Hugging Face's logo

masakhane
/
m2m100_418M_en_tsn_rel

Text2Text Generation
Transformers
PyTorch
English
Tswana
m2m_100
Inference Endpoints
Model card Files Community
2
Davlan commited on
Commit
f761abc
•
1 Parent(s): 24ad653

add MT model

Browse files
Files changed (13) hide show
  1. all_results.json +22 -0
  2. config.json +38 -0
  3. eval_results.json +10 -0
  4. generated_predictions.txt +0 -0
  5. predict_results.json +9 -0
  6. pytorch_model.bin +3 -0
  7. sentencepiece.bpe.model +3 -0
  8. special_tokens_map.json +1 -0
  9. tokenizer_config.json +1 -0
  10. train_results.json +8 -0
  11. trainer_state.json +3151 -0
  12. training_args.bin +3 -0
  13. vocab.json +0 -0
all_results.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_bleu": 21.064,
4
+ "eval_gen_len": 55.7833,
5
+ "eval_loss": 2.461031198501587,
6
+ "eval_runtime": 230.0777,
7
+ "eval_samples": 540,
8
+ "eval_samples_per_second": 2.347,
9
+ "eval_steps_per_second": 0.235,
10
+ "predict_bleu": 19.8454,
11
+ "predict_gen_len": 53.8833,
12
+ "predict_loss": 2.305523157119751,
13
+ "predict_runtime": 632.5281,
14
+ "predict_samples": 1500,
15
+ "predict_samples_per_second": 2.371,
16
+ "predict_steps_per_second": 0.237,
17
+ "train_loss": 0.11540543585032806,
18
+ "train_runtime": 18089.1482,
19
+ "train_samples": 869199,
20
+ "train_samples_per_second": 144.153,
21
+ "train_steps_per_second": 14.415
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": 128092,
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": 21.064,
4
+ "eval_gen_len": 55.7833,
5
+ "eval_loss": 2.461031198501587,
6
+ "eval_runtime": 230.0777,
7
+ "eval_samples": 540,
8
+ "eval_samples_per_second": 2.347,
9
+ "eval_steps_per_second": 0.235
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": 19.8454,
3
+ "predict_gen_len": 53.8833,
4
+ "predict_loss": 2.305523157119751,
5
+ "predict_runtime": 632.5281,
6
+ "predict_samples": 1500,
7
+ "predict_samples_per_second": 2.371,
8
+ "predict_steps_per_second": 0.237
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2178903eac19d7f61352c8d4f58fa0a89e07209f1eead901850844136be7a3f7
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": 0.11540543585032806,
4
+ "train_runtime": 18089.1482,
5
+ "train_samples": 869199,
6
+ "train_samples_per_second": 144.153,
7
+ "train_steps_per_second": 14.415
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3151 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 260760,
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": 4.9904126399754563e-05,
13
+ "loss": 2.9959,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 4.980825279950913e-05,
19
+ "loss": 2.0327,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 4.971237919926369e-05,
25
+ "loss": 1.7134,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 4.961650559901826e-05,
31
+ "loss": 1.6017,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 4.952063199877282e-05,
37
+ "loss": 1.4858,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 4.942475839852738e-05,
43
+ "loss": 1.3887,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.04,
48
+ "learning_rate": 4.932888479828195e-05,
49
+ "loss": 1.348,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 4.923301119803651e-05,
55
+ "loss": 1.2965,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.05,
60
+ "learning_rate": 4.913713759779107e-05,
61
+ "loss": 1.2528,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.06,
66
+ "learning_rate": 4.904126399754564e-05,
67
+ "loss": 1.2444,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.06,
72
+ "learning_rate": 4.89453903973002e-05,
73
+ "loss": 1.2153,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.07,
78
+ "learning_rate": 4.884951679705477e-05,
79
+ "loss": 1.189,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.07,
84
+ "learning_rate": 4.875364319680933e-05,
85
+ "loss": 1.162,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.08,
90
+ "learning_rate": 4.865776959656389e-05,
91
+ "loss": 1.144,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.09,
96
+ "learning_rate": 4.856189599631846e-05,
97
+ "loss": 1.1463,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.09,
102
+ "learning_rate": 4.846602239607302e-05,
103
+ "loss": 1.1062,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.1,
108
+ "learning_rate": 4.837014879582759e-05,
109
+ "loss": 1.0981,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.1,
114
+ "learning_rate": 4.827427519558215e-05,
115
+ "loss": 1.0894,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.11,
120
+ "learning_rate": 4.817840159533671e-05,
121
+ "loss": 1.0681,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.12,
126
+ "learning_rate": 4.808252799509128e-05,
127
+ "loss": 1.0633,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.12,
132
+ "learning_rate": 4.798665439484584e-05,
133
+ "loss": 1.0484,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.13,
138
+ "learning_rate": 4.7890780794600406e-05,
139
+ "loss": 1.0343,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.13,
144
+ "learning_rate": 4.779490719435497e-05,
145
+ "loss": 1.0362,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.14,
150
+ "learning_rate": 4.769903359410953e-05,
151
+ "loss": 1.0078,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.14,
156
+ "learning_rate": 4.7603159993864096e-05,
157
+ "loss": 1.0099,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.15,
162
+ "learning_rate": 4.750728639361866e-05,
163
+ "loss": 0.987,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.16,
168
+ "learning_rate": 4.741141279337322e-05,
169
+ "loss": 0.9977,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.16,
174
+ "learning_rate": 4.7315539193127786e-05,
175
+ "loss": 0.9912,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.17,
180
+ "learning_rate": 4.721966559288235e-05,
181
+ "loss": 0.9769,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.17,
186
+ "learning_rate": 4.7123791992636915e-05,
187
+ "loss": 0.9701,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.18,
192
+ "learning_rate": 4.7027918392391476e-05,
193
+ "loss": 0.9703,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.18,
198
+ "learning_rate": 4.6932044792146037e-05,
199
+ "loss": 0.9531,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.19,
204
+ "learning_rate": 4.6836171191900604e-05,
205
+ "loss": 0.9525,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.2,
210
+ "learning_rate": 4.6740297591655165e-05,
211
+ "loss": 0.9588,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.2,
216
+ "learning_rate": 4.664442399140973e-05,
217
+ "loss": 0.9511,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.21,
222
+ "learning_rate": 4.6548550391164294e-05,
223
+ "loss": 0.9375,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.21,
228
+ "learning_rate": 4.6452676790918855e-05,
229
+ "loss": 0.9315,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.22,
234
+ "learning_rate": 4.635680319067342e-05,
235
+ "loss": 0.9313,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.22,
240
+ "learning_rate": 4.6260929590427984e-05,
241
+ "loss": 0.9272,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.23,
246
+ "learning_rate": 4.6165055990182545e-05,
247
+ "loss": 0.929,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.24,
252
+ "learning_rate": 4.606918238993711e-05,
253
+ "loss": 0.9215,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.24,
258
+ "learning_rate": 4.5973308789691674e-05,
259
+ "loss": 0.9066,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.25,
264
+ "learning_rate": 4.587743518944624e-05,
265
+ "loss": 0.8864,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.25,
270
+ "learning_rate": 4.57815615892008e-05,
271
+ "loss": 0.8892,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.26,
276
+ "learning_rate": 4.5685687988955364e-05,
277
+ "loss": 0.8988,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.26,
282
+ "learning_rate": 4.558981438870993e-05,
283
+ "loss": 0.902,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.27,
288
+ "learning_rate": 4.549394078846449e-05,
289
+ "loss": 0.8844,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.28,
294
+ "learning_rate": 4.5398067188219054e-05,
295
+ "loss": 0.8954,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.28,
300
+ "learning_rate": 4.530219358797362e-05,
301
+ "loss": 0.8988,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.29,
306
+ "learning_rate": 4.520631998772818e-05,
307
+ "loss": 0.8754,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.29,
312
+ "learning_rate": 4.5110446387482743e-05,
313
+ "loss": 0.8661,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.3,
318
+ "learning_rate": 4.501457278723731e-05,
319
+ "loss": 0.8818,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.3,
324
+ "learning_rate": 4.491869918699187e-05,
325
+ "loss": 0.8647,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.31,
330
+ "learning_rate": 4.482282558674643e-05,
331
+ "loss": 0.8867,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.32,
336
+ "learning_rate": 4.4726951986501e-05,
337
+ "loss": 0.8702,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.32,
342
+ "learning_rate": 4.463107838625556e-05,
343
+ "loss": 0.853,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.33,
348
+ "learning_rate": 4.453520478601012e-05,
349
+ "loss": 0.8685,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.33,
354
+ "learning_rate": 4.443933118576469e-05,
355
+ "loss": 0.8677,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.34,
360
+ "learning_rate": 4.434345758551925e-05,
361
+ "loss": 0.8651,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.35,
366
+ "learning_rate": 4.424758398527381e-05,
367
+ "loss": 0.8494,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.35,
372
+ "learning_rate": 4.415171038502838e-05,
373
+ "loss": 0.8409,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.36,
378
+ "learning_rate": 4.405583678478294e-05,
379
+ "loss": 0.8441,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.36,
384
+ "learning_rate": 4.39599631845375e-05,
385
+ "loss": 0.8382,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.37,
390
+ "learning_rate": 4.386408958429207e-05,
391
+ "loss": 0.8346,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.37,
396
+ "learning_rate": 4.376821598404663e-05,
397
+ "loss": 0.8507,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.38,
402
+ "learning_rate": 4.36723423838012e-05,
403
+ "loss": 0.8328,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.39,
408
+ "learning_rate": 4.357646878355576e-05,
409
+ "loss": 0.8281,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.39,
414
+ "learning_rate": 4.348059518331032e-05,
415
+ "loss": 0.8423,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.4,
420
+ "learning_rate": 4.338472158306489e-05,
421
+ "loss": 0.8161,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.4,
426
+ "learning_rate": 4.328884798281945e-05,
427
+ "loss": 0.8438,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.41,
432
+ "learning_rate": 4.319297438257401e-05,
433
+ "loss": 0.8278,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.41,
438
+ "learning_rate": 4.309710078232858e-05,
439
+ "loss": 0.8269,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.42,
444
+ "learning_rate": 4.300122718208314e-05,
445
+ "loss": 0.8266,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.43,
450
+ "learning_rate": 4.290535358183771e-05,
451
+ "loss": 0.8211,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.43,
456
+ "learning_rate": 4.280947998159227e-05,
457
+ "loss": 0.8376,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.44,
462
+ "learning_rate": 4.271360638134683e-05,
463
+ "loss": 0.8106,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.44,
468
+ "learning_rate": 4.26177327811014e-05,
469
+ "loss": 0.8044,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.45,
474
+ "learning_rate": 4.252185918085596e-05,
475
+ "loss": 0.8068,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.45,
480
+ "learning_rate": 4.242598558061053e-05,
481
+ "loss": 0.8109,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.46,
486
+ "learning_rate": 4.233011198036509e-05,
487
+ "loss": 0.8165,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.47,
492
+ "learning_rate": 4.223423838011965e-05,
493
+ "loss": 0.8064,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.47,
498
+ "learning_rate": 4.213836477987422e-05,
499
+ "loss": 0.7974,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.48,
504
+ "learning_rate": 4.204249117962878e-05,
505
+ "loss": 0.7813,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.48,
510
+ "learning_rate": 4.194661757938334e-05,
511
+ "loss": 0.8056,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.49,
516
+ "learning_rate": 4.1850743979137906e-05,
517
+ "loss": 0.7912,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.49,
522
+ "learning_rate": 4.175487037889247e-05,
523
+ "loss": 0.7834,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.5,
528
+ "learning_rate": 4.1658996778647035e-05,
529
+ "loss": 0.7965,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.51,
534
+ "learning_rate": 4.1563123178401596e-05,
535
+ "loss": 0.7965,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.51,
540
+ "learning_rate": 4.146724957815616e-05,
541
+ "loss": 0.7842,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.52,
546
+ "learning_rate": 4.1371375977910725e-05,
547
+ "loss": 0.7985,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.52,
552
+ "learning_rate": 4.1275502377665286e-05,
553
+ "loss": 0.796,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.53,
558
+ "learning_rate": 4.1179628777419854e-05,
559
+ "loss": 0.7936,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.53,
564
+ "learning_rate": 4.1083755177174415e-05,
565
+ "loss": 0.7953,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.54,
570
+ "learning_rate": 4.0987881576928976e-05,
571
+ "loss": 0.7797,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.55,
576
+ "learning_rate": 4.0892007976683544e-05,
577
+ "loss": 0.7824,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.55,
582
+ "learning_rate": 4.0796134376438105e-05,
583
+ "loss": 0.7743,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.56,
588
+ "learning_rate": 4.0700260776192666e-05,
589
+ "loss": 0.7841,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.56,
594
+ "learning_rate": 4.0604387175947234e-05,
595
+ "loss": 0.7778,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.57,
600
+ "learning_rate": 4.0508513575701795e-05,
601
+ "loss": 0.7811,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.58,
606
+ "learning_rate": 4.041263997545636e-05,
607
+ "loss": 0.7662,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.58,
612
+ "learning_rate": 4.0316766375210924e-05,
613
+ "loss": 0.7768,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.59,
618
+ "learning_rate": 4.0220892774965485e-05,
619
+ "loss": 0.7689,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.59,
624
+ "learning_rate": 4.012501917472005e-05,
625
+ "loss": 0.7785,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.6,
630
+ "learning_rate": 4.0029145574474613e-05,
631
+ "loss": 0.7688,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.6,
636
+ "learning_rate": 3.993327197422918e-05,
637
+ "loss": 0.7787,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.61,
642
+ "learning_rate": 3.983739837398374e-05,
643
+ "loss": 0.7649,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.62,
648
+ "learning_rate": 3.97415247737383e-05,
649
+ "loss": 0.7692,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.62,
654
+ "learning_rate": 3.964565117349287e-05,
655
+ "loss": 0.7682,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.63,
660
+ "learning_rate": 3.954977757324743e-05,
661
+ "loss": 0.7614,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.63,
666
+ "learning_rate": 3.9453903973002e-05,
667
+ "loss": 0.7726,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.64,
672
+ "learning_rate": 3.935803037275656e-05,
673
+ "loss": 0.7621,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.64,
678
+ "learning_rate": 3.926215677251112e-05,
679
+ "loss": 0.7552,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.65,
684
+ "learning_rate": 3.916628317226569e-05,
685
+ "loss": 0.7518,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.66,
690
+ "learning_rate": 3.907040957202025e-05,
691
+ "loss": 0.77,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.66,
696
+ "learning_rate": 3.897453597177481e-05,
697
+ "loss": 0.7494,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.67,
702
+ "learning_rate": 3.887866237152938e-05,
703
+ "loss": 0.7673,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.67,
708
+ "learning_rate": 3.878278877128394e-05,
709
+ "loss": 0.7541,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.68,
714
+ "learning_rate": 3.868691517103851e-05,
715
+ "loss": 0.7398,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.68,
720
+ "learning_rate": 3.859104157079307e-05,
721
+ "loss": 0.744,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.69,
726
+ "learning_rate": 3.849516797054763e-05,
727
+ "loss": 0.7443,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.7,
732
+ "learning_rate": 3.83992943703022e-05,
733
+ "loss": 0.7514,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.7,
738
+ "learning_rate": 3.830342077005676e-05,
739
+ "loss": 0.7395,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.71,
744
+ "learning_rate": 3.820754716981133e-05,
745
+ "loss": 0.7477,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.71,
750
+ "learning_rate": 3.811167356956589e-05,
751
+ "loss": 0.7485,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.72,
756
+ "learning_rate": 3.801579996932045e-05,
757
+ "loss": 0.74,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.72,
762
+ "learning_rate": 3.791992636907502e-05,
763
+ "loss": 0.7349,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.73,
768
+ "learning_rate": 3.782405276882958e-05,
769
+ "loss": 0.746,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.74,
774
+ "learning_rate": 3.772817916858414e-05,
775
+ "loss": 0.7446,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.74,
780
+ "learning_rate": 3.763230556833871e-05,
781
+ "loss": 0.7328,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 0.75,
786
+ "learning_rate": 3.753643196809327e-05,
787
+ "loss": 0.7449,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 0.75,
792
+ "learning_rate": 3.7440558367847836e-05,
793
+ "loss": 0.7202,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 0.76,
798
+ "learning_rate": 3.73446847676024e-05,
799
+ "loss": 0.7384,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 0.77,
804
+ "learning_rate": 3.724881116735696e-05,
805
+ "loss": 0.7249,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 0.77,
810
+ "learning_rate": 3.7152937567111526e-05,
811
+ "loss": 0.7286,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 0.78,
816
+ "learning_rate": 3.7057063966866087e-05,
817
+ "loss": 0.728,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 0.78,
822
+ "learning_rate": 3.6961190366620654e-05,
823
+ "loss": 0.7224,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 0.79,
828
+ "learning_rate": 3.6865316766375215e-05,
829
+ "loss": 0.7306,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 0.79,
834
+ "learning_rate": 3.6769443166129776e-05,
835
+ "loss": 0.7264,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 0.8,
840
+ "learning_rate": 3.6673569565884344e-05,
841
+ "loss": 0.7392,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 0.81,
846
+ "learning_rate": 3.6577695965638905e-05,
847
+ "loss": 0.7216,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 0.81,
852
+ "learning_rate": 3.6481822365393466e-05,
853
+ "loss": 0.7271,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 0.82,
858
+ "learning_rate": 3.6385948765148034e-05,
859
+ "loss": 0.7233,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 0.82,
864
+ "learning_rate": 3.6290075164902595e-05,
865
+ "loss": 0.7364,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 0.83,
870
+ "learning_rate": 3.619420156465716e-05,
871
+ "loss": 0.7186,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 0.83,
876
+ "learning_rate": 3.6098327964411724e-05,
877
+ "loss": 0.7246,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 0.84,
882
+ "learning_rate": 3.6002454364166285e-05,
883
+ "loss": 0.7228,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 0.85,
888
+ "learning_rate": 3.590658076392085e-05,
889
+ "loss": 0.7183,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 0.85,
894
+ "learning_rate": 3.5810707163675414e-05,
895
+ "loss": 0.7187,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 0.86,
900
+ "learning_rate": 3.5714833563429975e-05,
901
+ "loss": 0.7188,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 0.86,
906
+ "learning_rate": 3.561895996318454e-05,
907
+ "loss": 0.7234,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 0.87,
912
+ "learning_rate": 3.5523086362939104e-05,
913
+ "loss": 0.7102,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 0.87,
918
+ "learning_rate": 3.5427212762693665e-05,
919
+ "loss": 0.7201,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 0.88,
924
+ "learning_rate": 3.533133916244823e-05,
925
+ "loss": 0.7176,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 0.89,
930
+ "learning_rate": 3.5235465562202793e-05,
931
+ "loss": 0.7069,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 0.89,
936
+ "learning_rate": 3.5139591961957355e-05,
937
+ "loss": 0.7118,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 0.9,
942
+ "learning_rate": 3.504371836171192e-05,
943
+ "loss": 0.7147,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 0.9,
948
+ "learning_rate": 3.494784476146648e-05,
949
+ "loss": 0.7132,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 0.91,
954
+ "learning_rate": 3.4851971161221044e-05,
955
+ "loss": 0.7059,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 0.91,
960
+ "learning_rate": 3.475609756097561e-05,
961
+ "loss": 0.713,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 0.92,
966
+ "learning_rate": 3.466022396073017e-05,
967
+ "loss": 0.7154,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 0.93,
972
+ "learning_rate": 3.4564350360484734e-05,
973
+ "loss": 0.69,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 0.93,
978
+ "learning_rate": 3.44684767602393e-05,
979
+ "loss": 0.7152,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 0.94,
984
+ "learning_rate": 3.437260315999386e-05,
985
+ "loss": 0.6957,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 0.94,
990
+ "learning_rate": 3.4276729559748424e-05,
991
+ "loss": 0.7189,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 0.95,
996
+ "learning_rate": 3.418085595950299e-05,
997
+ "loss": 0.705,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 0.95,
1002
+ "learning_rate": 3.408498235925755e-05,
1003
+ "loss": 0.7036,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 0.96,
1008
+ "learning_rate": 3.398910875901212e-05,
1009
+ "loss": 0.7008,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 0.97,
1014
+ "learning_rate": 3.389323515876668e-05,
1015
+ "loss": 0.6923,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 0.97,
1020
+ "learning_rate": 3.379736155852124e-05,
1021
+ "loss": 0.7158,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 0.98,
1026
+ "learning_rate": 3.370148795827581e-05,
1027
+ "loss": 0.7009,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 0.98,
1032
+ "learning_rate": 3.360561435803037e-05,
1033
+ "loss": 0.7074,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 0.99,
1038
+ "learning_rate": 3.350974075778493e-05,
1039
+ "loss": 0.7037,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 1.0,
1044
+ "learning_rate": 3.34138671575395e-05,
1045
+ "loss": 0.6913,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 1.0,
1050
+ "learning_rate": 3.331799355729406e-05,
1051
+ "loss": 0.693,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 1.01,
1056
+ "learning_rate": 3.322211995704863e-05,
1057
+ "loss": 0.6385,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 1.01,
1062
+ "learning_rate": 3.312624635680319e-05,
1063
+ "loss": 0.6387,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 1.02,
1068
+ "learning_rate": 3.303037275655775e-05,
1069
+ "loss": 0.6385,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 1.02,
1074
+ "learning_rate": 3.293449915631232e-05,
1075
+ "loss": 0.6346,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 1.03,
1080
+ "learning_rate": 3.283862555606688e-05,
1081
+ "loss": 0.6373,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 1.04,
1086
+ "learning_rate": 3.274275195582145e-05,
1087
+ "loss": 0.6377,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 1.04,
1092
+ "learning_rate": 3.264687835557601e-05,
1093
+ "loss": 0.6451,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 1.05,
1098
+ "learning_rate": 3.255100475533057e-05,
1099
+ "loss": 0.6343,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 1.05,
1104
+ "learning_rate": 3.245513115508514e-05,
1105
+ "loss": 0.6436,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 1.06,
1110
+ "learning_rate": 3.23592575548397e-05,
1111
+ "loss": 0.6248,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 1.06,
1116
+ "learning_rate": 3.226338395459426e-05,
1117
+ "loss": 0.6393,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 1.07,
1122
+ "learning_rate": 3.216751035434883e-05,
1123
+ "loss": 0.6388,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 1.08,
1128
+ "learning_rate": 3.207163675410339e-05,
1129
+ "loss": 0.643,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 1.08,
1134
+ "learning_rate": 3.1975763153857956e-05,
1135
+ "loss": 0.6426,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 1.09,
1140
+ "learning_rate": 3.187988955361252e-05,
1141
+ "loss": 0.6269,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 1.09,
1146
+ "learning_rate": 3.178401595336708e-05,
1147
+ "loss": 0.6372,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 1.1,
1152
+ "learning_rate": 3.1688142353121646e-05,
1153
+ "loss": 0.6213,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 1.1,
1158
+ "learning_rate": 3.159226875287621e-05,
1159
+ "loss": 0.6396,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 1.11,
1164
+ "learning_rate": 3.1496395152630775e-05,
1165
+ "loss": 0.6388,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 1.12,
1170
+ "learning_rate": 3.1400521552385336e-05,
1171
+ "loss": 0.6325,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 1.12,
1176
+ "learning_rate": 3.13046479521399e-05,
1177
+ "loss": 0.6311,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 1.13,
1182
+ "learning_rate": 3.1208774351894465e-05,
1183
+ "loss": 0.635,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 1.13,
1188
+ "learning_rate": 3.1112900751649026e-05,
1189
+ "loss": 0.6311,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 1.14,
1194
+ "learning_rate": 3.1017027151403594e-05,
1195
+ "loss": 0.6406,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 1.14,
1200
+ "learning_rate": 3.0921153551158155e-05,
1201
+ "loss": 0.6436,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 1.15,
1206
+ "learning_rate": 3.0825279950912716e-05,
1207
+ "loss": 0.6294,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 1.16,
1212
+ "learning_rate": 3.0729406350667284e-05,
1213
+ "loss": 0.6307,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 1.16,
1218
+ "learning_rate": 3.0633532750421845e-05,
1219
+ "loss": 0.6325,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 1.17,
1224
+ "learning_rate": 3.0537659150176406e-05,
1225
+ "loss": 0.6343,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 1.17,
1230
+ "learning_rate": 3.0441785549930974e-05,
1231
+ "loss": 0.6296,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 1.18,
1236
+ "learning_rate": 3.0345911949685535e-05,
1237
+ "loss": 0.6244,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 1.18,
1242
+ "learning_rate": 3.0250038349440102e-05,
1243
+ "loss": 0.6342,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 1.19,
1248
+ "learning_rate": 3.0154164749194663e-05,
1249
+ "loss": 0.6259,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 1.2,
1254
+ "learning_rate": 3.0058291148949224e-05,
1255
+ "loss": 0.6332,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 1.2,
1260
+ "learning_rate": 2.9962417548703792e-05,
1261
+ "loss": 0.626,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 1.21,
1266
+ "learning_rate": 2.9866543948458353e-05,
1267
+ "loss": 0.6314,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 1.21,
1272
+ "learning_rate": 2.977067034821292e-05,
1273
+ "loss": 0.6276,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 1.22,
1278
+ "learning_rate": 2.9674796747967482e-05,
1279
+ "loss": 0.6274,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 1.23,
1284
+ "learning_rate": 2.9578923147722043e-05,
1285
+ "loss": 0.6231,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 1.23,
1290
+ "learning_rate": 2.948304954747661e-05,
1291
+ "loss": 0.6257,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 1.24,
1296
+ "learning_rate": 2.9387175947231172e-05,
1297
+ "loss": 0.6379,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 1.24,
1302
+ "learning_rate": 2.9291302346985733e-05,
1303
+ "loss": 0.6294,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 1.25,
1308
+ "learning_rate": 2.91954287467403e-05,
1309
+ "loss": 0.6297,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 1.25,
1314
+ "learning_rate": 2.9099555146494862e-05,
1315
+ "loss": 0.6273,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 1.26,
1320
+ "learning_rate": 2.900368154624943e-05,
1321
+ "loss": 0.6223,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 1.27,
1326
+ "learning_rate": 2.890780794600399e-05,
1327
+ "loss": 0.6352,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 1.27,
1332
+ "learning_rate": 2.881193434575855e-05,
1333
+ "loss": 0.6209,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 1.28,
1338
+ "learning_rate": 2.871606074551312e-05,
1339
+ "loss": 0.627,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 1.28,
1344
+ "learning_rate": 2.862018714526768e-05,
1345
+ "loss": 0.63,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 1.29,
1350
+ "learning_rate": 2.8524313545022245e-05,
1351
+ "loss": 0.6264,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 1.29,
1356
+ "learning_rate": 2.842843994477681e-05,
1357
+ "loss": 0.6212,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 1.3,
1362
+ "learning_rate": 2.833256634453137e-05,
1363
+ "loss": 0.6137,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 1.31,
1368
+ "learning_rate": 2.8236692744285935e-05,
1369
+ "loss": 0.6187,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 1.31,
1374
+ "learning_rate": 2.81408191440405e-05,
1375
+ "loss": 0.6266,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 1.32,
1380
+ "learning_rate": 2.804494554379506e-05,
1381
+ "loss": 0.6157,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 1.32,
1386
+ "learning_rate": 2.7949071943549625e-05,
1387
+ "loss": 0.623,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 1.33,
1392
+ "learning_rate": 2.785319834330419e-05,
1393
+ "loss": 0.6202,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 1.33,
1398
+ "learning_rate": 2.7757324743058753e-05,
1399
+ "loss": 0.6169,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 1.34,
1404
+ "learning_rate": 2.7661451142813314e-05,
1405
+ "loss": 0.6134,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 1.35,
1410
+ "learning_rate": 2.756557754256788e-05,
1411
+ "loss": 0.6189,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 1.35,
1416
+ "learning_rate": 2.7469703942322443e-05,
1417
+ "loss": 0.6263,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 1.36,
1422
+ "learning_rate": 2.7373830342077004e-05,
1423
+ "loss": 0.6331,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 1.36,
1428
+ "learning_rate": 2.7277956741831572e-05,
1429
+ "loss": 0.6204,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 1.37,
1434
+ "learning_rate": 2.7182083141586133e-05,
1435
+ "loss": 0.6187,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 1.37,
1440
+ "learning_rate": 2.7086209541340694e-05,
1441
+ "loss": 0.6099,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 1.38,
1446
+ "learning_rate": 2.6990335941095262e-05,
1447
+ "loss": 0.6184,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 1.39,
1452
+ "learning_rate": 2.6894462340849823e-05,
1453
+ "loss": 0.6134,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 1.39,
1458
+ "learning_rate": 2.679858874060439e-05,
1459
+ "loss": 0.6064,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 1.4,
1464
+ "learning_rate": 2.6702715140358952e-05,
1465
+ "loss": 0.6315,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 1.4,
1470
+ "learning_rate": 2.6606841540113513e-05,
1471
+ "loss": 0.6245,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 1.41,
1476
+ "learning_rate": 2.651096793986808e-05,
1477
+ "loss": 0.6069,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 1.42,
1482
+ "learning_rate": 2.641509433962264e-05,
1483
+ "loss": 0.6224,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 1.42,
1488
+ "learning_rate": 2.6319220739377203e-05,
1489
+ "loss": 0.6165,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 1.43,
1494
+ "learning_rate": 2.622334713913177e-05,
1495
+ "loss": 0.6124,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 1.43,
1500
+ "learning_rate": 2.612747353888633e-05,
1501
+ "loss": 0.6112,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 1.44,
1506
+ "learning_rate": 2.60315999386409e-05,
1507
+ "loss": 0.6093,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 1.44,
1512
+ "learning_rate": 2.593572633839546e-05,
1513
+ "loss": 0.6035,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 1.45,
1518
+ "learning_rate": 2.583985273815002e-05,
1519
+ "loss": 0.609,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 1.46,
1524
+ "learning_rate": 2.574397913790459e-05,
1525
+ "loss": 0.5955,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 1.46,
1530
+ "learning_rate": 2.564810553765915e-05,
1531
+ "loss": 0.6195,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 1.47,
1536
+ "learning_rate": 2.5552231937413718e-05,
1537
+ "loss": 0.6143,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 1.47,
1542
+ "learning_rate": 2.545635833716828e-05,
1543
+ "loss": 0.6083,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 1.48,
1548
+ "learning_rate": 2.536048473692284e-05,
1549
+ "loss": 0.6209,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 1.48,
1554
+ "learning_rate": 2.5264611136677408e-05,
1555
+ "loss": 0.6203,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 1.49,
1560
+ "learning_rate": 2.516873753643197e-05,
1561
+ "loss": 0.6001,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 1.5,
1566
+ "learning_rate": 2.507286393618653e-05,
1567
+ "loss": 0.6096,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 1.5,
1572
+ "learning_rate": 2.4976990335941098e-05,
1573
+ "loss": 0.6131,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 1.51,
1578
+ "learning_rate": 2.488111673569566e-05,
1579
+ "loss": 0.6104,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 1.51,
1584
+ "learning_rate": 2.4785243135450223e-05,
1585
+ "loss": 0.6044,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 1.52,
1590
+ "learning_rate": 2.4689369535204788e-05,
1591
+ "loss": 0.5952,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 1.52,
1596
+ "learning_rate": 2.4593495934959352e-05,
1597
+ "loss": 0.5971,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 1.53,
1602
+ "learning_rate": 2.4497622334713916e-05,
1603
+ "loss": 0.6026,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 1.54,
1608
+ "learning_rate": 2.4401748734468477e-05,
1609
+ "loss": 0.5951,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 1.54,
1614
+ "learning_rate": 2.4305875134223042e-05,
1615
+ "loss": 0.6101,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 1.55,
1620
+ "learning_rate": 2.4210001533977606e-05,
1621
+ "loss": 0.599,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 1.55,
1626
+ "learning_rate": 2.411412793373217e-05,
1627
+ "loss": 0.6086,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 1.56,
1632
+ "learning_rate": 2.4018254333486732e-05,
1633
+ "loss": 0.6069,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 1.56,
1638
+ "learning_rate": 2.3922380733241296e-05,
1639
+ "loss": 0.606,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 1.57,
1644
+ "learning_rate": 2.382650713299586e-05,
1645
+ "loss": 0.6106,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 1.58,
1650
+ "learning_rate": 2.3730633532750425e-05,
1651
+ "loss": 0.5935,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 1.58,
1656
+ "learning_rate": 2.363475993250499e-05,
1657
+ "loss": 0.6018,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 1.59,
1662
+ "learning_rate": 2.353888633225955e-05,
1663
+ "loss": 0.6004,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 1.59,
1668
+ "learning_rate": 2.3443012732014115e-05,
1669
+ "loss": 0.5984,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 1.6,
1674
+ "learning_rate": 2.334713913176868e-05,
1675
+ "loss": 0.6003,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 1.6,
1680
+ "learning_rate": 2.325126553152324e-05,
1681
+ "loss": 0.6122,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 1.61,
1686
+ "learning_rate": 2.3155391931277805e-05,
1687
+ "loss": 0.5897,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 1.62,
1692
+ "learning_rate": 2.305951833103237e-05,
1693
+ "loss": 0.604,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 1.62,
1698
+ "learning_rate": 2.296364473078693e-05,
1699
+ "loss": 0.5999,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 1.63,
1704
+ "learning_rate": 2.2867771130541495e-05,
1705
+ "loss": 0.5959,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 1.63,
1710
+ "learning_rate": 2.277189753029606e-05,
1711
+ "loss": 0.6054,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 1.64,
1716
+ "learning_rate": 2.267602393005062e-05,
1717
+ "loss": 0.6028,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 1.65,
1722
+ "learning_rate": 2.2580150329805184e-05,
1723
+ "loss": 0.6035,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 1.65,
1728
+ "learning_rate": 2.248427672955975e-05,
1729
+ "loss": 0.5916,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 1.66,
1734
+ "learning_rate": 2.2388403129314313e-05,
1735
+ "loss": 0.6072,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 1.66,
1740
+ "learning_rate": 2.2292529529068874e-05,
1741
+ "loss": 0.5871,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 1.67,
1746
+ "learning_rate": 2.219665592882344e-05,
1747
+ "loss": 0.601,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 1.67,
1752
+ "learning_rate": 2.2100782328578003e-05,
1753
+ "loss": 0.5954,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 1.68,
1758
+ "learning_rate": 2.2004908728332567e-05,
1759
+ "loss": 0.5971,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 1.69,
1764
+ "learning_rate": 2.190903512808713e-05,
1765
+ "loss": 0.5976,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 1.69,
1770
+ "learning_rate": 2.1813161527841693e-05,
1771
+ "loss": 0.5951,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 1.7,
1776
+ "learning_rate": 2.1717287927596257e-05,
1777
+ "loss": 0.5925,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 1.7,
1782
+ "learning_rate": 2.1621414327350822e-05,
1783
+ "loss": 0.5913,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 1.71,
1788
+ "learning_rate": 2.1525540727105386e-05,
1789
+ "loss": 0.6006,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 1.71,
1794
+ "learning_rate": 2.1429667126859947e-05,
1795
+ "loss": 0.5942,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 1.72,
1800
+ "learning_rate": 2.133379352661451e-05,
1801
+ "loss": 0.5825,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 1.73,
1806
+ "learning_rate": 2.1237919926369076e-05,
1807
+ "loss": 0.592,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 1.73,
1812
+ "learning_rate": 2.114204632612364e-05,
1813
+ "loss": 0.5848,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 1.74,
1818
+ "learning_rate": 2.10461727258782e-05,
1819
+ "loss": 0.594,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 1.74,
1824
+ "learning_rate": 2.0950299125632766e-05,
1825
+ "loss": 0.598,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 1.75,
1830
+ "learning_rate": 2.085442552538733e-05,
1831
+ "loss": 0.5973,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 1.75,
1836
+ "learning_rate": 2.0758551925141895e-05,
1837
+ "loss": 0.5906,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 1.76,
1842
+ "learning_rate": 2.0662678324896456e-05,
1843
+ "loss": 0.5904,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 1.77,
1848
+ "learning_rate": 2.056680472465102e-05,
1849
+ "loss": 0.5919,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 1.77,
1854
+ "learning_rate": 2.0470931124405585e-05,
1855
+ "loss": 0.5979,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 1.78,
1860
+ "learning_rate": 2.037505752416015e-05,
1861
+ "loss": 0.5892,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 1.78,
1866
+ "learning_rate": 2.0279183923914713e-05,
1867
+ "loss": 0.5922,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 1.79,
1872
+ "learning_rate": 2.0183310323669274e-05,
1873
+ "loss": 0.5945,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 1.79,
1878
+ "learning_rate": 2.008743672342384e-05,
1879
+ "loss": 0.5907,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 1.8,
1884
+ "learning_rate": 1.9991563123178403e-05,
1885
+ "loss": 0.5775,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 1.81,
1890
+ "learning_rate": 1.9895689522932968e-05,
1891
+ "loss": 0.5845,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 1.81,
1896
+ "learning_rate": 1.979981592268753e-05,
1897
+ "loss": 0.5941,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 1.82,
1902
+ "learning_rate": 1.9703942322442093e-05,
1903
+ "loss": 0.5806,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 1.82,
1908
+ "learning_rate": 1.9608068722196658e-05,
1909
+ "loss": 0.5867,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 1.83,
1914
+ "learning_rate": 1.9512195121951222e-05,
1915
+ "loss": 0.5862,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 1.84,
1920
+ "learning_rate": 1.9416321521705786e-05,
1921
+ "loss": 0.5726,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 1.84,
1926
+ "learning_rate": 1.9320447921460347e-05,
1927
+ "loss": 0.5832,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 1.85,
1932
+ "learning_rate": 1.9224574321214912e-05,
1933
+ "loss": 0.5858,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 1.85,
1938
+ "learning_rate": 1.9128700720969476e-05,
1939
+ "loss": 0.5805,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 1.86,
1944
+ "learning_rate": 1.903282712072404e-05,
1945
+ "loss": 0.5764,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 1.86,
1950
+ "learning_rate": 1.89369535204786e-05,
1951
+ "loss": 0.6018,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 1.87,
1956
+ "learning_rate": 1.8841079920233166e-05,
1957
+ "loss": 0.5776,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 1.88,
1962
+ "learning_rate": 1.874520631998773e-05,
1963
+ "loss": 0.595,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 1.88,
1968
+ "learning_rate": 1.8649332719742295e-05,
1969
+ "loss": 0.5891,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 1.89,
1974
+ "learning_rate": 1.8553459119496856e-05,
1975
+ "loss": 0.5952,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 1.89,
1980
+ "learning_rate": 1.845758551925142e-05,
1981
+ "loss": 0.5861,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 1.9,
1986
+ "learning_rate": 1.8361711919005985e-05,
1987
+ "loss": 0.5746,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 1.9,
1992
+ "learning_rate": 1.826583831876055e-05,
1993
+ "loss": 0.5905,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 1.91,
1998
+ "learning_rate": 1.816996471851511e-05,
1999
+ "loss": 0.5871,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 1.92,
2004
+ "learning_rate": 1.8074091118269675e-05,
2005
+ "loss": 0.5833,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 1.92,
2010
+ "learning_rate": 1.797821751802424e-05,
2011
+ "loss": 0.5864,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 1.93,
2016
+ "learning_rate": 1.78823439177788e-05,
2017
+ "loss": 0.5847,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 1.93,
2022
+ "learning_rate": 1.7786470317533364e-05,
2023
+ "loss": 0.5895,
2024
+ "step": 168000
2025
+ },
2026
+ {
2027
+ "epoch": 1.94,
2028
+ "learning_rate": 1.769059671728793e-05,
2029
+ "loss": 0.5862,
2030
+ "step": 168500
2031
+ },
2032
+ {
2033
+ "epoch": 1.94,
2034
+ "learning_rate": 1.759472311704249e-05,
2035
+ "loss": 0.5939,
2036
+ "step": 169000
2037
+ },
2038
+ {
2039
+ "epoch": 1.95,
2040
+ "learning_rate": 1.7498849516797054e-05,
2041
+ "loss": 0.5705,
2042
+ "step": 169500
2043
+ },
2044
+ {
2045
+ "epoch": 1.96,
2046
+ "learning_rate": 1.740297591655162e-05,
2047
+ "loss": 0.5706,
2048
+ "step": 170000
2049
+ },
2050
+ {
2051
+ "epoch": 1.96,
2052
+ "learning_rate": 1.7307102316306183e-05,
2053
+ "loss": 0.5667,
2054
+ "step": 170500
2055
+ },
2056
+ {
2057
+ "epoch": 1.97,
2058
+ "learning_rate": 1.7211228716060744e-05,
2059
+ "loss": 0.5772,
2060
+ "step": 171000
2061
+ },
2062
+ {
2063
+ "epoch": 1.97,
2064
+ "learning_rate": 1.711535511581531e-05,
2065
+ "loss": 0.5799,
2066
+ "step": 171500
2067
+ },
2068
+ {
2069
+ "epoch": 1.98,
2070
+ "learning_rate": 1.7019481515569873e-05,
2071
+ "loss": 0.5877,
2072
+ "step": 172000
2073
+ },
2074
+ {
2075
+ "epoch": 1.98,
2076
+ "learning_rate": 1.6923607915324437e-05,
2077
+ "loss": 0.5628,
2078
+ "step": 172500
2079
+ },
2080
+ {
2081
+ "epoch": 1.99,
2082
+ "learning_rate": 1.6827734315079e-05,
2083
+ "loss": 0.5812,
2084
+ "step": 173000
2085
+ },
2086
+ {
2087
+ "epoch": 2.0,
2088
+ "learning_rate": 1.6731860714833563e-05,
2089
+ "loss": 0.5675,
2090
+ "step": 173500
2091
+ },
2092
+ {
2093
+ "epoch": 2.0,
2094
+ "learning_rate": 1.6635987114588127e-05,
2095
+ "loss": 0.5465,
2096
+ "step": 174000
2097
+ },
2098
+ {
2099
+ "epoch": 2.01,
2100
+ "learning_rate": 1.654011351434269e-05,
2101
+ "loss": 0.5017,
2102
+ "step": 174500
2103
+ },
2104
+ {
2105
+ "epoch": 2.01,
2106
+ "learning_rate": 1.6444239914097253e-05,
2107
+ "loss": 0.4992,
2108
+ "step": 175000
2109
+ },
2110
+ {
2111
+ "epoch": 2.02,
2112
+ "learning_rate": 1.6348366313851817e-05,
2113
+ "loss": 0.4933,
2114
+ "step": 175500
2115
+ },
2116
+ {
2117
+ "epoch": 2.02,
2118
+ "learning_rate": 1.625249271360638e-05,
2119
+ "loss": 0.525,
2120
+ "step": 176000
2121
+ },
2122
+ {
2123
+ "epoch": 2.03,
2124
+ "learning_rate": 1.6156619113360946e-05,
2125
+ "loss": 0.5102,
2126
+ "step": 176500
2127
+ },
2128
+ {
2129
+ "epoch": 2.04,
2130
+ "learning_rate": 1.606074551311551e-05,
2131
+ "loss": 0.5177,
2132
+ "step": 177000
2133
+ },
2134
+ {
2135
+ "epoch": 2.04,
2136
+ "learning_rate": 1.596487191287007e-05,
2137
+ "loss": 0.5103,
2138
+ "step": 177500
2139
+ },
2140
+ {
2141
+ "epoch": 2.05,
2142
+ "learning_rate": 1.5868998312624636e-05,
2143
+ "loss": 0.5144,
2144
+ "step": 178000
2145
+ },
2146
+ {
2147
+ "epoch": 2.05,
2148
+ "learning_rate": 1.57731247123792e-05,
2149
+ "loss": 0.5163,
2150
+ "step": 178500
2151
+ },
2152
+ {
2153
+ "epoch": 2.06,
2154
+ "learning_rate": 1.5677251112133765e-05,
2155
+ "loss": 0.5015,
2156
+ "step": 179000
2157
+ },
2158
+ {
2159
+ "epoch": 2.07,
2160
+ "learning_rate": 1.5581377511888326e-05,
2161
+ "loss": 0.5067,
2162
+ "step": 179500
2163
+ },
2164
+ {
2165
+ "epoch": 2.07,
2166
+ "learning_rate": 1.548550391164289e-05,
2167
+ "loss": 0.5135,
2168
+ "step": 180000
2169
+ },
2170
+ {
2171
+ "epoch": 2.08,
2172
+ "learning_rate": 1.5389630311397454e-05,
2173
+ "loss": 0.5042,
2174
+ "step": 180500
2175
+ },
2176
+ {
2177
+ "epoch": 2.08,
2178
+ "learning_rate": 1.529375671115202e-05,
2179
+ "loss": 0.5123,
2180
+ "step": 181000
2181
+ },
2182
+ {
2183
+ "epoch": 2.09,
2184
+ "learning_rate": 1.5197883110906583e-05,
2185
+ "loss": 0.5128,
2186
+ "step": 181500
2187
+ },
2188
+ {
2189
+ "epoch": 2.09,
2190
+ "learning_rate": 1.5102009510661144e-05,
2191
+ "loss": 0.5003,
2192
+ "step": 182000
2193
+ },
2194
+ {
2195
+ "epoch": 2.1,
2196
+ "learning_rate": 1.5006135910415709e-05,
2197
+ "loss": 0.511,
2198
+ "step": 182500
2199
+ },
2200
+ {
2201
+ "epoch": 2.11,
2202
+ "learning_rate": 1.4910262310170273e-05,
2203
+ "loss": 0.5044,
2204
+ "step": 183000
2205
+ },
2206
+ {
2207
+ "epoch": 2.11,
2208
+ "learning_rate": 1.4814388709924838e-05,
2209
+ "loss": 0.5104,
2210
+ "step": 183500
2211
+ },
2212
+ {
2213
+ "epoch": 2.12,
2214
+ "learning_rate": 1.4718515109679399e-05,
2215
+ "loss": 0.512,
2216
+ "step": 184000
2217
+ },
2218
+ {
2219
+ "epoch": 2.12,
2220
+ "learning_rate": 1.4622641509433963e-05,
2221
+ "loss": 0.516,
2222
+ "step": 184500
2223
+ },
2224
+ {
2225
+ "epoch": 2.13,
2226
+ "learning_rate": 1.4526767909188527e-05,
2227
+ "loss": 0.5064,
2228
+ "step": 185000
2229
+ },
2230
+ {
2231
+ "epoch": 2.13,
2232
+ "learning_rate": 1.443089430894309e-05,
2233
+ "loss": 0.5139,
2234
+ "step": 185500
2235
+ },
2236
+ {
2237
+ "epoch": 2.14,
2238
+ "learning_rate": 1.4335020708697653e-05,
2239
+ "loss": 0.5052,
2240
+ "step": 186000
2241
+ },
2242
+ {
2243
+ "epoch": 2.15,
2244
+ "learning_rate": 1.4239147108452217e-05,
2245
+ "loss": 0.5083,
2246
+ "step": 186500
2247
+ },
2248
+ {
2249
+ "epoch": 2.15,
2250
+ "learning_rate": 1.414327350820678e-05,
2251
+ "loss": 0.4972,
2252
+ "step": 187000
2253
+ },
2254
+ {
2255
+ "epoch": 2.16,
2256
+ "learning_rate": 1.4047399907961344e-05,
2257
+ "loss": 0.5115,
2258
+ "step": 187500
2259
+ },
2260
+ {
2261
+ "epoch": 2.16,
2262
+ "learning_rate": 1.3951526307715909e-05,
2263
+ "loss": 0.5049,
2264
+ "step": 188000
2265
+ },
2266
+ {
2267
+ "epoch": 2.17,
2268
+ "learning_rate": 1.385565270747047e-05,
2269
+ "loss": 0.5188,
2270
+ "step": 188500
2271
+ },
2272
+ {
2273
+ "epoch": 2.17,
2274
+ "learning_rate": 1.3759779107225034e-05,
2275
+ "loss": 0.5102,
2276
+ "step": 189000
2277
+ },
2278
+ {
2279
+ "epoch": 2.18,
2280
+ "learning_rate": 1.3663905506979599e-05,
2281
+ "loss": 0.5034,
2282
+ "step": 189500
2283
+ },
2284
+ {
2285
+ "epoch": 2.19,
2286
+ "learning_rate": 1.3568031906734163e-05,
2287
+ "loss": 0.5158,
2288
+ "step": 190000
2289
+ },
2290
+ {
2291
+ "epoch": 2.19,
2292
+ "learning_rate": 1.3472158306488724e-05,
2293
+ "loss": 0.5021,
2294
+ "step": 190500
2295
+ },
2296
+ {
2297
+ "epoch": 2.2,
2298
+ "learning_rate": 1.3376284706243289e-05,
2299
+ "loss": 0.5116,
2300
+ "step": 191000
2301
+ },
2302
+ {
2303
+ "epoch": 2.2,
2304
+ "learning_rate": 1.3280411105997853e-05,
2305
+ "loss": 0.5078,
2306
+ "step": 191500
2307
+ },
2308
+ {
2309
+ "epoch": 2.21,
2310
+ "learning_rate": 1.3184537505752417e-05,
2311
+ "loss": 0.5182,
2312
+ "step": 192000
2313
+ },
2314
+ {
2315
+ "epoch": 2.21,
2316
+ "learning_rate": 1.3088663905506982e-05,
2317
+ "loss": 0.4994,
2318
+ "step": 192500
2319
+ },
2320
+ {
2321
+ "epoch": 2.22,
2322
+ "learning_rate": 1.2992790305261543e-05,
2323
+ "loss": 0.5111,
2324
+ "step": 193000
2325
+ },
2326
+ {
2327
+ "epoch": 2.23,
2328
+ "learning_rate": 1.2896916705016107e-05,
2329
+ "loss": 0.5058,
2330
+ "step": 193500
2331
+ },
2332
+ {
2333
+ "epoch": 2.23,
2334
+ "learning_rate": 1.2801043104770672e-05,
2335
+ "loss": 0.503,
2336
+ "step": 194000
2337
+ },
2338
+ {
2339
+ "epoch": 2.24,
2340
+ "learning_rate": 1.2705169504525236e-05,
2341
+ "loss": 0.5092,
2342
+ "step": 194500
2343
+ },
2344
+ {
2345
+ "epoch": 2.24,
2346
+ "learning_rate": 1.2609295904279797e-05,
2347
+ "loss": 0.5071,
2348
+ "step": 195000
2349
+ },
2350
+ {
2351
+ "epoch": 2.25,
2352
+ "learning_rate": 1.2513422304034362e-05,
2353
+ "loss": 0.5145,
2354
+ "step": 195500
2355
+ },
2356
+ {
2357
+ "epoch": 2.25,
2358
+ "learning_rate": 1.2417548703788926e-05,
2359
+ "loss": 0.5087,
2360
+ "step": 196000
2361
+ },
2362
+ {
2363
+ "epoch": 2.26,
2364
+ "learning_rate": 1.2321675103543489e-05,
2365
+ "loss": 0.5097,
2366
+ "step": 196500
2367
+ },
2368
+ {
2369
+ "epoch": 2.27,
2370
+ "learning_rate": 1.2225801503298053e-05,
2371
+ "loss": 0.5132,
2372
+ "step": 197000
2373
+ },
2374
+ {
2375
+ "epoch": 2.27,
2376
+ "learning_rate": 1.2129927903052617e-05,
2377
+ "loss": 0.5063,
2378
+ "step": 197500
2379
+ },
2380
+ {
2381
+ "epoch": 2.28,
2382
+ "learning_rate": 1.203405430280718e-05,
2383
+ "loss": 0.5143,
2384
+ "step": 198000
2385
+ },
2386
+ {
2387
+ "epoch": 2.28,
2388
+ "learning_rate": 1.1938180702561743e-05,
2389
+ "loss": 0.5013,
2390
+ "step": 198500
2391
+ },
2392
+ {
2393
+ "epoch": 2.29,
2394
+ "learning_rate": 1.1842307102316307e-05,
2395
+ "loss": 0.4953,
2396
+ "step": 199000
2397
+ },
2398
+ {
2399
+ "epoch": 2.3,
2400
+ "learning_rate": 1.174643350207087e-05,
2401
+ "loss": 0.5026,
2402
+ "step": 199500
2403
+ },
2404
+ {
2405
+ "epoch": 2.3,
2406
+ "learning_rate": 1.1650559901825433e-05,
2407
+ "loss": 0.5026,
2408
+ "step": 200000
2409
+ },
2410
+ {
2411
+ "epoch": 2.31,
2412
+ "learning_rate": 1.1554686301579997e-05,
2413
+ "loss": 0.5042,
2414
+ "step": 200500
2415
+ },
2416
+ {
2417
+ "epoch": 2.31,
2418
+ "learning_rate": 1.145881270133456e-05,
2419
+ "loss": 0.4996,
2420
+ "step": 201000
2421
+ },
2422
+ {
2423
+ "epoch": 2.32,
2424
+ "learning_rate": 1.1362939101089124e-05,
2425
+ "loss": 0.5076,
2426
+ "step": 201500
2427
+ },
2428
+ {
2429
+ "epoch": 2.32,
2430
+ "learning_rate": 1.1267065500843687e-05,
2431
+ "loss": 0.4997,
2432
+ "step": 202000
2433
+ },
2434
+ {
2435
+ "epoch": 2.33,
2436
+ "learning_rate": 1.1171191900598251e-05,
2437
+ "loss": 0.5052,
2438
+ "step": 202500
2439
+ },
2440
+ {
2441
+ "epoch": 2.34,
2442
+ "learning_rate": 1.1075318300352816e-05,
2443
+ "loss": 0.5025,
2444
+ "step": 203000
2445
+ },
2446
+ {
2447
+ "epoch": 2.34,
2448
+ "learning_rate": 1.0979444700107379e-05,
2449
+ "loss": 0.5061,
2450
+ "step": 203500
2451
+ },
2452
+ {
2453
+ "epoch": 2.35,
2454
+ "learning_rate": 1.0883571099861943e-05,
2455
+ "loss": 0.511,
2456
+ "step": 204000
2457
+ },
2458
+ {
2459
+ "epoch": 2.35,
2460
+ "learning_rate": 1.0787697499616506e-05,
2461
+ "loss": 0.5031,
2462
+ "step": 204500
2463
+ },
2464
+ {
2465
+ "epoch": 2.36,
2466
+ "learning_rate": 1.069182389937107e-05,
2467
+ "loss": 0.5071,
2468
+ "step": 205000
2469
+ },
2470
+ {
2471
+ "epoch": 2.36,
2472
+ "learning_rate": 1.0595950299125633e-05,
2473
+ "loss": 0.5003,
2474
+ "step": 205500
2475
+ },
2476
+ {
2477
+ "epoch": 2.37,
2478
+ "learning_rate": 1.0500076698880197e-05,
2479
+ "loss": 0.4956,
2480
+ "step": 206000
2481
+ },
2482
+ {
2483
+ "epoch": 2.38,
2484
+ "learning_rate": 1.040420309863476e-05,
2485
+ "loss": 0.5051,
2486
+ "step": 206500
2487
+ },
2488
+ {
2489
+ "epoch": 2.38,
2490
+ "learning_rate": 1.0308329498389324e-05,
2491
+ "loss": 0.5016,
2492
+ "step": 207000
2493
+ },
2494
+ {
2495
+ "epoch": 2.39,
2496
+ "learning_rate": 1.0212455898143887e-05,
2497
+ "loss": 0.5055,
2498
+ "step": 207500
2499
+ },
2500
+ {
2501
+ "epoch": 2.39,
2502
+ "learning_rate": 1.0116582297898452e-05,
2503
+ "loss": 0.5049,
2504
+ "step": 208000
2505
+ },
2506
+ {
2507
+ "epoch": 2.4,
2508
+ "learning_rate": 1.0020708697653016e-05,
2509
+ "loss": 0.5079,
2510
+ "step": 208500
2511
+ },
2512
+ {
2513
+ "epoch": 2.4,
2514
+ "learning_rate": 9.924835097407579e-06,
2515
+ "loss": 0.5163,
2516
+ "step": 209000
2517
+ },
2518
+ {
2519
+ "epoch": 2.41,
2520
+ "learning_rate": 9.828961497162143e-06,
2521
+ "loss": 0.5015,
2522
+ "step": 209500
2523
+ },
2524
+ {
2525
+ "epoch": 2.42,
2526
+ "learning_rate": 9.733087896916706e-06,
2527
+ "loss": 0.5029,
2528
+ "step": 210000
2529
+ },
2530
+ {
2531
+ "epoch": 2.42,
2532
+ "learning_rate": 9.63721429667127e-06,
2533
+ "loss": 0.5034,
2534
+ "step": 210500
2535
+ },
2536
+ {
2537
+ "epoch": 2.43,
2538
+ "learning_rate": 9.541340696425833e-06,
2539
+ "loss": 0.5001,
2540
+ "step": 211000
2541
+ },
2542
+ {
2543
+ "epoch": 2.43,
2544
+ "learning_rate": 9.445467096180396e-06,
2545
+ "loss": 0.5036,
2546
+ "step": 211500
2547
+ },
2548
+ {
2549
+ "epoch": 2.44,
2550
+ "learning_rate": 9.34959349593496e-06,
2551
+ "loss": 0.5091,
2552
+ "step": 212000
2553
+ },
2554
+ {
2555
+ "epoch": 2.44,
2556
+ "learning_rate": 9.253719895689523e-06,
2557
+ "loss": 0.505,
2558
+ "step": 212500
2559
+ },
2560
+ {
2561
+ "epoch": 2.45,
2562
+ "learning_rate": 9.157846295444086e-06,
2563
+ "loss": 0.4992,
2564
+ "step": 213000
2565
+ },
2566
+ {
2567
+ "epoch": 2.46,
2568
+ "learning_rate": 9.06197269519865e-06,
2569
+ "loss": 0.5052,
2570
+ "step": 213500
2571
+ },
2572
+ {
2573
+ "epoch": 2.46,
2574
+ "learning_rate": 8.966099094953214e-06,
2575
+ "loss": 0.4838,
2576
+ "step": 214000
2577
+ },
2578
+ {
2579
+ "epoch": 2.47,
2580
+ "learning_rate": 8.870225494707777e-06,
2581
+ "loss": 0.503,
2582
+ "step": 214500
2583
+ },
2584
+ {
2585
+ "epoch": 2.47,
2586
+ "learning_rate": 8.774351894462341e-06,
2587
+ "loss": 0.509,
2588
+ "step": 215000
2589
+ },
2590
+ {
2591
+ "epoch": 2.48,
2592
+ "learning_rate": 8.678478294216904e-06,
2593
+ "loss": 0.4943,
2594
+ "step": 215500
2595
+ },
2596
+ {
2597
+ "epoch": 2.49,
2598
+ "learning_rate": 8.582604693971469e-06,
2599
+ "loss": 0.5004,
2600
+ "step": 216000
2601
+ },
2602
+ {
2603
+ "epoch": 2.49,
2604
+ "learning_rate": 8.486731093726031e-06,
2605
+ "loss": 0.494,
2606
+ "step": 216500
2607
+ },
2608
+ {
2609
+ "epoch": 2.5,
2610
+ "learning_rate": 8.390857493480596e-06,
2611
+ "loss": 0.4979,
2612
+ "step": 217000
2613
+ },
2614
+ {
2615
+ "epoch": 2.5,
2616
+ "learning_rate": 8.294983893235158e-06,
2617
+ "loss": 0.506,
2618
+ "step": 217500
2619
+ },
2620
+ {
2621
+ "epoch": 2.51,
2622
+ "learning_rate": 8.199110292989723e-06,
2623
+ "loss": 0.5048,
2624
+ "step": 218000
2625
+ },
2626
+ {
2627
+ "epoch": 2.51,
2628
+ "learning_rate": 8.103236692744286e-06,
2629
+ "loss": 0.5033,
2630
+ "step": 218500
2631
+ },
2632
+ {
2633
+ "epoch": 2.52,
2634
+ "learning_rate": 8.00736309249885e-06,
2635
+ "loss": 0.4989,
2636
+ "step": 219000
2637
+ },
2638
+ {
2639
+ "epoch": 2.53,
2640
+ "learning_rate": 7.911489492253414e-06,
2641
+ "loss": 0.4928,
2642
+ "step": 219500
2643
+ },
2644
+ {
2645
+ "epoch": 2.53,
2646
+ "learning_rate": 7.815615892007977e-06,
2647
+ "loss": 0.4981,
2648
+ "step": 220000
2649
+ },
2650
+ {
2651
+ "epoch": 2.54,
2652
+ "learning_rate": 7.719742291762542e-06,
2653
+ "loss": 0.505,
2654
+ "step": 220500
2655
+ },
2656
+ {
2657
+ "epoch": 2.54,
2658
+ "learning_rate": 7.623868691517104e-06,
2659
+ "loss": 0.502,
2660
+ "step": 221000
2661
+ },
2662
+ {
2663
+ "epoch": 2.55,
2664
+ "learning_rate": 7.527995091271668e-06,
2665
+ "loss": 0.4872,
2666
+ "step": 221500
2667
+ },
2668
+ {
2669
+ "epoch": 2.55,
2670
+ "learning_rate": 7.432121491026231e-06,
2671
+ "loss": 0.4967,
2672
+ "step": 222000
2673
+ },
2674
+ {
2675
+ "epoch": 2.56,
2676
+ "learning_rate": 7.336247890780795e-06,
2677
+ "loss": 0.5011,
2678
+ "step": 222500
2679
+ },
2680
+ {
2681
+ "epoch": 2.57,
2682
+ "learning_rate": 7.240374290535358e-06,
2683
+ "loss": 0.5026,
2684
+ "step": 223000
2685
+ },
2686
+ {
2687
+ "epoch": 2.57,
2688
+ "learning_rate": 7.144500690289922e-06,
2689
+ "loss": 0.5013,
2690
+ "step": 223500
2691
+ },
2692
+ {
2693
+ "epoch": 2.58,
2694
+ "learning_rate": 7.048627090044485e-06,
2695
+ "loss": 0.4961,
2696
+ "step": 224000
2697
+ },
2698
+ {
2699
+ "epoch": 2.58,
2700
+ "learning_rate": 6.952753489799049e-06,
2701
+ "loss": 0.4968,
2702
+ "step": 224500
2703
+ },
2704
+ {
2705
+ "epoch": 2.59,
2706
+ "learning_rate": 6.856879889553614e-06,
2707
+ "loss": 0.4882,
2708
+ "step": 225000
2709
+ },
2710
+ {
2711
+ "epoch": 2.59,
2712
+ "learning_rate": 6.761006289308176e-06,
2713
+ "loss": 0.4921,
2714
+ "step": 225500
2715
+ },
2716
+ {
2717
+ "epoch": 2.6,
2718
+ "learning_rate": 6.665132689062741e-06,
2719
+ "loss": 0.4931,
2720
+ "step": 226000
2721
+ },
2722
+ {
2723
+ "epoch": 2.61,
2724
+ "learning_rate": 6.5692590888173035e-06,
2725
+ "loss": 0.4924,
2726
+ "step": 226500
2727
+ },
2728
+ {
2729
+ "epoch": 2.61,
2730
+ "learning_rate": 6.473385488571868e-06,
2731
+ "loss": 0.5035,
2732
+ "step": 227000
2733
+ },
2734
+ {
2735
+ "epoch": 2.62,
2736
+ "learning_rate": 6.377511888326431e-06,
2737
+ "loss": 0.4953,
2738
+ "step": 227500
2739
+ },
2740
+ {
2741
+ "epoch": 2.62,
2742
+ "learning_rate": 6.281638288080994e-06,
2743
+ "loss": 0.4973,
2744
+ "step": 228000
2745
+ },
2746
+ {
2747
+ "epoch": 2.63,
2748
+ "learning_rate": 6.185764687835558e-06,
2749
+ "loss": 0.4982,
2750
+ "step": 228500
2751
+ },
2752
+ {
2753
+ "epoch": 2.63,
2754
+ "learning_rate": 6.089891087590121e-06,
2755
+ "loss": 0.4866,
2756
+ "step": 229000
2757
+ },
2758
+ {
2759
+ "epoch": 2.64,
2760
+ "learning_rate": 5.994017487344685e-06,
2761
+ "loss": 0.5021,
2762
+ "step": 229500
2763
+ },
2764
+ {
2765
+ "epoch": 2.65,
2766
+ "learning_rate": 5.8981438870992485e-06,
2767
+ "loss": 0.4912,
2768
+ "step": 230000
2769
+ },
2770
+ {
2771
+ "epoch": 2.65,
2772
+ "learning_rate": 5.802270286853812e-06,
2773
+ "loss": 0.4966,
2774
+ "step": 230500
2775
+ },
2776
+ {
2777
+ "epoch": 2.66,
2778
+ "learning_rate": 5.706396686608376e-06,
2779
+ "loss": 0.4913,
2780
+ "step": 231000
2781
+ },
2782
+ {
2783
+ "epoch": 2.66,
2784
+ "learning_rate": 5.610523086362939e-06,
2785
+ "loss": 0.4875,
2786
+ "step": 231500
2787
+ },
2788
+ {
2789
+ "epoch": 2.67,
2790
+ "learning_rate": 5.514649486117503e-06,
2791
+ "loss": 0.4958,
2792
+ "step": 232000
2793
+ },
2794
+ {
2795
+ "epoch": 2.67,
2796
+ "learning_rate": 5.418775885872066e-06,
2797
+ "loss": 0.496,
2798
+ "step": 232500
2799
+ },
2800
+ {
2801
+ "epoch": 2.68,
2802
+ "learning_rate": 5.322902285626631e-06,
2803
+ "loss": 0.4889,
2804
+ "step": 233000
2805
+ },
2806
+ {
2807
+ "epoch": 2.69,
2808
+ "learning_rate": 5.227028685381194e-06,
2809
+ "loss": 0.4899,
2810
+ "step": 233500
2811
+ },
2812
+ {
2813
+ "epoch": 2.69,
2814
+ "learning_rate": 5.131155085135757e-06,
2815
+ "loss": 0.4972,
2816
+ "step": 234000
2817
+ },
2818
+ {
2819
+ "epoch": 2.7,
2820
+ "learning_rate": 5.035281484890321e-06,
2821
+ "loss": 0.5017,
2822
+ "step": 234500
2823
+ },
2824
+ {
2825
+ "epoch": 2.7,
2826
+ "learning_rate": 4.939407884644884e-06,
2827
+ "loss": 0.4922,
2828
+ "step": 235000
2829
+ },
2830
+ {
2831
+ "epoch": 2.71,
2832
+ "learning_rate": 4.843534284399448e-06,
2833
+ "loss": 0.5002,
2834
+ "step": 235500
2835
+ },
2836
+ {
2837
+ "epoch": 2.72,
2838
+ "learning_rate": 4.747660684154011e-06,
2839
+ "loss": 0.4961,
2840
+ "step": 236000
2841
+ },
2842
+ {
2843
+ "epoch": 2.72,
2844
+ "learning_rate": 4.651787083908575e-06,
2845
+ "loss": 0.4859,
2846
+ "step": 236500
2847
+ },
2848
+ {
2849
+ "epoch": 2.73,
2850
+ "learning_rate": 4.5559134836631385e-06,
2851
+ "loss": 0.4967,
2852
+ "step": 237000
2853
+ },
2854
+ {
2855
+ "epoch": 2.73,
2856
+ "learning_rate": 4.460039883417702e-06,
2857
+ "loss": 0.5026,
2858
+ "step": 237500
2859
+ },
2860
+ {
2861
+ "epoch": 2.74,
2862
+ "learning_rate": 4.364166283172266e-06,
2863
+ "loss": 0.4883,
2864
+ "step": 238000
2865
+ },
2866
+ {
2867
+ "epoch": 2.74,
2868
+ "learning_rate": 4.26829268292683e-06,
2869
+ "loss": 0.4866,
2870
+ "step": 238500
2871
+ },
2872
+ {
2873
+ "epoch": 2.75,
2874
+ "learning_rate": 4.172419082681394e-06,
2875
+ "loss": 0.4926,
2876
+ "step": 239000
2877
+ },
2878
+ {
2879
+ "epoch": 2.76,
2880
+ "learning_rate": 4.076545482435957e-06,
2881
+ "loss": 0.4991,
2882
+ "step": 239500
2883
+ },
2884
+ {
2885
+ "epoch": 2.76,
2886
+ "learning_rate": 3.980671882190521e-06,
2887
+ "loss": 0.4936,
2888
+ "step": 240000
2889
+ },
2890
+ {
2891
+ "epoch": 2.77,
2892
+ "learning_rate": 3.8847982819450834e-06,
2893
+ "loss": 0.4907,
2894
+ "step": 240500
2895
+ },
2896
+ {
2897
+ "epoch": 2.77,
2898
+ "learning_rate": 3.7889246816996474e-06,
2899
+ "loss": 0.4962,
2900
+ "step": 241000
2901
+ },
2902
+ {
2903
+ "epoch": 2.78,
2904
+ "learning_rate": 3.693051081454211e-06,
2905
+ "loss": 0.4909,
2906
+ "step": 241500
2907
+ },
2908
+ {
2909
+ "epoch": 2.78,
2910
+ "learning_rate": 3.5971774812087746e-06,
2911
+ "loss": 0.4886,
2912
+ "step": 242000
2913
+ },
2914
+ {
2915
+ "epoch": 2.79,
2916
+ "learning_rate": 3.501303880963338e-06,
2917
+ "loss": 0.489,
2918
+ "step": 242500
2919
+ },
2920
+ {
2921
+ "epoch": 2.8,
2922
+ "learning_rate": 3.4054302807179013e-06,
2923
+ "loss": 0.486,
2924
+ "step": 243000
2925
+ },
2926
+ {
2927
+ "epoch": 2.8,
2928
+ "learning_rate": 3.309556680472465e-06,
2929
+ "loss": 0.4891,
2930
+ "step": 243500
2931
+ },
2932
+ {
2933
+ "epoch": 2.81,
2934
+ "learning_rate": 3.2136830802270293e-06,
2935
+ "loss": 0.4843,
2936
+ "step": 244000
2937
+ },
2938
+ {
2939
+ "epoch": 2.81,
2940
+ "learning_rate": 3.1178094799815924e-06,
2941
+ "loss": 0.4943,
2942
+ "step": 244500
2943
+ },
2944
+ {
2945
+ "epoch": 2.82,
2946
+ "learning_rate": 3.021935879736156e-06,
2947
+ "loss": 0.4828,
2948
+ "step": 245000
2949
+ },
2950
+ {
2951
+ "epoch": 2.82,
2952
+ "learning_rate": 2.9260622794907195e-06,
2953
+ "loss": 0.4961,
2954
+ "step": 245500
2955
+ },
2956
+ {
2957
+ "epoch": 2.83,
2958
+ "learning_rate": 2.830188679245283e-06,
2959
+ "loss": 0.4975,
2960
+ "step": 246000
2961
+ },
2962
+ {
2963
+ "epoch": 2.84,
2964
+ "learning_rate": 2.7343150789998467e-06,
2965
+ "loss": 0.4924,
2966
+ "step": 246500
2967
+ },
2968
+ {
2969
+ "epoch": 2.84,
2970
+ "learning_rate": 2.6384414787544102e-06,
2971
+ "loss": 0.4941,
2972
+ "step": 247000
2973
+ },
2974
+ {
2975
+ "epoch": 2.85,
2976
+ "learning_rate": 2.542567878508974e-06,
2977
+ "loss": 0.4773,
2978
+ "step": 247500
2979
+ },
2980
+ {
2981
+ "epoch": 2.85,
2982
+ "learning_rate": 2.446694278263538e-06,
2983
+ "loss": 0.4858,
2984
+ "step": 248000
2985
+ },
2986
+ {
2987
+ "epoch": 2.86,
2988
+ "learning_rate": 2.350820678018101e-06,
2989
+ "loss": 0.4967,
2990
+ "step": 248500
2991
+ },
2992
+ {
2993
+ "epoch": 2.86,
2994
+ "learning_rate": 2.2549470777726645e-06,
2995
+ "loss": 0.4899,
2996
+ "step": 249000
2997
+ },
2998
+ {
2999
+ "epoch": 2.87,
3000
+ "learning_rate": 2.159073477527228e-06,
3001
+ "loss": 0.4959,
3002
+ "step": 249500
3003
+ },
3004
+ {
3005
+ "epoch": 2.88,
3006
+ "learning_rate": 2.0631998772817916e-06,
3007
+ "loss": 0.4907,
3008
+ "step": 250000
3009
+ },
3010
+ {
3011
+ "epoch": 2.88,
3012
+ "learning_rate": 1.967326277036355e-06,
3013
+ "loss": 0.4921,
3014
+ "step": 250500
3015
+ },
3016
+ {
3017
+ "epoch": 2.89,
3018
+ "learning_rate": 1.871452676790919e-06,
3019
+ "loss": 0.4896,
3020
+ "step": 251000
3021
+ },
3022
+ {
3023
+ "epoch": 2.89,
3024
+ "learning_rate": 1.7755790765454826e-06,
3025
+ "loss": 0.4888,
3026
+ "step": 251500
3027
+ },
3028
+ {
3029
+ "epoch": 2.9,
3030
+ "learning_rate": 1.6797054763000461e-06,
3031
+ "loss": 0.4903,
3032
+ "step": 252000
3033
+ },
3034
+ {
3035
+ "epoch": 2.9,
3036
+ "learning_rate": 1.5838318760546097e-06,
3037
+ "loss": 0.4881,
3038
+ "step": 252500
3039
+ },
3040
+ {
3041
+ "epoch": 2.91,
3042
+ "learning_rate": 1.4879582758091733e-06,
3043
+ "loss": 0.4895,
3044
+ "step": 253000
3045
+ },
3046
+ {
3047
+ "epoch": 2.92,
3048
+ "learning_rate": 1.3920846755637368e-06,
3049
+ "loss": 0.4815,
3050
+ "step": 253500
3051
+ },
3052
+ {
3053
+ "epoch": 2.92,
3054
+ "learning_rate": 1.2962110753183004e-06,
3055
+ "loss": 0.4906,
3056
+ "step": 254000
3057
+ },
3058
+ {
3059
+ "epoch": 2.93,
3060
+ "learning_rate": 1.200337475072864e-06,
3061
+ "loss": 0.4936,
3062
+ "step": 254500
3063
+ },
3064
+ {
3065
+ "epoch": 2.93,
3066
+ "learning_rate": 1.1044638748274275e-06,
3067
+ "loss": 0.4753,
3068
+ "step": 255000
3069
+ },
3070
+ {
3071
+ "epoch": 2.94,
3072
+ "learning_rate": 1.0085902745819913e-06,
3073
+ "loss": 0.4883,
3074
+ "step": 255500
3075
+ },
3076
+ {
3077
+ "epoch": 2.95,
3078
+ "learning_rate": 9.127166743365548e-07,
3079
+ "loss": 0.4933,
3080
+ "step": 256000
3081
+ },
3082
+ {
3083
+ "epoch": 2.95,
3084
+ "learning_rate": 8.168430740911182e-07,
3085
+ "loss": 0.4831,
3086
+ "step": 256500
3087
+ },
3088
+ {
3089
+ "epoch": 2.96,
3090
+ "learning_rate": 7.209694738456819e-07,
3091
+ "loss": 0.4861,
3092
+ "step": 257000
3093
+ },
3094
+ {
3095
+ "epoch": 2.96,
3096
+ "learning_rate": 6.250958736002455e-07,
3097
+ "loss": 0.4806,
3098
+ "step": 257500
3099
+ },
3100
+ {
3101
+ "epoch": 2.97,
3102
+ "learning_rate": 5.29222273354809e-07,
3103
+ "loss": 0.4767,
3104
+ "step": 258000
3105
+ },
3106
+ {
3107
+ "epoch": 2.97,
3108
+ "learning_rate": 4.333486731093726e-07,
3109
+ "loss": 0.4845,
3110
+ "step": 258500
3111
+ },
3112
+ {
3113
+ "epoch": 2.98,
3114
+ "learning_rate": 3.3747507286393617e-07,
3115
+ "loss": 0.4924,
3116
+ "step": 259000
3117
+ },
3118
+ {
3119
+ "epoch": 2.99,
3120
+ "learning_rate": 2.416014726184998e-07,
3121
+ "loss": 0.4937,
3122
+ "step": 259500
3123
+ },
3124
+ {
3125
+ "epoch": 2.99,
3126
+ "learning_rate": 1.4572787237306336e-07,
3127
+ "loss": 0.4803,
3128
+ "step": 260000
3129
+ },
3130
+ {
3131
+ "epoch": 3.0,
3132
+ "learning_rate": 4.985427212762693e-08,
3133
+ "loss": 0.4802,
3134
+ "step": 260500
3135
+ },
3136
+ {
3137
+ "epoch": 3.0,
3138
+ "step": 260760,
3139
+ "total_flos": 2.990387392931758e+17,
3140
+ "train_loss": 0.11540543585032806,
3141
+ "train_runtime": 18089.1482,
3142
+ "train_samples_per_second": 144.153,
3143
+ "train_steps_per_second": 14.415
3144
+ }
3145
+ ],
3146
+ "max_steps": 260760,
3147
+ "num_train_epochs": 3,
3148
+ "total_flos": 2.990387392931758e+17,
3149
+ "trial_name": null,
3150
+ "trial_params": null
3151
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ec330993befe2ae284311862967bbdf6be39d3ecb72dce07dc1ce2ae83842e9
3
+ size 2799
vocab.json ADDED
The diff for this file is too large to render. See raw diff