoch": 4.119808306709265, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2579 + }, + { + "epoch": 4.121405750798722, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2580 + }, + { + "epoch": 4.123003194888179, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2581 + }, + { + "epoch": 4.124600638977636, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2582 + }, + { + "epoch": 4.126198083067092, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2583 + }, + { + "epoch": 4.127795527156549, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 2584 + }, + { + "epoch": 4.1293929712460065, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2585 + }, + { + "epoch": 4.1309904153354635, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 2586 + }, + { + "epoch": 4.13258785942492, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2587 + }, + { + "epoch": 4.134185303514377, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2588 + }, + { + "epoch": 4.135782747603834, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 2589 + }, + { + "epoch": 4.137380191693291, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 2590 + }, + { + "epoch": 4.138977635782748, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 2591 + }, + { + "epoch": 4.140575079872204, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2592 + }, + { + "epoch": 4.142172523961661, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2593 + }, + { + "epoch": 4.143769968051118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2594 + }, + { + "epoch": 4.145367412140575, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2595 + }, + { + "epoch": 4.146964856230032, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2596 + }, + { + "epoch": 4.1485623003194885, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2597 + }, + { + "epoch": 4.1501597444089455, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 2598 + }, + { + "epoch": 4.151757188498403, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2599 + }, + { + "epoch": 4.15335463258786, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 2600 + }, + { + "epoch": 4.154952076677317, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2601 + }, + { + "epoch": 4.156549520766773, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2602 + }, + { + "epoch": 4.15814696485623, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2603 + }, + { + "epoch": 4.159744408945687, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 2604 + }, + { + "epoch": 4.161341853035144, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2605 + }, + { + "epoch": 4.1629392971246, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2606 + }, + { + "epoch": 4.164536741214057, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2607 + }, + { + "epoch": 4.166134185303514, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2608 + }, + { + "epoch": 4.167731629392971, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2609 + }, + { + "epoch": 4.169329073482428, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2610 + }, + { + "epoch": 4.170926517571885, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2611 + }, + { + "epoch": 4.172523961661342, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2612 + }, + { + "epoch": 4.174121405750799, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2613 + }, + { + "epoch": 4.175718849840256, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 2614 + }, + { + "epoch": 4.177316293929713, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2615 + }, + { + "epoch": 4.178913738019169, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2616 + }, + { + "epoch": 4.180511182108626, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 2617 + }, + { + "epoch": 4.182108626198083, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2618 + }, + { + "epoch": 4.18370607028754, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2619 + }, + { + "epoch": 4.185303514376997, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2620 + }, + { + "epoch": 4.186900958466453, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2621 + }, + { + "epoch": 4.18849840255591, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2622 + }, + { + "epoch": 4.1900958466453675, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2623 + }, + { + "epoch": 4.1916932907348246, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2624 + }, + { + "epoch": 4.193290734824281, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2625 + }, + { + "epoch": 4.194888178913738, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 2626 + }, + { + "epoch": 4.196485623003195, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2627 + }, + { + "epoch": 4.198083067092652, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 2628 + }, + { + "epoch": 4.199680511182109, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2629 + }, + { + "epoch": 4.201277955271565, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2630 + }, + { + "epoch": 4.202875399361022, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2631 + }, + { + "epoch": 4.204472843450479, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2632 + }, + { + "epoch": 4.206070287539936, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2633 + }, + { + "epoch": 4.207667731629393, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2634 + }, + { + "epoch": 4.2092651757188495, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2635 + }, + { + "epoch": 4.210862619808307, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2636 + }, + { + "epoch": 4.212460063897764, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2637 + }, + { + "epoch": 4.214057507987221, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 2638 + }, + { + "epoch": 4.215654952076678, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2639 + }, + { + "epoch": 4.217252396166134, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2640 + }, + { + "epoch": 4.218849840255591, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2641 + }, + { + "epoch": 4.220447284345048, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2642 + }, + { + "epoch": 4.222044728434505, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2643 + }, + { + "epoch": 4.223642172523961, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2644 + }, + { + "epoch": 4.225239616613418, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2645 + }, + { + "epoch": 4.226837060702875, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 2646 + }, + { + "epoch": 4.228434504792332, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2647 + }, + { + "epoch": 4.2300319488817895, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2648 + }, + { + "epoch": 4.231629392971246, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2649 + }, + { + "epoch": 4.233226837060703, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2650 + }, + { + "epoch": 4.23482428115016, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2651 + }, + { + "epoch": 4.236421725239617, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2652 + }, + { + "epoch": 4.238019169329074, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2653 + }, + { + "epoch": 4.23961661341853, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 2654 + }, + { + "epoch": 4.241214057507987, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 2655 + }, + { + "epoch": 4.242811501597444, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 2656 + }, + { + "epoch": 4.244408945686901, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2657 + }, + { + "epoch": 4.246006389776358, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2658 + }, + { + "epoch": 4.247603833865814, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2659 + }, + { + "epoch": 4.2492012779552715, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 2660 + }, + { + "epoch": 4.2507987220447285, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2661 + }, + { + "epoch": 4.252396166134186, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2662 + }, + { + "epoch": 4.253993610223642, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2663 + }, + { + "epoch": 4.255591054313099, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 2664 + }, + { + "epoch": 4.257188498402556, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2665 + }, + { + "epoch": 4.258785942492013, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2666 + }, + { + "epoch": 4.26038338658147, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2667 + }, + { + "epoch": 4.261980830670926, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2668 + }, + { + "epoch": 4.263578274760383, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2669 + }, + { + "epoch": 4.26517571884984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 2670 + }, + { + "epoch": 4.266773162939297, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2671 + }, + { + "epoch": 4.268370607028754, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 2672 + }, + { + "epoch": 4.2699680511182105, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 2673 + }, + { + "epoch": 4.271565495207668, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 2674 + }, + { + "epoch": 4.273162939297125, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 2675 + }, + { + "epoch": 4.274760383386582, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2676 + }, + { + "epoch": 4.276357827476039, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2677 + }, + { + "epoch": 4.277955271565495, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2678 + }, + { + "epoch": 4.279552715654952, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2679 + }, + { + "epoch": 4.281150159744409, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2680 + }, + { + "epoch": 4.282747603833866, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2681 + }, + { + "epoch": 4.284345047923322, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2682 + }, + { + "epoch": 4.285942492012779, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2683 + }, + { + "epoch": 4.287539936102236, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2684 + }, + { + "epoch": 4.289137380191693, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2685 + }, + { + "epoch": 4.2907348242811505, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2686 + }, + { + "epoch": 4.292332268370607, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2687 + }, + { + "epoch": 4.293929712460064, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2688 + }, + { + "epoch": 4.295527156549521, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2689 + }, + { + "epoch": 4.297124600638978, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2690 + }, + { + "epoch": 4.298722044728435, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2691 + }, + { + "epoch": 4.300319488817891, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2692 + }, + { + "epoch": 4.301916932907348, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2693 + }, + { + "epoch": 4.303514376996805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2694 + }, + { + "epoch": 4.305111821086262, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2695 + }, + { + "epoch": 4.306709265175719, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 2696 + }, + { + "epoch": 4.3083067092651754, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2697 + }, + { + "epoch": 4.3099041533546325, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2698 + }, + { + "epoch": 4.31150159744409, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2699 + }, + { + "epoch": 4.313099041533547, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 2700 + }, + { + "epoch": 4.314696485623003, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 2701 + }, + { + "epoch": 4.31629392971246, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 2702 + }, + { + "epoch": 4.317891373801917, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 2703 + }, + { + "epoch": 4.319488817891374, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2704 + }, + { + "epoch": 4.321086261980831, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 2705 + }, + { + "epoch": 4.322683706070287, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2706 + }, + { + "epoch": 4.324281150159744, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2707 + }, + { + "epoch": 4.325878594249201, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2708 + }, + { + "epoch": 4.327476038338658, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 2709 + }, + { + "epoch": 4.329073482428115, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 2710 + }, + { + "epoch": 4.330670926517572, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2711 + }, + { + "epoch": 4.332268370607029, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2712 + }, + { + "epoch": 4.333865814696486, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2713 + }, + { + "epoch": 4.335463258785943, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2714 + }, + { + "epoch": 4.3370607028754, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2715 + }, + { + "epoch": 4.338658146964856, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2716 + }, + { + "epoch": 4.340255591054313, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 2717 + }, + { + "epoch": 4.34185303514377, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 2718 + }, + { + "epoch": 4.343450479233227, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2719 + }, + { + "epoch": 4.345047923322683, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2720 + }, + { + "epoch": 4.34664536741214, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2721 + }, + { + "epoch": 4.348242811501597, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2722 + }, + { + "epoch": 4.3498402555910545, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2723 + }, + { + "epoch": 4.3514376996805115, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2724 + }, + { + "epoch": 4.353035143769968, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2725 + }, + { + "epoch": 4.354632587859425, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2726 + }, + { + "epoch": 4.356230031948882, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2727 + }, + { + "epoch": 4.357827476038339, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 2728 + }, + { + "epoch": 4.359424920127796, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2729 + }, + { + "epoch": 4.361022364217252, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 2730 + }, + { + "epoch": 4.362619808306709, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2731 + }, + { + "epoch": 4.364217252396166, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2732 + }, + { + "epoch": 4.365814696485623, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2733 + }, + { + "epoch": 4.36741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 2734 + }, + { + "epoch": 4.3690095846645365, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2735 + }, + { + "epoch": 4.3706070287539935, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 2736 + }, + { + "epoch": 4.372204472843451, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2737 + }, + { + "epoch": 4.373801916932908, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2738 + }, + { + "epoch": 4.375399361022364, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2739 + }, + { + "epoch": 4.376996805111821, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 2740 + }, + { + "epoch": 4.378594249201278, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2741 + }, + { + "epoch": 4.380191693290735, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2742 + }, + { + "epoch": 4.381789137380192, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2743 + }, + { + "epoch": 4.383386581469648, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 2744 + }, + { + "epoch": 4.384984025559105, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2745 + }, + { + "epoch": 4.386581469648562, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2746 + }, + { + "epoch": 4.388178913738019, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 2747 + }, + { + "epoch": 4.389776357827476, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 2748 + }, + { + "epoch": 4.391373801916933, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2749 + }, + { + "epoch": 4.39297124600639, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2750 + }, + { + "epoch": 4.394568690095847, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2751 + }, + { + "epoch": 4.396166134185304, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2752 + }, + { + "epoch": 4.397763578274761, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2753 + }, + { + "epoch": 4.399361022364217, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2754 + }, + { + "epoch": 4.400958466453674, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 2755 + }, + { + "epoch": 4.402555910543131, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 2756 + }, + { + "epoch": 4.404153354632588, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2757 + }, + { + "epoch": 4.405750798722044, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2758 + }, + { + "epoch": 4.407348242811501, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2759 + }, + { + "epoch": 4.4089456869009584, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2760 + }, + { + "epoch": 4.4105431309904155, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2761 + }, + { + "epoch": 4.412140575079873, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2762 + }, + { + "epoch": 4.413738019169329, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2763 + }, + { + "epoch": 4.415335463258786, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2764 + }, + { + "epoch": 4.416932907348243, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2765 + }, + { + "epoch": 4.4185303514377, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2766 + }, + { + "epoch": 4.420127795527157, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 2767 + }, + { + "epoch": 4.421725239616613, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 2768 + }, + { + "epoch": 4.42332268370607, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2769 + }, + { + "epoch": 4.424920127795527, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2770 + }, + { + "epoch": 4.426517571884984, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 2771 + }, + { + "epoch": 4.428115015974441, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2772 + }, + { + "epoch": 4.4297124600638975, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2773 + }, + { + "epoch": 4.431309904153355, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2774 + }, + { + "epoch": 4.432907348242812, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2775 + }, + { + "epoch": 4.434504792332269, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 2776 + }, + { + "epoch": 4.436102236421725, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 2777 + }, + { + "epoch": 4.437699680511182, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2778 + }, + { + "epoch": 4.439297124600639, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 2779 + }, + { + "epoch": 4.440894568690096, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2780 + }, + { + "epoch": 4.442492012779553, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2781 + }, + { + "epoch": 4.444089456869009, + "grad_norm": 0.458984375, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 2782 + }, + { + "epoch": 4.445686900958466, + "grad_norm": 0.95703125, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2783 + }, + { + "epoch": 4.447284345047923, + "grad_norm": 1.859375, + "learning_rate": 0.0005, + "loss": 1.135, + "step": 2784 + }, + { + "epoch": 4.44888178913738, + "grad_norm": 0.37109375, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 2785 + }, + { + "epoch": 4.4504792332268375, + "grad_norm": 1.6015625, + "learning_rate": 0.0005, + "loss": 1.1374, + "step": 2786 + }, + { + "epoch": 4.452076677316294, + "grad_norm": 0.298828125, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 2787 + }, + { + "epoch": 4.453674121405751, + "grad_norm": 0.828125, + "learning_rate": 0.0005, + "loss": 1.1162, + "step": 2788 + }, + { + "epoch": 4.455271565495208, + "grad_norm": 0.4140625, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 2789 + }, + { + "epoch": 4.456869009584665, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 2790 + }, + { + "epoch": 4.458466453674122, + "grad_norm": 0.35546875, + "learning_rate": 0.0005, + "loss": 1.0979, + "step": 2791 + }, + { + "epoch": 4.460063897763578, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 2792 + }, + { + "epoch": 4.461661341853035, + "grad_norm": 0.384765625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 2793 + }, + { + "epoch": 4.463258785942492, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 2794 + }, + { + "epoch": 4.464856230031949, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2795 + }, + { + "epoch": 4.466453674121405, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 2796 + }, + { + "epoch": 4.468051118210862, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 2797 + }, + { + "epoch": 4.4696485623003195, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2798 + }, + { + "epoch": 4.4712460063897765, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 2799 + }, + { + "epoch": 4.472843450479234, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2800 + }, + { + "epoch": 4.47444089456869, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 2801 + }, + { + "epoch": 4.476038338658147, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2802 + }, + { + "epoch": 4.477635782747604, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2803 + }, + { + "epoch": 4.479233226837061, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2804 + }, + { + "epoch": 4.480830670926517, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 2805 + }, + { + "epoch": 4.482428115015974, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 2806 + }, + { + "epoch": 4.484025559105431, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 2807 + }, + { + "epoch": 4.485623003194888, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2808 + }, + { + "epoch": 4.487220447284345, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2809 + }, + { + "epoch": 4.488817891373802, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2810 + }, + { + "epoch": 4.4904153354632586, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2811 + }, + { + "epoch": 4.492012779552716, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 2812 + }, + { + "epoch": 4.493610223642173, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 2813 + }, + { + "epoch": 4.49520766773163, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 2814 + }, + { + "epoch": 4.496805111821086, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2815 + }, + { + "epoch": 4.498402555910543, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2816 + }, + { + "epoch": 4.5, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2817 + }, + { + "epoch": 4.501597444089457, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 2818 + }, + { + "epoch": 4.503194888178914, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.087, + "step": 2819 + }, + { + "epoch": 4.50479233226837, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2820 + }, + { + "epoch": 4.506389776357827, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 2821 + }, + { + "epoch": 4.507987220447284, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 2822 + }, + { + "epoch": 4.5095846645367414, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2823 + }, + { + "epoch": 4.511182108626198, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2824 + }, + { + "epoch": 4.512779552715655, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2825 + }, + { + "epoch": 4.514376996805112, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2826 + }, + { + "epoch": 4.515974440894569, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 2827 + }, + { + "epoch": 4.517571884984026, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2828 + }, + { + "epoch": 4.519169329073483, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2829 + }, + { + "epoch": 4.520766773162939, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2830 + }, + { + "epoch": 4.522364217252396, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 2831 + }, + { + "epoch": 4.523961661341853, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2832 + }, + { + "epoch": 4.52555910543131, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 2833 + }, + { + "epoch": 4.527156549520766, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2834 + }, + { + "epoch": 4.5287539936102235, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2835 + }, + { + "epoch": 4.5303514376996805, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 2836 + }, + { + "epoch": 4.531948881789138, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 2837 + }, + { + "epoch": 4.533546325878595, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 2838 + }, + { + "epoch": 4.535143769968051, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2839 + }, + { + "epoch": 4.536741214057508, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2840 + }, + { + "epoch": 4.538338658146965, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 2841 + }, + { + "epoch": 4.539936102236422, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2842 + }, + { + "epoch": 4.541533546325878, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 2843 + }, + { + "epoch": 4.543130990415335, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2844 + }, + { + "epoch": 4.544728434504792, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2845 + }, + { + "epoch": 4.546325878594249, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 2846 + }, + { + "epoch": 4.547923322683706, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2847 + }, + { + "epoch": 4.549520766773163, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2848 + }, + { + "epoch": 4.55111821086262, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2849 + }, + { + "epoch": 4.552715654952077, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2850 + }, + { + "epoch": 4.554313099041534, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2851 + }, + { + "epoch": 4.555910543130991, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 2852 + }, + { + "epoch": 4.557507987220447, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 2853 + }, + { + "epoch": 4.559105431309904, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2854 + }, + { + "epoch": 4.560702875399361, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2855 + }, + { + "epoch": 4.562300319488818, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2856 + }, + { + "epoch": 4.563897763578275, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 2857 + }, + { + "epoch": 4.565495207667731, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 2858 + }, + { + "epoch": 4.567092651757188, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2859 + }, + { + "epoch": 4.568690095846645, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 2860 + }, + { + "epoch": 4.5702875399361025, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 2861 + }, + { + "epoch": 4.571884984025559, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2862 + }, + { + "epoch": 4.573482428115016, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2863 + }, + { + "epoch": 4.575079872204473, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 2864 + }, + { + "epoch": 4.57667731629393, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 2865 + }, + { + "epoch": 4.578274760383387, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 2866 + }, + { + "epoch": 4.579872204472844, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2867 + }, + { + "epoch": 4.5814696485623, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 2868 + }, + { + "epoch": 4.583067092651757, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 2869 + }, + { + "epoch": 4.584664536741214, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 2870 + }, + { + "epoch": 4.586261980830671, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 2871 + }, + { + "epoch": 4.587859424920127, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2872 + }, + { + "epoch": 4.5894568690095845, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 2873 + }, + { + "epoch": 4.5910543130990416, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 2874 + }, + { + "epoch": 4.592651757188499, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 2875 + }, + { + "epoch": 4.594249201277956, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2876 + }, + { + "epoch": 4.595846645367412, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 2877 + }, + { + "epoch": 4.597444089456869, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2878 + }, + { + "epoch": 4.599041533546326, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2879 + }, + { + "epoch": 4.600638977635783, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2880 + }, + { + "epoch": 4.602236421725239, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2881 + }, + { + "epoch": 4.603833865814696, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 2882 + }, + { + "epoch": 4.605431309904153, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2883 + }, + { + "epoch": 4.60702875399361, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2884 + }, + { + "epoch": 4.608626198083067, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 2885 + }, + { + "epoch": 4.6102236421725244, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 2886 + }, + { + "epoch": 4.611821086261981, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 2887 + }, + { + "epoch": 4.613418530351438, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2888 + }, + { + "epoch": 4.615015974440895, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 2889 + }, + { + "epoch": 4.616613418530352, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2890 + }, + { + "epoch": 4.618210862619808, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2891 + }, + { + "epoch": 4.619808306709265, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 2892 + }, + { + "epoch": 4.621405750798722, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2893 + }, + { + "epoch": 4.623003194888179, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 2894 + }, + { + "epoch": 4.624600638977636, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2895 + }, + { + "epoch": 4.626198083067092, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 2896 + }, + { + "epoch": 4.627795527156549, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 2897 + }, + { + "epoch": 4.6293929712460065, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2898 + }, + { + "epoch": 4.6309904153354635, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2899 + }, + { + "epoch": 4.63258785942492, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2900 + }, + { + "epoch": 4.634185303514377, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2901 + }, + { + "epoch": 4.635782747603834, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2902 + }, + { + "epoch": 4.637380191693291, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2903 + }, + { + "epoch": 4.638977635782748, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 2904 + }, + { + "epoch": 4.640575079872205, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2905 + }, + { + "epoch": 4.642172523961661, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 2906 + }, + { + "epoch": 4.643769968051118, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2907 + }, + { + "epoch": 4.645367412140575, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 2908 + }, + { + "epoch": 4.646964856230032, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2909 + }, + { + "epoch": 4.6485623003194885, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 2910 + }, + { + "epoch": 4.6501597444089455, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2911 + }, + { + "epoch": 4.651757188498403, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2912 + }, + { + "epoch": 4.65335463258786, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 2913 + }, + { + "epoch": 4.654952076677317, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2914 + }, + { + "epoch": 4.656549520766773, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2915 + }, + { + "epoch": 4.65814696485623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2916 + }, + { + "epoch": 4.659744408945687, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 2917 + }, + { + "epoch": 4.661341853035144, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2918 + }, + { + "epoch": 4.6629392971246, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2919 + }, + { + "epoch": 4.664536741214057, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 2920 + }, + { + "epoch": 4.666134185303514, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 2921 + }, + { + "epoch": 4.667731629392971, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2922 + }, + { + "epoch": 4.669329073482428, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 2923 + }, + { + "epoch": 4.6709265175718855, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 2924 + }, + { + "epoch": 4.672523961661342, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 2925 + }, + { + "epoch": 4.674121405750799, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 2926 + }, + { + "epoch": 4.675718849840256, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 2927 + }, + { + "epoch": 4.677316293929713, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 2928 + }, + { + "epoch": 4.678913738019169, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 2929 + }, + { + "epoch": 4.680511182108626, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 2930 + }, + { + "epoch": 4.682108626198083, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 2931 + }, + { + "epoch": 4.68370607028754, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 2932 + }, + { + "epoch": 4.685303514376997, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 2933 + }, + { + "epoch": 4.686900958466453, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 2934 + }, + { + "epoch": 4.68849840255591, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2935 + }, + { + "epoch": 4.6900958466453675, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 2936 + }, + { + "epoch": 4.6916932907348246, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 2937 + }, + { + "epoch": 4.693290734824281, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2938 + }, + { + "epoch": 4.694888178913738, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2939 + }, + { + "epoch": 4.696485623003195, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2940 + }, + { + "epoch": 4.698083067092652, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 2941 + }, + { + "epoch": 4.699680511182109, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 2942 + }, + { + "epoch": 4.701277955271565, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2943 + }, + { + "epoch": 4.702875399361022, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 2944 + }, + { + "epoch": 4.704472843450479, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 2945 + }, + { + "epoch": 4.706070287539936, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 2946 + }, + { + "epoch": 4.707667731629393, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 2947 + }, + { + "epoch": 4.7092651757188495, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2948 + }, + { + "epoch": 4.710862619808307, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 2949 + }, + { + "epoch": 4.712460063897764, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2950 + }, + { + "epoch": 4.714057507987221, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2951 + }, + { + "epoch": 4.715654952076678, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 2952 + }, + { + "epoch": 4.717252396166134, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 2953 + }, + { + "epoch": 4.718849840255591, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2954 + }, + { + "epoch": 4.720447284345048, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 2955 + }, + { + "epoch": 4.722044728434505, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 2956 + }, + { + "epoch": 4.723642172523961, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 2957 + }, + { + "epoch": 4.725239616613418, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 2958 + }, + { + "epoch": 4.726837060702875, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 2959 + }, + { + "epoch": 4.728434504792332, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2960 + }, + { + "epoch": 4.7300319488817895, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 2961 + }, + { + "epoch": 4.731629392971246, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 2962 + }, + { + "epoch": 4.733226837060703, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 2963 + }, + { + "epoch": 4.73482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2964 + }, + { + "epoch": 4.736421725239617, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 2965 + }, + { + "epoch": 4.738019169329074, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 2966 + }, + { + "epoch": 4.73961661341853, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2967 + }, + { + "epoch": 4.741214057507987, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 2968 + }, + { + "epoch": 4.742811501597444, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 2969 + }, + { + "epoch": 4.744408945686901, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 2970 + }, + { + "epoch": 4.746006389776358, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 2971 + }, + { + "epoch": 4.747603833865814, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 2972 + }, + { + "epoch": 4.7492012779552715, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2973 + }, + { + "epoch": 4.7507987220447285, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2974 + }, + { + "epoch": 4.752396166134186, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 2975 + }, + { + "epoch": 4.753993610223642, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 2976 + }, + { + "epoch": 4.755591054313099, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 2977 + }, + { + "epoch": 4.757188498402556, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 2978 + }, + { + "epoch": 4.758785942492013, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 2979 + }, + { + "epoch": 4.76038338658147, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 2980 + }, + { + "epoch": 4.761980830670926, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 2981 + }, + { + "epoch": 4.763578274760383, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 2982 + }, + { + "epoch": 4.76517571884984, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2983 + }, + { + "epoch": 4.766773162939297, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 2984 + }, + { + "epoch": 4.768370607028754, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 2985 + }, + { + "epoch": 4.7699680511182105, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 2986 + }, + { + "epoch": 4.771565495207668, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 2987 + }, + { + "epoch": 4.773162939297125, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 2988 + }, + { + "epoch": 4.774760383386582, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 2989 + }, + { + "epoch": 4.776357827476039, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 2990 + }, + { + "epoch": 4.777955271565495, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2991 + }, + { + "epoch": 4.779552715654952, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 2992 + }, + { + "epoch": 4.781150159744409, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 2993 + }, + { + "epoch": 4.782747603833866, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2994 + }, + { + "epoch": 4.784345047923322, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 2995 + }, + { + "epoch": 4.785942492012779, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 2996 + }, + { + "epoch": 4.787539936102236, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 2997 + }, + { + "epoch": 4.789137380191693, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 2998 + }, + { + "epoch": 4.7907348242811505, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 2999 + }, + { + "epoch": 4.792332268370607, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3000 + }, + { + "epoch": 4.793929712460064, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3001 + }, + { + "epoch": 4.795527156549521, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3002 + }, + { + "epoch": 4.797124600638978, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3003 + }, + { + "epoch": 4.798722044728435, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 3004 + }, + { + "epoch": 4.800319488817891, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3005 + }, + { + "epoch": 4.801916932907348, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3006 + }, + { + "epoch": 4.803514376996805, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3007 + }, + { + "epoch": 4.805111821086262, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 3008 + }, + { + "epoch": 4.806709265175719, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 3009 + }, + { + "epoch": 4.8083067092651754, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3010 + }, + { + "epoch": 4.8099041533546325, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3011 + }, + { + "epoch": 4.81150159744409, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3012 + }, + { + "epoch": 4.813099041533547, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3013 + }, + { + "epoch": 4.814696485623003, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3014 + }, + { + "epoch": 4.81629392971246, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 3015 + }, + { + "epoch": 4.817891373801917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3016 + }, + { + "epoch": 4.819488817891374, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 3017 + }, + { + "epoch": 4.821086261980831, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3018 + }, + { + "epoch": 4.822683706070287, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3019 + }, + { + "epoch": 4.824281150159744, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3020 + }, + { + "epoch": 4.825878594249201, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3021 + }, + { + "epoch": 4.827476038338658, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3022 + }, + { + "epoch": 4.8290734824281145, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3023 + }, + { + "epoch": 4.830670926517572, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3024 + }, + { + "epoch": 4.832268370607029, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 3025 + }, + { + "epoch": 4.833865814696486, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3026 + }, + { + "epoch": 4.835463258785943, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3027 + }, + { + "epoch": 4.8370607028754, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3028 + }, + { + "epoch": 4.838658146964856, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3029 + }, + { + "epoch": 4.840255591054313, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3030 + }, + { + "epoch": 4.84185303514377, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3031 + }, + { + "epoch": 4.843450479233227, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3032 + }, + { + "epoch": 4.845047923322683, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3033 + }, + { + "epoch": 4.84664536741214, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3034 + }, + { + "epoch": 4.848242811501597, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3035 + }, + { + "epoch": 4.8498402555910545, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 3036 + }, + { + "epoch": 4.8514376996805115, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3037 + }, + { + "epoch": 4.853035143769968, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 3038 + }, + { + "epoch": 4.854632587859425, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3039 + }, + { + "epoch": 4.856230031948882, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3040 + }, + { + "epoch": 4.857827476038339, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3041 + }, + { + "epoch": 4.859424920127795, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3042 + }, + { + "epoch": 4.861022364217252, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3043 + }, + { + "epoch": 4.862619808306709, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3044 + }, + { + "epoch": 4.864217252396166, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3045 + }, + { + "epoch": 4.865814696485623, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3046 + }, + { + "epoch": 4.86741214057508, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3047 + }, + { + "epoch": 4.8690095846645365, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3048 + }, + { + "epoch": 4.8706070287539935, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3049 + }, + { + "epoch": 4.872204472843451, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 3050 + }, + { + "epoch": 4.873801916932908, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 3051 + }, + { + "epoch": 4.875399361022364, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3052 + }, + { + "epoch": 4.876996805111821, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3053 + }, + { + "epoch": 4.878594249201278, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3054 + }, + { + "epoch": 4.880191693290735, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3055 + }, + { + "epoch": 4.881789137380192, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 3056 + }, + { + "epoch": 4.883386581469648, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3057 + }, + { + "epoch": 4.884984025559105, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3058 + }, + { + "epoch": 4.886581469648562, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3059 + }, + { + "epoch": 4.888178913738019, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3060 + }, + { + "epoch": 4.8897763578274756, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3061 + }, + { + "epoch": 4.891373801916933, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3062 + }, + { + "epoch": 4.89297124600639, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3063 + }, + { + "epoch": 4.894568690095847, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3064 + }, + { + "epoch": 4.896166134185304, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 3065 + }, + { + "epoch": 4.897763578274761, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3066 + }, + { + "epoch": 4.899361022364217, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 3067 + }, + { + "epoch": 4.900958466453674, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3068 + }, + { + "epoch": 4.902555910543131, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3069 + }, + { + "epoch": 4.904153354632588, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3070 + }, + { + "epoch": 4.905750798722044, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 3071 + }, + { + "epoch": 4.907348242811501, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3072 + }, + { + "epoch": 4.9089456869009584, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3073 + }, + { + "epoch": 4.9105431309904155, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3074 + }, + { + "epoch": 4.912140575079873, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3075 + }, + { + "epoch": 4.913738019169329, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 3076 + }, + { + "epoch": 4.915335463258786, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3077 + }, + { + "epoch": 4.916932907348243, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3078 + }, + { + "epoch": 4.9185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3079 + }, + { + "epoch": 4.920127795527156, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 3080 + }, + { + "epoch": 4.921725239616613, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3081 + }, + { + "epoch": 4.92332268370607, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 3082 + }, + { + "epoch": 4.924920127795527, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 3083 + }, + { + "epoch": 4.926517571884984, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3084 + }, + { + "epoch": 4.928115015974441, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3085 + }, + { + "epoch": 4.9297124600638975, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3086 + }, + { + "epoch": 4.931309904153355, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3087 + }, + { + "epoch": 4.932907348242812, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3088 + }, + { + "epoch": 4.934504792332269, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3089 + }, + { + "epoch": 4.936102236421725, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3090 + }, + { + "epoch": 4.937699680511182, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3091 + }, + { + "epoch": 4.939297124600639, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 3092 + }, + { + "epoch": 4.940894568690096, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 3093 + }, + { + "epoch": 4.942492012779553, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3094 + }, + { + "epoch": 4.944089456869009, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3095 + }, + { + "epoch": 4.945686900958466, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3096 + }, + { + "epoch": 4.947284345047923, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3097 + }, + { + "epoch": 4.94888178913738, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3098 + }, + { + "epoch": 4.950479233226837, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3099 + }, + { + "epoch": 4.952076677316294, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3100 + }, + { + "epoch": 4.953674121405751, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3101 + }, + { + "epoch": 4.955271565495208, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3102 + }, + { + "epoch": 4.956869009584665, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3103 + }, + { + "epoch": 4.958466453674122, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3104 + }, + { + "epoch": 4.960063897763578, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3105 + }, + { + "epoch": 4.961661341853035, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 3106 + }, + { + "epoch": 4.963258785942492, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3107 + }, + { + "epoch": 4.964856230031949, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3108 + }, + { + "epoch": 4.966453674121405, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 3109 + }, + { + "epoch": 4.968051118210862, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3110 + }, + { + "epoch": 4.9696485623003195, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 3111 + }, + { + "epoch": 4.9712460063897765, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3112 + }, + { + "epoch": 4.972843450479234, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3113 + }, + { + "epoch": 4.97444089456869, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3114 + }, + { + "epoch": 4.976038338658147, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 3115 + }, + { + "epoch": 4.977635782747604, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 3116 + }, + { + "epoch": 4.979233226837061, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3117 + }, + { + "epoch": 4.980830670926517, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3118 + }, + { + "epoch": 4.982428115015974, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3119 + }, + { + "epoch": 4.984025559105431, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3120 + }, + { + "epoch": 4.985623003194888, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3121 + }, + { + "epoch": 4.987220447284345, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3122 + }, + { + "epoch": 4.988817891373802, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3123 + }, + { + "epoch": 4.9904153354632586, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3124 + }, + { + "epoch": 4.992012779552716, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3125 + }, + { + "epoch": 4.993610223642173, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3126 + }, + { + "epoch": 4.99520766773163, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3127 + }, + { + "epoch": 4.996805111821086, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3128 + }, + { + "epoch": 4.998402555910543, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3129 + }, + { + "epoch": 5.0, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3130 + }, + { + "epoch": 5.001597444089457, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3131 + }, + { + "epoch": 5.003194888178914, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3132 + }, + { + "epoch": 5.00479233226837, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3133 + }, + { + "epoch": 5.006389776357827, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3134 + }, + { + "epoch": 5.007987220447284, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3135 + }, + { + "epoch": 5.0095846645367414, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3136 + }, + { + "epoch": 5.0111821086261985, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3137 + }, + { + "epoch": 5.012779552715655, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 3138 + }, + { + "epoch": 5.014376996805112, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3139 + }, + { + "epoch": 5.015974440894569, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3140 + }, + { + "epoch": 5.017571884984026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3141 + }, + { + "epoch": 5.019169329073482, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3142 + }, + { + "epoch": 5.020766773162939, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3143 + }, + { + "epoch": 5.022364217252396, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3144 + }, + { + "epoch": 5.023961661341853, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 3145 + }, + { + "epoch": 5.02555910543131, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3146 + }, + { + "epoch": 5.027156549520766, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3147 + }, + { + "epoch": 5.0287539936102235, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3148 + }, + { + "epoch": 5.0303514376996805, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3149 + }, + { + "epoch": 5.031948881789138, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3150 + }, + { + "epoch": 5.033546325878595, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3151 + }, + { + "epoch": 5.035143769968051, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3152 + }, + { + "epoch": 5.036741214057508, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3153 + }, + { + "epoch": 5.038338658146965, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3154 + }, + { + "epoch": 5.039936102236422, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3155 + }, + { + "epoch": 5.041533546325879, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3156 + }, + { + "epoch": 5.043130990415335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3157 + }, + { + "epoch": 5.044728434504792, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 3158 + }, + { + "epoch": 5.046325878594249, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0857, + "step": 3159 + }, + { + "epoch": 5.047923322683706, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3160 + }, + { + "epoch": 5.0495207667731625, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3161 + }, + { + "epoch": 5.05111821086262, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3162 + }, + { + "epoch": 5.052715654952077, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3163 + }, + { + "epoch": 5.054313099041534, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3164 + }, + { + "epoch": 5.055910543130991, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3165 + }, + { + "epoch": 5.057507987220447, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3166 + }, + { + "epoch": 5.059105431309904, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3167 + }, + { + "epoch": 5.060702875399361, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3168 + }, + { + "epoch": 5.062300319488818, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3169 + }, + { + "epoch": 5.063897763578275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3170 + }, + { + "epoch": 5.065495207667731, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3171 + }, + { + "epoch": 5.067092651757188, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3172 + }, + { + "epoch": 5.068690095846645, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 3173 + }, + { + "epoch": 5.0702875399361025, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3174 + }, + { + "epoch": 5.0718849840255595, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3175 + }, + { + "epoch": 5.073482428115016, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 3176 + }, + { + "epoch": 5.075079872204473, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 3177 + }, + { + "epoch": 5.07667731629393, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3178 + }, + { + "epoch": 5.078274760383387, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 3179 + }, + { + "epoch": 5.079872204472843, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3180 + }, + { + "epoch": 5.0814696485623, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3181 + }, + { + "epoch": 5.083067092651757, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3182 + }, + { + "epoch": 5.084664536741214, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3183 + }, + { + "epoch": 5.086261980830671, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 3184 + }, + { + "epoch": 5.087859424920127, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 3185 + }, + { + "epoch": 5.0894568690095845, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3186 + }, + { + "epoch": 5.0910543130990416, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3187 + }, + { + "epoch": 5.092651757188499, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3188 + }, + { + "epoch": 5.094249201277956, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0879, + "step": 3189 + }, + { + "epoch": 5.095846645367412, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3190 + }, + { + "epoch": 5.097444089456869, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3191 + }, + { + "epoch": 5.099041533546326, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3192 + }, + { + "epoch": 5.100638977635783, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3193 + }, + { + "epoch": 5.102236421725239, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3194 + }, + { + "epoch": 5.103833865814696, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3195 + }, + { + "epoch": 5.105431309904153, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3196 + }, + { + "epoch": 5.10702875399361, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3197 + }, + { + "epoch": 5.108626198083067, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3198 + }, + { + "epoch": 5.110223642172524, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 3199 + }, + { + "epoch": 5.111821086261981, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3200 + }, + { + "epoch": 5.113418530351438, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 3201 + }, + { + "epoch": 5.115015974440895, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 3202 + }, + { + "epoch": 5.116613418530352, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3203 + }, + { + "epoch": 5.118210862619808, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3204 + }, + { + "epoch": 5.119808306709265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 3205 + }, + { + "epoch": 5.121405750798722, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3206 + }, + { + "epoch": 5.123003194888179, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3207 + }, + { + "epoch": 5.124600638977636, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3208 + }, + { + "epoch": 5.126198083067092, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3209 + }, + { + "epoch": 5.127795527156549, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3210 + }, + { + "epoch": 5.1293929712460065, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3211 + }, + { + "epoch": 5.1309904153354635, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3212 + }, + { + "epoch": 5.13258785942492, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3213 + }, + { + "epoch": 5.134185303514377, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3214 + }, + { + "epoch": 5.135782747603834, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3215 + }, + { + "epoch": 5.137380191693291, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3216 + }, + { + "epoch": 5.138977635782748, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3217 + }, + { + "epoch": 5.140575079872204, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0823, + "step": 3218 + }, + { + "epoch": 5.142172523961661, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3219 + }, + { + "epoch": 5.143769968051118, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3220 + }, + { + "epoch": 5.145367412140575, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3221 + }, + { + "epoch": 5.146964856230032, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3222 + }, + { + "epoch": 5.1485623003194885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3223 + }, + { + "epoch": 5.1501597444089455, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3224 + }, + { + "epoch": 5.151757188498403, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3225 + }, + { + "epoch": 5.15335463258786, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3226 + }, + { + "epoch": 5.154952076677317, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3227 + }, + { + "epoch": 5.156549520766773, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3228 + }, + { + "epoch": 5.15814696485623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3229 + }, + { + "epoch": 5.159744408945687, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3230 + }, + { + "epoch": 5.161341853035144, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3231 + }, + { + "epoch": 5.1629392971246, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3232 + }, + { + "epoch": 5.164536741214057, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 3233 + }, + { + "epoch": 5.166134185303514, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3234 + }, + { + "epoch": 5.167731629392971, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3235 + }, + { + "epoch": 5.169329073482428, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3236 + }, + { + "epoch": 5.170926517571885, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 3237 + }, + { + "epoch": 5.172523961661342, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3238 + }, + { + "epoch": 5.174121405750799, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3239 + }, + { + "epoch": 5.175718849840256, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3240 + }, + { + "epoch": 5.177316293929713, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3241 + }, + { + "epoch": 5.178913738019169, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 3242 + }, + { + "epoch": 5.180511182108626, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3243 + }, + { + "epoch": 5.182108626198083, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3244 + }, + { + "epoch": 5.18370607028754, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3245 + }, + { + "epoch": 5.185303514376997, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 3246 + }, + { + "epoch": 5.186900958466453, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3247 + }, + { + "epoch": 5.18849840255591, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3248 + }, + { + "epoch": 5.1900958466453675, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3249 + }, + { + "epoch": 5.1916932907348246, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3250 + }, + { + "epoch": 5.193290734824281, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3251 + }, + { + "epoch": 5.194888178913738, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3252 + }, + { + "epoch": 5.196485623003195, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 3253 + }, + { + "epoch": 5.198083067092652, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3254 + }, + { + "epoch": 5.199680511182109, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3255 + }, + { + "epoch": 5.201277955271565, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3256 + }, + { + "epoch": 5.202875399361022, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3257 + }, + { + "epoch": 5.204472843450479, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3258 + }, + { + "epoch": 5.206070287539936, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3259 + }, + { + "epoch": 5.207667731629393, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3260 + }, + { + "epoch": 5.2092651757188495, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3261 + }, + { + "epoch": 5.210862619808307, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3262 + }, + { + "epoch": 5.212460063897764, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3263 + }, + { + "epoch": 5.214057507987221, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3264 + }, + { + "epoch": 5.215654952076678, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 3265 + }, + { + "epoch": 5.217252396166134, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3266 + }, + { + "epoch": 5.218849840255591, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3267 + }, + { + "epoch": 5.220447284345048, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3268 + }, + { + "epoch": 5.222044728434505, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3269 + }, + { + "epoch": 5.223642172523961, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 3270 + }, + { + "epoch": 5.225239616613418, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3271 + }, + { + "epoch": 5.226837060702875, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3272 + }, + { + "epoch": 5.228434504792332, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3273 + }, + { + "epoch": 5.2300319488817895, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3274 + }, + { + "epoch": 5.231629392971246, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3275 + }, + { + "epoch": 5.233226837060703, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3276 + }, + { + "epoch": 5.23482428115016, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3277 + }, + { + "epoch": 5.236421725239617, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3278 + }, + { + "epoch": 5.238019169329074, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3279 + }, + { + "epoch": 5.23961661341853, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3280 + }, + { + "epoch": 5.241214057507987, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3281 + }, + { + "epoch": 5.242811501597444, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 3282 + }, + { + "epoch": 5.244408945686901, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3283 + }, + { + "epoch": 5.246006389776358, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3284 + }, + { + "epoch": 5.247603833865814, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3285 + }, + { + "epoch": 5.2492012779552715, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3286 + }, + { + "epoch": 5.2507987220447285, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3287 + }, + { + "epoch": 5.252396166134186, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3288 + }, + { + "epoch": 5.253993610223642, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3289 + }, + { + "epoch": 5.255591054313099, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3290 + }, + { + "epoch": 5.257188498402556, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3291 + }, + { + "epoch": 5.258785942492013, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3292 + }, + { + "epoch": 5.26038338658147, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3293 + }, + { + "epoch": 5.261980830670926, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3294 + }, + { + "epoch": 5.263578274760383, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3295 + }, + { + "epoch": 5.26517571884984, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3296 + }, + { + "epoch": 5.266773162939297, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3297 + }, + { + "epoch": 5.268370607028754, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3298 + }, + { + "epoch": 5.2699680511182105, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3299 + }, + { + "epoch": 5.271565495207668, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3300 + }, + { + "epoch": 5.273162939297125, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 3301 + }, + { + "epoch": 5.274760383386582, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 3302 + }, + { + "epoch": 5.276357827476039, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3303 + }, + { + "epoch": 5.277955271565495, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3304 + }, + { + "epoch": 5.279552715654952, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 3305 + }, + { + "epoch": 5.281150159744409, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 3306 + }, + { + "epoch": 5.282747603833866, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3307 + }, + { + "epoch": 5.284345047923322, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3308 + }, + { + "epoch": 5.285942492012779, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3309 + }, + { + "epoch": 5.287539936102236, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3310 + }, + { + "epoch": 5.289137380191693, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3311 + }, + { + "epoch": 5.2907348242811505, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3312 + }, + { + "epoch": 5.292332268370607, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 3313 + }, + { + "epoch": 5.293929712460064, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3314 + }, + { + "epoch": 5.295527156549521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3315 + }, + { + "epoch": 5.297124600638978, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3316 + }, + { + "epoch": 5.298722044728435, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3317 + }, + { + "epoch": 5.300319488817891, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3318 + }, + { + "epoch": 5.301916932907348, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3319 + }, + { + "epoch": 5.303514376996805, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3320 + }, + { + "epoch": 5.305111821086262, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3321 + }, + { + "epoch": 5.306709265175719, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3322 + }, + { + "epoch": 5.3083067092651754, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3323 + }, + { + "epoch": 5.3099041533546325, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3324 + }, + { + "epoch": 5.31150159744409, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3325 + }, + { + "epoch": 5.313099041533547, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3326 + }, + { + "epoch": 5.314696485623003, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3327 + }, + { + "epoch": 5.31629392971246, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3328 + }, + { + "epoch": 5.317891373801917, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3329 + }, + { + "epoch": 5.319488817891374, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3330 + }, + { + "epoch": 5.321086261980831, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 3331 + }, + { + "epoch": 5.322683706070287, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3332 + }, + { + "epoch": 5.324281150159744, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3333 + }, + { + "epoch": 5.325878594249201, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3334 + }, + { + "epoch": 5.327476038338658, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3335 + }, + { + "epoch": 5.329073482428115, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 3336 + }, + { + "epoch": 5.330670926517572, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3337 + }, + { + "epoch": 5.332268370607029, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3338 + }, + { + "epoch": 5.333865814696486, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3339 + }, + { + "epoch": 5.335463258785943, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3340 + }, + { + "epoch": 5.3370607028754, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3341 + }, + { + "epoch": 5.338658146964856, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3342 + }, + { + "epoch": 5.340255591054313, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3343 + }, + { + "epoch": 5.34185303514377, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3344 + }, + { + "epoch": 5.343450479233227, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3345 + }, + { + "epoch": 5.345047923322683, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3346 + }, + { + "epoch": 5.34664536741214, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3347 + }, + { + "epoch": 5.348242811501597, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3348 + }, + { + "epoch": 5.3498402555910545, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3349 + }, + { + "epoch": 5.3514376996805115, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3350 + }, + { + "epoch": 5.353035143769968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3351 + }, + { + "epoch": 5.354632587859425, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3352 + }, + { + "epoch": 5.356230031948882, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 3353 + }, + { + "epoch": 5.357827476038339, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3354 + }, + { + "epoch": 5.359424920127796, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3355 + }, + { + "epoch": 5.361022364217252, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3356 + }, + { + "epoch": 5.362619808306709, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3357 + }, + { + "epoch": 5.364217252396166, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3358 + }, + { + "epoch": 5.365814696485623, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3359 + }, + { + "epoch": 5.36741214057508, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3360 + }, + { + "epoch": 5.3690095846645365, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3361 + }, + { + "epoch": 5.3706070287539935, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 3362 + }, + { + "epoch": 5.372204472843451, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3363 + }, + { + "epoch": 5.373801916932908, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3364 + }, + { + "epoch": 5.375399361022364, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3365 + }, + { + "epoch": 5.376996805111821, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3366 + }, + { + "epoch": 5.378594249201278, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3367 + }, + { + "epoch": 5.380191693290735, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3368 + }, + { + "epoch": 5.381789137380192, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3369 + }, + { + "epoch": 5.383386581469648, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3370 + }, + { + "epoch": 5.384984025559105, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3371 + }, + { + "epoch": 5.386581469648562, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3372 + }, + { + "epoch": 5.388178913738019, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3373 + }, + { + "epoch": 5.389776357827476, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3374 + }, + { + "epoch": 5.391373801916933, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 3375 + }, + { + "epoch": 5.39297124600639, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 3376 + }, + { + "epoch": 5.394568690095847, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3377 + }, + { + "epoch": 5.396166134185304, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3378 + }, + { + "epoch": 5.397763578274761, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3379 + }, + { + "epoch": 5.399361022364217, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3380 + }, + { + "epoch": 5.400958466453674, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3381 + }, + { + "epoch": 5.402555910543131, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3382 + }, + { + "epoch": 5.404153354632588, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3383 + }, + { + "epoch": 5.405750798722044, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3384 + }, + { + "epoch": 5.407348242811501, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3385 + }, + { + "epoch": 5.4089456869009584, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3386 + }, + { + "epoch": 5.4105431309904155, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3387 + }, + { + "epoch": 5.412140575079873, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3388 + }, + { + "epoch": 5.413738019169329, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3389 + }, + { + "epoch": 5.415335463258786, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 3390 + }, + { + "epoch": 5.416932907348243, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3391 + }, + { + "epoch": 5.4185303514377, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3392 + }, + { + "epoch": 5.420127795527157, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 3393 + }, + { + "epoch": 5.421725239616613, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3394 + }, + { + "epoch": 5.42332268370607, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3395 + }, + { + "epoch": 5.424920127795527, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3396 + }, + { + "epoch": 5.426517571884984, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3397 + }, + { + "epoch": 5.428115015974441, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3398 + }, + { + "epoch": 5.4297124600638975, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3399 + }, + { + "epoch": 5.431309904153355, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3400 + }, + { + "epoch": 5.432907348242812, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3401 + }, + { + "epoch": 5.434504792332269, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3402 + }, + { + "epoch": 5.436102236421725, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3403 + }, + { + "epoch": 5.437699680511182, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3404 + }, + { + "epoch": 5.439297124600639, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3405 + }, + { + "epoch": 5.440894568690096, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3406 + }, + { + "epoch": 5.442492012779553, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3407 + }, + { + "epoch": 5.444089456869009, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3408 + }, + { + "epoch": 5.445686900958466, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3409 + }, + { + "epoch": 5.447284345047923, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3410 + }, + { + "epoch": 5.44888178913738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3411 + }, + { + "epoch": 5.4504792332268375, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3412 + }, + { + "epoch": 5.452076677316294, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3413 + }, + { + "epoch": 5.453674121405751, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3414 + }, + { + "epoch": 5.455271565495208, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 3415 + }, + { + "epoch": 5.456869009584665, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3416 + }, + { + "epoch": 5.458466453674122, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3417 + }, + { + "epoch": 5.460063897763578, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3418 + }, + { + "epoch": 5.461661341853035, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 3419 + }, + { + "epoch": 5.463258785942492, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3420 + }, + { + "epoch": 5.464856230031949, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3421 + }, + { + "epoch": 5.466453674121405, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3422 + }, + { + "epoch": 5.468051118210862, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3423 + }, + { + "epoch": 5.4696485623003195, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3424 + }, + { + "epoch": 5.4712460063897765, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3425 + }, + { + "epoch": 5.472843450479234, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 3426 + }, + { + "epoch": 5.47444089456869, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0812, + "step": 3427 + }, + { + "epoch": 5.476038338658147, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3428 + }, + { + "epoch": 5.477635782747604, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3429 + }, + { + "epoch": 5.479233226837061, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 3430 + }, + { + "epoch": 5.480830670926517, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3431 + }, + { + "epoch": 5.482428115015974, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3432 + }, + { + "epoch": 5.484025559105431, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3433 + }, + { + "epoch": 5.485623003194888, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3434 + }, + { + "epoch": 5.487220447284345, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 3435 + }, + { + "epoch": 5.488817891373802, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3436 + }, + { + "epoch": 5.4904153354632586, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3437 + }, + { + "epoch": 5.492012779552716, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3438 + }, + { + "epoch": 5.493610223642173, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3439 + }, + { + "epoch": 5.49520766773163, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3440 + }, + { + "epoch": 5.496805111821086, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3441 + }, + { + "epoch": 5.498402555910543, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3442 + }, + { + "epoch": 5.5, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3443 + }, + { + "epoch": 5.501597444089457, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3444 + }, + { + "epoch": 5.503194888178914, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3445 + }, + { + "epoch": 5.50479233226837, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3446 + }, + { + "epoch": 5.506389776357827, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3447 + }, + { + "epoch": 5.507987220447284, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3448 + }, + { + "epoch": 5.5095846645367414, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3449 + }, + { + "epoch": 5.511182108626198, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 3450 + }, + { + "epoch": 5.512779552715655, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3451 + }, + { + "epoch": 5.514376996805112, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3452 + }, + { + "epoch": 5.515974440894569, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3453 + }, + { + "epoch": 5.517571884984026, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3454 + }, + { + "epoch": 5.519169329073483, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3455 + }, + { + "epoch": 5.520766773162939, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3456 + }, + { + "epoch": 5.522364217252396, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3457 + }, + { + "epoch": 5.523961661341853, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3458 + }, + { + "epoch": 5.52555910543131, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 3459 + }, + { + "epoch": 5.527156549520766, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3460 + }, + { + "epoch": 5.5287539936102235, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3461 + }, + { + "epoch": 5.5303514376996805, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3462 + }, + { + "epoch": 5.531948881789138, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 3463 + }, + { + "epoch": 5.533546325878595, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3464 + }, + { + "epoch": 5.535143769968051, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3465 + }, + { + "epoch": 5.536741214057508, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 3466 + }, + { + "epoch": 5.538338658146965, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 3467 + }, + { + "epoch": 5.539936102236422, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3468 + }, + { + "epoch": 5.541533546325878, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3469 + }, + { + "epoch": 5.543130990415335, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 3470 + }, + { + "epoch": 5.544728434504792, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 3471 + }, + { + "epoch": 5.546325878594249, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3472 + }, + { + "epoch": 5.547923322683706, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3473 + }, + { + "epoch": 5.549520766773163, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 3474 + }, + { + "epoch": 5.55111821086262, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3475 + }, + { + "epoch": 5.552715654952077, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 3476 + }, + { + "epoch": 5.554313099041534, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3477 + }, + { + "epoch": 5.555910543130991, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3478 + }, + { + "epoch": 5.557507987220447, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3479 + }, + { + "epoch": 5.559105431309904, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3480 + }, + { + "epoch": 5.560702875399361, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3481 + }, + { + "epoch": 5.562300319488818, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3482 + }, + { + "epoch": 5.563897763578275, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3483 + }, + { + "epoch": 5.565495207667731, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3484 + }, + { + "epoch": 5.567092651757188, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3485 + }, + { + "epoch": 5.568690095846645, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3486 + }, + { + "epoch": 5.5702875399361025, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3487 + }, + { + "epoch": 5.571884984025559, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3488 + }, + { + "epoch": 5.573482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3489 + }, + { + "epoch": 5.575079872204473, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 3490 + }, + { + "epoch": 5.57667731629393, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3491 + }, + { + "epoch": 5.578274760383387, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 3492 + }, + { + "epoch": 5.579872204472844, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3493 + }, + { + "epoch": 5.5814696485623, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 3494 + }, + { + "epoch": 5.583067092651757, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 3495 + }, + { + "epoch": 5.584664536741214, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3496 + }, + { + "epoch": 5.586261980830671, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3497 + }, + { + "epoch": 5.587859424920127, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3498 + }, + { + "epoch": 5.5894568690095845, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3499 + }, + { + "epoch": 5.5910543130990416, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3500 + }, + { + "epoch": 5.592651757188499, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3501 + }, + { + "epoch": 5.594249201277956, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3502 + }, + { + "epoch": 5.595846645367412, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3503 + }, + { + "epoch": 5.597444089456869, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3504 + }, + { + "epoch": 5.599041533546326, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3505 + }, + { + "epoch": 5.600638977635783, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 3506 + }, + { + "epoch": 5.602236421725239, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3507 + }, + { + "epoch": 5.603833865814696, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3508 + }, + { + "epoch": 5.605431309904153, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 3509 + }, + { + "epoch": 5.60702875399361, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3510 + }, + { + "epoch": 5.608626198083067, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3511 + }, + { + "epoch": 5.6102236421725244, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 3512 + }, + { + "epoch": 5.611821086261981, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3513 + }, + { + "epoch": 5.613418530351438, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3514 + }, + { + "epoch": 5.615015974440895, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3515 + }, + { + "epoch": 5.616613418530352, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3516 + }, + { + "epoch": 5.618210862619808, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3517 + }, + { + "epoch": 5.619808306709265, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 3518 + }, + { + "epoch": 5.621405750798722, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3519 + }, + { + "epoch": 5.623003194888179, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3520 + }, + { + "epoch": 5.624600638977636, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3521 + }, + { + "epoch": 5.626198083067092, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3522 + }, + { + "epoch": 5.627795527156549, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3523 + }, + { + "epoch": 5.6293929712460065, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3524 + }, + { + "epoch": 5.6309904153354635, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3525 + }, + { + "epoch": 5.63258785942492, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3526 + }, + { + "epoch": 5.634185303514377, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 3527 + }, + { + "epoch": 5.635782747603834, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3528 + }, + { + "epoch": 5.637380191693291, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3529 + }, + { + "epoch": 5.638977635782748, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3530 + }, + { + "epoch": 5.640575079872205, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 3531 + }, + { + "epoch": 5.642172523961661, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3532 + }, + { + "epoch": 5.643769968051118, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 3533 + }, + { + "epoch": 5.645367412140575, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3534 + }, + { + "epoch": 5.646964856230032, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 3535 + }, + { + "epoch": 5.6485623003194885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3536 + }, + { + "epoch": 5.6501597444089455, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3537 + }, + { + "epoch": 5.651757188498403, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3538 + }, + { + "epoch": 5.65335463258786, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3539 + }, + { + "epoch": 5.654952076677317, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3540 + }, + { + "epoch": 5.656549520766773, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3541 + }, + { + "epoch": 5.65814696485623, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3542 + }, + { + "epoch": 5.659744408945687, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3543 + }, + { + "epoch": 5.661341853035144, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 3544 + }, + { + "epoch": 5.6629392971246, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3545 + }, + { + "epoch": 5.664536741214057, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 3546 + }, + { + "epoch": 5.666134185303514, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3547 + }, + { + "epoch": 5.667731629392971, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3548 + }, + { + "epoch": 5.669329073482428, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3549 + }, + { + "epoch": 5.6709265175718855, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3550 + }, + { + "epoch": 5.672523961661342, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3551 + }, + { + "epoch": 5.674121405750799, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3552 + }, + { + "epoch": 5.675718849840256, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3553 + }, + { + "epoch": 5.677316293929713, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3554 + }, + { + "epoch": 5.678913738019169, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3555 + }, + { + "epoch": 5.680511182108626, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 3556 + }, + { + "epoch": 5.682108626198083, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3557 + }, + { + "epoch": 5.68370607028754, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3558 + }, + { + "epoch": 5.685303514376997, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3559 + }, + { + "epoch": 5.686900958466453, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3560 + }, + { + "epoch": 5.68849840255591, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3561 + }, + { + "epoch": 5.6900958466453675, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 3562 + }, + { + "epoch": 5.6916932907348246, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 3563 + }, + { + "epoch": 5.693290734824281, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3564 + }, + { + "epoch": 5.694888178913738, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3565 + }, + { + "epoch": 5.696485623003195, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3566 + }, + { + "epoch": 5.698083067092652, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3567 + }, + { + "epoch": 5.699680511182109, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3568 + }, + { + "epoch": 5.701277955271565, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3569 + }, + { + "epoch": 5.702875399361022, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3570 + }, + { + "epoch": 5.704472843450479, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3571 + }, + { + "epoch": 5.706070287539936, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3572 + }, + { + "epoch": 5.707667731629393, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3573 + }, + { + "epoch": 5.7092651757188495, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3574 + }, + { + "epoch": 5.710862619808307, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 3575 + }, + { + "epoch": 5.712460063897764, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3576 + }, + { + "epoch": 5.714057507987221, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3577 + }, + { + "epoch": 5.715654952076678, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3578 + }, + { + "epoch": 5.717252396166134, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3579 + }, + { + "epoch": 5.718849840255591, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3580 + }, + { + "epoch": 5.720447284345048, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3581 + }, + { + "epoch": 5.722044728434505, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3582 + }, + { + "epoch": 5.723642172523961, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3583 + }, + { + "epoch": 5.725239616613418, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 3584 + }, + { + "epoch": 5.726837060702875, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3585 + }, + { + "epoch": 5.728434504792332, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3586 + }, + { + "epoch": 5.7300319488817895, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3587 + }, + { + "epoch": 5.731629392971246, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3588 + }, + { + "epoch": 5.733226837060703, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3589 + }, + { + "epoch": 5.73482428115016, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3590 + }, + { + "epoch": 5.736421725239617, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3591 + }, + { + "epoch": 5.738019169329074, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3592 + }, + { + "epoch": 5.73961661341853, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3593 + }, + { + "epoch": 5.741214057507987, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 3594 + }, + { + "epoch": 5.742811501597444, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 3595 + }, + { + "epoch": 5.744408945686901, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3596 + }, + { + "epoch": 5.746006389776358, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3597 + }, + { + "epoch": 5.747603833865814, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3598 + }, + { + "epoch": 5.7492012779552715, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3599 + }, + { + "epoch": 5.7507987220447285, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3600 + }, + { + "epoch": 5.752396166134186, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 3601 + }, + { + "epoch": 5.753993610223642, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3602 + }, + { + "epoch": 5.755591054313099, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3603 + }, + { + "epoch": 5.757188498402556, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3604 + }, + { + "epoch": 5.758785942492013, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3605 + }, + { + "epoch": 5.76038338658147, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3606 + }, + { + "epoch": 5.761980830670926, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3607 + }, + { + "epoch": 5.763578274760383, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3608 + }, + { + "epoch": 5.76517571884984, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3609 + }, + { + "epoch": 5.766773162939297, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3610 + }, + { + "epoch": 5.768370607028754, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3611 + }, + { + "epoch": 5.7699680511182105, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3612 + }, + { + "epoch": 5.771565495207668, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3613 + }, + { + "epoch": 5.773162939297125, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3614 + }, + { + "epoch": 5.774760383386582, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3615 + }, + { + "epoch": 5.776357827476039, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3616 + }, + { + "epoch": 5.777955271565495, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3617 + }, + { + "epoch": 5.779552715654952, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3618 + }, + { + "epoch": 5.781150159744409, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3619 + }, + { + "epoch": 5.782747603833866, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3620 + }, + { + "epoch": 5.784345047923322, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3621 + }, + { + "epoch": 5.785942492012779, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3622 + }, + { + "epoch": 5.787539936102236, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3623 + }, + { + "epoch": 5.789137380191693, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 3624 + }, + { + "epoch": 5.7907348242811505, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3625 + }, + { + "epoch": 5.792332268370607, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 3626 + }, + { + "epoch": 5.793929712460064, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3627 + }, + { + "epoch": 5.795527156549521, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3628 + }, + { + "epoch": 5.797124600638978, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3629 + }, + { + "epoch": 5.798722044728435, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 3630 + }, + { + "epoch": 5.800319488817891, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 3631 + }, + { + "epoch": 5.801916932907348, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3632 + }, + { + "epoch": 5.803514376996805, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3633 + }, + { + "epoch": 5.805111821086262, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 3634 + }, + { + "epoch": 5.806709265175719, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3635 + }, + { + "epoch": 5.8083067092651754, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3636 + }, + { + "epoch": 5.8099041533546325, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3637 + }, + { + "epoch": 5.81150159744409, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 3638 + }, + { + "epoch": 5.813099041533547, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 3639 + }, + { + "epoch": 5.814696485623003, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3640 + }, + { + "epoch": 5.81629392971246, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3641 + }, + { + "epoch": 5.817891373801917, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 3642 + }, + { + "epoch": 5.819488817891374, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 3643 + }, + { + "epoch": 5.821086261980831, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3644 + }, + { + "epoch": 5.822683706070287, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3645 + }, + { + "epoch": 5.824281150159744, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3646 + }, + { + "epoch": 5.825878594249201, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3647 + }, + { + "epoch": 5.827476038338658, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 3648 + }, + { + "epoch": 5.8290734824281145, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3649 + }, + { + "epoch": 5.830670926517572, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3650 + }, + { + "epoch": 5.832268370607029, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3651 + }, + { + "epoch": 5.833865814696486, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 3652 + }, + { + "epoch": 5.835463258785943, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 3653 + }, + { + "epoch": 5.8370607028754, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3654 + }, + { + "epoch": 5.838658146964856, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 3655 + }, + { + "epoch": 5.840255591054313, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3656 + }, + { + "epoch": 5.84185303514377, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3657 + }, + { + "epoch": 5.843450479233227, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3658 + }, + { + "epoch": 5.845047923322683, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3659 + }, + { + "epoch": 5.84664536741214, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3660 + }, + { + "epoch": 5.848242811501597, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3661 + }, + { + "epoch": 5.8498402555910545, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 3662 + }, + { + "epoch": 5.8514376996805115, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 3663 + }, + { + "epoch": 5.853035143769968, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 3664 + }, + { + "epoch": 5.854632587859425, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3665 + }, + { + "epoch": 5.856230031948882, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3666 + }, + { + "epoch": 5.857827476038339, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3667 + }, + { + "epoch": 5.859424920127795, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3668 + }, + { + "epoch": 5.861022364217252, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3669 + }, + { + "epoch": 5.862619808306709, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3670 + }, + { + "epoch": 5.864217252396166, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 3671 + }, + { + "epoch": 5.865814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3672 + }, + { + "epoch": 5.86741214057508, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3673 + }, + { + "epoch": 5.8690095846645365, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3674 + }, + { + "epoch": 5.8706070287539935, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 3675 + }, + { + "epoch": 5.872204472843451, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 3676 + }, + { + "epoch": 5.873801916932908, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3677 + }, + { + "epoch": 5.875399361022364, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 3678 + }, + { + "epoch": 5.876996805111821, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 3679 + }, + { + "epoch": 5.878594249201278, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3680 + }, + { + "epoch": 5.880191693290735, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3681 + }, + { + "epoch": 5.881789137380192, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3682 + }, + { + "epoch": 5.883386581469648, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3683 + }, + { + "epoch": 5.884984025559105, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 3684 + }, + { + "epoch": 5.886581469648562, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3685 + }, + { + "epoch": 5.888178913738019, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 3686 + }, + { + "epoch": 5.8897763578274756, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3687 + }, + { + "epoch": 5.891373801916933, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3688 + }, + { + "epoch": 5.89297124600639, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3689 + }, + { + "epoch": 5.894568690095847, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3690 + }, + { + "epoch": 5.896166134185304, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3691 + }, + { + "epoch": 5.897763578274761, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 3692 + }, + { + "epoch": 5.899361022364217, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3693 + }, + { + "epoch": 5.900958466453674, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 3694 + }, + { + "epoch": 5.902555910543131, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 3695 + }, + { + "epoch": 5.904153354632588, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 3696 + }, + { + "epoch": 5.905750798722044, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3697 + }, + { + "epoch": 5.907348242811501, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3698 + }, + { + "epoch": 5.9089456869009584, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3699 + }, + { + "epoch": 5.9105431309904155, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3700 + }, + { + "epoch": 5.912140575079873, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 3701 + }, + { + "epoch": 5.913738019169329, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3702 + }, + { + "epoch": 5.915335463258786, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 3703 + }, + { + "epoch": 5.916932907348243, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3704 + }, + { + "epoch": 5.9185303514377, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3705 + }, + { + "epoch": 5.920127795527156, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3706 + }, + { + "epoch": 5.921725239616613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3707 + }, + { + "epoch": 5.92332268370607, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3708 + }, + { + "epoch": 5.924920127795527, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3709 + }, + { + "epoch": 5.926517571884984, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 3710 + }, + { + "epoch": 5.928115015974441, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3711 + }, + { + "epoch": 5.9297124600638975, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 3712 + }, + { + "epoch": 5.931309904153355, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3713 + }, + { + "epoch": 5.932907348242812, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 3714 + }, + { + "epoch": 5.934504792332269, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3715 + }, + { + "epoch": 5.936102236421725, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3716 + }, + { + "epoch": 5.937699680511182, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3717 + }, + { + "epoch": 5.939297124600639, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3718 + }, + { + "epoch": 5.940894568690096, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3719 + }, + { + "epoch": 5.942492012779553, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3720 + }, + { + "epoch": 5.944089456869009, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3721 + }, + { + "epoch": 5.945686900958466, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3722 + }, + { + "epoch": 5.947284345047923, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3723 + }, + { + "epoch": 5.94888178913738, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3724 + }, + { + "epoch": 5.950479233226837, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 3725 + }, + { + "epoch": 5.952076677316294, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 3726 + }, + { + "epoch": 5.953674121405751, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3727 + }, + { + "epoch": 5.955271565495208, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 3728 + }, + { + "epoch": 5.956869009584665, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3729 + }, + { + "epoch": 5.958466453674122, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3730 + }, + { + "epoch": 5.960063897763578, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3731 + }, + { + "epoch": 5.961661341853035, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3732 + }, + { + "epoch": 5.963258785942492, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 3733 + }, + { + "epoch": 5.964856230031949, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3734 + }, + { + "epoch": 5.966453674121405, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 3735 + }, + { + "epoch": 5.968051118210862, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 3736 + }, + { + "epoch": 5.9696485623003195, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 3737 + }, + { + "epoch": 5.9712460063897765, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3738 + }, + { + "epoch": 5.972843450479234, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3739 + }, + { + "epoch": 5.97444089456869, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3740 + }, + { + "epoch": 5.976038338658147, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 3741 + }, + { + "epoch": 5.977635782747604, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3742 + }, + { + "epoch": 5.979233226837061, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3743 + }, + { + "epoch": 5.980830670926517, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 3744 + }, + { + "epoch": 5.982428115015974, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 3745 + }, + { + "epoch": 5.984025559105431, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 3746 + }, + { + "epoch": 5.985623003194888, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3747 + }, + { + "epoch": 5.987220447284345, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 3748 + }, + { + "epoch": 5.988817891373802, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3749 + }, + { + "epoch": 5.9904153354632586, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3750 + }, + { + "epoch": 5.992012779552716, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3751 + }, + { + "epoch": 5.993610223642173, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3752 + }, + { + "epoch": 5.99520766773163, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 3753 + }, + { + "epoch": 5.996805111821086, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3754 + }, + { + "epoch": 5.998402555910543, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3755 + }, + { + "epoch": 6.0, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3756 + }, + { + "epoch": 6.001597444089457, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3757 + }, + { + "epoch": 6.003194888178914, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3758 + }, + { + "epoch": 6.00479233226837, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3759 + }, + { + "epoch": 6.006389776357827, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 3760 + }, + { + "epoch": 6.007987220447284, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 3761 + }, + { + "epoch": 6.0095846645367414, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3762 + }, + { + "epoch": 6.0111821086261985, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3763 + }, + { + "epoch": 6.012779552715655, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3764 + }, + { + "epoch": 6.014376996805112, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3765 + }, + { + "epoch": 6.015974440894569, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3766 + }, + { + "epoch": 6.017571884984026, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 3767 + }, + { + "epoch": 6.019169329073482, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3768 + }, + { + "epoch": 6.020766773162939, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3769 + }, + { + "epoch": 6.022364217252396, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 3770 + }, + { + "epoch": 6.023961661341853, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3771 + }, + { + "epoch": 6.02555910543131, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3772 + }, + { + "epoch": 6.027156549520766, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3773 + }, + { + "epoch": 6.0287539936102235, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 3774 + }, + { + "epoch": 6.0303514376996805, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3775 + }, + { + "epoch": 6.031948881789138, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3776 + }, + { + "epoch": 6.033546325878595, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3777 + }, + { + "epoch": 6.035143769968051, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3778 + }, + { + "epoch": 6.036741214057508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3779 + }, + { + "epoch": 6.038338658146965, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3780 + }, + { + "epoch": 6.039936102236422, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3781 + }, + { + "epoch": 6.041533546325879, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 3782 + }, + { + "epoch": 6.043130990415335, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 3783 + }, + { + "epoch": 6.044728434504792, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3784 + }, + { + "epoch": 6.046325878594249, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3785 + }, + { + "epoch": 6.047923322683706, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3786 + }, + { + "epoch": 6.0495207667731625, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3787 + }, + { + "epoch": 6.05111821086262, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3788 + }, + { + "epoch": 6.052715654952077, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3789 + }, + { + "epoch": 6.054313099041534, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3790 + }, + { + "epoch": 6.055910543130991, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 3791 + }, + { + "epoch": 6.057507987220447, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 3792 + }, + { + "epoch": 6.059105431309904, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3793 + }, + { + "epoch": 6.060702875399361, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3794 + }, + { + "epoch": 6.062300319488818, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 3795 + }, + { + "epoch": 6.063897763578275, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 3796 + }, + { + "epoch": 6.065495207667731, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3797 + }, + { + "epoch": 6.067092651757188, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3798 + }, + { + "epoch": 6.068690095846645, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 3799 + }, + { + "epoch": 6.0702875399361025, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 3800 + }, + { + "epoch": 6.0718849840255595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 3801 + }, + { + "epoch": 6.073482428115016, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3802 + }, + { + "epoch": 6.075079872204473, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3803 + }, + { + "epoch": 6.07667731629393, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 3804 + }, + { + "epoch": 6.078274760383387, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3805 + }, + { + "epoch": 6.079872204472843, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3806 + }, + { + "epoch": 6.0814696485623, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 3807 + }, + { + "epoch": 6.083067092651757, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3808 + }, + { + "epoch": 6.084664536741214, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3809 + }, + { + "epoch": 6.086261980830671, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3810 + }, + { + "epoch": 6.087859424920127, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3811 + }, + { + "epoch": 6.0894568690095845, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3812 + }, + { + "epoch": 6.0910543130990416, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 3813 + }, + { + "epoch": 6.092651757188499, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3814 + }, + { + "epoch": 6.094249201277956, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3815 + }, + { + "epoch": 6.095846645367412, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3816 + }, + { + "epoch": 6.097444089456869, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3817 + }, + { + "epoch": 6.099041533546326, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3818 + }, + { + "epoch": 6.100638977635783, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 3819 + }, + { + "epoch": 6.102236421725239, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 3820 + }, + { + "epoch": 6.103833865814696, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 3821 + }, + { + "epoch": 6.105431309904153, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 3822 + }, + { + "epoch": 6.10702875399361, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3823 + }, + { + "epoch": 6.108626198083067, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 3824 + }, + { + "epoch": 6.110223642172524, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3825 + }, + { + "epoch": 6.111821086261981, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3826 + }, + { + "epoch": 6.113418530351438, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 3827 + }, + { + "epoch": 6.115015974440895, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 3828 + }, + { + "epoch": 6.116613418530352, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3829 + }, + { + "epoch": 6.118210862619808, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3830 + }, + { + "epoch": 6.119808306709265, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 3831 + }, + { + "epoch": 6.121405750798722, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3832 + }, + { + "epoch": 6.123003194888179, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3833 + }, + { + "epoch": 6.124600638977636, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 3834 + }, + { + "epoch": 6.126198083067092, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3835 + }, + { + "epoch": 6.127795527156549, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3836 + }, + { + "epoch": 6.1293929712460065, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3837 + }, + { + "epoch": 6.1309904153354635, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3838 + }, + { + "epoch": 6.13258785942492, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3839 + }, + { + "epoch": 6.134185303514377, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3840 + }, + { + "epoch": 6.135782747603834, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3841 + }, + { + "epoch": 6.137380191693291, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3842 + }, + { + "epoch": 6.138977635782748, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3843 + }, + { + "epoch": 6.140575079872204, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3844 + }, + { + "epoch": 6.142172523961661, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 3845 + }, + { + "epoch": 6.143769968051118, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3846 + }, + { + "epoch": 6.145367412140575, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3847 + }, + { + "epoch": 6.146964856230032, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 3848 + }, + { + "epoch": 6.1485623003194885, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 3849 + }, + { + "epoch": 6.1501597444089455, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 3850 + }, + { + "epoch": 6.151757188498403, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3851 + }, + { + "epoch": 6.15335463258786, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 3852 + }, + { + "epoch": 6.154952076677317, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 3853 + }, + { + "epoch": 6.156549520766773, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3854 + }, + { + "epoch": 6.15814696485623, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3855 + }, + { + "epoch": 6.159744408945687, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 3856 + }, + { + "epoch": 6.161341853035144, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3857 + }, + { + "epoch": 6.1629392971246, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3858 + }, + { + "epoch": 6.164536741214057, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 3859 + }, + { + "epoch": 6.166134185303514, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3860 + }, + { + "epoch": 6.167731629392971, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3861 + }, + { + "epoch": 6.169329073482428, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 3862 + }, + { + "epoch": 6.170926517571885, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 3863 + }, + { + "epoch": 6.172523961661342, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 3864 + }, + { + "epoch": 6.174121405750799, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3865 + }, + { + "epoch": 6.175718849840256, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3866 + }, + { + "epoch": 6.177316293929713, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 3867 + }, + { + "epoch": 6.178913738019169, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 3868 + }, + { + "epoch": 6.180511182108626, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3869 + }, + { + "epoch": 6.182108626198083, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 3870 + }, + { + "epoch": 6.18370607028754, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 3871 + }, + { + "epoch": 6.185303514376997, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 3872 + }, + { + "epoch": 6.186900958466453, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3873 + }, + { + "epoch": 6.18849840255591, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3874 + }, + { + "epoch": 6.1900958466453675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 3875 + }, + { + "epoch": 6.1916932907348246, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3876 + }, + { + "epoch": 6.193290734824281, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3877 + }, + { + "epoch": 6.194888178913738, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3878 + }, + { + "epoch": 6.196485623003195, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3879 + }, + { + "epoch": 6.198083067092652, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 3880 + }, + { + "epoch": 6.199680511182109, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3881 + }, + { + "epoch": 6.201277955271565, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 3882 + }, + { + "epoch": 6.202875399361022, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 3883 + }, + { + "epoch": 6.204472843450479, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 3884 + }, + { + "epoch": 6.206070287539936, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 3885 + }, + { + "epoch": 6.207667731629393, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 3886 + }, + { + "epoch": 6.2092651757188495, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3887 + }, + { + "epoch": 6.210862619808307, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3888 + }, + { + "epoch": 6.212460063897764, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 3889 + }, + { + "epoch": 6.214057507987221, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 3890 + }, + { + "epoch": 6.215654952076678, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 3891 + }, + { + "epoch": 6.217252396166134, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3892 + }, + { + "epoch": 6.218849840255591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3893 + }, + { + "epoch": 6.220447284345048, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3894 + }, + { + "epoch": 6.222044728434505, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3895 + }, + { + "epoch": 6.223642172523961, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3896 + }, + { + "epoch": 6.225239616613418, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3897 + }, + { + "epoch": 6.226837060702875, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 3898 + }, + { + "epoch": 6.228434504792332, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3899 + }, + { + "epoch": 6.2300319488817895, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3900 + }, + { + "epoch": 6.231629392971246, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3901 + }, + { + "epoch": 6.233226837060703, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 3902 + }, + { + "epoch": 6.23482428115016, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 3903 + }, + { + "epoch": 6.236421725239617, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3904 + }, + { + "epoch": 6.238019169329074, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 3905 + }, + { + "epoch": 6.23961661341853, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 3906 + }, + { + "epoch": 6.241214057507987, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 3907 + }, + { + "epoch": 6.242811501597444, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3908 + }, + { + "epoch": 6.244408945686901, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3909 + }, + { + "epoch": 6.246006389776358, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3910 + }, + { + "epoch": 6.247603833865814, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 3911 + }, + { + "epoch": 6.2492012779552715, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 3912 + }, + { + "epoch": 6.2507987220447285, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3913 + }, + { + "epoch": 6.252396166134186, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 3914 + }, + { + "epoch": 6.253993610223642, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3915 + }, + { + "epoch": 6.255591054313099, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3916 + }, + { + "epoch": 6.257188498402556, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 3917 + }, + { + "epoch": 6.258785942492013, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 3918 + }, + { + "epoch": 6.26038338658147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 3919 + }, + { + "epoch": 6.261980830670926, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3920 + }, + { + "epoch": 6.263578274760383, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 3921 + }, + { + "epoch": 6.26517571884984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 3922 + }, + { + "epoch": 6.266773162939297, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 3923 + }, + { + "epoch": 6.268370607028754, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 3924 + }, + { + "epoch": 6.2699680511182105, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3925 + }, + { + "epoch": 6.271565495207668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3926 + }, + { + "epoch": 6.273162939297125, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3927 + }, + { + "epoch": 6.274760383386582, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 3928 + }, + { + "epoch": 6.276357827476039, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 3929 + }, + { + "epoch": 6.277955271565495, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 3930 + }, + { + "epoch": 6.279552715654952, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 3931 + }, + { + "epoch": 6.281150159744409, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 3932 + }, + { + "epoch": 6.282747603833866, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 3933 + }, + { + "epoch": 6.284345047923322, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3934 + }, + { + "epoch": 6.285942492012779, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 3935 + }, + { + "epoch": 6.287539936102236, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 3936 + }, + { + "epoch": 6.289137380191693, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 3937 + }, + { + "epoch": 6.2907348242811505, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 3938 + }, + { + "epoch": 6.292332268370607, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3939 + }, + { + "epoch": 6.293929712460064, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 3940 + }, + { + "epoch": 6.295527156549521, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 3941 + }, + { + "epoch": 6.297124600638978, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 3942 + }, + { + "epoch": 6.298722044728435, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 3943 + }, + { + "epoch": 6.300319488817891, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 3944 + }, + { + "epoch": 6.301916932907348, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 3945 + }, + { + "epoch": 6.303514376996805, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3946 + }, + { + "epoch": 6.305111821086262, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 3947 + }, + { + "epoch": 6.306709265175719, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 3948 + }, + { + "epoch": 6.3083067092651754, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 3949 + }, + { + "epoch": 6.3099041533546325, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3950 + }, + { + "epoch": 6.31150159744409, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 3951 + }, + { + "epoch": 6.313099041533547, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 3952 + }, + { + "epoch": 6.314696485623003, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3953 + }, + { + "epoch": 6.31629392971246, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3954 + }, + { + "epoch": 6.317891373801917, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 3955 + }, + { + "epoch": 6.319488817891374, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3956 + }, + { + "epoch": 6.321086261980831, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 3957 + }, + { + "epoch": 6.322683706070287, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 3958 + }, + { + "epoch": 6.324281150159744, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 3959 + }, + { + "epoch": 6.325878594249201, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3960 + }, + { + "epoch": 6.327476038338658, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3961 + }, + { + "epoch": 6.329073482428115, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 3962 + }, + { + "epoch": 6.330670926517572, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3963 + }, + { + "epoch": 6.332268370607029, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 3964 + }, + { + "epoch": 6.333865814696486, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 3965 + }, + { + "epoch": 6.335463258785943, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 3966 + }, + { + "epoch": 6.3370607028754, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 3967 + }, + { + "epoch": 6.338658146964856, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 3968 + }, + { + "epoch": 6.340255591054313, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 3969 + }, + { + "epoch": 6.34185303514377, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3970 + }, + { + "epoch": 6.343450479233227, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3971 + }, + { + "epoch": 6.345047923322683, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 3972 + }, + { + "epoch": 6.34664536741214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 3973 + }, + { + "epoch": 6.348242811501597, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 3974 + }, + { + "epoch": 6.3498402555910545, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 3975 + }, + { + "epoch": 6.3514376996805115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 3976 + }, + { + "epoch": 6.353035143769968, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 3977 + }, + { + "epoch": 6.354632587859425, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 3978 + }, + { + "epoch": 6.356230031948882, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 3979 + }, + { + "epoch": 6.357827476038339, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 3980 + }, + { + "epoch": 6.359424920127796, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3981 + }, + { + "epoch": 6.361022364217252, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 3982 + }, + { + "epoch": 6.362619808306709, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 3983 + }, + { + "epoch": 6.364217252396166, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 3984 + }, + { + "epoch": 6.365814696485623, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 3985 + }, + { + "epoch": 6.36741214057508, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 3986 + }, + { + "epoch": 6.3690095846645365, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 3987 + }, + { + "epoch": 6.3706070287539935, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 3988 + }, + { + "epoch": 6.372204472843451, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 3989 + }, + { + "epoch": 6.373801916932908, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 3990 + }, + { + "epoch": 6.375399361022364, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 3991 + }, + { + "epoch": 6.376996805111821, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 3992 + }, + { + "epoch": 6.378594249201278, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 3993 + }, + { + "epoch": 6.380191693290735, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 3994 + }, + { + "epoch": 6.381789137380192, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 3995 + }, + { + "epoch": 6.383386581469648, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 3996 + }, + { + "epoch": 6.384984025559105, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 3997 + }, + { + "epoch": 6.386581469648562, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 3998 + }, + { + "epoch": 6.388178913738019, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 3999 + }, + { + "epoch": 6.389776357827476, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4000 + }, + { + "epoch": 6.391373801916933, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4001 + }, + { + "epoch": 6.39297124600639, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4002 + }, + { + "epoch": 6.394568690095847, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4003 + }, + { + "epoch": 6.396166134185304, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 4004 + }, + { + "epoch": 6.397763578274761, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4005 + }, + { + "epoch": 6.399361022364217, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4006 + }, + { + "epoch": 6.400958466453674, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4007 + }, + { + "epoch": 6.402555910543131, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4008 + }, + { + "epoch": 6.404153354632588, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4009 + }, + { + "epoch": 6.405750798722044, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4010 + }, + { + "epoch": 6.407348242811501, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 4011 + }, + { + "epoch": 6.4089456869009584, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4012 + }, + { + "epoch": 6.4105431309904155, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4013 + }, + { + "epoch": 6.412140575079873, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 4014 + }, + { + "epoch": 6.413738019169329, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4015 + }, + { + "epoch": 6.415335463258786, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4016 + }, + { + "epoch": 6.416932907348243, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4017 + }, + { + "epoch": 6.4185303514377, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 4018 + }, + { + "epoch": 6.420127795527157, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4019 + }, + { + "epoch": 6.421725239616613, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4020 + }, + { + "epoch": 6.42332268370607, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4021 + }, + { + "epoch": 6.424920127795527, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 4022 + }, + { + "epoch": 6.426517571884984, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4023 + }, + { + "epoch": 6.428115015974441, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4024 + }, + { + "epoch": 6.4297124600638975, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4025 + }, + { + "epoch": 6.431309904153355, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4026 + }, + { + "epoch": 6.432907348242812, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4027 + }, + { + "epoch": 6.434504792332269, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4028 + }, + { + "epoch": 6.436102236421725, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4029 + }, + { + "epoch": 6.437699680511182, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4030 + }, + { + "epoch": 6.439297124600639, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4031 + }, + { + "epoch": 6.440894568690096, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4032 + }, + { + "epoch": 6.442492012779553, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4033 + }, + { + "epoch": 6.444089456869009, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4034 + }, + { + "epoch": 6.445686900958466, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 4035 + }, + { + "epoch": 6.447284345047923, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4036 + }, + { + "epoch": 6.44888178913738, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4037 + }, + { + "epoch": 6.4504792332268375, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4038 + }, + { + "epoch": 6.452076677316294, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4039 + }, + { + "epoch": 6.453674121405751, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 4040 + }, + { + "epoch": 6.455271565495208, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4041 + }, + { + "epoch": 6.456869009584665, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4042 + }, + { + "epoch": 6.458466453674122, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4043 + }, + { + "epoch": 6.460063897763578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4044 + }, + { + "epoch": 6.461661341853035, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4045 + }, + { + "epoch": 6.463258785942492, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4046 + }, + { + "epoch": 6.464856230031949, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4047 + }, + { + "epoch": 6.466453674121405, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4048 + }, + { + "epoch": 6.468051118210862, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4049 + }, + { + "epoch": 6.4696485623003195, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4050 + }, + { + "epoch": 6.4712460063897765, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4051 + }, + { + "epoch": 6.472843450479234, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4052 + }, + { + "epoch": 6.47444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4053 + }, + { + "epoch": 6.476038338658147, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 4054 + }, + { + "epoch": 6.477635782747604, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4055 + }, + { + "epoch": 6.479233226837061, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4056 + }, + { + "epoch": 6.480830670926517, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4057 + }, + { + "epoch": 6.482428115015974, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4058 + }, + { + "epoch": 6.484025559105431, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4059 + }, + { + "epoch": 6.485623003194888, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4060 + }, + { + "epoch": 6.487220447284345, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4061 + }, + { + "epoch": 6.488817891373802, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4062 + }, + { + "epoch": 6.4904153354632586, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4063 + }, + { + "epoch": 6.492012779552716, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4064 + }, + { + "epoch": 6.493610223642173, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4065 + }, + { + "epoch": 6.49520766773163, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4066 + }, + { + "epoch": 6.496805111821086, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4067 + }, + { + "epoch": 6.498402555910543, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4068 + }, + { + "epoch": 6.5, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 4069 + }, + { + "epoch": 6.501597444089457, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4070 + }, + { + "epoch": 6.503194888178914, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4071 + }, + { + "epoch": 6.50479233226837, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4072 + }, + { + "epoch": 6.506389776357827, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 4073 + }, + { + "epoch": 6.507987220447284, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4074 + }, + { + "epoch": 6.5095846645367414, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4075 + }, + { + "epoch": 6.511182108626198, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4076 + }, + { + "epoch": 6.512779552715655, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4077 + }, + { + "epoch": 6.514376996805112, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 4078 + }, + { + "epoch": 6.515974440894569, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4079 + }, + { + "epoch": 6.517571884984026, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4080 + }, + { + "epoch": 6.519169329073483, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4081 + }, + { + "epoch": 6.520766773162939, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4082 + }, + { + "epoch": 6.522364217252396, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4083 + }, + { + "epoch": 6.523961661341853, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4084 + }, + { + "epoch": 6.52555910543131, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4085 + }, + { + "epoch": 6.527156549520766, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4086 + }, + { + "epoch": 6.5287539936102235, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4087 + }, + { + "epoch": 6.5303514376996805, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4088 + }, + { + "epoch": 6.531948881789138, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4089 + }, + { + "epoch": 6.533546325878595, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 4090 + }, + { + "epoch": 6.535143769968051, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4091 + }, + { + "epoch": 6.536741214057508, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4092 + }, + { + "epoch": 6.538338658146965, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4093 + }, + { + "epoch": 6.539936102236422, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4094 + }, + { + "epoch": 6.541533546325878, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4095 + }, + { + "epoch": 6.543130990415335, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4096 + }, + { + "epoch": 6.544728434504792, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4097 + }, + { + "epoch": 6.546325878594249, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4098 + }, + { + "epoch": 6.547923322683706, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4099 + }, + { + "epoch": 6.549520766773163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4100 + }, + { + "epoch": 6.55111821086262, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4101 + }, + { + "epoch": 6.552715654952077, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4102 + }, + { + "epoch": 6.554313099041534, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4103 + }, + { + "epoch": 6.555910543130991, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 4104 + }, + { + "epoch": 6.557507987220447, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 4105 + }, + { + "epoch": 6.559105431309904, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4106 + }, + { + "epoch": 6.560702875399361, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4107 + }, + { + "epoch": 6.562300319488818, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4108 + }, + { + "epoch": 6.563897763578275, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4109 + }, + { + "epoch": 6.565495207667731, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4110 + }, + { + "epoch": 6.567092651757188, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4111 + }, + { + "epoch": 6.568690095846645, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4112 + }, + { + "epoch": 6.5702875399361025, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4113 + }, + { + "epoch": 6.571884984025559, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4114 + }, + { + "epoch": 6.573482428115016, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4115 + }, + { + "epoch": 6.575079872204473, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4116 + }, + { + "epoch": 6.57667731629393, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4117 + }, + { + "epoch": 6.578274760383387, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4118 + }, + { + "epoch": 6.579872204472844, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4119 + }, + { + "epoch": 6.5814696485623, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4120 + }, + { + "epoch": 6.583067092651757, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4121 + }, + { + "epoch": 6.584664536741214, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4122 + }, + { + "epoch": 6.586261980830671, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4123 + }, + { + "epoch": 6.587859424920127, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4124 + }, + { + "epoch": 6.5894568690095845, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4125 + }, + { + "epoch": 6.5910543130990416, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4126 + }, + { + "epoch": 6.592651757188499, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 4127 + }, + { + "epoch": 6.594249201277956, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4128 + }, + { + "epoch": 6.595846645367412, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4129 + }, + { + "epoch": 6.597444089456869, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 4130 + }, + { + "epoch": 6.599041533546326, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 4131 + }, + { + "epoch": 6.600638977635783, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4132 + }, + { + "epoch": 6.602236421725239, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4133 + }, + { + "epoch": 6.603833865814696, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4134 + }, + { + "epoch": 6.605431309904153, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4135 + }, + { + "epoch": 6.60702875399361, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 4136 + }, + { + "epoch": 6.608626198083067, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4137 + }, + { + "epoch": 6.6102236421725244, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4138 + }, + { + "epoch": 6.611821086261981, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 4139 + }, + { + "epoch": 6.613418530351438, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 4140 + }, + { + "epoch": 6.615015974440895, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4141 + }, + { + "epoch": 6.616613418530352, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4142 + }, + { + "epoch": 6.618210862619808, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4143 + }, + { + "epoch": 6.619808306709265, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4144 + }, + { + "epoch": 6.621405750798722, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4145 + }, + { + "epoch": 6.623003194888179, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4146 + }, + { + "epoch": 6.624600638977636, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4147 + }, + { + "epoch": 6.626198083067092, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4148 + }, + { + "epoch": 6.627795527156549, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4149 + }, + { + "epoch": 6.6293929712460065, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4150 + }, + { + "epoch": 6.6309904153354635, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4151 + }, + { + "epoch": 6.63258785942492, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 4152 + }, + { + "epoch": 6.634185303514377, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4153 + }, + { + "epoch": 6.635782747603834, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4154 + }, + { + "epoch": 6.637380191693291, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4155 + }, + { + "epoch": 6.638977635782748, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 4156 + }, + { + "epoch": 6.640575079872205, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4157 + }, + { + "epoch": 6.642172523961661, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4158 + }, + { + "epoch": 6.643769968051118, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4159 + }, + { + "epoch": 6.645367412140575, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4160 + }, + { + "epoch": 6.646964856230032, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4161 + }, + { + "epoch": 6.6485623003194885, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4162 + }, + { + "epoch": 6.6501597444089455, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4163 + }, + { + "epoch": 6.651757188498403, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4164 + }, + { + "epoch": 6.65335463258786, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4165 + }, + { + "epoch": 6.654952076677317, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4166 + }, + { + "epoch": 6.656549520766773, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4167 + }, + { + "epoch": 6.65814696485623, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4168 + }, + { + "epoch": 6.659744408945687, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4169 + }, + { + "epoch": 6.661341853035144, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4170 + }, + { + "epoch": 6.6629392971246, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4171 + }, + { + "epoch": 6.664536741214057, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 4172 + }, + { + "epoch": 6.666134185303514, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4173 + }, + { + "epoch": 6.667731629392971, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4174 + }, + { + "epoch": 6.669329073482428, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4175 + }, + { + "epoch": 6.6709265175718855, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4176 + }, + { + "epoch": 6.672523961661342, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4177 + }, + { + "epoch": 6.674121405750799, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4178 + }, + { + "epoch": 6.675718849840256, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4179 + }, + { + "epoch": 6.677316293929713, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4180 + }, + { + "epoch": 6.678913738019169, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4181 + }, + { + "epoch": 6.680511182108626, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4182 + }, + { + "epoch": 6.682108626198083, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4183 + }, + { + "epoch": 6.68370607028754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 4184 + }, + { + "epoch": 6.685303514376997, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 4185 + }, + { + "epoch": 6.686900958466453, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4186 + }, + { + "epoch": 6.68849840255591, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4187 + }, + { + "epoch": 6.6900958466453675, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4188 + }, + { + "epoch": 6.6916932907348246, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4189 + }, + { + "epoch": 6.693290734824281, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4190 + }, + { + "epoch": 6.694888178913738, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4191 + }, + { + "epoch": 6.696485623003195, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4192 + }, + { + "epoch": 6.698083067092652, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4193 + }, + { + "epoch": 6.699680511182109, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4194 + }, + { + "epoch": 6.701277955271565, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4195 + }, + { + "epoch": 6.702875399361022, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 4196 + }, + { + "epoch": 6.704472843450479, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4197 + }, + { + "epoch": 6.706070287539936, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4198 + }, + { + "epoch": 6.707667731629393, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4199 + }, + { + "epoch": 6.7092651757188495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4200 + }, + { + "epoch": 6.710862619808307, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 4201 + }, + { + "epoch": 6.712460063897764, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4202 + }, + { + "epoch": 6.714057507987221, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4203 + }, + { + "epoch": 6.715654952076678, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4204 + }, + { + "epoch": 6.717252396166134, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4205 + }, + { + "epoch": 6.718849840255591, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4206 + }, + { + "epoch": 6.720447284345048, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4207 + }, + { + "epoch": 6.722044728434505, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4208 + }, + { + "epoch": 6.723642172523961, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4209 + }, + { + "epoch": 6.725239616613418, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4210 + }, + { + "epoch": 6.726837060702875, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4211 + }, + { + "epoch": 6.728434504792332, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4212 + }, + { + "epoch": 6.7300319488817895, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4213 + }, + { + "epoch": 6.731629392971246, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4214 + }, + { + "epoch": 6.733226837060703, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4215 + }, + { + "epoch": 6.73482428115016, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4216 + }, + { + "epoch": 6.736421725239617, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4217 + }, + { + "epoch": 6.738019169329074, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 4218 + }, + { + "epoch": 6.73961661341853, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4219 + }, + { + "epoch": 6.741214057507987, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4220 + }, + { + "epoch": 6.742811501597444, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4221 + }, + { + "epoch": 6.744408945686901, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4222 + }, + { + "epoch": 6.746006389776358, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 4223 + }, + { + "epoch": 6.747603833865814, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4224 + }, + { + "epoch": 6.7492012779552715, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4225 + }, + { + "epoch": 6.7507987220447285, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4226 + }, + { + "epoch": 6.752396166134186, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 4227 + }, + { + "epoch": 6.753993610223642, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 4228 + }, + { + "epoch": 6.755591054313099, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 4229 + }, + { + "epoch": 6.757188498402556, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4230 + }, + { + "epoch": 6.758785942492013, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4231 + }, + { + "epoch": 6.76038338658147, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4232 + }, + { + "epoch": 6.761980830670926, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4233 + }, + { + "epoch": 6.763578274760383, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4234 + }, + { + "epoch": 6.76517571884984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4235 + }, + { + "epoch": 6.766773162939297, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4236 + }, + { + "epoch": 6.768370607028754, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 4237 + }, + { + "epoch": 6.7699680511182105, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4238 + }, + { + "epoch": 6.771565495207668, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4239 + }, + { + "epoch": 6.773162939297125, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4240 + }, + { + "epoch": 6.774760383386582, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 4241 + }, + { + "epoch": 6.776357827476039, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4242 + }, + { + "epoch": 6.777955271565495, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4243 + }, + { + "epoch": 6.779552715654952, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4244 + }, + { + "epoch": 6.781150159744409, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4245 + }, + { + "epoch": 6.782747603833866, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4246 + }, + { + "epoch": 6.784345047923322, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4247 + }, + { + "epoch": 6.785942492012779, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4248 + }, + { + "epoch": 6.787539936102236, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4249 + }, + { + "epoch": 6.789137380191693, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4250 + }, + { + "epoch": 6.7907348242811505, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4251 + }, + { + "epoch": 6.792332268370607, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4252 + }, + { + "epoch": 6.793929712460064, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4253 + }, + { + "epoch": 6.795527156549521, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4254 + }, + { + "epoch": 6.797124600638978, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4255 + }, + { + "epoch": 6.798722044728435, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4256 + }, + { + "epoch": 6.800319488817891, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4257 + }, + { + "epoch": 6.801916932907348, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4258 + }, + { + "epoch": 6.803514376996805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4259 + }, + { + "epoch": 6.805111821086262, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4260 + }, + { + "epoch": 6.806709265175719, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 4261 + }, + { + "epoch": 6.8083067092651754, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4262 + }, + { + "epoch": 6.8099041533546325, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4263 + }, + { + "epoch": 6.81150159744409, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4264 + }, + { + "epoch": 6.813099041533547, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4265 + }, + { + "epoch": 6.814696485623003, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4266 + }, + { + "epoch": 6.81629392971246, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4267 + }, + { + "epoch": 6.817891373801917, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4268 + }, + { + "epoch": 6.819488817891374, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4269 + }, + { + "epoch": 6.821086261980831, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 4270 + }, + { + "epoch": 6.822683706070287, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4271 + }, + { + "epoch": 6.824281150159744, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4272 + }, + { + "epoch": 6.825878594249201, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4273 + }, + { + "epoch": 6.827476038338658, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4274 + }, + { + "epoch": 6.8290734824281145, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4275 + }, + { + "epoch": 6.830670926517572, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4276 + }, + { + "epoch": 6.832268370607029, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4277 + }, + { + "epoch": 6.833865814696486, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4278 + }, + { + "epoch": 6.835463258785943, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4279 + }, + { + "epoch": 6.8370607028754, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4280 + }, + { + "epoch": 6.838658146964856, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4281 + }, + { + "epoch": 6.840255591054313, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4282 + }, + { + "epoch": 6.84185303514377, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4283 + }, + { + "epoch": 6.843450479233227, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4284 + }, + { + "epoch": 6.845047923322683, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4285 + }, + { + "epoch": 6.84664536741214, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4286 + }, + { + "epoch": 6.848242811501597, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4287 + }, + { + "epoch": 6.8498402555910545, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4288 + }, + { + "epoch": 6.8514376996805115, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4289 + }, + { + "epoch": 6.853035143769968, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4290 + }, + { + "epoch": 6.854632587859425, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4291 + }, + { + "epoch": 6.856230031948882, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4292 + }, + { + "epoch": 6.857827476038339, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4293 + }, + { + "epoch": 6.859424920127795, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4294 + }, + { + "epoch": 6.861022364217252, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4295 + }, + { + "epoch": 6.862619808306709, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4296 + }, + { + "epoch": 6.864217252396166, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4297 + }, + { + "epoch": 6.865814696485623, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4298 + }, + { + "epoch": 6.86741214057508, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4299 + }, + { + "epoch": 6.8690095846645365, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4300 + }, + { + "epoch": 6.8706070287539935, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4301 + }, + { + "epoch": 6.872204472843451, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 4302 + }, + { + "epoch": 6.873801916932908, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 4303 + }, + { + "epoch": 6.875399361022364, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4304 + }, + { + "epoch": 6.876996805111821, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4305 + }, + { + "epoch": 6.878594249201278, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4306 + }, + { + "epoch": 6.880191693290735, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4307 + }, + { + "epoch": 6.881789137380192, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4308 + }, + { + "epoch": 6.883386581469648, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4309 + }, + { + "epoch": 6.884984025559105, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4310 + }, + { + "epoch": 6.886581469648562, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4311 + }, + { + "epoch": 6.888178913738019, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4312 + }, + { + "epoch": 6.8897763578274756, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4313 + }, + { + "epoch": 6.891373801916933, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4314 + }, + { + "epoch": 6.89297124600639, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4315 + }, + { + "epoch": 6.894568690095847, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4316 + }, + { + "epoch": 6.896166134185304, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4317 + }, + { + "epoch": 6.897763578274761, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4318 + }, + { + "epoch": 6.899361022364217, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4319 + }, + { + "epoch": 6.900958466453674, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4320 + }, + { + "epoch": 6.902555910543131, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4321 + }, + { + "epoch": 6.904153354632588, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4322 + }, + { + "epoch": 6.905750798722044, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4323 + }, + { + "epoch": 6.907348242811501, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4324 + }, + { + "epoch": 6.9089456869009584, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4325 + }, + { + "epoch": 6.9105431309904155, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4326 + }, + { + "epoch": 6.912140575079873, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4327 + }, + { + "epoch": 6.913738019169329, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 4328 + }, + { + "epoch": 6.915335463258786, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4329 + }, + { + "epoch": 6.916932907348243, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4330 + }, + { + "epoch": 6.9185303514377, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4331 + }, + { + "epoch": 6.920127795527156, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4332 + }, + { + "epoch": 6.921725239616613, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4333 + }, + { + "epoch": 6.92332268370607, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4334 + }, + { + "epoch": 6.924920127795527, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 4335 + }, + { + "epoch": 6.926517571884984, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4336 + }, + { + "epoch": 6.928115015974441, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4337 + }, + { + "epoch": 6.9297124600638975, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4338 + }, + { + "epoch": 6.931309904153355, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4339 + }, + { + "epoch": 6.932907348242812, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4340 + }, + { + "epoch": 6.934504792332269, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4341 + }, + { + "epoch": 6.936102236421725, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4342 + }, + { + "epoch": 6.937699680511182, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4343 + }, + { + "epoch": 6.939297124600639, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4344 + }, + { + "epoch": 6.940894568690096, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4345 + }, + { + "epoch": 6.942492012779553, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 4346 + }, + { + "epoch": 6.944089456869009, + "grad_norm": 0.34375, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4347 + }, + { + "epoch": 6.945686900958466, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4348 + }, + { + "epoch": 6.947284345047923, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4349 + }, + { + "epoch": 6.94888178913738, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4350 + }, + { + "epoch": 6.950479233226837, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4351 + }, + { + "epoch": 6.952076677316294, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0777, + "step": 4352 + }, + { + "epoch": 6.953674121405751, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4353 + }, + { + "epoch": 6.955271565495208, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4354 + }, + { + "epoch": 6.956869009584665, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4355 + }, + { + "epoch": 6.958466453674122, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4356 + }, + { + "epoch": 6.960063897763578, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 4357 + }, + { + "epoch": 6.961661341853035, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4358 + }, + { + "epoch": 6.963258785942492, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4359 + }, + { + "epoch": 6.964856230031949, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4360 + }, + { + "epoch": 6.966453674121405, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4361 + }, + { + "epoch": 6.968051118210862, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4362 + }, + { + "epoch": 6.9696485623003195, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4363 + }, + { + "epoch": 6.9712460063897765, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4364 + }, + { + "epoch": 6.972843450479234, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4365 + }, + { + "epoch": 6.97444089456869, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4366 + }, + { + "epoch": 6.976038338658147, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4367 + }, + { + "epoch": 6.977635782747604, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4368 + }, + { + "epoch": 6.979233226837061, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4369 + }, + { + "epoch": 6.980830670926517, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4370 + }, + { + "epoch": 6.982428115015974, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4371 + }, + { + "epoch": 6.984025559105431, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4372 + }, + { + "epoch": 6.985623003194888, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4373 + }, + { + "epoch": 6.987220447284345, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4374 + }, + { + "epoch": 6.988817891373802, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4375 + }, + { + "epoch": 6.9904153354632586, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4376 + }, + { + "epoch": 6.992012779552716, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4377 + }, + { + "epoch": 6.993610223642173, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4378 + }, + { + "epoch": 6.99520766773163, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4379 + }, + { + "epoch": 6.996805111821086, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 4380 + }, + { + "epoch": 6.998402555910543, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4381 + }, + { + "epoch": 7.0, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4382 + }, + { + "epoch": 7.001597444089457, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4383 + }, + { + "epoch": 7.003194888178914, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 4384 + }, + { + "epoch": 7.00479233226837, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4385 + }, + { + "epoch": 7.006389776357827, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4386 + }, + { + "epoch": 7.007987220447284, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4387 + }, + { + "epoch": 7.0095846645367414, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4388 + }, + { + "epoch": 7.0111821086261985, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4389 + }, + { + "epoch": 7.012779552715655, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4390 + }, + { + "epoch": 7.014376996805112, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4391 + }, + { + "epoch": 7.015974440894569, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4392 + }, + { + "epoch": 7.017571884984026, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4393 + }, + { + "epoch": 7.019169329073482, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4394 + }, + { + "epoch": 7.020766773162939, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4395 + }, + { + "epoch": 7.022364217252396, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 4396 + }, + { + "epoch": 7.023961661341853, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4397 + }, + { + "epoch": 7.02555910543131, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4398 + }, + { + "epoch": 7.027156549520766, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4399 + }, + { + "epoch": 7.0287539936102235, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4400 + }, + { + "epoch": 7.0303514376996805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4401 + }, + { + "epoch": 7.031948881789138, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4402 + }, + { + "epoch": 7.033546325878595, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4403 + }, + { + "epoch": 7.035143769968051, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4404 + }, + { + "epoch": 7.036741214057508, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4405 + }, + { + "epoch": 7.038338658146965, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4406 + }, + { + "epoch": 7.039936102236422, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4407 + }, + { + "epoch": 7.041533546325879, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 4408 + }, + { + "epoch": 7.043130990415335, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4409 + }, + { + "epoch": 7.044728434504792, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4410 + }, + { + "epoch": 7.046325878594249, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4411 + }, + { + "epoch": 7.047923322683706, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4412 + }, + { + "epoch": 7.0495207667731625, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4413 + }, + { + "epoch": 7.05111821086262, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4414 + }, + { + "epoch": 7.052715654952077, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4415 + }, + { + "epoch": 7.054313099041534, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4416 + }, + { + "epoch": 7.055910543130991, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4417 + }, + { + "epoch": 7.057507987220447, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4418 + }, + { + "epoch": 7.059105431309904, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4419 + }, + { + "epoch": 7.060702875399361, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4420 + }, + { + "epoch": 7.062300319488818, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4421 + }, + { + "epoch": 7.063897763578275, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4422 + }, + { + "epoch": 7.065495207667731, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4423 + }, + { + "epoch": 7.067092651757188, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4424 + }, + { + "epoch": 7.068690095846645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 4425 + }, + { + "epoch": 7.0702875399361025, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4426 + }, + { + "epoch": 7.0718849840255595, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4427 + }, + { + "epoch": 7.073482428115016, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 4428 + }, + { + "epoch": 7.075079872204473, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4429 + }, + { + "epoch": 7.07667731629393, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4430 + }, + { + "epoch": 7.078274760383387, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4431 + }, + { + "epoch": 7.079872204472843, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4432 + }, + { + "epoch": 7.0814696485623, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4433 + }, + { + "epoch": 7.083067092651757, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4434 + }, + { + "epoch": 7.084664536741214, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 4435 + }, + { + "epoch": 7.086261980830671, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4436 + }, + { + "epoch": 7.087859424920127, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4437 + }, + { + "epoch": 7.0894568690095845, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4438 + }, + { + "epoch": 7.0910543130990416, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4439 + }, + { + "epoch": 7.092651757188499, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 4440 + }, + { + "epoch": 7.094249201277956, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4441 + }, + { + "epoch": 7.095846645367412, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4442 + }, + { + "epoch": 7.097444089456869, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4443 + }, + { + "epoch": 7.099041533546326, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4444 + }, + { + "epoch": 7.100638977635783, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4445 + }, + { + "epoch": 7.102236421725239, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4446 + }, + { + "epoch": 7.103833865814696, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4447 + }, + { + "epoch": 7.105431309904153, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4448 + }, + { + "epoch": 7.10702875399361, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4449 + }, + { + "epoch": 7.108626198083067, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4450 + }, + { + "epoch": 7.110223642172524, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4451 + }, + { + "epoch": 7.111821086261981, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4452 + }, + { + "epoch": 7.113418530351438, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4453 + }, + { + "epoch": 7.115015974440895, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4454 + }, + { + "epoch": 7.116613418530352, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4455 + }, + { + "epoch": 7.118210862619808, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4456 + }, + { + "epoch": 7.119808306709265, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4457 + }, + { + "epoch": 7.121405750798722, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4458 + }, + { + "epoch": 7.123003194888179, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 4459 + }, + { + "epoch": 7.124600638977636, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4460 + }, + { + "epoch": 7.126198083067092, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4461 + }, + { + "epoch": 7.127795527156549, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4462 + }, + { + "epoch": 7.1293929712460065, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4463 + }, + { + "epoch": 7.1309904153354635, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4464 + }, + { + "epoch": 7.13258785942492, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4465 + }, + { + "epoch": 7.134185303514377, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4466 + }, + { + "epoch": 7.135782747603834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4467 + }, + { + "epoch": 7.137380191693291, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4468 + }, + { + "epoch": 7.138977635782748, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4469 + }, + { + "epoch": 7.140575079872204, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 4470 + }, + { + "epoch": 7.142172523961661, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4471 + }, + { + "epoch": 7.143769968051118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4472 + }, + { + "epoch": 7.145367412140575, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4473 + }, + { + "epoch": 7.146964856230032, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4474 + }, + { + "epoch": 7.1485623003194885, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4475 + }, + { + "epoch": 7.1501597444089455, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4476 + }, + { + "epoch": 7.151757188498403, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4477 + }, + { + "epoch": 7.15335463258786, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 4478 + }, + { + "epoch": 7.154952076677317, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4479 + }, + { + "epoch": 7.156549520766773, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 4480 + }, + { + "epoch": 7.15814696485623, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4481 + }, + { + "epoch": 7.159744408945687, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4482 + }, + { + "epoch": 7.161341853035144, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4483 + }, + { + "epoch": 7.1629392971246, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4484 + }, + { + "epoch": 7.164536741214057, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4485 + }, + { + "epoch": 7.166134185303514, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4486 + }, + { + "epoch": 7.167731629392971, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4487 + }, + { + "epoch": 7.169329073482428, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4488 + }, + { + "epoch": 7.170926517571885, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4489 + }, + { + "epoch": 7.172523961661342, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4490 + }, + { + "epoch": 7.174121405750799, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4491 + }, + { + "epoch": 7.175718849840256, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 4492 + }, + { + "epoch": 7.177316293929713, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4493 + }, + { + "epoch": 7.178913738019169, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4494 + }, + { + "epoch": 7.180511182108626, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4495 + }, + { + "epoch": 7.182108626198083, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 4496 + }, + { + "epoch": 7.18370607028754, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4497 + }, + { + "epoch": 7.185303514376997, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4498 + }, + { + "epoch": 7.186900958466453, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4499 + }, + { + "epoch": 7.18849840255591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4500 + }, + { + "epoch": 7.1900958466453675, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4501 + }, + { + "epoch": 7.1916932907348246, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 4502 + }, + { + "epoch": 7.193290734824281, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4503 + }, + { + "epoch": 7.194888178913738, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 4504 + }, + { + "epoch": 7.196485623003195, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4505 + }, + { + "epoch": 7.198083067092652, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4506 + }, + { + "epoch": 7.199680511182109, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4507 + }, + { + "epoch": 7.201277955271565, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4508 + }, + { + "epoch": 7.202875399361022, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4509 + }, + { + "epoch": 7.204472843450479, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4510 + }, + { + "epoch": 7.206070287539936, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4511 + }, + { + "epoch": 7.207667731629393, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4512 + }, + { + "epoch": 7.2092651757188495, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 4513 + }, + { + "epoch": 7.210862619808307, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4514 + }, + { + "epoch": 7.212460063897764, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4515 + }, + { + "epoch": 7.214057507987221, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 4516 + }, + { + "epoch": 7.215654952076678, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4517 + }, + { + "epoch": 7.217252396166134, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4518 + }, + { + "epoch": 7.218849840255591, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4519 + }, + { + "epoch": 7.220447284345048, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4520 + }, + { + "epoch": 7.222044728434505, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4521 + }, + { + "epoch": 7.223642172523961, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4522 + }, + { + "epoch": 7.225239616613418, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 4523 + }, + { + "epoch": 7.226837060702875, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4524 + }, + { + "epoch": 7.228434504792332, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 4525 + }, + { + "epoch": 7.2300319488817895, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4526 + }, + { + "epoch": 7.231629392971246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4527 + }, + { + "epoch": 7.233226837060703, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4528 + }, + { + "epoch": 7.23482428115016, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4529 + }, + { + "epoch": 7.236421725239617, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4530 + }, + { + "epoch": 7.238019169329074, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4531 + }, + { + "epoch": 7.23961661341853, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4532 + }, + { + "epoch": 7.241214057507987, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4533 + }, + { + "epoch": 7.242811501597444, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4534 + }, + { + "epoch": 7.244408945686901, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4535 + }, + { + "epoch": 7.246006389776358, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4536 + }, + { + "epoch": 7.247603833865814, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4537 + }, + { + "epoch": 7.2492012779552715, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4538 + }, + { + "epoch": 7.2507987220447285, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4539 + }, + { + "epoch": 7.252396166134186, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4540 + }, + { + "epoch": 7.253993610223642, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4541 + }, + { + "epoch": 7.255591054313099, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4542 + }, + { + "epoch": 7.257188498402556, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 4543 + }, + { + "epoch": 7.258785942492013, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 4544 + }, + { + "epoch": 7.26038338658147, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 4545 + }, + { + "epoch": 7.261980830670926, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4546 + }, + { + "epoch": 7.263578274760383, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4547 + }, + { + "epoch": 7.26517571884984, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4548 + }, + { + "epoch": 7.266773162939297, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4549 + }, + { + "epoch": 7.268370607028754, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4550 + }, + { + "epoch": 7.2699680511182105, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4551 + }, + { + "epoch": 7.271565495207668, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4552 + }, + { + "epoch": 7.273162939297125, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4553 + }, + { + "epoch": 7.274760383386582, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4554 + }, + { + "epoch": 7.276357827476039, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4555 + }, + { + "epoch": 7.277955271565495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 4556 + }, + { + "epoch": 7.279552715654952, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4557 + }, + { + "epoch": 7.281150159744409, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4558 + }, + { + "epoch": 7.282747603833866, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4559 + }, + { + "epoch": 7.284345047923322, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4560 + }, + { + "epoch": 7.285942492012779, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 4561 + }, + { + "epoch": 7.287539936102236, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4562 + }, + { + "epoch": 7.289137380191693, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 4563 + }, + { + "epoch": 7.2907348242811505, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 4564 + }, + { + "epoch": 7.292332268370607, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4565 + }, + { + "epoch": 7.293929712460064, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4566 + }, + { + "epoch": 7.295527156549521, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 4567 + }, + { + "epoch": 7.297124600638978, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4568 + }, + { + "epoch": 7.298722044728435, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4569 + }, + { + "epoch": 7.300319488817891, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4570 + }, + { + "epoch": 7.301916932907348, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4571 + }, + { + "epoch": 7.303514376996805, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4572 + }, + { + "epoch": 7.305111821086262, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 4573 + }, + { + "epoch": 7.306709265175719, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4574 + }, + { + "epoch": 7.3083067092651754, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4575 + }, + { + "epoch": 7.3099041533546325, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4576 + }, + { + "epoch": 7.31150159744409, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 4577 + }, + { + "epoch": 7.313099041533547, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4578 + }, + { + "epoch": 7.314696485623003, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 4579 + }, + { + "epoch": 7.31629392971246, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4580 + }, + { + "epoch": 7.317891373801917, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4581 + }, + { + "epoch": 7.319488817891374, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4582 + }, + { + "epoch": 7.321086261980831, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4583 + }, + { + "epoch": 7.322683706070287, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4584 + }, + { + "epoch": 7.324281150159744, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4585 + }, + { + "epoch": 7.325878594249201, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4586 + }, + { + "epoch": 7.327476038338658, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 4587 + }, + { + "epoch": 7.329073482428115, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4588 + }, + { + "epoch": 7.330670926517572, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4589 + }, + { + "epoch": 7.332268370607029, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 4590 + }, + { + "epoch": 7.333865814696486, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4591 + }, + { + "epoch": 7.335463258785943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 4592 + }, + { + "epoch": 7.3370607028754, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4593 + }, + { + "epoch": 7.338658146964856, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4594 + }, + { + "epoch": 7.340255591054313, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 4595 + }, + { + "epoch": 7.34185303514377, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 4596 + }, + { + "epoch": 7.343450479233227, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4597 + }, + { + "epoch": 7.345047923322683, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4598 + }, + { + "epoch": 7.34664536741214, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4599 + }, + { + "epoch": 7.348242811501597, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4600 + }, + { + "epoch": 7.3498402555910545, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4601 + }, + { + "epoch": 7.3514376996805115, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4602 + }, + { + "epoch": 7.353035143769968, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4603 + }, + { + "epoch": 7.354632587859425, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4604 + }, + { + "epoch": 7.356230031948882, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4605 + }, + { + "epoch": 7.357827476038339, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4606 + }, + { + "epoch": 7.359424920127796, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4607 + }, + { + "epoch": 7.361022364217252, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4608 + }, + { + "epoch": 7.362619808306709, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 4609 + }, + { + "epoch": 7.364217252396166, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 4610 + }, + { + "epoch": 7.365814696485623, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4611 + }, + { + "epoch": 7.36741214057508, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4612 + }, + { + "epoch": 7.3690095846645365, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4613 + }, + { + "epoch": 7.3706070287539935, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4614 + }, + { + "epoch": 7.372204472843451, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4615 + }, + { + "epoch": 7.373801916932908, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4616 + }, + { + "epoch": 7.375399361022364, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4617 + }, + { + "epoch": 7.376996805111821, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4618 + }, + { + "epoch": 7.378594249201278, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4619 + }, + { + "epoch": 7.380191693290735, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4620 + }, + { + "epoch": 7.381789137380192, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4621 + }, + { + "epoch": 7.383386581469648, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4622 + }, + { + "epoch": 7.384984025559105, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 4623 + }, + { + "epoch": 7.386581469648562, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4624 + }, + { + "epoch": 7.388178913738019, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4625 + }, + { + "epoch": 7.389776357827476, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4626 + }, + { + "epoch": 7.391373801916933, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 4627 + }, + { + "epoch": 7.39297124600639, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 4628 + }, + { + "epoch": 7.394568690095847, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4629 + }, + { + "epoch": 7.396166134185304, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4630 + }, + { + "epoch": 7.397763578274761, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4631 + }, + { + "epoch": 7.399361022364217, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4632 + }, + { + "epoch": 7.400958466453674, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4633 + }, + { + "epoch": 7.402555910543131, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4634 + }, + { + "epoch": 7.404153354632588, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4635 + }, + { + "epoch": 7.405750798722044, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 4636 + }, + { + "epoch": 7.407348242811501, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 4637 + }, + { + "epoch": 7.4089456869009584, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4638 + }, + { + "epoch": 7.4105431309904155, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4639 + }, + { + "epoch": 7.412140575079873, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4640 + }, + { + "epoch": 7.413738019169329, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4641 + }, + { + "epoch": 7.415335463258786, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4642 + }, + { + "epoch": 7.416932907348243, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4643 + }, + { + "epoch": 7.4185303514377, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4644 + }, + { + "epoch": 7.420127795527157, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4645 + }, + { + "epoch": 7.421725239616613, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4646 + }, + { + "epoch": 7.42332268370607, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4647 + }, + { + "epoch": 7.424920127795527, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4648 + }, + { + "epoch": 7.426517571884984, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4649 + }, + { + "epoch": 7.428115015974441, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4650 + }, + { + "epoch": 7.4297124600638975, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4651 + }, + { + "epoch": 7.431309904153355, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4652 + }, + { + "epoch": 7.432907348242812, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 4653 + }, + { + "epoch": 7.434504792332269, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4654 + }, + { + "epoch": 7.436102236421725, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 4655 + }, + { + "epoch": 7.437699680511182, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4656 + }, + { + "epoch": 7.439297124600639, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4657 + }, + { + "epoch": 7.440894568690096, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4658 + }, + { + "epoch": 7.442492012779553, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4659 + }, + { + "epoch": 7.444089456869009, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4660 + }, + { + "epoch": 7.445686900958466, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4661 + }, + { + "epoch": 7.447284345047923, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4662 + }, + { + "epoch": 7.44888178913738, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4663 + }, + { + "epoch": 7.4504792332268375, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4664 + }, + { + "epoch": 7.452076677316294, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4665 + }, + { + "epoch": 7.453674121405751, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4666 + }, + { + "epoch": 7.455271565495208, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4667 + }, + { + "epoch": 7.456869009584665, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 4668 + }, + { + "epoch": 7.458466453674122, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 4669 + }, + { + "epoch": 7.460063897763578, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 4670 + }, + { + "epoch": 7.461661341853035, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 4671 + }, + { + "epoch": 7.463258785942492, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4672 + }, + { + "epoch": 7.464856230031949, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4673 + }, + { + "epoch": 7.466453674121405, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4674 + }, + { + "epoch": 7.468051118210862, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4675 + }, + { + "epoch": 7.4696485623003195, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4676 + }, + { + "epoch": 7.4712460063897765, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4677 + }, + { + "epoch": 7.472843450479234, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4678 + }, + { + "epoch": 7.47444089456869, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4679 + }, + { + "epoch": 7.476038338658147, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4680 + }, + { + "epoch": 7.477635782747604, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4681 + }, + { + "epoch": 7.479233226837061, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4682 + }, + { + "epoch": 7.480830670926517, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4683 + }, + { + "epoch": 7.482428115015974, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4684 + }, + { + "epoch": 7.484025559105431, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4685 + }, + { + "epoch": 7.485623003194888, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4686 + }, + { + "epoch": 7.487220447284345, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4687 + }, + { + "epoch": 7.488817891373802, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4688 + }, + { + "epoch": 7.4904153354632586, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4689 + }, + { + "epoch": 7.492012779552716, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4690 + }, + { + "epoch": 7.493610223642173, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4691 + }, + { + "epoch": 7.49520766773163, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 4692 + }, + { + "epoch": 7.496805111821086, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 4693 + }, + { + "epoch": 7.498402555910543, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 4694 + }, + { + "epoch": 7.5, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4695 + }, + { + "epoch": 7.501597444089457, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4696 + }, + { + "epoch": 7.503194888178914, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4697 + }, + { + "epoch": 7.50479233226837, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4698 + }, + { + "epoch": 7.506389776357827, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4699 + }, + { + "epoch": 7.507987220447284, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4700 + }, + { + "epoch": 7.5095846645367414, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4701 + }, + { + "epoch": 7.511182108626198, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 4702 + }, + { + "epoch": 7.512779552715655, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4703 + }, + { + "epoch": 7.514376996805112, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4704 + }, + { + "epoch": 7.515974440894569, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4705 + }, + { + "epoch": 7.517571884984026, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4706 + }, + { + "epoch": 7.519169329073483, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4707 + }, + { + "epoch": 7.520766773162939, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4708 + }, + { + "epoch": 7.522364217252396, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4709 + }, + { + "epoch": 7.523961661341853, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 4710 + }, + { + "epoch": 7.52555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 4711 + }, + { + "epoch": 7.527156549520766, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4712 + }, + { + "epoch": 7.5287539936102235, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4713 + }, + { + "epoch": 7.5303514376996805, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4714 + }, + { + "epoch": 7.531948881789138, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4715 + }, + { + "epoch": 7.533546325878595, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4716 + }, + { + "epoch": 7.535143769968051, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4717 + }, + { + "epoch": 7.536741214057508, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4718 + }, + { + "epoch": 7.538338658146965, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4719 + }, + { + "epoch": 7.539936102236422, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4720 + }, + { + "epoch": 7.541533546325878, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4721 + }, + { + "epoch": 7.543130990415335, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 4722 + }, + { + "epoch": 7.544728434504792, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4723 + }, + { + "epoch": 7.546325878594249, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4724 + }, + { + "epoch": 7.547923322683706, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4725 + }, + { + "epoch": 7.549520766773163, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4726 + }, + { + "epoch": 7.55111821086262, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4727 + }, + { + "epoch": 7.552715654952077, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 4728 + }, + { + "epoch": 7.554313099041534, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4729 + }, + { + "epoch": 7.555910543130991, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4730 + }, + { + "epoch": 7.557507987220447, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4731 + }, + { + "epoch": 7.559105431309904, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4732 + }, + { + "epoch": 7.560702875399361, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4733 + }, + { + "epoch": 7.562300319488818, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4734 + }, + { + "epoch": 7.563897763578275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4735 + }, + { + "epoch": 7.565495207667731, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 4736 + }, + { + "epoch": 7.567092651757188, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4737 + }, + { + "epoch": 7.568690095846645, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 4738 + }, + { + "epoch": 7.5702875399361025, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 4739 + }, + { + "epoch": 7.571884984025559, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 4740 + }, + { + "epoch": 7.573482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 4741 + }, + { + "epoch": 7.575079872204473, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4742 + }, + { + "epoch": 7.57667731629393, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 4743 + }, + { + "epoch": 7.578274760383387, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 4744 + }, + { + "epoch": 7.579872204472844, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 4745 + }, + { + "epoch": 7.5814696485623, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 4746 + }, + { + "epoch": 7.583067092651757, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4747 + }, + { + "epoch": 7.584664536741214, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4748 + }, + { + "epoch": 7.586261980830671, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4749 + }, + { + "epoch": 7.587859424920127, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4750 + }, + { + "epoch": 7.5894568690095845, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4751 + }, + { + "epoch": 7.5910543130990416, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4752 + }, + { + "epoch": 7.592651757188499, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4753 + }, + { + "epoch": 7.594249201277956, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4754 + }, + { + "epoch": 7.595846645367412, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4755 + }, + { + "epoch": 7.597444089456869, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4756 + }, + { + "epoch": 7.599041533546326, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4757 + }, + { + "epoch": 7.600638977635783, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 4758 + }, + { + "epoch": 7.602236421725239, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4759 + }, + { + "epoch": 7.603833865814696, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4760 + }, + { + "epoch": 7.605431309904153, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 4761 + }, + { + "epoch": 7.60702875399361, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4762 + }, + { + "epoch": 7.608626198083067, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4763 + }, + { + "epoch": 7.6102236421725244, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4764 + }, + { + "epoch": 7.611821086261981, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4765 + }, + { + "epoch": 7.613418530351438, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4766 + }, + { + "epoch": 7.615015974440895, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 4767 + }, + { + "epoch": 7.616613418530352, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4768 + }, + { + "epoch": 7.618210862619808, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4769 + }, + { + "epoch": 7.619808306709265, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4770 + }, + { + "epoch": 7.621405750798722, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4771 + }, + { + "epoch": 7.623003194888179, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 4772 + }, + { + "epoch": 7.624600638977636, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4773 + }, + { + "epoch": 7.626198083067092, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4774 + }, + { + "epoch": 7.627795527156549, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 4775 + }, + { + "epoch": 7.6293929712460065, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4776 + }, + { + "epoch": 7.6309904153354635, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4777 + }, + { + "epoch": 7.63258785942492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4778 + }, + { + "epoch": 7.634185303514377, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4779 + }, + { + "epoch": 7.635782747603834, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 4780 + }, + { + "epoch": 7.637380191693291, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4781 + }, + { + "epoch": 7.638977635782748, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 4782 + }, + { + "epoch": 7.640575079872205, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4783 + }, + { + "epoch": 7.642172523961661, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4784 + }, + { + "epoch": 7.643769968051118, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4785 + }, + { + "epoch": 7.645367412140575, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4786 + }, + { + "epoch": 7.646964856230032, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 4787 + }, + { + "epoch": 7.6485623003194885, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 4788 + }, + { + "epoch": 7.6501597444089455, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4789 + }, + { + "epoch": 7.651757188498403, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 4790 + }, + { + "epoch": 7.65335463258786, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 4791 + }, + { + "epoch": 7.654952076677317, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4792 + }, + { + "epoch": 7.656549520766773, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4793 + }, + { + "epoch": 7.65814696485623, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 4794 + }, + { + "epoch": 7.659744408945687, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4795 + }, + { + "epoch": 7.661341853035144, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4796 + }, + { + "epoch": 7.6629392971246, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4797 + }, + { + "epoch": 7.664536741214057, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4798 + }, + { + "epoch": 7.666134185303514, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4799 + }, + { + "epoch": 7.667731629392971, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4800 + }, + { + "epoch": 7.669329073482428, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4801 + }, + { + "epoch": 7.6709265175718855, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4802 + }, + { + "epoch": 7.672523961661342, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 4803 + }, + { + "epoch": 7.674121405750799, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 4804 + }, + { + "epoch": 7.675718849840256, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 4805 + }, + { + "epoch": 7.677316293929713, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 4806 + }, + { + "epoch": 7.678913738019169, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4807 + }, + { + "epoch": 7.680511182108626, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 4808 + }, + { + "epoch": 7.682108626198083, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 4809 + }, + { + "epoch": 7.68370607028754, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4810 + }, + { + "epoch": 7.685303514376997, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 4811 + }, + { + "epoch": 7.686900958466453, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4812 + }, + { + "epoch": 7.68849840255591, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4813 + }, + { + "epoch": 7.6900958466453675, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4814 + }, + { + "epoch": 7.6916932907348246, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4815 + }, + { + "epoch": 7.693290734824281, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 4816 + }, + { + "epoch": 7.694888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4817 + }, + { + "epoch": 7.696485623003195, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 4818 + }, + { + "epoch": 7.698083067092652, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 4819 + }, + { + "epoch": 7.699680511182109, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4820 + }, + { + "epoch": 7.701277955271565, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4821 + }, + { + "epoch": 7.702875399361022, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4822 + }, + { + "epoch": 7.704472843450479, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4823 + }, + { + "epoch": 7.706070287539936, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4824 + }, + { + "epoch": 7.707667731629393, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4825 + }, + { + "epoch": 7.7092651757188495, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4826 + }, + { + "epoch": 7.710862619808307, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4827 + }, + { + "epoch": 7.712460063897764, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4828 + }, + { + "epoch": 7.714057507987221, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 4829 + }, + { + "epoch": 7.715654952076678, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4830 + }, + { + "epoch": 7.717252396166134, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 4831 + }, + { + "epoch": 7.718849840255591, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4832 + }, + { + "epoch": 7.720447284345048, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4833 + }, + { + "epoch": 7.722044728434505, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4834 + }, + { + "epoch": 7.723642172523961, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4835 + }, + { + "epoch": 7.725239616613418, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4836 + }, + { + "epoch": 7.726837060702875, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4837 + }, + { + "epoch": 7.728434504792332, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 4838 + }, + { + "epoch": 7.7300319488817895, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4839 + }, + { + "epoch": 7.731629392971246, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 4840 + }, + { + "epoch": 7.733226837060703, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4841 + }, + { + "epoch": 7.73482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 4842 + }, + { + "epoch": 7.736421725239617, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4843 + }, + { + "epoch": 7.738019169329074, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4844 + }, + { + "epoch": 7.73961661341853, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 4845 + }, + { + "epoch": 7.741214057507987, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 4846 + }, + { + "epoch": 7.742811501597444, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4847 + }, + { + "epoch": 7.744408945686901, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4848 + }, + { + "epoch": 7.746006389776358, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 4849 + }, + { + "epoch": 7.747603833865814, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4850 + }, + { + "epoch": 7.7492012779552715, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 4851 + }, + { + "epoch": 7.7507987220447285, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 4852 + }, + { + "epoch": 7.752396166134186, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 4853 + }, + { + "epoch": 7.753993610223642, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4854 + }, + { + "epoch": 7.755591054313099, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4855 + }, + { + "epoch": 7.757188498402556, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4856 + }, + { + "epoch": 7.758785942492013, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 4857 + }, + { + "epoch": 7.76038338658147, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4858 + }, + { + "epoch": 7.761980830670926, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4859 + }, + { + "epoch": 7.763578274760383, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4860 + }, + { + "epoch": 7.76517571884984, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4861 + }, + { + "epoch": 7.766773162939297, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4862 + }, + { + "epoch": 7.768370607028754, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4863 + }, + { + "epoch": 7.7699680511182105, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4864 + }, + { + "epoch": 7.771565495207668, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 4865 + }, + { + "epoch": 7.773162939297125, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4866 + }, + { + "epoch": 7.774760383386582, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 4867 + }, + { + "epoch": 7.776357827476039, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4868 + }, + { + "epoch": 7.777955271565495, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 4869 + }, + { + "epoch": 7.779552715654952, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 4870 + }, + { + "epoch": 7.781150159744409, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 4871 + }, + { + "epoch": 7.782747603833866, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4872 + }, + { + "epoch": 7.784345047923322, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4873 + }, + { + "epoch": 7.785942492012779, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 4874 + }, + { + "epoch": 7.787539936102236, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 4875 + }, + { + "epoch": 7.789137380191693, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 4876 + }, + { + "epoch": 7.7907348242811505, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 4877 + }, + { + "epoch": 7.792332268370607, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4878 + }, + { + "epoch": 7.793929712460064, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4879 + }, + { + "epoch": 7.795527156549521, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 4880 + }, + { + "epoch": 7.797124600638978, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4881 + }, + { + "epoch": 7.798722044728435, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4882 + }, + { + "epoch": 7.800319488817891, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4883 + }, + { + "epoch": 7.801916932907348, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 4884 + }, + { + "epoch": 7.803514376996805, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 4885 + }, + { + "epoch": 7.805111821086262, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 4886 + }, + { + "epoch": 7.806709265175719, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4887 + }, + { + "epoch": 7.8083067092651754, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 4888 + }, + { + "epoch": 7.8099041533546325, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4889 + }, + { + "epoch": 7.81150159744409, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 4890 + }, + { + "epoch": 7.813099041533547, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4891 + }, + { + "epoch": 7.814696485623003, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 4892 + }, + { + "epoch": 7.81629392971246, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 4893 + }, + { + "epoch": 7.817891373801917, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 4894 + }, + { + "epoch": 7.819488817891374, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 4895 + }, + { + "epoch": 7.821086261980831, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4896 + }, + { + "epoch": 7.822683706070287, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 4897 + }, + { + "epoch": 7.824281150159744, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 4898 + }, + { + "epoch": 7.825878594249201, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4899 + }, + { + "epoch": 7.827476038338658, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4900 + }, + { + "epoch": 7.8290734824281145, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 4901 + }, + { + "epoch": 7.830670926517572, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 4902 + }, + { + "epoch": 7.832268370607029, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 4903 + }, + { + "epoch": 7.833865814696486, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4904 + }, + { + "epoch": 7.835463258785943, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 4905 + }, + { + "epoch": 7.8370607028754, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4906 + }, + { + "epoch": 7.838658146964856, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 4907 + }, + { + "epoch": 7.840255591054313, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 4908 + }, + { + "epoch": 7.84185303514377, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 4909 + }, + { + "epoch": 7.843450479233227, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 4910 + }, + { + "epoch": 7.845047923322683, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 4911 + }, + { + "epoch": 7.84664536741214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4912 + }, + { + "epoch": 7.848242811501597, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 4913 + }, + { + "epoch": 7.8498402555910545, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 4914 + }, + { + "epoch": 7.8514376996805115, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 4915 + }, + { + "epoch": 7.853035143769968, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 4916 + }, + { + "epoch": 7.854632587859425, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4917 + }, + { + "epoch": 7.856230031948882, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 4918 + }, + { + "epoch": 7.857827476038339, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 4919 + }, + { + "epoch": 7.859424920127795, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 4920 + }, + { + "epoch": 7.861022364217252, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4921 + }, + { + "epoch": 7.862619808306709, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 4922 + }, + { + "epoch": 7.864217252396166, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 4923 + }, + { + "epoch": 7.865814696485623, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 4924 + }, + { + "epoch": 7.86741214057508, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 4925 + }, + { + "epoch": 7.8690095846645365, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 4926 + }, + { + "epoch": 7.8706070287539935, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4927 + }, + { + "epoch": 7.872204472843451, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0784, + "step": 4928 + }, + { + "epoch": 7.873801916932908, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 4929 + }, + { + "epoch": 7.875399361022364, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 4930 + }, + { + "epoch": 7.876996805111821, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 4931 + }, + { + "epoch": 7.878594249201278, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 4932 + }, + { + "epoch": 7.880191693290735, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 4933 + }, + { + "epoch": 7.881789137380192, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4934 + }, + { + "epoch": 7.883386581469648, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 4935 + }, + { + "epoch": 7.884984025559105, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 4936 + }, + { + "epoch": 7.886581469648562, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 4937 + }, + { + "epoch": 7.888178913738019, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 4938 + }, + { + "epoch": 7.8897763578274756, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4939 + }, + { + "epoch": 7.891373801916933, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 4940 + }, + { + "epoch": 7.89297124600639, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4941 + }, + { + "epoch": 7.894568690095847, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 4942 + }, + { + "epoch": 7.896166134185304, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 4943 + }, + { + "epoch": 7.897763578274761, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 4944 + }, + { + "epoch": 7.899361022364217, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 4945 + }, + { + "epoch": 7.900958466453674, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4946 + }, + { + "epoch": 7.902555910543131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4947 + }, + { + "epoch": 7.904153354632588, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4948 + }, + { + "epoch": 7.905750798722044, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4949 + }, + { + "epoch": 7.907348242811501, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4950 + }, + { + "epoch": 7.9089456869009584, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 4951 + }, + { + "epoch": 7.9105431309904155, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 4952 + }, + { + "epoch": 7.912140575079873, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 4953 + }, + { + "epoch": 7.913738019169329, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 4954 + }, + { + "epoch": 7.915335463258786, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 4955 + }, + { + "epoch": 7.916932907348243, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 4956 + }, + { + "epoch": 7.9185303514377, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 4957 + }, + { + "epoch": 7.920127795527156, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4958 + }, + { + "epoch": 7.921725239616613, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 4959 + }, + { + "epoch": 7.92332268370607, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 4960 + }, + { + "epoch": 7.924920127795527, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 4961 + }, + { + "epoch": 7.926517571884984, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 4962 + }, + { + "epoch": 7.928115015974441, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 4963 + }, + { + "epoch": 7.9297124600638975, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 4964 + }, + { + "epoch": 7.931309904153355, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 4965 + }, + { + "epoch": 7.932907348242812, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 4966 + }, + { + "epoch": 7.934504792332269, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4967 + }, + { + "epoch": 7.936102236421725, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4968 + }, + { + "epoch": 7.937699680511182, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 4969 + }, + { + "epoch": 7.939297124600639, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 4970 + }, + { + "epoch": 7.940894568690096, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 4971 + }, + { + "epoch": 7.942492012779553, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 4972 + }, + { + "epoch": 7.944089456869009, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4973 + }, + { + "epoch": 7.945686900958466, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4974 + }, + { + "epoch": 7.947284345047923, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 4975 + }, + { + "epoch": 7.94888178913738, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 4976 + }, + { + "epoch": 7.950479233226837, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 4977 + }, + { + "epoch": 7.952076677316294, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4978 + }, + { + "epoch": 7.953674121405751, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 4979 + }, + { + "epoch": 7.955271565495208, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 4980 + }, + { + "epoch": 7.956869009584665, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 4981 + }, + { + "epoch": 7.958466453674122, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 4982 + }, + { + "epoch": 7.960063897763578, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4983 + }, + { + "epoch": 7.961661341853035, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 4984 + }, + { + "epoch": 7.963258785942492, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 4985 + }, + { + "epoch": 7.964856230031949, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 4986 + }, + { + "epoch": 7.966453674121405, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4987 + }, + { + "epoch": 7.968051118210862, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 4988 + }, + { + "epoch": 7.9696485623003195, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 4989 + }, + { + "epoch": 7.9712460063897765, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 4990 + }, + { + "epoch": 7.972843450479234, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 4991 + }, + { + "epoch": 7.97444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 4992 + }, + { + "epoch": 7.976038338658147, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 4993 + }, + { + "epoch": 7.977635782747604, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 4994 + }, + { + "epoch": 7.979233226837061, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 4995 + }, + { + "epoch": 7.980830670926517, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 4996 + }, + { + "epoch": 7.982428115015974, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 4997 + }, + { + "epoch": 7.984025559105431, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 4998 + }, + { + "epoch": 7.985623003194888, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 4999 + }, + { + "epoch": 7.987220447284345, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5000 + }, + { + "epoch": 7.988817891373802, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5001 + }, + { + "epoch": 7.9904153354632586, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5002 + }, + { + "epoch": 7.992012779552716, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5003 + }, + { + "epoch": 7.993610223642173, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5004 + }, + { + "epoch": 7.99520766773163, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5005 + }, + { + "epoch": 7.996805111821086, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 5006 + }, + { + "epoch": 7.998402555910543, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 5007 + }, + { + "epoch": 8.0, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5008 + }, + { + "epoch": 8.001597444089457, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5009 + }, + { + "epoch": 8.003194888178914, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5010 + }, + { + "epoch": 8.004792332268371, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5011 + }, + { + "epoch": 8.006389776357828, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5012 + }, + { + "epoch": 8.007987220447284, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5013 + }, + { + "epoch": 8.00958466453674, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5014 + }, + { + "epoch": 8.011182108626198, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5015 + }, + { + "epoch": 8.012779552715655, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 5016 + }, + { + "epoch": 8.014376996805112, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 5017 + }, + { + "epoch": 8.015974440894569, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5018 + }, + { + "epoch": 8.017571884984026, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 5019 + }, + { + "epoch": 8.019169329073483, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5020 + }, + { + "epoch": 8.02076677316294, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5021 + }, + { + "epoch": 8.022364217252397, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5022 + }, + { + "epoch": 8.023961661341852, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5023 + }, + { + "epoch": 8.02555910543131, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5024 + }, + { + "epoch": 8.027156549520766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 5025 + }, + { + "epoch": 8.028753993610223, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5026 + }, + { + "epoch": 8.03035143769968, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5027 + }, + { + "epoch": 8.031948881789138, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5028 + }, + { + "epoch": 8.033546325878595, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5029 + }, + { + "epoch": 8.035143769968052, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5030 + }, + { + "epoch": 8.036741214057509, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5031 + }, + { + "epoch": 8.038338658146964, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5032 + }, + { + "epoch": 8.039936102236421, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 5033 + }, + { + "epoch": 8.041533546325878, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5034 + }, + { + "epoch": 8.043130990415335, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5035 + }, + { + "epoch": 8.044728434504792, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5036 + }, + { + "epoch": 8.04632587859425, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 5037 + }, + { + "epoch": 8.047923322683706, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 5038 + }, + { + "epoch": 8.049520766773163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5039 + }, + { + "epoch": 8.05111821086262, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5040 + }, + { + "epoch": 8.052715654952078, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5041 + }, + { + "epoch": 8.054313099041533, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5042 + }, + { + "epoch": 8.05591054313099, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5043 + }, + { + "epoch": 8.057507987220447, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5044 + }, + { + "epoch": 8.059105431309904, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5045 + }, + { + "epoch": 8.060702875399361, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5046 + }, + { + "epoch": 8.062300319488818, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5047 + }, + { + "epoch": 8.063897763578275, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5048 + }, + { + "epoch": 8.065495207667732, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5049 + }, + { + "epoch": 8.06709265175719, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5050 + }, + { + "epoch": 8.068690095846645, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 5051 + }, + { + "epoch": 8.070287539936102, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5052 + }, + { + "epoch": 8.071884984025559, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5053 + }, + { + "epoch": 8.073482428115016, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5054 + }, + { + "epoch": 8.075079872204473, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5055 + }, + { + "epoch": 8.07667731629393, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5056 + }, + { + "epoch": 8.078274760383387, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 5057 + }, + { + "epoch": 8.079872204472844, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5058 + }, + { + "epoch": 8.081469648562301, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5059 + }, + { + "epoch": 8.083067092651758, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5060 + }, + { + "epoch": 8.084664536741213, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5061 + }, + { + "epoch": 8.08626198083067, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 5062 + }, + { + "epoch": 8.087859424920127, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5063 + }, + { + "epoch": 8.089456869009584, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5064 + }, + { + "epoch": 8.091054313099042, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5065 + }, + { + "epoch": 8.092651757188499, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5066 + }, + { + "epoch": 8.094249201277956, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5067 + }, + { + "epoch": 8.095846645367413, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5068 + }, + { + "epoch": 8.09744408945687, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5069 + }, + { + "epoch": 8.099041533546325, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5070 + }, + { + "epoch": 8.100638977635782, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5071 + }, + { + "epoch": 8.10223642172524, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5072 + }, + { + "epoch": 8.103833865814696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5073 + }, + { + "epoch": 8.105431309904153, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5074 + }, + { + "epoch": 8.10702875399361, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5075 + }, + { + "epoch": 8.108626198083067, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5076 + }, + { + "epoch": 8.110223642172524, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5077 + }, + { + "epoch": 8.111821086261982, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 5078 + }, + { + "epoch": 8.113418530351439, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5079 + }, + { + "epoch": 8.115015974440894, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5080 + }, + { + "epoch": 8.116613418530351, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5081 + }, + { + "epoch": 8.118210862619808, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 5082 + }, + { + "epoch": 8.119808306709265, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5083 + }, + { + "epoch": 8.121405750798722, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5084 + }, + { + "epoch": 8.12300319488818, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5085 + }, + { + "epoch": 8.124600638977636, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 5086 + }, + { + "epoch": 8.126198083067093, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5087 + }, + { + "epoch": 8.12779552715655, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 5088 + }, + { + "epoch": 8.129392971246006, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5089 + }, + { + "epoch": 8.130990415335463, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5090 + }, + { + "epoch": 8.13258785942492, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5091 + }, + { + "epoch": 8.134185303514377, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 5092 + }, + { + "epoch": 8.135782747603834, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5093 + }, + { + "epoch": 8.13738019169329, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5094 + }, + { + "epoch": 8.138977635782748, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5095 + }, + { + "epoch": 8.140575079872205, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5096 + }, + { + "epoch": 8.142172523961662, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5097 + }, + { + "epoch": 8.143769968051119, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5098 + }, + { + "epoch": 8.145367412140574, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5099 + }, + { + "epoch": 8.146964856230031, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5100 + }, + { + "epoch": 8.148562300319488, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5101 + }, + { + "epoch": 8.150159744408946, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 5102 + }, + { + "epoch": 8.151757188498403, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5103 + }, + { + "epoch": 8.15335463258786, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5104 + }, + { + "epoch": 8.154952076677317, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5105 + }, + { + "epoch": 8.156549520766774, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5106 + }, + { + "epoch": 8.15814696485623, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5107 + }, + { + "epoch": 8.159744408945686, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5108 + }, + { + "epoch": 8.161341853035143, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5109 + }, + { + "epoch": 8.1629392971246, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5110 + }, + { + "epoch": 8.164536741214057, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5111 + }, + { + "epoch": 8.166134185303514, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5112 + }, + { + "epoch": 8.167731629392971, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5113 + }, + { + "epoch": 8.169329073482428, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5114 + }, + { + "epoch": 8.170926517571885, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5115 + }, + { + "epoch": 8.172523961661343, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5116 + }, + { + "epoch": 8.1741214057508, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5117 + }, + { + "epoch": 8.175718849840255, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5118 + }, + { + "epoch": 8.177316293929712, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5119 + }, + { + "epoch": 8.178913738019169, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5120 + }, + { + "epoch": 8.180511182108626, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5121 + }, + { + "epoch": 8.182108626198083, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5122 + }, + { + "epoch": 8.18370607028754, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5123 + }, + { + "epoch": 8.185303514376997, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 5124 + }, + { + "epoch": 8.186900958466454, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5125 + }, + { + "epoch": 8.188498402555911, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5126 + }, + { + "epoch": 8.190095846645367, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5127 + }, + { + "epoch": 8.191693290734824, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5128 + }, + { + "epoch": 8.19329073482428, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5129 + }, + { + "epoch": 8.194888178913738, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5130 + }, + { + "epoch": 8.196485623003195, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5131 + }, + { + "epoch": 8.198083067092652, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5132 + }, + { + "epoch": 8.199680511182109, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5133 + }, + { + "epoch": 8.201277955271566, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5134 + }, + { + "epoch": 8.202875399361023, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5135 + }, + { + "epoch": 8.204472843450478, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5136 + }, + { + "epoch": 8.206070287539935, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5137 + }, + { + "epoch": 8.207667731629392, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5138 + }, + { + "epoch": 8.20926517571885, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 5139 + }, + { + "epoch": 8.210862619808307, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5140 + }, + { + "epoch": 8.212460063897764, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5141 + }, + { + "epoch": 8.21405750798722, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5142 + }, + { + "epoch": 8.215654952076678, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5143 + }, + { + "epoch": 8.217252396166135, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5144 + }, + { + "epoch": 8.218849840255592, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 5145 + }, + { + "epoch": 8.220447284345047, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 5146 + }, + { + "epoch": 8.222044728434504, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5147 + }, + { + "epoch": 8.223642172523961, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5148 + }, + { + "epoch": 8.225239616613418, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5149 + }, + { + "epoch": 8.226837060702875, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5150 + }, + { + "epoch": 8.228434504792332, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5151 + }, + { + "epoch": 8.23003194888179, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5152 + }, + { + "epoch": 8.231629392971247, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5153 + }, + { + "epoch": 8.233226837060704, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 5154 + }, + { + "epoch": 8.23482428115016, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5155 + }, + { + "epoch": 8.236421725239616, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5156 + }, + { + "epoch": 8.238019169329073, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5157 + }, + { + "epoch": 8.23961661341853, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5158 + }, + { + "epoch": 8.241214057507987, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5159 + }, + { + "epoch": 8.242811501597444, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5160 + }, + { + "epoch": 8.244408945686901, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5161 + }, + { + "epoch": 8.246006389776358, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5162 + }, + { + "epoch": 8.247603833865815, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5163 + }, + { + "epoch": 8.249201277955272, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5164 + }, + { + "epoch": 8.250798722044728, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5165 + }, + { + "epoch": 8.252396166134185, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5166 + }, + { + "epoch": 8.253993610223642, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5167 + }, + { + "epoch": 8.255591054313099, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 5168 + }, + { + "epoch": 8.257188498402556, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5169 + }, + { + "epoch": 8.258785942492013, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5170 + }, + { + "epoch": 8.26038338658147, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5171 + }, + { + "epoch": 8.261980830670927, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 5172 + }, + { + "epoch": 8.263578274760384, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5173 + }, + { + "epoch": 8.26517571884984, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5174 + }, + { + "epoch": 8.266773162939296, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5175 + }, + { + "epoch": 8.268370607028753, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5176 + }, + { + "epoch": 8.26996805111821, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5177 + }, + { + "epoch": 8.271565495207668, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5178 + }, + { + "epoch": 8.273162939297125, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5179 + }, + { + "epoch": 8.274760383386582, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5180 + }, + { + "epoch": 8.276357827476039, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5181 + }, + { + "epoch": 8.277955271565496, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5182 + }, + { + "epoch": 8.279552715654953, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5183 + }, + { + "epoch": 8.281150159744408, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5184 + }, + { + "epoch": 8.282747603833865, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5185 + }, + { + "epoch": 8.284345047923322, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5186 + }, + { + "epoch": 8.28594249201278, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5187 + }, + { + "epoch": 8.287539936102236, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5188 + }, + { + "epoch": 8.289137380191693, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5189 + }, + { + "epoch": 8.29073482428115, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 5190 + }, + { + "epoch": 8.292332268370608, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 5191 + }, + { + "epoch": 8.293929712460065, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5192 + }, + { + "epoch": 8.295527156549522, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5193 + }, + { + "epoch": 8.297124600638977, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5194 + }, + { + "epoch": 8.298722044728434, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 5195 + }, + { + "epoch": 8.300319488817891, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5196 + }, + { + "epoch": 8.301916932907348, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5197 + }, + { + "epoch": 8.303514376996805, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5198 + }, + { + "epoch": 8.305111821086262, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5199 + }, + { + "epoch": 8.30670926517572, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5200 + }, + { + "epoch": 8.308306709265176, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 5201 + }, + { + "epoch": 8.309904153354633, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5202 + }, + { + "epoch": 8.311501597444089, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5203 + }, + { + "epoch": 8.313099041533546, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5204 + }, + { + "epoch": 8.314696485623003, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5205 + }, + { + "epoch": 8.31629392971246, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5206 + }, + { + "epoch": 8.317891373801917, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 5207 + }, + { + "epoch": 8.319488817891374, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5208 + }, + { + "epoch": 8.321086261980831, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5209 + }, + { + "epoch": 8.322683706070288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5210 + }, + { + "epoch": 8.324281150159745, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5211 + }, + { + "epoch": 8.3258785942492, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5212 + }, + { + "epoch": 8.327476038338657, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5213 + }, + { + "epoch": 8.329073482428115, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 5214 + }, + { + "epoch": 8.330670926517572, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5215 + }, + { + "epoch": 8.332268370607029, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5216 + }, + { + "epoch": 8.333865814696486, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5217 + }, + { + "epoch": 8.335463258785943, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5218 + }, + { + "epoch": 8.3370607028754, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5219 + }, + { + "epoch": 8.338658146964857, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5220 + }, + { + "epoch": 8.340255591054314, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 5221 + }, + { + "epoch": 8.34185303514377, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5222 + }, + { + "epoch": 8.343450479233226, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5223 + }, + { + "epoch": 8.345047923322683, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5224 + }, + { + "epoch": 8.34664536741214, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5225 + }, + { + "epoch": 8.348242811501597, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5226 + }, + { + "epoch": 8.349840255591054, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5227 + }, + { + "epoch": 8.351437699680512, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 5228 + }, + { + "epoch": 8.353035143769969, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5229 + }, + { + "epoch": 8.354632587859426, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5230 + }, + { + "epoch": 8.356230031948883, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 5231 + }, + { + "epoch": 8.357827476038338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 5232 + }, + { + "epoch": 8.359424920127795, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5233 + }, + { + "epoch": 8.361022364217252, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5234 + }, + { + "epoch": 8.36261980830671, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5235 + }, + { + "epoch": 8.364217252396166, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5236 + }, + { + "epoch": 8.365814696485623, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5237 + }, + { + "epoch": 8.36741214057508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5238 + }, + { + "epoch": 8.369009584664537, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5239 + }, + { + "epoch": 8.370607028753994, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5240 + }, + { + "epoch": 8.37220447284345, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5241 + }, + { + "epoch": 8.373801916932907, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5242 + }, + { + "epoch": 8.375399361022364, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5243 + }, + { + "epoch": 8.37699680511182, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5244 + }, + { + "epoch": 8.378594249201278, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 5245 + }, + { + "epoch": 8.380191693290735, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5246 + }, + { + "epoch": 8.381789137380192, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5247 + }, + { + "epoch": 8.383386581469649, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5248 + }, + { + "epoch": 8.384984025559106, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5249 + }, + { + "epoch": 8.386581469648561, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5250 + }, + { + "epoch": 8.388178913738018, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5251 + }, + { + "epoch": 8.389776357827476, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5252 + }, + { + "epoch": 8.391373801916933, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5253 + }, + { + "epoch": 8.39297124600639, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5254 + }, + { + "epoch": 8.394568690095847, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5255 + }, + { + "epoch": 8.396166134185304, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5256 + }, + { + "epoch": 8.39776357827476, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5257 + }, + { + "epoch": 8.399361022364218, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5258 + }, + { + "epoch": 8.400958466453675, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5259 + }, + { + "epoch": 8.40255591054313, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5260 + }, + { + "epoch": 8.404153354632587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5261 + }, + { + "epoch": 8.405750798722044, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 5262 + }, + { + "epoch": 8.407348242811501, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 5263 + }, + { + "epoch": 8.408945686900958, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5264 + }, + { + "epoch": 8.410543130990416, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5265 + }, + { + "epoch": 8.412140575079873, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5266 + }, + { + "epoch": 8.41373801916933, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5267 + }, + { + "epoch": 8.415335463258787, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5268 + }, + { + "epoch": 8.416932907348242, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5269 + }, + { + "epoch": 8.418530351437699, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5270 + }, + { + "epoch": 8.420127795527156, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5271 + }, + { + "epoch": 8.421725239616613, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5272 + }, + { + "epoch": 8.42332268370607, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5273 + }, + { + "epoch": 8.424920127795527, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5274 + }, + { + "epoch": 8.426517571884984, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5275 + }, + { + "epoch": 8.428115015974441, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5276 + }, + { + "epoch": 8.429712460063898, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5277 + }, + { + "epoch": 8.431309904153355, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5278 + }, + { + "epoch": 8.43290734824281, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5279 + }, + { + "epoch": 8.434504792332268, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5280 + }, + { + "epoch": 8.436102236421725, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5281 + }, + { + "epoch": 8.437699680511182, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 5282 + }, + { + "epoch": 8.439297124600639, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5283 + }, + { + "epoch": 8.440894568690096, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5284 + }, + { + "epoch": 8.442492012779553, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5285 + }, + { + "epoch": 8.44408945686901, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5286 + }, + { + "epoch": 8.445686900958467, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5287 + }, + { + "epoch": 8.447284345047922, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5288 + }, + { + "epoch": 8.44888178913738, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 5289 + }, + { + "epoch": 8.450479233226837, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 5290 + }, + { + "epoch": 8.452076677316294, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5291 + }, + { + "epoch": 8.45367412140575, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5292 + }, + { + "epoch": 8.455271565495208, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5293 + }, + { + "epoch": 8.456869009584665, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5294 + }, + { + "epoch": 8.458466453674122, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5295 + }, + { + "epoch": 8.460063897763579, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5296 + }, + { + "epoch": 8.461661341853036, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5297 + }, + { + "epoch": 8.463258785942491, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5298 + }, + { + "epoch": 8.464856230031948, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5299 + }, + { + "epoch": 8.466453674121405, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5300 + }, + { + "epoch": 8.468051118210862, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 5301 + }, + { + "epoch": 8.46964856230032, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5302 + }, + { + "epoch": 8.471246006389777, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5303 + }, + { + "epoch": 8.472843450479234, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5304 + }, + { + "epoch": 8.47444089456869, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 5305 + }, + { + "epoch": 8.476038338658148, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5306 + }, + { + "epoch": 8.477635782747603, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5307 + }, + { + "epoch": 8.47923322683706, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5308 + }, + { + "epoch": 8.480830670926517, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 5309 + }, + { + "epoch": 8.482428115015974, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5310 + }, + { + "epoch": 8.484025559105431, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5311 + }, + { + "epoch": 8.485623003194888, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 5312 + }, + { + "epoch": 8.487220447284345, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5313 + }, + { + "epoch": 8.488817891373802, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 5314 + }, + { + "epoch": 8.49041533546326, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 5315 + }, + { + "epoch": 8.492012779552716, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5316 + }, + { + "epoch": 8.493610223642172, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5317 + }, + { + "epoch": 8.495207667731629, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5318 + }, + { + "epoch": 8.496805111821086, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5319 + }, + { + "epoch": 8.498402555910543, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5320 + }, + { + "epoch": 8.5, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 5321 + }, + { + "epoch": 8.501597444089457, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 5322 + }, + { + "epoch": 8.503194888178914, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5323 + }, + { + "epoch": 8.504792332268371, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5324 + }, + { + "epoch": 8.506389776357828, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5325 + }, + { + "epoch": 8.507987220447284, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5326 + }, + { + "epoch": 8.50958466453674, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5327 + }, + { + "epoch": 8.511182108626198, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5328 + }, + { + "epoch": 8.512779552715655, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5329 + }, + { + "epoch": 8.514376996805112, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5330 + }, + { + "epoch": 8.515974440894569, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5331 + }, + { + "epoch": 8.517571884984026, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5332 + }, + { + "epoch": 8.519169329073483, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5333 + }, + { + "epoch": 8.52076677316294, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5334 + }, + { + "epoch": 8.522364217252395, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5335 + }, + { + "epoch": 8.523961661341852, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5336 + }, + { + "epoch": 8.52555910543131, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5337 + }, + { + "epoch": 8.527156549520766, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5338 + }, + { + "epoch": 8.528753993610223, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5339 + }, + { + "epoch": 8.53035143769968, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 5340 + }, + { + "epoch": 8.531948881789138, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5341 + }, + { + "epoch": 8.533546325878595, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5342 + }, + { + "epoch": 8.535143769968052, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5343 + }, + { + "epoch": 8.536741214057509, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5344 + }, + { + "epoch": 8.538338658146966, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5345 + }, + { + "epoch": 8.539936102236421, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 5346 + }, + { + "epoch": 8.541533546325878, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5347 + }, + { + "epoch": 8.543130990415335, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5348 + }, + { + "epoch": 8.544728434504792, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 5349 + }, + { + "epoch": 8.54632587859425, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5350 + }, + { + "epoch": 8.547923322683706, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5351 + }, + { + "epoch": 8.549520766773163, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 5352 + }, + { + "epoch": 8.55111821086262, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5353 + }, + { + "epoch": 8.552715654952078, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 5354 + }, + { + "epoch": 8.554313099041533, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 5355 + }, + { + "epoch": 8.55591054313099, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 5356 + }, + { + "epoch": 8.557507987220447, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5357 + }, + { + "epoch": 8.559105431309904, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5358 + }, + { + "epoch": 8.560702875399361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 5359 + }, + { + "epoch": 8.562300319488818, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5360 + }, + { + "epoch": 8.563897763578275, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5361 + }, + { + "epoch": 8.565495207667732, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 5362 + }, + { + "epoch": 8.56709265175719, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 5363 + }, + { + "epoch": 8.568690095846645, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5364 + }, + { + "epoch": 8.570287539936102, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 5365 + }, + { + "epoch": 8.571884984025559, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5366 + }, + { + "epoch": 8.573482428115016, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5367 + }, + { + "epoch": 8.575079872204473, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5368 + }, + { + "epoch": 8.57667731629393, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5369 + }, + { + "epoch": 8.578274760383387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5370 + }, + { + "epoch": 8.579872204472844, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 5371 + }, + { + "epoch": 8.581469648562301, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5372 + }, + { + "epoch": 8.583067092651756, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5373 + }, + { + "epoch": 8.584664536741213, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5374 + }, + { + "epoch": 8.58626198083067, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5375 + }, + { + "epoch": 8.587859424920127, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5376 + }, + { + "epoch": 8.589456869009584, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5377 + }, + { + "epoch": 8.591054313099042, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5378 + }, + { + "epoch": 8.592651757188499, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5379 + }, + { + "epoch": 8.594249201277956, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5380 + }, + { + "epoch": 8.595846645367413, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5381 + }, + { + "epoch": 8.59744408945687, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5382 + }, + { + "epoch": 8.599041533546325, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5383 + }, + { + "epoch": 8.600638977635782, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5384 + }, + { + "epoch": 8.60223642172524, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5385 + }, + { + "epoch": 8.603833865814696, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5386 + }, + { + "epoch": 8.605431309904153, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5387 + }, + { + "epoch": 8.60702875399361, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5388 + }, + { + "epoch": 8.608626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 5389 + }, + { + "epoch": 8.610223642172524, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5390 + }, + { + "epoch": 8.611821086261982, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 5391 + }, + { + "epoch": 8.613418530351439, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5392 + }, + { + "epoch": 8.615015974440894, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5393 + }, + { + "epoch": 8.616613418530351, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 5394 + }, + { + "epoch": 8.618210862619808, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5395 + }, + { + "epoch": 8.619808306709265, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5396 + }, + { + "epoch": 8.621405750798722, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5397 + }, + { + "epoch": 8.62300319488818, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5398 + }, + { + "epoch": 8.624600638977636, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5399 + }, + { + "epoch": 8.626198083067093, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5400 + }, + { + "epoch": 8.62779552715655, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5401 + }, + { + "epoch": 8.629392971246006, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5402 + }, + { + "epoch": 8.630990415335463, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5403 + }, + { + "epoch": 8.63258785942492, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5404 + }, + { + "epoch": 8.634185303514377, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5405 + }, + { + "epoch": 8.635782747603834, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5406 + }, + { + "epoch": 8.63738019169329, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5407 + }, + { + "epoch": 8.638977635782748, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5408 + }, + { + "epoch": 8.640575079872205, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5409 + }, + { + "epoch": 8.642172523961662, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5410 + }, + { + "epoch": 8.643769968051117, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5411 + }, + { + "epoch": 8.645367412140574, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5412 + }, + { + "epoch": 8.646964856230031, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 5413 + }, + { + "epoch": 8.648562300319488, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5414 + }, + { + "epoch": 8.650159744408946, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5415 + }, + { + "epoch": 8.651757188498403, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5416 + }, + { + "epoch": 8.65335463258786, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5417 + }, + { + "epoch": 8.654952076677317, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5418 + }, + { + "epoch": 8.656549520766774, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5419 + }, + { + "epoch": 8.65814696485623, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5420 + }, + { + "epoch": 8.659744408945686, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 5421 + }, + { + "epoch": 8.661341853035143, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5422 + }, + { + "epoch": 8.6629392971246, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 5423 + }, + { + "epoch": 8.664536741214057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 5424 + }, + { + "epoch": 8.666134185303514, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5425 + }, + { + "epoch": 8.667731629392971, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5426 + }, + { + "epoch": 8.669329073482428, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5427 + }, + { + "epoch": 8.670926517571885, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5428 + }, + { + "epoch": 8.672523961661343, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5429 + }, + { + "epoch": 8.6741214057508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 5430 + }, + { + "epoch": 8.675718849840255, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5431 + }, + { + "epoch": 8.677316293929712, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5432 + }, + { + "epoch": 8.678913738019169, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5433 + }, + { + "epoch": 8.680511182108626, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 5434 + }, + { + "epoch": 8.682108626198083, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 5435 + }, + { + "epoch": 8.68370607028754, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5436 + }, + { + "epoch": 8.685303514376997, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5437 + }, + { + "epoch": 8.686900958466454, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5438 + }, + { + "epoch": 8.688498402555911, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5439 + }, + { + "epoch": 8.690095846645367, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5440 + }, + { + "epoch": 8.691693290734824, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5441 + }, + { + "epoch": 8.69329073482428, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5442 + }, + { + "epoch": 8.694888178913738, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5443 + }, + { + "epoch": 8.696485623003195, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5444 + }, + { + "epoch": 8.698083067092652, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5445 + }, + { + "epoch": 8.699680511182109, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5446 + }, + { + "epoch": 8.701277955271566, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5447 + }, + { + "epoch": 8.702875399361023, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5448 + }, + { + "epoch": 8.704472843450478, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5449 + }, + { + "epoch": 8.706070287539935, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5450 + }, + { + "epoch": 8.707667731629392, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5451 + }, + { + "epoch": 8.70926517571885, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5452 + }, + { + "epoch": 8.710862619808307, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5453 + }, + { + "epoch": 8.712460063897764, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5454 + }, + { + "epoch": 8.71405750798722, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5455 + }, + { + "epoch": 8.715654952076678, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5456 + }, + { + "epoch": 8.717252396166135, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5457 + }, + { + "epoch": 8.718849840255592, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5458 + }, + { + "epoch": 8.720447284345047, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5459 + }, + { + "epoch": 8.722044728434504, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 5460 + }, + { + "epoch": 8.723642172523961, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5461 + }, + { + "epoch": 8.725239616613418, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5462 + }, + { + "epoch": 8.726837060702875, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5463 + }, + { + "epoch": 8.728434504792332, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5464 + }, + { + "epoch": 8.73003194888179, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5465 + }, + { + "epoch": 8.731629392971247, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5466 + }, + { + "epoch": 8.733226837060704, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5467 + }, + { + "epoch": 8.73482428115016, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5468 + }, + { + "epoch": 8.736421725239616, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5469 + }, + { + "epoch": 8.738019169329073, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5470 + }, + { + "epoch": 8.73961661341853, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5471 + }, + { + "epoch": 8.741214057507987, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5472 + }, + { + "epoch": 8.742811501597444, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 5473 + }, + { + "epoch": 8.744408945686901, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5474 + }, + { + "epoch": 8.746006389776358, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5475 + }, + { + "epoch": 8.747603833865815, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 5476 + }, + { + "epoch": 8.749201277955272, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5477 + }, + { + "epoch": 8.750798722044728, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5478 + }, + { + "epoch": 8.752396166134185, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5479 + }, + { + "epoch": 8.753993610223642, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5480 + }, + { + "epoch": 8.755591054313099, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5481 + }, + { + "epoch": 8.757188498402556, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5482 + }, + { + "epoch": 8.758785942492013, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5483 + }, + { + "epoch": 8.76038338658147, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5484 + }, + { + "epoch": 8.761980830670927, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5485 + }, + { + "epoch": 8.763578274760384, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5486 + }, + { + "epoch": 8.76517571884984, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5487 + }, + { + "epoch": 8.766773162939296, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5488 + }, + { + "epoch": 8.768370607028753, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5489 + }, + { + "epoch": 8.76996805111821, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5490 + }, + { + "epoch": 8.771565495207668, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5491 + }, + { + "epoch": 8.773162939297125, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5492 + }, + { + "epoch": 8.774760383386582, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5493 + }, + { + "epoch": 8.776357827476039, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5494 + }, + { + "epoch": 8.777955271565496, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5495 + }, + { + "epoch": 8.779552715654953, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 5496 + }, + { + "epoch": 8.781150159744408, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5497 + }, + { + "epoch": 8.782747603833865, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5498 + }, + { + "epoch": 8.784345047923322, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5499 + }, + { + "epoch": 8.78594249201278, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5500 + }, + { + "epoch": 8.787539936102236, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 5501 + }, + { + "epoch": 8.789137380191693, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 5502 + }, + { + "epoch": 8.79073482428115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 5503 + }, + { + "epoch": 8.792332268370608, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5504 + }, + { + "epoch": 8.793929712460065, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5505 + }, + { + "epoch": 8.795527156549522, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5506 + }, + { + "epoch": 8.797124600638977, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5507 + }, + { + "epoch": 8.798722044728434, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5508 + }, + { + "epoch": 8.800319488817891, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5509 + }, + { + "epoch": 8.801916932907348, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5510 + }, + { + "epoch": 8.803514376996805, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5511 + }, + { + "epoch": 8.805111821086262, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5512 + }, + { + "epoch": 8.80670926517572, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 5513 + }, + { + "epoch": 8.808306709265176, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5514 + }, + { + "epoch": 8.809904153354633, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5515 + }, + { + "epoch": 8.811501597444089, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5516 + }, + { + "epoch": 8.813099041533546, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5517 + }, + { + "epoch": 8.814696485623003, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5518 + }, + { + "epoch": 8.81629392971246, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5519 + }, + { + "epoch": 8.817891373801917, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5520 + }, + { + "epoch": 8.819488817891374, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5521 + }, + { + "epoch": 8.821086261980831, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5522 + }, + { + "epoch": 8.822683706070288, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 5523 + }, + { + "epoch": 8.824281150159745, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 5524 + }, + { + "epoch": 8.8258785942492, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5525 + }, + { + "epoch": 8.827476038338657, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5526 + }, + { + "epoch": 8.829073482428115, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5527 + }, + { + "epoch": 8.830670926517572, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5528 + }, + { + "epoch": 8.832268370607029, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5529 + }, + { + "epoch": 8.833865814696486, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5530 + }, + { + "epoch": 8.835463258785943, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5531 + }, + { + "epoch": 8.8370607028754, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 5532 + }, + { + "epoch": 8.838658146964857, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5533 + }, + { + "epoch": 8.840255591054314, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5534 + }, + { + "epoch": 8.84185303514377, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5535 + }, + { + "epoch": 8.843450479233226, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5536 + }, + { + "epoch": 8.845047923322683, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5537 + }, + { + "epoch": 8.84664536741214, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5538 + }, + { + "epoch": 8.848242811501597, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5539 + }, + { + "epoch": 8.849840255591054, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5540 + }, + { + "epoch": 8.851437699680512, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5541 + }, + { + "epoch": 8.853035143769969, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5542 + }, + { + "epoch": 8.854632587859426, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5543 + }, + { + "epoch": 8.856230031948883, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5544 + }, + { + "epoch": 8.857827476038338, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5545 + }, + { + "epoch": 8.859424920127795, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5546 + }, + { + "epoch": 8.861022364217252, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 5547 + }, + { + "epoch": 8.86261980830671, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5548 + }, + { + "epoch": 8.864217252396166, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5549 + }, + { + "epoch": 8.865814696485623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5550 + }, + { + "epoch": 8.86741214057508, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5551 + }, + { + "epoch": 8.869009584664537, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5552 + }, + { + "epoch": 8.870607028753994, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5553 + }, + { + "epoch": 8.87220447284345, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5554 + }, + { + "epoch": 8.873801916932907, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5555 + }, + { + "epoch": 8.875399361022364, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5556 + }, + { + "epoch": 8.87699680511182, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5557 + }, + { + "epoch": 8.878594249201278, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5558 + }, + { + "epoch": 8.880191693290735, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5559 + }, + { + "epoch": 8.881789137380192, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 5560 + }, + { + "epoch": 8.883386581469649, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5561 + }, + { + "epoch": 8.884984025559106, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5562 + }, + { + "epoch": 8.886581469648561, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5563 + }, + { + "epoch": 8.888178913738018, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5564 + }, + { + "epoch": 8.889776357827476, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5565 + }, + { + "epoch": 8.891373801916933, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5566 + }, + { + "epoch": 8.89297124600639, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5567 + }, + { + "epoch": 8.894568690095847, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5568 + }, + { + "epoch": 8.896166134185304, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5569 + }, + { + "epoch": 8.89776357827476, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5570 + }, + { + "epoch": 8.899361022364218, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5571 + }, + { + "epoch": 8.900958466453675, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5572 + }, + { + "epoch": 8.90255591054313, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5573 + }, + { + "epoch": 8.904153354632587, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5574 + }, + { + "epoch": 8.905750798722044, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5575 + }, + { + "epoch": 8.907348242811501, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5576 + }, + { + "epoch": 8.908945686900958, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5577 + }, + { + "epoch": 8.910543130990416, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5578 + }, + { + "epoch": 8.912140575079873, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5579 + }, + { + "epoch": 8.91373801916933, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5580 + }, + { + "epoch": 8.915335463258787, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5581 + }, + { + "epoch": 8.916932907348244, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5582 + }, + { + "epoch": 8.918530351437699, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5583 + }, + { + "epoch": 8.920127795527156, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 5584 + }, + { + "epoch": 8.921725239616613, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5585 + }, + { + "epoch": 8.92332268370607, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5586 + }, + { + "epoch": 8.924920127795527, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5587 + }, + { + "epoch": 8.926517571884984, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5588 + }, + { + "epoch": 8.928115015974441, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5589 + }, + { + "epoch": 8.929712460063898, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5590 + }, + { + "epoch": 8.931309904153355, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5591 + }, + { + "epoch": 8.93290734824281, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 5592 + }, + { + "epoch": 8.934504792332268, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5593 + }, + { + "epoch": 8.936102236421725, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 5594 + }, + { + "epoch": 8.937699680511182, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5595 + }, + { + "epoch": 8.939297124600639, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5596 + }, + { + "epoch": 8.940894568690096, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5597 + }, + { + "epoch": 8.942492012779553, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5598 + }, + { + "epoch": 8.94408945686901, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 5599 + }, + { + "epoch": 8.945686900958467, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5600 + }, + { + "epoch": 8.947284345047922, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5601 + }, + { + "epoch": 8.94888178913738, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5602 + }, + { + "epoch": 8.950479233226837, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5603 + }, + { + "epoch": 8.952076677316294, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5604 + }, + { + "epoch": 8.95367412140575, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5605 + }, + { + "epoch": 8.955271565495208, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5606 + }, + { + "epoch": 8.956869009584665, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 5607 + }, + { + "epoch": 8.958466453674122, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5608 + }, + { + "epoch": 8.960063897763579, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5609 + }, + { + "epoch": 8.961661341853034, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 5610 + }, + { + "epoch": 8.963258785942491, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5611 + }, + { + "epoch": 8.964856230031948, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5612 + }, + { + "epoch": 8.966453674121405, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5613 + }, + { + "epoch": 8.968051118210862, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5614 + }, + { + "epoch": 8.96964856230032, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5615 + }, + { + "epoch": 8.971246006389777, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5616 + }, + { + "epoch": 8.972843450479234, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5617 + }, + { + "epoch": 8.97444089456869, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5618 + }, + { + "epoch": 8.976038338658148, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5619 + }, + { + "epoch": 8.977635782747605, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5620 + }, + { + "epoch": 8.97923322683706, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5621 + }, + { + "epoch": 8.980830670926517, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5622 + }, + { + "epoch": 8.982428115015974, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5623 + }, + { + "epoch": 8.984025559105431, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5624 + }, + { + "epoch": 8.985623003194888, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5625 + }, + { + "epoch": 8.987220447284345, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5626 + }, + { + "epoch": 8.988817891373802, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5627 + }, + { + "epoch": 8.99041533546326, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 5628 + }, + { + "epoch": 8.992012779552716, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5629 + }, + { + "epoch": 8.993610223642172, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5630 + }, + { + "epoch": 8.995207667731629, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 5631 + }, + { + "epoch": 8.996805111821086, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 5632 + }, + { + "epoch": 8.998402555910543, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5633 + }, + { + "epoch": 9.0, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 5634 + }, + { + "epoch": 9.001597444089457, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 5635 + }, + { + "epoch": 9.003194888178914, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5636 + }, + { + "epoch": 9.004792332268371, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5637 + }, + { + "epoch": 9.006389776357828, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5638 + }, + { + "epoch": 9.007987220447284, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5639 + }, + { + "epoch": 9.00958466453674, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5640 + }, + { + "epoch": 9.011182108626198, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5641 + }, + { + "epoch": 9.012779552715655, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5642 + }, + { + "epoch": 9.014376996805112, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 5643 + }, + { + "epoch": 9.015974440894569, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 5644 + }, + { + "epoch": 9.017571884984026, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 5645 + }, + { + "epoch": 9.019169329073483, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5646 + }, + { + "epoch": 9.02076677316294, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5647 + }, + { + "epoch": 9.022364217252397, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5648 + }, + { + "epoch": 9.023961661341852, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5649 + }, + { + "epoch": 9.02555910543131, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5650 + }, + { + "epoch": 9.027156549520766, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 5651 + }, + { + "epoch": 9.028753993610223, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 5652 + }, + { + "epoch": 9.03035143769968, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5653 + }, + { + "epoch": 9.031948881789138, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 5654 + }, + { + "epoch": 9.033546325878595, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5655 + }, + { + "epoch": 9.035143769968052, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5656 + }, + { + "epoch": 9.036741214057509, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5657 + }, + { + "epoch": 9.038338658146964, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 5658 + }, + { + "epoch": 9.039936102236421, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5659 + }, + { + "epoch": 9.041533546325878, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5660 + }, + { + "epoch": 9.043130990415335, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5661 + }, + { + "epoch": 9.044728434504792, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 5662 + }, + { + "epoch": 9.04632587859425, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5663 + }, + { + "epoch": 9.047923322683706, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5664 + }, + { + "epoch": 9.049520766773163, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5665 + }, + { + "epoch": 9.05111821086262, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5666 + }, + { + "epoch": 9.052715654952078, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5667 + }, + { + "epoch": 9.054313099041533, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5668 + }, + { + "epoch": 9.05591054313099, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5669 + }, + { + "epoch": 9.057507987220447, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5670 + }, + { + "epoch": 9.059105431309904, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5671 + }, + { + "epoch": 9.060702875399361, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 5672 + }, + { + "epoch": 9.062300319488818, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5673 + }, + { + "epoch": 9.063897763578275, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 5674 + }, + { + "epoch": 9.065495207667732, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 5675 + }, + { + "epoch": 9.06709265175719, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5676 + }, + { + "epoch": 9.068690095846645, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5677 + }, + { + "epoch": 9.070287539936102, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5678 + }, + { + "epoch": 9.071884984025559, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5679 + }, + { + "epoch": 9.073482428115016, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5680 + }, + { + "epoch": 9.075079872204473, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5681 + }, + { + "epoch": 9.07667731629393, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5682 + }, + { + "epoch": 9.078274760383387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5683 + }, + { + "epoch": 9.079872204472844, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 5684 + }, + { + "epoch": 9.081469648562301, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5685 + }, + { + "epoch": 9.083067092651758, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 5686 + }, + { + "epoch": 9.084664536741213, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5687 + }, + { + "epoch": 9.08626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 5688 + }, + { + "epoch": 9.087859424920127, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5689 + }, + { + "epoch": 9.089456869009584, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5690 + }, + { + "epoch": 9.091054313099042, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5691 + }, + { + "epoch": 9.092651757188499, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5692 + }, + { + "epoch": 9.094249201277956, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5693 + }, + { + "epoch": 9.095846645367413, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5694 + }, + { + "epoch": 9.09744408945687, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 5695 + }, + { + "epoch": 9.099041533546325, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 5696 + }, + { + "epoch": 9.100638977635782, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 5697 + }, + { + "epoch": 9.10223642172524, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5698 + }, + { + "epoch": 9.103833865814696, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5699 + }, + { + "epoch": 9.105431309904153, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5700 + }, + { + "epoch": 9.10702875399361, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 5701 + }, + { + "epoch": 9.108626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5702 + }, + { + "epoch": 9.110223642172524, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5703 + }, + { + "epoch": 9.111821086261982, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 5704 + }, + { + "epoch": 9.113418530351439, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 5705 + }, + { + "epoch": 9.115015974440894, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 5706 + }, + { + "epoch": 9.116613418530351, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5707 + }, + { + "epoch": 9.118210862619808, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 5708 + }, + { + "epoch": 9.119808306709265, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5709 + }, + { + "epoch": 9.121405750798722, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5710 + }, + { + "epoch": 9.12300319488818, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 5711 + }, + { + "epoch": 9.124600638977636, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5712 + }, + { + "epoch": 9.126198083067093, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5713 + }, + { + "epoch": 9.12779552715655, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5714 + }, + { + "epoch": 9.129392971246006, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5715 + }, + { + "epoch": 9.130990415335463, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5716 + }, + { + "epoch": 9.13258785942492, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5717 + }, + { + "epoch": 9.134185303514377, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 5718 + }, + { + "epoch": 9.135782747603834, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5719 + }, + { + "epoch": 9.13738019169329, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5720 + }, + { + "epoch": 9.138977635782748, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 5721 + }, + { + "epoch": 9.140575079872205, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5722 + }, + { + "epoch": 9.142172523961662, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5723 + }, + { + "epoch": 9.143769968051119, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5724 + }, + { + "epoch": 9.145367412140574, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5725 + }, + { + "epoch": 9.146964856230031, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5726 + }, + { + "epoch": 9.148562300319488, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5727 + }, + { + "epoch": 9.150159744408946, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5728 + }, + { + "epoch": 9.151757188498403, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5729 + }, + { + "epoch": 9.15335463258786, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 5730 + }, + { + "epoch": 9.154952076677317, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5731 + }, + { + "epoch": 9.156549520766774, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5732 + }, + { + "epoch": 9.15814696485623, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5733 + }, + { + "epoch": 9.159744408945686, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5734 + }, + { + "epoch": 9.161341853035143, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5735 + }, + { + "epoch": 9.1629392971246, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5736 + }, + { + "epoch": 9.164536741214057, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5737 + }, + { + "epoch": 9.166134185303514, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5738 + }, + { + "epoch": 9.167731629392971, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 5739 + }, + { + "epoch": 9.169329073482428, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5740 + }, + { + "epoch": 9.170926517571885, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 5741 + }, + { + "epoch": 9.172523961661343, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 5742 + }, + { + "epoch": 9.1741214057508, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5743 + }, + { + "epoch": 9.175718849840255, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5744 + }, + { + "epoch": 9.177316293929712, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5745 + }, + { + "epoch": 9.178913738019169, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5746 + }, + { + "epoch": 9.180511182108626, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5747 + }, + { + "epoch": 9.182108626198083, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5748 + }, + { + "epoch": 9.18370607028754, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 5749 + }, + { + "epoch": 9.185303514376997, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5750 + }, + { + "epoch": 9.186900958466454, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5751 + }, + { + "epoch": 9.188498402555911, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5752 + }, + { + "epoch": 9.190095846645367, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5753 + }, + { + "epoch": 9.191693290734824, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5754 + }, + { + "epoch": 9.19329073482428, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5755 + }, + { + "epoch": 9.194888178913738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5756 + }, + { + "epoch": 9.196485623003195, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 5757 + }, + { + "epoch": 9.198083067092652, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5758 + }, + { + "epoch": 9.199680511182109, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 5759 + }, + { + "epoch": 9.201277955271566, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 5760 + }, + { + "epoch": 9.202875399361023, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5761 + }, + { + "epoch": 9.204472843450478, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5762 + }, + { + "epoch": 9.206070287539935, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5763 + }, + { + "epoch": 9.207667731629392, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 5764 + }, + { + "epoch": 9.20926517571885, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5765 + }, + { + "epoch": 9.210862619808307, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 5766 + }, + { + "epoch": 9.212460063897764, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 5767 + }, + { + "epoch": 9.21405750798722, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 5768 + }, + { + "epoch": 9.215654952076678, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5769 + }, + { + "epoch": 9.217252396166135, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5770 + }, + { + "epoch": 9.218849840255592, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5771 + }, + { + "epoch": 9.220447284345047, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5772 + }, + { + "epoch": 9.222044728434504, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5773 + }, + { + "epoch": 9.223642172523961, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5774 + }, + { + "epoch": 9.225239616613418, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5775 + }, + { + "epoch": 9.226837060702875, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5776 + }, + { + "epoch": 9.228434504792332, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5777 + }, + { + "epoch": 9.23003194888179, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5778 + }, + { + "epoch": 9.231629392971247, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5779 + }, + { + "epoch": 9.233226837060704, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5780 + }, + { + "epoch": 9.23482428115016, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5781 + }, + { + "epoch": 9.236421725239616, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5782 + }, + { + "epoch": 9.238019169329073, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 5783 + }, + { + "epoch": 9.23961661341853, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5784 + }, + { + "epoch": 9.241214057507987, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5785 + }, + { + "epoch": 9.242811501597444, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5786 + }, + { + "epoch": 9.244408945686901, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5787 + }, + { + "epoch": 9.246006389776358, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5788 + }, + { + "epoch": 9.247603833865815, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5789 + }, + { + "epoch": 9.249201277955272, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5790 + }, + { + "epoch": 9.250798722044728, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 5791 + }, + { + "epoch": 9.252396166134185, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 5792 + }, + { + "epoch": 9.253993610223642, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 5793 + }, + { + "epoch": 9.255591054313099, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 5794 + }, + { + "epoch": 9.257188498402556, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5795 + }, + { + "epoch": 9.258785942492013, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5796 + }, + { + "epoch": 9.26038338658147, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5797 + }, + { + "epoch": 9.261980830670927, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 5798 + }, + { + "epoch": 9.263578274760384, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 5799 + }, + { + "epoch": 9.26517571884984, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5800 + }, + { + "epoch": 9.266773162939296, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5801 + }, + { + "epoch": 9.268370607028753, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 5802 + }, + { + "epoch": 9.26996805111821, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5803 + }, + { + "epoch": 9.271565495207668, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5804 + }, + { + "epoch": 9.273162939297125, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 5805 + }, + { + "epoch": 9.274760383386582, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5806 + }, + { + "epoch": 9.276357827476039, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5807 + }, + { + "epoch": 9.277955271565496, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 5808 + }, + { + "epoch": 9.279552715654953, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5809 + }, + { + "epoch": 9.281150159744408, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5810 + }, + { + "epoch": 9.282747603833865, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5811 + }, + { + "epoch": 9.284345047923322, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5812 + }, + { + "epoch": 9.28594249201278, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5813 + }, + { + "epoch": 9.287539936102236, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 5814 + }, + { + "epoch": 9.289137380191693, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5815 + }, + { + "epoch": 9.29073482428115, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5816 + }, + { + "epoch": 9.292332268370608, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 5817 + }, + { + "epoch": 9.293929712460065, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 5818 + }, + { + "epoch": 9.295527156549522, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5819 + }, + { + "epoch": 9.297124600638977, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 5820 + }, + { + "epoch": 9.298722044728434, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5821 + }, + { + "epoch": 9.300319488817891, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 5822 + }, + { + "epoch": 9.301916932907348, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5823 + }, + { + "epoch": 9.303514376996805, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 5824 + }, + { + "epoch": 9.305111821086262, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5825 + }, + { + "epoch": 9.30670926517572, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5826 + }, + { + "epoch": 9.308306709265176, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 5827 + }, + { + "epoch": 9.309904153354633, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5828 + }, + { + "epoch": 9.311501597444089, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5829 + }, + { + "epoch": 9.313099041533546, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 5830 + }, + { + "epoch": 9.314696485623003, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5831 + }, + { + "epoch": 9.31629392971246, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5832 + }, + { + "epoch": 9.317891373801917, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5833 + }, + { + "epoch": 9.319488817891374, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5834 + }, + { + "epoch": 9.321086261980831, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5835 + }, + { + "epoch": 9.322683706070288, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5836 + }, + { + "epoch": 9.324281150159745, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5837 + }, + { + "epoch": 9.3258785942492, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5838 + }, + { + "epoch": 9.327476038338657, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5839 + }, + { + "epoch": 9.329073482428115, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 5840 + }, + { + "epoch": 9.330670926517572, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5841 + }, + { + "epoch": 9.332268370607029, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5842 + }, + { + "epoch": 9.333865814696486, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 5843 + }, + { + "epoch": 9.335463258785943, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5844 + }, + { + "epoch": 9.3370607028754, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 5845 + }, + { + "epoch": 9.338658146964857, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5846 + }, + { + "epoch": 9.340255591054314, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 5847 + }, + { + "epoch": 9.34185303514377, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 5848 + }, + { + "epoch": 9.343450479233226, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 5849 + }, + { + "epoch": 9.345047923322683, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5850 + }, + { + "epoch": 9.34664536741214, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 5851 + }, + { + "epoch": 9.348242811501597, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5852 + }, + { + "epoch": 9.349840255591054, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 5853 + }, + { + "epoch": 9.351437699680512, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 5854 + }, + { + "epoch": 9.353035143769969, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5855 + }, + { + "epoch": 9.354632587859426, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5856 + }, + { + "epoch": 9.356230031948883, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5857 + }, + { + "epoch": 9.357827476038338, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5858 + }, + { + "epoch": 9.359424920127795, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 5859 + }, + { + "epoch": 9.361022364217252, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5860 + }, + { + "epoch": 9.36261980830671, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 5861 + }, + { + "epoch": 9.364217252396166, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5862 + }, + { + "epoch": 9.365814696485623, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5863 + }, + { + "epoch": 9.36741214057508, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 5864 + }, + { + "epoch": 9.369009584664537, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5865 + }, + { + "epoch": 9.370607028753994, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5866 + }, + { + "epoch": 9.37220447284345, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 5867 + }, + { + "epoch": 9.373801916932907, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 5868 + }, + { + "epoch": 9.375399361022364, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5869 + }, + { + "epoch": 9.37699680511182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5870 + }, + { + "epoch": 9.378594249201278, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5871 + }, + { + "epoch": 9.380191693290735, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 5872 + }, + { + "epoch": 9.381789137380192, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 5873 + }, + { + "epoch": 9.383386581469649, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5874 + }, + { + "epoch": 9.384984025559106, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5875 + }, + { + "epoch": 9.386581469648561, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 5876 + }, + { + "epoch": 9.388178913738018, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5877 + }, + { + "epoch": 9.389776357827476, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5878 + }, + { + "epoch": 9.391373801916933, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 5879 + }, + { + "epoch": 9.39297124600639, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5880 + }, + { + "epoch": 9.394568690095847, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5881 + }, + { + "epoch": 9.396166134185304, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 5882 + }, + { + "epoch": 9.39776357827476, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5883 + }, + { + "epoch": 9.399361022364218, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5884 + }, + { + "epoch": 9.400958466453675, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 5885 + }, + { + "epoch": 9.40255591054313, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5886 + }, + { + "epoch": 9.404153354632587, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5887 + }, + { + "epoch": 9.405750798722044, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 5888 + }, + { + "epoch": 9.407348242811501, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 5889 + }, + { + "epoch": 9.408945686900958, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5890 + }, + { + "epoch": 9.410543130990416, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5891 + }, + { + "epoch": 9.412140575079873, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 5892 + }, + { + "epoch": 9.41373801916933, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5893 + }, + { + "epoch": 9.415335463258787, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5894 + }, + { + "epoch": 9.416932907348242, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5895 + }, + { + "epoch": 9.418530351437699, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5896 + }, + { + "epoch": 9.420127795527156, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 5897 + }, + { + "epoch": 9.421725239616613, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 5898 + }, + { + "epoch": 9.42332268370607, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 5899 + }, + { + "epoch": 9.424920127795527, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 5900 + }, + { + "epoch": 9.426517571884984, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5901 + }, + { + "epoch": 9.428115015974441, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5902 + }, + { + "epoch": 9.429712460063898, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5903 + }, + { + "epoch": 9.431309904153355, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5904 + }, + { + "epoch": 9.43290734824281, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 5905 + }, + { + "epoch": 9.434504792332268, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 5906 + }, + { + "epoch": 9.436102236421725, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5907 + }, + { + "epoch": 9.437699680511182, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 5908 + }, + { + "epoch": 9.439297124600639, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5909 + }, + { + "epoch": 9.440894568690096, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 5910 + }, + { + "epoch": 9.442492012779553, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 5911 + }, + { + "epoch": 9.44408945686901, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5912 + }, + { + "epoch": 9.445686900958467, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5913 + }, + { + "epoch": 9.447284345047922, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 5914 + }, + { + "epoch": 9.44888178913738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5915 + }, + { + "epoch": 9.450479233226837, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 5916 + }, + { + "epoch": 9.452076677316294, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 5917 + }, + { + "epoch": 9.45367412140575, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5918 + }, + { + "epoch": 9.455271565495208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 5919 + }, + { + "epoch": 9.456869009584665, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5920 + }, + { + "epoch": 9.458466453674122, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 5921 + }, + { + "epoch": 9.460063897763579, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 5922 + }, + { + "epoch": 9.461661341853036, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5923 + }, + { + "epoch": 9.463258785942491, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5924 + }, + { + "epoch": 9.464856230031948, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 5925 + }, + { + "epoch": 9.466453674121405, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5926 + }, + { + "epoch": 9.468051118210862, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5927 + }, + { + "epoch": 9.46964856230032, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 5928 + }, + { + "epoch": 9.471246006389777, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5929 + }, + { + "epoch": 9.472843450479234, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 5930 + }, + { + "epoch": 9.47444089456869, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5931 + }, + { + "epoch": 9.476038338658148, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5932 + }, + { + "epoch": 9.477635782747603, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5933 + }, + { + "epoch": 9.47923322683706, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5934 + }, + { + "epoch": 9.480830670926517, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 5935 + }, + { + "epoch": 9.482428115015974, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 5936 + }, + { + "epoch": 9.484025559105431, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5937 + }, + { + "epoch": 9.485623003194888, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5938 + }, + { + "epoch": 9.487220447284345, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 5939 + }, + { + "epoch": 9.488817891373802, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 5940 + }, + { + "epoch": 9.49041533546326, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 5941 + }, + { + "epoch": 9.492012779552716, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5942 + }, + { + "epoch": 9.493610223642172, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5943 + }, + { + "epoch": 9.495207667731629, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 5944 + }, + { + "epoch": 9.496805111821086, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 5945 + }, + { + "epoch": 9.498402555910543, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5946 + }, + { + "epoch": 9.5, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 5947 + }, + { + "epoch": 9.501597444089457, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5948 + }, + { + "epoch": 9.503194888178914, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 5949 + }, + { + "epoch": 9.504792332268371, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 5950 + }, + { + "epoch": 9.506389776357828, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 5951 + }, + { + "epoch": 9.507987220447284, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5952 + }, + { + "epoch": 9.50958466453674, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 5953 + }, + { + "epoch": 9.511182108626198, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 5954 + }, + { + "epoch": 9.512779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5955 + }, + { + "epoch": 9.514376996805112, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 5956 + }, + { + "epoch": 9.515974440894569, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5957 + }, + { + "epoch": 9.517571884984026, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5958 + }, + { + "epoch": 9.519169329073483, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 5959 + }, + { + "epoch": 9.52076677316294, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 5960 + }, + { + "epoch": 9.522364217252395, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 5961 + }, + { + "epoch": 9.523961661341852, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5962 + }, + { + "epoch": 9.52555910543131, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5963 + }, + { + "epoch": 9.527156549520766, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 5964 + }, + { + "epoch": 9.528753993610223, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 5965 + }, + { + "epoch": 9.53035143769968, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5966 + }, + { + "epoch": 9.531948881789138, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 5967 + }, + { + "epoch": 9.533546325878595, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 5968 + }, + { + "epoch": 9.535143769968052, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 5969 + }, + { + "epoch": 9.536741214057509, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5970 + }, + { + "epoch": 9.538338658146966, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 5971 + }, + { + "epoch": 9.539936102236421, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 5972 + }, + { + "epoch": 9.541533546325878, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5973 + }, + { + "epoch": 9.543130990415335, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 5974 + }, + { + "epoch": 9.544728434504792, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 5975 + }, + { + "epoch": 9.54632587859425, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 5976 + }, + { + "epoch": 9.547923322683706, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 5977 + }, + { + "epoch": 9.549520766773163, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 5978 + }, + { + "epoch": 9.55111821086262, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 5979 + }, + { + "epoch": 9.552715654952078, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 5980 + }, + { + "epoch": 9.554313099041533, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5981 + }, + { + "epoch": 9.55591054313099, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 5982 + }, + { + "epoch": 9.557507987220447, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 5983 + }, + { + "epoch": 9.559105431309904, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 5984 + }, + { + "epoch": 9.560702875399361, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 5985 + }, + { + "epoch": 9.562300319488818, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 5986 + }, + { + "epoch": 9.563897763578275, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 5987 + }, + { + "epoch": 9.565495207667732, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 5988 + }, + { + "epoch": 9.56709265175719, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 5989 + }, + { + "epoch": 9.568690095846645, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 5990 + }, + { + "epoch": 9.570287539936102, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 5991 + }, + { + "epoch": 9.571884984025559, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 5992 + }, + { + "epoch": 9.573482428115016, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 5993 + }, + { + "epoch": 9.575079872204473, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 5994 + }, + { + "epoch": 9.57667731629393, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 5995 + }, + { + "epoch": 9.578274760383387, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 5996 + }, + { + "epoch": 9.579872204472844, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 5997 + }, + { + "epoch": 9.581469648562301, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 5998 + }, + { + "epoch": 9.583067092651756, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 5999 + }, + { + "epoch": 9.584664536741213, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6000 + }, + { + "epoch": 9.58626198083067, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 6001 + }, + { + "epoch": 9.587859424920127, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6002 + }, + { + "epoch": 9.589456869009584, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 6003 + }, + { + "epoch": 9.591054313099042, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6004 + }, + { + "epoch": 9.592651757188499, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6005 + }, + { + "epoch": 9.594249201277956, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6006 + }, + { + "epoch": 9.595846645367413, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6007 + }, + { + "epoch": 9.59744408945687, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6008 + }, + { + "epoch": 9.599041533546325, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6009 + }, + { + "epoch": 9.600638977635782, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6010 + }, + { + "epoch": 9.60223642172524, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6011 + }, + { + "epoch": 9.603833865814696, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6012 + }, + { + "epoch": 9.605431309904153, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6013 + }, + { + "epoch": 9.60702875399361, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6014 + }, + { + "epoch": 9.608626198083067, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6015 + }, + { + "epoch": 9.610223642172524, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 6016 + }, + { + "epoch": 9.611821086261982, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 6017 + }, + { + "epoch": 9.613418530351439, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6018 + }, + { + "epoch": 9.615015974440894, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6019 + }, + { + "epoch": 9.616613418530351, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6020 + }, + { + "epoch": 9.618210862619808, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6021 + }, + { + "epoch": 9.619808306709265, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 6022 + }, + { + "epoch": 9.621405750798722, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6023 + }, + { + "epoch": 9.62300319488818, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6024 + }, + { + "epoch": 9.624600638977636, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6025 + }, + { + "epoch": 9.626198083067093, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6026 + }, + { + "epoch": 9.62779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6027 + }, + { + "epoch": 9.629392971246006, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6028 + }, + { + "epoch": 9.630990415335463, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6029 + }, + { + "epoch": 9.63258785942492, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6030 + }, + { + "epoch": 9.634185303514377, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6031 + }, + { + "epoch": 9.635782747603834, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6032 + }, + { + "epoch": 9.63738019169329, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6033 + }, + { + "epoch": 9.638977635782748, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6034 + }, + { + "epoch": 9.640575079872205, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6035 + }, + { + "epoch": 9.642172523961662, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6036 + }, + { + "epoch": 9.643769968051117, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6037 + }, + { + "epoch": 9.645367412140574, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6038 + }, + { + "epoch": 9.646964856230031, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6039 + }, + { + "epoch": 9.648562300319488, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 6040 + }, + { + "epoch": 9.650159744408946, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 6041 + }, + { + "epoch": 9.651757188498403, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6042 + }, + { + "epoch": 9.65335463258786, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6043 + }, + { + "epoch": 9.654952076677317, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6044 + }, + { + "epoch": 9.656549520766774, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6045 + }, + { + "epoch": 9.65814696485623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 6046 + }, + { + "epoch": 9.659744408945686, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6047 + }, + { + "epoch": 9.661341853035143, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6048 + }, + { + "epoch": 9.6629392971246, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6049 + }, + { + "epoch": 9.664536741214057, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6050 + }, + { + "epoch": 9.666134185303514, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6051 + }, + { + "epoch": 9.667731629392971, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6052 + }, + { + "epoch": 9.669329073482428, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6053 + }, + { + "epoch": 9.670926517571885, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6054 + }, + { + "epoch": 9.672523961661343, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 6055 + }, + { + "epoch": 9.6741214057508, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6056 + }, + { + "epoch": 9.675718849840255, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6057 + }, + { + "epoch": 9.677316293929712, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 6058 + }, + { + "epoch": 9.678913738019169, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6059 + }, + { + "epoch": 9.680511182108626, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6060 + }, + { + "epoch": 9.682108626198083, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6061 + }, + { + "epoch": 9.68370607028754, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6062 + }, + { + "epoch": 9.685303514376997, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6063 + }, + { + "epoch": 9.686900958466454, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 6064 + }, + { + "epoch": 9.688498402555911, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6065 + }, + { + "epoch": 9.690095846645367, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6066 + }, + { + "epoch": 9.691693290734824, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6067 + }, + { + "epoch": 9.69329073482428, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 6068 + }, + { + "epoch": 9.694888178913738, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6069 + }, + { + "epoch": 9.696485623003195, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6070 + }, + { + "epoch": 9.698083067092652, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6071 + }, + { + "epoch": 9.699680511182109, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6072 + }, + { + "epoch": 9.701277955271566, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6073 + }, + { + "epoch": 9.702875399361023, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6074 + }, + { + "epoch": 9.704472843450478, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6075 + }, + { + "epoch": 9.706070287539935, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6076 + }, + { + "epoch": 9.707667731629392, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6077 + }, + { + "epoch": 9.70926517571885, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6078 + }, + { + "epoch": 9.710862619808307, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6079 + }, + { + "epoch": 9.712460063897764, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6080 + }, + { + "epoch": 9.71405750798722, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6081 + }, + { + "epoch": 9.715654952076678, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6082 + }, + { + "epoch": 9.717252396166135, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6083 + }, + { + "epoch": 9.718849840255592, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 6084 + }, + { + "epoch": 9.720447284345047, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 6085 + }, + { + "epoch": 9.722044728434504, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 6086 + }, + { + "epoch": 9.723642172523961, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6087 + }, + { + "epoch": 9.725239616613418, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6088 + }, + { + "epoch": 9.726837060702875, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6089 + }, + { + "epoch": 9.728434504792332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 6090 + }, + { + "epoch": 9.73003194888179, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6091 + }, + { + "epoch": 9.731629392971247, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6092 + }, + { + "epoch": 9.733226837060704, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6093 + }, + { + "epoch": 9.73482428115016, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6094 + }, + { + "epoch": 9.736421725239616, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6095 + }, + { + "epoch": 9.738019169329073, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6096 + }, + { + "epoch": 9.73961661341853, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6097 + }, + { + "epoch": 9.741214057507987, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 6098 + }, + { + "epoch": 9.742811501597444, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6099 + }, + { + "epoch": 9.744408945686901, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6100 + }, + { + "epoch": 9.746006389776358, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6101 + }, + { + "epoch": 9.747603833865815, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6102 + }, + { + "epoch": 9.749201277955272, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6103 + }, + { + "epoch": 9.750798722044728, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6104 + }, + { + "epoch": 9.752396166134185, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6105 + }, + { + "epoch": 9.753993610223642, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6106 + }, + { + "epoch": 9.755591054313099, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 6107 + }, + { + "epoch": 9.757188498402556, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6108 + }, + { + "epoch": 9.758785942492013, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6109 + }, + { + "epoch": 9.76038338658147, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6110 + }, + { + "epoch": 9.761980830670927, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6111 + }, + { + "epoch": 9.763578274760384, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6112 + }, + { + "epoch": 9.76517571884984, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6113 + }, + { + "epoch": 9.766773162939296, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6114 + }, + { + "epoch": 9.768370607028753, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6115 + }, + { + "epoch": 9.76996805111821, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6116 + }, + { + "epoch": 9.771565495207668, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6117 + }, + { + "epoch": 9.773162939297125, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6118 + }, + { + "epoch": 9.774760383386582, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6119 + }, + { + "epoch": 9.776357827476039, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6120 + }, + { + "epoch": 9.777955271565496, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6121 + }, + { + "epoch": 9.779552715654953, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6122 + }, + { + "epoch": 9.781150159744408, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6123 + }, + { + "epoch": 9.782747603833865, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6124 + }, + { + "epoch": 9.784345047923322, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6125 + }, + { + "epoch": 9.78594249201278, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6126 + }, + { + "epoch": 9.787539936102236, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6127 + }, + { + "epoch": 9.789137380191693, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6128 + }, + { + "epoch": 9.79073482428115, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6129 + }, + { + "epoch": 9.792332268370608, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6130 + }, + { + "epoch": 9.793929712460065, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6131 + }, + { + "epoch": 9.795527156549522, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6132 + }, + { + "epoch": 9.797124600638977, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6133 + }, + { + "epoch": 9.798722044728434, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6134 + }, + { + "epoch": 9.800319488817891, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6135 + }, + { + "epoch": 9.801916932907348, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6136 + }, + { + "epoch": 9.803514376996805, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6137 + }, + { + "epoch": 9.805111821086262, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6138 + }, + { + "epoch": 9.80670926517572, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6139 + }, + { + "epoch": 9.808306709265176, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6140 + }, + { + "epoch": 9.809904153354633, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6141 + }, + { + "epoch": 9.811501597444089, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6142 + }, + { + "epoch": 9.813099041533546, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6143 + }, + { + "epoch": 9.814696485623003, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 6144 + }, + { + "epoch": 9.81629392971246, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 6145 + }, + { + "epoch": 9.817891373801917, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6146 + }, + { + "epoch": 9.819488817891374, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6147 + }, + { + "epoch": 9.821086261980831, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6148 + }, + { + "epoch": 9.822683706070288, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 6149 + }, + { + "epoch": 9.824281150159745, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6150 + }, + { + "epoch": 9.8258785942492, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6151 + }, + { + "epoch": 9.827476038338657, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6152 + }, + { + "epoch": 9.829073482428115, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6153 + }, + { + "epoch": 9.830670926517572, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6154 + }, + { + "epoch": 9.832268370607029, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6155 + }, + { + "epoch": 9.833865814696486, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6156 + }, + { + "epoch": 9.835463258785943, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6157 + }, + { + "epoch": 9.8370607028754, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6158 + }, + { + "epoch": 9.838658146964857, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6159 + }, + { + "epoch": 9.840255591054314, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6160 + }, + { + "epoch": 9.84185303514377, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6161 + }, + { + "epoch": 9.843450479233226, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6162 + }, + { + "epoch": 9.845047923322683, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6163 + }, + { + "epoch": 9.84664536741214, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6164 + }, + { + "epoch": 9.848242811501597, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 6165 + }, + { + "epoch": 9.849840255591054, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 6166 + }, + { + "epoch": 9.851437699680512, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6167 + }, + { + "epoch": 9.853035143769969, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6168 + }, + { + "epoch": 9.854632587859426, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6169 + }, + { + "epoch": 9.856230031948883, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6170 + }, + { + "epoch": 9.857827476038338, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6171 + }, + { + "epoch": 9.859424920127795, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6172 + }, + { + "epoch": 9.861022364217252, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6173 + }, + { + "epoch": 9.86261980830671, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6174 + }, + { + "epoch": 9.864217252396166, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 6175 + }, + { + "epoch": 9.865814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6176 + }, + { + "epoch": 9.86741214057508, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6177 + }, + { + "epoch": 9.869009584664537, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6178 + }, + { + "epoch": 9.870607028753994, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6179 + }, + { + "epoch": 9.87220447284345, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6180 + }, + { + "epoch": 9.873801916932907, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6181 + }, + { + "epoch": 9.875399361022364, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6182 + }, + { + "epoch": 9.87699680511182, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6183 + }, + { + "epoch": 9.878594249201278, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 6184 + }, + { + "epoch": 9.880191693290735, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6185 + }, + { + "epoch": 9.881789137380192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6186 + }, + { + "epoch": 9.883386581469649, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6187 + }, + { + "epoch": 9.884984025559106, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6188 + }, + { + "epoch": 9.886581469648561, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6189 + }, + { + "epoch": 9.888178913738018, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6190 + }, + { + "epoch": 9.889776357827476, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6191 + }, + { + "epoch": 9.891373801916933, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6192 + }, + { + "epoch": 9.89297124600639, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6193 + }, + { + "epoch": 9.894568690095847, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 6194 + }, + { + "epoch": 9.896166134185304, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6195 + }, + { + "epoch": 9.89776357827476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6196 + }, + { + "epoch": 9.899361022364218, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6197 + }, + { + "epoch": 9.900958466453675, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6198 + }, + { + "epoch": 9.90255591054313, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6199 + }, + { + "epoch": 9.904153354632587, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6200 + }, + { + "epoch": 9.905750798722044, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6201 + }, + { + "epoch": 9.907348242811501, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6202 + }, + { + "epoch": 9.908945686900958, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6203 + }, + { + "epoch": 9.910543130990416, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6204 + }, + { + "epoch": 9.912140575079873, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 6205 + }, + { + "epoch": 9.91373801916933, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6206 + }, + { + "epoch": 9.915335463258787, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6207 + }, + { + "epoch": 9.916932907348244, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6208 + }, + { + "epoch": 9.918530351437699, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 6209 + }, + { + "epoch": 9.920127795527156, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6210 + }, + { + "epoch": 9.921725239616613, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 6211 + }, + { + "epoch": 9.92332268370607, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6212 + }, + { + "epoch": 9.924920127795527, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6213 + }, + { + "epoch": 9.926517571884984, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6214 + }, + { + "epoch": 9.928115015974441, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6215 + }, + { + "epoch": 9.929712460063898, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6216 + }, + { + "epoch": 9.931309904153355, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6217 + }, + { + "epoch": 9.93290734824281, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6218 + }, + { + "epoch": 9.934504792332268, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6219 + }, + { + "epoch": 9.936102236421725, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6220 + }, + { + "epoch": 9.937699680511182, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6221 + }, + { + "epoch": 9.939297124600639, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6222 + }, + { + "epoch": 9.940894568690096, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6223 + }, + { + "epoch": 9.942492012779553, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6224 + }, + { + "epoch": 9.94408945686901, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6225 + }, + { + "epoch": 9.945686900958467, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6226 + }, + { + "epoch": 9.947284345047922, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6227 + }, + { + "epoch": 9.94888178913738, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6228 + }, + { + "epoch": 9.950479233226837, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6229 + }, + { + "epoch": 9.952076677316294, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6230 + }, + { + "epoch": 9.95367412140575, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6231 + }, + { + "epoch": 9.955271565495208, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6232 + }, + { + "epoch": 9.956869009584665, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6233 + }, + { + "epoch": 9.958466453674122, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6234 + }, + { + "epoch": 9.960063897763579, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 6235 + }, + { + "epoch": 9.961661341853034, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 6236 + }, + { + "epoch": 9.963258785942491, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6237 + }, + { + "epoch": 9.964856230031948, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6238 + }, + { + "epoch": 9.966453674121405, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6239 + }, + { + "epoch": 9.968051118210862, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6240 + }, + { + "epoch": 9.96964856230032, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6241 + }, + { + "epoch": 9.971246006389777, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6242 + }, + { + "epoch": 9.972843450479234, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6243 + }, + { + "epoch": 9.97444089456869, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6244 + }, + { + "epoch": 9.976038338658148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6245 + }, + { + "epoch": 9.977635782747605, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6246 + }, + { + "epoch": 9.97923322683706, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6247 + }, + { + "epoch": 9.980830670926517, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6248 + }, + { + "epoch": 9.982428115015974, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6249 + }, + { + "epoch": 9.984025559105431, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6250 + }, + { + "epoch": 9.985623003194888, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 6251 + }, + { + "epoch": 9.987220447284345, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6252 + }, + { + "epoch": 9.988817891373802, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6253 + }, + { + "epoch": 9.99041533546326, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6254 + }, + { + "epoch": 9.992012779552716, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 6255 + }, + { + "epoch": 9.993610223642172, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6256 + }, + { + "epoch": 9.995207667731629, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 6257 + }, + { + "epoch": 9.996805111821086, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6258 + }, + { + "epoch": 9.998402555910543, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6259 + }, + { + "epoch": 10.0, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6260 + }, + { + "epoch": 10.001597444089457, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6261 + }, + { + "epoch": 10.003194888178914, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6262 + }, + { + "epoch": 10.004792332268371, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6263 + }, + { + "epoch": 10.006389776357828, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6264 + }, + { + "epoch": 10.007987220447284, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6265 + }, + { + "epoch": 10.00958466453674, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6266 + }, + { + "epoch": 10.011182108626198, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6267 + }, + { + "epoch": 10.012779552715655, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6268 + }, + { + "epoch": 10.014376996805112, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6269 + }, + { + "epoch": 10.015974440894569, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6270 + }, + { + "epoch": 10.017571884984026, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6271 + }, + { + "epoch": 10.019169329073483, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6272 + }, + { + "epoch": 10.02076677316294, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6273 + }, + { + "epoch": 10.022364217252397, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6274 + }, + { + "epoch": 10.023961661341852, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6275 + }, + { + "epoch": 10.02555910543131, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6276 + }, + { + "epoch": 10.027156549520766, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6277 + }, + { + "epoch": 10.028753993610223, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6278 + }, + { + "epoch": 10.03035143769968, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6279 + }, + { + "epoch": 10.031948881789138, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6280 + }, + { + "epoch": 10.033546325878595, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 6281 + }, + { + "epoch": 10.035143769968052, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6282 + }, + { + "epoch": 10.036741214057509, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6283 + }, + { + "epoch": 10.038338658146964, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6284 + }, + { + "epoch": 10.039936102236421, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6285 + }, + { + "epoch": 10.041533546325878, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6286 + }, + { + "epoch": 10.043130990415335, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 6287 + }, + { + "epoch": 10.044728434504792, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6288 + }, + { + "epoch": 10.04632587859425, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6289 + }, + { + "epoch": 10.047923322683706, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 6290 + }, + { + "epoch": 10.049520766773163, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6291 + }, + { + "epoch": 10.05111821086262, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6292 + }, + { + "epoch": 10.052715654952078, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 6293 + }, + { + "epoch": 10.054313099041533, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6294 + }, + { + "epoch": 10.05591054313099, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 6295 + }, + { + "epoch": 10.057507987220447, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6296 + }, + { + "epoch": 10.059105431309904, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6297 + }, + { + "epoch": 10.060702875399361, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 6298 + }, + { + "epoch": 10.062300319488818, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6299 + }, + { + "epoch": 10.063897763578275, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6300 + }, + { + "epoch": 10.065495207667732, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6301 + }, + { + "epoch": 10.06709265175719, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6302 + }, + { + "epoch": 10.068690095846645, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6303 + }, + { + "epoch": 10.070287539936102, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6304 + }, + { + "epoch": 10.071884984025559, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 6305 + }, + { + "epoch": 10.073482428115016, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6306 + }, + { + "epoch": 10.075079872204473, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6307 + }, + { + "epoch": 10.07667731629393, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6308 + }, + { + "epoch": 10.078274760383387, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6309 + }, + { + "epoch": 10.079872204472844, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6310 + }, + { + "epoch": 10.081469648562301, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6311 + }, + { + "epoch": 10.083067092651758, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6312 + }, + { + "epoch": 10.084664536741213, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6313 + }, + { + "epoch": 10.08626198083067, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6314 + }, + { + "epoch": 10.087859424920127, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6315 + }, + { + "epoch": 10.089456869009584, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6316 + }, + { + "epoch": 10.091054313099042, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6317 + }, + { + "epoch": 10.092651757188499, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6318 + }, + { + "epoch": 10.094249201277956, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 6319 + }, + { + "epoch": 10.095846645367413, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6320 + }, + { + "epoch": 10.09744408945687, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6321 + }, + { + "epoch": 10.099041533546325, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6322 + }, + { + "epoch": 10.100638977635782, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6323 + }, + { + "epoch": 10.10223642172524, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6324 + }, + { + "epoch": 10.103833865814696, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6325 + }, + { + "epoch": 10.105431309904153, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6326 + }, + { + "epoch": 10.10702875399361, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 6327 + }, + { + "epoch": 10.108626198083067, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6328 + }, + { + "epoch": 10.110223642172524, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6329 + }, + { + "epoch": 10.111821086261982, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6330 + }, + { + "epoch": 10.113418530351439, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6331 + }, + { + "epoch": 10.115015974440894, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6332 + }, + { + "epoch": 10.116613418530351, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 6333 + }, + { + "epoch": 10.118210862619808, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6334 + }, + { + "epoch": 10.119808306709265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 6335 + }, + { + "epoch": 10.121405750798722, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6336 + }, + { + "epoch": 10.12300319488818, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6337 + }, + { + "epoch": 10.124600638977636, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6338 + }, + { + "epoch": 10.126198083067093, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6339 + }, + { + "epoch": 10.12779552715655, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6340 + }, + { + "epoch": 10.129392971246006, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 6341 + }, + { + "epoch": 10.130990415335463, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6342 + }, + { + "epoch": 10.13258785942492, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6343 + }, + { + "epoch": 10.134185303514377, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6344 + }, + { + "epoch": 10.135782747603834, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6345 + }, + { + "epoch": 10.13738019169329, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6346 + }, + { + "epoch": 10.138977635782748, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6347 + }, + { + "epoch": 10.140575079872205, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6348 + }, + { + "epoch": 10.142172523961662, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6349 + }, + { + "epoch": 10.143769968051119, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6350 + }, + { + "epoch": 10.145367412140574, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6351 + }, + { + "epoch": 10.146964856230031, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6352 + }, + { + "epoch": 10.148562300319488, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6353 + }, + { + "epoch": 10.150159744408946, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6354 + }, + { + "epoch": 10.151757188498403, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6355 + }, + { + "epoch": 10.15335463258786, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6356 + }, + { + "epoch": 10.154952076677317, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6357 + }, + { + "epoch": 10.156549520766774, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6358 + }, + { + "epoch": 10.15814696485623, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6359 + }, + { + "epoch": 10.159744408945686, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6360 + }, + { + "epoch": 10.161341853035143, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6361 + }, + { + "epoch": 10.1629392971246, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6362 + }, + { + "epoch": 10.164536741214057, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6363 + }, + { + "epoch": 10.166134185303514, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6364 + }, + { + "epoch": 10.167731629392971, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6365 + }, + { + "epoch": 10.169329073482428, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6366 + }, + { + "epoch": 10.170926517571885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6367 + }, + { + "epoch": 10.172523961661343, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6368 + }, + { + "epoch": 10.1741214057508, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6369 + }, + { + "epoch": 10.175718849840255, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6370 + }, + { + "epoch": 10.177316293929712, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6371 + }, + { + "epoch": 10.178913738019169, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6372 + }, + { + "epoch": 10.180511182108626, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6373 + }, + { + "epoch": 10.182108626198083, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6374 + }, + { + "epoch": 10.18370607028754, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6375 + }, + { + "epoch": 10.185303514376997, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6376 + }, + { + "epoch": 10.186900958466454, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6377 + }, + { + "epoch": 10.188498402555911, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6378 + }, + { + "epoch": 10.190095846645367, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6379 + }, + { + "epoch": 10.191693290734824, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6380 + }, + { + "epoch": 10.19329073482428, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6381 + }, + { + "epoch": 10.194888178913738, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 6382 + }, + { + "epoch": 10.196485623003195, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6383 + }, + { + "epoch": 10.198083067092652, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6384 + }, + { + "epoch": 10.199680511182109, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6385 + }, + { + "epoch": 10.201277955271566, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6386 + }, + { + "epoch": 10.202875399361023, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6387 + }, + { + "epoch": 10.204472843450478, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6388 + }, + { + "epoch": 10.206070287539935, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6389 + }, + { + "epoch": 10.207667731629392, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6390 + }, + { + "epoch": 10.20926517571885, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6391 + }, + { + "epoch": 10.210862619808307, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 6392 + }, + { + "epoch": 10.212460063897764, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 6393 + }, + { + "epoch": 10.21405750798722, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6394 + }, + { + "epoch": 10.215654952076678, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6395 + }, + { + "epoch": 10.217252396166135, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6396 + }, + { + "epoch": 10.218849840255592, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6397 + }, + { + "epoch": 10.220447284345047, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 6398 + }, + { + "epoch": 10.222044728434504, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6399 + }, + { + "epoch": 10.223642172523961, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6400 + }, + { + "epoch": 10.225239616613418, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6401 + }, + { + "epoch": 10.226837060702875, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6402 + }, + { + "epoch": 10.228434504792332, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6403 + }, + { + "epoch": 10.23003194888179, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6404 + }, + { + "epoch": 10.231629392971247, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6405 + }, + { + "epoch": 10.233226837060704, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6406 + }, + { + "epoch": 10.23482428115016, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6407 + }, + { + "epoch": 10.236421725239616, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6408 + }, + { + "epoch": 10.238019169329073, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6409 + }, + { + "epoch": 10.23961661341853, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6410 + }, + { + "epoch": 10.241214057507987, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6411 + }, + { + "epoch": 10.242811501597444, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6412 + }, + { + "epoch": 10.244408945686901, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 6413 + }, + { + "epoch": 10.246006389776358, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6414 + }, + { + "epoch": 10.247603833865815, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6415 + }, + { + "epoch": 10.249201277955272, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6416 + }, + { + "epoch": 10.250798722044728, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6417 + }, + { + "epoch": 10.252396166134185, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6418 + }, + { + "epoch": 10.253993610223642, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6419 + }, + { + "epoch": 10.255591054313099, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6420 + }, + { + "epoch": 10.257188498402556, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6421 + }, + { + "epoch": 10.258785942492013, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6422 + }, + { + "epoch": 10.26038338658147, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6423 + }, + { + "epoch": 10.261980830670927, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6424 + }, + { + "epoch": 10.263578274760384, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6425 + }, + { + "epoch": 10.26517571884984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6426 + }, + { + "epoch": 10.266773162939296, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6427 + }, + { + "epoch": 10.268370607028753, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6428 + }, + { + "epoch": 10.26996805111821, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 6429 + }, + { + "epoch": 10.271565495207668, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6430 + }, + { + "epoch": 10.273162939297125, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6431 + }, + { + "epoch": 10.274760383386582, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6432 + }, + { + "epoch": 10.276357827476039, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 6433 + }, + { + "epoch": 10.277955271565496, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6434 + }, + { + "epoch": 10.279552715654953, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6435 + }, + { + "epoch": 10.281150159744408, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6436 + }, + { + "epoch": 10.282747603833865, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6437 + }, + { + "epoch": 10.284345047923322, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6438 + }, + { + "epoch": 10.28594249201278, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6439 + }, + { + "epoch": 10.287539936102236, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6440 + }, + { + "epoch": 10.289137380191693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6441 + }, + { + "epoch": 10.29073482428115, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6442 + }, + { + "epoch": 10.292332268370608, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6443 + }, + { + "epoch": 10.293929712460065, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6444 + }, + { + "epoch": 10.295527156549522, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6445 + }, + { + "epoch": 10.297124600638977, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 6446 + }, + { + "epoch": 10.298722044728434, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6447 + }, + { + "epoch": 10.300319488817891, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6448 + }, + { + "epoch": 10.301916932907348, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 6449 + }, + { + "epoch": 10.303514376996805, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6450 + }, + { + "epoch": 10.305111821086262, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6451 + }, + { + "epoch": 10.30670926517572, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6452 + }, + { + "epoch": 10.308306709265176, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6453 + }, + { + "epoch": 10.309904153354633, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6454 + }, + { + "epoch": 10.311501597444089, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6455 + }, + { + "epoch": 10.313099041533546, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6456 + }, + { + "epoch": 10.314696485623003, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6457 + }, + { + "epoch": 10.31629392971246, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6458 + }, + { + "epoch": 10.317891373801917, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6459 + }, + { + "epoch": 10.319488817891374, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6460 + }, + { + "epoch": 10.321086261980831, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6461 + }, + { + "epoch": 10.322683706070288, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6462 + }, + { + "epoch": 10.324281150159745, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6463 + }, + { + "epoch": 10.3258785942492, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 6464 + }, + { + "epoch": 10.327476038338657, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6465 + }, + { + "epoch": 10.329073482428115, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6466 + }, + { + "epoch": 10.330670926517572, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6467 + }, + { + "epoch": 10.332268370607029, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6468 + }, + { + "epoch": 10.333865814696486, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6469 + }, + { + "epoch": 10.335463258785943, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 6470 + }, + { + "epoch": 10.3370607028754, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6471 + }, + { + "epoch": 10.338658146964857, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6472 + }, + { + "epoch": 10.340255591054314, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6473 + }, + { + "epoch": 10.34185303514377, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6474 + }, + { + "epoch": 10.343450479233226, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 6475 + }, + { + "epoch": 10.345047923322683, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6476 + }, + { + "epoch": 10.34664536741214, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6477 + }, + { + "epoch": 10.348242811501597, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6478 + }, + { + "epoch": 10.349840255591054, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6479 + }, + { + "epoch": 10.351437699680512, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6480 + }, + { + "epoch": 10.353035143769969, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6481 + }, + { + "epoch": 10.354632587859426, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 6482 + }, + { + "epoch": 10.356230031948883, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6483 + }, + { + "epoch": 10.357827476038338, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6484 + }, + { + "epoch": 10.359424920127795, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6485 + }, + { + "epoch": 10.361022364217252, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6486 + }, + { + "epoch": 10.36261980830671, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6487 + }, + { + "epoch": 10.364217252396166, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6488 + }, + { + "epoch": 10.365814696485623, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6489 + }, + { + "epoch": 10.36741214057508, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6490 + }, + { + "epoch": 10.369009584664537, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 6491 + }, + { + "epoch": 10.370607028753994, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6492 + }, + { + "epoch": 10.37220447284345, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6493 + }, + { + "epoch": 10.373801916932907, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 6494 + }, + { + "epoch": 10.375399361022364, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 6495 + }, + { + "epoch": 10.37699680511182, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6496 + }, + { + "epoch": 10.378594249201278, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 6497 + }, + { + "epoch": 10.380191693290735, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6498 + }, + { + "epoch": 10.381789137380192, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6499 + }, + { + "epoch": 10.383386581469649, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 6500 + }, + { + "epoch": 10.384984025559106, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6501 + }, + { + "epoch": 10.386581469648561, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6502 + }, + { + "epoch": 10.388178913738018, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6503 + }, + { + "epoch": 10.389776357827476, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6504 + }, + { + "epoch": 10.391373801916933, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6505 + }, + { + "epoch": 10.39297124600639, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6506 + }, + { + "epoch": 10.394568690095847, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6507 + }, + { + "epoch": 10.396166134185304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6508 + }, + { + "epoch": 10.39776357827476, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6509 + }, + { + "epoch": 10.399361022364218, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6510 + }, + { + "epoch": 10.400958466453675, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 6511 + }, + { + "epoch": 10.40255591054313, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6512 + }, + { + "epoch": 10.404153354632587, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6513 + }, + { + "epoch": 10.405750798722044, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6514 + }, + { + "epoch": 10.407348242811501, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6515 + }, + { + "epoch": 10.408945686900958, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6516 + }, + { + "epoch": 10.410543130990416, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 6517 + }, + { + "epoch": 10.412140575079873, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6518 + }, + { + "epoch": 10.41373801916933, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6519 + }, + { + "epoch": 10.415335463258787, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6520 + }, + { + "epoch": 10.416932907348242, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6521 + }, + { + "epoch": 10.418530351437699, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6522 + }, + { + "epoch": 10.420127795527156, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6523 + }, + { + "epoch": 10.421725239616613, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6524 + }, + { + "epoch": 10.42332268370607, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6525 + }, + { + "epoch": 10.424920127795527, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 6526 + }, + { + "epoch": 10.426517571884984, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6527 + }, + { + "epoch": 10.428115015974441, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6528 + }, + { + "epoch": 10.429712460063898, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6529 + }, + { + "epoch": 10.431309904153355, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6530 + }, + { + "epoch": 10.43290734824281, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6531 + }, + { + "epoch": 10.434504792332268, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 6532 + }, + { + "epoch": 10.436102236421725, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6533 + }, + { + "epoch": 10.437699680511182, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6534 + }, + { + "epoch": 10.439297124600639, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6535 + }, + { + "epoch": 10.440894568690096, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6536 + }, + { + "epoch": 10.442492012779553, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6537 + }, + { + "epoch": 10.44408945686901, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6538 + }, + { + "epoch": 10.445686900958467, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6539 + }, + { + "epoch": 10.447284345047922, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6540 + }, + { + "epoch": 10.44888178913738, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6541 + }, + { + "epoch": 10.450479233226837, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6542 + }, + { + "epoch": 10.452076677316294, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6543 + }, + { + "epoch": 10.45367412140575, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6544 + }, + { + "epoch": 10.455271565495208, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 6545 + }, + { + "epoch": 10.456869009584665, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6546 + }, + { + "epoch": 10.458466453674122, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 6547 + }, + { + "epoch": 10.460063897763579, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6548 + }, + { + "epoch": 10.461661341853036, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6549 + }, + { + "epoch": 10.463258785942491, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6550 + }, + { + "epoch": 10.464856230031948, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6551 + }, + { + "epoch": 10.466453674121405, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6552 + }, + { + "epoch": 10.468051118210862, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 6553 + }, + { + "epoch": 10.46964856230032, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6554 + }, + { + "epoch": 10.471246006389777, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6555 + }, + { + "epoch": 10.472843450479234, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6556 + }, + { + "epoch": 10.47444089456869, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6557 + }, + { + "epoch": 10.476038338658148, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6558 + }, + { + "epoch": 10.477635782747603, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6559 + }, + { + "epoch": 10.47923322683706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6560 + }, + { + "epoch": 10.480830670926517, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6561 + }, + { + "epoch": 10.482428115015974, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6562 + }, + { + "epoch": 10.484025559105431, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 6563 + }, + { + "epoch": 10.485623003194888, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6564 + }, + { + "epoch": 10.487220447284345, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6565 + }, + { + "epoch": 10.488817891373802, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6566 + }, + { + "epoch": 10.49041533546326, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6567 + }, + { + "epoch": 10.492012779552716, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6568 + }, + { + "epoch": 10.493610223642172, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6569 + }, + { + "epoch": 10.495207667731629, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6570 + }, + { + "epoch": 10.496805111821086, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6571 + }, + { + "epoch": 10.498402555910543, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6572 + }, + { + "epoch": 10.5, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6573 + }, + { + "epoch": 10.501597444089457, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 6574 + }, + { + "epoch": 10.503194888178914, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6575 + }, + { + "epoch": 10.504792332268371, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6576 + }, + { + "epoch": 10.506389776357828, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6577 + }, + { + "epoch": 10.507987220447284, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6578 + }, + { + "epoch": 10.50958466453674, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6579 + }, + { + "epoch": 10.511182108626198, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6580 + }, + { + "epoch": 10.512779552715655, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6581 + }, + { + "epoch": 10.514376996805112, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6582 + }, + { + "epoch": 10.515974440894569, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 6583 + }, + { + "epoch": 10.517571884984026, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6584 + }, + { + "epoch": 10.519169329073483, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6585 + }, + { + "epoch": 10.52076677316294, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6586 + }, + { + "epoch": 10.522364217252395, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6587 + }, + { + "epoch": 10.523961661341852, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6588 + }, + { + "epoch": 10.52555910543131, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 6589 + }, + { + "epoch": 10.527156549520766, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6590 + }, + { + "epoch": 10.528753993610223, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6591 + }, + { + "epoch": 10.53035143769968, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6592 + }, + { + "epoch": 10.531948881789138, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6593 + }, + { + "epoch": 10.533546325878595, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6594 + }, + { + "epoch": 10.535143769968052, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6595 + }, + { + "epoch": 10.536741214057509, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6596 + }, + { + "epoch": 10.538338658146966, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6597 + }, + { + "epoch": 10.539936102236421, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6598 + }, + { + "epoch": 10.541533546325878, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6599 + }, + { + "epoch": 10.543130990415335, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6600 + }, + { + "epoch": 10.544728434504792, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6601 + }, + { + "epoch": 10.54632587859425, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6602 + }, + { + "epoch": 10.547923322683706, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6603 + }, + { + "epoch": 10.549520766773163, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 6604 + }, + { + "epoch": 10.55111821086262, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6605 + }, + { + "epoch": 10.552715654952078, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 6606 + }, + { + "epoch": 10.554313099041533, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6607 + }, + { + "epoch": 10.55591054313099, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6608 + }, + { + "epoch": 10.557507987220447, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 6609 + }, + { + "epoch": 10.559105431309904, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6610 + }, + { + "epoch": 10.560702875399361, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6611 + }, + { + "epoch": 10.562300319488818, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6612 + }, + { + "epoch": 10.563897763578275, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6613 + }, + { + "epoch": 10.565495207667732, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6614 + }, + { + "epoch": 10.56709265175719, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 6615 + }, + { + "epoch": 10.568690095846645, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 6616 + }, + { + "epoch": 10.570287539936102, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6617 + }, + { + "epoch": 10.571884984025559, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 6618 + }, + { + "epoch": 10.573482428115016, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 6619 + }, + { + "epoch": 10.575079872204473, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 6620 + }, + { + "epoch": 10.57667731629393, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6621 + }, + { + "epoch": 10.578274760383387, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 6622 + }, + { + "epoch": 10.579872204472844, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 6623 + }, + { + "epoch": 10.581469648562301, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6624 + }, + { + "epoch": 10.583067092651756, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6625 + }, + { + "epoch": 10.584664536741213, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 6626 + }, + { + "epoch": 10.58626198083067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6627 + }, + { + "epoch": 10.587859424920127, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 6628 + }, + { + "epoch": 10.589456869009584, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6629 + }, + { + "epoch": 10.591054313099042, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6630 + }, + { + "epoch": 10.592651757188499, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6631 + }, + { + "epoch": 10.594249201277956, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 6632 + }, + { + "epoch": 10.595846645367413, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6633 + }, + { + "epoch": 10.59744408945687, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6634 + }, + { + "epoch": 10.599041533546325, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6635 + }, + { + "epoch": 10.600638977635782, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6636 + }, + { + "epoch": 10.60223642172524, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6637 + }, + { + "epoch": 10.603833865814696, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6638 + }, + { + "epoch": 10.605431309904153, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6639 + }, + { + "epoch": 10.60702875399361, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6640 + }, + { + "epoch": 10.608626198083067, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6641 + }, + { + "epoch": 10.610223642172524, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6642 + }, + { + "epoch": 10.611821086261982, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6643 + }, + { + "epoch": 10.613418530351439, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6644 + }, + { + "epoch": 10.615015974440894, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6645 + }, + { + "epoch": 10.616613418530351, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 6646 + }, + { + "epoch": 10.618210862619808, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6647 + }, + { + "epoch": 10.619808306709265, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6648 + }, + { + "epoch": 10.621405750798722, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6649 + }, + { + "epoch": 10.62300319488818, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 6650 + }, + { + "epoch": 10.624600638977636, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 6651 + }, + { + "epoch": 10.626198083067093, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6652 + }, + { + "epoch": 10.62779552715655, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6653 + }, + { + "epoch": 10.629392971246006, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6654 + }, + { + "epoch": 10.630990415335463, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6655 + }, + { + "epoch": 10.63258785942492, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6656 + }, + { + "epoch": 10.634185303514377, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6657 + }, + { + "epoch": 10.635782747603834, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6658 + }, + { + "epoch": 10.63738019169329, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 6659 + }, + { + "epoch": 10.638977635782748, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6660 + }, + { + "epoch": 10.640575079872205, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6661 + }, + { + "epoch": 10.642172523961662, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 6662 + }, + { + "epoch": 10.643769968051117, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6663 + }, + { + "epoch": 10.645367412140574, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6664 + }, + { + "epoch": 10.646964856230031, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6665 + }, + { + "epoch": 10.648562300319488, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6666 + }, + { + "epoch": 10.650159744408946, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 6667 + }, + { + "epoch": 10.651757188498403, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6668 + }, + { + "epoch": 10.65335463258786, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 6669 + }, + { + "epoch": 10.654952076677317, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6670 + }, + { + "epoch": 10.656549520766774, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6671 + }, + { + "epoch": 10.65814696485623, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6672 + }, + { + "epoch": 10.659744408945686, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6673 + }, + { + "epoch": 10.661341853035143, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6674 + }, + { + "epoch": 10.6629392971246, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 6675 + }, + { + "epoch": 10.664536741214057, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6676 + }, + { + "epoch": 10.666134185303514, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6677 + }, + { + "epoch": 10.667731629392971, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6678 + }, + { + "epoch": 10.669329073482428, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6679 + }, + { + "epoch": 10.670926517571885, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 6680 + }, + { + "epoch": 10.672523961661343, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 6681 + }, + { + "epoch": 10.6741214057508, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6682 + }, + { + "epoch": 10.675718849840255, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6683 + }, + { + "epoch": 10.677316293929712, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6684 + }, + { + "epoch": 10.678913738019169, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6685 + }, + { + "epoch": 10.680511182108626, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6686 + }, + { + "epoch": 10.682108626198083, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6687 + }, + { + "epoch": 10.68370607028754, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6688 + }, + { + "epoch": 10.685303514376997, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6689 + }, + { + "epoch": 10.686900958466454, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6690 + }, + { + "epoch": 10.688498402555911, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6691 + }, + { + "epoch": 10.690095846645367, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6692 + }, + { + "epoch": 10.691693290734824, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6693 + }, + { + "epoch": 10.69329073482428, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6694 + }, + { + "epoch": 10.694888178913738, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6695 + }, + { + "epoch": 10.696485623003195, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6696 + }, + { + "epoch": 10.698083067092652, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6697 + }, + { + "epoch": 10.699680511182109, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6698 + }, + { + "epoch": 10.701277955271566, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 6699 + }, + { + "epoch": 10.702875399361023, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6700 + }, + { + "epoch": 10.704472843450478, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6701 + }, + { + "epoch": 10.706070287539935, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6702 + }, + { + "epoch": 10.707667731629392, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6703 + }, + { + "epoch": 10.70926517571885, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6704 + }, + { + "epoch": 10.710862619808307, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6705 + }, + { + "epoch": 10.712460063897764, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6706 + }, + { + "epoch": 10.71405750798722, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6707 + }, + { + "epoch": 10.715654952076678, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6708 + }, + { + "epoch": 10.717252396166135, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6709 + }, + { + "epoch": 10.718849840255592, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6710 + }, + { + "epoch": 10.720447284345047, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6711 + }, + { + "epoch": 10.722044728434504, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6712 + }, + { + "epoch": 10.723642172523961, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6713 + }, + { + "epoch": 10.725239616613418, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6714 + }, + { + "epoch": 10.726837060702875, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6715 + }, + { + "epoch": 10.728434504792332, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6716 + }, + { + "epoch": 10.73003194888179, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 6717 + }, + { + "epoch": 10.731629392971247, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6718 + }, + { + "epoch": 10.733226837060704, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6719 + }, + { + "epoch": 10.73482428115016, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6720 + }, + { + "epoch": 10.736421725239616, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6721 + }, + { + "epoch": 10.738019169329073, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6722 + }, + { + "epoch": 10.73961661341853, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 6723 + }, + { + "epoch": 10.741214057507987, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6724 + }, + { + "epoch": 10.742811501597444, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 6725 + }, + { + "epoch": 10.744408945686901, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6726 + }, + { + "epoch": 10.746006389776358, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6727 + }, + { + "epoch": 10.747603833865815, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6728 + }, + { + "epoch": 10.749201277955272, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6729 + }, + { + "epoch": 10.750798722044728, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6730 + }, + { + "epoch": 10.752396166134185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6731 + }, + { + "epoch": 10.753993610223642, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6732 + }, + { + "epoch": 10.755591054313099, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 6733 + }, + { + "epoch": 10.757188498402556, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6734 + }, + { + "epoch": 10.758785942492013, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6735 + }, + { + "epoch": 10.76038338658147, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6736 + }, + { + "epoch": 10.761980830670927, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6737 + }, + { + "epoch": 10.763578274760384, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6738 + }, + { + "epoch": 10.76517571884984, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 6739 + }, + { + "epoch": 10.766773162939296, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6740 + }, + { + "epoch": 10.768370607028753, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6741 + }, + { + "epoch": 10.76996805111821, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6742 + }, + { + "epoch": 10.771565495207668, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 6743 + }, + { + "epoch": 10.773162939297125, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6744 + }, + { + "epoch": 10.774760383386582, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 6745 + }, + { + "epoch": 10.776357827476039, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6746 + }, + { + "epoch": 10.777955271565496, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6747 + }, + { + "epoch": 10.779552715654953, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6748 + }, + { + "epoch": 10.781150159744408, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6749 + }, + { + "epoch": 10.782747603833865, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 6750 + }, + { + "epoch": 10.784345047923322, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6751 + }, + { + "epoch": 10.78594249201278, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 6752 + }, + { + "epoch": 10.787539936102236, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6753 + }, + { + "epoch": 10.789137380191693, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6754 + }, + { + "epoch": 10.79073482428115, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6755 + }, + { + "epoch": 10.792332268370608, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 6756 + }, + { + "epoch": 10.793929712460065, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 6757 + }, + { + "epoch": 10.795527156549522, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6758 + }, + { + "epoch": 10.797124600638977, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6759 + }, + { + "epoch": 10.798722044728434, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6760 + }, + { + "epoch": 10.800319488817891, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 6761 + }, + { + "epoch": 10.801916932907348, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 6762 + }, + { + "epoch": 10.803514376996805, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6763 + }, + { + "epoch": 10.805111821086262, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6764 + }, + { + "epoch": 10.80670926517572, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6765 + }, + { + "epoch": 10.808306709265176, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 6766 + }, + { + "epoch": 10.809904153354633, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6767 + }, + { + "epoch": 10.811501597444089, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6768 + }, + { + "epoch": 10.813099041533546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6769 + }, + { + "epoch": 10.814696485623003, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6770 + }, + { + "epoch": 10.81629392971246, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 6771 + }, + { + "epoch": 10.817891373801917, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6772 + }, + { + "epoch": 10.819488817891374, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6773 + }, + { + "epoch": 10.821086261980831, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6774 + }, + { + "epoch": 10.822683706070288, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6775 + }, + { + "epoch": 10.824281150159745, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6776 + }, + { + "epoch": 10.8258785942492, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 6777 + }, + { + "epoch": 10.827476038338657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6778 + }, + { + "epoch": 10.829073482428115, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6779 + }, + { + "epoch": 10.830670926517572, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 6780 + }, + { + "epoch": 10.832268370607029, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6781 + }, + { + "epoch": 10.833865814696486, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6782 + }, + { + "epoch": 10.835463258785943, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 6783 + }, + { + "epoch": 10.8370607028754, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6784 + }, + { + "epoch": 10.838658146964857, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6785 + }, + { + "epoch": 10.840255591054314, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6786 + }, + { + "epoch": 10.84185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6787 + }, + { + "epoch": 10.843450479233226, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6788 + }, + { + "epoch": 10.845047923322683, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6789 + }, + { + "epoch": 10.84664536741214, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 6790 + }, + { + "epoch": 10.848242811501597, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6791 + }, + { + "epoch": 10.849840255591054, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 6792 + }, + { + "epoch": 10.851437699680512, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6793 + }, + { + "epoch": 10.853035143769969, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6794 + }, + { + "epoch": 10.854632587859426, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 6795 + }, + { + "epoch": 10.856230031948883, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6796 + }, + { + "epoch": 10.857827476038338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6797 + }, + { + "epoch": 10.859424920127795, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 6798 + }, + { + "epoch": 10.861022364217252, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6799 + }, + { + "epoch": 10.86261980830671, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 6800 + }, + { + "epoch": 10.864217252396166, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 6801 + }, + { + "epoch": 10.865814696485623, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6802 + }, + { + "epoch": 10.86741214057508, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 6803 + }, + { + "epoch": 10.869009584664537, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6804 + }, + { + "epoch": 10.870607028753994, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 6805 + }, + { + "epoch": 10.87220447284345, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6806 + }, + { + "epoch": 10.873801916932907, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 6807 + }, + { + "epoch": 10.875399361022364, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6808 + }, + { + "epoch": 10.87699680511182, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 6809 + }, + { + "epoch": 10.878594249201278, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6810 + }, + { + "epoch": 10.880191693290735, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 6811 + }, + { + "epoch": 10.881789137380192, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6812 + }, + { + "epoch": 10.883386581469649, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6813 + }, + { + "epoch": 10.884984025559106, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6814 + }, + { + "epoch": 10.886581469648561, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6815 + }, + { + "epoch": 10.888178913738018, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6816 + }, + { + "epoch": 10.889776357827476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6817 + }, + { + "epoch": 10.891373801916933, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6818 + }, + { + "epoch": 10.89297124600639, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6819 + }, + { + "epoch": 10.894568690095847, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 6820 + }, + { + "epoch": 10.896166134185304, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 6821 + }, + { + "epoch": 10.89776357827476, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 6822 + }, + { + "epoch": 10.899361022364218, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6823 + }, + { + "epoch": 10.900958466453675, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 6824 + }, + { + "epoch": 10.90255591054313, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6825 + }, + { + "epoch": 10.904153354632587, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6826 + }, + { + "epoch": 10.905750798722044, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6827 + }, + { + "epoch": 10.907348242811501, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 6828 + }, + { + "epoch": 10.908945686900958, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 6829 + }, + { + "epoch": 10.910543130990416, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 6830 + }, + { + "epoch": 10.912140575079873, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6831 + }, + { + "epoch": 10.91373801916933, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 6832 + }, + { + "epoch": 10.915335463258787, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 6833 + }, + { + "epoch": 10.916932907348244, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6834 + }, + { + "epoch": 10.918530351437699, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6835 + }, + { + "epoch": 10.920127795527156, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6836 + }, + { + "epoch": 10.921725239616613, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 6837 + }, + { + "epoch": 10.92332268370607, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6838 + }, + { + "epoch": 10.924920127795527, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6839 + }, + { + "epoch": 10.926517571884984, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 6840 + }, + { + "epoch": 10.928115015974441, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6841 + }, + { + "epoch": 10.929712460063898, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 6842 + }, + { + "epoch": 10.931309904153355, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6843 + }, + { + "epoch": 10.93290734824281, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6844 + }, + { + "epoch": 10.934504792332268, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6845 + }, + { + "epoch": 10.936102236421725, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6846 + }, + { + "epoch": 10.937699680511182, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6847 + }, + { + "epoch": 10.939297124600639, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6848 + }, + { + "epoch": 10.940894568690096, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 6849 + }, + { + "epoch": 10.942492012779553, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 6850 + }, + { + "epoch": 10.94408945686901, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6851 + }, + { + "epoch": 10.945686900958467, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6852 + }, + { + "epoch": 10.947284345047922, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6853 + }, + { + "epoch": 10.94888178913738, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6854 + }, + { + "epoch": 10.950479233226837, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 6855 + }, + { + "epoch": 10.952076677316294, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6856 + }, + { + "epoch": 10.95367412140575, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6857 + }, + { + "epoch": 10.955271565495208, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6858 + }, + { + "epoch": 10.956869009584665, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6859 + }, + { + "epoch": 10.958466453674122, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6860 + }, + { + "epoch": 10.960063897763579, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 6861 + }, + { + "epoch": 10.961661341853034, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6862 + }, + { + "epoch": 10.963258785942491, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6863 + }, + { + "epoch": 10.964856230031948, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 6864 + }, + { + "epoch": 10.966453674121405, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6865 + }, + { + "epoch": 10.968051118210862, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6866 + }, + { + "epoch": 10.96964856230032, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 6867 + }, + { + "epoch": 10.971246006389777, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6868 + }, + { + "epoch": 10.972843450479234, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 6869 + }, + { + "epoch": 10.97444089456869, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6870 + }, + { + "epoch": 10.976038338658148, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6871 + }, + { + "epoch": 10.977635782747605, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6872 + }, + { + "epoch": 10.97923322683706, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6873 + }, + { + "epoch": 10.980830670926517, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 6874 + }, + { + "epoch": 10.982428115015974, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 6875 + }, + { + "epoch": 10.984025559105431, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6876 + }, + { + "epoch": 10.985623003194888, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6877 + }, + { + "epoch": 10.987220447284345, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 6878 + }, + { + "epoch": 10.988817891373802, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6879 + }, + { + "epoch": 10.99041533546326, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 6880 + }, + { + "epoch": 10.992012779552716, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6881 + }, + { + "epoch": 10.993610223642172, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6882 + }, + { + "epoch": 10.995207667731629, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6883 + }, + { + "epoch": 10.996805111821086, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6884 + }, + { + "epoch": 10.998402555910543, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 6885 + }, + { + "epoch": 11.0, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6886 + }, + { + "epoch": 11.001597444089457, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6887 + }, + { + "epoch": 11.003194888178914, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6888 + }, + { + "epoch": 11.004792332268371, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 6889 + }, + { + "epoch": 11.006389776357828, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 6890 + }, + { + "epoch": 11.007987220447284, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 6891 + }, + { + "epoch": 11.00958466453674, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6892 + }, + { + "epoch": 11.011182108626198, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 6893 + }, + { + "epoch": 11.012779552715655, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 6894 + }, + { + "epoch": 11.014376996805112, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 6895 + }, + { + "epoch": 11.015974440894569, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 6896 + }, + { + "epoch": 11.017571884984026, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 6897 + }, + { + "epoch": 11.019169329073483, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6898 + }, + { + "epoch": 11.02076677316294, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 6899 + }, + { + "epoch": 11.022364217252397, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 6900 + }, + { + "epoch": 11.023961661341852, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 6901 + }, + { + "epoch": 11.02555910543131, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6902 + }, + { + "epoch": 11.027156549520766, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 6903 + }, + { + "epoch": 11.028753993610223, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 6904 + }, + { + "epoch": 11.03035143769968, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6905 + }, + { + "epoch": 11.031948881789138, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 6906 + }, + { + "epoch": 11.033546325878595, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6907 + }, + { + "epoch": 11.035143769968052, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6908 + }, + { + "epoch": 11.036741214057509, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 6909 + }, + { + "epoch": 11.038338658146964, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6910 + }, + { + "epoch": 11.039936102236421, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 6911 + }, + { + "epoch": 11.041533546325878, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 6912 + }, + { + "epoch": 11.043130990415335, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 6913 + }, + { + "epoch": 11.044728434504792, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6914 + }, + { + "epoch": 11.04632587859425, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 6915 + }, + { + "epoch": 11.047923322683706, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 6916 + }, + { + "epoch": 11.049520766773163, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 6917 + }, + { + "epoch": 11.05111821086262, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 6918 + }, + { + "epoch": 11.052715654952078, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6919 + }, + { + "epoch": 11.054313099041533, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6920 + }, + { + "epoch": 11.05591054313099, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 6921 + }, + { + "epoch": 11.057507987220447, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 6922 + }, + { + "epoch": 11.059105431309904, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6923 + }, + { + "epoch": 11.060702875399361, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6924 + }, + { + "epoch": 11.062300319488818, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6925 + }, + { + "epoch": 11.063897763578275, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6926 + }, + { + "epoch": 11.065495207667732, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 6927 + }, + { + "epoch": 11.06709265175719, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6928 + }, + { + "epoch": 11.068690095846645, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 6929 + }, + { + "epoch": 11.070287539936102, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6930 + }, + { + "epoch": 11.071884984025559, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6931 + }, + { + "epoch": 11.073482428115016, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6932 + }, + { + "epoch": 11.075079872204473, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6933 + }, + { + "epoch": 11.07667731629393, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6934 + }, + { + "epoch": 11.078274760383387, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6935 + }, + { + "epoch": 11.079872204472844, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6936 + }, + { + "epoch": 11.081469648562301, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 6937 + }, + { + "epoch": 11.083067092651758, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6938 + }, + { + "epoch": 11.084664536741213, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 6939 + }, + { + "epoch": 11.08626198083067, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6940 + }, + { + "epoch": 11.087859424920127, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 6941 + }, + { + "epoch": 11.089456869009584, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 6942 + }, + { + "epoch": 11.091054313099042, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 6943 + }, + { + "epoch": 11.092651757188499, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 6944 + }, + { + "epoch": 11.094249201277956, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6945 + }, + { + "epoch": 11.095846645367413, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6946 + }, + { + "epoch": 11.09744408945687, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 6947 + }, + { + "epoch": 11.099041533546325, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 6948 + }, + { + "epoch": 11.100638977635782, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 6949 + }, + { + "epoch": 11.10223642172524, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 6950 + }, + { + "epoch": 11.103833865814696, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6951 + }, + { + "epoch": 11.105431309904153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6952 + }, + { + "epoch": 11.10702875399361, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 6953 + }, + { + "epoch": 11.108626198083067, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6954 + }, + { + "epoch": 11.110223642172524, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 6955 + }, + { + "epoch": 11.111821086261982, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6956 + }, + { + "epoch": 11.113418530351439, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 6957 + }, + { + "epoch": 11.115015974440894, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 6958 + }, + { + "epoch": 11.116613418530351, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 6959 + }, + { + "epoch": 11.118210862619808, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6960 + }, + { + "epoch": 11.119808306709265, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 6961 + }, + { + "epoch": 11.121405750798722, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 6962 + }, + { + "epoch": 11.12300319488818, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 6963 + }, + { + "epoch": 11.124600638977636, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 6964 + }, + { + "epoch": 11.126198083067093, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 6965 + }, + { + "epoch": 11.12779552715655, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 6966 + }, + { + "epoch": 11.129392971246006, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 6967 + }, + { + "epoch": 11.130990415335463, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 6968 + }, + { + "epoch": 11.13258785942492, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 6969 + }, + { + "epoch": 11.134185303514377, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6970 + }, + { + "epoch": 11.135782747603834, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 6971 + }, + { + "epoch": 11.13738019169329, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 6972 + }, + { + "epoch": 11.138977635782748, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 6973 + }, + { + "epoch": 11.140575079872205, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 6974 + }, + { + "epoch": 11.142172523961662, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6975 + }, + { + "epoch": 11.143769968051119, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6976 + }, + { + "epoch": 11.145367412140574, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 6977 + }, + { + "epoch": 11.146964856230031, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 6978 + }, + { + "epoch": 11.148562300319488, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 6979 + }, + { + "epoch": 11.150159744408946, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 6980 + }, + { + "epoch": 11.151757188498403, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 6981 + }, + { + "epoch": 11.15335463258786, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 6982 + }, + { + "epoch": 11.154952076677317, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 6983 + }, + { + "epoch": 11.156549520766774, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 6984 + }, + { + "epoch": 11.15814696485623, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 6985 + }, + { + "epoch": 11.159744408945686, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 6986 + }, + { + "epoch": 11.161341853035143, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 6987 + }, + { + "epoch": 11.1629392971246, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 6988 + }, + { + "epoch": 11.164536741214057, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 6989 + }, + { + "epoch": 11.166134185303514, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 6990 + }, + { + "epoch": 11.167731629392971, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 6991 + }, + { + "epoch": 11.169329073482428, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 6992 + }, + { + "epoch": 11.170926517571885, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 6993 + }, + { + "epoch": 11.172523961661343, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 6994 + }, + { + "epoch": 11.1741214057508, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 6995 + }, + { + "epoch": 11.175718849840255, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 6996 + }, + { + "epoch": 11.177316293929712, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 6997 + }, + { + "epoch": 11.178913738019169, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 6998 + }, + { + "epoch": 11.180511182108626, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 6999 + }, + { + "epoch": 11.182108626198083, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7000 + }, + { + "epoch": 11.18370607028754, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7001 + }, + { + "epoch": 11.185303514376997, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7002 + }, + { + "epoch": 11.186900958466454, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7003 + }, + { + "epoch": 11.188498402555911, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7004 + }, + { + "epoch": 11.190095846645367, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7005 + }, + { + "epoch": 11.191693290734824, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7006 + }, + { + "epoch": 11.19329073482428, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7007 + }, + { + "epoch": 11.194888178913738, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7008 + }, + { + "epoch": 11.196485623003195, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7009 + }, + { + "epoch": 11.198083067092652, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7010 + }, + { + "epoch": 11.199680511182109, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7011 + }, + { + "epoch": 11.201277955271566, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7012 + }, + { + "epoch": 11.202875399361023, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7013 + }, + { + "epoch": 11.204472843450478, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7014 + }, + { + "epoch": 11.206070287539935, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7015 + }, + { + "epoch": 11.207667731629392, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7016 + }, + { + "epoch": 11.20926517571885, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7017 + }, + { + "epoch": 11.210862619808307, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7018 + }, + { + "epoch": 11.212460063897764, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7019 + }, + { + "epoch": 11.21405750798722, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7020 + }, + { + "epoch": 11.215654952076678, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7021 + }, + { + "epoch": 11.217252396166135, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7022 + }, + { + "epoch": 11.218849840255592, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7023 + }, + { + "epoch": 11.220447284345047, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 7024 + }, + { + "epoch": 11.222044728434504, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7025 + }, + { + "epoch": 11.223642172523961, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7026 + }, + { + "epoch": 11.225239616613418, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7027 + }, + { + "epoch": 11.226837060702875, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7028 + }, + { + "epoch": 11.228434504792332, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7029 + }, + { + "epoch": 11.23003194888179, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7030 + }, + { + "epoch": 11.231629392971247, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7031 + }, + { + "epoch": 11.233226837060704, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7032 + }, + { + "epoch": 11.23482428115016, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7033 + }, + { + "epoch": 11.236421725239616, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7034 + }, + { + "epoch": 11.238019169329073, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7035 + }, + { + "epoch": 11.23961661341853, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 7036 + }, + { + "epoch": 11.241214057507987, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7037 + }, + { + "epoch": 11.242811501597444, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7038 + }, + { + "epoch": 11.244408945686901, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7039 + }, + { + "epoch": 11.246006389776358, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7040 + }, + { + "epoch": 11.247603833865815, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7041 + }, + { + "epoch": 11.249201277955272, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 7042 + }, + { + "epoch": 11.250798722044728, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 7043 + }, + { + "epoch": 11.252396166134185, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7044 + }, + { + "epoch": 11.253993610223642, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7045 + }, + { + "epoch": 11.255591054313099, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7046 + }, + { + "epoch": 11.257188498402556, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7047 + }, + { + "epoch": 11.258785942492013, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7048 + }, + { + "epoch": 11.26038338658147, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7049 + }, + { + "epoch": 11.261980830670927, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7050 + }, + { + "epoch": 11.263578274760384, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7051 + }, + { + "epoch": 11.26517571884984, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7052 + }, + { + "epoch": 11.266773162939296, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7053 + }, + { + "epoch": 11.268370607028753, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7054 + }, + { + "epoch": 11.26996805111821, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7055 + }, + { + "epoch": 11.271565495207668, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7056 + }, + { + "epoch": 11.273162939297125, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 7057 + }, + { + "epoch": 11.274760383386582, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 7058 + }, + { + "epoch": 11.276357827476039, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7059 + }, + { + "epoch": 11.277955271565496, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7060 + }, + { + "epoch": 11.279552715654953, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7061 + }, + { + "epoch": 11.281150159744408, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7062 + }, + { + "epoch": 11.282747603833865, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7063 + }, + { + "epoch": 11.284345047923322, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7064 + }, + { + "epoch": 11.28594249201278, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7065 + }, + { + "epoch": 11.287539936102236, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 7066 + }, + { + "epoch": 11.289137380191693, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7067 + }, + { + "epoch": 11.29073482428115, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7068 + }, + { + "epoch": 11.292332268370608, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 7069 + }, + { + "epoch": 11.293929712460065, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 7070 + }, + { + "epoch": 11.295527156549522, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7071 + }, + { + "epoch": 11.297124600638977, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 7072 + }, + { + "epoch": 11.298722044728434, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7073 + }, + { + "epoch": 11.300319488817891, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7074 + }, + { + "epoch": 11.301916932907348, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7075 + }, + { + "epoch": 11.303514376996805, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7076 + }, + { + "epoch": 11.305111821086262, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7077 + }, + { + "epoch": 11.30670926517572, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7078 + }, + { + "epoch": 11.308306709265176, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7079 + }, + { + "epoch": 11.309904153354633, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7080 + }, + { + "epoch": 11.311501597444089, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7081 + }, + { + "epoch": 11.313099041533546, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 7082 + }, + { + "epoch": 11.314696485623003, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7083 + }, + { + "epoch": 11.31629392971246, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7084 + }, + { + "epoch": 11.317891373801917, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7085 + }, + { + "epoch": 11.319488817891374, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7086 + }, + { + "epoch": 11.321086261980831, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7087 + }, + { + "epoch": 11.322683706070288, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 7088 + }, + { + "epoch": 11.324281150159745, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7089 + }, + { + "epoch": 11.3258785942492, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7090 + }, + { + "epoch": 11.327476038338657, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7091 + }, + { + "epoch": 11.329073482428115, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7092 + }, + { + "epoch": 11.330670926517572, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7093 + }, + { + "epoch": 11.332268370607029, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7094 + }, + { + "epoch": 11.333865814696486, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7095 + }, + { + "epoch": 11.335463258785943, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7096 + }, + { + "epoch": 11.3370607028754, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7097 + }, + { + "epoch": 11.338658146964857, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7098 + }, + { + "epoch": 11.340255591054314, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7099 + }, + { + "epoch": 11.34185303514377, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7100 + }, + { + "epoch": 11.343450479233226, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7101 + }, + { + "epoch": 11.345047923322683, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7102 + }, + { + "epoch": 11.34664536741214, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7103 + }, + { + "epoch": 11.348242811501597, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7104 + }, + { + "epoch": 11.349840255591054, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7105 + }, + { + "epoch": 11.351437699680512, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7106 + }, + { + "epoch": 11.353035143769969, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7107 + }, + { + "epoch": 11.354632587859426, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7108 + }, + { + "epoch": 11.356230031948883, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7109 + }, + { + "epoch": 11.357827476038338, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7110 + }, + { + "epoch": 11.359424920127795, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7111 + }, + { + "epoch": 11.361022364217252, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7112 + }, + { + "epoch": 11.36261980830671, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7113 + }, + { + "epoch": 11.364217252396166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7114 + }, + { + "epoch": 11.365814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7115 + }, + { + "epoch": 11.36741214057508, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7116 + }, + { + "epoch": 11.369009584664537, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 7117 + }, + { + "epoch": 11.370607028753994, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7118 + }, + { + "epoch": 11.37220447284345, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7119 + }, + { + "epoch": 11.373801916932907, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7120 + }, + { + "epoch": 11.375399361022364, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 7121 + }, + { + "epoch": 11.37699680511182, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7122 + }, + { + "epoch": 11.378594249201278, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7123 + }, + { + "epoch": 11.380191693290735, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7124 + }, + { + "epoch": 11.381789137380192, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7125 + }, + { + "epoch": 11.383386581469649, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7126 + }, + { + "epoch": 11.384984025559106, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7127 + }, + { + "epoch": 11.386581469648561, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7128 + }, + { + "epoch": 11.388178913738018, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7129 + }, + { + "epoch": 11.389776357827476, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7130 + }, + { + "epoch": 11.391373801916933, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7131 + }, + { + "epoch": 11.39297124600639, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7132 + }, + { + "epoch": 11.394568690095847, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 7133 + }, + { + "epoch": 11.396166134185304, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7134 + }, + { + "epoch": 11.39776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7135 + }, + { + "epoch": 11.399361022364218, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7136 + }, + { + "epoch": 11.400958466453675, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7137 + }, + { + "epoch": 11.40255591054313, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7138 + }, + { + "epoch": 11.404153354632587, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7139 + }, + { + "epoch": 11.405750798722044, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7140 + }, + { + "epoch": 11.407348242811501, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7141 + }, + { + "epoch": 11.408945686900958, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7142 + }, + { + "epoch": 11.410543130990416, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7143 + }, + { + "epoch": 11.412140575079873, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7144 + }, + { + "epoch": 11.41373801916933, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7145 + }, + { + "epoch": 11.415335463258787, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 7146 + }, + { + "epoch": 11.416932907348242, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7147 + }, + { + "epoch": 11.418530351437699, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 7148 + }, + { + "epoch": 11.420127795527156, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7149 + }, + { + "epoch": 11.421725239616613, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7150 + }, + { + "epoch": 11.42332268370607, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7151 + }, + { + "epoch": 11.424920127795527, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7152 + }, + { + "epoch": 11.426517571884984, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7153 + }, + { + "epoch": 11.428115015974441, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7154 + }, + { + "epoch": 11.429712460063898, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7155 + }, + { + "epoch": 11.431309904153355, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7156 + }, + { + "epoch": 11.43290734824281, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7157 + }, + { + "epoch": 11.434504792332268, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7158 + }, + { + "epoch": 11.436102236421725, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7159 + }, + { + "epoch": 11.437699680511182, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7160 + }, + { + "epoch": 11.439297124600639, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7161 + }, + { + "epoch": 11.440894568690096, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7162 + }, + { + "epoch": 11.442492012779553, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7163 + }, + { + "epoch": 11.44408945686901, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7164 + }, + { + "epoch": 11.445686900958467, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7165 + }, + { + "epoch": 11.447284345047922, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 7166 + }, + { + "epoch": 11.44888178913738, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7167 + }, + { + "epoch": 11.450479233226837, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7168 + }, + { + "epoch": 11.452076677316294, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7169 + }, + { + "epoch": 11.45367412140575, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7170 + }, + { + "epoch": 11.455271565495208, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7171 + }, + { + "epoch": 11.456869009584665, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7172 + }, + { + "epoch": 11.458466453674122, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7173 + }, + { + "epoch": 11.460063897763579, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7174 + }, + { + "epoch": 11.461661341853036, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7175 + }, + { + "epoch": 11.463258785942491, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7176 + }, + { + "epoch": 11.464856230031948, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7177 + }, + { + "epoch": 11.466453674121405, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7178 + }, + { + "epoch": 11.468051118210862, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7179 + }, + { + "epoch": 11.46964856230032, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7180 + }, + { + "epoch": 11.471246006389777, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 7181 + }, + { + "epoch": 11.472843450479234, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 7182 + }, + { + "epoch": 11.47444089456869, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7183 + }, + { + "epoch": 11.476038338658148, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7184 + }, + { + "epoch": 11.477635782747603, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 7185 + }, + { + "epoch": 11.47923322683706, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7186 + }, + { + "epoch": 11.480830670926517, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7187 + }, + { + "epoch": 11.482428115015974, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7188 + }, + { + "epoch": 11.484025559105431, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7189 + }, + { + "epoch": 11.485623003194888, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7190 + }, + { + "epoch": 11.487220447284345, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 7191 + }, + { + "epoch": 11.488817891373802, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7192 + }, + { + "epoch": 11.49041533546326, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7193 + }, + { + "epoch": 11.492012779552716, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7194 + }, + { + "epoch": 11.493610223642172, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7195 + }, + { + "epoch": 11.495207667731629, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7196 + }, + { + "epoch": 11.496805111821086, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 7197 + }, + { + "epoch": 11.498402555910543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7198 + }, + { + "epoch": 11.5, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7199 + }, + { + "epoch": 11.501597444089457, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7200 + }, + { + "epoch": 11.503194888178914, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 7201 + }, + { + "epoch": 11.504792332268371, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7202 + }, + { + "epoch": 11.506389776357828, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7203 + }, + { + "epoch": 11.507987220447284, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7204 + }, + { + "epoch": 11.50958466453674, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7205 + }, + { + "epoch": 11.511182108626198, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7206 + }, + { + "epoch": 11.512779552715655, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 7207 + }, + { + "epoch": 11.514376996805112, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7208 + }, + { + "epoch": 11.515974440894569, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7209 + }, + { + "epoch": 11.517571884984026, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7210 + }, + { + "epoch": 11.519169329073483, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 7211 + }, + { + "epoch": 11.52076677316294, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7212 + }, + { + "epoch": 11.522364217252395, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7213 + }, + { + "epoch": 11.523961661341852, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 7214 + }, + { + "epoch": 11.52555910543131, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7215 + }, + { + "epoch": 11.527156549520766, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7216 + }, + { + "epoch": 11.528753993610223, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7217 + }, + { + "epoch": 11.53035143769968, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7218 + }, + { + "epoch": 11.531948881789138, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7219 + }, + { + "epoch": 11.533546325878595, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7220 + }, + { + "epoch": 11.535143769968052, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7221 + }, + { + "epoch": 11.536741214057509, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7222 + }, + { + "epoch": 11.538338658146966, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7223 + }, + { + "epoch": 11.539936102236421, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7224 + }, + { + "epoch": 11.541533546325878, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 7225 + }, + { + "epoch": 11.543130990415335, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7226 + }, + { + "epoch": 11.544728434504792, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7227 + }, + { + "epoch": 11.54632587859425, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7228 + }, + { + "epoch": 11.547923322683706, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7229 + }, + { + "epoch": 11.549520766773163, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7230 + }, + { + "epoch": 11.55111821086262, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7231 + }, + { + "epoch": 11.552715654952078, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7232 + }, + { + "epoch": 11.554313099041533, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7233 + }, + { + "epoch": 11.55591054313099, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7234 + }, + { + "epoch": 11.557507987220447, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7235 + }, + { + "epoch": 11.559105431309904, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7236 + }, + { + "epoch": 11.560702875399361, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7237 + }, + { + "epoch": 11.562300319488818, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7238 + }, + { + "epoch": 11.563897763578275, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7239 + }, + { + "epoch": 11.565495207667732, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7240 + }, + { + "epoch": 11.56709265175719, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 7241 + }, + { + "epoch": 11.568690095846645, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7242 + }, + { + "epoch": 11.570287539936102, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7243 + }, + { + "epoch": 11.571884984025559, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7244 + }, + { + "epoch": 11.573482428115016, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7245 + }, + { + "epoch": 11.575079872204473, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7246 + }, + { + "epoch": 11.57667731629393, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7247 + }, + { + "epoch": 11.578274760383387, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 7248 + }, + { + "epoch": 11.579872204472844, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7249 + }, + { + "epoch": 11.581469648562301, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7250 + }, + { + "epoch": 11.583067092651756, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7251 + }, + { + "epoch": 11.584664536741213, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 7252 + }, + { + "epoch": 11.58626198083067, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7253 + }, + { + "epoch": 11.587859424920127, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7254 + }, + { + "epoch": 11.589456869009584, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7255 + }, + { + "epoch": 11.591054313099042, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7256 + }, + { + "epoch": 11.592651757188499, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7257 + }, + { + "epoch": 11.594249201277956, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7258 + }, + { + "epoch": 11.595846645367413, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7259 + }, + { + "epoch": 11.59744408945687, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7260 + }, + { + "epoch": 11.599041533546325, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7261 + }, + { + "epoch": 11.600638977635782, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 7262 + }, + { + "epoch": 11.60223642172524, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7263 + }, + { + "epoch": 11.603833865814696, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7264 + }, + { + "epoch": 11.605431309904153, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 7265 + }, + { + "epoch": 11.60702875399361, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7266 + }, + { + "epoch": 11.608626198083067, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7267 + }, + { + "epoch": 11.610223642172524, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7268 + }, + { + "epoch": 11.611821086261982, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7269 + }, + { + "epoch": 11.613418530351439, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7270 + }, + { + "epoch": 11.615015974440894, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7271 + }, + { + "epoch": 11.616613418530351, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7272 + }, + { + "epoch": 11.618210862619808, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7273 + }, + { + "epoch": 11.619808306709265, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7274 + }, + { + "epoch": 11.621405750798722, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7275 + }, + { + "epoch": 11.62300319488818, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7276 + }, + { + "epoch": 11.624600638977636, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7277 + }, + { + "epoch": 11.626198083067093, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 7278 + }, + { + "epoch": 11.62779552715655, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7279 + }, + { + "epoch": 11.629392971246006, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7280 + }, + { + "epoch": 11.630990415335463, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7281 + }, + { + "epoch": 11.63258785942492, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7282 + }, + { + "epoch": 11.634185303514377, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7283 + }, + { + "epoch": 11.635782747603834, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7284 + }, + { + "epoch": 11.63738019169329, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7285 + }, + { + "epoch": 11.638977635782748, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7286 + }, + { + "epoch": 11.640575079872205, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7287 + }, + { + "epoch": 11.642172523961662, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7288 + }, + { + "epoch": 11.643769968051117, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7289 + }, + { + "epoch": 11.645367412140574, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7290 + }, + { + "epoch": 11.646964856230031, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7291 + }, + { + "epoch": 11.648562300319488, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 7292 + }, + { + "epoch": 11.650159744408946, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7293 + }, + { + "epoch": 11.651757188498403, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7294 + }, + { + "epoch": 11.65335463258786, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 7295 + }, + { + "epoch": 11.654952076677317, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7296 + }, + { + "epoch": 11.656549520766774, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7297 + }, + { + "epoch": 11.65814696485623, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7298 + }, + { + "epoch": 11.659744408945686, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7299 + }, + { + "epoch": 11.661341853035143, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 7300 + }, + { + "epoch": 11.6629392971246, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7301 + }, + { + "epoch": 11.664536741214057, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7302 + }, + { + "epoch": 11.666134185303514, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7303 + }, + { + "epoch": 11.667731629392971, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7304 + }, + { + "epoch": 11.669329073482428, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7305 + }, + { + "epoch": 11.670926517571885, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 7306 + }, + { + "epoch": 11.672523961661343, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7307 + }, + { + "epoch": 11.6741214057508, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7308 + }, + { + "epoch": 11.675718849840255, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7309 + }, + { + "epoch": 11.677316293929712, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7310 + }, + { + "epoch": 11.678913738019169, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7311 + }, + { + "epoch": 11.680511182108626, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7312 + }, + { + "epoch": 11.682108626198083, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7313 + }, + { + "epoch": 11.68370607028754, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7314 + }, + { + "epoch": 11.685303514376997, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7315 + }, + { + "epoch": 11.686900958466454, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7316 + }, + { + "epoch": 11.688498402555911, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7317 + }, + { + "epoch": 11.690095846645367, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7318 + }, + { + "epoch": 11.691693290734824, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7319 + }, + { + "epoch": 11.69329073482428, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7320 + }, + { + "epoch": 11.694888178913738, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 7321 + }, + { + "epoch": 11.696485623003195, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7322 + }, + { + "epoch": 11.698083067092652, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7323 + }, + { + "epoch": 11.699680511182109, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7324 + }, + { + "epoch": 11.701277955271566, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7325 + }, + { + "epoch": 11.702875399361023, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 7326 + }, + { + "epoch": 11.704472843450478, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7327 + }, + { + "epoch": 11.706070287539935, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7328 + }, + { + "epoch": 11.707667731629392, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7329 + }, + { + "epoch": 11.70926517571885, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7330 + }, + { + "epoch": 11.710862619808307, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7331 + }, + { + "epoch": 11.712460063897764, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7332 + }, + { + "epoch": 11.71405750798722, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7333 + }, + { + "epoch": 11.715654952076678, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7334 + }, + { + "epoch": 11.717252396166135, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7335 + }, + { + "epoch": 11.718849840255592, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7336 + }, + { + "epoch": 11.720447284345047, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7337 + }, + { + "epoch": 11.722044728434504, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7338 + }, + { + "epoch": 11.723642172523961, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7339 + }, + { + "epoch": 11.725239616613418, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7340 + }, + { + "epoch": 11.726837060702875, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7341 + }, + { + "epoch": 11.728434504792332, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7342 + }, + { + "epoch": 11.73003194888179, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7343 + }, + { + "epoch": 11.731629392971247, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7344 + }, + { + "epoch": 11.733226837060704, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7345 + }, + { + "epoch": 11.73482428115016, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7346 + }, + { + "epoch": 11.736421725239616, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7347 + }, + { + "epoch": 11.738019169329073, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7348 + }, + { + "epoch": 11.73961661341853, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7349 + }, + { + "epoch": 11.741214057507987, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7350 + }, + { + "epoch": 11.742811501597444, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7351 + }, + { + "epoch": 11.744408945686901, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7352 + }, + { + "epoch": 11.746006389776358, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7353 + }, + { + "epoch": 11.747603833865815, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7354 + }, + { + "epoch": 11.749201277955272, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7355 + }, + { + "epoch": 11.750798722044728, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7356 + }, + { + "epoch": 11.752396166134185, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7357 + }, + { + "epoch": 11.753993610223642, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7358 + }, + { + "epoch": 11.755591054313099, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7359 + }, + { + "epoch": 11.757188498402556, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7360 + }, + { + "epoch": 11.758785942492013, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7361 + }, + { + "epoch": 11.76038338658147, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7362 + }, + { + "epoch": 11.761980830670927, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7363 + }, + { + "epoch": 11.763578274760384, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7364 + }, + { + "epoch": 11.76517571884984, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7365 + }, + { + "epoch": 11.766773162939296, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7366 + }, + { + "epoch": 11.768370607028753, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7367 + }, + { + "epoch": 11.76996805111821, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7368 + }, + { + "epoch": 11.771565495207668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7369 + }, + { + "epoch": 11.773162939297125, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7370 + }, + { + "epoch": 11.774760383386582, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7371 + }, + { + "epoch": 11.776357827476039, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 7372 + }, + { + "epoch": 11.777955271565496, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7373 + }, + { + "epoch": 11.779552715654953, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7374 + }, + { + "epoch": 11.781150159744408, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7375 + }, + { + "epoch": 11.782747603833865, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7376 + }, + { + "epoch": 11.784345047923322, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7377 + }, + { + "epoch": 11.78594249201278, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7378 + }, + { + "epoch": 11.787539936102236, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7379 + }, + { + "epoch": 11.789137380191693, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7380 + }, + { + "epoch": 11.79073482428115, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7381 + }, + { + "epoch": 11.792332268370608, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7382 + }, + { + "epoch": 11.793929712460065, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7383 + }, + { + "epoch": 11.795527156549522, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7384 + }, + { + "epoch": 11.797124600638977, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7385 + }, + { + "epoch": 11.798722044728434, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7386 + }, + { + "epoch": 11.800319488817891, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7387 + }, + { + "epoch": 11.801916932907348, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7388 + }, + { + "epoch": 11.803514376996805, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7389 + }, + { + "epoch": 11.805111821086262, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7390 + }, + { + "epoch": 11.80670926517572, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7391 + }, + { + "epoch": 11.808306709265176, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7392 + }, + { + "epoch": 11.809904153354633, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7393 + }, + { + "epoch": 11.811501597444089, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7394 + }, + { + "epoch": 11.813099041533546, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7395 + }, + { + "epoch": 11.814696485623003, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7396 + }, + { + "epoch": 11.81629392971246, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 7397 + }, + { + "epoch": 11.817891373801917, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7398 + }, + { + "epoch": 11.819488817891374, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7399 + }, + { + "epoch": 11.821086261980831, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7400 + }, + { + "epoch": 11.822683706070288, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7401 + }, + { + "epoch": 11.824281150159745, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7402 + }, + { + "epoch": 11.8258785942492, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7403 + }, + { + "epoch": 11.827476038338657, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7404 + }, + { + "epoch": 11.829073482428115, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7405 + }, + { + "epoch": 11.830670926517572, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7406 + }, + { + "epoch": 11.832268370607029, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7407 + }, + { + "epoch": 11.833865814696486, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 7408 + }, + { + "epoch": 11.835463258785943, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7409 + }, + { + "epoch": 11.8370607028754, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7410 + }, + { + "epoch": 11.838658146964857, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7411 + }, + { + "epoch": 11.840255591054314, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7412 + }, + { + "epoch": 11.84185303514377, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7413 + }, + { + "epoch": 11.843450479233226, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 7414 + }, + { + "epoch": 11.845047923322683, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7415 + }, + { + "epoch": 11.84664536741214, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7416 + }, + { + "epoch": 11.848242811501597, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7417 + }, + { + "epoch": 11.849840255591054, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7418 + }, + { + "epoch": 11.851437699680512, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7419 + }, + { + "epoch": 11.853035143769969, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7420 + }, + { + "epoch": 11.854632587859426, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7421 + }, + { + "epoch": 11.856230031948883, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7422 + }, + { + "epoch": 11.857827476038338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7423 + }, + { + "epoch": 11.859424920127795, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7424 + }, + { + "epoch": 11.861022364217252, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7425 + }, + { + "epoch": 11.86261980830671, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7426 + }, + { + "epoch": 11.864217252396166, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7427 + }, + { + "epoch": 11.865814696485623, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7428 + }, + { + "epoch": 11.86741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 7429 + }, + { + "epoch": 11.869009584664537, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7430 + }, + { + "epoch": 11.870607028753994, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7431 + }, + { + "epoch": 11.87220447284345, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7432 + }, + { + "epoch": 11.873801916932907, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7433 + }, + { + "epoch": 11.875399361022364, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7434 + }, + { + "epoch": 11.87699680511182, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7435 + }, + { + "epoch": 11.878594249201278, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7436 + }, + { + "epoch": 11.880191693290735, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7437 + }, + { + "epoch": 11.881789137380192, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7438 + }, + { + "epoch": 11.883386581469649, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7439 + }, + { + "epoch": 11.884984025559106, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7440 + }, + { + "epoch": 11.886581469648561, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7441 + }, + { + "epoch": 11.888178913738018, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7442 + }, + { + "epoch": 11.889776357827476, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7443 + }, + { + "epoch": 11.891373801916933, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7444 + }, + { + "epoch": 11.89297124600639, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7445 + }, + { + "epoch": 11.894568690095847, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7446 + }, + { + "epoch": 11.896166134185304, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7447 + }, + { + "epoch": 11.89776357827476, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7448 + }, + { + "epoch": 11.899361022364218, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7449 + }, + { + "epoch": 11.900958466453675, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7450 + }, + { + "epoch": 11.90255591054313, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7451 + }, + { + "epoch": 11.904153354632587, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7452 + }, + { + "epoch": 11.905750798722044, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7453 + }, + { + "epoch": 11.907348242811501, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7454 + }, + { + "epoch": 11.908945686900958, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7455 + }, + { + "epoch": 11.910543130990416, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7456 + }, + { + "epoch": 11.912140575079873, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7457 + }, + { + "epoch": 11.91373801916933, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 7458 + }, + { + "epoch": 11.915335463258787, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7459 + }, + { + "epoch": 11.916932907348244, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7460 + }, + { + "epoch": 11.918530351437699, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7461 + }, + { + "epoch": 11.920127795527156, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7462 + }, + { + "epoch": 11.921725239616613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7463 + }, + { + "epoch": 11.92332268370607, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7464 + }, + { + "epoch": 11.924920127795527, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7465 + }, + { + "epoch": 11.926517571884984, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7466 + }, + { + "epoch": 11.928115015974441, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7467 + }, + { + "epoch": 11.929712460063898, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7468 + }, + { + "epoch": 11.931309904153355, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7469 + }, + { + "epoch": 11.93290734824281, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7470 + }, + { + "epoch": 11.934504792332268, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 7471 + }, + { + "epoch": 11.936102236421725, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7472 + }, + { + "epoch": 11.937699680511182, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 7473 + }, + { + "epoch": 11.939297124600639, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7474 + }, + { + "epoch": 11.940894568690096, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7475 + }, + { + "epoch": 11.942492012779553, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7476 + }, + { + "epoch": 11.94408945686901, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7477 + }, + { + "epoch": 11.945686900958467, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 7478 + }, + { + "epoch": 11.947284345047922, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7479 + }, + { + "epoch": 11.94888178913738, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7480 + }, + { + "epoch": 11.950479233226837, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7481 + }, + { + "epoch": 11.952076677316294, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7482 + }, + { + "epoch": 11.95367412140575, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7483 + }, + { + "epoch": 11.955271565495208, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7484 + }, + { + "epoch": 11.956869009584665, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7485 + }, + { + "epoch": 11.958466453674122, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7486 + }, + { + "epoch": 11.960063897763579, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7487 + }, + { + "epoch": 11.961661341853034, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7488 + }, + { + "epoch": 11.963258785942491, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7489 + }, + { + "epoch": 11.964856230031948, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 7490 + }, + { + "epoch": 11.966453674121405, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 7491 + }, + { + "epoch": 11.968051118210862, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7492 + }, + { + "epoch": 11.96964856230032, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7493 + }, + { + "epoch": 11.971246006389777, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7494 + }, + { + "epoch": 11.972843450479234, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7495 + }, + { + "epoch": 11.97444089456869, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7496 + }, + { + "epoch": 11.976038338658148, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7497 + }, + { + "epoch": 11.977635782747605, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7498 + }, + { + "epoch": 11.97923322683706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7499 + }, + { + "epoch": 11.980830670926517, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7500 + }, + { + "epoch": 11.982428115015974, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7501 + }, + { + "epoch": 11.984025559105431, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7502 + }, + { + "epoch": 11.985623003194888, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7503 + }, + { + "epoch": 11.987220447284345, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7504 + }, + { + "epoch": 11.988817891373802, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7505 + }, + { + "epoch": 11.99041533546326, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7506 + }, + { + "epoch": 11.992012779552716, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7507 + }, + { + "epoch": 11.993610223642172, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7508 + }, + { + "epoch": 11.995207667731629, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7509 + }, + { + "epoch": 11.996805111821086, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7510 + }, + { + "epoch": 11.998402555910543, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7511 + }, + { + "epoch": 12.0, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7512 + }, + { + "epoch": 12.001597444089457, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7513 + }, + { + "epoch": 12.003194888178914, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7514 + }, + { + "epoch": 12.004792332268371, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7515 + }, + { + "epoch": 12.006389776357828, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7516 + }, + { + "epoch": 12.007987220447284, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7517 + }, + { + "epoch": 12.00958466453674, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7518 + }, + { + "epoch": 12.011182108626198, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7519 + }, + { + "epoch": 12.012779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 7520 + }, + { + "epoch": 12.014376996805112, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7521 + }, + { + "epoch": 12.015974440894569, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7522 + }, + { + "epoch": 12.017571884984026, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 7523 + }, + { + "epoch": 12.019169329073483, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7524 + }, + { + "epoch": 12.02076677316294, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7525 + }, + { + "epoch": 12.022364217252397, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7526 + }, + { + "epoch": 12.023961661341852, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7527 + }, + { + "epoch": 12.02555910543131, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7528 + }, + { + "epoch": 12.027156549520766, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7529 + }, + { + "epoch": 12.028753993610223, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7530 + }, + { + "epoch": 12.03035143769968, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7531 + }, + { + "epoch": 12.031948881789138, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7532 + }, + { + "epoch": 12.033546325878595, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 7533 + }, + { + "epoch": 12.035143769968052, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7534 + }, + { + "epoch": 12.036741214057509, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7535 + }, + { + "epoch": 12.038338658146964, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7536 + }, + { + "epoch": 12.039936102236421, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7537 + }, + { + "epoch": 12.041533546325878, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7538 + }, + { + "epoch": 12.043130990415335, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 7539 + }, + { + "epoch": 12.044728434504792, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 7540 + }, + { + "epoch": 12.04632587859425, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 7541 + }, + { + "epoch": 12.047923322683706, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7542 + }, + { + "epoch": 12.049520766773163, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7543 + }, + { + "epoch": 12.05111821086262, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7544 + }, + { + "epoch": 12.052715654952078, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7545 + }, + { + "epoch": 12.054313099041533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7546 + }, + { + "epoch": 12.05591054313099, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7547 + }, + { + "epoch": 12.057507987220447, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7548 + }, + { + "epoch": 12.059105431309904, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7549 + }, + { + "epoch": 12.060702875399361, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7550 + }, + { + "epoch": 12.062300319488818, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7551 + }, + { + "epoch": 12.063897763578275, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7552 + }, + { + "epoch": 12.065495207667732, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7553 + }, + { + "epoch": 12.06709265175719, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7554 + }, + { + "epoch": 12.068690095846645, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7555 + }, + { + "epoch": 12.070287539936102, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7556 + }, + { + "epoch": 12.071884984025559, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7557 + }, + { + "epoch": 12.073482428115016, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7558 + }, + { + "epoch": 12.075079872204473, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 7559 + }, + { + "epoch": 12.07667731629393, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7560 + }, + { + "epoch": 12.078274760383387, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7561 + }, + { + "epoch": 12.079872204472844, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 7562 + }, + { + "epoch": 12.081469648562301, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7563 + }, + { + "epoch": 12.083067092651758, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7564 + }, + { + "epoch": 12.084664536741213, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7565 + }, + { + "epoch": 12.08626198083067, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7566 + }, + { + "epoch": 12.087859424920127, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 7567 + }, + { + "epoch": 12.089456869009584, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7568 + }, + { + "epoch": 12.091054313099042, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7569 + }, + { + "epoch": 12.092651757188499, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7570 + }, + { + "epoch": 12.094249201277956, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7571 + }, + { + "epoch": 12.095846645367413, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 7572 + }, + { + "epoch": 12.09744408945687, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7573 + }, + { + "epoch": 12.099041533546325, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7574 + }, + { + "epoch": 12.100638977635782, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7575 + }, + { + "epoch": 12.10223642172524, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7576 + }, + { + "epoch": 12.103833865814696, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7577 + }, + { + "epoch": 12.105431309904153, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7578 + }, + { + "epoch": 12.10702875399361, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7579 + }, + { + "epoch": 12.108626198083067, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7580 + }, + { + "epoch": 12.110223642172524, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7581 + }, + { + "epoch": 12.111821086261982, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7582 + }, + { + "epoch": 12.113418530351439, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7583 + }, + { + "epoch": 12.115015974440894, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7584 + }, + { + "epoch": 12.116613418530351, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 7585 + }, + { + "epoch": 12.118210862619808, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7586 + }, + { + "epoch": 12.119808306709265, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7587 + }, + { + "epoch": 12.121405750798722, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7588 + }, + { + "epoch": 12.12300319488818, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7589 + }, + { + "epoch": 12.124600638977636, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7590 + }, + { + "epoch": 12.126198083067093, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7591 + }, + { + "epoch": 12.12779552715655, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7592 + }, + { + "epoch": 12.129392971246006, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7593 + }, + { + "epoch": 12.130990415335463, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7594 + }, + { + "epoch": 12.13258785942492, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7595 + }, + { + "epoch": 12.134185303514377, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7596 + }, + { + "epoch": 12.135782747603834, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7597 + }, + { + "epoch": 12.13738019169329, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7598 + }, + { + "epoch": 12.138977635782748, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7599 + }, + { + "epoch": 12.140575079872205, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7600 + }, + { + "epoch": 12.142172523961662, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7601 + }, + { + "epoch": 12.143769968051119, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7602 + }, + { + "epoch": 12.145367412140574, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7603 + }, + { + "epoch": 12.146964856230031, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7604 + }, + { + "epoch": 12.148562300319488, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7605 + }, + { + "epoch": 12.150159744408946, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7606 + }, + { + "epoch": 12.151757188498403, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 7607 + }, + { + "epoch": 12.15335463258786, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7608 + }, + { + "epoch": 12.154952076677317, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7609 + }, + { + "epoch": 12.156549520766774, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7610 + }, + { + "epoch": 12.15814696485623, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 7611 + }, + { + "epoch": 12.159744408945686, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7612 + }, + { + "epoch": 12.161341853035143, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7613 + }, + { + "epoch": 12.1629392971246, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7614 + }, + { + "epoch": 12.164536741214057, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7615 + }, + { + "epoch": 12.166134185303514, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7616 + }, + { + "epoch": 12.167731629392971, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7617 + }, + { + "epoch": 12.169329073482428, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7618 + }, + { + "epoch": 12.170926517571885, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7619 + }, + { + "epoch": 12.172523961661343, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7620 + }, + { + "epoch": 12.1741214057508, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7621 + }, + { + "epoch": 12.175718849840255, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 7622 + }, + { + "epoch": 12.177316293929712, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7623 + }, + { + "epoch": 12.178913738019169, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7624 + }, + { + "epoch": 12.180511182108626, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7625 + }, + { + "epoch": 12.182108626198083, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 7626 + }, + { + "epoch": 12.18370607028754, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7627 + }, + { + "epoch": 12.185303514376997, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7628 + }, + { + "epoch": 12.186900958466454, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7629 + }, + { + "epoch": 12.188498402555911, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7630 + }, + { + "epoch": 12.190095846645367, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7631 + }, + { + "epoch": 12.191693290734824, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7632 + }, + { + "epoch": 12.19329073482428, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7633 + }, + { + "epoch": 12.194888178913738, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7634 + }, + { + "epoch": 12.196485623003195, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7635 + }, + { + "epoch": 12.198083067092652, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7636 + }, + { + "epoch": 12.199680511182109, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 7637 + }, + { + "epoch": 12.201277955271566, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7638 + }, + { + "epoch": 12.202875399361023, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7639 + }, + { + "epoch": 12.204472843450478, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 7640 + }, + { + "epoch": 12.206070287539935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7641 + }, + { + "epoch": 12.207667731629392, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7642 + }, + { + "epoch": 12.20926517571885, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7643 + }, + { + "epoch": 12.210862619808307, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7644 + }, + { + "epoch": 12.212460063897764, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 7645 + }, + { + "epoch": 12.21405750798722, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7646 + }, + { + "epoch": 12.215654952076678, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7647 + }, + { + "epoch": 12.217252396166135, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7648 + }, + { + "epoch": 12.218849840255592, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 7649 + }, + { + "epoch": 12.220447284345047, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7650 + }, + { + "epoch": 12.222044728434504, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7651 + }, + { + "epoch": 12.223642172523961, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7652 + }, + { + "epoch": 12.225239616613418, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7653 + }, + { + "epoch": 12.226837060702875, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7654 + }, + { + "epoch": 12.228434504792332, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7655 + }, + { + "epoch": 12.23003194888179, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7656 + }, + { + "epoch": 12.231629392971247, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7657 + }, + { + "epoch": 12.233226837060704, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7658 + }, + { + "epoch": 12.23482428115016, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 7659 + }, + { + "epoch": 12.236421725239616, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7660 + }, + { + "epoch": 12.238019169329073, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7661 + }, + { + "epoch": 12.23961661341853, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7662 + }, + { + "epoch": 12.241214057507987, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7663 + }, + { + "epoch": 12.242811501597444, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 7664 + }, + { + "epoch": 12.244408945686901, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7665 + }, + { + "epoch": 12.246006389776358, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7666 + }, + { + "epoch": 12.247603833865815, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7667 + }, + { + "epoch": 12.249201277955272, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7668 + }, + { + "epoch": 12.250798722044728, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7669 + }, + { + "epoch": 12.252396166134185, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7670 + }, + { + "epoch": 12.253993610223642, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7671 + }, + { + "epoch": 12.255591054313099, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7672 + }, + { + "epoch": 12.257188498402556, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 7673 + }, + { + "epoch": 12.258785942492013, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7674 + }, + { + "epoch": 12.26038338658147, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7675 + }, + { + "epoch": 12.261980830670927, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7676 + }, + { + "epoch": 12.263578274760384, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 7677 + }, + { + "epoch": 12.26517571884984, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7678 + }, + { + "epoch": 12.266773162939296, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7679 + }, + { + "epoch": 12.268370607028753, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7680 + }, + { + "epoch": 12.26996805111821, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7681 + }, + { + "epoch": 12.271565495207668, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7682 + }, + { + "epoch": 12.273162939297125, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7683 + }, + { + "epoch": 12.274760383386582, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7684 + }, + { + "epoch": 12.276357827476039, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7685 + }, + { + "epoch": 12.277955271565496, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7686 + }, + { + "epoch": 12.279552715654953, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7687 + }, + { + "epoch": 12.281150159744408, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7688 + }, + { + "epoch": 12.282747603833865, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7689 + }, + { + "epoch": 12.284345047923322, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7690 + }, + { + "epoch": 12.28594249201278, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7691 + }, + { + "epoch": 12.287539936102236, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7692 + }, + { + "epoch": 12.289137380191693, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7693 + }, + { + "epoch": 12.29073482428115, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7694 + }, + { + "epoch": 12.292332268370608, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7695 + }, + { + "epoch": 12.293929712460065, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7696 + }, + { + "epoch": 12.295527156549522, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7697 + }, + { + "epoch": 12.297124600638977, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7698 + }, + { + "epoch": 12.298722044728434, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 7699 + }, + { + "epoch": 12.300319488817891, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7700 + }, + { + "epoch": 12.301916932907348, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7701 + }, + { + "epoch": 12.303514376996805, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 7702 + }, + { + "epoch": 12.305111821086262, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 7703 + }, + { + "epoch": 12.30670926517572, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7704 + }, + { + "epoch": 12.308306709265176, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7705 + }, + { + "epoch": 12.309904153354633, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7706 + }, + { + "epoch": 12.311501597444089, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7707 + }, + { + "epoch": 12.313099041533546, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7708 + }, + { + "epoch": 12.314696485623003, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7709 + }, + { + "epoch": 12.31629392971246, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7710 + }, + { + "epoch": 12.317891373801917, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7711 + }, + { + "epoch": 12.319488817891374, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7712 + }, + { + "epoch": 12.321086261980831, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7713 + }, + { + "epoch": 12.322683706070288, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7714 + }, + { + "epoch": 12.324281150159745, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7715 + }, + { + "epoch": 12.3258785942492, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7716 + }, + { + "epoch": 12.327476038338657, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7717 + }, + { + "epoch": 12.329073482428115, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7718 + }, + { + "epoch": 12.330670926517572, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7719 + }, + { + "epoch": 12.332268370607029, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 7720 + }, + { + "epoch": 12.333865814696486, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7721 + }, + { + "epoch": 12.335463258785943, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7722 + }, + { + "epoch": 12.3370607028754, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7723 + }, + { + "epoch": 12.338658146964857, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7724 + }, + { + "epoch": 12.340255591054314, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7725 + }, + { + "epoch": 12.34185303514377, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7726 + }, + { + "epoch": 12.343450479233226, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7727 + }, + { + "epoch": 12.345047923322683, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 7728 + }, + { + "epoch": 12.34664536741214, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 7729 + }, + { + "epoch": 12.348242811501597, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7730 + }, + { + "epoch": 12.349840255591054, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7731 + }, + { + "epoch": 12.351437699680512, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 7732 + }, + { + "epoch": 12.353035143769969, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7733 + }, + { + "epoch": 12.354632587859426, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 7734 + }, + { + "epoch": 12.356230031948883, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7735 + }, + { + "epoch": 12.357827476038338, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7736 + }, + { + "epoch": 12.359424920127795, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7737 + }, + { + "epoch": 12.361022364217252, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7738 + }, + { + "epoch": 12.36261980830671, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 7739 + }, + { + "epoch": 12.364217252396166, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7740 + }, + { + "epoch": 12.365814696485623, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7741 + }, + { + "epoch": 12.36741214057508, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7742 + }, + { + "epoch": 12.369009584664537, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 7743 + }, + { + "epoch": 12.370607028753994, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7744 + }, + { + "epoch": 12.37220447284345, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 7745 + }, + { + "epoch": 12.373801916932907, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7746 + }, + { + "epoch": 12.375399361022364, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 7747 + }, + { + "epoch": 12.37699680511182, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7748 + }, + { + "epoch": 12.378594249201278, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7749 + }, + { + "epoch": 12.380191693290735, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7750 + }, + { + "epoch": 12.381789137380192, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7751 + }, + { + "epoch": 12.383386581469649, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7752 + }, + { + "epoch": 12.384984025559106, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7753 + }, + { + "epoch": 12.386581469648561, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7754 + }, + { + "epoch": 12.388178913738018, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7755 + }, + { + "epoch": 12.389776357827476, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7756 + }, + { + "epoch": 12.391373801916933, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7757 + }, + { + "epoch": 12.39297124600639, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 7758 + }, + { + "epoch": 12.394568690095847, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7759 + }, + { + "epoch": 12.396166134185304, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7760 + }, + { + "epoch": 12.39776357827476, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7761 + }, + { + "epoch": 12.399361022364218, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7762 + }, + { + "epoch": 12.400958466453675, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 7763 + }, + { + "epoch": 12.40255591054313, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7764 + }, + { + "epoch": 12.404153354632587, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7765 + }, + { + "epoch": 12.405750798722044, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7766 + }, + { + "epoch": 12.407348242811501, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7767 + }, + { + "epoch": 12.408945686900958, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7768 + }, + { + "epoch": 12.410543130990416, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 7769 + }, + { + "epoch": 12.412140575079873, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 7770 + }, + { + "epoch": 12.41373801916933, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 7771 + }, + { + "epoch": 12.415335463258787, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7772 + }, + { + "epoch": 12.416932907348242, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7773 + }, + { + "epoch": 12.418530351437699, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 7774 + }, + { + "epoch": 12.420127795527156, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7775 + }, + { + "epoch": 12.421725239616613, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7776 + }, + { + "epoch": 12.42332268370607, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7777 + }, + { + "epoch": 12.424920127795527, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7778 + }, + { + "epoch": 12.426517571884984, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7779 + }, + { + "epoch": 12.428115015974441, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7780 + }, + { + "epoch": 12.429712460063898, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7781 + }, + { + "epoch": 12.431309904153355, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7782 + }, + { + "epoch": 12.43290734824281, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7783 + }, + { + "epoch": 12.434504792332268, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7784 + }, + { + "epoch": 12.436102236421725, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7785 + }, + { + "epoch": 12.437699680511182, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7786 + }, + { + "epoch": 12.439297124600639, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7787 + }, + { + "epoch": 12.440894568690096, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7788 + }, + { + "epoch": 12.442492012779553, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7789 + }, + { + "epoch": 12.44408945686901, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 7790 + }, + { + "epoch": 12.445686900958467, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 7791 + }, + { + "epoch": 12.447284345047922, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7792 + }, + { + "epoch": 12.44888178913738, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7793 + }, + { + "epoch": 12.450479233226837, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 7794 + }, + { + "epoch": 12.452076677316294, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 7795 + }, + { + "epoch": 12.45367412140575, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7796 + }, + { + "epoch": 12.455271565495208, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7797 + }, + { + "epoch": 12.456869009584665, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7798 + }, + { + "epoch": 12.458466453674122, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 7799 + }, + { + "epoch": 12.460063897763579, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 7800 + }, + { + "epoch": 12.461661341853036, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7801 + }, + { + "epoch": 12.463258785942491, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7802 + }, + { + "epoch": 12.464856230031948, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7803 + }, + { + "epoch": 12.466453674121405, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7804 + }, + { + "epoch": 12.468051118210862, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7805 + }, + { + "epoch": 12.46964856230032, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 7806 + }, + { + "epoch": 12.471246006389777, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 7807 + }, + { + "epoch": 12.472843450479234, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 7808 + }, + { + "epoch": 12.47444089456869, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7809 + }, + { + "epoch": 12.476038338658148, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7810 + }, + { + "epoch": 12.477635782747603, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7811 + }, + { + "epoch": 12.47923322683706, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 7812 + }, + { + "epoch": 12.480830670926517, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 7813 + }, + { + "epoch": 12.482428115015974, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7814 + }, + { + "epoch": 12.484025559105431, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7815 + }, + { + "epoch": 12.485623003194888, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 7816 + }, + { + "epoch": 12.487220447284345, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7817 + }, + { + "epoch": 12.488817891373802, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 7818 + }, + { + "epoch": 12.49041533546326, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7819 + }, + { + "epoch": 12.492012779552716, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7820 + }, + { + "epoch": 12.493610223642172, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7821 + }, + { + "epoch": 12.495207667731629, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7822 + }, + { + "epoch": 12.496805111821086, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 7823 + }, + { + "epoch": 12.498402555910543, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7824 + }, + { + "epoch": 12.5, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7825 + }, + { + "epoch": 12.501597444089457, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7826 + }, + { + "epoch": 12.503194888178914, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7827 + }, + { + "epoch": 12.504792332268371, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7828 + }, + { + "epoch": 12.506389776357828, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7829 + }, + { + "epoch": 12.507987220447284, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 7830 + }, + { + "epoch": 12.50958466453674, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7831 + }, + { + "epoch": 12.511182108626198, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7832 + }, + { + "epoch": 12.512779552715655, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7833 + }, + { + "epoch": 12.514376996805112, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 7834 + }, + { + "epoch": 12.515974440894569, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 7835 + }, + { + "epoch": 12.517571884984026, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7836 + }, + { + "epoch": 12.519169329073483, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7837 + }, + { + "epoch": 12.52076677316294, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7838 + }, + { + "epoch": 12.522364217252395, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7839 + }, + { + "epoch": 12.523961661341852, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 7840 + }, + { + "epoch": 12.52555910543131, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7841 + }, + { + "epoch": 12.527156549520766, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 7842 + }, + { + "epoch": 12.528753993610223, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 7843 + }, + { + "epoch": 12.53035143769968, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7844 + }, + { + "epoch": 12.531948881789138, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7845 + }, + { + "epoch": 12.533546325878595, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7846 + }, + { + "epoch": 12.535143769968052, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7847 + }, + { + "epoch": 12.536741214057509, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7848 + }, + { + "epoch": 12.538338658146966, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7849 + }, + { + "epoch": 12.539936102236421, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 7850 + }, + { + "epoch": 12.541533546325878, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7851 + }, + { + "epoch": 12.543130990415335, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 7852 + }, + { + "epoch": 12.544728434504792, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 7853 + }, + { + "epoch": 12.54632587859425, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7854 + }, + { + "epoch": 12.547923322683706, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7855 + }, + { + "epoch": 12.549520766773163, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7856 + }, + { + "epoch": 12.55111821086262, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7857 + }, + { + "epoch": 12.552715654952078, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7858 + }, + { + "epoch": 12.554313099041533, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 7859 + }, + { + "epoch": 12.55591054313099, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 7860 + }, + { + "epoch": 12.557507987220447, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7861 + }, + { + "epoch": 12.559105431309904, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 7862 + }, + { + "epoch": 12.560702875399361, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7863 + }, + { + "epoch": 12.562300319488818, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7864 + }, + { + "epoch": 12.563897763578275, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7865 + }, + { + "epoch": 12.565495207667732, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7866 + }, + { + "epoch": 12.56709265175719, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7867 + }, + { + "epoch": 12.568690095846645, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 7868 + }, + { + "epoch": 12.570287539936102, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 7869 + }, + { + "epoch": 12.571884984025559, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7870 + }, + { + "epoch": 12.573482428115016, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7871 + }, + { + "epoch": 12.575079872204473, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 7872 + }, + { + "epoch": 12.57667731629393, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 7873 + }, + { + "epoch": 12.578274760383387, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 7874 + }, + { + "epoch": 12.579872204472844, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 7875 + }, + { + "epoch": 12.581469648562301, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 7876 + }, + { + "epoch": 12.583067092651756, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7877 + }, + { + "epoch": 12.584664536741213, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 7878 + }, + { + "epoch": 12.58626198083067, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7879 + }, + { + "epoch": 12.587859424920127, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 7880 + }, + { + "epoch": 12.589456869009584, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 7881 + }, + { + "epoch": 12.591054313099042, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7882 + }, + { + "epoch": 12.592651757188499, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 7883 + }, + { + "epoch": 12.594249201277956, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7884 + }, + { + "epoch": 12.595846645367413, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7885 + }, + { + "epoch": 12.59744408945687, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7886 + }, + { + "epoch": 12.599041533546325, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 7887 + }, + { + "epoch": 12.600638977635782, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7888 + }, + { + "epoch": 12.60223642172524, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 7889 + }, + { + "epoch": 12.603833865814696, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7890 + }, + { + "epoch": 12.605431309904153, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7891 + }, + { + "epoch": 12.60702875399361, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 7892 + }, + { + "epoch": 12.608626198083067, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7893 + }, + { + "epoch": 12.610223642172524, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 7894 + }, + { + "epoch": 12.611821086261982, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7895 + }, + { + "epoch": 12.613418530351439, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7896 + }, + { + "epoch": 12.615015974440894, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 7897 + }, + { + "epoch": 12.616613418530351, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 7898 + }, + { + "epoch": 12.618210862619808, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7899 + }, + { + "epoch": 12.619808306709265, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 7900 + }, + { + "epoch": 12.621405750798722, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 7901 + }, + { + "epoch": 12.62300319488818, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7902 + }, + { + "epoch": 12.624600638977636, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 7903 + }, + { + "epoch": 12.626198083067093, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7904 + }, + { + "epoch": 12.62779552715655, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 7905 + }, + { + "epoch": 12.629392971246006, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 7906 + }, + { + "epoch": 12.630990415335463, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 7907 + }, + { + "epoch": 12.63258785942492, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7908 + }, + { + "epoch": 12.634185303514377, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 7909 + }, + { + "epoch": 12.635782747603834, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7910 + }, + { + "epoch": 12.63738019169329, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 7911 + }, + { + "epoch": 12.638977635782748, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7912 + }, + { + "epoch": 12.640575079872205, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7913 + }, + { + "epoch": 12.642172523961662, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7914 + }, + { + "epoch": 12.643769968051117, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 7915 + }, + { + "epoch": 12.645367412140574, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 7916 + }, + { + "epoch": 12.646964856230031, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7917 + }, + { + "epoch": 12.648562300319488, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7918 + }, + { + "epoch": 12.650159744408946, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7919 + }, + { + "epoch": 12.651757188498403, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 7920 + }, + { + "epoch": 12.65335463258786, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 7921 + }, + { + "epoch": 12.654952076677317, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 7922 + }, + { + "epoch": 12.656549520766774, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 7923 + }, + { + "epoch": 12.65814696485623, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 7924 + }, + { + "epoch": 12.659744408945686, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7925 + }, + { + "epoch": 12.661341853035143, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 7926 + }, + { + "epoch": 12.6629392971246, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7927 + }, + { + "epoch": 12.664536741214057, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 7928 + }, + { + "epoch": 12.666134185303514, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 7929 + }, + { + "epoch": 12.667731629392971, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7930 + }, + { + "epoch": 12.669329073482428, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7931 + }, + { + "epoch": 12.670926517571885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 7932 + }, + { + "epoch": 12.672523961661343, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 7933 + }, + { + "epoch": 12.6741214057508, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 7934 + }, + { + "epoch": 12.675718849840255, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7935 + }, + { + "epoch": 12.677316293929712, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 7936 + }, + { + "epoch": 12.678913738019169, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 7937 + }, + { + "epoch": 12.680511182108626, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 7938 + }, + { + "epoch": 12.682108626198083, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7939 + }, + { + "epoch": 12.68370607028754, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7940 + }, + { + "epoch": 12.685303514376997, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7941 + }, + { + "epoch": 12.686900958466454, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 7942 + }, + { + "epoch": 12.688498402555911, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 7943 + }, + { + "epoch": 12.690095846645367, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 7944 + }, + { + "epoch": 12.691693290734824, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 7945 + }, + { + "epoch": 12.69329073482428, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7946 + }, + { + "epoch": 12.694888178913738, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 7947 + }, + { + "epoch": 12.696485623003195, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 7948 + }, + { + "epoch": 12.698083067092652, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7949 + }, + { + "epoch": 12.699680511182109, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 7950 + }, + { + "epoch": 12.701277955271566, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 7951 + }, + { + "epoch": 12.702875399361023, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 7952 + }, + { + "epoch": 12.704472843450478, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7953 + }, + { + "epoch": 12.706070287539935, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 7954 + }, + { + "epoch": 12.707667731629392, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 7955 + }, + { + "epoch": 12.70926517571885, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 7956 + }, + { + "epoch": 12.710862619808307, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 7957 + }, + { + "epoch": 12.712460063897764, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 7958 + }, + { + "epoch": 12.71405750798722, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 7959 + }, + { + "epoch": 12.715654952076678, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 7960 + }, + { + "epoch": 12.717252396166135, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 7961 + }, + { + "epoch": 12.718849840255592, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7962 + }, + { + "epoch": 12.720447284345047, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 7963 + }, + { + "epoch": 12.722044728434504, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7964 + }, + { + "epoch": 12.723642172523961, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7965 + }, + { + "epoch": 12.725239616613418, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 7966 + }, + { + "epoch": 12.726837060702875, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 7967 + }, + { + "epoch": 12.728434504792332, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 7968 + }, + { + "epoch": 12.73003194888179, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 7969 + }, + { + "epoch": 12.731629392971247, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7970 + }, + { + "epoch": 12.733226837060704, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 7971 + }, + { + "epoch": 12.73482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 7972 + }, + { + "epoch": 12.736421725239616, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 7973 + }, + { + "epoch": 12.738019169329073, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7974 + }, + { + "epoch": 12.73961661341853, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 7975 + }, + { + "epoch": 12.741214057507987, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 7976 + }, + { + "epoch": 12.742811501597444, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 7977 + }, + { + "epoch": 12.744408945686901, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7978 + }, + { + "epoch": 12.746006389776358, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7979 + }, + { + "epoch": 12.747603833865815, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 7980 + }, + { + "epoch": 12.749201277955272, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 7981 + }, + { + "epoch": 12.750798722044728, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 7982 + }, + { + "epoch": 12.752396166134185, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7983 + }, + { + "epoch": 12.753993610223642, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 7984 + }, + { + "epoch": 12.755591054313099, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 7985 + }, + { + "epoch": 12.757188498402556, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 7986 + }, + { + "epoch": 12.758785942492013, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 7987 + }, + { + "epoch": 12.76038338658147, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 7988 + }, + { + "epoch": 12.761980830670927, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 7989 + }, + { + "epoch": 12.763578274760384, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 7990 + }, + { + "epoch": 12.76517571884984, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 7991 + }, + { + "epoch": 12.766773162939296, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 7992 + }, + { + "epoch": 12.768370607028753, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 7993 + }, + { + "epoch": 12.76996805111821, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 7994 + }, + { + "epoch": 12.771565495207668, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 7995 + }, + { + "epoch": 12.773162939297125, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 7996 + }, + { + "epoch": 12.774760383386582, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 7997 + }, + { + "epoch": 12.776357827476039, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 7998 + }, + { + "epoch": 12.777955271565496, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 7999 + }, + { + "epoch": 12.779552715654953, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8000 + }, + { + "epoch": 12.781150159744408, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8001 + }, + { + "epoch": 12.782747603833865, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8002 + }, + { + "epoch": 12.784345047923322, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 8003 + }, + { + "epoch": 12.78594249201278, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 8004 + }, + { + "epoch": 12.787539936102236, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8005 + }, + { + "epoch": 12.789137380191693, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 8006 + }, + { + "epoch": 12.79073482428115, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8007 + }, + { + "epoch": 12.792332268370608, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8008 + }, + { + "epoch": 12.793929712460065, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8009 + }, + { + "epoch": 12.795527156549522, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 8010 + }, + { + "epoch": 12.797124600638977, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8011 + }, + { + "epoch": 12.798722044728434, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 8012 + }, + { + "epoch": 12.800319488817891, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8013 + }, + { + "epoch": 12.801916932907348, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8014 + }, + { + "epoch": 12.803514376996805, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8015 + }, + { + "epoch": 12.805111821086262, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8016 + }, + { + "epoch": 12.80670926517572, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8017 + }, + { + "epoch": 12.808306709265176, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8018 + }, + { + "epoch": 12.809904153354633, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8019 + }, + { + "epoch": 12.811501597444089, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8020 + }, + { + "epoch": 12.813099041533546, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 8021 + }, + { + "epoch": 12.814696485623003, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8022 + }, + { + "epoch": 12.81629392971246, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8023 + }, + { + "epoch": 12.817891373801917, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8024 + }, + { + "epoch": 12.819488817891374, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8025 + }, + { + "epoch": 12.821086261980831, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 8026 + }, + { + "epoch": 12.822683706070288, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 8027 + }, + { + "epoch": 12.824281150159745, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8028 + }, + { + "epoch": 12.8258785942492, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8029 + }, + { + "epoch": 12.827476038338657, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8030 + }, + { + "epoch": 12.829073482428115, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8031 + }, + { + "epoch": 12.830670926517572, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8032 + }, + { + "epoch": 12.832268370607029, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8033 + }, + { + "epoch": 12.833865814696486, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 8034 + }, + { + "epoch": 12.835463258785943, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 8035 + }, + { + "epoch": 12.8370607028754, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8036 + }, + { + "epoch": 12.838658146964857, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8037 + }, + { + "epoch": 12.840255591054314, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8038 + }, + { + "epoch": 12.84185303514377, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8039 + }, + { + "epoch": 12.843450479233226, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8040 + }, + { + "epoch": 12.845047923322683, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8041 + }, + { + "epoch": 12.84664536741214, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8042 + }, + { + "epoch": 12.848242811501597, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8043 + }, + { + "epoch": 12.849840255591054, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 8044 + }, + { + "epoch": 12.851437699680512, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 8045 + }, + { + "epoch": 12.853035143769969, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8046 + }, + { + "epoch": 12.854632587859426, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8047 + }, + { + "epoch": 12.856230031948883, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8048 + }, + { + "epoch": 12.857827476038338, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8049 + }, + { + "epoch": 12.859424920127795, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8050 + }, + { + "epoch": 12.861022364217252, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8051 + }, + { + "epoch": 12.86261980830671, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8052 + }, + { + "epoch": 12.864217252396166, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8053 + }, + { + "epoch": 12.865814696485623, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8054 + }, + { + "epoch": 12.86741214057508, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8055 + }, + { + "epoch": 12.869009584664537, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8056 + }, + { + "epoch": 12.870607028753994, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8057 + }, + { + "epoch": 12.87220447284345, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8058 + }, + { + "epoch": 12.873801916932907, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8059 + }, + { + "epoch": 12.875399361022364, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8060 + }, + { + "epoch": 12.87699680511182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8061 + }, + { + "epoch": 12.878594249201278, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8062 + }, + { + "epoch": 12.880191693290735, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8063 + }, + { + "epoch": 12.881789137380192, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8064 + }, + { + "epoch": 12.883386581469649, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8065 + }, + { + "epoch": 12.884984025559106, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8066 + }, + { + "epoch": 12.886581469648561, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8067 + }, + { + "epoch": 12.888178913738018, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8068 + }, + { + "epoch": 12.889776357827476, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 8069 + }, + { + "epoch": 12.891373801916933, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8070 + }, + { + "epoch": 12.89297124600639, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8071 + }, + { + "epoch": 12.894568690095847, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8072 + }, + { + "epoch": 12.896166134185304, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8073 + }, + { + "epoch": 12.89776357827476, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8074 + }, + { + "epoch": 12.899361022364218, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8075 + }, + { + "epoch": 12.900958466453675, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8076 + }, + { + "epoch": 12.90255591054313, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8077 + }, + { + "epoch": 12.904153354632587, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 8078 + }, + { + "epoch": 12.905750798722044, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8079 + }, + { + "epoch": 12.907348242811501, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8080 + }, + { + "epoch": 12.908945686900958, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8081 + }, + { + "epoch": 12.910543130990416, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8082 + }, + { + "epoch": 12.912140575079873, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8083 + }, + { + "epoch": 12.91373801916933, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8084 + }, + { + "epoch": 12.915335463258787, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 8085 + }, + { + "epoch": 12.916932907348244, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8086 + }, + { + "epoch": 12.918530351437699, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8087 + }, + { + "epoch": 12.920127795527156, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 8088 + }, + { + "epoch": 12.921725239616613, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8089 + }, + { + "epoch": 12.92332268370607, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8090 + }, + { + "epoch": 12.924920127795527, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8091 + }, + { + "epoch": 12.926517571884984, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8092 + }, + { + "epoch": 12.928115015974441, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8093 + }, + { + "epoch": 12.929712460063898, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8094 + }, + { + "epoch": 12.931309904153355, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8095 + }, + { + "epoch": 12.93290734824281, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8096 + }, + { + "epoch": 12.934504792332268, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 8097 + }, + { + "epoch": 12.936102236421725, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8098 + }, + { + "epoch": 12.937699680511182, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8099 + }, + { + "epoch": 12.939297124600639, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8100 + }, + { + "epoch": 12.940894568690096, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8101 + }, + { + "epoch": 12.942492012779553, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8102 + }, + { + "epoch": 12.94408945686901, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8103 + }, + { + "epoch": 12.945686900958467, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8104 + }, + { + "epoch": 12.947284345047922, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8105 + }, + { + "epoch": 12.94888178913738, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8106 + }, + { + "epoch": 12.950479233226837, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8107 + }, + { + "epoch": 12.952076677316294, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8108 + }, + { + "epoch": 12.95367412140575, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8109 + }, + { + "epoch": 12.955271565495208, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 8110 + }, + { + "epoch": 12.956869009584665, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8111 + }, + { + "epoch": 12.958466453674122, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8112 + }, + { + "epoch": 12.960063897763579, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8113 + }, + { + "epoch": 12.961661341853034, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8114 + }, + { + "epoch": 12.963258785942491, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8115 + }, + { + "epoch": 12.964856230031948, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 8116 + }, + { + "epoch": 12.966453674121405, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8117 + }, + { + "epoch": 12.968051118210862, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8118 + }, + { + "epoch": 12.96964856230032, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8119 + }, + { + "epoch": 12.971246006389777, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 8120 + }, + { + "epoch": 12.972843450479234, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 8121 + }, + { + "epoch": 12.97444089456869, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 8122 + }, + { + "epoch": 12.976038338658148, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8123 + }, + { + "epoch": 12.977635782747605, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 8124 + }, + { + "epoch": 12.97923322683706, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8125 + }, + { + "epoch": 12.980830670926517, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8126 + }, + { + "epoch": 12.982428115015974, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8127 + }, + { + "epoch": 12.984025559105431, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8128 + }, + { + "epoch": 12.985623003194888, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8129 + }, + { + "epoch": 12.987220447284345, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8130 + }, + { + "epoch": 12.988817891373802, + "grad_norm": 0.51171875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8131 + }, + { + "epoch": 12.99041533546326, + "grad_norm": 1.25, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 8132 + }, + { + "epoch": 12.992012779552716, + "grad_norm": 1.515625, + "learning_rate": 0.0005, + "loss": 1.0956, + "step": 8133 + }, + { + "epoch": 12.993610223642172, + "grad_norm": 0.41015625, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 8134 + }, + { + "epoch": 12.995207667731629, + "grad_norm": 1.1640625, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 8135 + }, + { + "epoch": 12.996805111821086, + "grad_norm": 0.71875, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 8136 + }, + { + "epoch": 12.998402555910543, + "grad_norm": 0.8046875, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 8137 + }, + { + "epoch": 13.0, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 8138 + }, + { + "epoch": 13.001597444089457, + "grad_norm": 0.73046875, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 8139 + }, + { + "epoch": 13.003194888178914, + "grad_norm": 1.0546875, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 8140 + }, + { + "epoch": 13.004792332268371, + "grad_norm": 0.87109375, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 8141 + }, + { + "epoch": 13.006389776357828, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 8142 + }, + { + "epoch": 13.007987220447284, + "grad_norm": 0.6015625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 8143 + }, + { + "epoch": 13.00958466453674, + "grad_norm": 0.5, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 8144 + }, + { + "epoch": 13.011182108626198, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 8145 + }, + { + "epoch": 13.012779552715655, + "grad_norm": 0.38671875, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 8146 + }, + { + "epoch": 13.014376996805112, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 8147 + }, + { + "epoch": 13.015974440894569, + "grad_norm": 0.33203125, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 8148 + }, + { + "epoch": 13.017571884984026, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 8149 + }, + { + "epoch": 13.019169329073483, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 8150 + }, + { + "epoch": 13.02076677316294, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8151 + }, + { + "epoch": 13.022364217252397, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 8152 + }, + { + "epoch": 13.023961661341852, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 8153 + }, + { + "epoch": 13.02555910543131, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8154 + }, + { + "epoch": 13.027156549520766, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8155 + }, + { + "epoch": 13.028753993610223, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 8156 + }, + { + "epoch": 13.03035143769968, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8157 + }, + { + "epoch": 13.031948881789138, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 8158 + }, + { + "epoch": 13.033546325878595, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 8159 + }, + { + "epoch": 13.035143769968052, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8160 + }, + { + "epoch": 13.036741214057509, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 8161 + }, + { + "epoch": 13.038338658146964, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8162 + }, + { + "epoch": 13.039936102236421, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8163 + }, + { + "epoch": 13.041533546325878, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8164 + }, + { + "epoch": 13.043130990415335, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8165 + }, + { + "epoch": 13.044728434504792, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 8166 + }, + { + "epoch": 13.04632587859425, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8167 + }, + { + "epoch": 13.047923322683706, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8168 + }, + { + "epoch": 13.049520766773163, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8169 + }, + { + "epoch": 13.05111821086262, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8170 + }, + { + "epoch": 13.052715654952078, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8171 + }, + { + "epoch": 13.054313099041533, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8172 + }, + { + "epoch": 13.05591054313099, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8173 + }, + { + "epoch": 13.057507987220447, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8174 + }, + { + "epoch": 13.059105431309904, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8175 + }, + { + "epoch": 13.060702875399361, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 8176 + }, + { + "epoch": 13.062300319488818, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8177 + }, + { + "epoch": 13.063897763578275, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 8178 + }, + { + "epoch": 13.065495207667732, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8179 + }, + { + "epoch": 13.06709265175719, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8180 + }, + { + "epoch": 13.068690095846645, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8181 + }, + { + "epoch": 13.070287539936102, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8182 + }, + { + "epoch": 13.071884984025559, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8183 + }, + { + "epoch": 13.073482428115016, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8184 + }, + { + "epoch": 13.075079872204473, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8185 + }, + { + "epoch": 13.07667731629393, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8186 + }, + { + "epoch": 13.078274760383387, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8187 + }, + { + "epoch": 13.079872204472844, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8188 + }, + { + "epoch": 13.081469648562301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8189 + }, + { + "epoch": 13.083067092651758, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8190 + }, + { + "epoch": 13.084664536741213, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8191 + }, + { + "epoch": 13.08626198083067, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8192 + }, + { + "epoch": 13.087859424920127, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8193 + }, + { + "epoch": 13.089456869009584, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8194 + }, + { + "epoch": 13.091054313099042, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8195 + }, + { + "epoch": 13.092651757188499, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8196 + }, + { + "epoch": 13.094249201277956, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8197 + }, + { + "epoch": 13.095846645367413, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8198 + }, + { + "epoch": 13.09744408945687, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8199 + }, + { + "epoch": 13.099041533546325, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8200 + }, + { + "epoch": 13.100638977635782, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8201 + }, + { + "epoch": 13.10223642172524, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8202 + }, + { + "epoch": 13.103833865814696, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 8203 + }, + { + "epoch": 13.105431309904153, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8204 + }, + { + "epoch": 13.10702875399361, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8205 + }, + { + "epoch": 13.108626198083067, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8206 + }, + { + "epoch": 13.110223642172524, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8207 + }, + { + "epoch": 13.111821086261982, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8208 + }, + { + "epoch": 13.113418530351439, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8209 + }, + { + "epoch": 13.115015974440894, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 8210 + }, + { + "epoch": 13.116613418530351, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8211 + }, + { + "epoch": 13.118210862619808, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8212 + }, + { + "epoch": 13.119808306709265, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8213 + }, + { + "epoch": 13.121405750798722, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8214 + }, + { + "epoch": 13.12300319488818, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8215 + }, + { + "epoch": 13.124600638977636, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 8216 + }, + { + "epoch": 13.126198083067093, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8217 + }, + { + "epoch": 13.12779552715655, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8218 + }, + { + "epoch": 13.129392971246006, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8219 + }, + { + "epoch": 13.130990415335463, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8220 + }, + { + "epoch": 13.13258785942492, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8221 + }, + { + "epoch": 13.134185303514377, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8222 + }, + { + "epoch": 13.135782747603834, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8223 + }, + { + "epoch": 13.13738019169329, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8224 + }, + { + "epoch": 13.138977635782748, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8225 + }, + { + "epoch": 13.140575079872205, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8226 + }, + { + "epoch": 13.142172523961662, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8227 + }, + { + "epoch": 13.143769968051119, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8228 + }, + { + "epoch": 13.145367412140574, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8229 + }, + { + "epoch": 13.146964856230031, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 8230 + }, + { + "epoch": 13.148562300319488, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 8231 + }, + { + "epoch": 13.150159744408946, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8232 + }, + { + "epoch": 13.151757188498403, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8233 + }, + { + "epoch": 13.15335463258786, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8234 + }, + { + "epoch": 13.154952076677317, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8235 + }, + { + "epoch": 13.156549520766774, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8236 + }, + { + "epoch": 13.15814696485623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8237 + }, + { + "epoch": 13.159744408945686, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8238 + }, + { + "epoch": 13.161341853035143, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8239 + }, + { + "epoch": 13.1629392971246, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 8240 + }, + { + "epoch": 13.164536741214057, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 8241 + }, + { + "epoch": 13.166134185303514, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 8242 + }, + { + "epoch": 13.167731629392971, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8243 + }, + { + "epoch": 13.169329073482428, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8244 + }, + { + "epoch": 13.170926517571885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8245 + }, + { + "epoch": 13.172523961661343, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8246 + }, + { + "epoch": 13.1741214057508, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8247 + }, + { + "epoch": 13.175718849840255, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8248 + }, + { + "epoch": 13.177316293929712, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8249 + }, + { + "epoch": 13.178913738019169, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8250 + }, + { + "epoch": 13.180511182108626, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8251 + }, + { + "epoch": 13.182108626198083, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8252 + }, + { + "epoch": 13.18370607028754, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8253 + }, + { + "epoch": 13.185303514376997, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8254 + }, + { + "epoch": 13.186900958466454, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8255 + }, + { + "epoch": 13.188498402555911, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8256 + }, + { + "epoch": 13.190095846645367, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8257 + }, + { + "epoch": 13.191693290734824, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8258 + }, + { + "epoch": 13.19329073482428, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8259 + }, + { + "epoch": 13.194888178913738, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8260 + }, + { + "epoch": 13.196485623003195, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8261 + }, + { + "epoch": 13.198083067092652, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 8262 + }, + { + "epoch": 13.199680511182109, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8263 + }, + { + "epoch": 13.201277955271566, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8264 + }, + { + "epoch": 13.202875399361023, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8265 + }, + { + "epoch": 13.204472843450478, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 8266 + }, + { + "epoch": 13.206070287539935, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8267 + }, + { + "epoch": 13.207667731629392, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8268 + }, + { + "epoch": 13.20926517571885, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 8269 + }, + { + "epoch": 13.210862619808307, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8270 + }, + { + "epoch": 13.212460063897764, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8271 + }, + { + "epoch": 13.21405750798722, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8272 + }, + { + "epoch": 13.215654952076678, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 8273 + }, + { + "epoch": 13.217252396166135, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 8274 + }, + { + "epoch": 13.218849840255592, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8275 + }, + { + "epoch": 13.220447284345047, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8276 + }, + { + "epoch": 13.222044728434504, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8277 + }, + { + "epoch": 13.223642172523961, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 8278 + }, + { + "epoch": 13.225239616613418, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8279 + }, + { + "epoch": 13.226837060702875, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8280 + }, + { + "epoch": 13.228434504792332, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 8281 + }, + { + "epoch": 13.23003194888179, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8282 + }, + { + "epoch": 13.231629392971247, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8283 + }, + { + "epoch": 13.233226837060704, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8284 + }, + { + "epoch": 13.23482428115016, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8285 + }, + { + "epoch": 13.236421725239616, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8286 + }, + { + "epoch": 13.238019169329073, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8287 + }, + { + "epoch": 13.23961661341853, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8288 + }, + { + "epoch": 13.241214057507987, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8289 + }, + { + "epoch": 13.242811501597444, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8290 + }, + { + "epoch": 13.244408945686901, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8291 + }, + { + "epoch": 13.246006389776358, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8292 + }, + { + "epoch": 13.247603833865815, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8293 + }, + { + "epoch": 13.249201277955272, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8294 + }, + { + "epoch": 13.250798722044728, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8295 + }, + { + "epoch": 13.252396166134185, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8296 + }, + { + "epoch": 13.253993610223642, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8297 + }, + { + "epoch": 13.255591054313099, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8298 + }, + { + "epoch": 13.257188498402556, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8299 + }, + { + "epoch": 13.258785942492013, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8300 + }, + { + "epoch": 13.26038338658147, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8301 + }, + { + "epoch": 13.261980830670927, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8302 + }, + { + "epoch": 13.263578274760384, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8303 + }, + { + "epoch": 13.26517571884984, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8304 + }, + { + "epoch": 13.266773162939296, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8305 + }, + { + "epoch": 13.268370607028753, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8306 + }, + { + "epoch": 13.26996805111821, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8307 + }, + { + "epoch": 13.271565495207668, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8308 + }, + { + "epoch": 13.273162939297125, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8309 + }, + { + "epoch": 13.274760383386582, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8310 + }, + { + "epoch": 13.276357827476039, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8311 + }, + { + "epoch": 13.277955271565496, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8312 + }, + { + "epoch": 13.279552715654953, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8313 + }, + { + "epoch": 13.281150159744408, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8314 + }, + { + "epoch": 13.282747603833865, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 8315 + }, + { + "epoch": 13.284345047923322, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8316 + }, + { + "epoch": 13.28594249201278, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8317 + }, + { + "epoch": 13.287539936102236, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8318 + }, + { + "epoch": 13.289137380191693, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8319 + }, + { + "epoch": 13.29073482428115, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8320 + }, + { + "epoch": 13.292332268370608, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8321 + }, + { + "epoch": 13.293929712460065, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8322 + }, + { + "epoch": 13.295527156549522, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8323 + }, + { + "epoch": 13.297124600638977, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8324 + }, + { + "epoch": 13.298722044728434, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8325 + }, + { + "epoch": 13.300319488817891, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8326 + }, + { + "epoch": 13.301916932907348, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8327 + }, + { + "epoch": 13.303514376996805, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8328 + }, + { + "epoch": 13.305111821086262, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 8329 + }, + { + "epoch": 13.30670926517572, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8330 + }, + { + "epoch": 13.308306709265176, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8331 + }, + { + "epoch": 13.309904153354633, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8332 + }, + { + "epoch": 13.311501597444089, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8333 + }, + { + "epoch": 13.313099041533546, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 8334 + }, + { + "epoch": 13.314696485623003, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8335 + }, + { + "epoch": 13.31629392971246, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 8336 + }, + { + "epoch": 13.317891373801917, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8337 + }, + { + "epoch": 13.319488817891374, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8338 + }, + { + "epoch": 13.321086261980831, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8339 + }, + { + "epoch": 13.322683706070288, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 8340 + }, + { + "epoch": 13.324281150159745, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8341 + }, + { + "epoch": 13.3258785942492, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8342 + }, + { + "epoch": 13.327476038338657, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8343 + }, + { + "epoch": 13.329073482428115, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8344 + }, + { + "epoch": 13.330670926517572, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8345 + }, + { + "epoch": 13.332268370607029, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8346 + }, + { + "epoch": 13.333865814696486, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8347 + }, + { + "epoch": 13.335463258785943, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8348 + }, + { + "epoch": 13.3370607028754, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8349 + }, + { + "epoch": 13.338658146964857, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 8350 + }, + { + "epoch": 13.340255591054314, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8351 + }, + { + "epoch": 13.34185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8352 + }, + { + "epoch": 13.343450479233226, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8353 + }, + { + "epoch": 13.345047923322683, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8354 + }, + { + "epoch": 13.34664536741214, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8355 + }, + { + "epoch": 13.348242811501597, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8356 + }, + { + "epoch": 13.349840255591054, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8357 + }, + { + "epoch": 13.351437699680512, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8358 + }, + { + "epoch": 13.353035143769969, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8359 + }, + { + "epoch": 13.354632587859426, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8360 + }, + { + "epoch": 13.356230031948883, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8361 + }, + { + "epoch": 13.357827476038338, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8362 + }, + { + "epoch": 13.359424920127795, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8363 + }, + { + "epoch": 13.361022364217252, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8364 + }, + { + "epoch": 13.36261980830671, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8365 + }, + { + "epoch": 13.364217252396166, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8366 + }, + { + "epoch": 13.365814696485623, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8367 + }, + { + "epoch": 13.36741214057508, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8368 + }, + { + "epoch": 13.369009584664537, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 8369 + }, + { + "epoch": 13.370607028753994, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8370 + }, + { + "epoch": 13.37220447284345, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8371 + }, + { + "epoch": 13.373801916932907, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8372 + }, + { + "epoch": 13.375399361022364, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8373 + }, + { + "epoch": 13.37699680511182, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8374 + }, + { + "epoch": 13.378594249201278, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8375 + }, + { + "epoch": 13.380191693290735, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8376 + }, + { + "epoch": 13.381789137380192, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8377 + }, + { + "epoch": 13.383386581469649, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8378 + }, + { + "epoch": 13.384984025559106, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8379 + }, + { + "epoch": 13.386581469648561, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8380 + }, + { + "epoch": 13.388178913738018, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 8381 + }, + { + "epoch": 13.389776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8382 + }, + { + "epoch": 13.391373801916933, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8383 + }, + { + "epoch": 13.39297124600639, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8384 + }, + { + "epoch": 13.394568690095847, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8385 + }, + { + "epoch": 13.396166134185304, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 8386 + }, + { + "epoch": 13.39776357827476, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8387 + }, + { + "epoch": 13.399361022364218, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 8388 + }, + { + "epoch": 13.400958466453675, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8389 + }, + { + "epoch": 13.40255591054313, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8390 + }, + { + "epoch": 13.404153354632587, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8391 + }, + { + "epoch": 13.405750798722044, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8392 + }, + { + "epoch": 13.407348242811501, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8393 + }, + { + "epoch": 13.408945686900958, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 8394 + }, + { + "epoch": 13.410543130990416, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8395 + }, + { + "epoch": 13.412140575079873, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8396 + }, + { + "epoch": 13.41373801916933, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8397 + }, + { + "epoch": 13.415335463258787, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8398 + }, + { + "epoch": 13.416932907348242, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8399 + }, + { + "epoch": 13.418530351437699, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8400 + }, + { + "epoch": 13.420127795527156, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8401 + }, + { + "epoch": 13.421725239616613, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8402 + }, + { + "epoch": 13.42332268370607, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8403 + }, + { + "epoch": 13.424920127795527, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8404 + }, + { + "epoch": 13.426517571884984, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8405 + }, + { + "epoch": 13.428115015974441, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8406 + }, + { + "epoch": 13.429712460063898, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8407 + }, + { + "epoch": 13.431309904153355, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8408 + }, + { + "epoch": 13.43290734824281, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8409 + }, + { + "epoch": 13.434504792332268, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8410 + }, + { + "epoch": 13.436102236421725, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8411 + }, + { + "epoch": 13.437699680511182, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 8412 + }, + { + "epoch": 13.439297124600639, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8413 + }, + { + "epoch": 13.440894568690096, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8414 + }, + { + "epoch": 13.442492012779553, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8415 + }, + { + "epoch": 13.44408945686901, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8416 + }, + { + "epoch": 13.445686900958467, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8417 + }, + { + "epoch": 13.447284345047922, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8418 + }, + { + "epoch": 13.44888178913738, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8419 + }, + { + "epoch": 13.450479233226837, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8420 + }, + { + "epoch": 13.452076677316294, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8421 + }, + { + "epoch": 13.45367412140575, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 8422 + }, + { + "epoch": 13.455271565495208, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8423 + }, + { + "epoch": 13.456869009584665, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8424 + }, + { + "epoch": 13.458466453674122, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8425 + }, + { + "epoch": 13.460063897763579, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8426 + }, + { + "epoch": 13.461661341853036, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8427 + }, + { + "epoch": 13.463258785942491, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8428 + }, + { + "epoch": 13.464856230031948, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8429 + }, + { + "epoch": 13.466453674121405, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8430 + }, + { + "epoch": 13.468051118210862, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8431 + }, + { + "epoch": 13.46964856230032, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8432 + }, + { + "epoch": 13.471246006389777, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8433 + }, + { + "epoch": 13.472843450479234, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8434 + }, + { + "epoch": 13.47444089456869, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8435 + }, + { + "epoch": 13.476038338658148, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8436 + }, + { + "epoch": 13.477635782747603, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 8437 + }, + { + "epoch": 13.47923322683706, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8438 + }, + { + "epoch": 13.480830670926517, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8439 + }, + { + "epoch": 13.482428115015974, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8440 + }, + { + "epoch": 13.484025559105431, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8441 + }, + { + "epoch": 13.485623003194888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8442 + }, + { + "epoch": 13.487220447284345, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 8443 + }, + { + "epoch": 13.488817891373802, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8444 + }, + { + "epoch": 13.49041533546326, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8445 + }, + { + "epoch": 13.492012779552716, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8446 + }, + { + "epoch": 13.493610223642172, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8447 + }, + { + "epoch": 13.495207667731629, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 8448 + }, + { + "epoch": 13.496805111821086, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8449 + }, + { + "epoch": 13.498402555910543, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8450 + }, + { + "epoch": 13.5, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8451 + }, + { + "epoch": 13.501597444089457, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 8452 + }, + { + "epoch": 13.503194888178914, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8453 + }, + { + "epoch": 13.504792332268371, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8454 + }, + { + "epoch": 13.506389776357828, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8455 + }, + { + "epoch": 13.507987220447284, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8456 + }, + { + "epoch": 13.50958466453674, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8457 + }, + { + "epoch": 13.511182108626198, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8458 + }, + { + "epoch": 13.512779552715655, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8459 + }, + { + "epoch": 13.514376996805112, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 8460 + }, + { + "epoch": 13.515974440894569, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8461 + }, + { + "epoch": 13.517571884984026, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8462 + }, + { + "epoch": 13.519169329073483, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8463 + }, + { + "epoch": 13.52076677316294, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8464 + }, + { + "epoch": 13.522364217252395, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8465 + }, + { + "epoch": 13.523961661341852, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8466 + }, + { + "epoch": 13.52555910543131, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 8467 + }, + { + "epoch": 13.527156549520766, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8468 + }, + { + "epoch": 13.528753993610223, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 8469 + }, + { + "epoch": 13.53035143769968, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8470 + }, + { + "epoch": 13.531948881789138, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8471 + }, + { + "epoch": 13.533546325878595, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8472 + }, + { + "epoch": 13.535143769968052, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8473 + }, + { + "epoch": 13.536741214057509, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8474 + }, + { + "epoch": 13.538338658146966, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8475 + }, + { + "epoch": 13.539936102236421, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8476 + }, + { + "epoch": 13.541533546325878, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8477 + }, + { + "epoch": 13.543130990415335, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 8478 + }, + { + "epoch": 13.544728434504792, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8479 + }, + { + "epoch": 13.54632587859425, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8480 + }, + { + "epoch": 13.547923322683706, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8481 + }, + { + "epoch": 13.549520766773163, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8482 + }, + { + "epoch": 13.55111821086262, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8483 + }, + { + "epoch": 13.552715654952078, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 8484 + }, + { + "epoch": 13.554313099041533, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8485 + }, + { + "epoch": 13.55591054313099, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8486 + }, + { + "epoch": 13.557507987220447, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8487 + }, + { + "epoch": 13.559105431309904, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8488 + }, + { + "epoch": 13.560702875399361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8489 + }, + { + "epoch": 13.562300319488818, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8490 + }, + { + "epoch": 13.563897763578275, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8491 + }, + { + "epoch": 13.565495207667732, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8492 + }, + { + "epoch": 13.56709265175719, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8493 + }, + { + "epoch": 13.568690095846645, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8494 + }, + { + "epoch": 13.570287539936102, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8495 + }, + { + "epoch": 13.571884984025559, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8496 + }, + { + "epoch": 13.573482428115016, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8497 + }, + { + "epoch": 13.575079872204473, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8498 + }, + { + "epoch": 13.57667731629393, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8499 + }, + { + "epoch": 13.578274760383387, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8500 + }, + { + "epoch": 13.579872204472844, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 8501 + }, + { + "epoch": 13.581469648562301, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8502 + }, + { + "epoch": 13.583067092651756, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8503 + }, + { + "epoch": 13.584664536741213, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8504 + }, + { + "epoch": 13.58626198083067, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8505 + }, + { + "epoch": 13.587859424920127, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8506 + }, + { + "epoch": 13.589456869009584, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8507 + }, + { + "epoch": 13.591054313099042, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8508 + }, + { + "epoch": 13.592651757188499, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8509 + }, + { + "epoch": 13.594249201277956, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8510 + }, + { + "epoch": 13.595846645367413, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8511 + }, + { + "epoch": 13.59744408945687, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8512 + }, + { + "epoch": 13.599041533546325, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8513 + }, + { + "epoch": 13.600638977635782, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8514 + }, + { + "epoch": 13.60223642172524, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8515 + }, + { + "epoch": 13.603833865814696, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8516 + }, + { + "epoch": 13.605431309904153, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8517 + }, + { + "epoch": 13.60702875399361, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8518 + }, + { + "epoch": 13.608626198083067, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8519 + }, + { + "epoch": 13.610223642172524, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8520 + }, + { + "epoch": 13.611821086261982, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8521 + }, + { + "epoch": 13.613418530351439, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8522 + }, + { + "epoch": 13.615015974440894, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8523 + }, + { + "epoch": 13.616613418530351, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8524 + }, + { + "epoch": 13.618210862619808, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8525 + }, + { + "epoch": 13.619808306709265, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 8526 + }, + { + "epoch": 13.621405750798722, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8527 + }, + { + "epoch": 13.62300319488818, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8528 + }, + { + "epoch": 13.624600638977636, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8529 + }, + { + "epoch": 13.626198083067093, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8530 + }, + { + "epoch": 13.62779552715655, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8531 + }, + { + "epoch": 13.629392971246006, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 8532 + }, + { + "epoch": 13.630990415335463, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8533 + }, + { + "epoch": 13.63258785942492, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8534 + }, + { + "epoch": 13.634185303514377, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8535 + }, + { + "epoch": 13.635782747603834, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8536 + }, + { + "epoch": 13.63738019169329, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8537 + }, + { + "epoch": 13.638977635782748, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8538 + }, + { + "epoch": 13.640575079872205, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8539 + }, + { + "epoch": 13.642172523961662, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8540 + }, + { + "epoch": 13.643769968051117, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8541 + }, + { + "epoch": 13.645367412140574, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8542 + }, + { + "epoch": 13.646964856230031, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8543 + }, + { + "epoch": 13.648562300319488, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8544 + }, + { + "epoch": 13.650159744408946, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8545 + }, + { + "epoch": 13.651757188498403, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8546 + }, + { + "epoch": 13.65335463258786, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8547 + }, + { + "epoch": 13.654952076677317, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 8548 + }, + { + "epoch": 13.656549520766774, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8549 + }, + { + "epoch": 13.65814696485623, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8550 + }, + { + "epoch": 13.659744408945686, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 8551 + }, + { + "epoch": 13.661341853035143, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8552 + }, + { + "epoch": 13.6629392971246, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8553 + }, + { + "epoch": 13.664536741214057, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8554 + }, + { + "epoch": 13.666134185303514, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8555 + }, + { + "epoch": 13.667731629392971, + "grad_norm": 0.03173828125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8556 + }, + { + "epoch": 13.669329073482428, + "grad_norm": 0.031494140625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8557 + }, + { + "epoch": 13.670926517571885, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8558 + }, + { + "epoch": 13.672523961661343, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8559 + }, + { + "epoch": 13.6741214057508, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8560 + }, + { + "epoch": 13.675718849840255, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8561 + }, + { + "epoch": 13.677316293929712, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8562 + }, + { + "epoch": 13.678913738019169, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8563 + }, + { + "epoch": 13.680511182108626, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8564 + }, + { + "epoch": 13.682108626198083, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 8565 + }, + { + "epoch": 13.68370607028754, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8566 + }, + { + "epoch": 13.685303514376997, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 8567 + }, + { + "epoch": 13.686900958466454, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8568 + }, + { + "epoch": 13.688498402555911, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8569 + }, + { + "epoch": 13.690095846645367, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8570 + }, + { + "epoch": 13.691693290734824, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8571 + }, + { + "epoch": 13.69329073482428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 8572 + }, + { + "epoch": 13.694888178913738, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8573 + }, + { + "epoch": 13.696485623003195, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8574 + }, + { + "epoch": 13.698083067092652, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8575 + }, + { + "epoch": 13.699680511182109, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8576 + }, + { + "epoch": 13.701277955271566, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 8577 + }, + { + "epoch": 13.702875399361023, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8578 + }, + { + "epoch": 13.704472843450478, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8579 + }, + { + "epoch": 13.706070287539935, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8580 + }, + { + "epoch": 13.707667731629392, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8581 + }, + { + "epoch": 13.70926517571885, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8582 + }, + { + "epoch": 13.710862619808307, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8583 + }, + { + "epoch": 13.712460063897764, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8584 + }, + { + "epoch": 13.71405750798722, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8585 + }, + { + "epoch": 13.715654952076678, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8586 + }, + { + "epoch": 13.717252396166135, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 8587 + }, + { + "epoch": 13.718849840255592, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8588 + }, + { + "epoch": 13.720447284345047, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 8589 + }, + { + "epoch": 13.722044728434504, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8590 + }, + { + "epoch": 13.723642172523961, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8591 + }, + { + "epoch": 13.725239616613418, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8592 + }, + { + "epoch": 13.726837060702875, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8593 + }, + { + "epoch": 13.728434504792332, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8594 + }, + { + "epoch": 13.73003194888179, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8595 + }, + { + "epoch": 13.731629392971247, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8596 + }, + { + "epoch": 13.733226837060704, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8597 + }, + { + "epoch": 13.73482428115016, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 8598 + }, + { + "epoch": 13.736421725239616, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8599 + }, + { + "epoch": 13.738019169329073, + "grad_norm": 0.031494140625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8600 + }, + { + "epoch": 13.73961661341853, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8601 + }, + { + "epoch": 13.741214057507987, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8602 + }, + { + "epoch": 13.742811501597444, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8603 + }, + { + "epoch": 13.744408945686901, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8604 + }, + { + "epoch": 13.746006389776358, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8605 + }, + { + "epoch": 13.747603833865815, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 8606 + }, + { + "epoch": 13.749201277955272, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 8607 + }, + { + "epoch": 13.750798722044728, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8608 + }, + { + "epoch": 13.752396166134185, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8609 + }, + { + "epoch": 13.753993610223642, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8610 + }, + { + "epoch": 13.755591054313099, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8611 + }, + { + "epoch": 13.757188498402556, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8612 + }, + { + "epoch": 13.758785942492013, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8613 + }, + { + "epoch": 13.76038338658147, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8614 + }, + { + "epoch": 13.761980830670927, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8615 + }, + { + "epoch": 13.763578274760384, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8616 + }, + { + "epoch": 13.76517571884984, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8617 + }, + { + "epoch": 13.766773162939296, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8618 + }, + { + "epoch": 13.768370607028753, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8619 + }, + { + "epoch": 13.76996805111821, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8620 + }, + { + "epoch": 13.771565495207668, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8621 + }, + { + "epoch": 13.773162939297125, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8622 + }, + { + "epoch": 13.774760383386582, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 8623 + }, + { + "epoch": 13.776357827476039, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 8624 + }, + { + "epoch": 13.777955271565496, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8625 + }, + { + "epoch": 13.779552715654953, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 8626 + }, + { + "epoch": 13.781150159744408, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 8627 + }, + { + "epoch": 13.782747603833865, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8628 + }, + { + "epoch": 13.784345047923322, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8629 + }, + { + "epoch": 13.78594249201278, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8630 + }, + { + "epoch": 13.787539936102236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8631 + }, + { + "epoch": 13.789137380191693, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8632 + }, + { + "epoch": 13.79073482428115, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8633 + }, + { + "epoch": 13.792332268370608, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8634 + }, + { + "epoch": 13.793929712460065, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8635 + }, + { + "epoch": 13.795527156549522, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 8636 + }, + { + "epoch": 13.797124600638977, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8637 + }, + { + "epoch": 13.798722044728434, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8638 + }, + { + "epoch": 13.800319488817891, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8639 + }, + { + "epoch": 13.801916932907348, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 8640 + }, + { + "epoch": 13.803514376996805, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8641 + }, + { + "epoch": 13.805111821086262, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8642 + }, + { + "epoch": 13.80670926517572, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8643 + }, + { + "epoch": 13.808306709265176, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8644 + }, + { + "epoch": 13.809904153354633, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 8645 + }, + { + "epoch": 13.811501597444089, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8646 + }, + { + "epoch": 13.813099041533546, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8647 + }, + { + "epoch": 13.814696485623003, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8648 + }, + { + "epoch": 13.81629392971246, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8649 + }, + { + "epoch": 13.817891373801917, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8650 + }, + { + "epoch": 13.819488817891374, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 8651 + }, + { + "epoch": 13.821086261980831, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8652 + }, + { + "epoch": 13.822683706070288, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8653 + }, + { + "epoch": 13.824281150159745, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8654 + }, + { + "epoch": 13.8258785942492, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8655 + }, + { + "epoch": 13.827476038338657, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8656 + }, + { + "epoch": 13.829073482428115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8657 + }, + { + "epoch": 13.830670926517572, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8658 + }, + { + "epoch": 13.832268370607029, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 8659 + }, + { + "epoch": 13.833865814696486, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 8660 + }, + { + "epoch": 13.835463258785943, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8661 + }, + { + "epoch": 13.8370607028754, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8662 + }, + { + "epoch": 13.838658146964857, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8663 + }, + { + "epoch": 13.840255591054314, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8664 + }, + { + "epoch": 13.84185303514377, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8665 + }, + { + "epoch": 13.843450479233226, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8666 + }, + { + "epoch": 13.845047923322683, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8667 + }, + { + "epoch": 13.84664536741214, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8668 + }, + { + "epoch": 13.848242811501597, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8669 + }, + { + "epoch": 13.849840255591054, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8670 + }, + { + "epoch": 13.851437699680512, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8671 + }, + { + "epoch": 13.853035143769969, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8672 + }, + { + "epoch": 13.854632587859426, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8673 + }, + { + "epoch": 13.856230031948883, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8674 + }, + { + "epoch": 13.857827476038338, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8675 + }, + { + "epoch": 13.859424920127795, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8676 + }, + { + "epoch": 13.861022364217252, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8677 + }, + { + "epoch": 13.86261980830671, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8678 + }, + { + "epoch": 13.864217252396166, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8679 + }, + { + "epoch": 13.865814696485623, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8680 + }, + { + "epoch": 13.86741214057508, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8681 + }, + { + "epoch": 13.869009584664537, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8682 + }, + { + "epoch": 13.870607028753994, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8683 + }, + { + "epoch": 13.87220447284345, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 8684 + }, + { + "epoch": 13.873801916932907, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 8685 + }, + { + "epoch": 13.875399361022364, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8686 + }, + { + "epoch": 13.87699680511182, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8687 + }, + { + "epoch": 13.878594249201278, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8688 + }, + { + "epoch": 13.880191693290735, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8689 + }, + { + "epoch": 13.881789137380192, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 8690 + }, + { + "epoch": 13.883386581469649, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8691 + }, + { + "epoch": 13.884984025559106, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8692 + }, + { + "epoch": 13.886581469648561, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8693 + }, + { + "epoch": 13.888178913738018, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8694 + }, + { + "epoch": 13.889776357827476, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8695 + }, + { + "epoch": 13.891373801916933, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8696 + }, + { + "epoch": 13.89297124600639, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8697 + }, + { + "epoch": 13.894568690095847, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8698 + }, + { + "epoch": 13.896166134185304, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8699 + }, + { + "epoch": 13.89776357827476, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8700 + }, + { + "epoch": 13.899361022364218, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8701 + }, + { + "epoch": 13.900958466453675, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8702 + }, + { + "epoch": 13.90255591054313, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8703 + }, + { + "epoch": 13.904153354632587, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 8704 + }, + { + "epoch": 13.905750798722044, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8705 + }, + { + "epoch": 13.907348242811501, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 8706 + }, + { + "epoch": 13.908945686900958, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8707 + }, + { + "epoch": 13.910543130990416, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8708 + }, + { + "epoch": 13.912140575079873, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8709 + }, + { + "epoch": 13.91373801916933, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8710 + }, + { + "epoch": 13.915335463258787, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 8711 + }, + { + "epoch": 13.916932907348244, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 8712 + }, + { + "epoch": 13.918530351437699, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 8713 + }, + { + "epoch": 13.920127795527156, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8714 + }, + { + "epoch": 13.921725239616613, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8715 + }, + { + "epoch": 13.92332268370607, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8716 + }, + { + "epoch": 13.924920127795527, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8717 + }, + { + "epoch": 13.926517571884984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 8718 + }, + { + "epoch": 13.928115015974441, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8719 + }, + { + "epoch": 13.929712460063898, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8720 + }, + { + "epoch": 13.931309904153355, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8721 + }, + { + "epoch": 13.93290734824281, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8722 + }, + { + "epoch": 13.934504792332268, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8723 + }, + { + "epoch": 13.936102236421725, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8724 + }, + { + "epoch": 13.937699680511182, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 8725 + }, + { + "epoch": 13.939297124600639, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8726 + }, + { + "epoch": 13.940894568690096, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8727 + }, + { + "epoch": 13.942492012779553, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8728 + }, + { + "epoch": 13.94408945686901, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8729 + }, + { + "epoch": 13.945686900958467, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8730 + }, + { + "epoch": 13.947284345047922, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8731 + }, + { + "epoch": 13.94888178913738, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 8732 + }, + { + "epoch": 13.950479233226837, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 8733 + }, + { + "epoch": 13.952076677316294, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8734 + }, + { + "epoch": 13.95367412140575, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8735 + }, + { + "epoch": 13.955271565495208, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8736 + }, + { + "epoch": 13.956869009584665, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8737 + }, + { + "epoch": 13.958466453674122, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8738 + }, + { + "epoch": 13.960063897763579, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8739 + }, + { + "epoch": 13.961661341853034, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8740 + }, + { + "epoch": 13.963258785942491, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8741 + }, + { + "epoch": 13.964856230031948, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8742 + }, + { + "epoch": 13.966453674121405, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8743 + }, + { + "epoch": 13.968051118210862, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8744 + }, + { + "epoch": 13.96964856230032, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8745 + }, + { + "epoch": 13.971246006389777, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8746 + }, + { + "epoch": 13.972843450479234, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 8747 + }, + { + "epoch": 13.97444089456869, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8748 + }, + { + "epoch": 13.976038338658148, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8749 + }, + { + "epoch": 13.977635782747605, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8750 + }, + { + "epoch": 13.97923322683706, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8751 + }, + { + "epoch": 13.980830670926517, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8752 + }, + { + "epoch": 13.982428115015974, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8753 + }, + { + "epoch": 13.984025559105431, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8754 + }, + { + "epoch": 13.985623003194888, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 8755 + }, + { + "epoch": 13.987220447284345, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8756 + }, + { + "epoch": 13.988817891373802, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8757 + }, + { + "epoch": 13.99041533546326, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8758 + }, + { + "epoch": 13.992012779552716, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8759 + }, + { + "epoch": 13.993610223642172, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8760 + }, + { + "epoch": 13.995207667731629, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 8761 + }, + { + "epoch": 13.996805111821086, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8762 + }, + { + "epoch": 13.998402555910543, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8763 + }, + { + "epoch": 14.0, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8764 + }, + { + "epoch": 14.001597444089457, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8765 + }, + { + "epoch": 14.003194888178914, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 8766 + }, + { + "epoch": 14.004792332268371, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 8767 + }, + { + "epoch": 14.006389776357828, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8768 + }, + { + "epoch": 14.007987220447284, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8769 + }, + { + "epoch": 14.00958466453674, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 8770 + }, + { + "epoch": 14.011182108626198, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8771 + }, + { + "epoch": 14.012779552715655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8772 + }, + { + "epoch": 14.014376996805112, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8773 + }, + { + "epoch": 14.015974440894569, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8774 + }, + { + "epoch": 14.017571884984026, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8775 + }, + { + "epoch": 14.019169329073483, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8776 + }, + { + "epoch": 14.02076677316294, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 8777 + }, + { + "epoch": 14.022364217252397, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8778 + }, + { + "epoch": 14.023961661341852, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8779 + }, + { + "epoch": 14.02555910543131, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8780 + }, + { + "epoch": 14.027156549520766, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 8781 + }, + { + "epoch": 14.028753993610223, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 8782 + }, + { + "epoch": 14.03035143769968, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8783 + }, + { + "epoch": 14.031948881789138, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8784 + }, + { + "epoch": 14.033546325878595, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8785 + }, + { + "epoch": 14.035143769968052, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8786 + }, + { + "epoch": 14.036741214057509, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8787 + }, + { + "epoch": 14.038338658146964, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 8788 + }, + { + "epoch": 14.039936102236421, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8789 + }, + { + "epoch": 14.041533546325878, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8790 + }, + { + "epoch": 14.043130990415335, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8791 + }, + { + "epoch": 14.044728434504792, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 8792 + }, + { + "epoch": 14.04632587859425, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 8793 + }, + { + "epoch": 14.047923322683706, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8794 + }, + { + "epoch": 14.049520766773163, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8795 + }, + { + "epoch": 14.05111821086262, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8796 + }, + { + "epoch": 14.052715654952078, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8797 + }, + { + "epoch": 14.054313099041533, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8798 + }, + { + "epoch": 14.05591054313099, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 8799 + }, + { + "epoch": 14.057507987220447, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 8800 + }, + { + "epoch": 14.059105431309904, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8801 + }, + { + "epoch": 14.060702875399361, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8802 + }, + { + "epoch": 14.062300319488818, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8803 + }, + { + "epoch": 14.063897763578275, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 8804 + }, + { + "epoch": 14.065495207667732, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8805 + }, + { + "epoch": 14.06709265175719, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8806 + }, + { + "epoch": 14.068690095846645, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8807 + }, + { + "epoch": 14.070287539936102, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 8808 + }, + { + "epoch": 14.071884984025559, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 8809 + }, + { + "epoch": 14.073482428115016, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8810 + }, + { + "epoch": 14.075079872204473, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 8811 + }, + { + "epoch": 14.07667731629393, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8812 + }, + { + "epoch": 14.078274760383387, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8813 + }, + { + "epoch": 14.079872204472844, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 8814 + }, + { + "epoch": 14.081469648562301, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8815 + }, + { + "epoch": 14.083067092651758, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 8816 + }, + { + "epoch": 14.084664536741213, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8817 + }, + { + "epoch": 14.08626198083067, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 8818 + }, + { + "epoch": 14.087859424920127, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8819 + }, + { + "epoch": 14.089456869009584, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8820 + }, + { + "epoch": 14.091054313099042, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8821 + }, + { + "epoch": 14.092651757188499, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8822 + }, + { + "epoch": 14.094249201277956, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8823 + }, + { + "epoch": 14.095846645367413, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8824 + }, + { + "epoch": 14.09744408945687, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 8825 + }, + { + "epoch": 14.099041533546325, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8826 + }, + { + "epoch": 14.100638977635782, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8827 + }, + { + "epoch": 14.10223642172524, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8828 + }, + { + "epoch": 14.103833865814696, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 8829 + }, + { + "epoch": 14.105431309904153, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8830 + }, + { + "epoch": 14.10702875399361, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 8831 + }, + { + "epoch": 14.108626198083067, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8832 + }, + { + "epoch": 14.110223642172524, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8833 + }, + { + "epoch": 14.111821086261982, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8834 + }, + { + "epoch": 14.113418530351439, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 8835 + }, + { + "epoch": 14.115015974440894, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 8836 + }, + { + "epoch": 14.116613418530351, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 8837 + }, + { + "epoch": 14.118210862619808, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8838 + }, + { + "epoch": 14.119808306709265, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8839 + }, + { + "epoch": 14.121405750798722, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 8840 + }, + { + "epoch": 14.12300319488818, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8841 + }, + { + "epoch": 14.124600638977636, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8842 + }, + { + "epoch": 14.126198083067093, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8843 + }, + { + "epoch": 14.12779552715655, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8844 + }, + { + "epoch": 14.129392971246006, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 8845 + }, + { + "epoch": 14.130990415335463, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 8846 + }, + { + "epoch": 14.13258785942492, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8847 + }, + { + "epoch": 14.134185303514377, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 8848 + }, + { + "epoch": 14.135782747603834, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8849 + }, + { + "epoch": 14.13738019169329, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8850 + }, + { + "epoch": 14.138977635782748, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8851 + }, + { + "epoch": 14.140575079872205, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8852 + }, + { + "epoch": 14.142172523961662, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8853 + }, + { + "epoch": 14.143769968051119, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 8854 + }, + { + "epoch": 14.145367412140574, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8855 + }, + { + "epoch": 14.146964856230031, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 8856 + }, + { + "epoch": 14.148562300319488, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 8857 + }, + { + "epoch": 14.150159744408946, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 8858 + }, + { + "epoch": 14.151757188498403, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8859 + }, + { + "epoch": 14.15335463258786, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8860 + }, + { + "epoch": 14.154952076677317, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8861 + }, + { + "epoch": 14.156549520766774, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8862 + }, + { + "epoch": 14.15814696485623, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8863 + }, + { + "epoch": 14.159744408945686, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 8864 + }, + { + "epoch": 14.161341853035143, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8865 + }, + { + "epoch": 14.1629392971246, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8866 + }, + { + "epoch": 14.164536741214057, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 8867 + }, + { + "epoch": 14.166134185303514, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 8868 + }, + { + "epoch": 14.167731629392971, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8869 + }, + { + "epoch": 14.169329073482428, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 8870 + }, + { + "epoch": 14.170926517571885, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 8871 + }, + { + "epoch": 14.172523961661343, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8872 + }, + { + "epoch": 14.1741214057508, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8873 + }, + { + "epoch": 14.175718849840255, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 8874 + }, + { + "epoch": 14.177316293929712, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8875 + }, + { + "epoch": 14.178913738019169, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8876 + }, + { + "epoch": 14.180511182108626, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8877 + }, + { + "epoch": 14.182108626198083, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8878 + }, + { + "epoch": 14.18370607028754, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8879 + }, + { + "epoch": 14.185303514376997, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 8880 + }, + { + "epoch": 14.186900958466454, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 8881 + }, + { + "epoch": 14.188498402555911, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8882 + }, + { + "epoch": 14.190095846645367, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8883 + }, + { + "epoch": 14.191693290734824, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 8884 + }, + { + "epoch": 14.19329073482428, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 8885 + }, + { + "epoch": 14.194888178913738, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8886 + }, + { + "epoch": 14.196485623003195, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 8887 + }, + { + "epoch": 14.198083067092652, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 8888 + }, + { + "epoch": 14.199680511182109, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 8889 + }, + { + "epoch": 14.201277955271566, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 8890 + }, + { + "epoch": 14.202875399361023, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8891 + }, + { + "epoch": 14.204472843450478, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8892 + }, + { + "epoch": 14.206070287539935, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8893 + }, + { + "epoch": 14.207667731629392, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 8894 + }, + { + "epoch": 14.20926517571885, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8895 + }, + { + "epoch": 14.210862619808307, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 8896 + }, + { + "epoch": 14.212460063897764, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8897 + }, + { + "epoch": 14.21405750798722, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 8898 + }, + { + "epoch": 14.215654952076678, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 8899 + }, + { + "epoch": 14.217252396166135, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 8900 + }, + { + "epoch": 14.218849840255592, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8901 + }, + { + "epoch": 14.220447284345047, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 8902 + }, + { + "epoch": 14.222044728434504, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 8903 + }, + { + "epoch": 14.223642172523961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8904 + }, + { + "epoch": 14.225239616613418, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8905 + }, + { + "epoch": 14.226837060702875, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 8906 + }, + { + "epoch": 14.228434504792332, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 8907 + }, + { + "epoch": 14.23003194888179, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8908 + }, + { + "epoch": 14.231629392971247, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8909 + }, + { + "epoch": 14.233226837060704, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8910 + }, + { + "epoch": 14.23482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8911 + }, + { + "epoch": 14.236421725239616, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 8912 + }, + { + "epoch": 14.238019169329073, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8913 + }, + { + "epoch": 14.23961661341853, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 8914 + }, + { + "epoch": 14.241214057507987, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 8915 + }, + { + "epoch": 14.242811501597444, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8916 + }, + { + "epoch": 14.244408945686901, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8917 + }, + { + "epoch": 14.246006389776358, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8918 + }, + { + "epoch": 14.247603833865815, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8919 + }, + { + "epoch": 14.249201277955272, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8920 + }, + { + "epoch": 14.250798722044728, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 8921 + }, + { + "epoch": 14.252396166134185, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8922 + }, + { + "epoch": 14.253993610223642, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 8923 + }, + { + "epoch": 14.255591054313099, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8924 + }, + { + "epoch": 14.257188498402556, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8925 + }, + { + "epoch": 14.258785942492013, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 8926 + }, + { + "epoch": 14.26038338658147, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 8927 + }, + { + "epoch": 14.261980830670927, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8928 + }, + { + "epoch": 14.263578274760384, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 8929 + }, + { + "epoch": 14.26517571884984, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 8930 + }, + { + "epoch": 14.266773162939296, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 8931 + }, + { + "epoch": 14.268370607028753, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8932 + }, + { + "epoch": 14.26996805111821, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8933 + }, + { + "epoch": 14.271565495207668, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8934 + }, + { + "epoch": 14.273162939297125, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 8935 + }, + { + "epoch": 14.274760383386582, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 8936 + }, + { + "epoch": 14.276357827476039, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 8937 + }, + { + "epoch": 14.277955271565496, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8938 + }, + { + "epoch": 14.279552715654953, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 8939 + }, + { + "epoch": 14.281150159744408, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 8940 + }, + { + "epoch": 14.282747603833865, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8941 + }, + { + "epoch": 14.284345047923322, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 8942 + }, + { + "epoch": 14.28594249201278, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 8943 + }, + { + "epoch": 14.287539936102236, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 8944 + }, + { + "epoch": 14.289137380191693, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 8945 + }, + { + "epoch": 14.29073482428115, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 8946 + }, + { + "epoch": 14.292332268370608, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8947 + }, + { + "epoch": 14.293929712460065, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8948 + }, + { + "epoch": 14.295527156549522, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 8949 + }, + { + "epoch": 14.297124600638977, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 8950 + }, + { + "epoch": 14.298722044728434, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 8951 + }, + { + "epoch": 14.300319488817891, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 8952 + }, + { + "epoch": 14.301916932907348, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 8953 + }, + { + "epoch": 14.303514376996805, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 8954 + }, + { + "epoch": 14.305111821086262, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 8955 + }, + { + "epoch": 14.30670926517572, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 8956 + }, + { + "epoch": 14.308306709265176, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 8957 + }, + { + "epoch": 14.309904153354633, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 8958 + }, + { + "epoch": 14.311501597444089, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 8959 + }, + { + "epoch": 14.313099041533546, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 8960 + }, + { + "epoch": 14.314696485623003, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8961 + }, + { + "epoch": 14.31629392971246, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 8962 + }, + { + "epoch": 14.317891373801917, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 8963 + }, + { + "epoch": 14.319488817891374, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8964 + }, + { + "epoch": 14.321086261980831, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8965 + }, + { + "epoch": 14.322683706070288, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 8966 + }, + { + "epoch": 14.324281150159745, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 8967 + }, + { + "epoch": 14.3258785942492, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8968 + }, + { + "epoch": 14.327476038338657, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 8969 + }, + { + "epoch": 14.329073482428115, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 8970 + }, + { + "epoch": 14.330670926517572, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 8971 + }, + { + "epoch": 14.332268370607029, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 8972 + }, + { + "epoch": 14.333865814696486, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 8973 + }, + { + "epoch": 14.335463258785943, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 8974 + }, + { + "epoch": 14.3370607028754, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 8975 + }, + { + "epoch": 14.338658146964857, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 8976 + }, + { + "epoch": 14.340255591054314, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 8977 + }, + { + "epoch": 14.34185303514377, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 8978 + }, + { + "epoch": 14.343450479233226, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 8979 + }, + { + "epoch": 14.345047923322683, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8980 + }, + { + "epoch": 14.34664536741214, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 8981 + }, + { + "epoch": 14.348242811501597, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 8982 + }, + { + "epoch": 14.349840255591054, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 8983 + }, + { + "epoch": 14.351437699680512, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8984 + }, + { + "epoch": 14.353035143769969, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 8985 + }, + { + "epoch": 14.354632587859426, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 8986 + }, + { + "epoch": 14.356230031948883, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8987 + }, + { + "epoch": 14.357827476038338, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 8988 + }, + { + "epoch": 14.359424920127795, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 8989 + }, + { + "epoch": 14.361022364217252, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 8990 + }, + { + "epoch": 14.36261980830671, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 8991 + }, + { + "epoch": 14.364217252396166, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 8992 + }, + { + "epoch": 14.365814696485623, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 8993 + }, + { + "epoch": 14.36741214057508, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 8994 + }, + { + "epoch": 14.369009584664537, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 8995 + }, + { + "epoch": 14.370607028753994, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 8996 + }, + { + "epoch": 14.37220447284345, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 8997 + }, + { + "epoch": 14.373801916932907, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 8998 + }, + { + "epoch": 14.375399361022364, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 8999 + }, + { + "epoch": 14.37699680511182, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9000 + }, + { + "epoch": 14.378594249201278, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9001 + }, + { + "epoch": 14.380191693290735, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9002 + }, + { + "epoch": 14.381789137380192, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9003 + }, + { + "epoch": 14.383386581469649, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9004 + }, + { + "epoch": 14.384984025559106, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9005 + }, + { + "epoch": 14.386581469648561, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9006 + }, + { + "epoch": 14.388178913738018, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 9007 + }, + { + "epoch": 14.389776357827476, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9008 + }, + { + "epoch": 14.391373801916933, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9009 + }, + { + "epoch": 14.39297124600639, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 9010 + }, + { + "epoch": 14.394568690095847, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9011 + }, + { + "epoch": 14.396166134185304, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9012 + }, + { + "epoch": 14.39776357827476, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9013 + }, + { + "epoch": 14.399361022364218, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9014 + }, + { + "epoch": 14.400958466453675, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9015 + }, + { + "epoch": 14.40255591054313, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9016 + }, + { + "epoch": 14.404153354632587, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9017 + }, + { + "epoch": 14.405750798722044, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9018 + }, + { + "epoch": 14.407348242811501, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9019 + }, + { + "epoch": 14.408945686900958, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9020 + }, + { + "epoch": 14.410543130990416, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9021 + }, + { + "epoch": 14.412140575079873, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9022 + }, + { + "epoch": 14.41373801916933, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9023 + }, + { + "epoch": 14.415335463258787, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9024 + }, + { + "epoch": 14.416932907348242, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9025 + }, + { + "epoch": 14.418530351437699, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9026 + }, + { + "epoch": 14.420127795527156, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9027 + }, + { + "epoch": 14.421725239616613, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9028 + }, + { + "epoch": 14.42332268370607, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9029 + }, + { + "epoch": 14.424920127795527, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9030 + }, + { + "epoch": 14.426517571884984, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9031 + }, + { + "epoch": 14.428115015974441, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9032 + }, + { + "epoch": 14.429712460063898, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9033 + }, + { + "epoch": 14.431309904153355, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9034 + }, + { + "epoch": 14.43290734824281, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9035 + }, + { + "epoch": 14.434504792332268, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9036 + }, + { + "epoch": 14.436102236421725, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9037 + }, + { + "epoch": 14.437699680511182, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9038 + }, + { + "epoch": 14.439297124600639, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 9039 + }, + { + "epoch": 14.440894568690096, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9040 + }, + { + "epoch": 14.442492012779553, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9041 + }, + { + "epoch": 14.44408945686901, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9042 + }, + { + "epoch": 14.445686900958467, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9043 + }, + { + "epoch": 14.447284345047922, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 9044 + }, + { + "epoch": 14.44888178913738, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9045 + }, + { + "epoch": 14.450479233226837, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9046 + }, + { + "epoch": 14.452076677316294, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9047 + }, + { + "epoch": 14.45367412140575, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 9048 + }, + { + "epoch": 14.455271565495208, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9049 + }, + { + "epoch": 14.456869009584665, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9050 + }, + { + "epoch": 14.458466453674122, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9051 + }, + { + "epoch": 14.460063897763579, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9052 + }, + { + "epoch": 14.461661341853036, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9053 + }, + { + "epoch": 14.463258785942491, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9054 + }, + { + "epoch": 14.464856230031948, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9055 + }, + { + "epoch": 14.466453674121405, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9056 + }, + { + "epoch": 14.468051118210862, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9057 + }, + { + "epoch": 14.46964856230032, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9058 + }, + { + "epoch": 14.471246006389777, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9059 + }, + { + "epoch": 14.472843450479234, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9060 + }, + { + "epoch": 14.47444089456869, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9061 + }, + { + "epoch": 14.476038338658148, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9062 + }, + { + "epoch": 14.477635782747603, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9063 + }, + { + "epoch": 14.47923322683706, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9064 + }, + { + "epoch": 14.480830670926517, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9065 + }, + { + "epoch": 14.482428115015974, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9066 + }, + { + "epoch": 14.484025559105431, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9067 + }, + { + "epoch": 14.485623003194888, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9068 + }, + { + "epoch": 14.487220447284345, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9069 + }, + { + "epoch": 14.488817891373802, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9070 + }, + { + "epoch": 14.49041533546326, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 9071 + }, + { + "epoch": 14.492012779552716, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9072 + }, + { + "epoch": 14.493610223642172, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9073 + }, + { + "epoch": 14.495207667731629, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9074 + }, + { + "epoch": 14.496805111821086, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9075 + }, + { + "epoch": 14.498402555910543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 9076 + }, + { + "epoch": 14.5, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9077 + }, + { + "epoch": 14.501597444089457, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9078 + }, + { + "epoch": 14.503194888178914, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9079 + }, + { + "epoch": 14.504792332268371, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9080 + }, + { + "epoch": 14.506389776357828, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9081 + }, + { + "epoch": 14.507987220447284, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9082 + }, + { + "epoch": 14.50958466453674, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9083 + }, + { + "epoch": 14.511182108626198, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9084 + }, + { + "epoch": 14.512779552715655, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9085 + }, + { + "epoch": 14.514376996805112, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9086 + }, + { + "epoch": 14.515974440894569, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9087 + }, + { + "epoch": 14.517571884984026, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9088 + }, + { + "epoch": 14.519169329073483, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9089 + }, + { + "epoch": 14.52076677316294, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9090 + }, + { + "epoch": 14.522364217252395, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9091 + }, + { + "epoch": 14.523961661341852, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9092 + }, + { + "epoch": 14.52555910543131, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9093 + }, + { + "epoch": 14.527156549520766, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9094 + }, + { + "epoch": 14.528753993610223, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9095 + }, + { + "epoch": 14.53035143769968, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9096 + }, + { + "epoch": 14.531948881789138, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9097 + }, + { + "epoch": 14.533546325878595, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9098 + }, + { + "epoch": 14.535143769968052, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9099 + }, + { + "epoch": 14.536741214057509, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9100 + }, + { + "epoch": 14.538338658146966, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9101 + }, + { + "epoch": 14.539936102236421, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9102 + }, + { + "epoch": 14.541533546325878, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9103 + }, + { + "epoch": 14.543130990415335, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9104 + }, + { + "epoch": 14.544728434504792, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9105 + }, + { + "epoch": 14.54632587859425, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9106 + }, + { + "epoch": 14.547923322683706, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9107 + }, + { + "epoch": 14.549520766773163, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9108 + }, + { + "epoch": 14.55111821086262, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9109 + }, + { + "epoch": 14.552715654952078, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9110 + }, + { + "epoch": 14.554313099041533, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9111 + }, + { + "epoch": 14.55591054313099, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9112 + }, + { + "epoch": 14.557507987220447, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9113 + }, + { + "epoch": 14.559105431309904, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 9114 + }, + { + "epoch": 14.560702875399361, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9115 + }, + { + "epoch": 14.562300319488818, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9116 + }, + { + "epoch": 14.563897763578275, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9117 + }, + { + "epoch": 14.565495207667732, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9118 + }, + { + "epoch": 14.56709265175719, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9119 + }, + { + "epoch": 14.568690095846645, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9120 + }, + { + "epoch": 14.570287539936102, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9121 + }, + { + "epoch": 14.571884984025559, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 9122 + }, + { + "epoch": 14.573482428115016, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9123 + }, + { + "epoch": 14.575079872204473, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9124 + }, + { + "epoch": 14.57667731629393, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9125 + }, + { + "epoch": 14.578274760383387, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9126 + }, + { + "epoch": 14.579872204472844, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 9127 + }, + { + "epoch": 14.581469648562301, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9128 + }, + { + "epoch": 14.583067092651756, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 9129 + }, + { + "epoch": 14.584664536741213, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9130 + }, + { + "epoch": 14.58626198083067, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9131 + }, + { + "epoch": 14.587859424920127, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9132 + }, + { + "epoch": 14.589456869009584, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9133 + }, + { + "epoch": 14.591054313099042, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 9134 + }, + { + "epoch": 14.592651757188499, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9135 + }, + { + "epoch": 14.594249201277956, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9136 + }, + { + "epoch": 14.595846645367413, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9137 + }, + { + "epoch": 14.59744408945687, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9138 + }, + { + "epoch": 14.599041533546325, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9139 + }, + { + "epoch": 14.600638977635782, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9140 + }, + { + "epoch": 14.60223642172524, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 9141 + }, + { + "epoch": 14.603833865814696, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9142 + }, + { + "epoch": 14.605431309904153, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9143 + }, + { + "epoch": 14.60702875399361, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9144 + }, + { + "epoch": 14.608626198083067, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9145 + }, + { + "epoch": 14.610223642172524, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9146 + }, + { + "epoch": 14.611821086261982, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9147 + }, + { + "epoch": 14.613418530351439, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9148 + }, + { + "epoch": 14.615015974440894, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 9149 + }, + { + "epoch": 14.616613418530351, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9150 + }, + { + "epoch": 14.618210862619808, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9151 + }, + { + "epoch": 14.619808306709265, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9152 + }, + { + "epoch": 14.621405750798722, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9153 + }, + { + "epoch": 14.62300319488818, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9154 + }, + { + "epoch": 14.624600638977636, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9155 + }, + { + "epoch": 14.626198083067093, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9156 + }, + { + "epoch": 14.62779552715655, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9157 + }, + { + "epoch": 14.629392971246006, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9158 + }, + { + "epoch": 14.630990415335463, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9159 + }, + { + "epoch": 14.63258785942492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9160 + }, + { + "epoch": 14.634185303514377, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9161 + }, + { + "epoch": 14.635782747603834, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9162 + }, + { + "epoch": 14.63738019169329, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9163 + }, + { + "epoch": 14.638977635782748, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9164 + }, + { + "epoch": 14.640575079872205, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 9165 + }, + { + "epoch": 14.642172523961662, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9166 + }, + { + "epoch": 14.643769968051117, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9167 + }, + { + "epoch": 14.645367412140574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9168 + }, + { + "epoch": 14.646964856230031, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9169 + }, + { + "epoch": 14.648562300319488, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9170 + }, + { + "epoch": 14.650159744408946, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9171 + }, + { + "epoch": 14.651757188498403, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9172 + }, + { + "epoch": 14.65335463258786, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 9173 + }, + { + "epoch": 14.654952076677317, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9174 + }, + { + "epoch": 14.656549520766774, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9175 + }, + { + "epoch": 14.65814696485623, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9176 + }, + { + "epoch": 14.659744408945686, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9177 + }, + { + "epoch": 14.661341853035143, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9178 + }, + { + "epoch": 14.6629392971246, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9179 + }, + { + "epoch": 14.664536741214057, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9180 + }, + { + "epoch": 14.666134185303514, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9181 + }, + { + "epoch": 14.667731629392971, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9182 + }, + { + "epoch": 14.669329073482428, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9183 + }, + { + "epoch": 14.670926517571885, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9184 + }, + { + "epoch": 14.672523961661343, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9185 + }, + { + "epoch": 14.6741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9186 + }, + { + "epoch": 14.675718849840255, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9187 + }, + { + "epoch": 14.677316293929712, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9188 + }, + { + "epoch": 14.678913738019169, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9189 + }, + { + "epoch": 14.680511182108626, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9190 + }, + { + "epoch": 14.682108626198083, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9191 + }, + { + "epoch": 14.68370607028754, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9192 + }, + { + "epoch": 14.685303514376997, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9193 + }, + { + "epoch": 14.686900958466454, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 9194 + }, + { + "epoch": 14.688498402555911, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9195 + }, + { + "epoch": 14.690095846645367, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9196 + }, + { + "epoch": 14.691693290734824, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9197 + }, + { + "epoch": 14.69329073482428, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9198 + }, + { + "epoch": 14.694888178913738, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9199 + }, + { + "epoch": 14.696485623003195, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 9200 + }, + { + "epoch": 14.698083067092652, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9201 + }, + { + "epoch": 14.699680511182109, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9202 + }, + { + "epoch": 14.701277955271566, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9203 + }, + { + "epoch": 14.702875399361023, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9204 + }, + { + "epoch": 14.704472843450478, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 9205 + }, + { + "epoch": 14.706070287539935, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9206 + }, + { + "epoch": 14.707667731629392, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9207 + }, + { + "epoch": 14.70926517571885, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 9208 + }, + { + "epoch": 14.710862619808307, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9209 + }, + { + "epoch": 14.712460063897764, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9210 + }, + { + "epoch": 14.71405750798722, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9211 + }, + { + "epoch": 14.715654952076678, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9212 + }, + { + "epoch": 14.717252396166135, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9213 + }, + { + "epoch": 14.718849840255592, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 9214 + }, + { + "epoch": 14.720447284345047, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9215 + }, + { + "epoch": 14.722044728434504, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9216 + }, + { + "epoch": 14.723642172523961, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9217 + }, + { + "epoch": 14.725239616613418, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9218 + }, + { + "epoch": 14.726837060702875, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9219 + }, + { + "epoch": 14.728434504792332, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9220 + }, + { + "epoch": 14.73003194888179, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9221 + }, + { + "epoch": 14.731629392971247, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9222 + }, + { + "epoch": 14.733226837060704, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9223 + }, + { + "epoch": 14.73482428115016, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9224 + }, + { + "epoch": 14.736421725239616, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9225 + }, + { + "epoch": 14.738019169329073, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9226 + }, + { + "epoch": 14.73961661341853, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 9227 + }, + { + "epoch": 14.741214057507987, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 9228 + }, + { + "epoch": 14.742811501597444, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 9229 + }, + { + "epoch": 14.744408945686901, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9230 + }, + { + "epoch": 14.746006389776358, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9231 + }, + { + "epoch": 14.747603833865815, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9232 + }, + { + "epoch": 14.749201277955272, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9233 + }, + { + "epoch": 14.750798722044728, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9234 + }, + { + "epoch": 14.752396166134185, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9235 + }, + { + "epoch": 14.753993610223642, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9236 + }, + { + "epoch": 14.755591054313099, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9237 + }, + { + "epoch": 14.757188498402556, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 9238 + }, + { + "epoch": 14.758785942492013, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9239 + }, + { + "epoch": 14.76038338658147, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 9240 + }, + { + "epoch": 14.761980830670927, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9241 + }, + { + "epoch": 14.763578274760384, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9242 + }, + { + "epoch": 14.76517571884984, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9243 + }, + { + "epoch": 14.766773162939296, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9244 + }, + { + "epoch": 14.768370607028753, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9245 + }, + { + "epoch": 14.76996805111821, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9246 + }, + { + "epoch": 14.771565495207668, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 9247 + }, + { + "epoch": 14.773162939297125, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9248 + }, + { + "epoch": 14.774760383386582, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9249 + }, + { + "epoch": 14.776357827476039, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9250 + }, + { + "epoch": 14.777955271565496, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9251 + }, + { + "epoch": 14.779552715654953, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9252 + }, + { + "epoch": 14.781150159744408, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9253 + }, + { + "epoch": 14.782747603833865, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9254 + }, + { + "epoch": 14.784345047923322, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9255 + }, + { + "epoch": 14.78594249201278, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9256 + }, + { + "epoch": 14.787539936102236, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9257 + }, + { + "epoch": 14.789137380191693, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9258 + }, + { + "epoch": 14.79073482428115, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9259 + }, + { + "epoch": 14.792332268370608, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9260 + }, + { + "epoch": 14.793929712460065, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9261 + }, + { + "epoch": 14.795527156549522, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9262 + }, + { + "epoch": 14.797124600638977, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9263 + }, + { + "epoch": 14.798722044728434, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 9264 + }, + { + "epoch": 14.800319488817891, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9265 + }, + { + "epoch": 14.801916932907348, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9266 + }, + { + "epoch": 14.803514376996805, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9267 + }, + { + "epoch": 14.805111821086262, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9268 + }, + { + "epoch": 14.80670926517572, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9269 + }, + { + "epoch": 14.808306709265176, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9270 + }, + { + "epoch": 14.809904153354633, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9271 + }, + { + "epoch": 14.811501597444089, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9272 + }, + { + "epoch": 14.813099041533546, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9273 + }, + { + "epoch": 14.814696485623003, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9274 + }, + { + "epoch": 14.81629392971246, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9275 + }, + { + "epoch": 14.817891373801917, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9276 + }, + { + "epoch": 14.819488817891374, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9277 + }, + { + "epoch": 14.821086261980831, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9278 + }, + { + "epoch": 14.822683706070288, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9279 + }, + { + "epoch": 14.824281150159745, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 9280 + }, + { + "epoch": 14.8258785942492, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9281 + }, + { + "epoch": 14.827476038338657, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9282 + }, + { + "epoch": 14.829073482428115, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9283 + }, + { + "epoch": 14.830670926517572, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9284 + }, + { + "epoch": 14.832268370607029, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9285 + }, + { + "epoch": 14.833865814696486, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9286 + }, + { + "epoch": 14.835463258785943, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9287 + }, + { + "epoch": 14.8370607028754, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 9288 + }, + { + "epoch": 14.838658146964857, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9289 + }, + { + "epoch": 14.840255591054314, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9290 + }, + { + "epoch": 14.84185303514377, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9291 + }, + { + "epoch": 14.843450479233226, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9292 + }, + { + "epoch": 14.845047923322683, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9293 + }, + { + "epoch": 14.84664536741214, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9294 + }, + { + "epoch": 14.848242811501597, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9295 + }, + { + "epoch": 14.849840255591054, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9296 + }, + { + "epoch": 14.851437699680512, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9297 + }, + { + "epoch": 14.853035143769969, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9298 + }, + { + "epoch": 14.854632587859426, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9299 + }, + { + "epoch": 14.856230031948883, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9300 + }, + { + "epoch": 14.857827476038338, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9301 + }, + { + "epoch": 14.859424920127795, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9302 + }, + { + "epoch": 14.861022364217252, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9303 + }, + { + "epoch": 14.86261980830671, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9304 + }, + { + "epoch": 14.864217252396166, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9305 + }, + { + "epoch": 14.865814696485623, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9306 + }, + { + "epoch": 14.86741214057508, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9307 + }, + { + "epoch": 14.869009584664537, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9308 + }, + { + "epoch": 14.870607028753994, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9309 + }, + { + "epoch": 14.87220447284345, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9310 + }, + { + "epoch": 14.873801916932907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9311 + }, + { + "epoch": 14.875399361022364, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9312 + }, + { + "epoch": 14.87699680511182, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9313 + }, + { + "epoch": 14.878594249201278, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9314 + }, + { + "epoch": 14.880191693290735, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9315 + }, + { + "epoch": 14.881789137380192, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9316 + }, + { + "epoch": 14.883386581469649, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9317 + }, + { + "epoch": 14.884984025559106, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9318 + }, + { + "epoch": 14.886581469648561, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9319 + }, + { + "epoch": 14.888178913738018, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9320 + }, + { + "epoch": 14.889776357827476, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9321 + }, + { + "epoch": 14.891373801916933, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 9322 + }, + { + "epoch": 14.89297124600639, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9323 + }, + { + "epoch": 14.894568690095847, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9324 + }, + { + "epoch": 14.896166134185304, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9325 + }, + { + "epoch": 14.89776357827476, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 9326 + }, + { + "epoch": 14.899361022364218, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9327 + }, + { + "epoch": 14.900958466453675, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9328 + }, + { + "epoch": 14.90255591054313, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 9329 + }, + { + "epoch": 14.904153354632587, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9330 + }, + { + "epoch": 14.905750798722044, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9331 + }, + { + "epoch": 14.907348242811501, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9332 + }, + { + "epoch": 14.908945686900958, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9333 + }, + { + "epoch": 14.910543130990416, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9334 + }, + { + "epoch": 14.912140575079873, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9335 + }, + { + "epoch": 14.91373801916933, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9336 + }, + { + "epoch": 14.915335463258787, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9337 + }, + { + "epoch": 14.916932907348244, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9338 + }, + { + "epoch": 14.918530351437699, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9339 + }, + { + "epoch": 14.920127795527156, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9340 + }, + { + "epoch": 14.921725239616613, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9341 + }, + { + "epoch": 14.92332268370607, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9342 + }, + { + "epoch": 14.924920127795527, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9343 + }, + { + "epoch": 14.926517571884984, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9344 + }, + { + "epoch": 14.928115015974441, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9345 + }, + { + "epoch": 14.929712460063898, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9346 + }, + { + "epoch": 14.931309904153355, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9347 + }, + { + "epoch": 14.93290734824281, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 9348 + }, + { + "epoch": 14.934504792332268, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9349 + }, + { + "epoch": 14.936102236421725, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9350 + }, + { + "epoch": 14.937699680511182, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9351 + }, + { + "epoch": 14.939297124600639, + "grad_norm": 0.031982421875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9352 + }, + { + "epoch": 14.940894568690096, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9353 + }, + { + "epoch": 14.942492012779553, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 9354 + }, + { + "epoch": 14.94408945686901, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9355 + }, + { + "epoch": 14.945686900958467, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9356 + }, + { + "epoch": 14.947284345047922, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9357 + }, + { + "epoch": 14.94888178913738, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9358 + }, + { + "epoch": 14.950479233226837, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9359 + }, + { + "epoch": 14.952076677316294, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9360 + }, + { + "epoch": 14.95367412140575, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9361 + }, + { + "epoch": 14.955271565495208, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9362 + }, + { + "epoch": 14.956869009584665, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9363 + }, + { + "epoch": 14.958466453674122, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 9364 + }, + { + "epoch": 14.960063897763579, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9365 + }, + { + "epoch": 14.961661341853034, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9366 + }, + { + "epoch": 14.963258785942491, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9367 + }, + { + "epoch": 14.964856230031948, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9368 + }, + { + "epoch": 14.966453674121405, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9369 + }, + { + "epoch": 14.968051118210862, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9370 + }, + { + "epoch": 14.96964856230032, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9371 + }, + { + "epoch": 14.971246006389777, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9372 + }, + { + "epoch": 14.972843450479234, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9373 + }, + { + "epoch": 14.97444089456869, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9374 + }, + { + "epoch": 14.976038338658148, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9375 + }, + { + "epoch": 14.977635782747605, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9376 + }, + { + "epoch": 14.97923322683706, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9377 + }, + { + "epoch": 14.980830670926517, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9378 + }, + { + "epoch": 14.982428115015974, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9379 + }, + { + "epoch": 14.984025559105431, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9380 + }, + { + "epoch": 14.985623003194888, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9381 + }, + { + "epoch": 14.987220447284345, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9382 + }, + { + "epoch": 14.988817891373802, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9383 + }, + { + "epoch": 14.99041533546326, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9384 + }, + { + "epoch": 14.992012779552716, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9385 + }, + { + "epoch": 14.993610223642172, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9386 + }, + { + "epoch": 14.995207667731629, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9387 + }, + { + "epoch": 14.996805111821086, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9388 + }, + { + "epoch": 14.998402555910543, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9389 + }, + { + "epoch": 15.0, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9390 + }, + { + "epoch": 15.001597444089457, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 9391 + }, + { + "epoch": 15.003194888178914, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9392 + }, + { + "epoch": 15.004792332268371, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9393 + }, + { + "epoch": 15.006389776357828, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9394 + }, + { + "epoch": 15.007987220447284, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9395 + }, + { + "epoch": 15.00958466453674, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9396 + }, + { + "epoch": 15.011182108626198, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9397 + }, + { + "epoch": 15.012779552715655, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9398 + }, + { + "epoch": 15.014376996805112, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9399 + }, + { + "epoch": 15.015974440894569, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9400 + }, + { + "epoch": 15.017571884984026, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9401 + }, + { + "epoch": 15.019169329073483, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 9402 + }, + { + "epoch": 15.02076677316294, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9403 + }, + { + "epoch": 15.022364217252397, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9404 + }, + { + "epoch": 15.023961661341852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9405 + }, + { + "epoch": 15.02555910543131, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 9406 + }, + { + "epoch": 15.027156549520766, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9407 + }, + { + "epoch": 15.028753993610223, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9408 + }, + { + "epoch": 15.03035143769968, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9409 + }, + { + "epoch": 15.031948881789138, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9410 + }, + { + "epoch": 15.033546325878595, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9411 + }, + { + "epoch": 15.035143769968052, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9412 + }, + { + "epoch": 15.036741214057509, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9413 + }, + { + "epoch": 15.038338658146964, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9414 + }, + { + "epoch": 15.039936102236421, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 9415 + }, + { + "epoch": 15.041533546325878, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9416 + }, + { + "epoch": 15.043130990415335, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9417 + }, + { + "epoch": 15.044728434504792, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9418 + }, + { + "epoch": 15.04632587859425, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9419 + }, + { + "epoch": 15.047923322683706, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9420 + }, + { + "epoch": 15.049520766773163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9421 + }, + { + "epoch": 15.05111821086262, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9422 + }, + { + "epoch": 15.052715654952078, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9423 + }, + { + "epoch": 15.054313099041533, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9424 + }, + { + "epoch": 15.05591054313099, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9425 + }, + { + "epoch": 15.057507987220447, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9426 + }, + { + "epoch": 15.059105431309904, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9427 + }, + { + "epoch": 15.060702875399361, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9428 + }, + { + "epoch": 15.062300319488818, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 9429 + }, + { + "epoch": 15.063897763578275, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9430 + }, + { + "epoch": 15.065495207667732, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 9431 + }, + { + "epoch": 15.06709265175719, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9432 + }, + { + "epoch": 15.068690095846645, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9433 + }, + { + "epoch": 15.070287539936102, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9434 + }, + { + "epoch": 15.071884984025559, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9435 + }, + { + "epoch": 15.073482428115016, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9436 + }, + { + "epoch": 15.075079872204473, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9437 + }, + { + "epoch": 15.07667731629393, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9438 + }, + { + "epoch": 15.078274760383387, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 9439 + }, + { + "epoch": 15.079872204472844, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9440 + }, + { + "epoch": 15.081469648562301, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9441 + }, + { + "epoch": 15.083067092651758, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9442 + }, + { + "epoch": 15.084664536741213, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9443 + }, + { + "epoch": 15.08626198083067, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9444 + }, + { + "epoch": 15.087859424920127, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9445 + }, + { + "epoch": 15.089456869009584, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9446 + }, + { + "epoch": 15.091054313099042, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 9447 + }, + { + "epoch": 15.092651757188499, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9448 + }, + { + "epoch": 15.094249201277956, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9449 + }, + { + "epoch": 15.095846645367413, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9450 + }, + { + "epoch": 15.09744408945687, + "grad_norm": 0.031982421875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9451 + }, + { + "epoch": 15.099041533546325, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9452 + }, + { + "epoch": 15.100638977635782, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9453 + }, + { + "epoch": 15.10223642172524, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9454 + }, + { + "epoch": 15.103833865814696, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9455 + }, + { + "epoch": 15.105431309904153, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9456 + }, + { + "epoch": 15.10702875399361, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9457 + }, + { + "epoch": 15.108626198083067, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9458 + }, + { + "epoch": 15.110223642172524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9459 + }, + { + "epoch": 15.111821086261982, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9460 + }, + { + "epoch": 15.113418530351439, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9461 + }, + { + "epoch": 15.115015974440894, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9462 + }, + { + "epoch": 15.116613418530351, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9463 + }, + { + "epoch": 15.118210862619808, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9464 + }, + { + "epoch": 15.119808306709265, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9465 + }, + { + "epoch": 15.121405750798722, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9466 + }, + { + "epoch": 15.12300319488818, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9467 + }, + { + "epoch": 15.124600638977636, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9468 + }, + { + "epoch": 15.126198083067093, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9469 + }, + { + "epoch": 15.12779552715655, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 9470 + }, + { + "epoch": 15.129392971246006, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9471 + }, + { + "epoch": 15.130990415335463, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9472 + }, + { + "epoch": 15.13258785942492, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 9473 + }, + { + "epoch": 15.134185303514377, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9474 + }, + { + "epoch": 15.135782747603834, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9475 + }, + { + "epoch": 15.13738019169329, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9476 + }, + { + "epoch": 15.138977635782748, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9477 + }, + { + "epoch": 15.140575079872205, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9478 + }, + { + "epoch": 15.142172523961662, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9479 + }, + { + "epoch": 15.143769968051119, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9480 + }, + { + "epoch": 15.145367412140574, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9481 + }, + { + "epoch": 15.146964856230031, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9482 + }, + { + "epoch": 15.148562300319488, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9483 + }, + { + "epoch": 15.150159744408946, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9484 + }, + { + "epoch": 15.151757188498403, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9485 + }, + { + "epoch": 15.15335463258786, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9486 + }, + { + "epoch": 15.154952076677317, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9487 + }, + { + "epoch": 15.156549520766774, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9488 + }, + { + "epoch": 15.15814696485623, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9489 + }, + { + "epoch": 15.159744408945686, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 9490 + }, + { + "epoch": 15.161341853035143, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9491 + }, + { + "epoch": 15.1629392971246, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9492 + }, + { + "epoch": 15.164536741214057, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9493 + }, + { + "epoch": 15.166134185303514, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9494 + }, + { + "epoch": 15.167731629392971, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9495 + }, + { + "epoch": 15.169329073482428, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9496 + }, + { + "epoch": 15.170926517571885, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9497 + }, + { + "epoch": 15.172523961661343, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9498 + }, + { + "epoch": 15.1741214057508, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9499 + }, + { + "epoch": 15.175718849840255, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9500 + }, + { + "epoch": 15.177316293929712, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9501 + }, + { + "epoch": 15.178913738019169, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9502 + }, + { + "epoch": 15.180511182108626, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9503 + }, + { + "epoch": 15.182108626198083, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9504 + }, + { + "epoch": 15.18370607028754, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9505 + }, + { + "epoch": 15.185303514376997, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9506 + }, + { + "epoch": 15.186900958466454, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9507 + }, + { + "epoch": 15.188498402555911, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9508 + }, + { + "epoch": 15.190095846645367, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9509 + }, + { + "epoch": 15.191693290734824, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9510 + }, + { + "epoch": 15.19329073482428, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9511 + }, + { + "epoch": 15.194888178913738, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9512 + }, + { + "epoch": 15.196485623003195, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9513 + }, + { + "epoch": 15.198083067092652, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9514 + }, + { + "epoch": 15.199680511182109, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 9515 + }, + { + "epoch": 15.201277955271566, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 9516 + }, + { + "epoch": 15.202875399361023, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9517 + }, + { + "epoch": 15.204472843450478, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9518 + }, + { + "epoch": 15.206070287539935, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9519 + }, + { + "epoch": 15.207667731629392, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9520 + }, + { + "epoch": 15.20926517571885, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 9521 + }, + { + "epoch": 15.210862619808307, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9522 + }, + { + "epoch": 15.212460063897764, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9523 + }, + { + "epoch": 15.21405750798722, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9524 + }, + { + "epoch": 15.215654952076678, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9525 + }, + { + "epoch": 15.217252396166135, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9526 + }, + { + "epoch": 15.218849840255592, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9527 + }, + { + "epoch": 15.220447284345047, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9528 + }, + { + "epoch": 15.222044728434504, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9529 + }, + { + "epoch": 15.223642172523961, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9530 + }, + { + "epoch": 15.225239616613418, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 9531 + }, + { + "epoch": 15.226837060702875, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9532 + }, + { + "epoch": 15.228434504792332, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9533 + }, + { + "epoch": 15.23003194888179, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9534 + }, + { + "epoch": 15.231629392971247, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9535 + }, + { + "epoch": 15.233226837060704, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9536 + }, + { + "epoch": 15.23482428115016, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9537 + }, + { + "epoch": 15.236421725239616, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9538 + }, + { + "epoch": 15.238019169329073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9539 + }, + { + "epoch": 15.23961661341853, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9540 + }, + { + "epoch": 15.241214057507987, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9541 + }, + { + "epoch": 15.242811501597444, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9542 + }, + { + "epoch": 15.244408945686901, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9543 + }, + { + "epoch": 15.246006389776358, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9544 + }, + { + "epoch": 15.247603833865815, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 9545 + }, + { + "epoch": 15.249201277955272, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9546 + }, + { + "epoch": 15.250798722044728, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9547 + }, + { + "epoch": 15.252396166134185, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9548 + }, + { + "epoch": 15.253993610223642, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9549 + }, + { + "epoch": 15.255591054313099, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9550 + }, + { + "epoch": 15.257188498402556, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9551 + }, + { + "epoch": 15.258785942492013, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 9552 + }, + { + "epoch": 15.26038338658147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9553 + }, + { + "epoch": 15.261980830670927, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9554 + }, + { + "epoch": 15.263578274760384, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9555 + }, + { + "epoch": 15.26517571884984, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 9556 + }, + { + "epoch": 15.266773162939296, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9557 + }, + { + "epoch": 15.268370607028753, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9558 + }, + { + "epoch": 15.26996805111821, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9559 + }, + { + "epoch": 15.271565495207668, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9560 + }, + { + "epoch": 15.273162939297125, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9561 + }, + { + "epoch": 15.274760383386582, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9562 + }, + { + "epoch": 15.276357827476039, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9563 + }, + { + "epoch": 15.277955271565496, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9564 + }, + { + "epoch": 15.279552715654953, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9565 + }, + { + "epoch": 15.281150159744408, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9566 + }, + { + "epoch": 15.282747603833865, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9567 + }, + { + "epoch": 15.284345047923322, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9568 + }, + { + "epoch": 15.28594249201278, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9569 + }, + { + "epoch": 15.287539936102236, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9570 + }, + { + "epoch": 15.289137380191693, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9571 + }, + { + "epoch": 15.29073482428115, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 9572 + }, + { + "epoch": 15.292332268370608, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9573 + }, + { + "epoch": 15.293929712460065, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9574 + }, + { + "epoch": 15.295527156549522, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9575 + }, + { + "epoch": 15.297124600638977, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9576 + }, + { + "epoch": 15.298722044728434, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9577 + }, + { + "epoch": 15.300319488817891, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 9578 + }, + { + "epoch": 15.301916932907348, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9579 + }, + { + "epoch": 15.303514376996805, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9580 + }, + { + "epoch": 15.305111821086262, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9581 + }, + { + "epoch": 15.30670926517572, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9582 + }, + { + "epoch": 15.308306709265176, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9583 + }, + { + "epoch": 15.309904153354633, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9584 + }, + { + "epoch": 15.311501597444089, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 9585 + }, + { + "epoch": 15.313099041533546, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 9586 + }, + { + "epoch": 15.314696485623003, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 9587 + }, + { + "epoch": 15.31629392971246, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9588 + }, + { + "epoch": 15.317891373801917, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9589 + }, + { + "epoch": 15.319488817891374, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9590 + }, + { + "epoch": 15.321086261980831, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9591 + }, + { + "epoch": 15.322683706070288, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9592 + }, + { + "epoch": 15.324281150159745, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9593 + }, + { + "epoch": 15.3258785942492, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9594 + }, + { + "epoch": 15.327476038338657, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9595 + }, + { + "epoch": 15.329073482428115, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9596 + }, + { + "epoch": 15.330670926517572, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9597 + }, + { + "epoch": 15.332268370607029, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9598 + }, + { + "epoch": 15.333865814696486, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 9599 + }, + { + "epoch": 15.335463258785943, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9600 + }, + { + "epoch": 15.3370607028754, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9601 + }, + { + "epoch": 15.338658146964857, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9602 + }, + { + "epoch": 15.340255591054314, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9603 + }, + { + "epoch": 15.34185303514377, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 9604 + }, + { + "epoch": 15.343450479233226, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9605 + }, + { + "epoch": 15.345047923322683, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9606 + }, + { + "epoch": 15.34664536741214, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9607 + }, + { + "epoch": 15.348242811501597, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9608 + }, + { + "epoch": 15.349840255591054, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 9609 + }, + { + "epoch": 15.351437699680512, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9610 + }, + { + "epoch": 15.353035143769969, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9611 + }, + { + "epoch": 15.354632587859426, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9612 + }, + { + "epoch": 15.356230031948883, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9613 + }, + { + "epoch": 15.357827476038338, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9614 + }, + { + "epoch": 15.359424920127795, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9615 + }, + { + "epoch": 15.361022364217252, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9616 + }, + { + "epoch": 15.36261980830671, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9617 + }, + { + "epoch": 15.364217252396166, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9618 + }, + { + "epoch": 15.365814696485623, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9619 + }, + { + "epoch": 15.36741214057508, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9620 + }, + { + "epoch": 15.369009584664537, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9621 + }, + { + "epoch": 15.370607028753994, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9622 + }, + { + "epoch": 15.37220447284345, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9623 + }, + { + "epoch": 15.373801916932907, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9624 + }, + { + "epoch": 15.375399361022364, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9625 + }, + { + "epoch": 15.37699680511182, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9626 + }, + { + "epoch": 15.378594249201278, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9627 + }, + { + "epoch": 15.380191693290735, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9628 + }, + { + "epoch": 15.381789137380192, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9629 + }, + { + "epoch": 15.383386581469649, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9630 + }, + { + "epoch": 15.384984025559106, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 9631 + }, + { + "epoch": 15.386581469648561, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9632 + }, + { + "epoch": 15.388178913738018, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9633 + }, + { + "epoch": 15.389776357827476, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9634 + }, + { + "epoch": 15.391373801916933, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9635 + }, + { + "epoch": 15.39297124600639, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9636 + }, + { + "epoch": 15.394568690095847, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9637 + }, + { + "epoch": 15.396166134185304, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9638 + }, + { + "epoch": 15.39776357827476, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9639 + }, + { + "epoch": 15.399361022364218, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9640 + }, + { + "epoch": 15.400958466453675, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9641 + }, + { + "epoch": 15.40255591054313, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9642 + }, + { + "epoch": 15.404153354632587, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9643 + }, + { + "epoch": 15.405750798722044, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9644 + }, + { + "epoch": 15.407348242811501, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9645 + }, + { + "epoch": 15.408945686900958, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9646 + }, + { + "epoch": 15.410543130990416, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 9647 + }, + { + "epoch": 15.412140575079873, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 9648 + }, + { + "epoch": 15.41373801916933, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9649 + }, + { + "epoch": 15.415335463258787, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9650 + }, + { + "epoch": 15.416932907348242, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9651 + }, + { + "epoch": 15.418530351437699, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9652 + }, + { + "epoch": 15.420127795527156, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9653 + }, + { + "epoch": 15.421725239616613, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9654 + }, + { + "epoch": 15.42332268370607, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 9655 + }, + { + "epoch": 15.424920127795527, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 9656 + }, + { + "epoch": 15.426517571884984, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9657 + }, + { + "epoch": 15.428115015974441, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9658 + }, + { + "epoch": 15.429712460063898, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9659 + }, + { + "epoch": 15.431309904153355, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9660 + }, + { + "epoch": 15.43290734824281, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9661 + }, + { + "epoch": 15.434504792332268, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9662 + }, + { + "epoch": 15.436102236421725, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9663 + }, + { + "epoch": 15.437699680511182, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9664 + }, + { + "epoch": 15.439297124600639, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9665 + }, + { + "epoch": 15.440894568690096, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 9666 + }, + { + "epoch": 15.442492012779553, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9667 + }, + { + "epoch": 15.44408945686901, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9668 + }, + { + "epoch": 15.445686900958467, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9669 + }, + { + "epoch": 15.447284345047922, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9670 + }, + { + "epoch": 15.44888178913738, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 9671 + }, + { + "epoch": 15.450479233226837, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9672 + }, + { + "epoch": 15.452076677316294, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9673 + }, + { + "epoch": 15.45367412140575, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9674 + }, + { + "epoch": 15.455271565495208, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9675 + }, + { + "epoch": 15.456869009584665, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9676 + }, + { + "epoch": 15.458466453674122, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9677 + }, + { + "epoch": 15.460063897763579, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9678 + }, + { + "epoch": 15.461661341853036, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 9679 + }, + { + "epoch": 15.463258785942491, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9680 + }, + { + "epoch": 15.464856230031948, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9681 + }, + { + "epoch": 15.466453674121405, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9682 + }, + { + "epoch": 15.468051118210862, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9683 + }, + { + "epoch": 15.46964856230032, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 9684 + }, + { + "epoch": 15.471246006389777, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9685 + }, + { + "epoch": 15.472843450479234, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 9686 + }, + { + "epoch": 15.47444089456869, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9687 + }, + { + "epoch": 15.476038338658148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9688 + }, + { + "epoch": 15.477635782747603, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9689 + }, + { + "epoch": 15.47923322683706, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9690 + }, + { + "epoch": 15.480830670926517, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9691 + }, + { + "epoch": 15.482428115015974, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9692 + }, + { + "epoch": 15.484025559105431, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9693 + }, + { + "epoch": 15.485623003194888, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 9694 + }, + { + "epoch": 15.487220447284345, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9695 + }, + { + "epoch": 15.488817891373802, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 9696 + }, + { + "epoch": 15.49041533546326, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9697 + }, + { + "epoch": 15.492012779552716, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9698 + }, + { + "epoch": 15.493610223642172, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 9699 + }, + { + "epoch": 15.495207667731629, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9700 + }, + { + "epoch": 15.496805111821086, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9701 + }, + { + "epoch": 15.498402555910543, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9702 + }, + { + "epoch": 15.5, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9703 + }, + { + "epoch": 15.501597444089457, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 9704 + }, + { + "epoch": 15.503194888178914, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9705 + }, + { + "epoch": 15.504792332268371, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 9706 + }, + { + "epoch": 15.506389776357828, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9707 + }, + { + "epoch": 15.507987220447284, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 9708 + }, + { + "epoch": 15.50958466453674, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9709 + }, + { + "epoch": 15.511182108626198, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 9710 + }, + { + "epoch": 15.512779552715655, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9711 + }, + { + "epoch": 15.514376996805112, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9712 + }, + { + "epoch": 15.515974440894569, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9713 + }, + { + "epoch": 15.517571884984026, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9714 + }, + { + "epoch": 15.519169329073483, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9715 + }, + { + "epoch": 15.52076677316294, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9716 + }, + { + "epoch": 15.522364217252395, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9717 + }, + { + "epoch": 15.523961661341852, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9718 + }, + { + "epoch": 15.52555910543131, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 9719 + }, + { + "epoch": 15.527156549520766, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9720 + }, + { + "epoch": 15.528753993610223, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9721 + }, + { + "epoch": 15.53035143769968, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9722 + }, + { + "epoch": 15.531948881789138, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9723 + }, + { + "epoch": 15.533546325878595, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9724 + }, + { + "epoch": 15.535143769968052, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9725 + }, + { + "epoch": 15.536741214057509, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9726 + }, + { + "epoch": 15.538338658146966, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9727 + }, + { + "epoch": 15.539936102236421, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 9728 + }, + { + "epoch": 15.541533546325878, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9729 + }, + { + "epoch": 15.543130990415335, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 9730 + }, + { + "epoch": 15.544728434504792, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 9731 + }, + { + "epoch": 15.54632587859425, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 9732 + }, + { + "epoch": 15.547923322683706, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9733 + }, + { + "epoch": 15.549520766773163, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9734 + }, + { + "epoch": 15.55111821086262, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 9735 + }, + { + "epoch": 15.552715654952078, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 9736 + }, + { + "epoch": 15.554313099041533, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 9737 + }, + { + "epoch": 15.55591054313099, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9738 + }, + { + "epoch": 15.557507987220447, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9739 + }, + { + "epoch": 15.559105431309904, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9740 + }, + { + "epoch": 15.560702875399361, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9741 + }, + { + "epoch": 15.562300319488818, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9742 + }, + { + "epoch": 15.563897763578275, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9743 + }, + { + "epoch": 15.565495207667732, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9744 + }, + { + "epoch": 15.56709265175719, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 9745 + }, + { + "epoch": 15.568690095846645, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 9746 + }, + { + "epoch": 15.570287539936102, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9747 + }, + { + "epoch": 15.571884984025559, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9748 + }, + { + "epoch": 15.573482428115016, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9749 + }, + { + "epoch": 15.575079872204473, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9750 + }, + { + "epoch": 15.57667731629393, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9751 + }, + { + "epoch": 15.578274760383387, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9752 + }, + { + "epoch": 15.579872204472844, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9753 + }, + { + "epoch": 15.581469648562301, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9754 + }, + { + "epoch": 15.583067092651756, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9755 + }, + { + "epoch": 15.584664536741213, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 9756 + }, + { + "epoch": 15.58626198083067, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9757 + }, + { + "epoch": 15.587859424920127, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 9758 + }, + { + "epoch": 15.589456869009584, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9759 + }, + { + "epoch": 15.591054313099042, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9760 + }, + { + "epoch": 15.592651757188499, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9761 + }, + { + "epoch": 15.594249201277956, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9762 + }, + { + "epoch": 15.595846645367413, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 9763 + }, + { + "epoch": 15.59744408945687, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9764 + }, + { + "epoch": 15.599041533546325, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9765 + }, + { + "epoch": 15.600638977635782, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9766 + }, + { + "epoch": 15.60223642172524, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9767 + }, + { + "epoch": 15.603833865814696, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9768 + }, + { + "epoch": 15.605431309904153, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9769 + }, + { + "epoch": 15.60702875399361, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9770 + }, + { + "epoch": 15.608626198083067, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 9771 + }, + { + "epoch": 15.610223642172524, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9772 + }, + { + "epoch": 15.611821086261982, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9773 + }, + { + "epoch": 15.613418530351439, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9774 + }, + { + "epoch": 15.615015974440894, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9775 + }, + { + "epoch": 15.616613418530351, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 9776 + }, + { + "epoch": 15.618210862619808, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9777 + }, + { + "epoch": 15.619808306709265, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9778 + }, + { + "epoch": 15.621405750798722, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9779 + }, + { + "epoch": 15.62300319488818, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9780 + }, + { + "epoch": 15.624600638977636, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 9781 + }, + { + "epoch": 15.626198083067093, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 9782 + }, + { + "epoch": 15.62779552715655, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 9783 + }, + { + "epoch": 15.629392971246006, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9784 + }, + { + "epoch": 15.630990415335463, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 9785 + }, + { + "epoch": 15.63258785942492, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9786 + }, + { + "epoch": 15.634185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 9787 + }, + { + "epoch": 15.635782747603834, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9788 + }, + { + "epoch": 15.63738019169329, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9789 + }, + { + "epoch": 15.638977635782748, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9790 + }, + { + "epoch": 15.640575079872205, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9791 + }, + { + "epoch": 15.642172523961662, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9792 + }, + { + "epoch": 15.643769968051117, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9793 + }, + { + "epoch": 15.645367412140574, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9794 + }, + { + "epoch": 15.646964856230031, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 9795 + }, + { + "epoch": 15.648562300319488, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9796 + }, + { + "epoch": 15.650159744408946, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 9797 + }, + { + "epoch": 15.651757188498403, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 9798 + }, + { + "epoch": 15.65335463258786, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9799 + }, + { + "epoch": 15.654952076677317, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 9800 + }, + { + "epoch": 15.656549520766774, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 9801 + }, + { + "epoch": 15.65814696485623, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 9802 + }, + { + "epoch": 15.659744408945686, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 9803 + }, + { + "epoch": 15.661341853035143, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9804 + }, + { + "epoch": 15.6629392971246, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9805 + }, + { + "epoch": 15.664536741214057, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 9806 + }, + { + "epoch": 15.666134185303514, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9807 + }, + { + "epoch": 15.667731629392971, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 9808 + }, + { + "epoch": 15.669329073482428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9809 + }, + { + "epoch": 15.670926517571885, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9810 + }, + { + "epoch": 15.672523961661343, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 9811 + }, + { + "epoch": 15.6741214057508, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9812 + }, + { + "epoch": 15.675718849840255, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9813 + }, + { + "epoch": 15.677316293929712, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9814 + }, + { + "epoch": 15.678913738019169, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 9815 + }, + { + "epoch": 15.680511182108626, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 9816 + }, + { + "epoch": 15.682108626198083, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9817 + }, + { + "epoch": 15.68370607028754, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9818 + }, + { + "epoch": 15.685303514376997, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9819 + }, + { + "epoch": 15.686900958466454, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9820 + }, + { + "epoch": 15.688498402555911, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 9821 + }, + { + "epoch": 15.690095846645367, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 9822 + }, + { + "epoch": 15.691693290734824, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9823 + }, + { + "epoch": 15.69329073482428, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9824 + }, + { + "epoch": 15.694888178913738, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9825 + }, + { + "epoch": 15.696485623003195, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9826 + }, + { + "epoch": 15.698083067092652, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 9827 + }, + { + "epoch": 15.699680511182109, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 9828 + }, + { + "epoch": 15.701277955271566, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9829 + }, + { + "epoch": 15.702875399361023, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9830 + }, + { + "epoch": 15.704472843450478, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9831 + }, + { + "epoch": 15.706070287539935, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9832 + }, + { + "epoch": 15.707667731629392, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9833 + }, + { + "epoch": 15.70926517571885, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 9834 + }, + { + "epoch": 15.710862619808307, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 9835 + }, + { + "epoch": 15.712460063897764, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9836 + }, + { + "epoch": 15.71405750798722, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9837 + }, + { + "epoch": 15.715654952076678, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 9838 + }, + { + "epoch": 15.717252396166135, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9839 + }, + { + "epoch": 15.718849840255592, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9840 + }, + { + "epoch": 15.720447284345047, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 9841 + }, + { + "epoch": 15.722044728434504, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9842 + }, + { + "epoch": 15.723642172523961, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9843 + }, + { + "epoch": 15.725239616613418, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9844 + }, + { + "epoch": 15.726837060702875, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9845 + }, + { + "epoch": 15.728434504792332, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9846 + }, + { + "epoch": 15.73003194888179, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 9847 + }, + { + "epoch": 15.731629392971247, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9848 + }, + { + "epoch": 15.733226837060704, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9849 + }, + { + "epoch": 15.73482428115016, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9850 + }, + { + "epoch": 15.736421725239616, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 9851 + }, + { + "epoch": 15.738019169329073, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 9852 + }, + { + "epoch": 15.73961661341853, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 9853 + }, + { + "epoch": 15.741214057507987, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9854 + }, + { + "epoch": 15.742811501597444, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9855 + }, + { + "epoch": 15.744408945686901, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9856 + }, + { + "epoch": 15.746006389776358, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9857 + }, + { + "epoch": 15.747603833865815, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 9858 + }, + { + "epoch": 15.749201277955272, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9859 + }, + { + "epoch": 15.750798722044728, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9860 + }, + { + "epoch": 15.752396166134185, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 9861 + }, + { + "epoch": 15.753993610223642, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9862 + }, + { + "epoch": 15.755591054313099, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 9863 + }, + { + "epoch": 15.757188498402556, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9864 + }, + { + "epoch": 15.758785942492013, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 9865 + }, + { + "epoch": 15.76038338658147, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 9866 + }, + { + "epoch": 15.761980830670927, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 9867 + }, + { + "epoch": 15.763578274760384, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 9868 + }, + { + "epoch": 15.76517571884984, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9869 + }, + { + "epoch": 15.766773162939296, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 9870 + }, + { + "epoch": 15.768370607028753, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 9871 + }, + { + "epoch": 15.76996805111821, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 9872 + }, + { + "epoch": 15.771565495207668, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9873 + }, + { + "epoch": 15.773162939297125, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9874 + }, + { + "epoch": 15.774760383386582, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9875 + }, + { + "epoch": 15.776357827476039, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9876 + }, + { + "epoch": 15.777955271565496, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 9877 + }, + { + "epoch": 15.779552715654953, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9878 + }, + { + "epoch": 15.781150159744408, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9879 + }, + { + "epoch": 15.782747603833865, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 9880 + }, + { + "epoch": 15.784345047923322, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9881 + }, + { + "epoch": 15.78594249201278, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 9882 + }, + { + "epoch": 15.787539936102236, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 9883 + }, + { + "epoch": 15.789137380191693, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9884 + }, + { + "epoch": 15.79073482428115, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9885 + }, + { + "epoch": 15.792332268370608, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9886 + }, + { + "epoch": 15.793929712460065, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9887 + }, + { + "epoch": 15.795527156549522, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9888 + }, + { + "epoch": 15.797124600638977, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 9889 + }, + { + "epoch": 15.798722044728434, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9890 + }, + { + "epoch": 15.800319488817891, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 9891 + }, + { + "epoch": 15.801916932907348, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9892 + }, + { + "epoch": 15.803514376996805, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 9893 + }, + { + "epoch": 15.805111821086262, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9894 + }, + { + "epoch": 15.80670926517572, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 9895 + }, + { + "epoch": 15.808306709265176, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 9896 + }, + { + "epoch": 15.809904153354633, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9897 + }, + { + "epoch": 15.811501597444089, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9898 + }, + { + "epoch": 15.813099041533546, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 9899 + }, + { + "epoch": 15.814696485623003, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 9900 + }, + { + "epoch": 15.81629392971246, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 9901 + }, + { + "epoch": 15.817891373801917, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 9902 + }, + { + "epoch": 15.819488817891374, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9903 + }, + { + "epoch": 15.821086261980831, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9904 + }, + { + "epoch": 15.822683706070288, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9905 + }, + { + "epoch": 15.824281150159745, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9906 + }, + { + "epoch": 15.8258785942492, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 9907 + }, + { + "epoch": 15.827476038338657, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 9908 + }, + { + "epoch": 15.829073482428115, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9909 + }, + { + "epoch": 15.830670926517572, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9910 + }, + { + "epoch": 15.832268370607029, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 9911 + }, + { + "epoch": 15.833865814696486, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9912 + }, + { + "epoch": 15.835463258785943, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9913 + }, + { + "epoch": 15.8370607028754, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 9914 + }, + { + "epoch": 15.838658146964857, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 9915 + }, + { + "epoch": 15.840255591054314, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 9916 + }, + { + "epoch": 15.84185303514377, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9917 + }, + { + "epoch": 15.843450479233226, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 9918 + }, + { + "epoch": 15.845047923322683, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 9919 + }, + { + "epoch": 15.84664536741214, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 9920 + }, + { + "epoch": 15.848242811501597, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 9921 + }, + { + "epoch": 15.849840255591054, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 9922 + }, + { + "epoch": 15.851437699680512, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9923 + }, + { + "epoch": 15.853035143769969, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 9924 + }, + { + "epoch": 15.854632587859426, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 9925 + }, + { + "epoch": 15.856230031948883, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9926 + }, + { + "epoch": 15.857827476038338, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 9927 + }, + { + "epoch": 15.859424920127795, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 9928 + }, + { + "epoch": 15.861022364217252, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9929 + }, + { + "epoch": 15.86261980830671, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 9930 + }, + { + "epoch": 15.864217252396166, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 9931 + }, + { + "epoch": 15.865814696485623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 9932 + }, + { + "epoch": 15.86741214057508, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9933 + }, + { + "epoch": 15.869009584664537, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 9934 + }, + { + "epoch": 15.870607028753994, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 9935 + }, + { + "epoch": 15.87220447284345, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 9936 + }, + { + "epoch": 15.873801916932907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9937 + }, + { + "epoch": 15.875399361022364, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 9938 + }, + { + "epoch": 15.87699680511182, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9939 + }, + { + "epoch": 15.878594249201278, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 9940 + }, + { + "epoch": 15.880191693290735, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9941 + }, + { + "epoch": 15.881789137380192, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 9942 + }, + { + "epoch": 15.883386581469649, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 9943 + }, + { + "epoch": 15.884984025559106, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9944 + }, + { + "epoch": 15.886581469648561, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 9945 + }, + { + "epoch": 15.888178913738018, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 9946 + }, + { + "epoch": 15.889776357827476, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 9947 + }, + { + "epoch": 15.891373801916933, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 9948 + }, + { + "epoch": 15.89297124600639, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9949 + }, + { + "epoch": 15.894568690095847, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 9950 + }, + { + "epoch": 15.896166134185304, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 9951 + }, + { + "epoch": 15.89776357827476, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9952 + }, + { + "epoch": 15.899361022364218, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9953 + }, + { + "epoch": 15.900958466453675, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9954 + }, + { + "epoch": 15.90255591054313, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9955 + }, + { + "epoch": 15.904153354632587, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9956 + }, + { + "epoch": 15.905750798722044, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 9957 + }, + { + "epoch": 15.907348242811501, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 9958 + }, + { + "epoch": 15.908945686900958, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 9959 + }, + { + "epoch": 15.910543130990416, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 9960 + }, + { + "epoch": 15.912140575079873, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 9961 + }, + { + "epoch": 15.91373801916933, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 9962 + }, + { + "epoch": 15.915335463258787, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9963 + }, + { + "epoch": 15.916932907348244, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9964 + }, + { + "epoch": 15.918530351437699, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 9965 + }, + { + "epoch": 15.920127795527156, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9966 + }, + { + "epoch": 15.921725239616613, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 9967 + }, + { + "epoch": 15.92332268370607, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9968 + }, + { + "epoch": 15.924920127795527, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 9969 + }, + { + "epoch": 15.926517571884984, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9970 + }, + { + "epoch": 15.928115015974441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 9971 + }, + { + "epoch": 15.929712460063898, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 9972 + }, + { + "epoch": 15.931309904153355, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 9973 + }, + { + "epoch": 15.93290734824281, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 9974 + }, + { + "epoch": 15.934504792332268, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9975 + }, + { + "epoch": 15.936102236421725, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 9976 + }, + { + "epoch": 15.937699680511182, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 9977 + }, + { + "epoch": 15.939297124600639, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 9978 + }, + { + "epoch": 15.940894568690096, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 9979 + }, + { + "epoch": 15.942492012779553, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 9980 + }, + { + "epoch": 15.94408945686901, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 9981 + }, + { + "epoch": 15.945686900958467, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 9982 + }, + { + "epoch": 15.947284345047922, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 9983 + }, + { + "epoch": 15.94888178913738, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 9984 + }, + { + "epoch": 15.950479233226837, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 9985 + }, + { + "epoch": 15.952076677316294, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9986 + }, + { + "epoch": 15.95367412140575, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 9987 + }, + { + "epoch": 15.955271565495208, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 9988 + }, + { + "epoch": 15.956869009584665, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 9989 + }, + { + "epoch": 15.958466453674122, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 9990 + }, + { + "epoch": 15.960063897763579, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 9991 + }, + { + "epoch": 15.961661341853034, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 9992 + }, + { + "epoch": 15.963258785942491, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 9993 + }, + { + "epoch": 15.964856230031948, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 9994 + }, + { + "epoch": 15.966453674121405, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 9995 + }, + { + "epoch": 15.968051118210862, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 9996 + }, + { + "epoch": 15.96964856230032, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 9997 + }, + { + "epoch": 15.971246006389777, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 9998 + }, + { + "epoch": 15.972843450479234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 9999 + }, + { + "epoch": 15.97444089456869, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10000 + }, + { + "epoch": 15.976038338658148, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10001 + }, + { + "epoch": 15.977635782747605, + "grad_norm": 0.48828125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10002 + }, + { + "epoch": 15.97923322683706, + "grad_norm": 0.56640625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 10003 + }, + { + "epoch": 15.980830670926517, + "grad_norm": 1.171875, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 10004 + }, + { + "epoch": 15.982428115015974, + "grad_norm": 1.703125, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 10005 + }, + { + "epoch": 15.984025559105431, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 10006 + }, + { + "epoch": 15.985623003194888, + "grad_norm": 0.9609375, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 10007 + }, + { + "epoch": 15.987220447284345, + "grad_norm": 1.1015625, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 10008 + }, + { + "epoch": 15.988817891373802, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 10009 + }, + { + "epoch": 15.99041533546326, + "grad_norm": 0.8125, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 10010 + }, + { + "epoch": 15.992012779552716, + "grad_norm": 0.5703125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 10011 + }, + { + "epoch": 15.993610223642172, + "grad_norm": 0.333984375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 10012 + }, + { + "epoch": 15.995207667731629, + "grad_norm": 0.54296875, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 10013 + }, + { + "epoch": 15.996805111821086, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 10014 + }, + { + "epoch": 15.998402555910543, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 10015 + }, + { + "epoch": 16.0, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 10016 + }, + { + "epoch": 16.001597444089455, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 10017 + }, + { + "epoch": 16.003194888178914, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 10018 + }, + { + "epoch": 16.00479233226837, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 10019 + }, + { + "epoch": 16.00638977635783, + "grad_norm": 0.333984375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10020 + }, + { + "epoch": 16.007987220447284, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 10021 + }, + { + "epoch": 16.009584664536742, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 10022 + }, + { + "epoch": 16.011182108626198, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10023 + }, + { + "epoch": 16.012779552715656, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 10024 + }, + { + "epoch": 16.01437699680511, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 10025 + }, + { + "epoch": 16.015974440894567, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10026 + }, + { + "epoch": 16.017571884984026, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 10027 + }, + { + "epoch": 16.01916932907348, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10028 + }, + { + "epoch": 16.02076677316294, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 10029 + }, + { + "epoch": 16.022364217252395, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10030 + }, + { + "epoch": 16.023961661341854, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10031 + }, + { + "epoch": 16.02555910543131, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10032 + }, + { + "epoch": 16.027156549520768, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10033 + }, + { + "epoch": 16.028753993610223, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 10034 + }, + { + "epoch": 16.03035143769968, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10035 + }, + { + "epoch": 16.031948881789138, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10036 + }, + { + "epoch": 16.033546325878593, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10037 + }, + { + "epoch": 16.03514376996805, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 10038 + }, + { + "epoch": 16.036741214057507, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10039 + }, + { + "epoch": 16.038338658146966, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10040 + }, + { + "epoch": 16.03993610223642, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10041 + }, + { + "epoch": 16.04153354632588, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 10042 + }, + { + "epoch": 16.043130990415335, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10043 + }, + { + "epoch": 16.044728434504794, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10044 + }, + { + "epoch": 16.04632587859425, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10045 + }, + { + "epoch": 16.047923322683705, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10046 + }, + { + "epoch": 16.049520766773163, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 10047 + }, + { + "epoch": 16.05111821086262, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10048 + }, + { + "epoch": 16.052715654952078, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10049 + }, + { + "epoch": 16.054313099041533, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10050 + }, + { + "epoch": 16.05591054313099, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10051 + }, + { + "epoch": 16.057507987220447, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10052 + }, + { + "epoch": 16.059105431309906, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 10053 + }, + { + "epoch": 16.06070287539936, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10054 + }, + { + "epoch": 16.062300319488816, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10055 + }, + { + "epoch": 16.063897763578275, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10056 + }, + { + "epoch": 16.06549520766773, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10057 + }, + { + "epoch": 16.06709265175719, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10058 + }, + { + "epoch": 16.068690095846645, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10059 + }, + { + "epoch": 16.070287539936103, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10060 + }, + { + "epoch": 16.07188498402556, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 10061 + }, + { + "epoch": 16.073482428115017, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10062 + }, + { + "epoch": 16.075079872204473, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10063 + }, + { + "epoch": 16.076677316293928, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10064 + }, + { + "epoch": 16.078274760383387, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10065 + }, + { + "epoch": 16.079872204472842, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10066 + }, + { + "epoch": 16.0814696485623, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10067 + }, + { + "epoch": 16.083067092651756, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10068 + }, + { + "epoch": 16.084664536741215, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10069 + }, + { + "epoch": 16.08626198083067, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 10070 + }, + { + "epoch": 16.08785942492013, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10071 + }, + { + "epoch": 16.089456869009584, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10072 + }, + { + "epoch": 16.09105431309904, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10073 + }, + { + "epoch": 16.0926517571885, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10074 + }, + { + "epoch": 16.094249201277954, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10075 + }, + { + "epoch": 16.095846645367413, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10076 + }, + { + "epoch": 16.097444089456868, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10077 + }, + { + "epoch": 16.099041533546327, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10078 + }, + { + "epoch": 16.100638977635782, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10079 + }, + { + "epoch": 16.10223642172524, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10080 + }, + { + "epoch": 16.103833865814696, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10081 + }, + { + "epoch": 16.105431309904155, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10082 + }, + { + "epoch": 16.10702875399361, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10083 + }, + { + "epoch": 16.108626198083066, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10084 + }, + { + "epoch": 16.110223642172524, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10085 + }, + { + "epoch": 16.11182108626198, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10086 + }, + { + "epoch": 16.11341853035144, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10087 + }, + { + "epoch": 16.115015974440894, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10088 + }, + { + "epoch": 16.116613418530353, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10089 + }, + { + "epoch": 16.118210862619808, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10090 + }, + { + "epoch": 16.119808306709267, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10091 + }, + { + "epoch": 16.121405750798722, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10092 + }, + { + "epoch": 16.123003194888177, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10093 + }, + { + "epoch": 16.124600638977636, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10094 + }, + { + "epoch": 16.12619808306709, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10095 + }, + { + "epoch": 16.12779552715655, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10096 + }, + { + "epoch": 16.129392971246006, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10097 + }, + { + "epoch": 16.130990415335464, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10098 + }, + { + "epoch": 16.13258785942492, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10099 + }, + { + "epoch": 16.13418530351438, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10100 + }, + { + "epoch": 16.135782747603834, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10101 + }, + { + "epoch": 16.13738019169329, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10102 + }, + { + "epoch": 16.138977635782748, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10103 + }, + { + "epoch": 16.140575079872203, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10104 + }, + { + "epoch": 16.142172523961662, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 10105 + }, + { + "epoch": 16.143769968051117, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10106 + }, + { + "epoch": 16.145367412140576, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10107 + }, + { + "epoch": 16.14696485623003, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10108 + }, + { + "epoch": 16.14856230031949, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10109 + }, + { + "epoch": 16.150159744408946, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10110 + }, + { + "epoch": 16.1517571884984, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10111 + }, + { + "epoch": 16.15335463258786, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10112 + }, + { + "epoch": 16.154952076677315, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10113 + }, + { + "epoch": 16.156549520766774, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10114 + }, + { + "epoch": 16.15814696485623, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10115 + }, + { + "epoch": 16.159744408945688, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10116 + }, + { + "epoch": 16.161341853035143, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10117 + }, + { + "epoch": 16.162939297124602, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10118 + }, + { + "epoch": 16.164536741214057, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 10119 + }, + { + "epoch": 16.166134185303516, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10120 + }, + { + "epoch": 16.16773162939297, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10121 + }, + { + "epoch": 16.169329073482427, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10122 + }, + { + "epoch": 16.170926517571885, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10123 + }, + { + "epoch": 16.17252396166134, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10124 + }, + { + "epoch": 16.1741214057508, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10125 + }, + { + "epoch": 16.175718849840255, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10126 + }, + { + "epoch": 16.177316293929714, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10127 + }, + { + "epoch": 16.17891373801917, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10128 + }, + { + "epoch": 16.180511182108628, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10129 + }, + { + "epoch": 16.182108626198083, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10130 + }, + { + "epoch": 16.18370607028754, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10131 + }, + { + "epoch": 16.185303514376997, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10132 + }, + { + "epoch": 16.186900958466452, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10133 + }, + { + "epoch": 16.18849840255591, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 10134 + }, + { + "epoch": 16.190095846645367, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10135 + }, + { + "epoch": 16.191693290734825, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 10136 + }, + { + "epoch": 16.19329073482428, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 10137 + }, + { + "epoch": 16.19488817891374, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10138 + }, + { + "epoch": 16.196485623003195, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10139 + }, + { + "epoch": 16.19808306709265, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10140 + }, + { + "epoch": 16.19968051118211, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10141 + }, + { + "epoch": 16.201277955271564, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10142 + }, + { + "epoch": 16.202875399361023, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10143 + }, + { + "epoch": 16.20447284345048, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10144 + }, + { + "epoch": 16.206070287539937, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10145 + }, + { + "epoch": 16.207667731629392, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10146 + }, + { + "epoch": 16.20926517571885, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 10147 + }, + { + "epoch": 16.210862619808307, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10148 + }, + { + "epoch": 16.212460063897762, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10149 + }, + { + "epoch": 16.21405750798722, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 10150 + }, + { + "epoch": 16.215654952076676, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10151 + }, + { + "epoch": 16.217252396166135, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10152 + }, + { + "epoch": 16.21884984025559, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10153 + }, + { + "epoch": 16.22044728434505, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10154 + }, + { + "epoch": 16.222044728434504, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 10155 + }, + { + "epoch": 16.223642172523963, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10156 + }, + { + "epoch": 16.22523961661342, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10157 + }, + { + "epoch": 16.226837060702877, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10158 + }, + { + "epoch": 16.228434504792332, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 10159 + }, + { + "epoch": 16.230031948881788, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10160 + }, + { + "epoch": 16.231629392971247, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10161 + }, + { + "epoch": 16.233226837060702, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10162 + }, + { + "epoch": 16.23482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10163 + }, + { + "epoch": 16.236421725239616, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10164 + }, + { + "epoch": 16.238019169329075, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10165 + }, + { + "epoch": 16.23961661341853, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10166 + }, + { + "epoch": 16.24121405750799, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10167 + }, + { + "epoch": 16.242811501597444, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10168 + }, + { + "epoch": 16.2444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10169 + }, + { + "epoch": 16.24600638977636, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10170 + }, + { + "epoch": 16.247603833865814, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10171 + }, + { + "epoch": 16.249201277955272, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10172 + }, + { + "epoch": 16.250798722044728, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10173 + }, + { + "epoch": 16.252396166134186, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 10174 + }, + { + "epoch": 16.25399361022364, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10175 + }, + { + "epoch": 16.2555910543131, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 10176 + }, + { + "epoch": 16.257188498402556, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10177 + }, + { + "epoch": 16.25878594249201, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10178 + }, + { + "epoch": 16.26038338658147, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10179 + }, + { + "epoch": 16.261980830670925, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10180 + }, + { + "epoch": 16.263578274760384, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10181 + }, + { + "epoch": 16.26517571884984, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10182 + }, + { + "epoch": 16.266773162939298, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10183 + }, + { + "epoch": 16.268370607028753, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10184 + }, + { + "epoch": 16.269968051118212, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10185 + }, + { + "epoch": 16.271565495207668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10186 + }, + { + "epoch": 16.273162939297123, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10187 + }, + { + "epoch": 16.27476038338658, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10188 + }, + { + "epoch": 16.276357827476037, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10189 + }, + { + "epoch": 16.277955271565496, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10190 + }, + { + "epoch": 16.27955271565495, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10191 + }, + { + "epoch": 16.28115015974441, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10192 + }, + { + "epoch": 16.282747603833865, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10193 + }, + { + "epoch": 16.284345047923324, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10194 + }, + { + "epoch": 16.28594249201278, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10195 + }, + { + "epoch": 16.287539936102238, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10196 + }, + { + "epoch": 16.289137380191693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10197 + }, + { + "epoch": 16.29073482428115, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10198 + }, + { + "epoch": 16.292332268370608, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10199 + }, + { + "epoch": 16.293929712460063, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10200 + }, + { + "epoch": 16.29552715654952, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 10201 + }, + { + "epoch": 16.297124600638977, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10202 + }, + { + "epoch": 16.298722044728436, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10203 + }, + { + "epoch": 16.30031948881789, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10204 + }, + { + "epoch": 16.30191693290735, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10205 + }, + { + "epoch": 16.303514376996805, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 10206 + }, + { + "epoch": 16.30511182108626, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10207 + }, + { + "epoch": 16.30670926517572, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10208 + }, + { + "epoch": 16.308306709265175, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10209 + }, + { + "epoch": 16.309904153354633, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10210 + }, + { + "epoch": 16.31150159744409, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10211 + }, + { + "epoch": 16.313099041533548, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10212 + }, + { + "epoch": 16.314696485623003, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10213 + }, + { + "epoch": 16.31629392971246, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10214 + }, + { + "epoch": 16.317891373801917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 10215 + }, + { + "epoch": 16.319488817891372, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10216 + }, + { + "epoch": 16.32108626198083, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10217 + }, + { + "epoch": 16.322683706070286, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10218 + }, + { + "epoch": 16.324281150159745, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10219 + }, + { + "epoch": 16.3258785942492, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10220 + }, + { + "epoch": 16.32747603833866, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10221 + }, + { + "epoch": 16.329073482428115, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10222 + }, + { + "epoch": 16.330670926517573, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10223 + }, + { + "epoch": 16.33226837060703, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10224 + }, + { + "epoch": 16.333865814696484, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10225 + }, + { + "epoch": 16.335463258785943, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10226 + }, + { + "epoch": 16.337060702875398, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 10227 + }, + { + "epoch": 16.338658146964857, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10228 + }, + { + "epoch": 16.340255591054312, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10229 + }, + { + "epoch": 16.34185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10230 + }, + { + "epoch": 16.343450479233226, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10231 + }, + { + "epoch": 16.345047923322685, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 10232 + }, + { + "epoch": 16.34664536741214, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10233 + }, + { + "epoch": 16.3482428115016, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10234 + }, + { + "epoch": 16.349840255591054, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10235 + }, + { + "epoch": 16.35143769968051, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10236 + }, + { + "epoch": 16.35303514376997, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10237 + }, + { + "epoch": 16.354632587859424, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 10238 + }, + { + "epoch": 16.356230031948883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10239 + }, + { + "epoch": 16.357827476038338, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10240 + }, + { + "epoch": 16.359424920127797, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 10241 + }, + { + "epoch": 16.361022364217252, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 10242 + }, + { + "epoch": 16.36261980830671, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10243 + }, + { + "epoch": 16.364217252396166, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10244 + }, + { + "epoch": 16.36581469648562, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10245 + }, + { + "epoch": 16.36741214057508, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10246 + }, + { + "epoch": 16.369009584664536, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10247 + }, + { + "epoch": 16.370607028753994, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10248 + }, + { + "epoch": 16.37220447284345, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10249 + }, + { + "epoch": 16.37380191693291, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10250 + }, + { + "epoch": 16.375399361022364, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10251 + }, + { + "epoch": 16.376996805111823, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10252 + }, + { + "epoch": 16.378594249201278, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10253 + }, + { + "epoch": 16.380191693290733, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 10254 + }, + { + "epoch": 16.381789137380192, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 10255 + }, + { + "epoch": 16.383386581469647, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10256 + }, + { + "epoch": 16.384984025559106, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10257 + }, + { + "epoch": 16.38658146964856, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10258 + }, + { + "epoch": 16.38817891373802, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10259 + }, + { + "epoch": 16.389776357827476, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10260 + }, + { + "epoch": 16.391373801916934, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10261 + }, + { + "epoch": 16.39297124600639, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10262 + }, + { + "epoch": 16.394568690095845, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10263 + }, + { + "epoch": 16.396166134185304, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10264 + }, + { + "epoch": 16.39776357827476, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10265 + }, + { + "epoch": 16.399361022364218, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10266 + }, + { + "epoch": 16.400958466453673, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10267 + }, + { + "epoch": 16.402555910543132, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10268 + }, + { + "epoch": 16.404153354632587, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10269 + }, + { + "epoch": 16.405750798722046, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10270 + }, + { + "epoch": 16.4073482428115, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10271 + }, + { + "epoch": 16.408945686900957, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10272 + }, + { + "epoch": 16.410543130990416, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10273 + }, + { + "epoch": 16.41214057507987, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10274 + }, + { + "epoch": 16.41373801916933, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10275 + }, + { + "epoch": 16.415335463258785, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10276 + }, + { + "epoch": 16.416932907348244, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10277 + }, + { + "epoch": 16.4185303514377, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10278 + }, + { + "epoch": 16.420127795527158, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10279 + }, + { + "epoch": 16.421725239616613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 10280 + }, + { + "epoch": 16.423322683706072, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10281 + }, + { + "epoch": 16.424920127795527, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10282 + }, + { + "epoch": 16.426517571884983, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10283 + }, + { + "epoch": 16.42811501597444, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10284 + }, + { + "epoch": 16.429712460063897, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10285 + }, + { + "epoch": 16.431309904153355, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10286 + }, + { + "epoch": 16.43290734824281, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10287 + }, + { + "epoch": 16.43450479233227, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10288 + }, + { + "epoch": 16.436102236421725, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10289 + }, + { + "epoch": 16.437699680511184, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 10290 + }, + { + "epoch": 16.43929712460064, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10291 + }, + { + "epoch": 16.440894568690094, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10292 + }, + { + "epoch": 16.442492012779553, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 10293 + }, + { + "epoch": 16.44408945686901, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10294 + }, + { + "epoch": 16.445686900958467, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10295 + }, + { + "epoch": 16.447284345047922, + "grad_norm": 0.031494140625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10296 + }, + { + "epoch": 16.44888178913738, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10297 + }, + { + "epoch": 16.450479233226837, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 10298 + }, + { + "epoch": 16.452076677316295, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 10299 + }, + { + "epoch": 16.45367412140575, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10300 + }, + { + "epoch": 16.455271565495206, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 10301 + }, + { + "epoch": 16.456869009584665, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 10302 + }, + { + "epoch": 16.45846645367412, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10303 + }, + { + "epoch": 16.46006389776358, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10304 + }, + { + "epoch": 16.461661341853034, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10305 + }, + { + "epoch": 16.463258785942493, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10306 + }, + { + "epoch": 16.46485623003195, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10307 + }, + { + "epoch": 16.466453674121407, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10308 + }, + { + "epoch": 16.468051118210862, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10309 + }, + { + "epoch": 16.46964856230032, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10310 + }, + { + "epoch": 16.471246006389777, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 10311 + }, + { + "epoch": 16.472843450479232, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10312 + }, + { + "epoch": 16.47444089456869, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10313 + }, + { + "epoch": 16.476038338658146, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10314 + }, + { + "epoch": 16.477635782747605, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10315 + }, + { + "epoch": 16.47923322683706, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10316 + }, + { + "epoch": 16.48083067092652, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10317 + }, + { + "epoch": 16.482428115015974, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10318 + }, + { + "epoch": 16.484025559105433, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10319 + }, + { + "epoch": 16.48562300319489, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 10320 + }, + { + "epoch": 16.487220447284344, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 10321 + }, + { + "epoch": 16.488817891373802, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10322 + }, + { + "epoch": 16.490415335463258, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 10323 + }, + { + "epoch": 16.492012779552716, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10324 + }, + { + "epoch": 16.49361022364217, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10325 + }, + { + "epoch": 16.49520766773163, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10326 + }, + { + "epoch": 16.496805111821086, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10327 + }, + { + "epoch": 16.498402555910545, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 10328 + }, + { + "epoch": 16.5, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10329 + }, + { + "epoch": 16.501597444089455, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10330 + }, + { + "epoch": 16.503194888178914, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10331 + }, + { + "epoch": 16.50479233226837, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10332 + }, + { + "epoch": 16.50638977635783, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10333 + }, + { + "epoch": 16.507987220447284, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10334 + }, + { + "epoch": 16.509584664536742, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10335 + }, + { + "epoch": 16.511182108626198, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10336 + }, + { + "epoch": 16.512779552715656, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10337 + }, + { + "epoch": 16.51437699680511, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 10338 + }, + { + "epoch": 16.515974440894567, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10339 + }, + { + "epoch": 16.517571884984026, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10340 + }, + { + "epoch": 16.51916932907348, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 10341 + }, + { + "epoch": 16.52076677316294, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10342 + }, + { + "epoch": 16.522364217252395, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10343 + }, + { + "epoch": 16.523961661341854, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 10344 + }, + { + "epoch": 16.52555910543131, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10345 + }, + { + "epoch": 16.527156549520768, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 10346 + }, + { + "epoch": 16.528753993610223, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 10347 + }, + { + "epoch": 16.53035143769968, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 10348 + }, + { + "epoch": 16.531948881789138, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10349 + }, + { + "epoch": 16.533546325878593, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10350 + }, + { + "epoch": 16.53514376996805, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10351 + }, + { + "epoch": 16.536741214057507, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10352 + }, + { + "epoch": 16.538338658146966, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10353 + }, + { + "epoch": 16.53993610223642, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10354 + }, + { + "epoch": 16.54153354632588, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10355 + }, + { + "epoch": 16.543130990415335, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10356 + }, + { + "epoch": 16.544728434504794, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10357 + }, + { + "epoch": 16.54632587859425, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10358 + }, + { + "epoch": 16.547923322683705, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10359 + }, + { + "epoch": 16.549520766773163, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10360 + }, + { + "epoch": 16.55111821086262, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10361 + }, + { + "epoch": 16.552715654952078, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10362 + }, + { + "epoch": 16.554313099041533, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 10363 + }, + { + "epoch": 16.55591054313099, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10364 + }, + { + "epoch": 16.557507987220447, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10365 + }, + { + "epoch": 16.559105431309906, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10366 + }, + { + "epoch": 16.56070287539936, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10367 + }, + { + "epoch": 16.562300319488816, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10368 + }, + { + "epoch": 16.563897763578275, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10369 + }, + { + "epoch": 16.56549520766773, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 10370 + }, + { + "epoch": 16.56709265175719, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10371 + }, + { + "epoch": 16.568690095846645, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10372 + }, + { + "epoch": 16.570287539936103, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 10373 + }, + { + "epoch": 16.57188498402556, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10374 + }, + { + "epoch": 16.573482428115017, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10375 + }, + { + "epoch": 16.575079872204473, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10376 + }, + { + "epoch": 16.576677316293928, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10377 + }, + { + "epoch": 16.578274760383387, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10378 + }, + { + "epoch": 16.579872204472842, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10379 + }, + { + "epoch": 16.5814696485623, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 10380 + }, + { + "epoch": 16.583067092651756, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10381 + }, + { + "epoch": 16.584664536741215, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10382 + }, + { + "epoch": 16.58626198083067, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10383 + }, + { + "epoch": 16.58785942492013, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10384 + }, + { + "epoch": 16.589456869009584, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10385 + }, + { + "epoch": 16.591054313099043, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10386 + }, + { + "epoch": 16.5926517571885, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10387 + }, + { + "epoch": 16.594249201277954, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10388 + }, + { + "epoch": 16.595846645367413, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10389 + }, + { + "epoch": 16.597444089456868, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10390 + }, + { + "epoch": 16.599041533546327, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 10391 + }, + { + "epoch": 16.600638977635782, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10392 + }, + { + "epoch": 16.60223642172524, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10393 + }, + { + "epoch": 16.603833865814696, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10394 + }, + { + "epoch": 16.605431309904155, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10395 + }, + { + "epoch": 16.60702875399361, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10396 + }, + { + "epoch": 16.608626198083066, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10397 + }, + { + "epoch": 16.610223642172524, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10398 + }, + { + "epoch": 16.61182108626198, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10399 + }, + { + "epoch": 16.61341853035144, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10400 + }, + { + "epoch": 16.615015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 10401 + }, + { + "epoch": 16.616613418530353, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10402 + }, + { + "epoch": 16.618210862619808, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10403 + }, + { + "epoch": 16.619808306709267, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10404 + }, + { + "epoch": 16.621405750798722, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10405 + }, + { + "epoch": 16.623003194888177, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10406 + }, + { + "epoch": 16.624600638977636, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10407 + }, + { + "epoch": 16.62619808306709, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10408 + }, + { + "epoch": 16.62779552715655, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10409 + }, + { + "epoch": 16.629392971246006, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10410 + }, + { + "epoch": 16.630990415335464, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10411 + }, + { + "epoch": 16.63258785942492, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10412 + }, + { + "epoch": 16.63418530351438, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10413 + }, + { + "epoch": 16.635782747603834, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 10414 + }, + { + "epoch": 16.63738019169329, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10415 + }, + { + "epoch": 16.638977635782748, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 10416 + }, + { + "epoch": 16.640575079872203, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 10417 + }, + { + "epoch": 16.642172523961662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10418 + }, + { + "epoch": 16.643769968051117, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10419 + }, + { + "epoch": 16.645367412140576, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10420 + }, + { + "epoch": 16.64696485623003, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10421 + }, + { + "epoch": 16.64856230031949, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10422 + }, + { + "epoch": 16.650159744408946, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10423 + }, + { + "epoch": 16.6517571884984, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10424 + }, + { + "epoch": 16.65335463258786, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10425 + }, + { + "epoch": 16.654952076677315, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10426 + }, + { + "epoch": 16.656549520766774, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10427 + }, + { + "epoch": 16.65814696485623, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 10428 + }, + { + "epoch": 16.659744408945688, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10429 + }, + { + "epoch": 16.661341853035143, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10430 + }, + { + "epoch": 16.662939297124602, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10431 + }, + { + "epoch": 16.664536741214057, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 10432 + }, + { + "epoch": 16.666134185303516, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10433 + }, + { + "epoch": 16.66773162939297, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10434 + }, + { + "epoch": 16.669329073482427, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10435 + }, + { + "epoch": 16.670926517571885, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10436 + }, + { + "epoch": 16.67252396166134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10437 + }, + { + "epoch": 16.6741214057508, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10438 + }, + { + "epoch": 16.675718849840255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10439 + }, + { + "epoch": 16.677316293929714, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10440 + }, + { + "epoch": 16.67891373801917, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10441 + }, + { + "epoch": 16.680511182108628, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10442 + }, + { + "epoch": 16.682108626198083, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10443 + }, + { + "epoch": 16.68370607028754, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10444 + }, + { + "epoch": 16.685303514376997, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 10445 + }, + { + "epoch": 16.686900958466452, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10446 + }, + { + "epoch": 16.68849840255591, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10447 + }, + { + "epoch": 16.690095846645367, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10448 + }, + { + "epoch": 16.691693290734825, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 10449 + }, + { + "epoch": 16.69329073482428, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10450 + }, + { + "epoch": 16.69488817891374, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10451 + }, + { + "epoch": 16.696485623003195, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10452 + }, + { + "epoch": 16.69808306709265, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10453 + }, + { + "epoch": 16.69968051118211, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10454 + }, + { + "epoch": 16.701277955271564, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10455 + }, + { + "epoch": 16.702875399361023, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10456 + }, + { + "epoch": 16.70447284345048, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10457 + }, + { + "epoch": 16.706070287539937, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10458 + }, + { + "epoch": 16.707667731629392, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 10459 + }, + { + "epoch": 16.70926517571885, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10460 + }, + { + "epoch": 16.710862619808307, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10461 + }, + { + "epoch": 16.712460063897765, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10462 + }, + { + "epoch": 16.71405750798722, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10463 + }, + { + "epoch": 16.715654952076676, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10464 + }, + { + "epoch": 16.717252396166135, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10465 + }, + { + "epoch": 16.71884984025559, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10466 + }, + { + "epoch": 16.72044728434505, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 10467 + }, + { + "epoch": 16.722044728434504, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10468 + }, + { + "epoch": 16.723642172523963, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10469 + }, + { + "epoch": 16.72523961661342, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10470 + }, + { + "epoch": 16.726837060702877, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10471 + }, + { + "epoch": 16.728434504792332, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10472 + }, + { + "epoch": 16.730031948881788, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10473 + }, + { + "epoch": 16.731629392971247, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10474 + }, + { + "epoch": 16.733226837060702, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10475 + }, + { + "epoch": 16.73482428115016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10476 + }, + { + "epoch": 16.736421725239616, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10477 + }, + { + "epoch": 16.738019169329075, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10478 + }, + { + "epoch": 16.73961661341853, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10479 + }, + { + "epoch": 16.74121405750799, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 10480 + }, + { + "epoch": 16.742811501597444, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10481 + }, + { + "epoch": 16.7444089456869, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10482 + }, + { + "epoch": 16.74600638977636, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 10483 + }, + { + "epoch": 16.747603833865814, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10484 + }, + { + "epoch": 16.749201277955272, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10485 + }, + { + "epoch": 16.750798722044728, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10486 + }, + { + "epoch": 16.752396166134186, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10487 + }, + { + "epoch": 16.75399361022364, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 10488 + }, + { + "epoch": 16.7555910543131, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10489 + }, + { + "epoch": 16.757188498402556, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10490 + }, + { + "epoch": 16.75878594249201, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10491 + }, + { + "epoch": 16.76038338658147, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10492 + }, + { + "epoch": 16.761980830670925, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10493 + }, + { + "epoch": 16.763578274760384, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10494 + }, + { + "epoch": 16.76517571884984, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10495 + }, + { + "epoch": 16.766773162939298, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10496 + }, + { + "epoch": 16.768370607028753, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10497 + }, + { + "epoch": 16.769968051118212, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10498 + }, + { + "epoch": 16.771565495207668, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10499 + }, + { + "epoch": 16.773162939297123, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10500 + }, + { + "epoch": 16.77476038338658, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10501 + }, + { + "epoch": 16.776357827476037, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10502 + }, + { + "epoch": 16.777955271565496, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10503 + }, + { + "epoch": 16.77955271565495, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10504 + }, + { + "epoch": 16.78115015974441, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10505 + }, + { + "epoch": 16.782747603833865, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10506 + }, + { + "epoch": 16.784345047923324, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10507 + }, + { + "epoch": 16.78594249201278, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10508 + }, + { + "epoch": 16.787539936102235, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10509 + }, + { + "epoch": 16.789137380191693, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10510 + }, + { + "epoch": 16.79073482428115, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10511 + }, + { + "epoch": 16.792332268370608, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10512 + }, + { + "epoch": 16.793929712460063, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 10513 + }, + { + "epoch": 16.79552715654952, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 10514 + }, + { + "epoch": 16.797124600638977, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10515 + }, + { + "epoch": 16.798722044728436, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 10516 + }, + { + "epoch": 16.80031948881789, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10517 + }, + { + "epoch": 16.80191693290735, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 10518 + }, + { + "epoch": 16.803514376996805, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10519 + }, + { + "epoch": 16.80511182108626, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10520 + }, + { + "epoch": 16.80670926517572, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10521 + }, + { + "epoch": 16.808306709265175, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10522 + }, + { + "epoch": 16.809904153354633, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10523 + }, + { + "epoch": 16.81150159744409, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10524 + }, + { + "epoch": 16.813099041533548, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10525 + }, + { + "epoch": 16.814696485623003, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 10526 + }, + { + "epoch": 16.81629392971246, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10527 + }, + { + "epoch": 16.817891373801917, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10528 + }, + { + "epoch": 16.819488817891372, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10529 + }, + { + "epoch": 16.82108626198083, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10530 + }, + { + "epoch": 16.822683706070286, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10531 + }, + { + "epoch": 16.824281150159745, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 10532 + }, + { + "epoch": 16.8258785942492, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10533 + }, + { + "epoch": 16.82747603833866, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10534 + }, + { + "epoch": 16.829073482428115, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10535 + }, + { + "epoch": 16.830670926517573, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10536 + }, + { + "epoch": 16.83226837060703, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10537 + }, + { + "epoch": 16.833865814696484, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10538 + }, + { + "epoch": 16.835463258785943, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10539 + }, + { + "epoch": 16.837060702875398, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10540 + }, + { + "epoch": 16.838658146964857, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10541 + }, + { + "epoch": 16.840255591054312, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 10542 + }, + { + "epoch": 16.84185303514377, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 10543 + }, + { + "epoch": 16.843450479233226, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10544 + }, + { + "epoch": 16.845047923322685, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10545 + }, + { + "epoch": 16.84664536741214, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10546 + }, + { + "epoch": 16.8482428115016, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10547 + }, + { + "epoch": 16.849840255591054, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 10548 + }, + { + "epoch": 16.85143769968051, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10549 + }, + { + "epoch": 16.85303514376997, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10550 + }, + { + "epoch": 16.854632587859424, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10551 + }, + { + "epoch": 16.856230031948883, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10552 + }, + { + "epoch": 16.857827476038338, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10553 + }, + { + "epoch": 16.859424920127797, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10554 + }, + { + "epoch": 16.861022364217252, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10555 + }, + { + "epoch": 16.86261980830671, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10556 + }, + { + "epoch": 16.864217252396166, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 10557 + }, + { + "epoch": 16.86581469648562, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10558 + }, + { + "epoch": 16.86741214057508, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10559 + }, + { + "epoch": 16.869009584664536, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10560 + }, + { + "epoch": 16.870607028753994, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 10561 + }, + { + "epoch": 16.87220447284345, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 10562 + }, + { + "epoch": 16.87380191693291, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10563 + }, + { + "epoch": 16.875399361022364, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 10564 + }, + { + "epoch": 16.876996805111823, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10565 + }, + { + "epoch": 16.878594249201278, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10566 + }, + { + "epoch": 16.880191693290733, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10567 + }, + { + "epoch": 16.881789137380192, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 10568 + }, + { + "epoch": 16.883386581469647, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10569 + }, + { + "epoch": 16.884984025559106, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10570 + }, + { + "epoch": 16.88658146964856, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10571 + }, + { + "epoch": 16.88817891373802, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10572 + }, + { + "epoch": 16.889776357827476, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10573 + }, + { + "epoch": 16.891373801916934, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10574 + }, + { + "epoch": 16.89297124600639, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10575 + }, + { + "epoch": 16.894568690095845, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 10576 + }, + { + "epoch": 16.896166134185304, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 10577 + }, + { + "epoch": 16.89776357827476, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 10578 + }, + { + "epoch": 16.899361022364218, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10579 + }, + { + "epoch": 16.900958466453673, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10580 + }, + { + "epoch": 16.902555910543132, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10581 + }, + { + "epoch": 16.904153354632587, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10582 + }, + { + "epoch": 16.905750798722046, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10583 + }, + { + "epoch": 16.9073482428115, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 10584 + }, + { + "epoch": 16.908945686900957, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 10585 + }, + { + "epoch": 16.910543130990416, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10586 + }, + { + "epoch": 16.91214057507987, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10587 + }, + { + "epoch": 16.91373801916933, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10588 + }, + { + "epoch": 16.915335463258785, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10589 + }, + { + "epoch": 16.916932907348244, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 10590 + }, + { + "epoch": 16.9185303514377, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10591 + }, + { + "epoch": 16.920127795527158, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10592 + }, + { + "epoch": 16.921725239616613, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 10593 + }, + { + "epoch": 16.923322683706072, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10594 + }, + { + "epoch": 16.924920127795527, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10595 + }, + { + "epoch": 16.926517571884983, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10596 + }, + { + "epoch": 16.92811501597444, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10597 + }, + { + "epoch": 16.929712460063897, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10598 + }, + { + "epoch": 16.931309904153355, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10599 + }, + { + "epoch": 16.93290734824281, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10600 + }, + { + "epoch": 16.93450479233227, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10601 + }, + { + "epoch": 16.936102236421725, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10602 + }, + { + "epoch": 16.937699680511184, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 10603 + }, + { + "epoch": 16.93929712460064, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10604 + }, + { + "epoch": 16.940894568690094, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10605 + }, + { + "epoch": 16.942492012779553, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10606 + }, + { + "epoch": 16.94408945686901, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10607 + }, + { + "epoch": 16.945686900958467, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 10608 + }, + { + "epoch": 16.947284345047922, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10609 + }, + { + "epoch": 16.94888178913738, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10610 + }, + { + "epoch": 16.950479233226837, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10611 + }, + { + "epoch": 16.952076677316295, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10612 + }, + { + "epoch": 16.95367412140575, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10613 + }, + { + "epoch": 16.955271565495206, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10614 + }, + { + "epoch": 16.956869009584665, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10615 + }, + { + "epoch": 16.95846645367412, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10616 + }, + { + "epoch": 16.96006389776358, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10617 + }, + { + "epoch": 16.961661341853034, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10618 + }, + { + "epoch": 16.963258785942493, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 10619 + }, + { + "epoch": 16.96485623003195, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10620 + }, + { + "epoch": 16.966453674121407, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10621 + }, + { + "epoch": 16.968051118210862, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10622 + }, + { + "epoch": 16.96964856230032, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10623 + }, + { + "epoch": 16.971246006389777, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10624 + }, + { + "epoch": 16.972843450479232, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10625 + }, + { + "epoch": 16.97444089456869, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10626 + }, + { + "epoch": 16.976038338658146, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10627 + }, + { + "epoch": 16.977635782747605, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10628 + }, + { + "epoch": 16.97923322683706, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10629 + }, + { + "epoch": 16.98083067092652, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10630 + }, + { + "epoch": 16.982428115015974, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10631 + }, + { + "epoch": 16.984025559105433, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10632 + }, + { + "epoch": 16.98562300319489, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10633 + }, + { + "epoch": 16.987220447284344, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10634 + }, + { + "epoch": 16.988817891373802, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 10635 + }, + { + "epoch": 16.990415335463258, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10636 + }, + { + "epoch": 16.992012779552716, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10637 + }, + { + "epoch": 16.99361022364217, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 10638 + }, + { + "epoch": 16.99520766773163, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10639 + }, + { + "epoch": 16.996805111821086, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10640 + }, + { + "epoch": 16.998402555910545, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10641 + }, + { + "epoch": 17.0, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10642 + }, + { + "epoch": 17.001597444089455, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10643 + }, + { + "epoch": 17.003194888178914, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10644 + }, + { + "epoch": 17.00479233226837, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10645 + }, + { + "epoch": 17.00638977635783, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10646 + }, + { + "epoch": 17.007987220447284, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10647 + }, + { + "epoch": 17.009584664536742, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10648 + }, + { + "epoch": 17.011182108626198, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10649 + }, + { + "epoch": 17.012779552715656, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10650 + }, + { + "epoch": 17.01437699680511, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10651 + }, + { + "epoch": 17.015974440894567, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 10652 + }, + { + "epoch": 17.017571884984026, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10653 + }, + { + "epoch": 17.01916932907348, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10654 + }, + { + "epoch": 17.02076677316294, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10655 + }, + { + "epoch": 17.022364217252395, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 10656 + }, + { + "epoch": 17.023961661341854, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10657 + }, + { + "epoch": 17.02555910543131, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10658 + }, + { + "epoch": 17.027156549520768, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10659 + }, + { + "epoch": 17.028753993610223, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10660 + }, + { + "epoch": 17.03035143769968, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10661 + }, + { + "epoch": 17.031948881789138, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10662 + }, + { + "epoch": 17.033546325878593, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10663 + }, + { + "epoch": 17.03514376996805, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10664 + }, + { + "epoch": 17.036741214057507, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 10665 + }, + { + "epoch": 17.038338658146966, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10666 + }, + { + "epoch": 17.03993610223642, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10667 + }, + { + "epoch": 17.04153354632588, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 10668 + }, + { + "epoch": 17.043130990415335, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10669 + }, + { + "epoch": 17.044728434504794, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10670 + }, + { + "epoch": 17.04632587859425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10671 + }, + { + "epoch": 17.047923322683705, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10672 + }, + { + "epoch": 17.049520766773163, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10673 + }, + { + "epoch": 17.05111821086262, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10674 + }, + { + "epoch": 17.052715654952078, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10675 + }, + { + "epoch": 17.054313099041533, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10676 + }, + { + "epoch": 17.05591054313099, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10677 + }, + { + "epoch": 17.057507987220447, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10678 + }, + { + "epoch": 17.059105431309906, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10679 + }, + { + "epoch": 17.06070287539936, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10680 + }, + { + "epoch": 17.062300319488816, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10681 + }, + { + "epoch": 17.063897763578275, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10682 + }, + { + "epoch": 17.06549520766773, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10683 + }, + { + "epoch": 17.06709265175719, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10684 + }, + { + "epoch": 17.068690095846645, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10685 + }, + { + "epoch": 17.070287539936103, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10686 + }, + { + "epoch": 17.07188498402556, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 10687 + }, + { + "epoch": 17.073482428115017, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10688 + }, + { + "epoch": 17.075079872204473, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10689 + }, + { + "epoch": 17.076677316293928, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10690 + }, + { + "epoch": 17.078274760383387, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10691 + }, + { + "epoch": 17.079872204472842, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10692 + }, + { + "epoch": 17.0814696485623, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 10693 + }, + { + "epoch": 17.083067092651756, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 10694 + }, + { + "epoch": 17.084664536741215, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10695 + }, + { + "epoch": 17.08626198083067, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10696 + }, + { + "epoch": 17.08785942492013, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10697 + }, + { + "epoch": 17.089456869009584, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 10698 + }, + { + "epoch": 17.09105431309904, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10699 + }, + { + "epoch": 17.0926517571885, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10700 + }, + { + "epoch": 17.094249201277954, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 10701 + }, + { + "epoch": 17.095846645367413, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 10702 + }, + { + "epoch": 17.097444089456868, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10703 + }, + { + "epoch": 17.099041533546327, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10704 + }, + { + "epoch": 17.100638977635782, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10705 + }, + { + "epoch": 17.10223642172524, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10706 + }, + { + "epoch": 17.103833865814696, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10707 + }, + { + "epoch": 17.105431309904155, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10708 + }, + { + "epoch": 17.10702875399361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10709 + }, + { + "epoch": 17.108626198083066, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10710 + }, + { + "epoch": 17.110223642172524, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10711 + }, + { + "epoch": 17.11182108626198, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10712 + }, + { + "epoch": 17.11341853035144, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10713 + }, + { + "epoch": 17.115015974440894, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10714 + }, + { + "epoch": 17.116613418530353, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10715 + }, + { + "epoch": 17.118210862619808, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10716 + }, + { + "epoch": 17.119808306709267, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10717 + }, + { + "epoch": 17.121405750798722, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 10718 + }, + { + "epoch": 17.123003194888177, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 10719 + }, + { + "epoch": 17.124600638977636, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10720 + }, + { + "epoch": 17.12619808306709, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10721 + }, + { + "epoch": 17.12779552715655, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10722 + }, + { + "epoch": 17.129392971246006, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10723 + }, + { + "epoch": 17.130990415335464, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10724 + }, + { + "epoch": 17.13258785942492, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10725 + }, + { + "epoch": 17.13418530351438, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 10726 + }, + { + "epoch": 17.135782747603834, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10727 + }, + { + "epoch": 17.13738019169329, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10728 + }, + { + "epoch": 17.138977635782748, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10729 + }, + { + "epoch": 17.140575079872203, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 10730 + }, + { + "epoch": 17.142172523961662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10731 + }, + { + "epoch": 17.143769968051117, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10732 + }, + { + "epoch": 17.145367412140576, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10733 + }, + { + "epoch": 17.14696485623003, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10734 + }, + { + "epoch": 17.14856230031949, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 10735 + }, + { + "epoch": 17.150159744408946, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10736 + }, + { + "epoch": 17.1517571884984, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 10737 + }, + { + "epoch": 17.15335463258786, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10738 + }, + { + "epoch": 17.154952076677315, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10739 + }, + { + "epoch": 17.156549520766774, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 10740 + }, + { + "epoch": 17.15814696485623, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10741 + }, + { + "epoch": 17.159744408945688, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 10742 + }, + { + "epoch": 17.161341853035143, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 10743 + }, + { + "epoch": 17.162939297124602, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10744 + }, + { + "epoch": 17.164536741214057, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10745 + }, + { + "epoch": 17.166134185303516, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 10746 + }, + { + "epoch": 17.16773162939297, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 10747 + }, + { + "epoch": 17.169329073482427, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10748 + }, + { + "epoch": 17.170926517571885, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10749 + }, + { + "epoch": 17.17252396166134, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10750 + }, + { + "epoch": 17.1741214057508, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10751 + }, + { + "epoch": 17.175718849840255, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10752 + }, + { + "epoch": 17.177316293929714, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 10753 + }, + { + "epoch": 17.17891373801917, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10754 + }, + { + "epoch": 17.180511182108628, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10755 + }, + { + "epoch": 17.182108626198083, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 10756 + }, + { + "epoch": 17.18370607028754, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 10757 + }, + { + "epoch": 17.185303514376997, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10758 + }, + { + "epoch": 17.186900958466452, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10759 + }, + { + "epoch": 17.18849840255591, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10760 + }, + { + "epoch": 17.190095846645367, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 10761 + }, + { + "epoch": 17.191693290734825, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10762 + }, + { + "epoch": 17.19329073482428, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10763 + }, + { + "epoch": 17.19488817891374, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10764 + }, + { + "epoch": 17.196485623003195, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10765 + }, + { + "epoch": 17.19808306709265, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 10766 + }, + { + "epoch": 17.19968051118211, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10767 + }, + { + "epoch": 17.201277955271564, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10768 + }, + { + "epoch": 17.202875399361023, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10769 + }, + { + "epoch": 17.20447284345048, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10770 + }, + { + "epoch": 17.206070287539937, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10771 + }, + { + "epoch": 17.207667731629392, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10772 + }, + { + "epoch": 17.20926517571885, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 10773 + }, + { + "epoch": 17.210862619808307, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10774 + }, + { + "epoch": 17.212460063897762, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10775 + }, + { + "epoch": 17.21405750798722, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 10776 + }, + { + "epoch": 17.215654952076676, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10777 + }, + { + "epoch": 17.217252396166135, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10778 + }, + { + "epoch": 17.21884984025559, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10779 + }, + { + "epoch": 17.22044728434505, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 10780 + }, + { + "epoch": 17.222044728434504, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10781 + }, + { + "epoch": 17.223642172523963, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 10782 + }, + { + "epoch": 17.22523961661342, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10783 + }, + { + "epoch": 17.226837060702877, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 10784 + }, + { + "epoch": 17.228434504792332, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10785 + }, + { + "epoch": 17.230031948881788, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 10786 + }, + { + "epoch": 17.231629392971247, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 10787 + }, + { + "epoch": 17.233226837060702, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10788 + }, + { + "epoch": 17.23482428115016, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 10789 + }, + { + "epoch": 17.236421725239616, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10790 + }, + { + "epoch": 17.238019169329075, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 10791 + }, + { + "epoch": 17.23961661341853, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10792 + }, + { + "epoch": 17.24121405750799, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 10793 + }, + { + "epoch": 17.242811501597444, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10794 + }, + { + "epoch": 17.2444089456869, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10795 + }, + { + "epoch": 17.24600638977636, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10796 + }, + { + "epoch": 17.247603833865814, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10797 + }, + { + "epoch": 17.249201277955272, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10798 + }, + { + "epoch": 17.250798722044728, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10799 + }, + { + "epoch": 17.252396166134186, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10800 + }, + { + "epoch": 17.25399361022364, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10801 + }, + { + "epoch": 17.2555910543131, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10802 + }, + { + "epoch": 17.257188498402556, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10803 + }, + { + "epoch": 17.25878594249201, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10804 + }, + { + "epoch": 17.26038338658147, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10805 + }, + { + "epoch": 17.261980830670925, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10806 + }, + { + "epoch": 17.263578274760384, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10807 + }, + { + "epoch": 17.26517571884984, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10808 + }, + { + "epoch": 17.266773162939298, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10809 + }, + { + "epoch": 17.268370607028753, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 10810 + }, + { + "epoch": 17.269968051118212, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10811 + }, + { + "epoch": 17.271565495207668, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 10812 + }, + { + "epoch": 17.273162939297123, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10813 + }, + { + "epoch": 17.27476038338658, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 10814 + }, + { + "epoch": 17.276357827476037, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10815 + }, + { + "epoch": 17.277955271565496, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 10816 + }, + { + "epoch": 17.27955271565495, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10817 + }, + { + "epoch": 17.28115015974441, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10818 + }, + { + "epoch": 17.282747603833865, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10819 + }, + { + "epoch": 17.284345047923324, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10820 + }, + { + "epoch": 17.28594249201278, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10821 + }, + { + "epoch": 17.287539936102238, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10822 + }, + { + "epoch": 17.289137380191693, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10823 + }, + { + "epoch": 17.29073482428115, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 10824 + }, + { + "epoch": 17.292332268370608, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 10825 + }, + { + "epoch": 17.293929712460063, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 10826 + }, + { + "epoch": 17.29552715654952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 10827 + }, + { + "epoch": 17.297124600638977, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 10828 + }, + { + "epoch": 17.298722044728436, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 10829 + }, + { + "epoch": 17.30031948881789, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 10830 + }, + { + "epoch": 17.30191693290735, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 10831 + }, + { + "epoch": 17.303514376996805, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 10832 + }, + { + "epoch": 17.30511182108626, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10833 + }, + { + "epoch": 17.30670926517572, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10834 + }, + { + "epoch": 17.308306709265175, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10835 + }, + { + "epoch": 17.309904153354633, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 10836 + }, + { + "epoch": 17.31150159744409, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10837 + }, + { + "epoch": 17.313099041533548, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10838 + }, + { + "epoch": 17.314696485623003, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 10839 + }, + { + "epoch": 17.31629392971246, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10840 + }, + { + "epoch": 17.317891373801917, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 10841 + }, + { + "epoch": 17.319488817891372, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 10842 + }, + { + "epoch": 17.32108626198083, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 10843 + }, + { + "epoch": 17.322683706070286, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10844 + }, + { + "epoch": 17.324281150159745, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10845 + }, + { + "epoch": 17.3258785942492, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10846 + }, + { + "epoch": 17.32747603833866, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10847 + }, + { + "epoch": 17.329073482428115, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 10848 + }, + { + "epoch": 17.330670926517573, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 10849 + }, + { + "epoch": 17.33226837060703, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10850 + }, + { + "epoch": 17.333865814696484, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10851 + }, + { + "epoch": 17.335463258785943, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 10852 + }, + { + "epoch": 17.337060702875398, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10853 + }, + { + "epoch": 17.338658146964857, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 10854 + }, + { + "epoch": 17.340255591054312, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 10855 + }, + { + "epoch": 17.34185303514377, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10856 + }, + { + "epoch": 17.343450479233226, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10857 + }, + { + "epoch": 17.345047923322685, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 10858 + }, + { + "epoch": 17.34664536741214, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10859 + }, + { + "epoch": 17.3482428115016, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 10860 + }, + { + "epoch": 17.349840255591054, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10861 + }, + { + "epoch": 17.35143769968051, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 10862 + }, + { + "epoch": 17.35303514376997, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10863 + }, + { + "epoch": 17.354632587859424, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10864 + }, + { + "epoch": 17.356230031948883, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10865 + }, + { + "epoch": 17.357827476038338, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10866 + }, + { + "epoch": 17.359424920127797, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 10867 + }, + { + "epoch": 17.361022364217252, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10868 + }, + { + "epoch": 17.36261980830671, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 10869 + }, + { + "epoch": 17.364217252396166, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10870 + }, + { + "epoch": 17.36581469648562, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 10871 + }, + { + "epoch": 17.36741214057508, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10872 + }, + { + "epoch": 17.369009584664536, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10873 + }, + { + "epoch": 17.370607028753994, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 10874 + }, + { + "epoch": 17.37220447284345, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10875 + }, + { + "epoch": 17.37380191693291, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 10876 + }, + { + "epoch": 17.375399361022364, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 10877 + }, + { + "epoch": 17.376996805111823, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10878 + }, + { + "epoch": 17.378594249201278, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10879 + }, + { + "epoch": 17.380191693290733, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10880 + }, + { + "epoch": 17.381789137380192, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 10881 + }, + { + "epoch": 17.383386581469647, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 10882 + }, + { + "epoch": 17.384984025559106, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 10883 + }, + { + "epoch": 17.38658146964856, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10884 + }, + { + "epoch": 17.38817891373802, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 10885 + }, + { + "epoch": 17.389776357827476, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 10886 + }, + { + "epoch": 17.391373801916934, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 10887 + }, + { + "epoch": 17.39297124600639, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10888 + }, + { + "epoch": 17.394568690095845, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10889 + }, + { + "epoch": 17.396166134185304, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 10890 + }, + { + "epoch": 17.39776357827476, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 10891 + }, + { + "epoch": 17.399361022364218, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 10892 + }, + { + "epoch": 17.400958466453673, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 10893 + }, + { + "epoch": 17.402555910543132, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 10894 + }, + { + "epoch": 17.404153354632587, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 10895 + }, + { + "epoch": 17.405750798722046, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 10896 + }, + { + "epoch": 17.4073482428115, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10897 + }, + { + "epoch": 17.408945686900957, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 10898 + }, + { + "epoch": 17.410543130990416, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 10899 + }, + { + "epoch": 17.41214057507987, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 10900 + }, + { + "epoch": 17.41373801916933, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 10901 + }, + { + "epoch": 17.415335463258785, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 10902 + }, + { + "epoch": 17.416932907348244, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 10903 + }, + { + "epoch": 17.4185303514377, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 10904 + }, + { + "epoch": 17.420127795527158, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10905 + }, + { + "epoch": 17.421725239616613, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 10906 + }, + { + "epoch": 17.423322683706072, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10907 + }, + { + "epoch": 17.424920127795527, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10908 + }, + { + "epoch": 17.426517571884983, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10909 + }, + { + "epoch": 17.42811501597444, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10910 + }, + { + "epoch": 17.429712460063897, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 10911 + }, + { + "epoch": 17.431309904153355, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10912 + }, + { + "epoch": 17.43290734824281, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10913 + }, + { + "epoch": 17.43450479233227, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 10914 + }, + { + "epoch": 17.436102236421725, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10915 + }, + { + "epoch": 17.437699680511184, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10916 + }, + { + "epoch": 17.43929712460064, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 10917 + }, + { + "epoch": 17.440894568690094, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 10918 + }, + { + "epoch": 17.442492012779553, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10919 + }, + { + "epoch": 17.44408945686901, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10920 + }, + { + "epoch": 17.445686900958467, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 10921 + }, + { + "epoch": 17.447284345047922, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10922 + }, + { + "epoch": 17.44888178913738, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 10923 + }, + { + "epoch": 17.450479233226837, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10924 + }, + { + "epoch": 17.452076677316295, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10925 + }, + { + "epoch": 17.45367412140575, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10926 + }, + { + "epoch": 17.455271565495206, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10927 + }, + { + "epoch": 17.456869009584665, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 10928 + }, + { + "epoch": 17.45846645367412, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 10929 + }, + { + "epoch": 17.46006389776358, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10930 + }, + { + "epoch": 17.461661341853034, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 10931 + }, + { + "epoch": 17.463258785942493, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 10932 + }, + { + "epoch": 17.46485623003195, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 10933 + }, + { + "epoch": 17.466453674121407, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 10934 + }, + { + "epoch": 17.468051118210862, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 10935 + }, + { + "epoch": 17.46964856230032, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10936 + }, + { + "epoch": 17.471246006389777, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 10937 + }, + { + "epoch": 17.472843450479232, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10938 + }, + { + "epoch": 17.47444089456869, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 10939 + }, + { + "epoch": 17.476038338658146, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10940 + }, + { + "epoch": 17.477635782747605, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 10941 + }, + { + "epoch": 17.47923322683706, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 10942 + }, + { + "epoch": 17.48083067092652, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10943 + }, + { + "epoch": 17.482428115015974, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10944 + }, + { + "epoch": 17.484025559105433, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10945 + }, + { + "epoch": 17.48562300319489, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 10946 + }, + { + "epoch": 17.487220447284344, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 10947 + }, + { + "epoch": 17.488817891373802, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 10948 + }, + { + "epoch": 17.490415335463258, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 10949 + }, + { + "epoch": 17.492012779552716, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10950 + }, + { + "epoch": 17.49361022364217, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 10951 + }, + { + "epoch": 17.49520766773163, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 10952 + }, + { + "epoch": 17.496805111821086, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 10953 + }, + { + "epoch": 17.498402555910545, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 10954 + }, + { + "epoch": 17.5, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 10955 + }, + { + "epoch": 17.501597444089455, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 10956 + }, + { + "epoch": 17.503194888178914, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 10957 + }, + { + "epoch": 17.50479233226837, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 10958 + }, + { + "epoch": 17.50638977635783, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10959 + }, + { + "epoch": 17.507987220447284, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 10960 + }, + { + "epoch": 17.509584664536742, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10961 + }, + { + "epoch": 17.511182108626198, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10962 + }, + { + "epoch": 17.512779552715656, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 10963 + }, + { + "epoch": 17.51437699680511, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 10964 + }, + { + "epoch": 17.515974440894567, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10965 + }, + { + "epoch": 17.517571884984026, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 10966 + }, + { + "epoch": 17.51916932907348, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 10967 + }, + { + "epoch": 17.52076677316294, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10968 + }, + { + "epoch": 17.522364217252395, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 10969 + }, + { + "epoch": 17.523961661341854, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10970 + }, + { + "epoch": 17.52555910543131, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 10971 + }, + { + "epoch": 17.527156549520768, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 10972 + }, + { + "epoch": 17.528753993610223, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 10973 + }, + { + "epoch": 17.53035143769968, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 10974 + }, + { + "epoch": 17.531948881789138, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 10975 + }, + { + "epoch": 17.533546325878593, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 10976 + }, + { + "epoch": 17.53514376996805, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 10977 + }, + { + "epoch": 17.536741214057507, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 10978 + }, + { + "epoch": 17.538338658146966, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 10979 + }, + { + "epoch": 17.53993610223642, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 10980 + }, + { + "epoch": 17.54153354632588, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 10981 + }, + { + "epoch": 17.543130990415335, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 10982 + }, + { + "epoch": 17.544728434504794, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 10983 + }, + { + "epoch": 17.54632587859425, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 10984 + }, + { + "epoch": 17.547923322683705, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 10985 + }, + { + "epoch": 17.549520766773163, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 10986 + }, + { + "epoch": 17.55111821086262, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 10987 + }, + { + "epoch": 17.552715654952078, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 10988 + }, + { + "epoch": 17.554313099041533, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 10989 + }, + { + "epoch": 17.55591054313099, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 10990 + }, + { + "epoch": 17.557507987220447, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 10991 + }, + { + "epoch": 17.559105431309906, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 10992 + }, + { + "epoch": 17.56070287539936, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 10993 + }, + { + "epoch": 17.562300319488816, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 10994 + }, + { + "epoch": 17.563897763578275, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 10995 + }, + { + "epoch": 17.56549520766773, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 10996 + }, + { + "epoch": 17.56709265175719, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 10997 + }, + { + "epoch": 17.568690095846645, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 10998 + }, + { + "epoch": 17.570287539936103, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 10999 + }, + { + "epoch": 17.57188498402556, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 11000 + }, + { + "epoch": 17.573482428115017, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11001 + }, + { + "epoch": 17.575079872204473, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11002 + }, + { + "epoch": 17.576677316293928, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11003 + }, + { + "epoch": 17.578274760383387, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11004 + }, + { + "epoch": 17.579872204472842, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11005 + }, + { + "epoch": 17.5814696485623, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11006 + }, + { + "epoch": 17.583067092651756, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11007 + }, + { + "epoch": 17.584664536741215, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11008 + }, + { + "epoch": 17.58626198083067, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 11009 + }, + { + "epoch": 17.58785942492013, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11010 + }, + { + "epoch": 17.589456869009584, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11011 + }, + { + "epoch": 17.591054313099043, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11012 + }, + { + "epoch": 17.5926517571885, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11013 + }, + { + "epoch": 17.594249201277954, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11014 + }, + { + "epoch": 17.595846645367413, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11015 + }, + { + "epoch": 17.597444089456868, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11016 + }, + { + "epoch": 17.599041533546327, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 11017 + }, + { + "epoch": 17.600638977635782, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 11018 + }, + { + "epoch": 17.60223642172524, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11019 + }, + { + "epoch": 17.603833865814696, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11020 + }, + { + "epoch": 17.605431309904155, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11021 + }, + { + "epoch": 17.60702875399361, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11022 + }, + { + "epoch": 17.608626198083066, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11023 + }, + { + "epoch": 17.610223642172524, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11024 + }, + { + "epoch": 17.61182108626198, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11025 + }, + { + "epoch": 17.61341853035144, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11026 + }, + { + "epoch": 17.615015974440894, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11027 + }, + { + "epoch": 17.616613418530353, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11028 + }, + { + "epoch": 17.618210862619808, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11029 + }, + { + "epoch": 17.619808306709267, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11030 + }, + { + "epoch": 17.621405750798722, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11031 + }, + { + "epoch": 17.623003194888177, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11032 + }, + { + "epoch": 17.624600638977636, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11033 + }, + { + "epoch": 17.62619808306709, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11034 + }, + { + "epoch": 17.62779552715655, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11035 + }, + { + "epoch": 17.629392971246006, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11036 + }, + { + "epoch": 17.630990415335464, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 11037 + }, + { + "epoch": 17.63258785942492, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11038 + }, + { + "epoch": 17.63418530351438, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11039 + }, + { + "epoch": 17.635782747603834, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 11040 + }, + { + "epoch": 17.63738019169329, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11041 + }, + { + "epoch": 17.638977635782748, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11042 + }, + { + "epoch": 17.640575079872203, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 11043 + }, + { + "epoch": 17.642172523961662, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11044 + }, + { + "epoch": 17.643769968051117, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11045 + }, + { + "epoch": 17.645367412140576, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11046 + }, + { + "epoch": 17.64696485623003, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11047 + }, + { + "epoch": 17.64856230031949, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11048 + }, + { + "epoch": 17.650159744408946, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11049 + }, + { + "epoch": 17.6517571884984, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11050 + }, + { + "epoch": 17.65335463258786, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11051 + }, + { + "epoch": 17.654952076677315, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11052 + }, + { + "epoch": 17.656549520766774, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11053 + }, + { + "epoch": 17.65814696485623, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11054 + }, + { + "epoch": 17.659744408945688, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11055 + }, + { + "epoch": 17.661341853035143, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11056 + }, + { + "epoch": 17.662939297124602, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11057 + }, + { + "epoch": 17.664536741214057, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11058 + }, + { + "epoch": 17.666134185303516, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11059 + }, + { + "epoch": 17.66773162939297, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11060 + }, + { + "epoch": 17.669329073482427, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11061 + }, + { + "epoch": 17.670926517571885, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11062 + }, + { + "epoch": 17.67252396166134, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11063 + }, + { + "epoch": 17.6741214057508, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 11064 + }, + { + "epoch": 17.675718849840255, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11065 + }, + { + "epoch": 17.677316293929714, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11066 + }, + { + "epoch": 17.67891373801917, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11067 + }, + { + "epoch": 17.680511182108628, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11068 + }, + { + "epoch": 17.682108626198083, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11069 + }, + { + "epoch": 17.68370607028754, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11070 + }, + { + "epoch": 17.685303514376997, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11071 + }, + { + "epoch": 17.686900958466452, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11072 + }, + { + "epoch": 17.68849840255591, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11073 + }, + { + "epoch": 17.690095846645367, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11074 + }, + { + "epoch": 17.691693290734825, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11075 + }, + { + "epoch": 17.69329073482428, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11076 + }, + { + "epoch": 17.69488817891374, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11077 + }, + { + "epoch": 17.696485623003195, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11078 + }, + { + "epoch": 17.69808306709265, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 11079 + }, + { + "epoch": 17.69968051118211, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 11080 + }, + { + "epoch": 17.701277955271564, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11081 + }, + { + "epoch": 17.702875399361023, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11082 + }, + { + "epoch": 17.70447284345048, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11083 + }, + { + "epoch": 17.706070287539937, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 11084 + }, + { + "epoch": 17.707667731629392, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11085 + }, + { + "epoch": 17.70926517571885, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11086 + }, + { + "epoch": 17.710862619808307, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 11087 + }, + { + "epoch": 17.712460063897765, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11088 + }, + { + "epoch": 17.71405750798722, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 11089 + }, + { + "epoch": 17.715654952076676, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11090 + }, + { + "epoch": 17.717252396166135, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11091 + }, + { + "epoch": 17.71884984025559, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11092 + }, + { + "epoch": 17.72044728434505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11093 + }, + { + "epoch": 17.722044728434504, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11094 + }, + { + "epoch": 17.723642172523963, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 11095 + }, + { + "epoch": 17.72523961661342, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11096 + }, + { + "epoch": 17.726837060702877, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11097 + }, + { + "epoch": 17.728434504792332, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11098 + }, + { + "epoch": 17.730031948881788, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11099 + }, + { + "epoch": 17.731629392971247, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11100 + }, + { + "epoch": 17.733226837060702, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11101 + }, + { + "epoch": 17.73482428115016, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11102 + }, + { + "epoch": 17.736421725239616, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11103 + }, + { + "epoch": 17.738019169329075, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11104 + }, + { + "epoch": 17.73961661341853, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11105 + }, + { + "epoch": 17.74121405750799, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11106 + }, + { + "epoch": 17.742811501597444, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11107 + }, + { + "epoch": 17.7444089456869, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11108 + }, + { + "epoch": 17.74600638977636, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11109 + }, + { + "epoch": 17.747603833865814, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11110 + }, + { + "epoch": 17.749201277955272, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11111 + }, + { + "epoch": 17.750798722044728, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11112 + }, + { + "epoch": 17.752396166134186, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11113 + }, + { + "epoch": 17.75399361022364, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11114 + }, + { + "epoch": 17.7555910543131, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11115 + }, + { + "epoch": 17.757188498402556, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11116 + }, + { + "epoch": 17.75878594249201, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11117 + }, + { + "epoch": 17.76038338658147, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11118 + }, + { + "epoch": 17.761980830670925, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11119 + }, + { + "epoch": 17.763578274760384, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11120 + }, + { + "epoch": 17.76517571884984, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11121 + }, + { + "epoch": 17.766773162939298, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11122 + }, + { + "epoch": 17.768370607028753, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11123 + }, + { + "epoch": 17.769968051118212, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 11124 + }, + { + "epoch": 17.771565495207668, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11125 + }, + { + "epoch": 17.773162939297123, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 11126 + }, + { + "epoch": 17.77476038338658, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11127 + }, + { + "epoch": 17.776357827476037, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 11128 + }, + { + "epoch": 17.777955271565496, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11129 + }, + { + "epoch": 17.77955271565495, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11130 + }, + { + "epoch": 17.78115015974441, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11131 + }, + { + "epoch": 17.782747603833865, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11132 + }, + { + "epoch": 17.784345047923324, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 11133 + }, + { + "epoch": 17.78594249201278, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11134 + }, + { + "epoch": 17.787539936102235, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11135 + }, + { + "epoch": 17.789137380191693, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11136 + }, + { + "epoch": 17.79073482428115, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11137 + }, + { + "epoch": 17.792332268370608, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11138 + }, + { + "epoch": 17.793929712460063, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11139 + }, + { + "epoch": 17.79552715654952, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11140 + }, + { + "epoch": 17.797124600638977, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11141 + }, + { + "epoch": 17.798722044728436, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11142 + }, + { + "epoch": 17.80031948881789, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11143 + }, + { + "epoch": 17.80191693290735, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11144 + }, + { + "epoch": 17.803514376996805, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11145 + }, + { + "epoch": 17.80511182108626, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11146 + }, + { + "epoch": 17.80670926517572, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11147 + }, + { + "epoch": 17.808306709265175, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 11148 + }, + { + "epoch": 17.809904153354633, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11149 + }, + { + "epoch": 17.81150159744409, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11150 + }, + { + "epoch": 17.813099041533548, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 11151 + }, + { + "epoch": 17.814696485623003, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 11152 + }, + { + "epoch": 17.81629392971246, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11153 + }, + { + "epoch": 17.817891373801917, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11154 + }, + { + "epoch": 17.819488817891372, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11155 + }, + { + "epoch": 17.82108626198083, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11156 + }, + { + "epoch": 17.822683706070286, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11157 + }, + { + "epoch": 17.824281150159745, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11158 + }, + { + "epoch": 17.8258785942492, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11159 + }, + { + "epoch": 17.82747603833866, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11160 + }, + { + "epoch": 17.829073482428115, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11161 + }, + { + "epoch": 17.830670926517573, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 11162 + }, + { + "epoch": 17.83226837060703, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11163 + }, + { + "epoch": 17.833865814696484, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11164 + }, + { + "epoch": 17.835463258785943, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 11165 + }, + { + "epoch": 17.837060702875398, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11166 + }, + { + "epoch": 17.838658146964857, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11167 + }, + { + "epoch": 17.840255591054312, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11168 + }, + { + "epoch": 17.84185303514377, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11169 + }, + { + "epoch": 17.843450479233226, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11170 + }, + { + "epoch": 17.845047923322685, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11171 + }, + { + "epoch": 17.84664536741214, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11172 + }, + { + "epoch": 17.8482428115016, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11173 + }, + { + "epoch": 17.849840255591054, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11174 + }, + { + "epoch": 17.85143769968051, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 11175 + }, + { + "epoch": 17.85303514376997, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11176 + }, + { + "epoch": 17.854632587859424, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11177 + }, + { + "epoch": 17.856230031948883, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11178 + }, + { + "epoch": 17.857827476038338, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11179 + }, + { + "epoch": 17.859424920127797, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11180 + }, + { + "epoch": 17.861022364217252, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11181 + }, + { + "epoch": 17.86261980830671, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11182 + }, + { + "epoch": 17.864217252396166, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11183 + }, + { + "epoch": 17.86581469648562, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 11184 + }, + { + "epoch": 17.86741214057508, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11185 + }, + { + "epoch": 17.869009584664536, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11186 + }, + { + "epoch": 17.870607028753994, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11187 + }, + { + "epoch": 17.87220447284345, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11188 + }, + { + "epoch": 17.87380191693291, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11189 + }, + { + "epoch": 17.875399361022364, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11190 + }, + { + "epoch": 17.876996805111823, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11191 + }, + { + "epoch": 17.878594249201278, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11192 + }, + { + "epoch": 17.880191693290733, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11193 + }, + { + "epoch": 17.881789137380192, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11194 + }, + { + "epoch": 17.883386581469647, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11195 + }, + { + "epoch": 17.884984025559106, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 11196 + }, + { + "epoch": 17.88658146964856, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11197 + }, + { + "epoch": 17.88817891373802, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11198 + }, + { + "epoch": 17.889776357827476, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11199 + }, + { + "epoch": 17.891373801916934, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 11200 + }, + { + "epoch": 17.89297124600639, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11201 + }, + { + "epoch": 17.894568690095845, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11202 + }, + { + "epoch": 17.896166134185304, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11203 + }, + { + "epoch": 17.89776357827476, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11204 + }, + { + "epoch": 17.899361022364218, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11205 + }, + { + "epoch": 17.900958466453673, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11206 + }, + { + "epoch": 17.902555910543132, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11207 + }, + { + "epoch": 17.904153354632587, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11208 + }, + { + "epoch": 17.905750798722046, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11209 + }, + { + "epoch": 17.9073482428115, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11210 + }, + { + "epoch": 17.908945686900957, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11211 + }, + { + "epoch": 17.910543130990416, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11212 + }, + { + "epoch": 17.91214057507987, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11213 + }, + { + "epoch": 17.91373801916933, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11214 + }, + { + "epoch": 17.915335463258785, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11215 + }, + { + "epoch": 17.916932907348244, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11216 + }, + { + "epoch": 17.9185303514377, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11217 + }, + { + "epoch": 17.920127795527158, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11218 + }, + { + "epoch": 17.921725239616613, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11219 + }, + { + "epoch": 17.923322683706072, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 11220 + }, + { + "epoch": 17.924920127795527, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 11221 + }, + { + "epoch": 17.926517571884983, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11222 + }, + { + "epoch": 17.92811501597444, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11223 + }, + { + "epoch": 17.929712460063897, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 11224 + }, + { + "epoch": 17.931309904153355, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11225 + }, + { + "epoch": 17.93290734824281, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11226 + }, + { + "epoch": 17.93450479233227, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11227 + }, + { + "epoch": 17.936102236421725, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11228 + }, + { + "epoch": 17.937699680511184, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11229 + }, + { + "epoch": 17.93929712460064, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11230 + }, + { + "epoch": 17.940894568690094, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11231 + }, + { + "epoch": 17.942492012779553, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 11232 + }, + { + "epoch": 17.94408945686901, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11233 + }, + { + "epoch": 17.945686900958467, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11234 + }, + { + "epoch": 17.947284345047922, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11235 + }, + { + "epoch": 17.94888178913738, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11236 + }, + { + "epoch": 17.950479233226837, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11237 + }, + { + "epoch": 17.952076677316295, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11238 + }, + { + "epoch": 17.95367412140575, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 11239 + }, + { + "epoch": 17.955271565495206, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11240 + }, + { + "epoch": 17.956869009584665, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11241 + }, + { + "epoch": 17.95846645367412, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 11242 + }, + { + "epoch": 17.96006389776358, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11243 + }, + { + "epoch": 17.961661341853034, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11244 + }, + { + "epoch": 17.963258785942493, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11245 + }, + { + "epoch": 17.96485623003195, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11246 + }, + { + "epoch": 17.966453674121407, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 11247 + }, + { + "epoch": 17.968051118210862, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 11248 + }, + { + "epoch": 17.96964856230032, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11249 + }, + { + "epoch": 17.971246006389777, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11250 + }, + { + "epoch": 17.972843450479232, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11251 + }, + { + "epoch": 17.97444089456869, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11252 + }, + { + "epoch": 17.976038338658146, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11253 + }, + { + "epoch": 17.977635782747605, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11254 + }, + { + "epoch": 17.97923322683706, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11255 + }, + { + "epoch": 17.98083067092652, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11256 + }, + { + "epoch": 17.982428115015974, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11257 + }, + { + "epoch": 17.984025559105433, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11258 + }, + { + "epoch": 17.98562300319489, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11259 + }, + { + "epoch": 17.987220447284344, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11260 + }, + { + "epoch": 17.988817891373802, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11261 + }, + { + "epoch": 17.990415335463258, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11262 + }, + { + "epoch": 17.992012779552716, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11263 + }, + { + "epoch": 17.99361022364217, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11264 + }, + { + "epoch": 17.99520766773163, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11265 + }, + { + "epoch": 17.996805111821086, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 11266 + }, + { + "epoch": 17.998402555910545, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11267 + }, + { + "epoch": 18.0, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11268 + }, + { + "epoch": 18.001597444089455, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11269 + }, + { + "epoch": 18.003194888178914, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11270 + }, + { + "epoch": 18.00479233226837, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11271 + }, + { + "epoch": 18.00638977635783, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11272 + }, + { + "epoch": 18.007987220447284, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11273 + }, + { + "epoch": 18.009584664536742, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11274 + }, + { + "epoch": 18.011182108626198, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11275 + }, + { + "epoch": 18.012779552715656, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11276 + }, + { + "epoch": 18.01437699680511, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 11277 + }, + { + "epoch": 18.015974440894567, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11278 + }, + { + "epoch": 18.017571884984026, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11279 + }, + { + "epoch": 18.01916932907348, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11280 + }, + { + "epoch": 18.02076677316294, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11281 + }, + { + "epoch": 18.022364217252395, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11282 + }, + { + "epoch": 18.023961661341854, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11283 + }, + { + "epoch": 18.02555910543131, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11284 + }, + { + "epoch": 18.027156549520768, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11285 + }, + { + "epoch": 18.028753993610223, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11286 + }, + { + "epoch": 18.03035143769968, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11287 + }, + { + "epoch": 18.031948881789138, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11288 + }, + { + "epoch": 18.033546325878593, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11289 + }, + { + "epoch": 18.03514376996805, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11290 + }, + { + "epoch": 18.036741214057507, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11291 + }, + { + "epoch": 18.038338658146966, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11292 + }, + { + "epoch": 18.03993610223642, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11293 + }, + { + "epoch": 18.04153354632588, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11294 + }, + { + "epoch": 18.043130990415335, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11295 + }, + { + "epoch": 18.044728434504794, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11296 + }, + { + "epoch": 18.04632587859425, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11297 + }, + { + "epoch": 18.047923322683705, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11298 + }, + { + "epoch": 18.049520766773163, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11299 + }, + { + "epoch": 18.05111821086262, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11300 + }, + { + "epoch": 18.052715654952078, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11301 + }, + { + "epoch": 18.054313099041533, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11302 + }, + { + "epoch": 18.05591054313099, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11303 + }, + { + "epoch": 18.057507987220447, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11304 + }, + { + "epoch": 18.059105431309906, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11305 + }, + { + "epoch": 18.06070287539936, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11306 + }, + { + "epoch": 18.062300319488816, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11307 + }, + { + "epoch": 18.063897763578275, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11308 + }, + { + "epoch": 18.06549520766773, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11309 + }, + { + "epoch": 18.06709265175719, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11310 + }, + { + "epoch": 18.068690095846645, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11311 + }, + { + "epoch": 18.070287539936103, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11312 + }, + { + "epoch": 18.07188498402556, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11313 + }, + { + "epoch": 18.073482428115017, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11314 + }, + { + "epoch": 18.075079872204473, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11315 + }, + { + "epoch": 18.076677316293928, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11316 + }, + { + "epoch": 18.078274760383387, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11317 + }, + { + "epoch": 18.079872204472842, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11318 + }, + { + "epoch": 18.0814696485623, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11319 + }, + { + "epoch": 18.083067092651756, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11320 + }, + { + "epoch": 18.084664536741215, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11321 + }, + { + "epoch": 18.08626198083067, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11322 + }, + { + "epoch": 18.08785942492013, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11323 + }, + { + "epoch": 18.089456869009584, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11324 + }, + { + "epoch": 18.09105431309904, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11325 + }, + { + "epoch": 18.0926517571885, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11326 + }, + { + "epoch": 18.094249201277954, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11327 + }, + { + "epoch": 18.095846645367413, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11328 + }, + { + "epoch": 18.097444089456868, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11329 + }, + { + "epoch": 18.099041533546327, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11330 + }, + { + "epoch": 18.100638977635782, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11331 + }, + { + "epoch": 18.10223642172524, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11332 + }, + { + "epoch": 18.103833865814696, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11333 + }, + { + "epoch": 18.105431309904155, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11334 + }, + { + "epoch": 18.10702875399361, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11335 + }, + { + "epoch": 18.108626198083066, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11336 + }, + { + "epoch": 18.110223642172524, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11337 + }, + { + "epoch": 18.11182108626198, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11338 + }, + { + "epoch": 18.11341853035144, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11339 + }, + { + "epoch": 18.115015974440894, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11340 + }, + { + "epoch": 18.116613418530353, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11341 + }, + { + "epoch": 18.118210862619808, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11342 + }, + { + "epoch": 18.119808306709267, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11343 + }, + { + "epoch": 18.121405750798722, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11344 + }, + { + "epoch": 18.123003194888177, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 11345 + }, + { + "epoch": 18.124600638977636, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11346 + }, + { + "epoch": 18.12619808306709, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11347 + }, + { + "epoch": 18.12779552715655, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11348 + }, + { + "epoch": 18.129392971246006, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11349 + }, + { + "epoch": 18.130990415335464, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11350 + }, + { + "epoch": 18.13258785942492, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11351 + }, + { + "epoch": 18.13418530351438, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11352 + }, + { + "epoch": 18.135782747603834, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11353 + }, + { + "epoch": 18.13738019169329, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11354 + }, + { + "epoch": 18.138977635782748, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 11355 + }, + { + "epoch": 18.140575079872203, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11356 + }, + { + "epoch": 18.142172523961662, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11357 + }, + { + "epoch": 18.143769968051117, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 11358 + }, + { + "epoch": 18.145367412140576, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11359 + }, + { + "epoch": 18.14696485623003, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 11360 + }, + { + "epoch": 18.14856230031949, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11361 + }, + { + "epoch": 18.150159744408946, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11362 + }, + { + "epoch": 18.1517571884984, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11363 + }, + { + "epoch": 18.15335463258786, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11364 + }, + { + "epoch": 18.154952076677315, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 11365 + }, + { + "epoch": 18.156549520766774, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11366 + }, + { + "epoch": 18.15814696485623, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11367 + }, + { + "epoch": 18.159744408945688, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11368 + }, + { + "epoch": 18.161341853035143, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11369 + }, + { + "epoch": 18.162939297124602, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 11370 + }, + { + "epoch": 18.164536741214057, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11371 + }, + { + "epoch": 18.166134185303516, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11372 + }, + { + "epoch": 18.16773162939297, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11373 + }, + { + "epoch": 18.169329073482427, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11374 + }, + { + "epoch": 18.170926517571885, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11375 + }, + { + "epoch": 18.17252396166134, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11376 + }, + { + "epoch": 18.1741214057508, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11377 + }, + { + "epoch": 18.175718849840255, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11378 + }, + { + "epoch": 18.177316293929714, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11379 + }, + { + "epoch": 18.17891373801917, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11380 + }, + { + "epoch": 18.180511182108628, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11381 + }, + { + "epoch": 18.182108626198083, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11382 + }, + { + "epoch": 18.18370607028754, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11383 + }, + { + "epoch": 18.185303514376997, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11384 + }, + { + "epoch": 18.186900958466452, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11385 + }, + { + "epoch": 18.18849840255591, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11386 + }, + { + "epoch": 18.190095846645367, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 11387 + }, + { + "epoch": 18.191693290734825, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11388 + }, + { + "epoch": 18.19329073482428, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11389 + }, + { + "epoch": 18.19488817891374, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11390 + }, + { + "epoch": 18.196485623003195, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11391 + }, + { + "epoch": 18.19808306709265, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11392 + }, + { + "epoch": 18.19968051118211, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11393 + }, + { + "epoch": 18.201277955271564, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11394 + }, + { + "epoch": 18.202875399361023, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11395 + }, + { + "epoch": 18.20447284345048, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 11396 + }, + { + "epoch": 18.206070287539937, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11397 + }, + { + "epoch": 18.207667731629392, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11398 + }, + { + "epoch": 18.20926517571885, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11399 + }, + { + "epoch": 18.210862619808307, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11400 + }, + { + "epoch": 18.212460063897762, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11401 + }, + { + "epoch": 18.21405750798722, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11402 + }, + { + "epoch": 18.215654952076676, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 11403 + }, + { + "epoch": 18.217252396166135, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11404 + }, + { + "epoch": 18.21884984025559, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11405 + }, + { + "epoch": 18.22044728434505, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11406 + }, + { + "epoch": 18.222044728434504, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11407 + }, + { + "epoch": 18.223642172523963, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11408 + }, + { + "epoch": 18.22523961661342, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11409 + }, + { + "epoch": 18.226837060702877, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 11410 + }, + { + "epoch": 18.228434504792332, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11411 + }, + { + "epoch": 18.230031948881788, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11412 + }, + { + "epoch": 18.231629392971247, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11413 + }, + { + "epoch": 18.233226837060702, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11414 + }, + { + "epoch": 18.23482428115016, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11415 + }, + { + "epoch": 18.236421725239616, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11416 + }, + { + "epoch": 18.238019169329075, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11417 + }, + { + "epoch": 18.23961661341853, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11418 + }, + { + "epoch": 18.24121405750799, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11419 + }, + { + "epoch": 18.242811501597444, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11420 + }, + { + "epoch": 18.2444089456869, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11421 + }, + { + "epoch": 18.24600638977636, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11422 + }, + { + "epoch": 18.247603833865814, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11423 + }, + { + "epoch": 18.249201277955272, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11424 + }, + { + "epoch": 18.250798722044728, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11425 + }, + { + "epoch": 18.252396166134186, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 11426 + }, + { + "epoch": 18.25399361022364, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11427 + }, + { + "epoch": 18.2555910543131, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11428 + }, + { + "epoch": 18.257188498402556, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11429 + }, + { + "epoch": 18.25878594249201, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11430 + }, + { + "epoch": 18.26038338658147, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11431 + }, + { + "epoch": 18.261980830670925, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11432 + }, + { + "epoch": 18.263578274760384, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11433 + }, + { + "epoch": 18.26517571884984, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11434 + }, + { + "epoch": 18.266773162939298, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11435 + }, + { + "epoch": 18.268370607028753, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11436 + }, + { + "epoch": 18.269968051118212, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11437 + }, + { + "epoch": 18.271565495207668, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11438 + }, + { + "epoch": 18.273162939297123, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11439 + }, + { + "epoch": 18.27476038338658, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11440 + }, + { + "epoch": 18.276357827476037, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11441 + }, + { + "epoch": 18.277955271565496, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11442 + }, + { + "epoch": 18.27955271565495, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11443 + }, + { + "epoch": 18.28115015974441, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 11444 + }, + { + "epoch": 18.282747603833865, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11445 + }, + { + "epoch": 18.284345047923324, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11446 + }, + { + "epoch": 18.28594249201278, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11447 + }, + { + "epoch": 18.287539936102238, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11448 + }, + { + "epoch": 18.289137380191693, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11449 + }, + { + "epoch": 18.29073482428115, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11450 + }, + { + "epoch": 18.292332268370608, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11451 + }, + { + "epoch": 18.293929712460063, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11452 + }, + { + "epoch": 18.29552715654952, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11453 + }, + { + "epoch": 18.297124600638977, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11454 + }, + { + "epoch": 18.298722044728436, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11455 + }, + { + "epoch": 18.30031948881789, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 11456 + }, + { + "epoch": 18.30191693290735, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11457 + }, + { + "epoch": 18.303514376996805, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11458 + }, + { + "epoch": 18.30511182108626, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11459 + }, + { + "epoch": 18.30670926517572, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11460 + }, + { + "epoch": 18.308306709265175, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11461 + }, + { + "epoch": 18.309904153354633, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 11462 + }, + { + "epoch": 18.31150159744409, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11463 + }, + { + "epoch": 18.313099041533548, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11464 + }, + { + "epoch": 18.314696485623003, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11465 + }, + { + "epoch": 18.31629392971246, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 11466 + }, + { + "epoch": 18.317891373801917, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 11467 + }, + { + "epoch": 18.319488817891372, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11468 + }, + { + "epoch": 18.32108626198083, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11469 + }, + { + "epoch": 18.322683706070286, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11470 + }, + { + "epoch": 18.324281150159745, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11471 + }, + { + "epoch": 18.3258785942492, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11472 + }, + { + "epoch": 18.32747603833866, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11473 + }, + { + "epoch": 18.329073482428115, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11474 + }, + { + "epoch": 18.330670926517573, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11475 + }, + { + "epoch": 18.33226837060703, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11476 + }, + { + "epoch": 18.333865814696484, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11477 + }, + { + "epoch": 18.335463258785943, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 11478 + }, + { + "epoch": 18.337060702875398, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11479 + }, + { + "epoch": 18.338658146964857, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11480 + }, + { + "epoch": 18.340255591054312, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11481 + }, + { + "epoch": 18.34185303514377, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11482 + }, + { + "epoch": 18.343450479233226, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11483 + }, + { + "epoch": 18.345047923322685, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11484 + }, + { + "epoch": 18.34664536741214, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11485 + }, + { + "epoch": 18.3482428115016, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11486 + }, + { + "epoch": 18.349840255591054, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 11487 + }, + { + "epoch": 18.35143769968051, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11488 + }, + { + "epoch": 18.35303514376997, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11489 + }, + { + "epoch": 18.354632587859424, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11490 + }, + { + "epoch": 18.356230031948883, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11491 + }, + { + "epoch": 18.357827476038338, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11492 + }, + { + "epoch": 18.359424920127797, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11493 + }, + { + "epoch": 18.361022364217252, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11494 + }, + { + "epoch": 18.36261980830671, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11495 + }, + { + "epoch": 18.364217252396166, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11496 + }, + { + "epoch": 18.36581469648562, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11497 + }, + { + "epoch": 18.36741214057508, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11498 + }, + { + "epoch": 18.369009584664536, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 11499 + }, + { + "epoch": 18.370607028753994, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11500 + }, + { + "epoch": 18.37220447284345, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11501 + }, + { + "epoch": 18.37380191693291, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 11502 + }, + { + "epoch": 18.375399361022364, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 11503 + }, + { + "epoch": 18.376996805111823, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11504 + }, + { + "epoch": 18.378594249201278, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11505 + }, + { + "epoch": 18.380191693290733, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11506 + }, + { + "epoch": 18.381789137380192, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11507 + }, + { + "epoch": 18.383386581469647, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11508 + }, + { + "epoch": 18.384984025559106, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11509 + }, + { + "epoch": 18.38658146964856, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11510 + }, + { + "epoch": 18.38817891373802, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11511 + }, + { + "epoch": 18.389776357827476, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 11512 + }, + { + "epoch": 18.391373801916934, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11513 + }, + { + "epoch": 18.39297124600639, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11514 + }, + { + "epoch": 18.394568690095845, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11515 + }, + { + "epoch": 18.396166134185304, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11516 + }, + { + "epoch": 18.39776357827476, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11517 + }, + { + "epoch": 18.399361022364218, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11518 + }, + { + "epoch": 18.400958466453673, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11519 + }, + { + "epoch": 18.402555910543132, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 11520 + }, + { + "epoch": 18.404153354632587, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11521 + }, + { + "epoch": 18.405750798722046, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 11522 + }, + { + "epoch": 18.4073482428115, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11523 + }, + { + "epoch": 18.408945686900957, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11524 + }, + { + "epoch": 18.410543130990416, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11525 + }, + { + "epoch": 18.41214057507987, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11526 + }, + { + "epoch": 18.41373801916933, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11527 + }, + { + "epoch": 18.415335463258785, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11528 + }, + { + "epoch": 18.416932907348244, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 11529 + }, + { + "epoch": 18.4185303514377, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11530 + }, + { + "epoch": 18.420127795527158, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11531 + }, + { + "epoch": 18.421725239616613, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 11532 + }, + { + "epoch": 18.423322683706072, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11533 + }, + { + "epoch": 18.424920127795527, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11534 + }, + { + "epoch": 18.426517571884983, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 11535 + }, + { + "epoch": 18.42811501597444, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11536 + }, + { + "epoch": 18.429712460063897, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11537 + }, + { + "epoch": 18.431309904153355, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11538 + }, + { + "epoch": 18.43290734824281, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11539 + }, + { + "epoch": 18.43450479233227, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11540 + }, + { + "epoch": 18.436102236421725, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11541 + }, + { + "epoch": 18.437699680511184, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11542 + }, + { + "epoch": 18.43929712460064, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11543 + }, + { + "epoch": 18.440894568690094, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11544 + }, + { + "epoch": 18.442492012779553, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11545 + }, + { + "epoch": 18.44408945686901, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 11546 + }, + { + "epoch": 18.445686900958467, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 11547 + }, + { + "epoch": 18.447284345047922, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11548 + }, + { + "epoch": 18.44888178913738, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 11549 + }, + { + "epoch": 18.450479233226837, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11550 + }, + { + "epoch": 18.452076677316295, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11551 + }, + { + "epoch": 18.45367412140575, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11552 + }, + { + "epoch": 18.455271565495206, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11553 + }, + { + "epoch": 18.456869009584665, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11554 + }, + { + "epoch": 18.45846645367412, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11555 + }, + { + "epoch": 18.46006389776358, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11556 + }, + { + "epoch": 18.461661341853034, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 11557 + }, + { + "epoch": 18.463258785942493, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11558 + }, + { + "epoch": 18.46485623003195, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11559 + }, + { + "epoch": 18.466453674121407, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11560 + }, + { + "epoch": 18.468051118210862, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11561 + }, + { + "epoch": 18.46964856230032, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11562 + }, + { + "epoch": 18.471246006389777, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11563 + }, + { + "epoch": 18.472843450479232, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11564 + }, + { + "epoch": 18.47444089456869, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11565 + }, + { + "epoch": 18.476038338658146, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 11566 + }, + { + "epoch": 18.477635782747605, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11567 + }, + { + "epoch": 18.47923322683706, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11568 + }, + { + "epoch": 18.48083067092652, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11569 + }, + { + "epoch": 18.482428115015974, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 11570 + }, + { + "epoch": 18.484025559105433, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11571 + }, + { + "epoch": 18.48562300319489, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11572 + }, + { + "epoch": 18.487220447284344, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11573 + }, + { + "epoch": 18.488817891373802, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11574 + }, + { + "epoch": 18.490415335463258, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11575 + }, + { + "epoch": 18.492012779552716, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11576 + }, + { + "epoch": 18.49361022364217, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11577 + }, + { + "epoch": 18.49520766773163, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11578 + }, + { + "epoch": 18.496805111821086, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11579 + }, + { + "epoch": 18.498402555910545, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11580 + }, + { + "epoch": 18.5, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11581 + }, + { + "epoch": 18.501597444089455, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11582 + }, + { + "epoch": 18.503194888178914, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11583 + }, + { + "epoch": 18.50479233226837, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11584 + }, + { + "epoch": 18.50638977635783, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11585 + }, + { + "epoch": 18.507987220447284, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11586 + }, + { + "epoch": 18.509584664536742, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 11587 + }, + { + "epoch": 18.511182108626198, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11588 + }, + { + "epoch": 18.512779552715656, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11589 + }, + { + "epoch": 18.51437699680511, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11590 + }, + { + "epoch": 18.515974440894567, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11591 + }, + { + "epoch": 18.517571884984026, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11592 + }, + { + "epoch": 18.51916932907348, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11593 + }, + { + "epoch": 18.52076677316294, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11594 + }, + { + "epoch": 18.522364217252395, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 11595 + }, + { + "epoch": 18.523961661341854, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11596 + }, + { + "epoch": 18.52555910543131, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11597 + }, + { + "epoch": 18.527156549520768, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11598 + }, + { + "epoch": 18.528753993610223, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11599 + }, + { + "epoch": 18.53035143769968, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11600 + }, + { + "epoch": 18.531948881789138, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11601 + }, + { + "epoch": 18.533546325878593, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11602 + }, + { + "epoch": 18.53514376996805, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11603 + }, + { + "epoch": 18.536741214057507, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11604 + }, + { + "epoch": 18.538338658146966, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11605 + }, + { + "epoch": 18.53993610223642, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 11606 + }, + { + "epoch": 18.54153354632588, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11607 + }, + { + "epoch": 18.543130990415335, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11608 + }, + { + "epoch": 18.544728434504794, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11609 + }, + { + "epoch": 18.54632587859425, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11610 + }, + { + "epoch": 18.547923322683705, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 11611 + }, + { + "epoch": 18.549520766773163, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11612 + }, + { + "epoch": 18.55111821086262, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 11613 + }, + { + "epoch": 18.552715654952078, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11614 + }, + { + "epoch": 18.554313099041533, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11615 + }, + { + "epoch": 18.55591054313099, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 11616 + }, + { + "epoch": 18.557507987220447, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11617 + }, + { + "epoch": 18.559105431309906, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11618 + }, + { + "epoch": 18.56070287539936, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11619 + }, + { + "epoch": 18.562300319488816, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11620 + }, + { + "epoch": 18.563897763578275, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 11621 + }, + { + "epoch": 18.56549520766773, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 11622 + }, + { + "epoch": 18.56709265175719, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11623 + }, + { + "epoch": 18.568690095846645, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 11624 + }, + { + "epoch": 18.570287539936103, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11625 + }, + { + "epoch": 18.57188498402556, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11626 + }, + { + "epoch": 18.573482428115017, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11627 + }, + { + "epoch": 18.575079872204473, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11628 + }, + { + "epoch": 18.576677316293928, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11629 + }, + { + "epoch": 18.578274760383387, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11630 + }, + { + "epoch": 18.579872204472842, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11631 + }, + { + "epoch": 18.5814696485623, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11632 + }, + { + "epoch": 18.583067092651756, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11633 + }, + { + "epoch": 18.584664536741215, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 11634 + }, + { + "epoch": 18.58626198083067, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11635 + }, + { + "epoch": 18.58785942492013, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 11636 + }, + { + "epoch": 18.589456869009584, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11637 + }, + { + "epoch": 18.591054313099043, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 11638 + }, + { + "epoch": 18.5926517571885, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11639 + }, + { + "epoch": 18.594249201277954, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11640 + }, + { + "epoch": 18.595846645367413, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11641 + }, + { + "epoch": 18.597444089456868, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 11642 + }, + { + "epoch": 18.599041533546327, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11643 + }, + { + "epoch": 18.600638977635782, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 11644 + }, + { + "epoch": 18.60223642172524, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11645 + }, + { + "epoch": 18.603833865814696, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11646 + }, + { + "epoch": 18.605431309904155, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11647 + }, + { + "epoch": 18.60702875399361, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 11648 + }, + { + "epoch": 18.608626198083066, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11649 + }, + { + "epoch": 18.610223642172524, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11650 + }, + { + "epoch": 18.61182108626198, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 11651 + }, + { + "epoch": 18.61341853035144, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11652 + }, + { + "epoch": 18.615015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11653 + }, + { + "epoch": 18.616613418530353, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11654 + }, + { + "epoch": 18.618210862619808, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11655 + }, + { + "epoch": 18.619808306709267, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11656 + }, + { + "epoch": 18.621405750798722, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11657 + }, + { + "epoch": 18.623003194888177, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11658 + }, + { + "epoch": 18.624600638977636, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 11659 + }, + { + "epoch": 18.62619808306709, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11660 + }, + { + "epoch": 18.62779552715655, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11661 + }, + { + "epoch": 18.629392971246006, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11662 + }, + { + "epoch": 18.630990415335464, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11663 + }, + { + "epoch": 18.63258785942492, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 11664 + }, + { + "epoch": 18.63418530351438, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 11665 + }, + { + "epoch": 18.635782747603834, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11666 + }, + { + "epoch": 18.63738019169329, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11667 + }, + { + "epoch": 18.638977635782748, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11668 + }, + { + "epoch": 18.640575079872203, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 11669 + }, + { + "epoch": 18.642172523961662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11670 + }, + { + "epoch": 18.643769968051117, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11671 + }, + { + "epoch": 18.645367412140576, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11672 + }, + { + "epoch": 18.64696485623003, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11673 + }, + { + "epoch": 18.64856230031949, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11674 + }, + { + "epoch": 18.650159744408946, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11675 + }, + { + "epoch": 18.6517571884984, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11676 + }, + { + "epoch": 18.65335463258786, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11677 + }, + { + "epoch": 18.654952076677315, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 11678 + }, + { + "epoch": 18.656549520766774, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11679 + }, + { + "epoch": 18.65814696485623, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11680 + }, + { + "epoch": 18.659744408945688, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11681 + }, + { + "epoch": 18.661341853035143, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11682 + }, + { + "epoch": 18.662939297124602, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 11683 + }, + { + "epoch": 18.664536741214057, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11684 + }, + { + "epoch": 18.666134185303516, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11685 + }, + { + "epoch": 18.66773162939297, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11686 + }, + { + "epoch": 18.669329073482427, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11687 + }, + { + "epoch": 18.670926517571885, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11688 + }, + { + "epoch": 18.67252396166134, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11689 + }, + { + "epoch": 18.6741214057508, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11690 + }, + { + "epoch": 18.675718849840255, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11691 + }, + { + "epoch": 18.677316293929714, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11692 + }, + { + "epoch": 18.67891373801917, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11693 + }, + { + "epoch": 18.680511182108628, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11694 + }, + { + "epoch": 18.682108626198083, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11695 + }, + { + "epoch": 18.68370607028754, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11696 + }, + { + "epoch": 18.685303514376997, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11697 + }, + { + "epoch": 18.686900958466452, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11698 + }, + { + "epoch": 18.68849840255591, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11699 + }, + { + "epoch": 18.690095846645367, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11700 + }, + { + "epoch": 18.691693290734825, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 11701 + }, + { + "epoch": 18.69329073482428, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11702 + }, + { + "epoch": 18.69488817891374, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 11703 + }, + { + "epoch": 18.696485623003195, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11704 + }, + { + "epoch": 18.69808306709265, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11705 + }, + { + "epoch": 18.69968051118211, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11706 + }, + { + "epoch": 18.701277955271564, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11707 + }, + { + "epoch": 18.702875399361023, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 11708 + }, + { + "epoch": 18.70447284345048, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 11709 + }, + { + "epoch": 18.706070287539937, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 11710 + }, + { + "epoch": 18.707667731629392, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11711 + }, + { + "epoch": 18.70926517571885, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11712 + }, + { + "epoch": 18.710862619808307, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11713 + }, + { + "epoch": 18.712460063897765, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 11714 + }, + { + "epoch": 18.71405750798722, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11715 + }, + { + "epoch": 18.715654952076676, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11716 + }, + { + "epoch": 18.717252396166135, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11717 + }, + { + "epoch": 18.71884984025559, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11718 + }, + { + "epoch": 18.72044728434505, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11719 + }, + { + "epoch": 18.722044728434504, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 11720 + }, + { + "epoch": 18.723642172523963, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11721 + }, + { + "epoch": 18.72523961661342, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11722 + }, + { + "epoch": 18.726837060702877, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 11723 + }, + { + "epoch": 18.728434504792332, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11724 + }, + { + "epoch": 18.730031948881788, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11725 + }, + { + "epoch": 18.731629392971247, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11726 + }, + { + "epoch": 18.733226837060702, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11727 + }, + { + "epoch": 18.73482428115016, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11728 + }, + { + "epoch": 18.736421725239616, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 11729 + }, + { + "epoch": 18.738019169329075, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 11730 + }, + { + "epoch": 18.73961661341853, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11731 + }, + { + "epoch": 18.74121405750799, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11732 + }, + { + "epoch": 18.742811501597444, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11733 + }, + { + "epoch": 18.7444089456869, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 11734 + }, + { + "epoch": 18.74600638977636, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11735 + }, + { + "epoch": 18.747603833865814, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11736 + }, + { + "epoch": 18.749201277955272, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11737 + }, + { + "epoch": 18.750798722044728, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11738 + }, + { + "epoch": 18.752396166134186, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11739 + }, + { + "epoch": 18.75399361022364, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11740 + }, + { + "epoch": 18.7555910543131, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11741 + }, + { + "epoch": 18.757188498402556, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11742 + }, + { + "epoch": 18.75878594249201, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11743 + }, + { + "epoch": 18.76038338658147, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11744 + }, + { + "epoch": 18.761980830670925, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 11745 + }, + { + "epoch": 18.763578274760384, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11746 + }, + { + "epoch": 18.76517571884984, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11747 + }, + { + "epoch": 18.766773162939298, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11748 + }, + { + "epoch": 18.768370607028753, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11749 + }, + { + "epoch": 18.769968051118212, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11750 + }, + { + "epoch": 18.771565495207668, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11751 + }, + { + "epoch": 18.773162939297123, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11752 + }, + { + "epoch": 18.77476038338658, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11753 + }, + { + "epoch": 18.776357827476037, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11754 + }, + { + "epoch": 18.777955271565496, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11755 + }, + { + "epoch": 18.77955271565495, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 11756 + }, + { + "epoch": 18.78115015974441, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11757 + }, + { + "epoch": 18.782747603833865, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 11758 + }, + { + "epoch": 18.784345047923324, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11759 + }, + { + "epoch": 18.78594249201278, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11760 + }, + { + "epoch": 18.787539936102235, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 11761 + }, + { + "epoch": 18.789137380191693, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 11762 + }, + { + "epoch": 18.79073482428115, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11763 + }, + { + "epoch": 18.792332268370608, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11764 + }, + { + "epoch": 18.793929712460063, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11765 + }, + { + "epoch": 18.79552715654952, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11766 + }, + { + "epoch": 18.797124600638977, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11767 + }, + { + "epoch": 18.798722044728436, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11768 + }, + { + "epoch": 18.80031948881789, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11769 + }, + { + "epoch": 18.80191693290735, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 11770 + }, + { + "epoch": 18.803514376996805, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11771 + }, + { + "epoch": 18.80511182108626, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 11772 + }, + { + "epoch": 18.80670926517572, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 11773 + }, + { + "epoch": 18.808306709265175, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11774 + }, + { + "epoch": 18.809904153354633, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 11775 + }, + { + "epoch": 18.81150159744409, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 11776 + }, + { + "epoch": 18.813099041533548, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11777 + }, + { + "epoch": 18.814696485623003, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11778 + }, + { + "epoch": 18.81629392971246, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11779 + }, + { + "epoch": 18.817891373801917, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 11780 + }, + { + "epoch": 18.819488817891372, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11781 + }, + { + "epoch": 18.82108626198083, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11782 + }, + { + "epoch": 18.822683706070286, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11783 + }, + { + "epoch": 18.824281150159745, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 11784 + }, + { + "epoch": 18.8258785942492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11785 + }, + { + "epoch": 18.82747603833866, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11786 + }, + { + "epoch": 18.829073482428115, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11787 + }, + { + "epoch": 18.830670926517573, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11788 + }, + { + "epoch": 18.83226837060703, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11789 + }, + { + "epoch": 18.833865814696484, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11790 + }, + { + "epoch": 18.835463258785943, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11791 + }, + { + "epoch": 18.837060702875398, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11792 + }, + { + "epoch": 18.838658146964857, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11793 + }, + { + "epoch": 18.840255591054312, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11794 + }, + { + "epoch": 18.84185303514377, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11795 + }, + { + "epoch": 18.843450479233226, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 11796 + }, + { + "epoch": 18.845047923322685, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11797 + }, + { + "epoch": 18.84664536741214, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 11798 + }, + { + "epoch": 18.8482428115016, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 11799 + }, + { + "epoch": 18.849840255591054, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 11800 + }, + { + "epoch": 18.85143769968051, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11801 + }, + { + "epoch": 18.85303514376997, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 11802 + }, + { + "epoch": 18.854632587859424, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 11803 + }, + { + "epoch": 18.856230031948883, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11804 + }, + { + "epoch": 18.857827476038338, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11805 + }, + { + "epoch": 18.859424920127797, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11806 + }, + { + "epoch": 18.861022364217252, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11807 + }, + { + "epoch": 18.86261980830671, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11808 + }, + { + "epoch": 18.864217252396166, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 11809 + }, + { + "epoch": 18.86581469648562, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11810 + }, + { + "epoch": 18.86741214057508, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11811 + }, + { + "epoch": 18.869009584664536, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11812 + }, + { + "epoch": 18.870607028753994, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11813 + }, + { + "epoch": 18.87220447284345, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11814 + }, + { + "epoch": 18.87380191693291, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11815 + }, + { + "epoch": 18.875399361022364, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11816 + }, + { + "epoch": 18.876996805111823, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11817 + }, + { + "epoch": 18.878594249201278, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11818 + }, + { + "epoch": 18.880191693290733, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 11819 + }, + { + "epoch": 18.881789137380192, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11820 + }, + { + "epoch": 18.883386581469647, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11821 + }, + { + "epoch": 18.884984025559106, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11822 + }, + { + "epoch": 18.88658146964856, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 11823 + }, + { + "epoch": 18.88817891373802, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11824 + }, + { + "epoch": 18.889776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11825 + }, + { + "epoch": 18.891373801916934, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 11826 + }, + { + "epoch": 18.89297124600639, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11827 + }, + { + "epoch": 18.894568690095845, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11828 + }, + { + "epoch": 18.896166134185304, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 11829 + }, + { + "epoch": 18.89776357827476, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 11830 + }, + { + "epoch": 18.899361022364218, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11831 + }, + { + "epoch": 18.900958466453673, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 11832 + }, + { + "epoch": 18.902555910543132, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 11833 + }, + { + "epoch": 18.904153354632587, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 11834 + }, + { + "epoch": 18.905750798722046, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11835 + }, + { + "epoch": 18.9073482428115, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11836 + }, + { + "epoch": 18.908945686900957, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 11837 + }, + { + "epoch": 18.910543130990416, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11838 + }, + { + "epoch": 18.91214057507987, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 11839 + }, + { + "epoch": 18.91373801916933, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11840 + }, + { + "epoch": 18.915335463258785, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11841 + }, + { + "epoch": 18.916932907348244, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11842 + }, + { + "epoch": 18.9185303514377, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 11843 + }, + { + "epoch": 18.920127795527158, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11844 + }, + { + "epoch": 18.921725239616613, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11845 + }, + { + "epoch": 18.923322683706072, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 11846 + }, + { + "epoch": 18.924920127795527, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11847 + }, + { + "epoch": 18.926517571884983, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11848 + }, + { + "epoch": 18.92811501597444, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11849 + }, + { + "epoch": 18.929712460063897, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11850 + }, + { + "epoch": 18.931309904153355, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11851 + }, + { + "epoch": 18.93290734824281, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 11852 + }, + { + "epoch": 18.93450479233227, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 11853 + }, + { + "epoch": 18.936102236421725, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11854 + }, + { + "epoch": 18.937699680511184, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11855 + }, + { + "epoch": 18.93929712460064, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11856 + }, + { + "epoch": 18.940894568690094, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11857 + }, + { + "epoch": 18.942492012779553, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 11858 + }, + { + "epoch": 18.94408945686901, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 11859 + }, + { + "epoch": 18.945686900958467, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11860 + }, + { + "epoch": 18.947284345047922, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11861 + }, + { + "epoch": 18.94888178913738, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11862 + }, + { + "epoch": 18.950479233226837, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 11863 + }, + { + "epoch": 18.952076677316295, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11864 + }, + { + "epoch": 18.95367412140575, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11865 + }, + { + "epoch": 18.955271565495206, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11866 + }, + { + "epoch": 18.956869009584665, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 11867 + }, + { + "epoch": 18.95846645367412, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11868 + }, + { + "epoch": 18.96006389776358, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 11869 + }, + { + "epoch": 18.961661341853034, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11870 + }, + { + "epoch": 18.963258785942493, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 11871 + }, + { + "epoch": 18.96485623003195, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 11872 + }, + { + "epoch": 18.966453674121407, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11873 + }, + { + "epoch": 18.968051118210862, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11874 + }, + { + "epoch": 18.96964856230032, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11875 + }, + { + "epoch": 18.971246006389777, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 11876 + }, + { + "epoch": 18.972843450479232, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11877 + }, + { + "epoch": 18.97444089456869, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 11878 + }, + { + "epoch": 18.976038338658146, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11879 + }, + { + "epoch": 18.977635782747605, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 11880 + }, + { + "epoch": 18.97923322683706, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 11881 + }, + { + "epoch": 18.98083067092652, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 11882 + }, + { + "epoch": 18.982428115015974, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11883 + }, + { + "epoch": 18.984025559105433, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11884 + }, + { + "epoch": 18.98562300319489, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 11885 + }, + { + "epoch": 18.987220447284344, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 11886 + }, + { + "epoch": 18.988817891373802, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11887 + }, + { + "epoch": 18.990415335463258, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11888 + }, + { + "epoch": 18.992012779552716, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11889 + }, + { + "epoch": 18.99361022364217, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11890 + }, + { + "epoch": 18.99520766773163, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11891 + }, + { + "epoch": 18.996805111821086, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 11892 + }, + { + "epoch": 18.998402555910545, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11893 + }, + { + "epoch": 19.0, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11894 + }, + { + "epoch": 19.001597444089455, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 11895 + }, + { + "epoch": 19.003194888178914, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11896 + }, + { + "epoch": 19.00479233226837, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11897 + }, + { + "epoch": 19.00638977635783, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11898 + }, + { + "epoch": 19.007987220447284, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 11899 + }, + { + "epoch": 19.009584664536742, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 11900 + }, + { + "epoch": 19.011182108626198, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 11901 + }, + { + "epoch": 19.012779552715656, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 11902 + }, + { + "epoch": 19.01437699680511, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 11903 + }, + { + "epoch": 19.015974440894567, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11904 + }, + { + "epoch": 19.017571884984026, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 11905 + }, + { + "epoch": 19.01916932907348, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11906 + }, + { + "epoch": 19.02076677316294, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 11907 + }, + { + "epoch": 19.022364217252395, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 11908 + }, + { + "epoch": 19.023961661341854, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11909 + }, + { + "epoch": 19.02555910543131, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11910 + }, + { + "epoch": 19.027156549520768, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 11911 + }, + { + "epoch": 19.028753993610223, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 11912 + }, + { + "epoch": 19.03035143769968, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11913 + }, + { + "epoch": 19.031948881789138, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11914 + }, + { + "epoch": 19.033546325878593, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 11915 + }, + { + "epoch": 19.03514376996805, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 11916 + }, + { + "epoch": 19.036741214057507, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11917 + }, + { + "epoch": 19.038338658146966, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 11918 + }, + { + "epoch": 19.03993610223642, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11919 + }, + { + "epoch": 19.04153354632588, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 11920 + }, + { + "epoch": 19.043130990415335, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 11921 + }, + { + "epoch": 19.044728434504794, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 11922 + }, + { + "epoch": 19.04632587859425, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11923 + }, + { + "epoch": 19.047923322683705, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 11924 + }, + { + "epoch": 19.049520766773163, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11925 + }, + { + "epoch": 19.05111821086262, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 11926 + }, + { + "epoch": 19.052715654952078, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 11927 + }, + { + "epoch": 19.054313099041533, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 11928 + }, + { + "epoch": 19.05591054313099, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 11929 + }, + { + "epoch": 19.057507987220447, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 11930 + }, + { + "epoch": 19.059105431309906, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 11931 + }, + { + "epoch": 19.06070287539936, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11932 + }, + { + "epoch": 19.062300319488816, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 11933 + }, + { + "epoch": 19.063897763578275, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 11934 + }, + { + "epoch": 19.06549520766773, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 11935 + }, + { + "epoch": 19.06709265175719, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11936 + }, + { + "epoch": 19.068690095846645, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11937 + }, + { + "epoch": 19.070287539936103, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 11938 + }, + { + "epoch": 19.07188498402556, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 11939 + }, + { + "epoch": 19.073482428115017, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 11940 + }, + { + "epoch": 19.075079872204473, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 11941 + }, + { + "epoch": 19.076677316293928, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 11942 + }, + { + "epoch": 19.078274760383387, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11943 + }, + { + "epoch": 19.079872204472842, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11944 + }, + { + "epoch": 19.0814696485623, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 11945 + }, + { + "epoch": 19.083067092651756, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 11946 + }, + { + "epoch": 19.084664536741215, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11947 + }, + { + "epoch": 19.08626198083067, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 11948 + }, + { + "epoch": 19.08785942492013, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11949 + }, + { + "epoch": 19.089456869009584, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 11950 + }, + { + "epoch": 19.09105431309904, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 11951 + }, + { + "epoch": 19.0926517571885, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11952 + }, + { + "epoch": 19.094249201277954, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11953 + }, + { + "epoch": 19.095846645367413, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 11954 + }, + { + "epoch": 19.097444089456868, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 11955 + }, + { + "epoch": 19.099041533546327, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 11956 + }, + { + "epoch": 19.100638977635782, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11957 + }, + { + "epoch": 19.10223642172524, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 11958 + }, + { + "epoch": 19.103833865814696, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11959 + }, + { + "epoch": 19.105431309904155, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 11960 + }, + { + "epoch": 19.10702875399361, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 11961 + }, + { + "epoch": 19.108626198083066, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 11962 + }, + { + "epoch": 19.110223642172524, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 11963 + }, + { + "epoch": 19.11182108626198, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 11964 + }, + { + "epoch": 19.11341853035144, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 11965 + }, + { + "epoch": 19.115015974440894, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 11966 + }, + { + "epoch": 19.116613418530353, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 11967 + }, + { + "epoch": 19.118210862619808, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 11968 + }, + { + "epoch": 19.119808306709267, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 11969 + }, + { + "epoch": 19.121405750798722, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 11970 + }, + { + "epoch": 19.123003194888177, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 11971 + }, + { + "epoch": 19.124600638977636, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 11972 + }, + { + "epoch": 19.12619808306709, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 11973 + }, + { + "epoch": 19.12779552715655, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 11974 + }, + { + "epoch": 19.129392971246006, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 11975 + }, + { + "epoch": 19.130990415335464, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 11976 + }, + { + "epoch": 19.13258785942492, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 11977 + }, + { + "epoch": 19.13418530351438, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 11978 + }, + { + "epoch": 19.135782747603834, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 11979 + }, + { + "epoch": 19.13738019169329, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11980 + }, + { + "epoch": 19.138977635782748, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 11981 + }, + { + "epoch": 19.140575079872203, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 11982 + }, + { + "epoch": 19.142172523961662, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 11983 + }, + { + "epoch": 19.143769968051117, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11984 + }, + { + "epoch": 19.145367412140576, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 11985 + }, + { + "epoch": 19.14696485623003, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 11986 + }, + { + "epoch": 19.14856230031949, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 11987 + }, + { + "epoch": 19.150159744408946, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 11988 + }, + { + "epoch": 19.1517571884984, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 11989 + }, + { + "epoch": 19.15335463258786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 11990 + }, + { + "epoch": 19.154952076677315, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 11991 + }, + { + "epoch": 19.156549520766774, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 11992 + }, + { + "epoch": 19.15814696485623, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 11993 + }, + { + "epoch": 19.159744408945688, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 11994 + }, + { + "epoch": 19.161341853035143, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11995 + }, + { + "epoch": 19.162939297124602, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 11996 + }, + { + "epoch": 19.164536741214057, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 11997 + }, + { + "epoch": 19.166134185303516, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 11998 + }, + { + "epoch": 19.16773162939297, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 11999 + }, + { + "epoch": 19.169329073482427, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12000 + }, + { + "epoch": 19.170926517571885, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12001 + }, + { + "epoch": 19.17252396166134, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12002 + }, + { + "epoch": 19.1741214057508, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12003 + }, + { + "epoch": 19.175718849840255, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12004 + }, + { + "epoch": 19.177316293929714, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12005 + }, + { + "epoch": 19.17891373801917, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12006 + }, + { + "epoch": 19.180511182108628, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 12007 + }, + { + "epoch": 19.182108626198083, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12008 + }, + { + "epoch": 19.18370607028754, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12009 + }, + { + "epoch": 19.185303514376997, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12010 + }, + { + "epoch": 19.186900958466452, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 12011 + }, + { + "epoch": 19.18849840255591, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 12012 + }, + { + "epoch": 19.190095846645367, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12013 + }, + { + "epoch": 19.191693290734825, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 12014 + }, + { + "epoch": 19.19329073482428, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12015 + }, + { + "epoch": 19.19488817891374, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 12016 + }, + { + "epoch": 19.196485623003195, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12017 + }, + { + "epoch": 19.19808306709265, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12018 + }, + { + "epoch": 19.19968051118211, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12019 + }, + { + "epoch": 19.201277955271564, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12020 + }, + { + "epoch": 19.202875399361023, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12021 + }, + { + "epoch": 19.20447284345048, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12022 + }, + { + "epoch": 19.206070287539937, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12023 + }, + { + "epoch": 19.207667731629392, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 12024 + }, + { + "epoch": 19.20926517571885, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12025 + }, + { + "epoch": 19.210862619808307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12026 + }, + { + "epoch": 19.212460063897762, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12027 + }, + { + "epoch": 19.21405750798722, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12028 + }, + { + "epoch": 19.215654952076676, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12029 + }, + { + "epoch": 19.217252396166135, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12030 + }, + { + "epoch": 19.21884984025559, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12031 + }, + { + "epoch": 19.22044728434505, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12032 + }, + { + "epoch": 19.222044728434504, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12033 + }, + { + "epoch": 19.223642172523963, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12034 + }, + { + "epoch": 19.22523961661342, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12035 + }, + { + "epoch": 19.226837060702877, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12036 + }, + { + "epoch": 19.228434504792332, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12037 + }, + { + "epoch": 19.230031948881788, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 12038 + }, + { + "epoch": 19.231629392971247, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12039 + }, + { + "epoch": 19.233226837060702, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12040 + }, + { + "epoch": 19.23482428115016, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12041 + }, + { + "epoch": 19.236421725239616, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12042 + }, + { + "epoch": 19.238019169329075, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12043 + }, + { + "epoch": 19.23961661341853, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12044 + }, + { + "epoch": 19.24121405750799, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12045 + }, + { + "epoch": 19.242811501597444, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12046 + }, + { + "epoch": 19.2444089456869, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12047 + }, + { + "epoch": 19.24600638977636, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12048 + }, + { + "epoch": 19.247603833865814, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 12049 + }, + { + "epoch": 19.249201277955272, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12050 + }, + { + "epoch": 19.250798722044728, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12051 + }, + { + "epoch": 19.252396166134186, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12052 + }, + { + "epoch": 19.25399361022364, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12053 + }, + { + "epoch": 19.2555910543131, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12054 + }, + { + "epoch": 19.257188498402556, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12055 + }, + { + "epoch": 19.25878594249201, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12056 + }, + { + "epoch": 19.26038338658147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12057 + }, + { + "epoch": 19.261980830670925, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12058 + }, + { + "epoch": 19.263578274760384, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12059 + }, + { + "epoch": 19.26517571884984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12060 + }, + { + "epoch": 19.266773162939298, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 12061 + }, + { + "epoch": 19.268370607028753, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12062 + }, + { + "epoch": 19.269968051118212, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12063 + }, + { + "epoch": 19.271565495207668, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12064 + }, + { + "epoch": 19.273162939297123, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12065 + }, + { + "epoch": 19.27476038338658, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12066 + }, + { + "epoch": 19.276357827476037, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12067 + }, + { + "epoch": 19.277955271565496, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12068 + }, + { + "epoch": 19.27955271565495, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12069 + }, + { + "epoch": 19.28115015974441, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12070 + }, + { + "epoch": 19.282747603833865, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12071 + }, + { + "epoch": 19.284345047923324, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12072 + }, + { + "epoch": 19.28594249201278, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12073 + }, + { + "epoch": 19.287539936102238, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12074 + }, + { + "epoch": 19.289137380191693, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12075 + }, + { + "epoch": 19.29073482428115, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12076 + }, + { + "epoch": 19.292332268370608, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12077 + }, + { + "epoch": 19.293929712460063, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12078 + }, + { + "epoch": 19.29552715654952, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12079 + }, + { + "epoch": 19.297124600638977, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 12080 + }, + { + "epoch": 19.298722044728436, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12081 + }, + { + "epoch": 19.30031948881789, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12082 + }, + { + "epoch": 19.30191693290735, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 12083 + }, + { + "epoch": 19.303514376996805, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12084 + }, + { + "epoch": 19.30511182108626, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12085 + }, + { + "epoch": 19.30670926517572, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12086 + }, + { + "epoch": 19.308306709265175, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 12087 + }, + { + "epoch": 19.309904153354633, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12088 + }, + { + "epoch": 19.31150159744409, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12089 + }, + { + "epoch": 19.313099041533548, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12090 + }, + { + "epoch": 19.314696485623003, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12091 + }, + { + "epoch": 19.31629392971246, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12092 + }, + { + "epoch": 19.317891373801917, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12093 + }, + { + "epoch": 19.319488817891372, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12094 + }, + { + "epoch": 19.32108626198083, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12095 + }, + { + "epoch": 19.322683706070286, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12096 + }, + { + "epoch": 19.324281150159745, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12097 + }, + { + "epoch": 19.3258785942492, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12098 + }, + { + "epoch": 19.32747603833866, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 12099 + }, + { + "epoch": 19.329073482428115, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12100 + }, + { + "epoch": 19.330670926517573, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 12101 + }, + { + "epoch": 19.33226837060703, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12102 + }, + { + "epoch": 19.333865814696484, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12103 + }, + { + "epoch": 19.335463258785943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12104 + }, + { + "epoch": 19.337060702875398, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12105 + }, + { + "epoch": 19.338658146964857, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12106 + }, + { + "epoch": 19.340255591054312, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12107 + }, + { + "epoch": 19.34185303514377, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12108 + }, + { + "epoch": 19.343450479233226, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12109 + }, + { + "epoch": 19.345047923322685, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12110 + }, + { + "epoch": 19.34664536741214, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12111 + }, + { + "epoch": 19.3482428115016, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12112 + }, + { + "epoch": 19.349840255591054, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12113 + }, + { + "epoch": 19.35143769968051, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12114 + }, + { + "epoch": 19.35303514376997, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12115 + }, + { + "epoch": 19.354632587859424, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12116 + }, + { + "epoch": 19.356230031948883, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12117 + }, + { + "epoch": 19.357827476038338, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12118 + }, + { + "epoch": 19.359424920127797, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12119 + }, + { + "epoch": 19.361022364217252, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12120 + }, + { + "epoch": 19.36261980830671, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12121 + }, + { + "epoch": 19.364217252396166, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12122 + }, + { + "epoch": 19.36581469648562, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12123 + }, + { + "epoch": 19.36741214057508, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12124 + }, + { + "epoch": 19.369009584664536, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 12125 + }, + { + "epoch": 19.370607028753994, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 12126 + }, + { + "epoch": 19.37220447284345, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12127 + }, + { + "epoch": 19.37380191693291, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 12128 + }, + { + "epoch": 19.375399361022364, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12129 + }, + { + "epoch": 19.376996805111823, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12130 + }, + { + "epoch": 19.378594249201278, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12131 + }, + { + "epoch": 19.380191693290733, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12132 + }, + { + "epoch": 19.381789137380192, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12133 + }, + { + "epoch": 19.383386581469647, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 12134 + }, + { + "epoch": 19.384984025559106, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12135 + }, + { + "epoch": 19.38658146964856, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12136 + }, + { + "epoch": 19.38817891373802, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12137 + }, + { + "epoch": 19.389776357827476, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12138 + }, + { + "epoch": 19.391373801916934, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12139 + }, + { + "epoch": 19.39297124600639, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12140 + }, + { + "epoch": 19.394568690095845, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 12141 + }, + { + "epoch": 19.396166134185304, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 12142 + }, + { + "epoch": 19.39776357827476, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 12143 + }, + { + "epoch": 19.399361022364218, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12144 + }, + { + "epoch": 19.400958466453673, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12145 + }, + { + "epoch": 19.402555910543132, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12146 + }, + { + "epoch": 19.404153354632587, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12147 + }, + { + "epoch": 19.405750798722046, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12148 + }, + { + "epoch": 19.4073482428115, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12149 + }, + { + "epoch": 19.408945686900957, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 12150 + }, + { + "epoch": 19.410543130990416, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12151 + }, + { + "epoch": 19.41214057507987, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12152 + }, + { + "epoch": 19.41373801916933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12153 + }, + { + "epoch": 19.415335463258785, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12154 + }, + { + "epoch": 19.416932907348244, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12155 + }, + { + "epoch": 19.4185303514377, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12156 + }, + { + "epoch": 19.420127795527158, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12157 + }, + { + "epoch": 19.421725239616613, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12158 + }, + { + "epoch": 19.423322683706072, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12159 + }, + { + "epoch": 19.424920127795527, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12160 + }, + { + "epoch": 19.426517571884983, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12161 + }, + { + "epoch": 19.42811501597444, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 12162 + }, + { + "epoch": 19.429712460063897, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12163 + }, + { + "epoch": 19.431309904153355, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 12164 + }, + { + "epoch": 19.43290734824281, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12165 + }, + { + "epoch": 19.43450479233227, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12166 + }, + { + "epoch": 19.436102236421725, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 12167 + }, + { + "epoch": 19.437699680511184, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12168 + }, + { + "epoch": 19.43929712460064, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12169 + }, + { + "epoch": 19.440894568690094, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12170 + }, + { + "epoch": 19.442492012779553, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12171 + }, + { + "epoch": 19.44408945686901, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12172 + }, + { + "epoch": 19.445686900958467, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12173 + }, + { + "epoch": 19.447284345047922, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12174 + }, + { + "epoch": 19.44888178913738, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 12175 + }, + { + "epoch": 19.450479233226837, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12176 + }, + { + "epoch": 19.452076677316295, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12177 + }, + { + "epoch": 19.45367412140575, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12178 + }, + { + "epoch": 19.455271565495206, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12179 + }, + { + "epoch": 19.456869009584665, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12180 + }, + { + "epoch": 19.45846645367412, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12181 + }, + { + "epoch": 19.46006389776358, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12182 + }, + { + "epoch": 19.461661341853034, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12183 + }, + { + "epoch": 19.463258785942493, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 12184 + }, + { + "epoch": 19.46485623003195, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12185 + }, + { + "epoch": 19.466453674121407, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12186 + }, + { + "epoch": 19.468051118210862, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12187 + }, + { + "epoch": 19.46964856230032, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12188 + }, + { + "epoch": 19.471246006389777, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12189 + }, + { + "epoch": 19.472843450479232, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12190 + }, + { + "epoch": 19.47444089456869, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12191 + }, + { + "epoch": 19.476038338658146, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12192 + }, + { + "epoch": 19.477635782747605, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12193 + }, + { + "epoch": 19.47923322683706, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12194 + }, + { + "epoch": 19.48083067092652, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12195 + }, + { + "epoch": 19.482428115015974, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12196 + }, + { + "epoch": 19.484025559105433, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12197 + }, + { + "epoch": 19.48562300319489, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12198 + }, + { + "epoch": 19.487220447284344, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12199 + }, + { + "epoch": 19.488817891373802, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12200 + }, + { + "epoch": 19.490415335463258, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12201 + }, + { + "epoch": 19.492012779552716, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 12202 + }, + { + "epoch": 19.49361022364217, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12203 + }, + { + "epoch": 19.49520766773163, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12204 + }, + { + "epoch": 19.496805111821086, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12205 + }, + { + "epoch": 19.498402555910545, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12206 + }, + { + "epoch": 19.5, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 12207 + }, + { + "epoch": 19.501597444089455, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12208 + }, + { + "epoch": 19.503194888178914, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12209 + }, + { + "epoch": 19.50479233226837, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12210 + }, + { + "epoch": 19.50638977635783, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12211 + }, + { + "epoch": 19.507987220447284, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 12212 + }, + { + "epoch": 19.509584664536742, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 12213 + }, + { + "epoch": 19.511182108626198, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12214 + }, + { + "epoch": 19.512779552715656, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12215 + }, + { + "epoch": 19.51437699680511, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12216 + }, + { + "epoch": 19.515974440894567, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12217 + }, + { + "epoch": 19.517571884984026, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12218 + }, + { + "epoch": 19.51916932907348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12219 + }, + { + "epoch": 19.52076677316294, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12220 + }, + { + "epoch": 19.522364217252395, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12221 + }, + { + "epoch": 19.523961661341854, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12222 + }, + { + "epoch": 19.52555910543131, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 12223 + }, + { + "epoch": 19.527156549520768, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12224 + }, + { + "epoch": 19.528753993610223, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 12225 + }, + { + "epoch": 19.53035143769968, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12226 + }, + { + "epoch": 19.531948881789138, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12227 + }, + { + "epoch": 19.533546325878593, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12228 + }, + { + "epoch": 19.53514376996805, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12229 + }, + { + "epoch": 19.536741214057507, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12230 + }, + { + "epoch": 19.538338658146966, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12231 + }, + { + "epoch": 19.53993610223642, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 12232 + }, + { + "epoch": 19.54153354632588, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12233 + }, + { + "epoch": 19.543130990415335, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12234 + }, + { + "epoch": 19.544728434504794, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12235 + }, + { + "epoch": 19.54632587859425, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12236 + }, + { + "epoch": 19.547923322683705, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12237 + }, + { + "epoch": 19.549520766773163, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12238 + }, + { + "epoch": 19.55111821086262, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12239 + }, + { + "epoch": 19.552715654952078, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 12240 + }, + { + "epoch": 19.554313099041533, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12241 + }, + { + "epoch": 19.55591054313099, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12242 + }, + { + "epoch": 19.557507987220447, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12243 + }, + { + "epoch": 19.559105431309906, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12244 + }, + { + "epoch": 19.56070287539936, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 12245 + }, + { + "epoch": 19.562300319488816, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12246 + }, + { + "epoch": 19.563897763578275, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12247 + }, + { + "epoch": 19.56549520766773, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12248 + }, + { + "epoch": 19.56709265175719, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12249 + }, + { + "epoch": 19.568690095846645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 12250 + }, + { + "epoch": 19.570287539936103, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12251 + }, + { + "epoch": 19.57188498402556, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12252 + }, + { + "epoch": 19.573482428115017, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 12253 + }, + { + "epoch": 19.575079872204473, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12254 + }, + { + "epoch": 19.576677316293928, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12255 + }, + { + "epoch": 19.578274760383387, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12256 + }, + { + "epoch": 19.579872204472842, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12257 + }, + { + "epoch": 19.5814696485623, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12258 + }, + { + "epoch": 19.583067092651756, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12259 + }, + { + "epoch": 19.584664536741215, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12260 + }, + { + "epoch": 19.58626198083067, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12261 + }, + { + "epoch": 19.58785942492013, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12262 + }, + { + "epoch": 19.589456869009584, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12263 + }, + { + "epoch": 19.591054313099043, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12264 + }, + { + "epoch": 19.5926517571885, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12265 + }, + { + "epoch": 19.594249201277954, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12266 + }, + { + "epoch": 19.595846645367413, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12267 + }, + { + "epoch": 19.597444089456868, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12268 + }, + { + "epoch": 19.599041533546327, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12269 + }, + { + "epoch": 19.600638977635782, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12270 + }, + { + "epoch": 19.60223642172524, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12271 + }, + { + "epoch": 19.603833865814696, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12272 + }, + { + "epoch": 19.605431309904155, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12273 + }, + { + "epoch": 19.60702875399361, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 12274 + }, + { + "epoch": 19.608626198083066, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12275 + }, + { + "epoch": 19.610223642172524, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12276 + }, + { + "epoch": 19.61182108626198, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12277 + }, + { + "epoch": 19.61341853035144, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12278 + }, + { + "epoch": 19.615015974440894, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12279 + }, + { + "epoch": 19.616613418530353, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12280 + }, + { + "epoch": 19.618210862619808, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 12281 + }, + { + "epoch": 19.619808306709267, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12282 + }, + { + "epoch": 19.621405750798722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12283 + }, + { + "epoch": 19.623003194888177, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12284 + }, + { + "epoch": 19.624600638977636, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12285 + }, + { + "epoch": 19.62619808306709, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12286 + }, + { + "epoch": 19.62779552715655, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12287 + }, + { + "epoch": 19.629392971246006, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 12288 + }, + { + "epoch": 19.630990415335464, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 12289 + }, + { + "epoch": 19.63258785942492, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12290 + }, + { + "epoch": 19.63418530351438, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12291 + }, + { + "epoch": 19.635782747603834, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12292 + }, + { + "epoch": 19.63738019169329, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12293 + }, + { + "epoch": 19.638977635782748, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12294 + }, + { + "epoch": 19.640575079872203, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12295 + }, + { + "epoch": 19.642172523961662, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 12296 + }, + { + "epoch": 19.643769968051117, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12297 + }, + { + "epoch": 19.645367412140576, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12298 + }, + { + "epoch": 19.64696485623003, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12299 + }, + { + "epoch": 19.64856230031949, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12300 + }, + { + "epoch": 19.650159744408946, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12301 + }, + { + "epoch": 19.6517571884984, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12302 + }, + { + "epoch": 19.65335463258786, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12303 + }, + { + "epoch": 19.654952076677315, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12304 + }, + { + "epoch": 19.656549520766774, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12305 + }, + { + "epoch": 19.65814696485623, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12306 + }, + { + "epoch": 19.659744408945688, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12307 + }, + { + "epoch": 19.661341853035143, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12308 + }, + { + "epoch": 19.662939297124602, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12309 + }, + { + "epoch": 19.664536741214057, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12310 + }, + { + "epoch": 19.666134185303516, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12311 + }, + { + "epoch": 19.66773162939297, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12312 + }, + { + "epoch": 19.669329073482427, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12313 + }, + { + "epoch": 19.670926517571885, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12314 + }, + { + "epoch": 19.67252396166134, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12315 + }, + { + "epoch": 19.6741214057508, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12316 + }, + { + "epoch": 19.675718849840255, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12317 + }, + { + "epoch": 19.677316293929714, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12318 + }, + { + "epoch": 19.67891373801917, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12319 + }, + { + "epoch": 19.680511182108628, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 12320 + }, + { + "epoch": 19.682108626198083, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12321 + }, + { + "epoch": 19.68370607028754, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12322 + }, + { + "epoch": 19.685303514376997, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12323 + }, + { + "epoch": 19.686900958466452, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 12324 + }, + { + "epoch": 19.68849840255591, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12325 + }, + { + "epoch": 19.690095846645367, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12326 + }, + { + "epoch": 19.691693290734825, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12327 + }, + { + "epoch": 19.69329073482428, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12328 + }, + { + "epoch": 19.69488817891374, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12329 + }, + { + "epoch": 19.696485623003195, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12330 + }, + { + "epoch": 19.69808306709265, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 12331 + }, + { + "epoch": 19.69968051118211, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12332 + }, + { + "epoch": 19.701277955271564, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12333 + }, + { + "epoch": 19.702875399361023, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12334 + }, + { + "epoch": 19.70447284345048, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12335 + }, + { + "epoch": 19.706070287539937, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12336 + }, + { + "epoch": 19.707667731629392, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 12337 + }, + { + "epoch": 19.70926517571885, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 12338 + }, + { + "epoch": 19.710862619808307, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12339 + }, + { + "epoch": 19.712460063897765, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12340 + }, + { + "epoch": 19.71405750798722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 12341 + }, + { + "epoch": 19.715654952076676, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12342 + }, + { + "epoch": 19.717252396166135, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12343 + }, + { + "epoch": 19.71884984025559, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12344 + }, + { + "epoch": 19.72044728434505, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 12345 + }, + { + "epoch": 19.722044728434504, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12346 + }, + { + "epoch": 19.723642172523963, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12347 + }, + { + "epoch": 19.72523961661342, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12348 + }, + { + "epoch": 19.726837060702877, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12349 + }, + { + "epoch": 19.728434504792332, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12350 + }, + { + "epoch": 19.730031948881788, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12351 + }, + { + "epoch": 19.731629392971247, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12352 + }, + { + "epoch": 19.733226837060702, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12353 + }, + { + "epoch": 19.73482428115016, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 12354 + }, + { + "epoch": 19.736421725239616, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12355 + }, + { + "epoch": 19.738019169329075, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12356 + }, + { + "epoch": 19.73961661341853, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12357 + }, + { + "epoch": 19.74121405750799, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12358 + }, + { + "epoch": 19.742811501597444, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12359 + }, + { + "epoch": 19.7444089456869, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12360 + }, + { + "epoch": 19.74600638977636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12361 + }, + { + "epoch": 19.747603833865814, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12362 + }, + { + "epoch": 19.749201277955272, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12363 + }, + { + "epoch": 19.750798722044728, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12364 + }, + { + "epoch": 19.752396166134186, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 12365 + }, + { + "epoch": 19.75399361022364, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12366 + }, + { + "epoch": 19.7555910543131, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12367 + }, + { + "epoch": 19.757188498402556, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12368 + }, + { + "epoch": 19.75878594249201, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12369 + }, + { + "epoch": 19.76038338658147, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 12370 + }, + { + "epoch": 19.761980830670925, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12371 + }, + { + "epoch": 19.763578274760384, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12372 + }, + { + "epoch": 19.76517571884984, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12373 + }, + { + "epoch": 19.766773162939298, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12374 + }, + { + "epoch": 19.768370607028753, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12375 + }, + { + "epoch": 19.769968051118212, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 12376 + }, + { + "epoch": 19.771565495207668, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12377 + }, + { + "epoch": 19.773162939297123, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12378 + }, + { + "epoch": 19.77476038338658, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12379 + }, + { + "epoch": 19.776357827476037, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12380 + }, + { + "epoch": 19.777955271565496, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12381 + }, + { + "epoch": 19.77955271565495, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12382 + }, + { + "epoch": 19.78115015974441, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12383 + }, + { + "epoch": 19.782747603833865, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12384 + }, + { + "epoch": 19.784345047923324, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12385 + }, + { + "epoch": 19.78594249201278, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12386 + }, + { + "epoch": 19.787539936102235, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12387 + }, + { + "epoch": 19.789137380191693, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12388 + }, + { + "epoch": 19.79073482428115, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12389 + }, + { + "epoch": 19.792332268370608, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12390 + }, + { + "epoch": 19.793929712460063, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12391 + }, + { + "epoch": 19.79552715654952, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12392 + }, + { + "epoch": 19.797124600638977, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 12393 + }, + { + "epoch": 19.798722044728436, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12394 + }, + { + "epoch": 19.80031948881789, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12395 + }, + { + "epoch": 19.80191693290735, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12396 + }, + { + "epoch": 19.803514376996805, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12397 + }, + { + "epoch": 19.80511182108626, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12398 + }, + { + "epoch": 19.80670926517572, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12399 + }, + { + "epoch": 19.808306709265175, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12400 + }, + { + "epoch": 19.809904153354633, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12401 + }, + { + "epoch": 19.81150159744409, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12402 + }, + { + "epoch": 19.813099041533548, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12403 + }, + { + "epoch": 19.814696485623003, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12404 + }, + { + "epoch": 19.81629392971246, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12405 + }, + { + "epoch": 19.817891373801917, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12406 + }, + { + "epoch": 19.819488817891372, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 12407 + }, + { + "epoch": 19.82108626198083, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12408 + }, + { + "epoch": 19.822683706070286, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12409 + }, + { + "epoch": 19.824281150159745, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12410 + }, + { + "epoch": 19.8258785942492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12411 + }, + { + "epoch": 19.82747603833866, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12412 + }, + { + "epoch": 19.829073482428115, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 12413 + }, + { + "epoch": 19.830670926517573, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12414 + }, + { + "epoch": 19.83226837060703, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12415 + }, + { + "epoch": 19.833865814696484, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12416 + }, + { + "epoch": 19.835463258785943, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12417 + }, + { + "epoch": 19.837060702875398, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12418 + }, + { + "epoch": 19.838658146964857, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12419 + }, + { + "epoch": 19.840255591054312, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12420 + }, + { + "epoch": 19.84185303514377, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12421 + }, + { + "epoch": 19.843450479233226, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12422 + }, + { + "epoch": 19.845047923322685, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12423 + }, + { + "epoch": 19.84664536741214, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12424 + }, + { + "epoch": 19.8482428115016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12425 + }, + { + "epoch": 19.849840255591054, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12426 + }, + { + "epoch": 19.85143769968051, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12427 + }, + { + "epoch": 19.85303514376997, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12428 + }, + { + "epoch": 19.854632587859424, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12429 + }, + { + "epoch": 19.856230031948883, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 12430 + }, + { + "epoch": 19.857827476038338, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12431 + }, + { + "epoch": 19.859424920127797, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12432 + }, + { + "epoch": 19.861022364217252, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12433 + }, + { + "epoch": 19.86261980830671, + "grad_norm": 0.032470703125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12434 + }, + { + "epoch": 19.864217252396166, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12435 + }, + { + "epoch": 19.86581469648562, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12436 + }, + { + "epoch": 19.86741214057508, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12437 + }, + { + "epoch": 19.869009584664536, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12438 + }, + { + "epoch": 19.870607028753994, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12439 + }, + { + "epoch": 19.87220447284345, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12440 + }, + { + "epoch": 19.87380191693291, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12441 + }, + { + "epoch": 19.875399361022364, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12442 + }, + { + "epoch": 19.876996805111823, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12443 + }, + { + "epoch": 19.878594249201278, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12444 + }, + { + "epoch": 19.880191693290733, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12445 + }, + { + "epoch": 19.881789137380192, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 12446 + }, + { + "epoch": 19.883386581469647, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12447 + }, + { + "epoch": 19.884984025559106, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12448 + }, + { + "epoch": 19.88658146964856, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 12449 + }, + { + "epoch": 19.88817891373802, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12450 + }, + { + "epoch": 19.889776357827476, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12451 + }, + { + "epoch": 19.891373801916934, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12452 + }, + { + "epoch": 19.89297124600639, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12453 + }, + { + "epoch": 19.894568690095845, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12454 + }, + { + "epoch": 19.896166134185304, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12455 + }, + { + "epoch": 19.89776357827476, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12456 + }, + { + "epoch": 19.899361022364218, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 12457 + }, + { + "epoch": 19.900958466453673, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12458 + }, + { + "epoch": 19.902555910543132, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12459 + }, + { + "epoch": 19.904153354632587, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12460 + }, + { + "epoch": 19.905750798722046, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12461 + }, + { + "epoch": 19.9073482428115, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12462 + }, + { + "epoch": 19.908945686900957, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 12463 + }, + { + "epoch": 19.910543130990416, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12464 + }, + { + "epoch": 19.91214057507987, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12465 + }, + { + "epoch": 19.91373801916933, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12466 + }, + { + "epoch": 19.915335463258785, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12467 + }, + { + "epoch": 19.916932907348244, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12468 + }, + { + "epoch": 19.9185303514377, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12469 + }, + { + "epoch": 19.920127795527158, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12470 + }, + { + "epoch": 19.921725239616613, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12471 + }, + { + "epoch": 19.923322683706072, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 12472 + }, + { + "epoch": 19.924920127795527, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12473 + }, + { + "epoch": 19.926517571884983, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 12474 + }, + { + "epoch": 19.92811501597444, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12475 + }, + { + "epoch": 19.929712460063897, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12476 + }, + { + "epoch": 19.931309904153355, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12477 + }, + { + "epoch": 19.93290734824281, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12478 + }, + { + "epoch": 19.93450479233227, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12479 + }, + { + "epoch": 19.936102236421725, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12480 + }, + { + "epoch": 19.937699680511184, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12481 + }, + { + "epoch": 19.93929712460064, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12482 + }, + { + "epoch": 19.940894568690094, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12483 + }, + { + "epoch": 19.942492012779553, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12484 + }, + { + "epoch": 19.94408945686901, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12485 + }, + { + "epoch": 19.945686900958467, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12486 + }, + { + "epoch": 19.947284345047922, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12487 + }, + { + "epoch": 19.94888178913738, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12488 + }, + { + "epoch": 19.950479233226837, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12489 + }, + { + "epoch": 19.952076677316295, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12490 + }, + { + "epoch": 19.95367412140575, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12491 + }, + { + "epoch": 19.955271565495206, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12492 + }, + { + "epoch": 19.956869009584665, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12493 + }, + { + "epoch": 19.95846645367412, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12494 + }, + { + "epoch": 19.96006389776358, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12495 + }, + { + "epoch": 19.961661341853034, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12496 + }, + { + "epoch": 19.963258785942493, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12497 + }, + { + "epoch": 19.96485623003195, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12498 + }, + { + "epoch": 19.966453674121407, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12499 + }, + { + "epoch": 19.968051118210862, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12500 + }, + { + "epoch": 19.96964856230032, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12501 + }, + { + "epoch": 19.971246006389777, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12502 + }, + { + "epoch": 19.972843450479232, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12503 + }, + { + "epoch": 19.97444089456869, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12504 + }, + { + "epoch": 19.976038338658146, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12505 + }, + { + "epoch": 19.977635782747605, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12506 + }, + { + "epoch": 19.97923322683706, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 12507 + }, + { + "epoch": 19.98083067092652, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12508 + }, + { + "epoch": 19.982428115015974, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12509 + }, + { + "epoch": 19.984025559105433, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12510 + }, + { + "epoch": 19.98562300319489, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 12511 + }, + { + "epoch": 19.987220447284344, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 12512 + }, + { + "epoch": 19.988817891373802, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12513 + }, + { + "epoch": 19.990415335463258, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12514 + }, + { + "epoch": 19.992012779552716, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12515 + }, + { + "epoch": 19.99361022364217, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12516 + }, + { + "epoch": 19.99520766773163, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12517 + }, + { + "epoch": 19.996805111821086, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 12518 + }, + { + "epoch": 19.998402555910545, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12519 + }, + { + "epoch": 20.0, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12520 + }, + { + "epoch": 20.001597444089455, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12521 + }, + { + "epoch": 20.003194888178914, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12522 + }, + { + "epoch": 20.00479233226837, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 12523 + }, + { + "epoch": 20.00638977635783, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12524 + }, + { + "epoch": 20.007987220447284, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12525 + }, + { + "epoch": 20.009584664536742, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12526 + }, + { + "epoch": 20.011182108626198, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12527 + }, + { + "epoch": 20.012779552715656, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12528 + }, + { + "epoch": 20.01437699680511, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12529 + }, + { + "epoch": 20.015974440894567, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12530 + }, + { + "epoch": 20.017571884984026, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12531 + }, + { + "epoch": 20.01916932907348, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12532 + }, + { + "epoch": 20.02076677316294, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12533 + }, + { + "epoch": 20.022364217252395, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 12534 + }, + { + "epoch": 20.023961661341854, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12535 + }, + { + "epoch": 20.02555910543131, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12536 + }, + { + "epoch": 20.027156549520768, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 12537 + }, + { + "epoch": 20.028753993610223, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12538 + }, + { + "epoch": 20.03035143769968, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12539 + }, + { + "epoch": 20.031948881789138, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 12540 + }, + { + "epoch": 20.033546325878593, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 12541 + }, + { + "epoch": 20.03514376996805, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12542 + }, + { + "epoch": 20.036741214057507, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12543 + }, + { + "epoch": 20.038338658146966, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12544 + }, + { + "epoch": 20.03993610223642, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12545 + }, + { + "epoch": 20.04153354632588, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 12546 + }, + { + "epoch": 20.043130990415335, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12547 + }, + { + "epoch": 20.044728434504794, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 12548 + }, + { + "epoch": 20.04632587859425, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12549 + }, + { + "epoch": 20.047923322683705, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12550 + }, + { + "epoch": 20.049520766773163, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12551 + }, + { + "epoch": 20.05111821086262, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12552 + }, + { + "epoch": 20.052715654952078, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12553 + }, + { + "epoch": 20.054313099041533, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12554 + }, + { + "epoch": 20.05591054313099, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12555 + }, + { + "epoch": 20.057507987220447, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12556 + }, + { + "epoch": 20.059105431309906, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12557 + }, + { + "epoch": 20.06070287539936, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12558 + }, + { + "epoch": 20.062300319488816, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12559 + }, + { + "epoch": 20.063897763578275, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12560 + }, + { + "epoch": 20.06549520766773, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12561 + }, + { + "epoch": 20.06709265175719, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12562 + }, + { + "epoch": 20.068690095846645, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12563 + }, + { + "epoch": 20.070287539936103, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12564 + }, + { + "epoch": 20.07188498402556, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12565 + }, + { + "epoch": 20.073482428115017, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 12566 + }, + { + "epoch": 20.075079872204473, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 12567 + }, + { + "epoch": 20.076677316293928, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12568 + }, + { + "epoch": 20.078274760383387, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12569 + }, + { + "epoch": 20.079872204472842, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12570 + }, + { + "epoch": 20.0814696485623, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12571 + }, + { + "epoch": 20.083067092651756, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12572 + }, + { + "epoch": 20.084664536741215, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12573 + }, + { + "epoch": 20.08626198083067, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 12574 + }, + { + "epoch": 20.08785942492013, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12575 + }, + { + "epoch": 20.089456869009584, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12576 + }, + { + "epoch": 20.09105431309904, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12577 + }, + { + "epoch": 20.0926517571885, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12578 + }, + { + "epoch": 20.094249201277954, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12579 + }, + { + "epoch": 20.095846645367413, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12580 + }, + { + "epoch": 20.097444089456868, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12581 + }, + { + "epoch": 20.099041533546327, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12582 + }, + { + "epoch": 20.100638977635782, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 12583 + }, + { + "epoch": 20.10223642172524, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12584 + }, + { + "epoch": 20.103833865814696, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12585 + }, + { + "epoch": 20.105431309904155, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12586 + }, + { + "epoch": 20.10702875399361, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12587 + }, + { + "epoch": 20.108626198083066, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12588 + }, + { + "epoch": 20.110223642172524, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12589 + }, + { + "epoch": 20.11182108626198, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 12590 + }, + { + "epoch": 20.11341853035144, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12591 + }, + { + "epoch": 20.115015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12592 + }, + { + "epoch": 20.116613418530353, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12593 + }, + { + "epoch": 20.118210862619808, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12594 + }, + { + "epoch": 20.119808306709267, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12595 + }, + { + "epoch": 20.121405750798722, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12596 + }, + { + "epoch": 20.123003194888177, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12597 + }, + { + "epoch": 20.124600638977636, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12598 + }, + { + "epoch": 20.12619808306709, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12599 + }, + { + "epoch": 20.12779552715655, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12600 + }, + { + "epoch": 20.129392971246006, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12601 + }, + { + "epoch": 20.130990415335464, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12602 + }, + { + "epoch": 20.13258785942492, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 12603 + }, + { + "epoch": 20.13418530351438, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 12604 + }, + { + "epoch": 20.135782747603834, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12605 + }, + { + "epoch": 20.13738019169329, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12606 + }, + { + "epoch": 20.138977635782748, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12607 + }, + { + "epoch": 20.140575079872203, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 12608 + }, + { + "epoch": 20.142172523961662, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12609 + }, + { + "epoch": 20.143769968051117, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12610 + }, + { + "epoch": 20.145367412140576, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 12611 + }, + { + "epoch": 20.14696485623003, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12612 + }, + { + "epoch": 20.14856230031949, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12613 + }, + { + "epoch": 20.150159744408946, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12614 + }, + { + "epoch": 20.1517571884984, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12615 + }, + { + "epoch": 20.15335463258786, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12616 + }, + { + "epoch": 20.154952076677315, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12617 + }, + { + "epoch": 20.156549520766774, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12618 + }, + { + "epoch": 20.15814696485623, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12619 + }, + { + "epoch": 20.159744408945688, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12620 + }, + { + "epoch": 20.161341853035143, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12621 + }, + { + "epoch": 20.162939297124602, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12622 + }, + { + "epoch": 20.164536741214057, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12623 + }, + { + "epoch": 20.166134185303516, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12624 + }, + { + "epoch": 20.16773162939297, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12625 + }, + { + "epoch": 20.169329073482427, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12626 + }, + { + "epoch": 20.170926517571885, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12627 + }, + { + "epoch": 20.17252396166134, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12628 + }, + { + "epoch": 20.1741214057508, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12629 + }, + { + "epoch": 20.175718849840255, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12630 + }, + { + "epoch": 20.177316293929714, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12631 + }, + { + "epoch": 20.17891373801917, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12632 + }, + { + "epoch": 20.180511182108628, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12633 + }, + { + "epoch": 20.182108626198083, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12634 + }, + { + "epoch": 20.18370607028754, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12635 + }, + { + "epoch": 20.185303514376997, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12636 + }, + { + "epoch": 20.186900958466452, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12637 + }, + { + "epoch": 20.18849840255591, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12638 + }, + { + "epoch": 20.190095846645367, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12639 + }, + { + "epoch": 20.191693290734825, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12640 + }, + { + "epoch": 20.19329073482428, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12641 + }, + { + "epoch": 20.19488817891374, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 12642 + }, + { + "epoch": 20.196485623003195, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12643 + }, + { + "epoch": 20.19808306709265, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12644 + }, + { + "epoch": 20.19968051118211, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 12645 + }, + { + "epoch": 20.201277955271564, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12646 + }, + { + "epoch": 20.202875399361023, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12647 + }, + { + "epoch": 20.20447284345048, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12648 + }, + { + "epoch": 20.206070287539937, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12649 + }, + { + "epoch": 20.207667731629392, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12650 + }, + { + "epoch": 20.20926517571885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12651 + }, + { + "epoch": 20.210862619808307, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 12652 + }, + { + "epoch": 20.212460063897762, + "grad_norm": 0.0537109375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12653 + }, + { + "epoch": 20.21405750798722, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12654 + }, + { + "epoch": 20.215654952076676, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12655 + }, + { + "epoch": 20.217252396166135, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12656 + }, + { + "epoch": 20.21884984025559, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 12657 + }, + { + "epoch": 20.22044728434505, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12658 + }, + { + "epoch": 20.222044728434504, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12659 + }, + { + "epoch": 20.223642172523963, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12660 + }, + { + "epoch": 20.22523961661342, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12661 + }, + { + "epoch": 20.226837060702877, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12662 + }, + { + "epoch": 20.228434504792332, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12663 + }, + { + "epoch": 20.230031948881788, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 12664 + }, + { + "epoch": 20.231629392971247, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12665 + }, + { + "epoch": 20.233226837060702, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 12666 + }, + { + "epoch": 20.23482428115016, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12667 + }, + { + "epoch": 20.236421725239616, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12668 + }, + { + "epoch": 20.238019169329075, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12669 + }, + { + "epoch": 20.23961661341853, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12670 + }, + { + "epoch": 20.24121405750799, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 12671 + }, + { + "epoch": 20.242811501597444, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12672 + }, + { + "epoch": 20.2444089456869, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12673 + }, + { + "epoch": 20.24600638977636, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 12674 + }, + { + "epoch": 20.247603833865814, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12675 + }, + { + "epoch": 20.249201277955272, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 12676 + }, + { + "epoch": 20.250798722044728, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12677 + }, + { + "epoch": 20.252396166134186, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12678 + }, + { + "epoch": 20.25399361022364, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12679 + }, + { + "epoch": 20.2555910543131, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12680 + }, + { + "epoch": 20.257188498402556, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12681 + }, + { + "epoch": 20.25878594249201, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 12682 + }, + { + "epoch": 20.26038338658147, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12683 + }, + { + "epoch": 20.261980830670925, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12684 + }, + { + "epoch": 20.263578274760384, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12685 + }, + { + "epoch": 20.26517571884984, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12686 + }, + { + "epoch": 20.266773162939298, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12687 + }, + { + "epoch": 20.268370607028753, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 12688 + }, + { + "epoch": 20.269968051118212, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12689 + }, + { + "epoch": 20.271565495207668, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 12690 + }, + { + "epoch": 20.273162939297123, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12691 + }, + { + "epoch": 20.27476038338658, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 12692 + }, + { + "epoch": 20.276357827476037, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12693 + }, + { + "epoch": 20.277955271565496, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12694 + }, + { + "epoch": 20.27955271565495, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12695 + }, + { + "epoch": 20.28115015974441, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12696 + }, + { + "epoch": 20.282747603833865, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12697 + }, + { + "epoch": 20.284345047923324, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12698 + }, + { + "epoch": 20.28594249201278, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 12699 + }, + { + "epoch": 20.287539936102238, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12700 + }, + { + "epoch": 20.289137380191693, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12701 + }, + { + "epoch": 20.29073482428115, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12702 + }, + { + "epoch": 20.292332268370608, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12703 + }, + { + "epoch": 20.293929712460063, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12704 + }, + { + "epoch": 20.29552715654952, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 12705 + }, + { + "epoch": 20.297124600638977, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12706 + }, + { + "epoch": 20.298722044728436, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12707 + }, + { + "epoch": 20.30031948881789, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 12708 + }, + { + "epoch": 20.30191693290735, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12709 + }, + { + "epoch": 20.303514376996805, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12710 + }, + { + "epoch": 20.30511182108626, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12711 + }, + { + "epoch": 20.30670926517572, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12712 + }, + { + "epoch": 20.308306709265175, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12713 + }, + { + "epoch": 20.309904153354633, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12714 + }, + { + "epoch": 20.31150159744409, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12715 + }, + { + "epoch": 20.313099041533548, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 12716 + }, + { + "epoch": 20.314696485623003, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12717 + }, + { + "epoch": 20.31629392971246, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12718 + }, + { + "epoch": 20.317891373801917, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12719 + }, + { + "epoch": 20.319488817891372, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12720 + }, + { + "epoch": 20.32108626198083, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12721 + }, + { + "epoch": 20.322683706070286, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 12722 + }, + { + "epoch": 20.324281150159745, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12723 + }, + { + "epoch": 20.3258785942492, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12724 + }, + { + "epoch": 20.32747603833866, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12725 + }, + { + "epoch": 20.329073482428115, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12726 + }, + { + "epoch": 20.330670926517573, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 12727 + }, + { + "epoch": 20.33226837060703, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 12728 + }, + { + "epoch": 20.333865814696484, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12729 + }, + { + "epoch": 20.335463258785943, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12730 + }, + { + "epoch": 20.337060702875398, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12731 + }, + { + "epoch": 20.338658146964857, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12732 + }, + { + "epoch": 20.340255591054312, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12733 + }, + { + "epoch": 20.34185303514377, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12734 + }, + { + "epoch": 20.343450479233226, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12735 + }, + { + "epoch": 20.345047923322685, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12736 + }, + { + "epoch": 20.34664536741214, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12737 + }, + { + "epoch": 20.3482428115016, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 12738 + }, + { + "epoch": 20.349840255591054, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12739 + }, + { + "epoch": 20.35143769968051, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12740 + }, + { + "epoch": 20.35303514376997, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12741 + }, + { + "epoch": 20.354632587859424, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12742 + }, + { + "epoch": 20.356230031948883, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12743 + }, + { + "epoch": 20.357827476038338, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12744 + }, + { + "epoch": 20.359424920127797, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12745 + }, + { + "epoch": 20.361022364217252, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12746 + }, + { + "epoch": 20.36261980830671, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12747 + }, + { + "epoch": 20.364217252396166, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12748 + }, + { + "epoch": 20.36581469648562, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12749 + }, + { + "epoch": 20.36741214057508, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 12750 + }, + { + "epoch": 20.369009584664536, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12751 + }, + { + "epoch": 20.370607028753994, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12752 + }, + { + "epoch": 20.37220447284345, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12753 + }, + { + "epoch": 20.37380191693291, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12754 + }, + { + "epoch": 20.375399361022364, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12755 + }, + { + "epoch": 20.376996805111823, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12756 + }, + { + "epoch": 20.378594249201278, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12757 + }, + { + "epoch": 20.380191693290733, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12758 + }, + { + "epoch": 20.381789137380192, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12759 + }, + { + "epoch": 20.383386581469647, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12760 + }, + { + "epoch": 20.384984025559106, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12761 + }, + { + "epoch": 20.38658146964856, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 12762 + }, + { + "epoch": 20.38817891373802, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12763 + }, + { + "epoch": 20.389776357827476, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12764 + }, + { + "epoch": 20.391373801916934, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12765 + }, + { + "epoch": 20.39297124600639, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12766 + }, + { + "epoch": 20.394568690095845, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12767 + }, + { + "epoch": 20.396166134185304, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12768 + }, + { + "epoch": 20.39776357827476, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12769 + }, + { + "epoch": 20.399361022364218, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 12770 + }, + { + "epoch": 20.400958466453673, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12771 + }, + { + "epoch": 20.402555910543132, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12772 + }, + { + "epoch": 20.404153354632587, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 12773 + }, + { + "epoch": 20.405750798722046, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12774 + }, + { + "epoch": 20.4073482428115, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12775 + }, + { + "epoch": 20.408945686900957, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 12776 + }, + { + "epoch": 20.410543130990416, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 12777 + }, + { + "epoch": 20.41214057507987, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 12778 + }, + { + "epoch": 20.41373801916933, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 12779 + }, + { + "epoch": 20.415335463258785, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12780 + }, + { + "epoch": 20.416932907348244, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 12781 + }, + { + "epoch": 20.4185303514377, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12782 + }, + { + "epoch": 20.420127795527158, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12783 + }, + { + "epoch": 20.421725239616613, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12784 + }, + { + "epoch": 20.423322683706072, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12785 + }, + { + "epoch": 20.424920127795527, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12786 + }, + { + "epoch": 20.426517571884983, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12787 + }, + { + "epoch": 20.42811501597444, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12788 + }, + { + "epoch": 20.429712460063897, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 12789 + }, + { + "epoch": 20.431309904153355, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12790 + }, + { + "epoch": 20.43290734824281, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12791 + }, + { + "epoch": 20.43450479233227, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12792 + }, + { + "epoch": 20.436102236421725, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 12793 + }, + { + "epoch": 20.437699680511184, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12794 + }, + { + "epoch": 20.43929712460064, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 12795 + }, + { + "epoch": 20.440894568690094, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 12796 + }, + { + "epoch": 20.442492012779553, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 12797 + }, + { + "epoch": 20.44408945686901, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12798 + }, + { + "epoch": 20.445686900958467, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12799 + }, + { + "epoch": 20.447284345047922, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12800 + }, + { + "epoch": 20.44888178913738, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12801 + }, + { + "epoch": 20.450479233226837, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12802 + }, + { + "epoch": 20.452076677316295, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12803 + }, + { + "epoch": 20.45367412140575, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12804 + }, + { + "epoch": 20.455271565495206, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 12805 + }, + { + "epoch": 20.456869009584665, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 12806 + }, + { + "epoch": 20.45846645367412, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 12807 + }, + { + "epoch": 20.46006389776358, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12808 + }, + { + "epoch": 20.461661341853034, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 12809 + }, + { + "epoch": 20.463258785942493, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 12810 + }, + { + "epoch": 20.46485623003195, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12811 + }, + { + "epoch": 20.466453674121407, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 12812 + }, + { + "epoch": 20.468051118210862, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12813 + }, + { + "epoch": 20.46964856230032, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12814 + }, + { + "epoch": 20.471246006389777, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12815 + }, + { + "epoch": 20.472843450479232, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12816 + }, + { + "epoch": 20.47444089456869, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12817 + }, + { + "epoch": 20.476038338658146, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12818 + }, + { + "epoch": 20.477635782747605, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12819 + }, + { + "epoch": 20.47923322683706, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12820 + }, + { + "epoch": 20.48083067092652, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12821 + }, + { + "epoch": 20.482428115015974, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12822 + }, + { + "epoch": 20.484025559105433, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12823 + }, + { + "epoch": 20.48562300319489, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 12824 + }, + { + "epoch": 20.487220447284344, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12825 + }, + { + "epoch": 20.488817891373802, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 12826 + }, + { + "epoch": 20.490415335463258, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 12827 + }, + { + "epoch": 20.492012779552716, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12828 + }, + { + "epoch": 20.49361022364217, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12829 + }, + { + "epoch": 20.49520766773163, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 12830 + }, + { + "epoch": 20.496805111821086, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12831 + }, + { + "epoch": 20.498402555910545, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12832 + }, + { + "epoch": 20.5, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12833 + }, + { + "epoch": 20.501597444089455, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12834 + }, + { + "epoch": 20.503194888178914, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12835 + }, + { + "epoch": 20.50479233226837, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12836 + }, + { + "epoch": 20.50638977635783, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12837 + }, + { + "epoch": 20.507987220447284, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 12838 + }, + { + "epoch": 20.509584664536742, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 12839 + }, + { + "epoch": 20.511182108626198, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 12840 + }, + { + "epoch": 20.512779552715656, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12841 + }, + { + "epoch": 20.51437699680511, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12842 + }, + { + "epoch": 20.515974440894567, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12843 + }, + { + "epoch": 20.517571884984026, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12844 + }, + { + "epoch": 20.51916932907348, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 12845 + }, + { + "epoch": 20.52076677316294, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 12846 + }, + { + "epoch": 20.522364217252395, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 12847 + }, + { + "epoch": 20.523961661341854, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12848 + }, + { + "epoch": 20.52555910543131, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12849 + }, + { + "epoch": 20.527156549520768, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12850 + }, + { + "epoch": 20.528753993610223, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12851 + }, + { + "epoch": 20.53035143769968, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12852 + }, + { + "epoch": 20.531948881789138, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12853 + }, + { + "epoch": 20.533546325878593, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12854 + }, + { + "epoch": 20.53514376996805, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12855 + }, + { + "epoch": 20.536741214057507, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12856 + }, + { + "epoch": 20.538338658146966, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12857 + }, + { + "epoch": 20.53993610223642, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12858 + }, + { + "epoch": 20.54153354632588, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12859 + }, + { + "epoch": 20.543130990415335, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 12860 + }, + { + "epoch": 20.544728434504794, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 12861 + }, + { + "epoch": 20.54632587859425, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 12862 + }, + { + "epoch": 20.547923322683705, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12863 + }, + { + "epoch": 20.549520766773163, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12864 + }, + { + "epoch": 20.55111821086262, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 12865 + }, + { + "epoch": 20.552715654952078, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12866 + }, + { + "epoch": 20.554313099041533, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12867 + }, + { + "epoch": 20.55591054313099, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12868 + }, + { + "epoch": 20.557507987220447, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12869 + }, + { + "epoch": 20.559105431309906, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 12870 + }, + { + "epoch": 20.56070287539936, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12871 + }, + { + "epoch": 20.562300319488816, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12872 + }, + { + "epoch": 20.563897763578275, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 12873 + }, + { + "epoch": 20.56549520766773, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 12874 + }, + { + "epoch": 20.56709265175719, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12875 + }, + { + "epoch": 20.568690095846645, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12876 + }, + { + "epoch": 20.570287539936103, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12877 + }, + { + "epoch": 20.57188498402556, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12878 + }, + { + "epoch": 20.573482428115017, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12879 + }, + { + "epoch": 20.575079872204473, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 12880 + }, + { + "epoch": 20.576677316293928, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 12881 + }, + { + "epoch": 20.578274760383387, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 12882 + }, + { + "epoch": 20.579872204472842, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12883 + }, + { + "epoch": 20.5814696485623, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12884 + }, + { + "epoch": 20.583067092651756, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12885 + }, + { + "epoch": 20.584664536741215, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12886 + }, + { + "epoch": 20.58626198083067, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 12887 + }, + { + "epoch": 20.58785942492013, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12888 + }, + { + "epoch": 20.589456869009584, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 12889 + }, + { + "epoch": 20.591054313099043, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 12890 + }, + { + "epoch": 20.5926517571885, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12891 + }, + { + "epoch": 20.594249201277954, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12892 + }, + { + "epoch": 20.595846645367413, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 12893 + }, + { + "epoch": 20.597444089456868, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12894 + }, + { + "epoch": 20.599041533546327, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 12895 + }, + { + "epoch": 20.600638977635782, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 12896 + }, + { + "epoch": 20.60223642172524, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 12897 + }, + { + "epoch": 20.603833865814696, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12898 + }, + { + "epoch": 20.605431309904155, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 12899 + }, + { + "epoch": 20.60702875399361, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12900 + }, + { + "epoch": 20.608626198083066, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12901 + }, + { + "epoch": 20.610223642172524, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12902 + }, + { + "epoch": 20.61182108626198, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12903 + }, + { + "epoch": 20.61341853035144, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12904 + }, + { + "epoch": 20.615015974440894, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 12905 + }, + { + "epoch": 20.616613418530353, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 12906 + }, + { + "epoch": 20.618210862619808, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12907 + }, + { + "epoch": 20.619808306709267, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 12908 + }, + { + "epoch": 20.621405750798722, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12909 + }, + { + "epoch": 20.623003194888177, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12910 + }, + { + "epoch": 20.624600638977636, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 12911 + }, + { + "epoch": 20.62619808306709, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 12912 + }, + { + "epoch": 20.62779552715655, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 12913 + }, + { + "epoch": 20.629392971246006, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 12914 + }, + { + "epoch": 20.630990415335464, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12915 + }, + { + "epoch": 20.63258785942492, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 12916 + }, + { + "epoch": 20.63418530351438, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 12917 + }, + { + "epoch": 20.635782747603834, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 12918 + }, + { + "epoch": 20.63738019169329, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12919 + }, + { + "epoch": 20.638977635782748, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 12920 + }, + { + "epoch": 20.640575079872203, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 12921 + }, + { + "epoch": 20.642172523961662, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12922 + }, + { + "epoch": 20.643769968051117, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12923 + }, + { + "epoch": 20.645367412140576, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12924 + }, + { + "epoch": 20.64696485623003, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 12925 + }, + { + "epoch": 20.64856230031949, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12926 + }, + { + "epoch": 20.650159744408946, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 12927 + }, + { + "epoch": 20.6517571884984, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12928 + }, + { + "epoch": 20.65335463258786, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 12929 + }, + { + "epoch": 20.654952076677315, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 12930 + }, + { + "epoch": 20.656549520766774, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12931 + }, + { + "epoch": 20.65814696485623, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12932 + }, + { + "epoch": 20.659744408945688, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 12933 + }, + { + "epoch": 20.661341853035143, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12934 + }, + { + "epoch": 20.662939297124602, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 12935 + }, + { + "epoch": 20.664536741214057, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12936 + }, + { + "epoch": 20.666134185303516, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12937 + }, + { + "epoch": 20.66773162939297, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 12938 + }, + { + "epoch": 20.669329073482427, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 12939 + }, + { + "epoch": 20.670926517571885, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 12940 + }, + { + "epoch": 20.67252396166134, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12941 + }, + { + "epoch": 20.6741214057508, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 12942 + }, + { + "epoch": 20.675718849840255, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12943 + }, + { + "epoch": 20.677316293929714, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 12944 + }, + { + "epoch": 20.67891373801917, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 12945 + }, + { + "epoch": 20.680511182108628, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 12946 + }, + { + "epoch": 20.682108626198083, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12947 + }, + { + "epoch": 20.68370607028754, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12948 + }, + { + "epoch": 20.685303514376997, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 12949 + }, + { + "epoch": 20.686900958466452, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 12950 + }, + { + "epoch": 20.68849840255591, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 12951 + }, + { + "epoch": 20.690095846645367, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12952 + }, + { + "epoch": 20.691693290734825, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 12953 + }, + { + "epoch": 20.69329073482428, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 12954 + }, + { + "epoch": 20.69488817891374, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 12955 + }, + { + "epoch": 20.696485623003195, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 12956 + }, + { + "epoch": 20.69808306709265, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12957 + }, + { + "epoch": 20.69968051118211, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 12958 + }, + { + "epoch": 20.701277955271564, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12959 + }, + { + "epoch": 20.702875399361023, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 12960 + }, + { + "epoch": 20.70447284345048, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 12961 + }, + { + "epoch": 20.706070287539937, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 12962 + }, + { + "epoch": 20.707667731629392, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12963 + }, + { + "epoch": 20.70926517571885, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 12964 + }, + { + "epoch": 20.710862619808307, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 12965 + }, + { + "epoch": 20.712460063897765, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12966 + }, + { + "epoch": 20.71405750798722, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 12967 + }, + { + "epoch": 20.715654952076676, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 12968 + }, + { + "epoch": 20.717252396166135, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 12969 + }, + { + "epoch": 20.71884984025559, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 12970 + }, + { + "epoch": 20.72044728434505, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 12971 + }, + { + "epoch": 20.722044728434504, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 12972 + }, + { + "epoch": 20.723642172523963, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 12973 + }, + { + "epoch": 20.72523961661342, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 12974 + }, + { + "epoch": 20.726837060702877, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 12975 + }, + { + "epoch": 20.728434504792332, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 12976 + }, + { + "epoch": 20.730031948881788, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12977 + }, + { + "epoch": 20.731629392971247, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 12978 + }, + { + "epoch": 20.733226837060702, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 12979 + }, + { + "epoch": 20.73482428115016, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 12980 + }, + { + "epoch": 20.736421725239616, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 12981 + }, + { + "epoch": 20.738019169329075, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 12982 + }, + { + "epoch": 20.73961661341853, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 12983 + }, + { + "epoch": 20.74121405750799, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 12984 + }, + { + "epoch": 20.742811501597444, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 12985 + }, + { + "epoch": 20.7444089456869, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 12986 + }, + { + "epoch": 20.74600638977636, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 12987 + }, + { + "epoch": 20.747603833865814, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 12988 + }, + { + "epoch": 20.749201277955272, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 12989 + }, + { + "epoch": 20.750798722044728, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12990 + }, + { + "epoch": 20.752396166134186, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12991 + }, + { + "epoch": 20.75399361022364, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 12992 + }, + { + "epoch": 20.7555910543131, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 12993 + }, + { + "epoch": 20.757188498402556, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 12994 + }, + { + "epoch": 20.75878594249201, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 12995 + }, + { + "epoch": 20.76038338658147, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 12996 + }, + { + "epoch": 20.761980830670925, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 12997 + }, + { + "epoch": 20.763578274760384, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 12998 + }, + { + "epoch": 20.76517571884984, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 12999 + }, + { + "epoch": 20.766773162939298, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 13000 + }, + { + "epoch": 20.768370607028753, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 13001 + }, + { + "epoch": 20.769968051118212, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13002 + }, + { + "epoch": 20.771565495207668, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13003 + }, + { + "epoch": 20.773162939297123, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 13004 + }, + { + "epoch": 20.77476038338658, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13005 + }, + { + "epoch": 20.776357827476037, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13006 + }, + { + "epoch": 20.777955271565496, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 13007 + }, + { + "epoch": 20.77955271565495, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 13008 + }, + { + "epoch": 20.78115015974441, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 13009 + }, + { + "epoch": 20.782747603833865, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 13010 + }, + { + "epoch": 20.784345047923324, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13011 + }, + { + "epoch": 20.78594249201278, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13012 + }, + { + "epoch": 20.787539936102235, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13013 + }, + { + "epoch": 20.789137380191693, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 13014 + }, + { + "epoch": 20.79073482428115, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 13015 + }, + { + "epoch": 20.792332268370608, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 13016 + }, + { + "epoch": 20.793929712460063, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 13017 + }, + { + "epoch": 20.79552715654952, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 13018 + }, + { + "epoch": 20.797124600638977, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 13019 + }, + { + "epoch": 20.798722044728436, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 13020 + }, + { + "epoch": 20.80031948881789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13021 + }, + { + "epoch": 20.80191693290735, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 13022 + }, + { + "epoch": 20.803514376996805, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 13023 + }, + { + "epoch": 20.80511182108626, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 13024 + }, + { + "epoch": 20.80670926517572, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 13025 + }, + { + "epoch": 20.808306709265175, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 13026 + }, + { + "epoch": 20.809904153354633, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 13027 + }, + { + "epoch": 20.81150159744409, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 13028 + }, + { + "epoch": 20.813099041533548, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 13029 + }, + { + "epoch": 20.814696485623003, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 13030 + }, + { + "epoch": 20.81629392971246, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 13031 + }, + { + "epoch": 20.817891373801917, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 13032 + }, + { + "epoch": 20.819488817891372, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 13033 + }, + { + "epoch": 20.82108626198083, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 13034 + }, + { + "epoch": 20.822683706070286, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 13035 + }, + { + "epoch": 20.824281150159745, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 13036 + }, + { + "epoch": 20.8258785942492, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 13037 + }, + { + "epoch": 20.82747603833866, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 13038 + }, + { + "epoch": 20.829073482428115, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 13039 + }, + { + "epoch": 20.830670926517573, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 13040 + }, + { + "epoch": 20.83226837060703, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 13041 + }, + { + "epoch": 20.833865814696484, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 13042 + }, + { + "epoch": 20.835463258785943, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 13043 + }, + { + "epoch": 20.837060702875398, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 13044 + }, + { + "epoch": 20.838658146964857, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13045 + }, + { + "epoch": 20.840255591054312, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 13046 + }, + { + "epoch": 20.84185303514377, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 13047 + }, + { + "epoch": 20.843450479233226, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 13048 + }, + { + "epoch": 20.845047923322685, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 13049 + }, + { + "epoch": 20.84664536741214, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 13050 + }, + { + "epoch": 20.8482428115016, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 13051 + }, + { + "epoch": 20.849840255591054, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13052 + }, + { + "epoch": 20.85143769968051, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13053 + }, + { + "epoch": 20.85303514376997, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 13054 + }, + { + "epoch": 20.854632587859424, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 13055 + }, + { + "epoch": 20.856230031948883, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 13056 + }, + { + "epoch": 20.857827476038338, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 13057 + }, + { + "epoch": 20.859424920127797, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 13058 + }, + { + "epoch": 20.861022364217252, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13059 + }, + { + "epoch": 20.86261980830671, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 13060 + }, + { + "epoch": 20.864217252396166, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 13061 + }, + { + "epoch": 20.86581469648562, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 13062 + }, + { + "epoch": 20.86741214057508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13063 + }, + { + "epoch": 20.869009584664536, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 13064 + }, + { + "epoch": 20.870607028753994, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13065 + }, + { + "epoch": 20.87220447284345, + "grad_norm": 0.053466796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 13066 + }, + { + "epoch": 20.87380191693291, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 13067 + }, + { + "epoch": 20.875399361022364, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 13068 + }, + { + "epoch": 20.876996805111823, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 13069 + }, + { + "epoch": 20.878594249201278, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 13070 + }, + { + "epoch": 20.880191693290733, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 13071 + }, + { + "epoch": 20.881789137380192, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 13072 + }, + { + "epoch": 20.883386581469647, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 13073 + }, + { + "epoch": 20.884984025559106, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13074 + }, + { + "epoch": 20.88658146964856, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 13075 + }, + { + "epoch": 20.88817891373802, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 13076 + }, + { + "epoch": 20.889776357827476, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13077 + }, + { + "epoch": 20.891373801916934, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 13078 + }, + { + "epoch": 20.89297124600639, + "grad_norm": 0.043212890625, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 13079 + }, + { + "epoch": 20.894568690095845, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 13080 + }, + { + "epoch": 20.896166134185304, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 13081 + }, + { + "epoch": 20.89776357827476, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 13082 + }, + { + "epoch": 20.899361022364218, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 13083 + }, + { + "epoch": 20.900958466453673, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 13084 + }, + { + "epoch": 20.902555910543132, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 13085 + }, + { + "epoch": 20.904153354632587, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 13086 + }, + { + "epoch": 20.905750798722046, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 13087 + }, + { + "epoch": 20.9073482428115, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 13088 + }, + { + "epoch": 20.908945686900957, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13089 + }, + { + "epoch": 20.910543130990416, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13090 + }, + { + "epoch": 20.91214057507987, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13091 + }, + { + "epoch": 20.91373801916933, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 13092 + }, + { + "epoch": 20.915335463258785, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 13093 + }, + { + "epoch": 20.916932907348244, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13094 + }, + { + "epoch": 20.9185303514377, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 13095 + }, + { + "epoch": 20.920127795527158, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 13096 + }, + { + "epoch": 20.921725239616613, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 13097 + }, + { + "epoch": 20.923322683706072, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 13098 + }, + { + "epoch": 20.924920127795527, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 13099 + }, + { + "epoch": 20.926517571884983, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13100 + }, + { + "epoch": 20.92811501597444, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 13101 + }, + { + "epoch": 20.929712460063897, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 13102 + }, + { + "epoch": 20.931309904153355, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 13103 + }, + { + "epoch": 20.93290734824281, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 13104 + }, + { + "epoch": 20.93450479233227, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13105 + }, + { + "epoch": 20.936102236421725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 13106 + }, + { + "epoch": 20.937699680511184, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 13107 + }, + { + "epoch": 20.93929712460064, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13108 + }, + { + "epoch": 20.940894568690094, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 13109 + }, + { + "epoch": 20.942492012779553, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13110 + }, + { + "epoch": 20.94408945686901, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13111 + }, + { + "epoch": 20.945686900958467, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 13112 + }, + { + "epoch": 20.947284345047922, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 13113 + }, + { + "epoch": 20.94888178913738, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 13114 + }, + { + "epoch": 20.950479233226837, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 13115 + }, + { + "epoch": 20.952076677316295, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 13116 + }, + { + "epoch": 20.95367412140575, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 13117 + }, + { + "epoch": 20.955271565495206, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 13118 + }, + { + "epoch": 20.956869009584665, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13119 + }, + { + "epoch": 20.95846645367412, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13120 + }, + { + "epoch": 20.96006389776358, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 13121 + }, + { + "epoch": 20.961661341853034, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 13122 + }, + { + "epoch": 20.963258785942493, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 13123 + }, + { + "epoch": 20.96485623003195, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 13124 + }, + { + "epoch": 20.966453674121407, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 13125 + }, + { + "epoch": 20.968051118210862, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 13126 + }, + { + "epoch": 20.96964856230032, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 13127 + }, + { + "epoch": 20.971246006389777, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 13128 + }, + { + "epoch": 20.972843450479232, + "grad_norm": 0.047607421875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 13129 + }, + { + "epoch": 20.97444089456869, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13130 + }, + { + "epoch": 20.976038338658146, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 13131 + }, + { + "epoch": 20.977635782747605, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 13132 + }, + { + "epoch": 20.97923322683706, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 13133 + }, + { + "epoch": 20.98083067092652, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 13134 + }, + { + "epoch": 20.982428115015974, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 13135 + }, + { + "epoch": 20.984025559105433, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 13136 + }, + { + "epoch": 20.98562300319489, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 13137 + }, + { + "epoch": 20.987220447284344, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 13138 + }, + { + "epoch": 20.988817891373802, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 13139 + }, + { + "epoch": 20.990415335463258, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 13140 + }, + { + "epoch": 20.992012779552716, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 13141 + }, + { + "epoch": 20.99361022364217, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 13142 + }, + { + "epoch": 20.99520766773163, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 13143 + }, + { + "epoch": 20.996805111821086, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 13144 + }, + { + "epoch": 20.998402555910545, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 13145 + }, + { + "epoch": 21.0, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 13146 + }, + { + "epoch": 21.001597444089455, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 13147 + }, + { + "epoch": 21.003194888178914, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 13148 + }, + { + "epoch": 21.00479233226837, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 13149 + }, + { + "epoch": 21.00638977635783, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 13150 + }, + { + "epoch": 21.007987220447284, + "grad_norm": 0.052490234375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 13151 + }, + { + "epoch": 21.009584664536742, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 13152 + }, + { + "epoch": 21.011182108626198, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 13153 + }, + { + "epoch": 21.012779552715656, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13154 + }, + { + "epoch": 21.01437699680511, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 13155 + }, + { + "epoch": 21.015974440894567, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 13156 + }, + { + "epoch": 21.017571884984026, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 13157 + }, + { + "epoch": 21.01916932907348, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 13158 + }, + { + "epoch": 21.02076677316294, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 13159 + }, + { + "epoch": 21.022364217252395, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 13160 + }, + { + "epoch": 21.023961661341854, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 13161 + }, + { + "epoch": 21.02555910543131, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 13162 + }, + { + "epoch": 21.027156549520768, + "grad_norm": 0.0322265625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 13163 + }, + { + "epoch": 21.028753993610223, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 13164 + }, + { + "epoch": 21.03035143769968, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 13165 + }, + { + "epoch": 21.031948881789138, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 13166 + }, + { + "epoch": 21.033546325878593, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 13167 + }, + { + "epoch": 21.03514376996805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 13168 + }, + { + "epoch": 21.036741214057507, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 13169 + }, + { + "epoch": 21.038338658146966, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 13170 + }, + { + "epoch": 21.03993610223642, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 13171 + }, + { + "epoch": 21.04153354632588, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 13172 + }, + { + "epoch": 21.043130990415335, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13173 + }, + { + "epoch": 21.044728434504794, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 13174 + }, + { + "epoch": 21.04632587859425, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 13175 + }, + { + "epoch": 21.047923322683705, + "grad_norm": 0.05078125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 13176 + }, + { + "epoch": 21.049520766773163, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 13177 + }, + { + "epoch": 21.05111821086262, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 13178 + }, + { + "epoch": 21.052715654952078, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 13179 + }, + { + "epoch": 21.054313099041533, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 13180 + }, + { + "epoch": 21.05591054313099, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 13181 + }, + { + "epoch": 21.057507987220447, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 13182 + }, + { + "epoch": 21.059105431309906, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 13183 + }, + { + "epoch": 21.06070287539936, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 13184 + }, + { + "epoch": 21.062300319488816, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 13185 + }, + { + "epoch": 21.063897763578275, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 13186 + }, + { + "epoch": 21.06549520766773, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 13187 + }, + { + "epoch": 21.06709265175719, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 13188 + }, + { + "epoch": 21.068690095846645, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 13189 + }, + { + "epoch": 21.070287539936103, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13190 + }, + { + "epoch": 21.07188498402556, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 13191 + }, + { + "epoch": 21.073482428115017, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 13192 + }, + { + "epoch": 21.075079872204473, + "grad_norm": 0.03466796875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13193 + }, + { + "epoch": 21.076677316293928, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 13194 + }, + { + "epoch": 21.078274760383387, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 13195 + }, + { + "epoch": 21.079872204472842, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 13196 + }, + { + "epoch": 21.0814696485623, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 13197 + }, + { + "epoch": 21.083067092651756, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 13198 + }, + { + "epoch": 21.084664536741215, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 13199 + }, + { + "epoch": 21.08626198083067, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13200 + }, + { + "epoch": 21.08785942492013, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 13201 + }, + { + "epoch": 21.089456869009584, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 13202 + }, + { + "epoch": 21.09105431309904, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 13203 + }, + { + "epoch": 21.0926517571885, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 13204 + }, + { + "epoch": 21.094249201277954, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 13205 + }, + { + "epoch": 21.095846645367413, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 13206 + }, + { + "epoch": 21.097444089456868, + "grad_norm": 0.04443359375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 13207 + }, + { + "epoch": 21.099041533546327, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 13208 + }, + { + "epoch": 21.100638977635782, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 13209 + }, + { + "epoch": 21.10223642172524, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 13210 + }, + { + "epoch": 21.103833865814696, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13211 + }, + { + "epoch": 21.105431309904155, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 13212 + }, + { + "epoch": 21.10702875399361, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 13213 + }, + { + "epoch": 21.108626198083066, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 13214 + }, + { + "epoch": 21.110223642172524, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 13215 + }, + { + "epoch": 21.11182108626198, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 13216 + }, + { + "epoch": 21.11341853035144, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13217 + }, + { + "epoch": 21.115015974440894, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13218 + }, + { + "epoch": 21.116613418530353, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 13219 + }, + { + "epoch": 21.118210862619808, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 13220 + }, + { + "epoch": 21.119808306709267, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 13221 + }, + { + "epoch": 21.121405750798722, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 13222 + }, + { + "epoch": 21.123003194888177, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 13223 + }, + { + "epoch": 21.124600638977636, + "grad_norm": 0.038330078125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 13224 + }, + { + "epoch": 21.12619808306709, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 13225 + }, + { + "epoch": 21.12779552715655, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13226 + }, + { + "epoch": 21.129392971246006, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 13227 + }, + { + "epoch": 21.130990415335464, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 13228 + }, + { + "epoch": 21.13258785942492, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 13229 + }, + { + "epoch": 21.13418530351438, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 13230 + }, + { + "epoch": 21.135782747603834, + "grad_norm": 0.051025390625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 13231 + }, + { + "epoch": 21.13738019169329, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 13232 + }, + { + "epoch": 21.138977635782748, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 13233 + }, + { + "epoch": 21.140575079872203, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 13234 + }, + { + "epoch": 21.142172523961662, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 13235 + }, + { + "epoch": 21.143769968051117, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 13236 + }, + { + "epoch": 21.145367412140576, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 13237 + }, + { + "epoch": 21.14696485623003, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 13238 + }, + { + "epoch": 21.14856230031949, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13239 + }, + { + "epoch": 21.150159744408946, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 13240 + }, + { + "epoch": 21.1517571884984, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 13241 + }, + { + "epoch": 21.15335463258786, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 13242 + }, + { + "epoch": 21.154952076677315, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 13243 + }, + { + "epoch": 21.156549520766774, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 13244 + }, + { + "epoch": 21.15814696485623, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13245 + }, + { + "epoch": 21.159744408945688, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 13246 + }, + { + "epoch": 21.161341853035143, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 13247 + }, + { + "epoch": 21.162939297124602, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13248 + }, + { + "epoch": 21.164536741214057, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13249 + }, + { + "epoch": 21.166134185303516, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 13250 + }, + { + "epoch": 21.16773162939297, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 13251 + }, + { + "epoch": 21.169329073482427, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 13252 + }, + { + "epoch": 21.170926517571885, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 13253 + }, + { + "epoch": 21.17252396166134, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 13254 + }, + { + "epoch": 21.1741214057508, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 13255 + }, + { + "epoch": 21.175718849840255, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 13256 + }, + { + "epoch": 21.177316293929714, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 13257 + }, + { + "epoch": 21.17891373801917, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 13258 + }, + { + "epoch": 21.180511182108628, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 13259 + }, + { + "epoch": 21.182108626198083, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 13260 + }, + { + "epoch": 21.18370607028754, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 13261 + }, + { + "epoch": 21.185303514376997, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 13262 + }, + { + "epoch": 21.186900958466452, + "grad_norm": 0.033935546875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 13263 + }, + { + "epoch": 21.18849840255591, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 13264 + }, + { + "epoch": 21.190095846645367, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 13265 + }, + { + "epoch": 21.191693290734825, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13266 + }, + { + "epoch": 21.19329073482428, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 13267 + }, + { + "epoch": 21.19488817891374, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 13268 + }, + { + "epoch": 21.196485623003195, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 13269 + }, + { + "epoch": 21.19808306709265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 13270 + }, + { + "epoch": 21.19968051118211, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 13271 + }, + { + "epoch": 21.201277955271564, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 13272 + }, + { + "epoch": 21.202875399361023, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 13273 + }, + { + "epoch": 21.20447284345048, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 13274 + }, + { + "epoch": 21.206070287539937, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 13275 + }, + { + "epoch": 21.207667731629392, + "grad_norm": 0.052734375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 13276 + }, + { + "epoch": 21.20926517571885, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 13277 + }, + { + "epoch": 21.210862619808307, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 13278 + }, + { + "epoch": 21.212460063897762, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 13279 + }, + { + "epoch": 21.21405750798722, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13280 + }, + { + "epoch": 21.215654952076676, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 13281 + }, + { + "epoch": 21.217252396166135, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 13282 + }, + { + "epoch": 21.21884984025559, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 13283 + }, + { + "epoch": 21.22044728434505, + "grad_norm": 0.042236328125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 13284 + }, + { + "epoch": 21.222044728434504, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13285 + }, + { + "epoch": 21.223642172523963, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 13286 + }, + { + "epoch": 21.22523961661342, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 13287 + }, + { + "epoch": 21.226837060702877, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13288 + }, + { + "epoch": 21.228434504792332, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 13289 + }, + { + "epoch": 21.230031948881788, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 13290 + }, + { + "epoch": 21.231629392971247, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 13291 + }, + { + "epoch": 21.233226837060702, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 13292 + }, + { + "epoch": 21.23482428115016, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 13293 + }, + { + "epoch": 21.236421725239616, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 13294 + }, + { + "epoch": 21.238019169329075, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 13295 + }, + { + "epoch": 21.23961661341853, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 13296 + }, + { + "epoch": 21.24121405750799, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 13297 + }, + { + "epoch": 21.242811501597444, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 13298 + }, + { + "epoch": 21.2444089456869, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 13299 + }, + { + "epoch": 21.24600638977636, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 13300 + }, + { + "epoch": 21.247603833865814, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 13301 + }, + { + "epoch": 21.249201277955272, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 13302 + }, + { + "epoch": 21.250798722044728, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 13303 + }, + { + "epoch": 21.252396166134186, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 13304 + }, + { + "epoch": 21.25399361022364, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 13305 + }, + { + "epoch": 21.2555910543131, + "grad_norm": 0.039794921875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 13306 + }, + { + "epoch": 21.257188498402556, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 13307 + }, + { + "epoch": 21.25878594249201, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 13308 + }, + { + "epoch": 21.26038338658147, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 13309 + }, + { + "epoch": 21.261980830670925, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 13310 + }, + { + "epoch": 21.263578274760384, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 13311 + }, + { + "epoch": 21.26517571884984, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13312 + }, + { + "epoch": 21.266773162939298, + "grad_norm": 0.033447265625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 13313 + }, + { + "epoch": 21.268370607028753, + "grad_norm": 0.05224609375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 13314 + }, + { + "epoch": 21.269968051118212, + "grad_norm": 0.03271484375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 13315 + }, + { + "epoch": 21.271565495207668, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 13316 + }, + { + "epoch": 21.273162939297123, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 13317 + }, + { + "epoch": 21.27476038338658, + "grad_norm": 0.033203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 13318 + }, + { + "epoch": 21.276357827476037, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 13319 + }, + { + "epoch": 21.277955271565496, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 13320 + }, + { + "epoch": 21.27955271565495, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 13321 + }, + { + "epoch": 21.28115015974441, + "grad_norm": 0.037841796875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 13322 + }, + { + "epoch": 21.282747603833865, + "grad_norm": 0.0361328125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13323 + }, + { + "epoch": 21.284345047923324, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 13324 + }, + { + "epoch": 21.28594249201278, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 13325 + }, + { + "epoch": 21.287539936102238, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 13326 + }, + { + "epoch": 21.289137380191693, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 13327 + }, + { + "epoch": 21.29073482428115, + "grad_norm": 0.0458984375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13328 + }, + { + "epoch": 21.292332268370608, + "grad_norm": 0.039306640625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 13329 + }, + { + "epoch": 21.293929712460063, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 13330 + }, + { + "epoch": 21.29552715654952, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 13331 + }, + { + "epoch": 21.297124600638977, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13332 + }, + { + "epoch": 21.298722044728436, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 13333 + }, + { + "epoch": 21.30031948881789, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 13334 + }, + { + "epoch": 21.30191693290735, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 13335 + }, + { + "epoch": 21.303514376996805, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 13336 + }, + { + "epoch": 21.30511182108626, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 13337 + }, + { + "epoch": 21.30670926517572, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 13338 + }, + { + "epoch": 21.308306709265175, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 13339 + }, + { + "epoch": 21.309904153354633, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 13340 + }, + { + "epoch": 21.31150159744409, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 13341 + }, + { + "epoch": 21.313099041533548, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 13342 + }, + { + "epoch": 21.314696485623003, + "grad_norm": 0.043701171875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 13343 + }, + { + "epoch": 21.31629392971246, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 13344 + }, + { + "epoch": 21.317891373801917, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 13345 + }, + { + "epoch": 21.319488817891372, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 13346 + }, + { + "epoch": 21.32108626198083, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 13347 + }, + { + "epoch": 21.322683706070286, + "grad_norm": 0.048583984375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 13348 + }, + { + "epoch": 21.324281150159745, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 13349 + }, + { + "epoch": 21.3258785942492, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 13350 + }, + { + "epoch": 21.32747603833866, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 13351 + }, + { + "epoch": 21.329073482428115, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 13352 + }, + { + "epoch": 21.330670926517573, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 13353 + }, + { + "epoch": 21.33226837060703, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 13354 + }, + { + "epoch": 21.333865814696484, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 13355 + }, + { + "epoch": 21.335463258785943, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 13356 + }, + { + "epoch": 21.337060702875398, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 13357 + }, + { + "epoch": 21.338658146964857, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 13358 + }, + { + "epoch": 21.340255591054312, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 13359 + }, + { + "epoch": 21.34185303514377, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 13360 + }, + { + "epoch": 21.343450479233226, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 13361 + }, + { + "epoch": 21.345047923322685, + "grad_norm": 0.04931640625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 13362 + }, + { + "epoch": 21.34664536741214, + "grad_norm": 0.034423828125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 13363 + }, + { + "epoch": 21.3482428115016, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 13364 + }, + { + "epoch": 21.349840255591054, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 13365 + }, + { + "epoch": 21.35143769968051, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 13366 + }, + { + "epoch": 21.35303514376997, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13367 + }, + { + "epoch": 21.354632587859424, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 13368 + }, + { + "epoch": 21.356230031948883, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 13369 + }, + { + "epoch": 21.357827476038338, + "grad_norm": 0.035888671875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 13370 + }, + { + "epoch": 21.359424920127797, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 13371 + }, + { + "epoch": 21.361022364217252, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 13372 + }, + { + "epoch": 21.36261980830671, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 13373 + }, + { + "epoch": 21.364217252396166, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 13374 + }, + { + "epoch": 21.36581469648562, + "grad_norm": 0.041259765625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 13375 + }, + { + "epoch": 21.36741214057508, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 13376 + }, + { + "epoch": 21.369009584664536, + "grad_norm": 0.0419921875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 13377 + }, + { + "epoch": 21.370607028753994, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 13378 + }, + { + "epoch": 21.37220447284345, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 13379 + }, + { + "epoch": 21.37380191693291, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13380 + }, + { + "epoch": 21.375399361022364, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 13381 + }, + { + "epoch": 21.376996805111823, + "grad_norm": 0.0498046875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 13382 + }, + { + "epoch": 21.378594249201278, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 13383 + }, + { + "epoch": 21.380191693290733, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 13384 + }, + { + "epoch": 21.381789137380192, + "grad_norm": 0.04736328125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 13385 + }, + { + "epoch": 21.383386581469647, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 13386 + }, + { + "epoch": 21.384984025559106, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13387 + }, + { + "epoch": 21.38658146964856, + "grad_norm": 0.05322265625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 13388 + }, + { + "epoch": 21.38817891373802, + "grad_norm": 0.048095703125, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 13389 + }, + { + "epoch": 21.389776357827476, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 13390 + }, + { + "epoch": 21.391373801916934, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 13391 + }, + { + "epoch": 21.39297124600639, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 13392 + }, + { + "epoch": 21.394568690095845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 13393 + }, + { + "epoch": 21.396166134185304, + "grad_norm": 0.04638671875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 13394 + }, + { + "epoch": 21.39776357827476, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 13395 + }, + { + "epoch": 21.399361022364218, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 13396 + }, + { + "epoch": 21.400958466453673, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 13397 + }, + { + "epoch": 21.402555910543132, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 13398 + }, + { + "epoch": 21.404153354632587, + "grad_norm": 0.046630859375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 13399 + }, + { + "epoch": 21.405750798722046, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 13400 + }, + { + "epoch": 21.4073482428115, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 13401 + }, + { + "epoch": 21.408945686900957, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13402 + }, + { + "epoch": 21.410543130990416, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 13403 + }, + { + "epoch": 21.41214057507987, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 13404 + }, + { + "epoch": 21.41373801916933, + "grad_norm": 0.040283203125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 13405 + }, + { + "epoch": 21.415335463258785, + "grad_norm": 0.050537109375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 13406 + }, + { + "epoch": 21.416932907348244, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 13407 + }, + { + "epoch": 21.4185303514377, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 13408 + }, + { + "epoch": 21.420127795527158, + "grad_norm": 0.045166015625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 13409 + }, + { + "epoch": 21.421725239616613, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 13410 + }, + { + "epoch": 21.423322683706072, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 13411 + }, + { + "epoch": 21.424920127795527, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 13412 + }, + { + "epoch": 21.426517571884983, + "grad_norm": 0.03955078125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 13413 + }, + { + "epoch": 21.42811501597444, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 13414 + }, + { + "epoch": 21.429712460063897, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 13415 + }, + { + "epoch": 21.431309904153355, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 13416 + }, + { + "epoch": 21.43290734824281, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 13417 + }, + { + "epoch": 21.43450479233227, + "grad_norm": 0.04345703125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 13418 + }, + { + "epoch": 21.436102236421725, + "grad_norm": 0.042724609375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13419 + }, + { + "epoch": 21.437699680511184, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 13420 + }, + { + "epoch": 21.43929712460064, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13421 + }, + { + "epoch": 21.440894568690094, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 13422 + }, + { + "epoch": 21.442492012779553, + "grad_norm": 0.0517578125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 13423 + }, + { + "epoch": 21.44408945686901, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 13424 + }, + { + "epoch": 21.445686900958467, + "grad_norm": 0.04052734375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 13425 + }, + { + "epoch": 21.447284345047922, + "grad_norm": 0.037109375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 13426 + }, + { + "epoch": 21.44888178913738, + "grad_norm": 0.036865234375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13427 + }, + { + "epoch": 21.450479233226837, + "grad_norm": 0.0400390625, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 13428 + }, + { + "epoch": 21.452076677316295, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 13429 + }, + { + "epoch": 21.45367412140575, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13430 + }, + { + "epoch": 21.455271565495206, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 13431 + }, + { + "epoch": 21.456869009584665, + "grad_norm": 0.040771484375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 13432 + }, + { + "epoch": 21.45846645367412, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 13433 + }, + { + "epoch": 21.46006389776358, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 13434 + }, + { + "epoch": 21.461661341853034, + "grad_norm": 0.03662109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 13435 + }, + { + "epoch": 21.463258785942493, + "grad_norm": 0.03759765625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 13436 + }, + { + "epoch": 21.46485623003195, + "grad_norm": 0.046142578125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 13437 + }, + { + "epoch": 21.466453674121407, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 13438 + }, + { + "epoch": 21.468051118210862, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 13439 + }, + { + "epoch": 21.46964856230032, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 13440 + }, + { + "epoch": 21.471246006389777, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 13441 + }, + { + "epoch": 21.472843450479232, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 13442 + }, + { + "epoch": 21.47444089456869, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 13443 + }, + { + "epoch": 21.476038338658146, + "grad_norm": 0.044677734375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 13444 + }, + { + "epoch": 21.477635782747605, + "grad_norm": 0.044921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 13445 + }, + { + "epoch": 21.47923322683706, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 13446 + }, + { + "epoch": 21.48083067092652, + "grad_norm": 0.04248046875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 13447 + }, + { + "epoch": 21.482428115015974, + "grad_norm": 0.038818359375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 13448 + }, + { + "epoch": 21.484025559105433, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 13449 + }, + { + "epoch": 21.48562300319489, + "grad_norm": 0.0478515625, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 13450 + }, + { + "epoch": 21.487220447284344, + "grad_norm": 0.037353515625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 13451 + }, + { + "epoch": 21.488817891373802, + "grad_norm": 0.05029296875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 13452 + }, + { + "epoch": 21.490415335463258, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 13453 + }, + { + "epoch": 21.492012779552716, + "grad_norm": 0.03564453125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 13454 + }, + { + "epoch": 21.49361022364217, + "grad_norm": 0.032958984375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 13455 + }, + { + "epoch": 21.49520766773163, + "grad_norm": 0.035400390625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 13456 + }, + { + "epoch": 21.496805111821086, + "grad_norm": 0.052978515625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 13457 + }, + { + "epoch": 21.498402555910545, + "grad_norm": 0.044189453125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 13458 + }, + { + "epoch": 21.5, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 13459 + }, + { + "epoch": 21.501597444089455, + "grad_norm": 0.04833984375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 13460 + }, + { + "epoch": 21.503194888178914, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 13461 + }, + { + "epoch": 21.50479233226837, + "grad_norm": 0.045654296875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 13462 + }, + { + "epoch": 21.50638977635783, + "grad_norm": 0.04296875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 13463 + }, + { + "epoch": 21.507987220447284, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 13464 + }, + { + "epoch": 21.509584664536742, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 13465 + }, + { + "epoch": 21.511182108626198, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 13466 + }, + { + "epoch": 21.512779552715656, + "grad_norm": 0.04150390625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 13467 + }, + { + "epoch": 21.51437699680511, + "grad_norm": 0.03515625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 13468 + }, + { + "epoch": 21.515974440894567, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13469 + }, + { + "epoch": 21.517571884984026, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 13470 + }, + { + "epoch": 21.51916932907348, + "grad_norm": 0.034912109375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 13471 + }, + { + "epoch": 21.52076677316294, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 13472 + }, + { + "epoch": 21.522364217252395, + "grad_norm": 0.049072265625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 13473 + }, + { + "epoch": 21.523961661341854, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 13474 + }, + { + "epoch": 21.52555910543131, + "grad_norm": 0.03369140625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 13475 + }, + { + "epoch": 21.527156549520768, + "grad_norm": 0.041015625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13476 + }, + { + "epoch": 21.528753993610223, + "grad_norm": 0.03857421875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 13477 + }, + { + "epoch": 21.53035143769968, + "grad_norm": 0.0341796875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 13478 + }, + { + "epoch": 21.531948881789138, + "grad_norm": 0.0390625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 13479 + }, + { + "epoch": 21.533546325878593, + "grad_norm": 0.0380859375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 13480 + }, + { + "epoch": 21.53514376996805, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 13481 + }, + { + "epoch": 21.536741214057507, + "grad_norm": 0.046875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 13482 + }, + { + "epoch": 21.538338658146966, + "grad_norm": 0.041748046875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 13483 + }, + { + "epoch": 21.53993610223642, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 13484 + }, + { + "epoch": 21.54153354632588, + "grad_norm": 0.047119140625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 13485 + }, + { + "epoch": 21.543130990415335, + "grad_norm": 0.0439453125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 13486 + }, + { + "epoch": 21.544728434504794, + "grad_norm": 0.04541015625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 13487 + }, + { + "epoch": 21.54632587859425, + "grad_norm": 0.048828125, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 13488 + }, + { + "epoch": 21.547923322683705, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 13489 + }, + { + "epoch": 21.549520766773163, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 13490 + }, + { + "epoch": 21.55111821086262, + "grad_norm": 0.05126953125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 13491 + }, + { + "epoch": 21.552715654952078, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 13492 + }, + { + "epoch": 21.554313099041533, + "grad_norm": 0.049560546875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 13493 + }, + { + "epoch": 21.55591054313099, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 13494 + }, + { + "epoch": 21.557507987220447, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 13495 + }, + { + "epoch": 21.559105431309906, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 13496 + }, + { + "epoch": 21.56070287539936, + "grad_norm": 0.050048828125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 13497 + }, + { + "epoch": 21.562300319488816, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 13498 + }, + { + "epoch": 21.563897763578275, + "grad_norm": 0.036376953125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 13499 + }, + { + "epoch": 21.56549520766773, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 13500 + } + ], + "logging_steps": 1.0, + "max_steps": 40064, + "num_input_tokens_seen": 0, + "num_train_epochs": 64, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 9.329580805984354e+18, + "train_batch_size": 256, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-13500/training_args.bin b/checkpoint-13500/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2da74eebfd8cfaccb04c98604a0d0d791ad85304 --- /dev/null +++ b/checkpoint-13500/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b2beee2fd37f6d9ac0c495475a630572dcdba494ec909a154ca692ab1b73833 +size 6008