add MT model
Browse files- all_results.json +22 -0
- config.json +38 -0
- eval_results.json +10 -0
- generated_predictions.txt +0 -0
- predict_results.json +9 -0
- pytorch_model.bin +3 -0
- sentencepiece.bpe.model +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- train_results.json +8 -0
- trainer_state.json +2197 -0
- training_args.bin +3 -0
- vocab.json +0 -0
all_results.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_bleu": 7.6292,
|
| 4 |
+
"eval_gen_len": 70.4128,
|
| 5 |
+
"eval_loss": 2.1154050827026367,
|
| 6 |
+
"eval_runtime": 724.3496,
|
| 7 |
+
"eval_samples": 1284,
|
| 8 |
+
"eval_samples_per_second": 1.773,
|
| 9 |
+
"eval_steps_per_second": 0.178,
|
| 10 |
+
"predict_bleu": 9.2277,
|
| 11 |
+
"predict_gen_len": 71.9527,
|
| 12 |
+
"predict_loss": 2.0788121223449707,
|
| 13 |
+
"predict_runtime": 851.6036,
|
| 14 |
+
"predict_samples": 1500,
|
| 15 |
+
"predict_samples_per_second": 1.761,
|
| 16 |
+
"predict_steps_per_second": 0.176,
|
| 17 |
+
"train_loss": 0.7403765432559616,
|
| 18 |
+
"train_runtime": 55016.1932,
|
| 19 |
+
"train_samples": 604315,
|
| 20 |
+
"train_samples_per_second": 32.953,
|
| 21 |
+
"train_steps_per_second": 3.295
|
| 22 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/m2m100_418M",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "relu",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"M2M100ForConditionalGeneration"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"d_model": 1024,
|
| 11 |
+
"decoder_attention_heads": 16,
|
| 12 |
+
"decoder_ffn_dim": 4096,
|
| 13 |
+
"decoder_layerdrop": 0.05,
|
| 14 |
+
"decoder_layers": 12,
|
| 15 |
+
"decoder_start_token_id": 2,
|
| 16 |
+
"dropout": 0.1,
|
| 17 |
+
"early_stopping": true,
|
| 18 |
+
"encoder_attention_heads": 16,
|
| 19 |
+
"encoder_ffn_dim": 4096,
|
| 20 |
+
"encoder_layerdrop": 0.05,
|
| 21 |
+
"encoder_layers": 12,
|
| 22 |
+
"eos_token_id": 2,
|
| 23 |
+
"forced_bos_token_id": 128088,
|
| 24 |
+
"gradient_checkpointing": false,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"max_length": 200,
|
| 28 |
+
"max_position_embeddings": 1024,
|
| 29 |
+
"model_type": "m2m_100",
|
| 30 |
+
"num_beams": 5,
|
| 31 |
+
"num_hidden_layers": 12,
|
| 32 |
+
"pad_token_id": 1,
|
| 33 |
+
"scale_embedding": true,
|
| 34 |
+
"torch_dtype": "float32",
|
| 35 |
+
"transformers_version": "4.10.3",
|
| 36 |
+
"use_cache": true,
|
| 37 |
+
"vocab_size": 128112
|
| 38 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_bleu": 7.6292,
|
| 4 |
+
"eval_gen_len": 70.4128,
|
| 5 |
+
"eval_loss": 2.1154050827026367,
|
| 6 |
+
"eval_runtime": 724.3496,
|
| 7 |
+
"eval_samples": 1284,
|
| 8 |
+
"eval_samples_per_second": 1.773,
|
| 9 |
+
"eval_steps_per_second": 0.178
|
| 10 |
+
}
|
generated_predictions.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
predict_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"predict_bleu": 9.2277,
|
| 3 |
+
"predict_gen_len": 71.9527,
|
| 4 |
+
"predict_loss": 2.0788121223449707,
|
| 5 |
+
"predict_runtime": 851.6036,
|
| 6 |
+
"predict_samples": 1500,
|
| 7 |
+
"predict_samples_per_second": 1.761,
|
| 8 |
+
"predict_steps_per_second": 0.176
|
| 9 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:106ffb20108c8ebcc5ce37e43f1ef3102df737c591cc1d9b31c0c8c1027befe1
|
| 3 |
+
size 1935792071
|
sentencepiece.bpe.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8f7c76ed2a5e0822be39f0a4f95a55eb19c78f4593ce609e2edbc2aea4d380a
|
| 3 |
+
size 2423393
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "additional_special_tokens": ["__af__", "__am__", "__ar__", "__ast__", "__az__", "__ba__", "__be__", "__bg__", "__bn__", "__br__", "__bs__", "__ca__", "__ceb__", "__cs__", "__cy__", "__da__", "__de__", "__el__", "__en__", "__es__", "__et__", "__fa__", "__ff__", "__fi__", "__fr__", "__fy__", "__ga__", "__gd__", "__gl__", "__gu__", "__ha__", "__he__", "__hi__", "__hr__", "__ht__", "__hu__", "__hy__", "__id__", "__ig__", "__ilo__", "__is__", "__it__", "__ja__", "__jv__", "__ka__", "__kk__", "__km__", "__kn__", "__ko__", "__lb__", "__lg__", "__ln__", "__lo__", "__lt__", "__lv__", "__mg__", "__mk__", "__ml__", "__mn__", "__mr__", "__ms__", "__my__", "__ne__", "__nl__", "__no__", "__ns__", "__oc__", "__or__", "__pa__", "__pl__", "__ps__", "__pt__", "__ro__", "__ru__", "__sd__", "__si__", "__sk__", "__sl__", "__so__", "__sq__", "__sr__", "__ss__", "__su__", "__sv__", "__sw__", "__ta__", "__th__", "__tl__", "__tn__", "__tr__", "__uk__", "__ur__", "__uz__", "__vi__", "__wo__", "__xh__", "__yi__", "__yo__", "__zh__", "__zu__"]}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"src_lang": null, "tgt_lang": null, "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "sp_model_kwargs": {}, "special_tokens_map_file": "m2m_100_1.2B_v2/special_tokens_map.json", "tokenizer_file": null, "name_or_path": "facebook/m2m100_418M", "model_max_length": 1024, "additional_special_tokens": ["__af__", "__am__", "__ar__", "__ast__", "__az__", "__ba__", "__be__", "__bg__", "__bn__", "__br__", "__bs__", "__ca__", "__ceb__", "__cs__", "__cy__", "__da__", "__de__", "__el__", "__en__", "__es__", "__et__", "__fa__", "__ff__", "__fi__", "__fr__", "__fy__", "__ga__", "__gd__", "__gl__", "__gu__", "__ha__", "__he__", "__hi__", "__hr__", "__ht__", "__hu__", "__hy__", "__id__", "__ig__", "__ilo__", "__is__", "__it__", "__ja__", "__jv__", "__ka__", "__kk__", "__km__", "__kn__", "__ko__", "__lb__", "__lg__", "__ln__", "__lo__", "__lt__", "__lv__", "__mg__", "__mk__", "__ml__", "__mn__", "__mr__", "__ms__", "__my__", "__ne__", "__nl__", "__no__", "__ns__", "__oc__", "__or__", "__pa__", "__pl__", "__ps__", "__pt__", "__ro__", "__ru__", "__sd__", "__si__", "__sk__", "__sl__", "__so__", "__sq__", "__sr__", "__ss__", "__su__", "__sv__", "__sw__", "__ta__", "__th__", "__tl__", "__tn__", "__tr__", "__uk__", "__ur__", "__uz__", "__vi__", "__wo__", "__xh__", "__yi__", "__yo__", "__zh__", "__zu__"], "tokenizer_class": "M2M100Tokenizer"}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 0.7403765432559616,
|
| 4 |
+
"train_runtime": 55016.1932,
|
| 5 |
+
"train_samples": 604315,
|
| 6 |
+
"train_samples_per_second": 32.953,
|
| 7 |
+
"train_steps_per_second": 3.295
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2197 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 181296,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 4.986210396258053e-05,
|
| 13 |
+
"loss": 2.8171,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 4.9724207925161064e-05,
|
| 19 |
+
"loss": 2.0072,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.02,
|
| 24 |
+
"learning_rate": 4.95863118877416e-05,
|
| 25 |
+
"loss": 1.7607,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.03,
|
| 30 |
+
"learning_rate": 4.9448415850322125e-05,
|
| 31 |
+
"loss": 1.6134,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.04,
|
| 36 |
+
"learning_rate": 4.931051981290266e-05,
|
| 37 |
+
"loss": 1.536,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.05,
|
| 42 |
+
"learning_rate": 4.917262377548319e-05,
|
| 43 |
+
"loss": 1.4503,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06,
|
| 48 |
+
"learning_rate": 4.9034727738063724e-05,
|
| 49 |
+
"loss": 1.4208,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.07,
|
| 54 |
+
"learning_rate": 4.8896831700644255e-05,
|
| 55 |
+
"loss": 1.3709,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.07,
|
| 60 |
+
"learning_rate": 4.8758935663224786e-05,
|
| 61 |
+
"loss": 1.3472,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.08,
|
| 66 |
+
"learning_rate": 4.862103962580532e-05,
|
| 67 |
+
"loss": 1.3117,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.09,
|
| 72 |
+
"learning_rate": 4.848314358838585e-05,
|
| 73 |
+
"loss": 1.2615,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.1,
|
| 78 |
+
"learning_rate": 4.834524755096638e-05,
|
| 79 |
+
"loss": 1.2461,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.11,
|
| 84 |
+
"learning_rate": 4.820735151354691e-05,
|
| 85 |
+
"loss": 1.2356,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12,
|
| 90 |
+
"learning_rate": 4.806945547612744e-05,
|
| 91 |
+
"loss": 1.218,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.12,
|
| 96 |
+
"learning_rate": 4.793155943870797e-05,
|
| 97 |
+
"loss": 1.2006,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.13,
|
| 102 |
+
"learning_rate": 4.77936634012885e-05,
|
| 103 |
+
"loss": 1.172,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.14,
|
| 108 |
+
"learning_rate": 4.765576736386904e-05,
|
| 109 |
+
"loss": 1.1617,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.15,
|
| 114 |
+
"learning_rate": 4.751787132644956e-05,
|
| 115 |
+
"loss": 1.1568,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.16,
|
| 120 |
+
"learning_rate": 4.73799752890301e-05,
|
| 121 |
+
"loss": 1.1374,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.17,
|
| 126 |
+
"learning_rate": 4.7242079251610624e-05,
|
| 127 |
+
"loss": 1.1234,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.17,
|
| 132 |
+
"learning_rate": 4.710418321419116e-05,
|
| 133 |
+
"loss": 1.1088,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.18,
|
| 138 |
+
"learning_rate": 4.696628717677169e-05,
|
| 139 |
+
"loss": 1.0915,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.19,
|
| 144 |
+
"learning_rate": 4.682839113935222e-05,
|
| 145 |
+
"loss": 1.0999,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.2,
|
| 150 |
+
"learning_rate": 4.6690495101932754e-05,
|
| 151 |
+
"loss": 1.0745,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.21,
|
| 156 |
+
"learning_rate": 4.6552599064513285e-05,
|
| 157 |
+
"loss": 1.0702,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.22,
|
| 162 |
+
"learning_rate": 4.6414703027093815e-05,
|
| 163 |
+
"loss": 1.0665,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.22,
|
| 168 |
+
"learning_rate": 4.6276806989674346e-05,
|
| 169 |
+
"loss": 1.0546,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.23,
|
| 174 |
+
"learning_rate": 4.613891095225488e-05,
|
| 175 |
+
"loss": 1.0611,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.24,
|
| 180 |
+
"learning_rate": 4.600101491483541e-05,
|
| 181 |
+
"loss": 1.0603,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.25,
|
| 186 |
+
"learning_rate": 4.586311887741594e-05,
|
| 187 |
+
"loss": 1.0343,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.26,
|
| 192 |
+
"learning_rate": 4.5725222839996476e-05,
|
| 193 |
+
"loss": 1.0327,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.26,
|
| 198 |
+
"learning_rate": 4.5587326802577e-05,
|
| 199 |
+
"loss": 1.0223,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.27,
|
| 204 |
+
"learning_rate": 4.544943076515754e-05,
|
| 205 |
+
"loss": 1.0156,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.28,
|
| 210 |
+
"learning_rate": 4.531153472773806e-05,
|
| 211 |
+
"loss": 1.0112,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.29,
|
| 216 |
+
"learning_rate": 4.51736386903186e-05,
|
| 217 |
+
"loss": 1.005,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.3,
|
| 222 |
+
"learning_rate": 4.503574265289913e-05,
|
| 223 |
+
"loss": 0.9836,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.31,
|
| 228 |
+
"learning_rate": 4.489784661547966e-05,
|
| 229 |
+
"loss": 0.9939,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.31,
|
| 234 |
+
"learning_rate": 4.475995057806019e-05,
|
| 235 |
+
"loss": 0.9986,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.32,
|
| 240 |
+
"learning_rate": 4.462205454064072e-05,
|
| 241 |
+
"loss": 0.9881,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.33,
|
| 246 |
+
"learning_rate": 4.448415850322125e-05,
|
| 247 |
+
"loss": 0.9716,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.34,
|
| 252 |
+
"learning_rate": 4.434626246580178e-05,
|
| 253 |
+
"loss": 0.9729,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.35,
|
| 258 |
+
"learning_rate": 4.4208366428382314e-05,
|
| 259 |
+
"loss": 0.9703,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.36,
|
| 264 |
+
"learning_rate": 4.4070470390962845e-05,
|
| 265 |
+
"loss": 0.9553,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.36,
|
| 270 |
+
"learning_rate": 4.3932574353543376e-05,
|
| 271 |
+
"loss": 0.9543,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.37,
|
| 276 |
+
"learning_rate": 4.379467831612391e-05,
|
| 277 |
+
"loss": 0.9639,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.38,
|
| 282 |
+
"learning_rate": 4.365678227870444e-05,
|
| 283 |
+
"loss": 0.9482,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.39,
|
| 288 |
+
"learning_rate": 4.3518886241284975e-05,
|
| 289 |
+
"loss": 0.9403,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.4,
|
| 294 |
+
"learning_rate": 4.33809902038655e-05,
|
| 295 |
+
"loss": 0.9338,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.41,
|
| 300 |
+
"learning_rate": 4.3243094166446036e-05,
|
| 301 |
+
"loss": 0.9471,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.41,
|
| 306 |
+
"learning_rate": 4.310519812902657e-05,
|
| 307 |
+
"loss": 0.9351,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.42,
|
| 312 |
+
"learning_rate": 4.29673020916071e-05,
|
| 313 |
+
"loss": 0.9187,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.43,
|
| 318 |
+
"learning_rate": 4.282940605418763e-05,
|
| 319 |
+
"loss": 0.9264,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.44,
|
| 324 |
+
"learning_rate": 4.269151001676816e-05,
|
| 325 |
+
"loss": 0.9177,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.45,
|
| 330 |
+
"learning_rate": 4.255361397934869e-05,
|
| 331 |
+
"loss": 0.9299,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.46,
|
| 336 |
+
"learning_rate": 4.241571794192922e-05,
|
| 337 |
+
"loss": 0.9184,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.46,
|
| 342 |
+
"learning_rate": 4.227782190450976e-05,
|
| 343 |
+
"loss": 0.9216,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.47,
|
| 348 |
+
"learning_rate": 4.213992586709028e-05,
|
| 349 |
+
"loss": 0.9113,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.48,
|
| 354 |
+
"learning_rate": 4.200202982967082e-05,
|
| 355 |
+
"loss": 0.8962,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.49,
|
| 360 |
+
"learning_rate": 4.186413379225135e-05,
|
| 361 |
+
"loss": 0.9086,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.5,
|
| 366 |
+
"learning_rate": 4.172623775483188e-05,
|
| 367 |
+
"loss": 0.9235,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.5,
|
| 372 |
+
"learning_rate": 4.158834171741241e-05,
|
| 373 |
+
"loss": 0.9018,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.51,
|
| 378 |
+
"learning_rate": 4.145044567999294e-05,
|
| 379 |
+
"loss": 0.903,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.52,
|
| 384 |
+
"learning_rate": 4.131254964257347e-05,
|
| 385 |
+
"loss": 0.8975,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.53,
|
| 390 |
+
"learning_rate": 4.1174653605154004e-05,
|
| 391 |
+
"loss": 0.8943,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.54,
|
| 396 |
+
"learning_rate": 4.1036757567734535e-05,
|
| 397 |
+
"loss": 0.8875,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.55,
|
| 402 |
+
"learning_rate": 4.089886153031507e-05,
|
| 403 |
+
"loss": 0.8919,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.55,
|
| 408 |
+
"learning_rate": 4.0760965492895596e-05,
|
| 409 |
+
"loss": 0.879,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.56,
|
| 414 |
+
"learning_rate": 4.0623069455476134e-05,
|
| 415 |
+
"loss": 0.8875,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.57,
|
| 420 |
+
"learning_rate": 4.048517341805666e-05,
|
| 421 |
+
"loss": 0.8633,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.58,
|
| 426 |
+
"learning_rate": 4.0347277380637195e-05,
|
| 427 |
+
"loss": 0.8776,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.59,
|
| 432 |
+
"learning_rate": 4.020938134321772e-05,
|
| 433 |
+
"loss": 0.8702,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.6,
|
| 438 |
+
"learning_rate": 4.007148530579826e-05,
|
| 439 |
+
"loss": 0.866,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.6,
|
| 444 |
+
"learning_rate": 3.993358926837879e-05,
|
| 445 |
+
"loss": 0.8634,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.61,
|
| 450 |
+
"learning_rate": 3.979569323095932e-05,
|
| 451 |
+
"loss": 0.883,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.62,
|
| 456 |
+
"learning_rate": 3.965779719353985e-05,
|
| 457 |
+
"loss": 0.8519,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.63,
|
| 462 |
+
"learning_rate": 3.951990115612038e-05,
|
| 463 |
+
"loss": 0.8643,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.64,
|
| 468 |
+
"learning_rate": 3.938200511870091e-05,
|
| 469 |
+
"loss": 0.8454,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.65,
|
| 474 |
+
"learning_rate": 3.924410908128144e-05,
|
| 475 |
+
"loss": 0.8666,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.65,
|
| 480 |
+
"learning_rate": 3.910621304386197e-05,
|
| 481 |
+
"loss": 0.8514,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.66,
|
| 486 |
+
"learning_rate": 3.896831700644251e-05,
|
| 487 |
+
"loss": 0.8454,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.67,
|
| 492 |
+
"learning_rate": 3.8830420969023034e-05,
|
| 493 |
+
"loss": 0.8615,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.68,
|
| 498 |
+
"learning_rate": 3.869252493160357e-05,
|
| 499 |
+
"loss": 0.8516,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.69,
|
| 504 |
+
"learning_rate": 3.8554628894184095e-05,
|
| 505 |
+
"loss": 0.8502,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.69,
|
| 510 |
+
"learning_rate": 3.841673285676463e-05,
|
| 511 |
+
"loss": 0.8455,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.7,
|
| 516 |
+
"learning_rate": 3.827883681934516e-05,
|
| 517 |
+
"loss": 0.8323,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.71,
|
| 522 |
+
"learning_rate": 3.8140940781925694e-05,
|
| 523 |
+
"loss": 0.8408,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.72,
|
| 528 |
+
"learning_rate": 3.8003044744506225e-05,
|
| 529 |
+
"loss": 0.8478,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.73,
|
| 534 |
+
"learning_rate": 3.7865148707086756e-05,
|
| 535 |
+
"loss": 0.8428,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.74,
|
| 540 |
+
"learning_rate": 3.7727252669667286e-05,
|
| 541 |
+
"loss": 0.838,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.74,
|
| 546 |
+
"learning_rate": 3.758935663224782e-05,
|
| 547 |
+
"loss": 0.843,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.75,
|
| 552 |
+
"learning_rate": 3.745146059482835e-05,
|
| 553 |
+
"loss": 0.8388,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.76,
|
| 558 |
+
"learning_rate": 3.731356455740888e-05,
|
| 559 |
+
"loss": 0.8256,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.77,
|
| 564 |
+
"learning_rate": 3.717566851998941e-05,
|
| 565 |
+
"loss": 0.8393,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.78,
|
| 570 |
+
"learning_rate": 3.703777248256994e-05,
|
| 571 |
+
"loss": 0.8295,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.79,
|
| 576 |
+
"learning_rate": 3.689987644515047e-05,
|
| 577 |
+
"loss": 0.8194,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.79,
|
| 582 |
+
"learning_rate": 3.676198040773101e-05,
|
| 583 |
+
"loss": 0.8262,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.8,
|
| 588 |
+
"learning_rate": 3.662408437031153e-05,
|
| 589 |
+
"loss": 0.8368,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.81,
|
| 594 |
+
"learning_rate": 3.648618833289207e-05,
|
| 595 |
+
"loss": 0.8226,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.82,
|
| 600 |
+
"learning_rate": 3.6348292295472594e-05,
|
| 601 |
+
"loss": 0.8207,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.83,
|
| 606 |
+
"learning_rate": 3.621039625805313e-05,
|
| 607 |
+
"loss": 0.8131,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.84,
|
| 612 |
+
"learning_rate": 3.607250022063366e-05,
|
| 613 |
+
"loss": 0.8253,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.84,
|
| 618 |
+
"learning_rate": 3.593460418321419e-05,
|
| 619 |
+
"loss": 0.8082,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.85,
|
| 624 |
+
"learning_rate": 3.5796708145794724e-05,
|
| 625 |
+
"loss": 0.8191,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.86,
|
| 630 |
+
"learning_rate": 3.5658812108375254e-05,
|
| 631 |
+
"loss": 0.8167,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.87,
|
| 636 |
+
"learning_rate": 3.5520916070955785e-05,
|
| 637 |
+
"loss": 0.7956,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.88,
|
| 642 |
+
"learning_rate": 3.5383020033536316e-05,
|
| 643 |
+
"loss": 0.804,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.89,
|
| 648 |
+
"learning_rate": 3.524512399611685e-05,
|
| 649 |
+
"loss": 0.7958,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.89,
|
| 654 |
+
"learning_rate": 3.510722795869738e-05,
|
| 655 |
+
"loss": 0.8147,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.9,
|
| 660 |
+
"learning_rate": 3.496933192127791e-05,
|
| 661 |
+
"loss": 0.8106,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.91,
|
| 666 |
+
"learning_rate": 3.4831435883858446e-05,
|
| 667 |
+
"loss": 0.8114,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.92,
|
| 672 |
+
"learning_rate": 3.4693539846438976e-05,
|
| 673 |
+
"loss": 0.8079,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.93,
|
| 678 |
+
"learning_rate": 3.455564380901951e-05,
|
| 679 |
+
"loss": 0.8037,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.93,
|
| 684 |
+
"learning_rate": 3.441774777160004e-05,
|
| 685 |
+
"loss": 0.7862,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.94,
|
| 690 |
+
"learning_rate": 3.427985173418057e-05,
|
| 691 |
+
"loss": 0.8013,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.95,
|
| 696 |
+
"learning_rate": 3.41419556967611e-05,
|
| 697 |
+
"loss": 0.7932,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.96,
|
| 702 |
+
"learning_rate": 3.400405965934163e-05,
|
| 703 |
+
"loss": 0.7843,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.97,
|
| 708 |
+
"learning_rate": 3.386616362192217e-05,
|
| 709 |
+
"loss": 0.7878,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.98,
|
| 714 |
+
"learning_rate": 3.372826758450269e-05,
|
| 715 |
+
"loss": 0.794,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.98,
|
| 720 |
+
"learning_rate": 3.359037154708323e-05,
|
| 721 |
+
"loss": 0.7875,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.99,
|
| 726 |
+
"learning_rate": 3.345247550966375e-05,
|
| 727 |
+
"loss": 0.7934,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.0,
|
| 732 |
+
"learning_rate": 3.331457947224429e-05,
|
| 733 |
+
"loss": 0.7802,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.01,
|
| 738 |
+
"learning_rate": 3.3176683434824815e-05,
|
| 739 |
+
"loss": 0.7117,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.02,
|
| 744 |
+
"learning_rate": 3.303878739740535e-05,
|
| 745 |
+
"loss": 0.7064,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.03,
|
| 750 |
+
"learning_rate": 3.290089135998588e-05,
|
| 751 |
+
"loss": 0.7223,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.03,
|
| 756 |
+
"learning_rate": 3.2762995322566414e-05,
|
| 757 |
+
"loss": 0.7159,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.04,
|
| 762 |
+
"learning_rate": 3.2625099285146944e-05,
|
| 763 |
+
"loss": 0.7211,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.05,
|
| 768 |
+
"learning_rate": 3.2487203247727475e-05,
|
| 769 |
+
"loss": 0.7173,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.06,
|
| 774 |
+
"learning_rate": 3.2349307210308006e-05,
|
| 775 |
+
"loss": 0.7173,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.07,
|
| 780 |
+
"learning_rate": 3.221141117288854e-05,
|
| 781 |
+
"loss": 0.7046,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.08,
|
| 786 |
+
"learning_rate": 3.207351513546907e-05,
|
| 787 |
+
"loss": 0.7106,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.08,
|
| 792 |
+
"learning_rate": 3.1935619098049605e-05,
|
| 793 |
+
"loss": 0.7023,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.09,
|
| 798 |
+
"learning_rate": 3.179772306063013e-05,
|
| 799 |
+
"loss": 0.7028,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.1,
|
| 804 |
+
"learning_rate": 3.1659827023210666e-05,
|
| 805 |
+
"loss": 0.7283,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.11,
|
| 810 |
+
"learning_rate": 3.152193098579119e-05,
|
| 811 |
+
"loss": 0.7025,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.12,
|
| 816 |
+
"learning_rate": 3.138403494837173e-05,
|
| 817 |
+
"loss": 0.716,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.13,
|
| 822 |
+
"learning_rate": 3.124613891095225e-05,
|
| 823 |
+
"loss": 0.7063,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.13,
|
| 828 |
+
"learning_rate": 3.110824287353279e-05,
|
| 829 |
+
"loss": 0.7079,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.14,
|
| 834 |
+
"learning_rate": 3.097034683611332e-05,
|
| 835 |
+
"loss": 0.7025,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.15,
|
| 840 |
+
"learning_rate": 3.083245079869385e-05,
|
| 841 |
+
"loss": 0.7029,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.16,
|
| 846 |
+
"learning_rate": 3.069455476127438e-05,
|
| 847 |
+
"loss": 0.7003,
|
| 848 |
+
"step": 70000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.17,
|
| 852 |
+
"learning_rate": 3.055665872385491e-05,
|
| 853 |
+
"loss": 0.7013,
|
| 854 |
+
"step": 70500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.17,
|
| 858 |
+
"learning_rate": 3.0418762686435443e-05,
|
| 859 |
+
"loss": 0.7154,
|
| 860 |
+
"step": 71000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.18,
|
| 864 |
+
"learning_rate": 3.0280866649015977e-05,
|
| 865 |
+
"loss": 0.7059,
|
| 866 |
+
"step": 71500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.19,
|
| 870 |
+
"learning_rate": 3.0142970611596505e-05,
|
| 871 |
+
"loss": 0.7047,
|
| 872 |
+
"step": 72000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.2,
|
| 876 |
+
"learning_rate": 3.000507457417704e-05,
|
| 877 |
+
"loss": 0.7047,
|
| 878 |
+
"step": 72500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.21,
|
| 882 |
+
"learning_rate": 2.9867178536757566e-05,
|
| 883 |
+
"loss": 0.6962,
|
| 884 |
+
"step": 73000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.22,
|
| 888 |
+
"learning_rate": 2.97292824993381e-05,
|
| 889 |
+
"loss": 0.708,
|
| 890 |
+
"step": 73500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.22,
|
| 894 |
+
"learning_rate": 2.959138646191863e-05,
|
| 895 |
+
"loss": 0.6973,
|
| 896 |
+
"step": 74000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.23,
|
| 900 |
+
"learning_rate": 2.9453490424499165e-05,
|
| 901 |
+
"loss": 0.7049,
|
| 902 |
+
"step": 74500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.24,
|
| 906 |
+
"learning_rate": 2.9315594387079692e-05,
|
| 907 |
+
"loss": 0.7052,
|
| 908 |
+
"step": 75000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.25,
|
| 912 |
+
"learning_rate": 2.9177698349660227e-05,
|
| 913 |
+
"loss": 0.6949,
|
| 914 |
+
"step": 75500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.26,
|
| 918 |
+
"learning_rate": 2.9039802312240754e-05,
|
| 919 |
+
"loss": 0.7017,
|
| 920 |
+
"step": 76000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.27,
|
| 924 |
+
"learning_rate": 2.8901906274821288e-05,
|
| 925 |
+
"loss": 0.702,
|
| 926 |
+
"step": 76500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.27,
|
| 930 |
+
"learning_rate": 2.876401023740182e-05,
|
| 931 |
+
"loss": 0.7055,
|
| 932 |
+
"step": 77000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 1.28,
|
| 936 |
+
"learning_rate": 2.862611419998235e-05,
|
| 937 |
+
"loss": 0.701,
|
| 938 |
+
"step": 77500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 1.29,
|
| 942 |
+
"learning_rate": 2.848821816256288e-05,
|
| 943 |
+
"loss": 0.7046,
|
| 944 |
+
"step": 78000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.3,
|
| 948 |
+
"learning_rate": 2.8350322125143414e-05,
|
| 949 |
+
"loss": 0.6899,
|
| 950 |
+
"step": 78500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.31,
|
| 954 |
+
"learning_rate": 2.8212426087723942e-05,
|
| 955 |
+
"loss": 0.7043,
|
| 956 |
+
"step": 79000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.32,
|
| 960 |
+
"learning_rate": 2.8074530050304476e-05,
|
| 961 |
+
"loss": 0.6923,
|
| 962 |
+
"step": 79500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.32,
|
| 966 |
+
"learning_rate": 2.7936634012885003e-05,
|
| 967 |
+
"loss": 0.7001,
|
| 968 |
+
"step": 80000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.33,
|
| 972 |
+
"learning_rate": 2.7798737975465537e-05,
|
| 973 |
+
"loss": 0.6947,
|
| 974 |
+
"step": 80500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.34,
|
| 978 |
+
"learning_rate": 2.7660841938046068e-05,
|
| 979 |
+
"loss": 0.6971,
|
| 980 |
+
"step": 81000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1.35,
|
| 984 |
+
"learning_rate": 2.7522945900626602e-05,
|
| 985 |
+
"loss": 0.6927,
|
| 986 |
+
"step": 81500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.36,
|
| 990 |
+
"learning_rate": 2.738504986320713e-05,
|
| 991 |
+
"loss": 0.6936,
|
| 992 |
+
"step": 82000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.37,
|
| 996 |
+
"learning_rate": 2.7247153825787664e-05,
|
| 997 |
+
"loss": 0.6914,
|
| 998 |
+
"step": 82500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.37,
|
| 1002 |
+
"learning_rate": 2.7109257788368198e-05,
|
| 1003 |
+
"loss": 0.6899,
|
| 1004 |
+
"step": 83000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.38,
|
| 1008 |
+
"learning_rate": 2.6971361750948725e-05,
|
| 1009 |
+
"loss": 0.6901,
|
| 1010 |
+
"step": 83500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.39,
|
| 1014 |
+
"learning_rate": 2.683346571352926e-05,
|
| 1015 |
+
"loss": 0.6909,
|
| 1016 |
+
"step": 84000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.4,
|
| 1020 |
+
"learning_rate": 2.6695569676109787e-05,
|
| 1021 |
+
"loss": 0.69,
|
| 1022 |
+
"step": 84500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.41,
|
| 1026 |
+
"learning_rate": 2.655767363869032e-05,
|
| 1027 |
+
"loss": 0.6937,
|
| 1028 |
+
"step": 85000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.41,
|
| 1032 |
+
"learning_rate": 2.6419777601270852e-05,
|
| 1033 |
+
"loss": 0.7011,
|
| 1034 |
+
"step": 85500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.42,
|
| 1038 |
+
"learning_rate": 2.6281881563851386e-05,
|
| 1039 |
+
"loss": 0.6917,
|
| 1040 |
+
"step": 86000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.43,
|
| 1044 |
+
"learning_rate": 2.6143985526431913e-05,
|
| 1045 |
+
"loss": 0.682,
|
| 1046 |
+
"step": 86500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.44,
|
| 1050 |
+
"learning_rate": 2.6006089489012447e-05,
|
| 1051 |
+
"loss": 0.6827,
|
| 1052 |
+
"step": 87000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.45,
|
| 1056 |
+
"learning_rate": 2.5868193451592975e-05,
|
| 1057 |
+
"loss": 0.6902,
|
| 1058 |
+
"step": 87500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1.46,
|
| 1062 |
+
"learning_rate": 2.573029741417351e-05,
|
| 1063 |
+
"loss": 0.6854,
|
| 1064 |
+
"step": 88000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 1.46,
|
| 1068 |
+
"learning_rate": 2.559240137675404e-05,
|
| 1069 |
+
"loss": 0.6903,
|
| 1070 |
+
"step": 88500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.47,
|
| 1074 |
+
"learning_rate": 2.5454505339334574e-05,
|
| 1075 |
+
"loss": 0.6813,
|
| 1076 |
+
"step": 89000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.48,
|
| 1080 |
+
"learning_rate": 2.53166093019151e-05,
|
| 1081 |
+
"loss": 0.6778,
|
| 1082 |
+
"step": 89500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.49,
|
| 1086 |
+
"learning_rate": 2.5178713264495635e-05,
|
| 1087 |
+
"loss": 0.6808,
|
| 1088 |
+
"step": 90000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.5,
|
| 1092 |
+
"learning_rate": 2.5040817227076163e-05,
|
| 1093 |
+
"loss": 0.6912,
|
| 1094 |
+
"step": 90500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.51,
|
| 1098 |
+
"learning_rate": 2.4902921189656693e-05,
|
| 1099 |
+
"loss": 0.684,
|
| 1100 |
+
"step": 91000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 1.51,
|
| 1104 |
+
"learning_rate": 2.4765025152237224e-05,
|
| 1105 |
+
"loss": 0.6771,
|
| 1106 |
+
"step": 91500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 1.52,
|
| 1110 |
+
"learning_rate": 2.4627129114817758e-05,
|
| 1111 |
+
"loss": 0.6751,
|
| 1112 |
+
"step": 92000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.53,
|
| 1116 |
+
"learning_rate": 2.448923307739829e-05,
|
| 1117 |
+
"loss": 0.6796,
|
| 1118 |
+
"step": 92500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.54,
|
| 1122 |
+
"learning_rate": 2.435133703997882e-05,
|
| 1123 |
+
"loss": 0.6758,
|
| 1124 |
+
"step": 93000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.55,
|
| 1128 |
+
"learning_rate": 2.421344100255935e-05,
|
| 1129 |
+
"loss": 0.6761,
|
| 1130 |
+
"step": 93500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.56,
|
| 1134 |
+
"learning_rate": 2.407554496513988e-05,
|
| 1135 |
+
"loss": 0.6821,
|
| 1136 |
+
"step": 94000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.56,
|
| 1140 |
+
"learning_rate": 2.3937648927720412e-05,
|
| 1141 |
+
"loss": 0.6711,
|
| 1142 |
+
"step": 94500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 1.57,
|
| 1146 |
+
"learning_rate": 2.3799752890300943e-05,
|
| 1147 |
+
"loss": 0.6862,
|
| 1148 |
+
"step": 95000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 1.58,
|
| 1152 |
+
"learning_rate": 2.3661856852881477e-05,
|
| 1153 |
+
"loss": 0.6844,
|
| 1154 |
+
"step": 95500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.59,
|
| 1158 |
+
"learning_rate": 2.3523960815462008e-05,
|
| 1159 |
+
"loss": 0.6841,
|
| 1160 |
+
"step": 96000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 1.6,
|
| 1164 |
+
"learning_rate": 2.338606477804254e-05,
|
| 1165 |
+
"loss": 0.6683,
|
| 1166 |
+
"step": 96500
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.61,
|
| 1170 |
+
"learning_rate": 2.3248168740623072e-05,
|
| 1171 |
+
"loss": 0.6622,
|
| 1172 |
+
"step": 97000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.61,
|
| 1176 |
+
"learning_rate": 2.3110272703203603e-05,
|
| 1177 |
+
"loss": 0.6732,
|
| 1178 |
+
"step": 97500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.62,
|
| 1182 |
+
"learning_rate": 2.2972376665784134e-05,
|
| 1183 |
+
"loss": 0.6839,
|
| 1184 |
+
"step": 98000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 1.63,
|
| 1188 |
+
"learning_rate": 2.2834480628364665e-05,
|
| 1189 |
+
"loss": 0.6773,
|
| 1190 |
+
"step": 98500
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 1.64,
|
| 1194 |
+
"learning_rate": 2.2696584590945195e-05,
|
| 1195 |
+
"loss": 0.6622,
|
| 1196 |
+
"step": 99000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.65,
|
| 1200 |
+
"learning_rate": 2.255868855352573e-05,
|
| 1201 |
+
"loss": 0.6745,
|
| 1202 |
+
"step": 99500
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 1.65,
|
| 1206 |
+
"learning_rate": 2.242079251610626e-05,
|
| 1207 |
+
"loss": 0.6776,
|
| 1208 |
+
"step": 100000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.66,
|
| 1212 |
+
"learning_rate": 2.228289647868679e-05,
|
| 1213 |
+
"loss": 0.6706,
|
| 1214 |
+
"step": 100500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.67,
|
| 1218 |
+
"learning_rate": 2.2145000441267322e-05,
|
| 1219 |
+
"loss": 0.6631,
|
| 1220 |
+
"step": 101000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.68,
|
| 1224 |
+
"learning_rate": 2.2007104403847853e-05,
|
| 1225 |
+
"loss": 0.6722,
|
| 1226 |
+
"step": 101500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.69,
|
| 1230 |
+
"learning_rate": 2.1869208366428383e-05,
|
| 1231 |
+
"loss": 0.6729,
|
| 1232 |
+
"step": 102000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 1.7,
|
| 1236 |
+
"learning_rate": 2.1731312329008914e-05,
|
| 1237 |
+
"loss": 0.6899,
|
| 1238 |
+
"step": 102500
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.7,
|
| 1242 |
+
"learning_rate": 2.1593416291589448e-05,
|
| 1243 |
+
"loss": 0.6523,
|
| 1244 |
+
"step": 103000
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 1.71,
|
| 1248 |
+
"learning_rate": 2.145552025416998e-05,
|
| 1249 |
+
"loss": 0.6654,
|
| 1250 |
+
"step": 103500
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.72,
|
| 1254 |
+
"learning_rate": 2.131762421675051e-05,
|
| 1255 |
+
"loss": 0.666,
|
| 1256 |
+
"step": 104000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.73,
|
| 1260 |
+
"learning_rate": 2.117972817933104e-05,
|
| 1261 |
+
"loss": 0.67,
|
| 1262 |
+
"step": 104500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.74,
|
| 1266 |
+
"learning_rate": 2.104183214191157e-05,
|
| 1267 |
+
"loss": 0.6795,
|
| 1268 |
+
"step": 105000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 1.75,
|
| 1272 |
+
"learning_rate": 2.0903936104492102e-05,
|
| 1273 |
+
"loss": 0.6666,
|
| 1274 |
+
"step": 105500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 1.75,
|
| 1278 |
+
"learning_rate": 2.0766040067072633e-05,
|
| 1279 |
+
"loss": 0.6661,
|
| 1280 |
+
"step": 106000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 1.76,
|
| 1284 |
+
"learning_rate": 2.0628144029653167e-05,
|
| 1285 |
+
"loss": 0.6701,
|
| 1286 |
+
"step": 106500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 1.77,
|
| 1290 |
+
"learning_rate": 2.0490247992233698e-05,
|
| 1291 |
+
"loss": 0.6628,
|
| 1292 |
+
"step": 107000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.78,
|
| 1296 |
+
"learning_rate": 2.035235195481423e-05,
|
| 1297 |
+
"loss": 0.6629,
|
| 1298 |
+
"step": 107500
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.79,
|
| 1302 |
+
"learning_rate": 2.021445591739476e-05,
|
| 1303 |
+
"loss": 0.6626,
|
| 1304 |
+
"step": 108000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.8,
|
| 1308 |
+
"learning_rate": 2.007655987997529e-05,
|
| 1309 |
+
"loss": 0.657,
|
| 1310 |
+
"step": 108500
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 1.8,
|
| 1314 |
+
"learning_rate": 1.993866384255582e-05,
|
| 1315 |
+
"loss": 0.6594,
|
| 1316 |
+
"step": 109000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 1.81,
|
| 1320 |
+
"learning_rate": 1.980076780513635e-05,
|
| 1321 |
+
"loss": 0.6637,
|
| 1322 |
+
"step": 109500
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 1.82,
|
| 1326 |
+
"learning_rate": 1.9662871767716885e-05,
|
| 1327 |
+
"loss": 0.6474,
|
| 1328 |
+
"step": 110000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 1.83,
|
| 1332 |
+
"learning_rate": 1.9524975730297416e-05,
|
| 1333 |
+
"loss": 0.6607,
|
| 1334 |
+
"step": 110500
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 1.84,
|
| 1338 |
+
"learning_rate": 1.9387079692877947e-05,
|
| 1339 |
+
"loss": 0.6569,
|
| 1340 |
+
"step": 111000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.85,
|
| 1344 |
+
"learning_rate": 1.9249183655458478e-05,
|
| 1345 |
+
"loss": 0.6644,
|
| 1346 |
+
"step": 111500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.85,
|
| 1350 |
+
"learning_rate": 1.911128761803901e-05,
|
| 1351 |
+
"loss": 0.6611,
|
| 1352 |
+
"step": 112000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 1.86,
|
| 1356 |
+
"learning_rate": 1.897339158061954e-05,
|
| 1357 |
+
"loss": 0.6597,
|
| 1358 |
+
"step": 112500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 1.87,
|
| 1362 |
+
"learning_rate": 1.883549554320007e-05,
|
| 1363 |
+
"loss": 0.665,
|
| 1364 |
+
"step": 113000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 1.88,
|
| 1368 |
+
"learning_rate": 1.8697599505780604e-05,
|
| 1369 |
+
"loss": 0.6434,
|
| 1370 |
+
"step": 113500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 1.89,
|
| 1374 |
+
"learning_rate": 1.8559703468361135e-05,
|
| 1375 |
+
"loss": 0.656,
|
| 1376 |
+
"step": 114000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.89,
|
| 1380 |
+
"learning_rate": 1.8421807430941666e-05,
|
| 1381 |
+
"loss": 0.6572,
|
| 1382 |
+
"step": 114500
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.9,
|
| 1386 |
+
"learning_rate": 1.8283911393522196e-05,
|
| 1387 |
+
"loss": 0.6424,
|
| 1388 |
+
"step": 115000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.91,
|
| 1392 |
+
"learning_rate": 1.8146015356102727e-05,
|
| 1393 |
+
"loss": 0.6396,
|
| 1394 |
+
"step": 115500
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 1.92,
|
| 1398 |
+
"learning_rate": 1.8008119318683258e-05,
|
| 1399 |
+
"loss": 0.644,
|
| 1400 |
+
"step": 116000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 1.93,
|
| 1404 |
+
"learning_rate": 1.787022328126379e-05,
|
| 1405 |
+
"loss": 0.6509,
|
| 1406 |
+
"step": 116500
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.94,
|
| 1410 |
+
"learning_rate": 1.7732327243844323e-05,
|
| 1411 |
+
"loss": 0.6475,
|
| 1412 |
+
"step": 117000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 1.94,
|
| 1416 |
+
"learning_rate": 1.7594431206424853e-05,
|
| 1417 |
+
"loss": 0.6514,
|
| 1418 |
+
"step": 117500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.95,
|
| 1422 |
+
"learning_rate": 1.7456535169005384e-05,
|
| 1423 |
+
"loss": 0.6419,
|
| 1424 |
+
"step": 118000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.96,
|
| 1428 |
+
"learning_rate": 1.7318639131585915e-05,
|
| 1429 |
+
"loss": 0.6415,
|
| 1430 |
+
"step": 118500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.97,
|
| 1434 |
+
"learning_rate": 1.7180743094166446e-05,
|
| 1435 |
+
"loss": 0.655,
|
| 1436 |
+
"step": 119000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 1.98,
|
| 1440 |
+
"learning_rate": 1.7042847056746976e-05,
|
| 1441 |
+
"loss": 0.6384,
|
| 1442 |
+
"step": 119500
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 1.99,
|
| 1446 |
+
"learning_rate": 1.6904951019327507e-05,
|
| 1447 |
+
"loss": 0.6439,
|
| 1448 |
+
"step": 120000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1.99,
|
| 1452 |
+
"learning_rate": 1.676705498190804e-05,
|
| 1453 |
+
"loss": 0.6484,
|
| 1454 |
+
"step": 120500
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 2.0,
|
| 1458 |
+
"learning_rate": 1.6629158944488572e-05,
|
| 1459 |
+
"loss": 0.6205,
|
| 1460 |
+
"step": 121000
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 2.01,
|
| 1464 |
+
"learning_rate": 1.6491262907069103e-05,
|
| 1465 |
+
"loss": 0.5833,
|
| 1466 |
+
"step": 121500
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.02,
|
| 1470 |
+
"learning_rate": 1.6353366869649634e-05,
|
| 1471 |
+
"loss": 0.5594,
|
| 1472 |
+
"step": 122000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.03,
|
| 1476 |
+
"learning_rate": 1.6215470832230164e-05,
|
| 1477 |
+
"loss": 0.5598,
|
| 1478 |
+
"step": 122500
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 2.04,
|
| 1482 |
+
"learning_rate": 1.6077574794810695e-05,
|
| 1483 |
+
"loss": 0.5795,
|
| 1484 |
+
"step": 123000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 2.04,
|
| 1488 |
+
"learning_rate": 1.5939678757391226e-05,
|
| 1489 |
+
"loss": 0.5747,
|
| 1490 |
+
"step": 123500
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 2.05,
|
| 1494 |
+
"learning_rate": 1.580178271997176e-05,
|
| 1495 |
+
"loss": 0.5647,
|
| 1496 |
+
"step": 124000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 2.06,
|
| 1500 |
+
"learning_rate": 1.566388668255229e-05,
|
| 1501 |
+
"loss": 0.5687,
|
| 1502 |
+
"step": 124500
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 2.07,
|
| 1506 |
+
"learning_rate": 1.5525990645132825e-05,
|
| 1507 |
+
"loss": 0.5627,
|
| 1508 |
+
"step": 125000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.08,
|
| 1512 |
+
"learning_rate": 1.5388094607713356e-05,
|
| 1513 |
+
"loss": 0.5649,
|
| 1514 |
+
"step": 125500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.08,
|
| 1518 |
+
"learning_rate": 1.5250198570293886e-05,
|
| 1519 |
+
"loss": 0.5696,
|
| 1520 |
+
"step": 126000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 2.09,
|
| 1524 |
+
"learning_rate": 1.5112302532874417e-05,
|
| 1525 |
+
"loss": 0.5581,
|
| 1526 |
+
"step": 126500
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 2.1,
|
| 1530 |
+
"learning_rate": 1.4974406495454948e-05,
|
| 1531 |
+
"loss": 0.5764,
|
| 1532 |
+
"step": 127000
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 2.11,
|
| 1536 |
+
"learning_rate": 1.483651045803548e-05,
|
| 1537 |
+
"loss": 0.5727,
|
| 1538 |
+
"step": 127500
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 2.12,
|
| 1542 |
+
"learning_rate": 1.4698614420616011e-05,
|
| 1543 |
+
"loss": 0.571,
|
| 1544 |
+
"step": 128000
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 2.13,
|
| 1548 |
+
"learning_rate": 1.4560718383196542e-05,
|
| 1549 |
+
"loss": 0.5697,
|
| 1550 |
+
"step": 128500
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.13,
|
| 1554 |
+
"learning_rate": 1.4422822345777073e-05,
|
| 1555 |
+
"loss": 0.5622,
|
| 1556 |
+
"step": 129000
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.14,
|
| 1560 |
+
"learning_rate": 1.4284926308357605e-05,
|
| 1561 |
+
"loss": 0.5686,
|
| 1562 |
+
"step": 129500
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 2.15,
|
| 1566 |
+
"learning_rate": 1.4147030270938136e-05,
|
| 1567 |
+
"loss": 0.5634,
|
| 1568 |
+
"step": 130000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 2.16,
|
| 1572 |
+
"learning_rate": 1.4009134233518666e-05,
|
| 1573 |
+
"loss": 0.5667,
|
| 1574 |
+
"step": 130500
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 2.17,
|
| 1578 |
+
"learning_rate": 1.3871238196099199e-05,
|
| 1579 |
+
"loss": 0.5726,
|
| 1580 |
+
"step": 131000
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 2.18,
|
| 1584 |
+
"learning_rate": 1.373334215867973e-05,
|
| 1585 |
+
"loss": 0.5589,
|
| 1586 |
+
"step": 131500
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 2.18,
|
| 1590 |
+
"learning_rate": 1.359544612126026e-05,
|
| 1591 |
+
"loss": 0.5704,
|
| 1592 |
+
"step": 132000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.19,
|
| 1596 |
+
"learning_rate": 1.3457550083840791e-05,
|
| 1597 |
+
"loss": 0.5682,
|
| 1598 |
+
"step": 132500
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.2,
|
| 1602 |
+
"learning_rate": 1.3319654046421324e-05,
|
| 1603 |
+
"loss": 0.5632,
|
| 1604 |
+
"step": 133000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 2.21,
|
| 1608 |
+
"learning_rate": 1.3181758009001854e-05,
|
| 1609 |
+
"loss": 0.5751,
|
| 1610 |
+
"step": 133500
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 2.22,
|
| 1614 |
+
"learning_rate": 1.3043861971582385e-05,
|
| 1615 |
+
"loss": 0.5628,
|
| 1616 |
+
"step": 134000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 2.23,
|
| 1620 |
+
"learning_rate": 1.2905965934162918e-05,
|
| 1621 |
+
"loss": 0.5564,
|
| 1622 |
+
"step": 134500
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 2.23,
|
| 1626 |
+
"learning_rate": 1.2768069896743448e-05,
|
| 1627 |
+
"loss": 0.5705,
|
| 1628 |
+
"step": 135000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 2.24,
|
| 1632 |
+
"learning_rate": 1.2630173859323979e-05,
|
| 1633 |
+
"loss": 0.5648,
|
| 1634 |
+
"step": 135500
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.25,
|
| 1638 |
+
"learning_rate": 1.249227782190451e-05,
|
| 1639 |
+
"loss": 0.5666,
|
| 1640 |
+
"step": 136000
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.26,
|
| 1644 |
+
"learning_rate": 1.2354381784485042e-05,
|
| 1645 |
+
"loss": 0.5694,
|
| 1646 |
+
"step": 136500
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 2.27,
|
| 1650 |
+
"learning_rate": 1.2216485747065573e-05,
|
| 1651 |
+
"loss": 0.5712,
|
| 1652 |
+
"step": 137000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 2.28,
|
| 1656 |
+
"learning_rate": 1.2078589709646104e-05,
|
| 1657 |
+
"loss": 0.5756,
|
| 1658 |
+
"step": 137500
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 2.28,
|
| 1662 |
+
"learning_rate": 1.1940693672226636e-05,
|
| 1663 |
+
"loss": 0.5672,
|
| 1664 |
+
"step": 138000
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 2.29,
|
| 1668 |
+
"learning_rate": 1.1802797634807167e-05,
|
| 1669 |
+
"loss": 0.5718,
|
| 1670 |
+
"step": 138500
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 2.3,
|
| 1674 |
+
"learning_rate": 1.1664901597387698e-05,
|
| 1675 |
+
"loss": 0.5551,
|
| 1676 |
+
"step": 139000
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.31,
|
| 1680 |
+
"learning_rate": 1.1527005559968228e-05,
|
| 1681 |
+
"loss": 0.569,
|
| 1682 |
+
"step": 139500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.32,
|
| 1686 |
+
"learning_rate": 1.138910952254876e-05,
|
| 1687 |
+
"loss": 0.5707,
|
| 1688 |
+
"step": 140000
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 2.32,
|
| 1692 |
+
"learning_rate": 1.1251213485129292e-05,
|
| 1693 |
+
"loss": 0.5576,
|
| 1694 |
+
"step": 140500
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 2.33,
|
| 1698 |
+
"learning_rate": 1.1113317447709822e-05,
|
| 1699 |
+
"loss": 0.5793,
|
| 1700 |
+
"step": 141000
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 2.34,
|
| 1704 |
+
"learning_rate": 1.0975421410290355e-05,
|
| 1705 |
+
"loss": 0.5605,
|
| 1706 |
+
"step": 141500
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 2.35,
|
| 1710 |
+
"learning_rate": 1.0837525372870886e-05,
|
| 1711 |
+
"loss": 0.5587,
|
| 1712 |
+
"step": 142000
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 2.36,
|
| 1716 |
+
"learning_rate": 1.0699629335451416e-05,
|
| 1717 |
+
"loss": 0.5652,
|
| 1718 |
+
"step": 142500
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.37,
|
| 1722 |
+
"learning_rate": 1.0561733298031947e-05,
|
| 1723 |
+
"loss": 0.5503,
|
| 1724 |
+
"step": 143000
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.37,
|
| 1728 |
+
"learning_rate": 1.042383726061248e-05,
|
| 1729 |
+
"loss": 0.5543,
|
| 1730 |
+
"step": 143500
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 2.38,
|
| 1734 |
+
"learning_rate": 1.028594122319301e-05,
|
| 1735 |
+
"loss": 0.5702,
|
| 1736 |
+
"step": 144000
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 2.39,
|
| 1740 |
+
"learning_rate": 1.0148045185773541e-05,
|
| 1741 |
+
"loss": 0.5558,
|
| 1742 |
+
"step": 144500
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 2.4,
|
| 1746 |
+
"learning_rate": 1.0010149148354073e-05,
|
| 1747 |
+
"loss": 0.5527,
|
| 1748 |
+
"step": 145000
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 2.41,
|
| 1752 |
+
"learning_rate": 9.872253110934606e-06,
|
| 1753 |
+
"loss": 0.5619,
|
| 1754 |
+
"step": 145500
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 2.42,
|
| 1758 |
+
"learning_rate": 9.734357073515137e-06,
|
| 1759 |
+
"loss": 0.5557,
|
| 1760 |
+
"step": 146000
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.42,
|
| 1764 |
+
"learning_rate": 9.596461036095667e-06,
|
| 1765 |
+
"loss": 0.5621,
|
| 1766 |
+
"step": 146500
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.43,
|
| 1770 |
+
"learning_rate": 9.4585649986762e-06,
|
| 1771 |
+
"loss": 0.5554,
|
| 1772 |
+
"step": 147000
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 2.44,
|
| 1776 |
+
"learning_rate": 9.32066896125673e-06,
|
| 1777 |
+
"loss": 0.5612,
|
| 1778 |
+
"step": 147500
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 2.45,
|
| 1782 |
+
"learning_rate": 9.182772923837261e-06,
|
| 1783 |
+
"loss": 0.5614,
|
| 1784 |
+
"step": 148000
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 2.46,
|
| 1788 |
+
"learning_rate": 9.044876886417792e-06,
|
| 1789 |
+
"loss": 0.5645,
|
| 1790 |
+
"step": 148500
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 2.47,
|
| 1794 |
+
"learning_rate": 8.906980848998324e-06,
|
| 1795 |
+
"loss": 0.5613,
|
| 1796 |
+
"step": 149000
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 2.47,
|
| 1800 |
+
"learning_rate": 8.769084811578855e-06,
|
| 1801 |
+
"loss": 0.5572,
|
| 1802 |
+
"step": 149500
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.48,
|
| 1806 |
+
"learning_rate": 8.631188774159386e-06,
|
| 1807 |
+
"loss": 0.5582,
|
| 1808 |
+
"step": 150000
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.49,
|
| 1812 |
+
"learning_rate": 8.493292736739918e-06,
|
| 1813 |
+
"loss": 0.5516,
|
| 1814 |
+
"step": 150500
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 2.5,
|
| 1818 |
+
"learning_rate": 8.355396699320449e-06,
|
| 1819 |
+
"loss": 0.5477,
|
| 1820 |
+
"step": 151000
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 2.51,
|
| 1824 |
+
"learning_rate": 8.21750066190098e-06,
|
| 1825 |
+
"loss": 0.5676,
|
| 1826 |
+
"step": 151500
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 2.52,
|
| 1830 |
+
"learning_rate": 8.07960462448151e-06,
|
| 1831 |
+
"loss": 0.5593,
|
| 1832 |
+
"step": 152000
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 2.52,
|
| 1836 |
+
"learning_rate": 7.941708587062043e-06,
|
| 1837 |
+
"loss": 0.565,
|
| 1838 |
+
"step": 152500
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 2.53,
|
| 1842 |
+
"learning_rate": 7.803812549642574e-06,
|
| 1843 |
+
"loss": 0.5651,
|
| 1844 |
+
"step": 153000
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.54,
|
| 1848 |
+
"learning_rate": 7.665916512223105e-06,
|
| 1849 |
+
"loss": 0.5492,
|
| 1850 |
+
"step": 153500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.55,
|
| 1854 |
+
"learning_rate": 7.528020474803636e-06,
|
| 1855 |
+
"loss": 0.5516,
|
| 1856 |
+
"step": 154000
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 2.56,
|
| 1860 |
+
"learning_rate": 7.390124437384168e-06,
|
| 1861 |
+
"loss": 0.5579,
|
| 1862 |
+
"step": 154500
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 2.56,
|
| 1866 |
+
"learning_rate": 7.2522283999646985e-06,
|
| 1867 |
+
"loss": 0.5543,
|
| 1868 |
+
"step": 155000
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 2.57,
|
| 1872 |
+
"learning_rate": 7.11433236254523e-06,
|
| 1873 |
+
"loss": 0.5518,
|
| 1874 |
+
"step": 155500
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 2.58,
|
| 1878 |
+
"learning_rate": 6.976436325125761e-06,
|
| 1879 |
+
"loss": 0.5446,
|
| 1880 |
+
"step": 156000
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 2.59,
|
| 1884 |
+
"learning_rate": 6.8385402877062925e-06,
|
| 1885 |
+
"loss": 0.5551,
|
| 1886 |
+
"step": 156500
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.6,
|
| 1890 |
+
"learning_rate": 6.700644250286823e-06,
|
| 1891 |
+
"loss": 0.5474,
|
| 1892 |
+
"step": 157000
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.61,
|
| 1896 |
+
"learning_rate": 6.562748212867355e-06,
|
| 1897 |
+
"loss": 0.55,
|
| 1898 |
+
"step": 157500
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 2.61,
|
| 1902 |
+
"learning_rate": 6.424852175447886e-06,
|
| 1903 |
+
"loss": 0.5522,
|
| 1904 |
+
"step": 158000
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 2.62,
|
| 1908 |
+
"learning_rate": 6.286956138028417e-06,
|
| 1909 |
+
"loss": 0.5631,
|
| 1910 |
+
"step": 158500
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 2.63,
|
| 1914 |
+
"learning_rate": 6.14906010060895e-06,
|
| 1915 |
+
"loss": 0.5435,
|
| 1916 |
+
"step": 159000
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 2.64,
|
| 1920 |
+
"learning_rate": 6.01116406318948e-06,
|
| 1921 |
+
"loss": 0.5492,
|
| 1922 |
+
"step": 159500
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 2.65,
|
| 1926 |
+
"learning_rate": 5.873268025770012e-06,
|
| 1927 |
+
"loss": 0.5424,
|
| 1928 |
+
"step": 160000
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.66,
|
| 1932 |
+
"learning_rate": 5.735371988350543e-06,
|
| 1933 |
+
"loss": 0.5493,
|
| 1934 |
+
"step": 160500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.66,
|
| 1938 |
+
"learning_rate": 5.597475950931074e-06,
|
| 1939 |
+
"loss": 0.5415,
|
| 1940 |
+
"step": 161000
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 2.67,
|
| 1944 |
+
"learning_rate": 5.459579913511605e-06,
|
| 1945 |
+
"loss": 0.5574,
|
| 1946 |
+
"step": 161500
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 2.68,
|
| 1950 |
+
"learning_rate": 5.321683876092137e-06,
|
| 1951 |
+
"loss": 0.5408,
|
| 1952 |
+
"step": 162000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 2.69,
|
| 1956 |
+
"learning_rate": 5.183787838672668e-06,
|
| 1957 |
+
"loss": 0.5486,
|
| 1958 |
+
"step": 162500
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 2.7,
|
| 1962 |
+
"learning_rate": 5.0458918012532e-06,
|
| 1963 |
+
"loss": 0.5493,
|
| 1964 |
+
"step": 163000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 2.71,
|
| 1968 |
+
"learning_rate": 4.9079957638337306e-06,
|
| 1969 |
+
"loss": 0.542,
|
| 1970 |
+
"step": 163500
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.71,
|
| 1974 |
+
"learning_rate": 4.770099726414262e-06,
|
| 1975 |
+
"loss": 0.5444,
|
| 1976 |
+
"step": 164000
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.72,
|
| 1980 |
+
"learning_rate": 4.632203688994794e-06,
|
| 1981 |
+
"loss": 0.5445,
|
| 1982 |
+
"step": 164500
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 2.73,
|
| 1986 |
+
"learning_rate": 4.4943076515753245e-06,
|
| 1987 |
+
"loss": 0.5449,
|
| 1988 |
+
"step": 165000
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 2.74,
|
| 1992 |
+
"learning_rate": 4.356411614155856e-06,
|
| 1993 |
+
"loss": 0.5479,
|
| 1994 |
+
"step": 165500
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 2.75,
|
| 1998 |
+
"learning_rate": 4.218515576736387e-06,
|
| 1999 |
+
"loss": 0.5568,
|
| 2000 |
+
"step": 166000
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 2.76,
|
| 2004 |
+
"learning_rate": 4.0806195393169184e-06,
|
| 2005 |
+
"loss": 0.5467,
|
| 2006 |
+
"step": 166500
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.76,
|
| 2010 |
+
"learning_rate": 3.942723501897449e-06,
|
| 2011 |
+
"loss": 0.5464,
|
| 2012 |
+
"step": 167000
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.77,
|
| 2016 |
+
"learning_rate": 3.8048274644779808e-06,
|
| 2017 |
+
"loss": 0.544,
|
| 2018 |
+
"step": 167500
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.78,
|
| 2022 |
+
"learning_rate": 3.666931427058512e-06,
|
| 2023 |
+
"loss": 0.5423,
|
| 2024 |
+
"step": 168000
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 2.79,
|
| 2028 |
+
"learning_rate": 3.529035389639043e-06,
|
| 2029 |
+
"loss": 0.5406,
|
| 2030 |
+
"step": 168500
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 2.8,
|
| 2034 |
+
"learning_rate": 3.3911393522195743e-06,
|
| 2035 |
+
"loss": 0.5413,
|
| 2036 |
+
"step": 169000
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 2.8,
|
| 2040 |
+
"learning_rate": 3.2532433148001063e-06,
|
| 2041 |
+
"loss": 0.543,
|
| 2042 |
+
"step": 169500
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 2.81,
|
| 2046 |
+
"learning_rate": 3.1153472773806375e-06,
|
| 2047 |
+
"loss": 0.538,
|
| 2048 |
+
"step": 170000
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 2.82,
|
| 2052 |
+
"learning_rate": 2.9774512399611686e-06,
|
| 2053 |
+
"loss": 0.5364,
|
| 2054 |
+
"step": 170500
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.83,
|
| 2058 |
+
"learning_rate": 2.8395552025417e-06,
|
| 2059 |
+
"loss": 0.5379,
|
| 2060 |
+
"step": 171000
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.84,
|
| 2064 |
+
"learning_rate": 2.7016591651222314e-06,
|
| 2065 |
+
"loss": 0.5479,
|
| 2066 |
+
"step": 171500
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 2.85,
|
| 2070 |
+
"learning_rate": 2.5637631277027626e-06,
|
| 2071 |
+
"loss": 0.5438,
|
| 2072 |
+
"step": 172000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 2.85,
|
| 2076 |
+
"learning_rate": 2.4258670902832938e-06,
|
| 2077 |
+
"loss": 0.5335,
|
| 2078 |
+
"step": 172500
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 2.86,
|
| 2082 |
+
"learning_rate": 2.287971052863825e-06,
|
| 2083 |
+
"loss": 0.5426,
|
| 2084 |
+
"step": 173000
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 2.87,
|
| 2088 |
+
"learning_rate": 2.150075015444356e-06,
|
| 2089 |
+
"loss": 0.5287,
|
| 2090 |
+
"step": 173500
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 2.88,
|
| 2094 |
+
"learning_rate": 2.0121789780248873e-06,
|
| 2095 |
+
"loss": 0.5423,
|
| 2096 |
+
"step": 174000
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.89,
|
| 2100 |
+
"learning_rate": 1.874282940605419e-06,
|
| 2101 |
+
"loss": 0.5384,
|
| 2102 |
+
"step": 174500
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.9,
|
| 2106 |
+
"learning_rate": 1.7363869031859502e-06,
|
| 2107 |
+
"loss": 0.5377,
|
| 2108 |
+
"step": 175000
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 2.9,
|
| 2112 |
+
"learning_rate": 1.5984908657664814e-06,
|
| 2113 |
+
"loss": 0.532,
|
| 2114 |
+
"step": 175500
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 2.91,
|
| 2118 |
+
"learning_rate": 1.4605948283470128e-06,
|
| 2119 |
+
"loss": 0.5463,
|
| 2120 |
+
"step": 176000
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 2.92,
|
| 2124 |
+
"learning_rate": 1.322698790927544e-06,
|
| 2125 |
+
"loss": 0.5345,
|
| 2126 |
+
"step": 176500
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 2.93,
|
| 2130 |
+
"learning_rate": 1.1848027535080751e-06,
|
| 2131 |
+
"loss": 0.5428,
|
| 2132 |
+
"step": 177000
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 2.94,
|
| 2136 |
+
"learning_rate": 1.0469067160886065e-06,
|
| 2137 |
+
"loss": 0.5413,
|
| 2138 |
+
"step": 177500
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.95,
|
| 2142 |
+
"learning_rate": 9.090106786691378e-07,
|
| 2143 |
+
"loss": 0.538,
|
| 2144 |
+
"step": 178000
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.95,
|
| 2148 |
+
"learning_rate": 7.711146412496691e-07,
|
| 2149 |
+
"loss": 0.536,
|
| 2150 |
+
"step": 178500
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 2.96,
|
| 2154 |
+
"learning_rate": 6.332186038302004e-07,
|
| 2155 |
+
"loss": 0.5484,
|
| 2156 |
+
"step": 179000
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 2.97,
|
| 2160 |
+
"learning_rate": 4.953225664107316e-07,
|
| 2161 |
+
"loss": 0.548,
|
| 2162 |
+
"step": 179500
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 2.98,
|
| 2166 |
+
"learning_rate": 3.574265289912629e-07,
|
| 2167 |
+
"loss": 0.5335,
|
| 2168 |
+
"step": 180000
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 2.99,
|
| 2172 |
+
"learning_rate": 2.195304915717942e-07,
|
| 2173 |
+
"loss": 0.5375,
|
| 2174 |
+
"step": 180500
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 3.0,
|
| 2178 |
+
"learning_rate": 8.163445415232549e-08,
|
| 2179 |
+
"loss": 0.5442,
|
| 2180 |
+
"step": 181000
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 3.0,
|
| 2184 |
+
"step": 181296,
|
| 2185 |
+
"total_flos": 2.0627822520336384e+17,
|
| 2186 |
+
"train_loss": 0.7403765432559616,
|
| 2187 |
+
"train_runtime": 55016.1932,
|
| 2188 |
+
"train_samples_per_second": 32.953,
|
| 2189 |
+
"train_steps_per_second": 3.295
|
| 2190 |
+
}
|
| 2191 |
+
],
|
| 2192 |
+
"max_steps": 181296,
|
| 2193 |
+
"num_train_epochs": 3,
|
| 2194 |
+
"total_flos": 2.0627822520336384e+17,
|
| 2195 |
+
"trial_name": null,
|
| 2196 |
+
"trial_params": null
|
| 2197 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13cfe1bd50c661a4251ae7d3ee79df5b3991525ef927ade20772c4e7ec8732fd
|
| 3 |
+
size 2799
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|