Hugging Face's logo

masakhane
/
m2m100_418M_en_tsn_rel_news

Text2Text Generation
Transformers
PyTorch
English
Tswana
m2m_100
Inference Endpoints
Model card Files Community
2
Davlan commited on
Commit
d2ad774
•
1 Parent(s): 69d4e66

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 +3157 -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": 28.4632,
4
+ "eval_gen_len": 52.3759,
5
+ "eval_loss": 1.208052158355713,
6
+ "eval_runtime": 216.6704,
7
+ "eval_samples": 540,
8
+ "eval_samples_per_second": 2.492,
9
+ "eval_steps_per_second": 0.249,
10
+ "predict_bleu": 28.9592,
11
+ "predict_gen_len": 50.6453,
12
+ "predict_loss": 1.1930155754089355,
13
+ "predict_runtime": 601.7871,
14
+ "predict_samples": 1500,
15
+ "predict_samples_per_second": 2.493,
16
+ "predict_steps_per_second": 0.249,
17
+ "train_loss": 0.6627257539563876,
18
+ "train_runtime": 78487.0606,
19
+ "train_samples": 871299,
20
+ "train_samples_per_second": 33.304,
21
+ "train_steps_per_second": 3.33
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": 28.4632,
4
+ "eval_gen_len": 52.3759,
5
+ "eval_loss": 1.208052158355713,
6
+ "eval_runtime": 216.6704,
7
+ "eval_samples": 540,
8
+ "eval_samples_per_second": 2.492,
9
+ "eval_steps_per_second": 0.249
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": 28.9592,
3
+ "predict_gen_len": 50.6453,
4
+ "predict_loss": 1.1930155754089355,
5
+ "predict_runtime": 601.7871,
6
+ "predict_samples": 1500,
7
+ "predict_samples_per_second": 2.493,
8
+ "predict_steps_per_second": 0.249
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd41ea46b6cc5091868f641758770a3ea9337da204be9ccc402d467835e5dd66
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": 0.6627257539563876,
4
+ "train_runtime": 78487.0606,
5
+ "train_samples": 871299,
6
+ "train_samples_per_second": 33.304,
7
+ "train_steps_per_second": 3.33
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3157 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 261390,
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": 4.990435747350702e-05,
13
+ "loss": 3.0142,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 4.980871494701404e-05,
19
+ "loss": 2.0464,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 4.971307242052106e-05,
25
+ "loss": 1.7157,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 4.9617429894028086e-05,
31
+ "loss": 1.5985,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 4.9521787367535106e-05,
37
+ "loss": 1.5032,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 4.942614484104212e-05,
43
+ "loss": 1.4132,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.04,
48
+ "learning_rate": 4.9330502314549144e-05,
49
+ "loss": 1.3507,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 4.9234859788056164e-05,
55
+ "loss": 1.3238,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.05,
60
+ "learning_rate": 4.913921726156318e-05,
61
+ "loss": 1.2726,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.06,
66
+ "learning_rate": 4.90435747350702e-05,
67
+ "loss": 1.2343,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.06,
72
+ "learning_rate": 4.894793220857722e-05,
73
+ "loss": 1.2175,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.07,
78
+ "learning_rate": 4.885228968208425e-05,
79
+ "loss": 1.1914,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.07,
84
+ "learning_rate": 4.875664715559126e-05,
85
+ "loss": 1.1483,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.08,
90
+ "learning_rate": 4.8661004629098287e-05,
91
+ "loss": 1.138,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.09,
96
+ "learning_rate": 4.8565362102605306e-05,
97
+ "loss": 1.124,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.09,
102
+ "learning_rate": 4.8469719576112325e-05,
103
+ "loss": 1.1216,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.1,
108
+ "learning_rate": 4.8374077049619345e-05,
109
+ "loss": 1.0871,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.1,
114
+ "learning_rate": 4.8278434523126364e-05,
115
+ "loss": 1.1015,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.11,
120
+ "learning_rate": 4.818279199663338e-05,
121
+ "loss": 1.0644,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.11,
126
+ "learning_rate": 4.808714947014041e-05,
127
+ "loss": 1.0549,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.12,
132
+ "learning_rate": 4.799150694364742e-05,
133
+ "loss": 1.0561,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.13,
138
+ "learning_rate": 4.789586441715445e-05,
139
+ "loss": 1.0315,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.13,
144
+ "learning_rate": 4.780022189066147e-05,
145
+ "loss": 1.0248,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.14,
150
+ "learning_rate": 4.7704579364168487e-05,
151
+ "loss": 1.0313,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.14,
156
+ "learning_rate": 4.7608936837675506e-05,
157
+ "loss": 1.0072,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.15,
162
+ "learning_rate": 4.7513294311182525e-05,
163
+ "loss": 1.0038,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.15,
168
+ "learning_rate": 4.741765178468955e-05,
169
+ "loss": 1.0004,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.16,
174
+ "learning_rate": 4.7322009258196564e-05,
175
+ "loss": 1.001,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.17,
180
+ "learning_rate": 4.722636673170358e-05,
181
+ "loss": 0.9823,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.17,
186
+ "learning_rate": 4.713072420521061e-05,
187
+ "loss": 0.9753,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.18,
192
+ "learning_rate": 4.703508167871763e-05,
193
+ "loss": 0.978,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.18,
198
+ "learning_rate": 4.693943915222465e-05,
199
+ "loss": 0.9699,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.19,
204
+ "learning_rate": 4.684379662573167e-05,
205
+ "loss": 0.9432,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.2,
210
+ "learning_rate": 4.674815409923869e-05,
211
+ "loss": 0.9581,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.2,
216
+ "learning_rate": 4.6652511572745706e-05,
217
+ "loss": 0.9526,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.21,
222
+ "learning_rate": 4.6556869046252725e-05,
223
+ "loss": 0.9385,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.21,
228
+ "learning_rate": 4.646122651975975e-05,
229
+ "loss": 0.9398,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.22,
234
+ "learning_rate": 4.636558399326677e-05,
235
+ "loss": 0.9368,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.22,
240
+ "learning_rate": 4.626994146677378e-05,
241
+ "loss": 0.9213,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.23,
246
+ "learning_rate": 4.617429894028081e-05,
247
+ "loss": 0.9399,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.24,
252
+ "learning_rate": 4.607865641378783e-05,
253
+ "loss": 0.9326,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.24,
258
+ "learning_rate": 4.598301388729485e-05,
259
+ "loss": 0.9334,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.25,
264
+ "learning_rate": 4.588737136080187e-05,
265
+ "loss": 0.9183,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.25,
270
+ "learning_rate": 4.579172883430889e-05,
271
+ "loss": 0.9175,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.26,
276
+ "learning_rate": 4.569608630781591e-05,
277
+ "loss": 0.9083,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.26,
282
+ "learning_rate": 4.560044378132293e-05,
283
+ "loss": 0.9111,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.27,
288
+ "learning_rate": 4.550480125482995e-05,
289
+ "loss": 0.8906,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.28,
294
+ "learning_rate": 4.540915872833697e-05,
295
+ "loss": 0.8879,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.28,
300
+ "learning_rate": 4.531351620184399e-05,
301
+ "loss": 0.8914,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.29,
306
+ "learning_rate": 4.521787367535101e-05,
307
+ "loss": 0.8783,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.29,
312
+ "learning_rate": 4.512223114885803e-05,
313
+ "loss": 0.879,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.3,
318
+ "learning_rate": 4.502658862236505e-05,
319
+ "loss": 0.8882,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.3,
324
+ "learning_rate": 4.4930946095872074e-05,
325
+ "loss": 0.8955,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.31,
330
+ "learning_rate": 4.483530356937909e-05,
331
+ "loss": 0.8973,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.32,
336
+ "learning_rate": 4.473966104288611e-05,
337
+ "loss": 0.8785,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.32,
342
+ "learning_rate": 4.464401851639313e-05,
343
+ "loss": 0.87,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.33,
348
+ "learning_rate": 4.454837598990015e-05,
349
+ "loss": 0.8668,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.33,
354
+ "learning_rate": 4.445273346340717e-05,
355
+ "loss": 0.8644,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.34,
360
+ "learning_rate": 4.435709093691419e-05,
361
+ "loss": 0.8532,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.34,
366
+ "learning_rate": 4.4261448410421216e-05,
367
+ "loss": 0.8529,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.35,
372
+ "learning_rate": 4.416580588392823e-05,
373
+ "loss": 0.8454,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.36,
378
+ "learning_rate": 4.407016335743525e-05,
379
+ "loss": 0.8585,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.36,
384
+ "learning_rate": 4.3974520830942274e-05,
385
+ "loss": 0.8393,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.37,
390
+ "learning_rate": 4.3878878304449294e-05,
391
+ "loss": 0.8612,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.37,
396
+ "learning_rate": 4.378323577795631e-05,
397
+ "loss": 0.8448,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.38,
402
+ "learning_rate": 4.368759325146333e-05,
403
+ "loss": 0.839,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.38,
408
+ "learning_rate": 4.359195072497035e-05,
409
+ "loss": 0.8311,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.39,
414
+ "learning_rate": 4.349630819847738e-05,
415
+ "loss": 0.826,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.4,
420
+ "learning_rate": 4.340066567198439e-05,
421
+ "loss": 0.834,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.4,
426
+ "learning_rate": 4.3305023145491416e-05,
427
+ "loss": 0.8266,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.41,
432
+ "learning_rate": 4.3209380618998436e-05,
433
+ "loss": 0.8284,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.41,
438
+ "learning_rate": 4.311373809250545e-05,
439
+ "loss": 0.8308,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.42,
444
+ "learning_rate": 4.3018095566012474e-05,
445
+ "loss": 0.8156,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.42,
450
+ "learning_rate": 4.2922453039519494e-05,
451
+ "loss": 0.8217,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.43,
456
+ "learning_rate": 4.282681051302651e-05,
457
+ "loss": 0.8274,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.44,
462
+ "learning_rate": 4.273116798653353e-05,
463
+ "loss": 0.825,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.44,
468
+ "learning_rate": 4.263552546004055e-05,
469
+ "loss": 0.8207,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.45,
474
+ "learning_rate": 4.253988293354758e-05,
475
+ "loss": 0.8264,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.45,
480
+ "learning_rate": 4.24442404070546e-05,
481
+ "loss": 0.8192,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.46,
486
+ "learning_rate": 4.2348597880561616e-05,
487
+ "loss": 0.8085,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.46,
492
+ "learning_rate": 4.2252955354068636e-05,
493
+ "loss": 0.8223,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.47,
498
+ "learning_rate": 4.2157312827575655e-05,
499
+ "loss": 0.8004,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.48,
504
+ "learning_rate": 4.2061670301082674e-05,
505
+ "loss": 0.8074,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.48,
510
+ "learning_rate": 4.1966027774589694e-05,
511
+ "loss": 0.813,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.49,
516
+ "learning_rate": 4.187038524809671e-05,
517
+ "loss": 0.7991,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.49,
522
+ "learning_rate": 4.177474272160374e-05,
523
+ "loss": 0.7942,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.5,
528
+ "learning_rate": 4.167910019511075e-05,
529
+ "loss": 0.8021,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.5,
534
+ "learning_rate": 4.158345766861778e-05,
535
+ "loss": 0.8121,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.51,
540
+ "learning_rate": 4.14878151421248e-05,
541
+ "loss": 0.802,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.52,
546
+ "learning_rate": 4.1392172615631816e-05,
547
+ "loss": 0.7872,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.52,
552
+ "learning_rate": 4.1296530089138836e-05,
553
+ "loss": 0.8041,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.53,
558
+ "learning_rate": 4.1200887562645855e-05,
559
+ "loss": 0.7901,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.53,
564
+ "learning_rate": 4.110524503615288e-05,
565
+ "loss": 0.7762,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.54,
570
+ "learning_rate": 4.1009602509659894e-05,
571
+ "loss": 0.7955,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.55,
576
+ "learning_rate": 4.091395998316691e-05,
577
+ "loss": 0.7851,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.55,
582
+ "learning_rate": 4.081831745667394e-05,
583
+ "loss": 0.7947,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.56,
588
+ "learning_rate": 4.072267493018096e-05,
589
+ "loss": 0.7956,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.56,
594
+ "learning_rate": 4.062703240368798e-05,
595
+ "loss": 0.7767,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.57,
600
+ "learning_rate": 4.0531389877195e-05,
601
+ "loss": 0.7884,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.57,
606
+ "learning_rate": 4.0435747350702016e-05,
607
+ "loss": 0.78,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.58,
612
+ "learning_rate": 4.034010482420904e-05,
613
+ "loss": 0.782,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.59,
618
+ "learning_rate": 4.0244462297716055e-05,
619
+ "loss": 0.783,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.59,
624
+ "learning_rate": 4.014881977122308e-05,
625
+ "loss": 0.769,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.6,
630
+ "learning_rate": 4.00531772447301e-05,
631
+ "loss": 0.7708,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.6,
636
+ "learning_rate": 3.995753471823711e-05,
637
+ "loss": 0.7611,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.61,
642
+ "learning_rate": 3.986189219174414e-05,
643
+ "loss": 0.7738,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.61,
648
+ "learning_rate": 3.976624966525116e-05,
649
+ "loss": 0.7758,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.62,
654
+ "learning_rate": 3.967060713875818e-05,
655
+ "loss": 0.7804,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.63,
660
+ "learning_rate": 3.95749646122652e-05,
661
+ "loss": 0.7703,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.63,
666
+ "learning_rate": 3.9479322085772217e-05,
667
+ "loss": 0.7743,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.64,
672
+ "learning_rate": 3.938367955927924e-05,
673
+ "loss": 0.7666,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.64,
678
+ "learning_rate": 3.928803703278626e-05,
679
+ "loss": 0.7633,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.65,
684
+ "learning_rate": 3.919239450629328e-05,
685
+ "loss": 0.7707,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.65,
690
+ "learning_rate": 3.90967519798003e-05,
691
+ "loss": 0.7553,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.66,
696
+ "learning_rate": 3.900110945330732e-05,
697
+ "loss": 0.7537,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.67,
702
+ "learning_rate": 3.890546692681434e-05,
703
+ "loss": 0.754,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.67,
708
+ "learning_rate": 3.880982440032136e-05,
709
+ "loss": 0.7448,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.68,
714
+ "learning_rate": 3.871418187382838e-05,
715
+ "loss": 0.747,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.68,
720
+ "learning_rate": 3.8618539347335404e-05,
721
+ "loss": 0.7593,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.69,
726
+ "learning_rate": 3.8522896820842417e-05,
727
+ "loss": 0.7612,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.69,
732
+ "learning_rate": 3.842725429434944e-05,
733
+ "loss": 0.7512,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.7,
738
+ "learning_rate": 3.833161176785646e-05,
739
+ "loss": 0.7438,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.71,
744
+ "learning_rate": 3.823596924136348e-05,
745
+ "loss": 0.7528,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.71,
750
+ "learning_rate": 3.81403267148705e-05,
751
+ "loss": 0.7306,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.72,
756
+ "learning_rate": 3.804468418837752e-05,
757
+ "loss": 0.7493,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.72,
762
+ "learning_rate": 3.7949041661884546e-05,
763
+ "loss": 0.7472,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.73,
768
+ "learning_rate": 3.785339913539156e-05,
769
+ "loss": 0.7412,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.73,
774
+ "learning_rate": 3.775775660889858e-05,
775
+ "loss": 0.7538,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.74,
780
+ "learning_rate": 3.7662114082405604e-05,
781
+ "loss": 0.7316,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 0.75,
786
+ "learning_rate": 3.7566471555912623e-05,
787
+ "loss": 0.754,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 0.75,
792
+ "learning_rate": 3.747082902941964e-05,
793
+ "loss": 0.7274,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 0.76,
798
+ "learning_rate": 3.737518650292666e-05,
799
+ "loss": 0.7258,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 0.76,
804
+ "learning_rate": 3.727954397643368e-05,
805
+ "loss": 0.7351,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 0.77,
810
+ "learning_rate": 3.718390144994071e-05,
811
+ "loss": 0.7626,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 0.77,
816
+ "learning_rate": 3.708825892344772e-05,
817
+ "loss": 0.7364,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 0.78,
822
+ "learning_rate": 3.6992616396954746e-05,
823
+ "loss": 0.7374,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 0.79,
828
+ "learning_rate": 3.6896973870461766e-05,
829
+ "loss": 0.7237,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 0.79,
834
+ "learning_rate": 3.6801331343968785e-05,
835
+ "loss": 0.7444,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 0.8,
840
+ "learning_rate": 3.6705688817475804e-05,
841
+ "loss": 0.7336,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 0.8,
846
+ "learning_rate": 3.6610046290982823e-05,
847
+ "loss": 0.7209,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 0.81,
852
+ "learning_rate": 3.651440376448985e-05,
853
+ "loss": 0.7277,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 0.81,
858
+ "learning_rate": 3.641876123799686e-05,
859
+ "loss": 0.7329,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 0.82,
864
+ "learning_rate": 3.632311871150388e-05,
865
+ "loss": 0.7197,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 0.83,
870
+ "learning_rate": 3.622747618501091e-05,
871
+ "loss": 0.729,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 0.83,
876
+ "learning_rate": 3.613183365851793e-05,
877
+ "loss": 0.7243,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 0.84,
882
+ "learning_rate": 3.6036191132024946e-05,
883
+ "loss": 0.7297,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 0.84,
888
+ "learning_rate": 3.5940548605531966e-05,
889
+ "loss": 0.7192,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 0.85,
894
+ "learning_rate": 3.5844906079038985e-05,
895
+ "loss": 0.716,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 0.86,
900
+ "learning_rate": 3.574926355254601e-05,
901
+ "loss": 0.7183,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 0.86,
906
+ "learning_rate": 3.5653621026053024e-05,
907
+ "loss": 0.7101,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 0.87,
912
+ "learning_rate": 3.555797849956004e-05,
913
+ "loss": 0.7158,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 0.87,
918
+ "learning_rate": 3.546233597306707e-05,
919
+ "loss": 0.709,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 0.88,
924
+ "learning_rate": 3.536669344657408e-05,
925
+ "loss": 0.7308,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 0.88,
930
+ "learning_rate": 3.527105092008111e-05,
931
+ "loss": 0.7171,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 0.89,
936
+ "learning_rate": 3.517540839358813e-05,
937
+ "loss": 0.7185,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 0.9,
942
+ "learning_rate": 3.5079765867095146e-05,
943
+ "loss": 0.7236,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 0.9,
948
+ "learning_rate": 3.4984123340602166e-05,
949
+ "loss": 0.7006,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 0.91,
954
+ "learning_rate": 3.4888480814109185e-05,
955
+ "loss": 0.7201,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 0.91,
960
+ "learning_rate": 3.479283828761621e-05,
961
+ "loss": 0.7258,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 0.92,
966
+ "learning_rate": 3.469719576112323e-05,
967
+ "loss": 0.7117,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 0.92,
972
+ "learning_rate": 3.460155323463024e-05,
973
+ "loss": 0.7279,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 0.93,
978
+ "learning_rate": 3.450591070813727e-05,
979
+ "loss": 0.7085,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 0.94,
984
+ "learning_rate": 3.441026818164429e-05,
985
+ "loss": 0.7076,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 0.94,
990
+ "learning_rate": 3.431462565515131e-05,
991
+ "loss": 0.7101,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 0.95,
996
+ "learning_rate": 3.421898312865833e-05,
997
+ "loss": 0.7102,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 0.95,
1002
+ "learning_rate": 3.4123340602165346e-05,
1003
+ "loss": 0.7012,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 0.96,
1008
+ "learning_rate": 3.402769807567237e-05,
1009
+ "loss": 0.7119,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 0.96,
1014
+ "learning_rate": 3.3932055549179385e-05,
1015
+ "loss": 0.7079,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 0.97,
1020
+ "learning_rate": 3.383641302268641e-05,
1021
+ "loss": 0.7056,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 0.98,
1026
+ "learning_rate": 3.374077049619343e-05,
1027
+ "loss": 0.7021,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 0.98,
1032
+ "learning_rate": 3.364512796970045e-05,
1033
+ "loss": 0.7188,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 0.99,
1038
+ "learning_rate": 3.354948544320747e-05,
1039
+ "loss": 0.6963,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 0.99,
1044
+ "learning_rate": 3.345384291671449e-05,
1045
+ "loss": 0.6971,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 1.0,
1050
+ "learning_rate": 3.3358200390221515e-05,
1051
+ "loss": 0.6967,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 1.0,
1056
+ "learning_rate": 3.326255786372853e-05,
1057
+ "loss": 0.6435,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 1.01,
1062
+ "learning_rate": 3.3166915337235546e-05,
1063
+ "loss": 0.6471,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 1.02,
1068
+ "learning_rate": 3.307127281074257e-05,
1069
+ "loss": 0.6381,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 1.02,
1074
+ "learning_rate": 3.297563028424959e-05,
1075
+ "loss": 0.6394,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 1.03,
1080
+ "learning_rate": 3.287998775775661e-05,
1081
+ "loss": 0.6357,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 1.03,
1086
+ "learning_rate": 3.278434523126363e-05,
1087
+ "loss": 0.6419,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 1.04,
1092
+ "learning_rate": 3.268870270477065e-05,
1093
+ "loss": 0.642,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 1.04,
1098
+ "learning_rate": 3.2593060178277676e-05,
1099
+ "loss": 0.6415,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 1.05,
1104
+ "learning_rate": 3.249741765178469e-05,
1105
+ "loss": 0.6439,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 1.06,
1110
+ "learning_rate": 3.2401775125291715e-05,
1111
+ "loss": 0.6374,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 1.06,
1116
+ "learning_rate": 3.2306132598798734e-05,
1117
+ "loss": 0.6303,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 1.07,
1122
+ "learning_rate": 3.2210490072305746e-05,
1123
+ "loss": 0.643,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 1.07,
1128
+ "learning_rate": 3.211484754581277e-05,
1129
+ "loss": 0.6355,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 1.08,
1134
+ "learning_rate": 3.201920501931979e-05,
1135
+ "loss": 0.6401,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 1.08,
1140
+ "learning_rate": 3.192356249282681e-05,
1141
+ "loss": 0.6384,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 1.09,
1146
+ "learning_rate": 3.182791996633383e-05,
1147
+ "loss": 0.6194,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 1.1,
1152
+ "learning_rate": 3.173227743984085e-05,
1153
+ "loss": 0.6294,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 1.1,
1158
+ "learning_rate": 3.1636634913347876e-05,
1159
+ "loss": 0.6356,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 1.11,
1164
+ "learning_rate": 3.1540992386854895e-05,
1165
+ "loss": 0.6409,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 1.11,
1170
+ "learning_rate": 3.144534986036191e-05,
1171
+ "loss": 0.63,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 1.12,
1176
+ "learning_rate": 3.1349707333868934e-05,
1177
+ "loss": 0.6353,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 1.12,
1182
+ "learning_rate": 3.125406480737595e-05,
1183
+ "loss": 0.6359,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 1.13,
1188
+ "learning_rate": 3.115842228088297e-05,
1189
+ "loss": 0.6345,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 1.14,
1194
+ "learning_rate": 3.106277975438999e-05,
1195
+ "loss": 0.6362,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 1.14,
1200
+ "learning_rate": 3.096713722789701e-05,
1201
+ "loss": 0.636,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 1.15,
1206
+ "learning_rate": 3.087149470140404e-05,
1207
+ "loss": 0.6427,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 1.15,
1212
+ "learning_rate": 3.077585217491105e-05,
1213
+ "loss": 0.6348,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 1.16,
1218
+ "learning_rate": 3.0680209648418076e-05,
1219
+ "loss": 0.6407,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 1.16,
1224
+ "learning_rate": 3.0584567121925095e-05,
1225
+ "loss": 0.6257,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 1.17,
1230
+ "learning_rate": 3.0488924595432118e-05,
1231
+ "loss": 0.6382,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 1.18,
1236
+ "learning_rate": 3.0393282068939134e-05,
1237
+ "loss": 0.6317,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 1.18,
1242
+ "learning_rate": 3.0297639542446153e-05,
1243
+ "loss": 0.642,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 1.19,
1248
+ "learning_rate": 3.0201997015953176e-05,
1249
+ "loss": 0.6272,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 1.19,
1254
+ "learning_rate": 3.0106354489460192e-05,
1255
+ "loss": 0.6395,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 1.2,
1260
+ "learning_rate": 3.0010711962967215e-05,
1261
+ "loss": 0.6204,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 1.21,
1266
+ "learning_rate": 2.9915069436474234e-05,
1267
+ "loss": 0.6241,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 1.21,
1272
+ "learning_rate": 2.9819426909981257e-05,
1273
+ "loss": 0.6422,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 1.22,
1278
+ "learning_rate": 2.9723784383488273e-05,
1279
+ "loss": 0.632,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 1.22,
1284
+ "learning_rate": 2.9628141856995295e-05,
1285
+ "loss": 0.6326,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 1.23,
1290
+ "learning_rate": 2.9532499330502318e-05,
1291
+ "loss": 0.631,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 1.23,
1296
+ "learning_rate": 2.9436856804009337e-05,
1297
+ "loss": 0.6369,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 1.24,
1302
+ "learning_rate": 2.9341214277516353e-05,
1303
+ "loss": 0.6192,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 1.25,
1308
+ "learning_rate": 2.9245571751023376e-05,
1309
+ "loss": 0.6422,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 1.25,
1314
+ "learning_rate": 2.91499292245304e-05,
1315
+ "loss": 0.6365,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 1.26,
1320
+ "learning_rate": 2.9054286698037415e-05,
1321
+ "loss": 0.6395,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 1.26,
1326
+ "learning_rate": 2.8958644171544434e-05,
1327
+ "loss": 0.6395,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 1.27,
1332
+ "learning_rate": 2.8863001645051457e-05,
1333
+ "loss": 0.6305,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 1.27,
1338
+ "learning_rate": 2.876735911855848e-05,
1339
+ "loss": 0.6224,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 1.28,
1344
+ "learning_rate": 2.8671716592065495e-05,
1345
+ "loss": 0.6189,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 1.29,
1350
+ "learning_rate": 2.8576074065572518e-05,
1351
+ "loss": 0.6183,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 1.29,
1356
+ "learning_rate": 2.8480431539079538e-05,
1357
+ "loss": 0.6451,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 1.3,
1362
+ "learning_rate": 2.838478901258656e-05,
1363
+ "loss": 0.6311,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 1.3,
1368
+ "learning_rate": 2.8289146486093576e-05,
1369
+ "loss": 0.631,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 1.31,
1374
+ "learning_rate": 2.81935039596006e-05,
1375
+ "loss": 0.6239,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 1.31,
1380
+ "learning_rate": 2.8097861433107618e-05,
1381
+ "loss": 0.6369,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 1.32,
1386
+ "learning_rate": 2.800221890661464e-05,
1387
+ "loss": 0.6278,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 1.33,
1392
+ "learning_rate": 2.7906576380121657e-05,
1393
+ "loss": 0.6256,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 1.33,
1398
+ "learning_rate": 2.781093385362868e-05,
1399
+ "loss": 0.621,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 1.34,
1404
+ "learning_rate": 2.7715291327135702e-05,
1405
+ "loss": 0.627,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 1.34,
1410
+ "learning_rate": 2.7619648800642718e-05,
1411
+ "loss": 0.6212,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 1.35,
1416
+ "learning_rate": 2.7524006274149738e-05,
1417
+ "loss": 0.6291,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 1.35,
1422
+ "learning_rate": 2.742836374765676e-05,
1423
+ "loss": 0.6246,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 1.36,
1428
+ "learning_rate": 2.7332721221163783e-05,
1429
+ "loss": 0.6194,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 1.37,
1434
+ "learning_rate": 2.72370786946708e-05,
1435
+ "loss": 0.6166,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 1.37,
1440
+ "learning_rate": 2.7141436168177818e-05,
1441
+ "loss": 0.6181,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 1.38,
1446
+ "learning_rate": 2.704579364168484e-05,
1447
+ "loss": 0.6226,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 1.38,
1452
+ "learning_rate": 2.6950151115191864e-05,
1453
+ "loss": 0.6198,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 1.39,
1458
+ "learning_rate": 2.685450858869888e-05,
1459
+ "loss": 0.6252,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 1.39,
1464
+ "learning_rate": 2.67588660622059e-05,
1465
+ "loss": 0.624,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 1.4,
1470
+ "learning_rate": 2.666322353571292e-05,
1471
+ "loss": 0.6279,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 1.41,
1476
+ "learning_rate": 2.6567581009219938e-05,
1477
+ "loss": 0.6156,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 1.41,
1482
+ "learning_rate": 2.647193848272696e-05,
1483
+ "loss": 0.6257,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 1.42,
1488
+ "learning_rate": 2.6376295956233983e-05,
1489
+ "loss": 0.6091,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 1.42,
1494
+ "learning_rate": 2.6280653429741002e-05,
1495
+ "loss": 0.6215,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 1.43,
1500
+ "learning_rate": 2.618501090324802e-05,
1501
+ "loss": 0.619,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 1.43,
1506
+ "learning_rate": 2.608936837675504e-05,
1507
+ "loss": 0.6097,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 1.44,
1512
+ "learning_rate": 2.5993725850262064e-05,
1513
+ "loss": 0.6267,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 1.45,
1518
+ "learning_rate": 2.5898083323769083e-05,
1519
+ "loss": 0.618,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 1.45,
1524
+ "learning_rate": 2.58024407972761e-05,
1525
+ "loss": 0.6072,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 1.46,
1530
+ "learning_rate": 2.5706798270783122e-05,
1531
+ "loss": 0.6112,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 1.46,
1536
+ "learning_rate": 2.5611155744290144e-05,
1537
+ "loss": 0.6135,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 1.47,
1542
+ "learning_rate": 2.551551321779716e-05,
1543
+ "loss": 0.6109,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 1.47,
1548
+ "learning_rate": 2.5419870691304183e-05,
1549
+ "loss": 0.6177,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 1.48,
1554
+ "learning_rate": 2.5324228164811202e-05,
1555
+ "loss": 0.6009,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 1.49,
1560
+ "learning_rate": 2.5228585638318225e-05,
1561
+ "loss": 0.6052,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 1.49,
1566
+ "learning_rate": 2.513294311182524e-05,
1567
+ "loss": 0.6166,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 1.5,
1572
+ "learning_rate": 2.5037300585332264e-05,
1573
+ "loss": 0.6126,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 1.5,
1578
+ "learning_rate": 2.4941658058839283e-05,
1579
+ "loss": 0.6214,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 1.51,
1584
+ "learning_rate": 2.4846015532346302e-05,
1585
+ "loss": 0.6171,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 1.51,
1590
+ "learning_rate": 2.4750373005853325e-05,
1591
+ "loss": 0.6037,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 1.52,
1596
+ "learning_rate": 2.4654730479360344e-05,
1597
+ "loss": 0.6085,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 1.53,
1602
+ "learning_rate": 2.4559087952867364e-05,
1603
+ "loss": 0.614,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 1.53,
1608
+ "learning_rate": 2.4463445426374383e-05,
1609
+ "loss": 0.6233,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 1.54,
1614
+ "learning_rate": 2.4367802899881402e-05,
1615
+ "loss": 0.6103,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 1.54,
1620
+ "learning_rate": 2.4272160373388425e-05,
1621
+ "loss": 0.6175,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 1.55,
1626
+ "learning_rate": 2.4176517846895445e-05,
1627
+ "loss": 0.6139,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 1.56,
1632
+ "learning_rate": 2.4080875320402467e-05,
1633
+ "loss": 0.6109,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 1.56,
1638
+ "learning_rate": 2.3985232793909483e-05,
1639
+ "loss": 0.6091,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 1.57,
1644
+ "learning_rate": 2.3889590267416506e-05,
1645
+ "loss": 0.6034,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 1.57,
1650
+ "learning_rate": 2.3793947740923525e-05,
1651
+ "loss": 0.6161,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 1.58,
1656
+ "learning_rate": 2.3698305214430548e-05,
1657
+ "loss": 0.6114,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 1.58,
1662
+ "learning_rate": 2.3602662687937567e-05,
1663
+ "loss": 0.6006,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 1.59,
1668
+ "learning_rate": 2.3507020161444583e-05,
1669
+ "loss": 0.5987,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 1.6,
1674
+ "learning_rate": 2.3411377634951606e-05,
1675
+ "loss": 0.6084,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 1.6,
1680
+ "learning_rate": 2.3315735108458625e-05,
1681
+ "loss": 0.6022,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 1.61,
1686
+ "learning_rate": 2.3220092581965648e-05,
1687
+ "loss": 0.6033,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 1.61,
1692
+ "learning_rate": 2.3124450055472667e-05,
1693
+ "loss": 0.608,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 1.62,
1698
+ "learning_rate": 2.3028807528979687e-05,
1699
+ "loss": 0.6088,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 1.62,
1704
+ "learning_rate": 2.2933165002486706e-05,
1705
+ "loss": 0.613,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 1.63,
1710
+ "learning_rate": 2.283752247599373e-05,
1711
+ "loss": 0.6048,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 1.64,
1716
+ "learning_rate": 2.2741879949500748e-05,
1717
+ "loss": 0.6156,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 1.64,
1722
+ "learning_rate": 2.2646237423007767e-05,
1723
+ "loss": 0.5922,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 1.65,
1728
+ "learning_rate": 2.2550594896514787e-05,
1729
+ "loss": 0.5913,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 1.65,
1734
+ "learning_rate": 2.245495237002181e-05,
1735
+ "loss": 0.6193,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 1.66,
1740
+ "learning_rate": 2.235930984352883e-05,
1741
+ "loss": 0.5952,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 1.66,
1746
+ "learning_rate": 2.2263667317035848e-05,
1747
+ "loss": 0.5864,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 1.67,
1752
+ "learning_rate": 2.2168024790542867e-05,
1753
+ "loss": 0.5968,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 1.68,
1758
+ "learning_rate": 2.2072382264049887e-05,
1759
+ "loss": 0.593,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 1.68,
1764
+ "learning_rate": 2.197673973755691e-05,
1765
+ "loss": 0.5931,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 1.69,
1770
+ "learning_rate": 2.188109721106393e-05,
1771
+ "loss": 0.6066,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 1.69,
1776
+ "learning_rate": 2.1785454684570948e-05,
1777
+ "loss": 0.598,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 1.7,
1782
+ "learning_rate": 2.1689812158077967e-05,
1783
+ "loss": 0.5971,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 1.7,
1788
+ "learning_rate": 2.159416963158499e-05,
1789
+ "loss": 0.6013,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 1.71,
1794
+ "learning_rate": 2.149852710509201e-05,
1795
+ "loss": 0.5954,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 1.72,
1800
+ "learning_rate": 2.1402884578599032e-05,
1801
+ "loss": 0.5965,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 1.72,
1806
+ "learning_rate": 2.1307242052106048e-05,
1807
+ "loss": 0.5851,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 1.73,
1812
+ "learning_rate": 2.1211599525613067e-05,
1813
+ "loss": 0.5922,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 1.73,
1818
+ "learning_rate": 2.111595699912009e-05,
1819
+ "loss": 0.5861,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 1.74,
1824
+ "learning_rate": 2.102031447262711e-05,
1825
+ "loss": 0.5923,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 1.74,
1830
+ "learning_rate": 2.0924671946134132e-05,
1831
+ "loss": 0.591,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 1.75,
1836
+ "learning_rate": 2.0829029419641148e-05,
1837
+ "loss": 0.5974,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 1.76,
1842
+ "learning_rate": 2.073338689314817e-05,
1843
+ "loss": 0.6003,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 1.76,
1848
+ "learning_rate": 2.063774436665519e-05,
1849
+ "loss": 0.6029,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 1.77,
1854
+ "learning_rate": 2.0542101840162213e-05,
1855
+ "loss": 0.5909,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 1.77,
1860
+ "learning_rate": 2.0446459313669232e-05,
1861
+ "loss": 0.5972,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 1.78,
1866
+ "learning_rate": 2.035081678717625e-05,
1867
+ "loss": 0.6074,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 1.78,
1872
+ "learning_rate": 2.025517426068327e-05,
1873
+ "loss": 0.5821,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 1.79,
1878
+ "learning_rate": 2.015953173419029e-05,
1879
+ "loss": 0.5863,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 1.8,
1884
+ "learning_rate": 2.0063889207697313e-05,
1885
+ "loss": 0.5801,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 1.8,
1890
+ "learning_rate": 1.9968246681204332e-05,
1891
+ "loss": 0.5914,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 1.81,
1896
+ "learning_rate": 1.987260415471135e-05,
1897
+ "loss": 0.6046,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 1.81,
1902
+ "learning_rate": 1.977696162821837e-05,
1903
+ "loss": 0.5955,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 1.82,
1908
+ "learning_rate": 1.9681319101725394e-05,
1909
+ "loss": 0.591,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 1.82,
1914
+ "learning_rate": 1.9585676575232413e-05,
1915
+ "loss": 0.588,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 1.83,
1920
+ "learning_rate": 1.9490034048739432e-05,
1921
+ "loss": 0.5932,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 1.84,
1926
+ "learning_rate": 1.939439152224645e-05,
1927
+ "loss": 0.5826,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 1.84,
1932
+ "learning_rate": 1.9298748995753474e-05,
1933
+ "loss": 0.5823,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 1.85,
1938
+ "learning_rate": 1.9203106469260494e-05,
1939
+ "loss": 0.5978,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 1.85,
1944
+ "learning_rate": 1.9107463942767513e-05,
1945
+ "loss": 0.5848,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 1.86,
1950
+ "learning_rate": 1.9011821416274532e-05,
1951
+ "loss": 0.5881,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 1.87,
1956
+ "learning_rate": 1.891617888978155e-05,
1957
+ "loss": 0.5879,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 1.87,
1962
+ "learning_rate": 1.8820536363288574e-05,
1963
+ "loss": 0.5926,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 1.88,
1968
+ "learning_rate": 1.8724893836795594e-05,
1969
+ "loss": 0.5869,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 1.88,
1974
+ "learning_rate": 1.8629251310302613e-05,
1975
+ "loss": 0.5831,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 1.89,
1980
+ "learning_rate": 1.8533608783809632e-05,
1981
+ "loss": 0.579,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 1.89,
1986
+ "learning_rate": 1.8437966257316655e-05,
1987
+ "loss": 0.5863,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 1.9,
1992
+ "learning_rate": 1.8342323730823674e-05,
1993
+ "loss": 0.5817,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 1.91,
1998
+ "learning_rate": 1.8246681204330697e-05,
1999
+ "loss": 0.5901,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 1.91,
2004
+ "learning_rate": 1.8151038677837713e-05,
2005
+ "loss": 0.5863,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 1.92,
2010
+ "learning_rate": 1.8055396151344736e-05,
2011
+ "loss": 0.5831,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 1.92,
2016
+ "learning_rate": 1.7959753624851755e-05,
2017
+ "loss": 0.5853,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 1.93,
2022
+ "learning_rate": 1.7864111098358774e-05,
2023
+ "loss": 0.5858,
2024
+ "step": 168000
2025
+ },
2026
+ {
2027
+ "epoch": 1.93,
2028
+ "learning_rate": 1.7768468571865797e-05,
2029
+ "loss": 0.5876,
2030
+ "step": 168500
2031
+ },
2032
+ {
2033
+ "epoch": 1.94,
2034
+ "learning_rate": 1.7672826045372813e-05,
2035
+ "loss": 0.5796,
2036
+ "step": 169000
2037
+ },
2038
+ {
2039
+ "epoch": 1.95,
2040
+ "learning_rate": 1.7577183518879836e-05,
2041
+ "loss": 0.5896,
2042
+ "step": 169500
2043
+ },
2044
+ {
2045
+ "epoch": 1.95,
2046
+ "learning_rate": 1.7481540992386855e-05,
2047
+ "loss": 0.5882,
2048
+ "step": 170000
2049
+ },
2050
+ {
2051
+ "epoch": 1.96,
2052
+ "learning_rate": 1.7385898465893878e-05,
2053
+ "loss": 0.5791,
2054
+ "step": 170500
2055
+ },
2056
+ {
2057
+ "epoch": 1.96,
2058
+ "learning_rate": 1.7290255939400897e-05,
2059
+ "loss": 0.5875,
2060
+ "step": 171000
2061
+ },
2062
+ {
2063
+ "epoch": 1.97,
2064
+ "learning_rate": 1.7194613412907916e-05,
2065
+ "loss": 0.5855,
2066
+ "step": 171500
2067
+ },
2068
+ {
2069
+ "epoch": 1.97,
2070
+ "learning_rate": 1.7098970886414936e-05,
2071
+ "loss": 0.5885,
2072
+ "step": 172000
2073
+ },
2074
+ {
2075
+ "epoch": 1.98,
2076
+ "learning_rate": 1.700332835992196e-05,
2077
+ "loss": 0.571,
2078
+ "step": 172500
2079
+ },
2080
+ {
2081
+ "epoch": 1.99,
2082
+ "learning_rate": 1.6907685833428978e-05,
2083
+ "loss": 0.5803,
2084
+ "step": 173000
2085
+ },
2086
+ {
2087
+ "epoch": 1.99,
2088
+ "learning_rate": 1.6812043306935997e-05,
2089
+ "loss": 0.5805,
2090
+ "step": 173500
2091
+ },
2092
+ {
2093
+ "epoch": 2.0,
2094
+ "learning_rate": 1.6716400780443016e-05,
2095
+ "loss": 0.5853,
2096
+ "step": 174000
2097
+ },
2098
+ {
2099
+ "epoch": 2.0,
2100
+ "learning_rate": 1.6620758253950036e-05,
2101
+ "loss": 0.5407,
2102
+ "step": 174500
2103
+ },
2104
+ {
2105
+ "epoch": 2.01,
2106
+ "learning_rate": 1.652511572745706e-05,
2107
+ "loss": 0.5119,
2108
+ "step": 175000
2109
+ },
2110
+ {
2111
+ "epoch": 2.01,
2112
+ "learning_rate": 1.6429473200964078e-05,
2113
+ "loss": 0.5168,
2114
+ "step": 175500
2115
+ },
2116
+ {
2117
+ "epoch": 2.02,
2118
+ "learning_rate": 1.6333830674471097e-05,
2119
+ "loss": 0.5199,
2120
+ "step": 176000
2121
+ },
2122
+ {
2123
+ "epoch": 2.03,
2124
+ "learning_rate": 1.6238188147978116e-05,
2125
+ "loss": 0.5095,
2126
+ "step": 176500
2127
+ },
2128
+ {
2129
+ "epoch": 2.03,
2130
+ "learning_rate": 1.614254562148514e-05,
2131
+ "loss": 0.5177,
2132
+ "step": 177000
2133
+ },
2134
+ {
2135
+ "epoch": 2.04,
2136
+ "learning_rate": 1.604690309499216e-05,
2137
+ "loss": 0.5131,
2138
+ "step": 177500
2139
+ },
2140
+ {
2141
+ "epoch": 2.04,
2142
+ "learning_rate": 1.5951260568499178e-05,
2143
+ "loss": 0.5133,
2144
+ "step": 178000
2145
+ },
2146
+ {
2147
+ "epoch": 2.05,
2148
+ "learning_rate": 1.5855618042006197e-05,
2149
+ "loss": 0.516,
2150
+ "step": 178500
2151
+ },
2152
+ {
2153
+ "epoch": 2.05,
2154
+ "learning_rate": 1.5759975515513217e-05,
2155
+ "loss": 0.5193,
2156
+ "step": 179000
2157
+ },
2158
+ {
2159
+ "epoch": 2.06,
2160
+ "learning_rate": 1.566433298902024e-05,
2161
+ "loss": 0.5109,
2162
+ "step": 179500
2163
+ },
2164
+ {
2165
+ "epoch": 2.07,
2166
+ "learning_rate": 1.556869046252726e-05,
2167
+ "loss": 0.5071,
2168
+ "step": 180000
2169
+ },
2170
+ {
2171
+ "epoch": 2.07,
2172
+ "learning_rate": 1.5473047936034278e-05,
2173
+ "loss": 0.5152,
2174
+ "step": 180500
2175
+ },
2176
+ {
2177
+ "epoch": 2.08,
2178
+ "learning_rate": 1.5377405409541297e-05,
2179
+ "loss": 0.5164,
2180
+ "step": 181000
2181
+ },
2182
+ {
2183
+ "epoch": 2.08,
2184
+ "learning_rate": 1.528176288304832e-05,
2185
+ "loss": 0.5195,
2186
+ "step": 181500
2187
+ },
2188
+ {
2189
+ "epoch": 2.09,
2190
+ "learning_rate": 1.518612035655534e-05,
2191
+ "loss": 0.5251,
2192
+ "step": 182000
2193
+ },
2194
+ {
2195
+ "epoch": 2.09,
2196
+ "learning_rate": 1.509047783006236e-05,
2197
+ "loss": 0.5204,
2198
+ "step": 182500
2199
+ },
2200
+ {
2201
+ "epoch": 2.1,
2202
+ "learning_rate": 1.499483530356938e-05,
2203
+ "loss": 0.5234,
2204
+ "step": 183000
2205
+ },
2206
+ {
2207
+ "epoch": 2.11,
2208
+ "learning_rate": 1.48991927770764e-05,
2209
+ "loss": 0.5029,
2210
+ "step": 183500
2211
+ },
2212
+ {
2213
+ "epoch": 2.11,
2214
+ "learning_rate": 1.480355025058342e-05,
2215
+ "loss": 0.5111,
2216
+ "step": 184000
2217
+ },
2218
+ {
2219
+ "epoch": 2.12,
2220
+ "learning_rate": 1.470790772409044e-05,
2221
+ "loss": 0.5099,
2222
+ "step": 184500
2223
+ },
2224
+ {
2225
+ "epoch": 2.12,
2226
+ "learning_rate": 1.461226519759746e-05,
2227
+ "loss": 0.519,
2228
+ "step": 185000
2229
+ },
2230
+ {
2231
+ "epoch": 2.13,
2232
+ "learning_rate": 1.451662267110448e-05,
2233
+ "loss": 0.5072,
2234
+ "step": 185500
2235
+ },
2236
+ {
2237
+ "epoch": 2.13,
2238
+ "learning_rate": 1.44209801446115e-05,
2239
+ "loss": 0.5064,
2240
+ "step": 186000
2241
+ },
2242
+ {
2243
+ "epoch": 2.14,
2244
+ "learning_rate": 1.432533761811852e-05,
2245
+ "loss": 0.511,
2246
+ "step": 186500
2247
+ },
2248
+ {
2249
+ "epoch": 2.15,
2250
+ "learning_rate": 1.4229695091625541e-05,
2251
+ "loss": 0.514,
2252
+ "step": 187000
2253
+ },
2254
+ {
2255
+ "epoch": 2.15,
2256
+ "learning_rate": 1.413405256513256e-05,
2257
+ "loss": 0.516,
2258
+ "step": 187500
2259
+ },
2260
+ {
2261
+ "epoch": 2.16,
2262
+ "learning_rate": 1.4038410038639583e-05,
2263
+ "loss": 0.5226,
2264
+ "step": 188000
2265
+ },
2266
+ {
2267
+ "epoch": 2.16,
2268
+ "learning_rate": 1.39427675121466e-05,
2269
+ "loss": 0.505,
2270
+ "step": 188500
2271
+ },
2272
+ {
2273
+ "epoch": 2.17,
2274
+ "learning_rate": 1.3847124985653623e-05,
2275
+ "loss": 0.5158,
2276
+ "step": 189000
2277
+ },
2278
+ {
2279
+ "epoch": 2.17,
2280
+ "learning_rate": 1.3751482459160641e-05,
2281
+ "loss": 0.5163,
2282
+ "step": 189500
2283
+ },
2284
+ {
2285
+ "epoch": 2.18,
2286
+ "learning_rate": 1.3655839932667664e-05,
2287
+ "loss": 0.5156,
2288
+ "step": 190000
2289
+ },
2290
+ {
2291
+ "epoch": 2.19,
2292
+ "learning_rate": 1.3560197406174683e-05,
2293
+ "loss": 0.5174,
2294
+ "step": 190500
2295
+ },
2296
+ {
2297
+ "epoch": 2.19,
2298
+ "learning_rate": 1.34645548796817e-05,
2299
+ "loss": 0.5163,
2300
+ "step": 191000
2301
+ },
2302
+ {
2303
+ "epoch": 2.2,
2304
+ "learning_rate": 1.3368912353188723e-05,
2305
+ "loss": 0.5079,
2306
+ "step": 191500
2307
+ },
2308
+ {
2309
+ "epoch": 2.2,
2310
+ "learning_rate": 1.3273269826695741e-05,
2311
+ "loss": 0.5082,
2312
+ "step": 192000
2313
+ },
2314
+ {
2315
+ "epoch": 2.21,
2316
+ "learning_rate": 1.3177627300202764e-05,
2317
+ "loss": 0.5132,
2318
+ "step": 192500
2319
+ },
2320
+ {
2321
+ "epoch": 2.22,
2322
+ "learning_rate": 1.3081984773709781e-05,
2323
+ "loss": 0.5174,
2324
+ "step": 193000
2325
+ },
2326
+ {
2327
+ "epoch": 2.22,
2328
+ "learning_rate": 1.2986342247216804e-05,
2329
+ "loss": 0.5105,
2330
+ "step": 193500
2331
+ },
2332
+ {
2333
+ "epoch": 2.23,
2334
+ "learning_rate": 1.2890699720723823e-05,
2335
+ "loss": 0.4967,
2336
+ "step": 194000
2337
+ },
2338
+ {
2339
+ "epoch": 2.23,
2340
+ "learning_rate": 1.2795057194230844e-05,
2341
+ "loss": 0.5078,
2342
+ "step": 194500
2343
+ },
2344
+ {
2345
+ "epoch": 2.24,
2346
+ "learning_rate": 1.2699414667737864e-05,
2347
+ "loss": 0.5116,
2348
+ "step": 195000
2349
+ },
2350
+ {
2351
+ "epoch": 2.24,
2352
+ "learning_rate": 1.2603772141244885e-05,
2353
+ "loss": 0.5141,
2354
+ "step": 195500
2355
+ },
2356
+ {
2357
+ "epoch": 2.25,
2358
+ "learning_rate": 1.2508129614751904e-05,
2359
+ "loss": 0.5174,
2360
+ "step": 196000
2361
+ },
2362
+ {
2363
+ "epoch": 2.26,
2364
+ "learning_rate": 1.2412487088258923e-05,
2365
+ "loss": 0.5151,
2366
+ "step": 196500
2367
+ },
2368
+ {
2369
+ "epoch": 2.26,
2370
+ "learning_rate": 1.2316844561765945e-05,
2371
+ "loss": 0.5078,
2372
+ "step": 197000
2373
+ },
2374
+ {
2375
+ "epoch": 2.27,
2376
+ "learning_rate": 1.2221202035272966e-05,
2377
+ "loss": 0.5187,
2378
+ "step": 197500
2379
+ },
2380
+ {
2381
+ "epoch": 2.27,
2382
+ "learning_rate": 1.2125559508779985e-05,
2383
+ "loss": 0.5154,
2384
+ "step": 198000
2385
+ },
2386
+ {
2387
+ "epoch": 2.28,
2388
+ "learning_rate": 1.2029916982287006e-05,
2389
+ "loss": 0.5165,
2390
+ "step": 198500
2391
+ },
2392
+ {
2393
+ "epoch": 2.28,
2394
+ "learning_rate": 1.1934274455794024e-05,
2395
+ "loss": 0.5042,
2396
+ "step": 199000
2397
+ },
2398
+ {
2399
+ "epoch": 2.29,
2400
+ "learning_rate": 1.1838631929301045e-05,
2401
+ "loss": 0.5042,
2402
+ "step": 199500
2403
+ },
2404
+ {
2405
+ "epoch": 2.3,
2406
+ "learning_rate": 1.1742989402808066e-05,
2407
+ "loss": 0.5185,
2408
+ "step": 200000
2409
+ },
2410
+ {
2411
+ "epoch": 2.3,
2412
+ "learning_rate": 1.1647346876315085e-05,
2413
+ "loss": 0.5047,
2414
+ "step": 200500
2415
+ },
2416
+ {
2417
+ "epoch": 2.31,
2418
+ "learning_rate": 1.1551704349822106e-05,
2419
+ "loss": 0.5055,
2420
+ "step": 201000
2421
+ },
2422
+ {
2423
+ "epoch": 2.31,
2424
+ "learning_rate": 1.1456061823329125e-05,
2425
+ "loss": 0.5035,
2426
+ "step": 201500
2427
+ },
2428
+ {
2429
+ "epoch": 2.32,
2430
+ "learning_rate": 1.1360419296836146e-05,
2431
+ "loss": 0.5121,
2432
+ "step": 202000
2433
+ },
2434
+ {
2435
+ "epoch": 2.32,
2436
+ "learning_rate": 1.1264776770343166e-05,
2437
+ "loss": 0.5112,
2438
+ "step": 202500
2439
+ },
2440
+ {
2441
+ "epoch": 2.33,
2442
+ "learning_rate": 1.1169134243850187e-05,
2443
+ "loss": 0.5035,
2444
+ "step": 203000
2445
+ },
2446
+ {
2447
+ "epoch": 2.34,
2448
+ "learning_rate": 1.1073491717357206e-05,
2449
+ "loss": 0.5009,
2450
+ "step": 203500
2451
+ },
2452
+ {
2453
+ "epoch": 2.34,
2454
+ "learning_rate": 1.0977849190864227e-05,
2455
+ "loss": 0.5067,
2456
+ "step": 204000
2457
+ },
2458
+ {
2459
+ "epoch": 2.35,
2460
+ "learning_rate": 1.0882206664371246e-05,
2461
+ "loss": 0.5055,
2462
+ "step": 204500
2463
+ },
2464
+ {
2465
+ "epoch": 2.35,
2466
+ "learning_rate": 1.0786564137878266e-05,
2467
+ "loss": 0.5025,
2468
+ "step": 205000
2469
+ },
2470
+ {
2471
+ "epoch": 2.36,
2472
+ "learning_rate": 1.0690921611385287e-05,
2473
+ "loss": 0.5044,
2474
+ "step": 205500
2475
+ },
2476
+ {
2477
+ "epoch": 2.36,
2478
+ "learning_rate": 1.0595279084892306e-05,
2479
+ "loss": 0.5091,
2480
+ "step": 206000
2481
+ },
2482
+ {
2483
+ "epoch": 2.37,
2484
+ "learning_rate": 1.0499636558399327e-05,
2485
+ "loss": 0.5074,
2486
+ "step": 206500
2487
+ },
2488
+ {
2489
+ "epoch": 2.38,
2490
+ "learning_rate": 1.0403994031906348e-05,
2491
+ "loss": 0.5142,
2492
+ "step": 207000
2493
+ },
2494
+ {
2495
+ "epoch": 2.38,
2496
+ "learning_rate": 1.0308351505413367e-05,
2497
+ "loss": 0.4958,
2498
+ "step": 207500
2499
+ },
2500
+ {
2501
+ "epoch": 2.39,
2502
+ "learning_rate": 1.0212708978920388e-05,
2503
+ "loss": 0.5208,
2504
+ "step": 208000
2505
+ },
2506
+ {
2507
+ "epoch": 2.39,
2508
+ "learning_rate": 1.0117066452427408e-05,
2509
+ "loss": 0.5013,
2510
+ "step": 208500
2511
+ },
2512
+ {
2513
+ "epoch": 2.4,
2514
+ "learning_rate": 1.0021423925934429e-05,
2515
+ "loss": 0.5009,
2516
+ "step": 209000
2517
+ },
2518
+ {
2519
+ "epoch": 2.4,
2520
+ "learning_rate": 9.925781399441448e-06,
2521
+ "loss": 0.5109,
2522
+ "step": 209500
2523
+ },
2524
+ {
2525
+ "epoch": 2.41,
2526
+ "learning_rate": 9.830138872948469e-06,
2527
+ "loss": 0.4979,
2528
+ "step": 210000
2529
+ },
2530
+ {
2531
+ "epoch": 2.42,
2532
+ "learning_rate": 9.734496346455488e-06,
2533
+ "loss": 0.5099,
2534
+ "step": 210500
2535
+ },
2536
+ {
2537
+ "epoch": 2.42,
2538
+ "learning_rate": 9.638853819962508e-06,
2539
+ "loss": 0.5016,
2540
+ "step": 211000
2541
+ },
2542
+ {
2543
+ "epoch": 2.43,
2544
+ "learning_rate": 9.543211293469529e-06,
2545
+ "loss": 0.5104,
2546
+ "step": 211500
2547
+ },
2548
+ {
2549
+ "epoch": 2.43,
2550
+ "learning_rate": 9.447568766976548e-06,
2551
+ "loss": 0.4964,
2552
+ "step": 212000
2553
+ },
2554
+ {
2555
+ "epoch": 2.44,
2556
+ "learning_rate": 9.351926240483569e-06,
2557
+ "loss": 0.5036,
2558
+ "step": 212500
2559
+ },
2560
+ {
2561
+ "epoch": 2.44,
2562
+ "learning_rate": 9.256283713990588e-06,
2563
+ "loss": 0.5175,
2564
+ "step": 213000
2565
+ },
2566
+ {
2567
+ "epoch": 2.45,
2568
+ "learning_rate": 9.16064118749761e-06,
2569
+ "loss": 0.5002,
2570
+ "step": 213500
2571
+ },
2572
+ {
2573
+ "epoch": 2.46,
2574
+ "learning_rate": 9.06499866100463e-06,
2575
+ "loss": 0.5012,
2576
+ "step": 214000
2577
+ },
2578
+ {
2579
+ "epoch": 2.46,
2580
+ "learning_rate": 8.96935613451165e-06,
2581
+ "loss": 0.5001,
2582
+ "step": 214500
2583
+ },
2584
+ {
2585
+ "epoch": 2.47,
2586
+ "learning_rate": 8.87371360801867e-06,
2587
+ "loss": 0.5012,
2588
+ "step": 215000
2589
+ },
2590
+ {
2591
+ "epoch": 2.47,
2592
+ "learning_rate": 8.77807108152569e-06,
2593
+ "loss": 0.5074,
2594
+ "step": 215500
2595
+ },
2596
+ {
2597
+ "epoch": 2.48,
2598
+ "learning_rate": 8.682428555032711e-06,
2599
+ "loss": 0.506,
2600
+ "step": 216000
2601
+ },
2602
+ {
2603
+ "epoch": 2.48,
2604
+ "learning_rate": 8.58678602853973e-06,
2605
+ "loss": 0.5022,
2606
+ "step": 216500
2607
+ },
2608
+ {
2609
+ "epoch": 2.49,
2610
+ "learning_rate": 8.49114350204675e-06,
2611
+ "loss": 0.5072,
2612
+ "step": 217000
2613
+ },
2614
+ {
2615
+ "epoch": 2.5,
2616
+ "learning_rate": 8.39550097555377e-06,
2617
+ "loss": 0.5027,
2618
+ "step": 217500
2619
+ },
2620
+ {
2621
+ "epoch": 2.5,
2622
+ "learning_rate": 8.29985844906079e-06,
2623
+ "loss": 0.5044,
2624
+ "step": 218000
2625
+ },
2626
+ {
2627
+ "epoch": 2.51,
2628
+ "learning_rate": 8.204215922567811e-06,
2629
+ "loss": 0.496,
2630
+ "step": 218500
2631
+ },
2632
+ {
2633
+ "epoch": 2.51,
2634
+ "learning_rate": 8.10857339607483e-06,
2635
+ "loss": 0.4946,
2636
+ "step": 219000
2637
+ },
2638
+ {
2639
+ "epoch": 2.52,
2640
+ "learning_rate": 8.012930869581852e-06,
2641
+ "loss": 0.5032,
2642
+ "step": 219500
2643
+ },
2644
+ {
2645
+ "epoch": 2.52,
2646
+ "learning_rate": 7.917288343088871e-06,
2647
+ "loss": 0.4993,
2648
+ "step": 220000
2649
+ },
2650
+ {
2651
+ "epoch": 2.53,
2652
+ "learning_rate": 7.821645816595892e-06,
2653
+ "loss": 0.5091,
2654
+ "step": 220500
2655
+ },
2656
+ {
2657
+ "epoch": 2.54,
2658
+ "learning_rate": 7.726003290102913e-06,
2659
+ "loss": 0.4929,
2660
+ "step": 221000
2661
+ },
2662
+ {
2663
+ "epoch": 2.54,
2664
+ "learning_rate": 7.630360763609932e-06,
2665
+ "loss": 0.4995,
2666
+ "step": 221500
2667
+ },
2668
+ {
2669
+ "epoch": 2.55,
2670
+ "learning_rate": 7.5347182371169516e-06,
2671
+ "loss": 0.504,
2672
+ "step": 222000
2673
+ },
2674
+ {
2675
+ "epoch": 2.55,
2676
+ "learning_rate": 7.439075710623972e-06,
2677
+ "loss": 0.5026,
2678
+ "step": 222500
2679
+ },
2680
+ {
2681
+ "epoch": 2.56,
2682
+ "learning_rate": 7.343433184130992e-06,
2683
+ "loss": 0.5056,
2684
+ "step": 223000
2685
+ },
2686
+ {
2687
+ "epoch": 2.57,
2688
+ "learning_rate": 7.247790657638012e-06,
2689
+ "loss": 0.4925,
2690
+ "step": 223500
2691
+ },
2692
+ {
2693
+ "epoch": 2.57,
2694
+ "learning_rate": 7.152148131145032e-06,
2695
+ "loss": 0.4944,
2696
+ "step": 224000
2697
+ },
2698
+ {
2699
+ "epoch": 2.58,
2700
+ "learning_rate": 7.0565056046520524e-06,
2701
+ "loss": 0.5029,
2702
+ "step": 224500
2703
+ },
2704
+ {
2705
+ "epoch": 2.58,
2706
+ "learning_rate": 6.9608630781590735e-06,
2707
+ "loss": 0.5002,
2708
+ "step": 225000
2709
+ },
2710
+ {
2711
+ "epoch": 2.59,
2712
+ "learning_rate": 6.865220551666094e-06,
2713
+ "loss": 0.5027,
2714
+ "step": 225500
2715
+ },
2716
+ {
2717
+ "epoch": 2.59,
2718
+ "learning_rate": 6.769578025173114e-06,
2719
+ "loss": 0.5052,
2720
+ "step": 226000
2721
+ },
2722
+ {
2723
+ "epoch": 2.6,
2724
+ "learning_rate": 6.673935498680134e-06,
2725
+ "loss": 0.5096,
2726
+ "step": 226500
2727
+ },
2728
+ {
2729
+ "epoch": 2.61,
2730
+ "learning_rate": 6.578292972187154e-06,
2731
+ "loss": 0.5041,
2732
+ "step": 227000
2733
+ },
2734
+ {
2735
+ "epoch": 2.61,
2736
+ "learning_rate": 6.482650445694174e-06,
2737
+ "loss": 0.4935,
2738
+ "step": 227500
2739
+ },
2740
+ {
2741
+ "epoch": 2.62,
2742
+ "learning_rate": 6.387007919201194e-06,
2743
+ "loss": 0.5016,
2744
+ "step": 228000
2745
+ },
2746
+ {
2747
+ "epoch": 2.62,
2748
+ "learning_rate": 6.291365392708214e-06,
2749
+ "loss": 0.4932,
2750
+ "step": 228500
2751
+ },
2752
+ {
2753
+ "epoch": 2.63,
2754
+ "learning_rate": 6.195722866215234e-06,
2755
+ "loss": 0.4973,
2756
+ "step": 229000
2757
+ },
2758
+ {
2759
+ "epoch": 2.63,
2760
+ "learning_rate": 6.100080339722254e-06,
2761
+ "loss": 0.4995,
2762
+ "step": 229500
2763
+ },
2764
+ {
2765
+ "epoch": 2.64,
2766
+ "learning_rate": 6.004437813229274e-06,
2767
+ "loss": 0.4871,
2768
+ "step": 230000
2769
+ },
2770
+ {
2771
+ "epoch": 2.65,
2772
+ "learning_rate": 5.9087952867362945e-06,
2773
+ "loss": 0.4925,
2774
+ "step": 230500
2775
+ },
2776
+ {
2777
+ "epoch": 2.65,
2778
+ "learning_rate": 5.813152760243315e-06,
2779
+ "loss": 0.4979,
2780
+ "step": 231000
2781
+ },
2782
+ {
2783
+ "epoch": 2.66,
2784
+ "learning_rate": 5.717510233750335e-06,
2785
+ "loss": 0.4907,
2786
+ "step": 231500
2787
+ },
2788
+ {
2789
+ "epoch": 2.66,
2790
+ "learning_rate": 5.621867707257355e-06,
2791
+ "loss": 0.5001,
2792
+ "step": 232000
2793
+ },
2794
+ {
2795
+ "epoch": 2.67,
2796
+ "learning_rate": 5.526225180764375e-06,
2797
+ "loss": 0.4919,
2798
+ "step": 232500
2799
+ },
2800
+ {
2801
+ "epoch": 2.67,
2802
+ "learning_rate": 5.430582654271395e-06,
2803
+ "loss": 0.5017,
2804
+ "step": 233000
2805
+ },
2806
+ {
2807
+ "epoch": 2.68,
2808
+ "learning_rate": 5.334940127778416e-06,
2809
+ "loss": 0.4891,
2810
+ "step": 233500
2811
+ },
2812
+ {
2813
+ "epoch": 2.69,
2814
+ "learning_rate": 5.239297601285436e-06,
2815
+ "loss": 0.4921,
2816
+ "step": 234000
2817
+ },
2818
+ {
2819
+ "epoch": 2.69,
2820
+ "learning_rate": 5.143655074792456e-06,
2821
+ "loss": 0.492,
2822
+ "step": 234500
2823
+ },
2824
+ {
2825
+ "epoch": 2.7,
2826
+ "learning_rate": 5.048012548299476e-06,
2827
+ "loss": 0.509,
2828
+ "step": 235000
2829
+ },
2830
+ {
2831
+ "epoch": 2.7,
2832
+ "learning_rate": 4.952370021806496e-06,
2833
+ "loss": 0.5025,
2834
+ "step": 235500
2835
+ },
2836
+ {
2837
+ "epoch": 2.71,
2838
+ "learning_rate": 4.8567274953135165e-06,
2839
+ "loss": 0.4915,
2840
+ "step": 236000
2841
+ },
2842
+ {
2843
+ "epoch": 2.71,
2844
+ "learning_rate": 4.761084968820537e-06,
2845
+ "loss": 0.4895,
2846
+ "step": 236500
2847
+ },
2848
+ {
2849
+ "epoch": 2.72,
2850
+ "learning_rate": 4.665442442327557e-06,
2851
+ "loss": 0.4965,
2852
+ "step": 237000
2853
+ },
2854
+ {
2855
+ "epoch": 2.73,
2856
+ "learning_rate": 4.569799915834577e-06,
2857
+ "loss": 0.4998,
2858
+ "step": 237500
2859
+ },
2860
+ {
2861
+ "epoch": 2.73,
2862
+ "learning_rate": 4.474157389341597e-06,
2863
+ "loss": 0.5009,
2864
+ "step": 238000
2865
+ },
2866
+ {
2867
+ "epoch": 2.74,
2868
+ "learning_rate": 4.378514862848617e-06,
2869
+ "loss": 0.4882,
2870
+ "step": 238500
2871
+ },
2872
+ {
2873
+ "epoch": 2.74,
2874
+ "learning_rate": 4.2828723363556375e-06,
2875
+ "loss": 0.4896,
2876
+ "step": 239000
2877
+ },
2878
+ {
2879
+ "epoch": 2.75,
2880
+ "learning_rate": 4.187229809862658e-06,
2881
+ "loss": 0.5006,
2882
+ "step": 239500
2883
+ },
2884
+ {
2885
+ "epoch": 2.75,
2886
+ "learning_rate": 4.091587283369678e-06,
2887
+ "loss": 0.4912,
2888
+ "step": 240000
2889
+ },
2890
+ {
2891
+ "epoch": 2.76,
2892
+ "learning_rate": 3.995944756876697e-06,
2893
+ "loss": 0.4915,
2894
+ "step": 240500
2895
+ },
2896
+ {
2897
+ "epoch": 2.77,
2898
+ "learning_rate": 3.900302230383718e-06,
2899
+ "loss": 0.4993,
2900
+ "step": 241000
2901
+ },
2902
+ {
2903
+ "epoch": 2.77,
2904
+ "learning_rate": 3.804659703890738e-06,
2905
+ "loss": 0.4902,
2906
+ "step": 241500
2907
+ },
2908
+ {
2909
+ "epoch": 2.78,
2910
+ "learning_rate": 3.709017177397758e-06,
2911
+ "loss": 0.5008,
2912
+ "step": 242000
2913
+ },
2914
+ {
2915
+ "epoch": 2.78,
2916
+ "learning_rate": 3.6133746509047787e-06,
2917
+ "loss": 0.4951,
2918
+ "step": 242500
2919
+ },
2920
+ {
2921
+ "epoch": 2.79,
2922
+ "learning_rate": 3.517732124411799e-06,
2923
+ "loss": 0.5028,
2924
+ "step": 243000
2925
+ },
2926
+ {
2927
+ "epoch": 2.79,
2928
+ "learning_rate": 3.4220895979188187e-06,
2929
+ "loss": 0.4866,
2930
+ "step": 243500
2931
+ },
2932
+ {
2933
+ "epoch": 2.8,
2934
+ "learning_rate": 3.326447071425839e-06,
2935
+ "loss": 0.4928,
2936
+ "step": 244000
2937
+ },
2938
+ {
2939
+ "epoch": 2.81,
2940
+ "learning_rate": 3.230804544932859e-06,
2941
+ "loss": 0.4833,
2942
+ "step": 244500
2943
+ },
2944
+ {
2945
+ "epoch": 2.81,
2946
+ "learning_rate": 3.135162018439879e-06,
2947
+ "loss": 0.4892,
2948
+ "step": 245000
2949
+ },
2950
+ {
2951
+ "epoch": 2.82,
2952
+ "learning_rate": 3.0395194919468998e-06,
2953
+ "loss": 0.4934,
2954
+ "step": 245500
2955
+ },
2956
+ {
2957
+ "epoch": 2.82,
2958
+ "learning_rate": 2.9438769654539195e-06,
2959
+ "loss": 0.4917,
2960
+ "step": 246000
2961
+ },
2962
+ {
2963
+ "epoch": 2.83,
2964
+ "learning_rate": 2.8482344389609397e-06,
2965
+ "loss": 0.501,
2966
+ "step": 246500
2967
+ },
2968
+ {
2969
+ "epoch": 2.83,
2970
+ "learning_rate": 2.75259191246796e-06,
2971
+ "loss": 0.4844,
2972
+ "step": 247000
2973
+ },
2974
+ {
2975
+ "epoch": 2.84,
2976
+ "learning_rate": 2.65694938597498e-06,
2977
+ "loss": 0.4945,
2978
+ "step": 247500
2979
+ },
2980
+ {
2981
+ "epoch": 2.85,
2982
+ "learning_rate": 2.5613068594820002e-06,
2983
+ "loss": 0.4906,
2984
+ "step": 248000
2985
+ },
2986
+ {
2987
+ "epoch": 2.85,
2988
+ "learning_rate": 2.4656643329890204e-06,
2989
+ "loss": 0.4933,
2990
+ "step": 248500
2991
+ },
2992
+ {
2993
+ "epoch": 2.86,
2994
+ "learning_rate": 2.3700218064960406e-06,
2995
+ "loss": 0.4926,
2996
+ "step": 249000
2997
+ },
2998
+ {
2999
+ "epoch": 2.86,
3000
+ "learning_rate": 2.2743792800030608e-06,
3001
+ "loss": 0.4892,
3002
+ "step": 249500
3003
+ },
3004
+ {
3005
+ "epoch": 2.87,
3006
+ "learning_rate": 2.1787367535100805e-06,
3007
+ "loss": 0.4965,
3008
+ "step": 250000
3009
+ },
3010
+ {
3011
+ "epoch": 2.88,
3012
+ "learning_rate": 2.083094227017101e-06,
3013
+ "loss": 0.4979,
3014
+ "step": 250500
3015
+ },
3016
+ {
3017
+ "epoch": 2.88,
3018
+ "learning_rate": 1.9874517005241213e-06,
3019
+ "loss": 0.4997,
3020
+ "step": 251000
3021
+ },
3022
+ {
3023
+ "epoch": 2.89,
3024
+ "learning_rate": 1.8918091740311412e-06,
3025
+ "loss": 0.4931,
3026
+ "step": 251500
3027
+ },
3028
+ {
3029
+ "epoch": 2.89,
3030
+ "learning_rate": 1.7961666475381614e-06,
3031
+ "loss": 0.4998,
3032
+ "step": 252000
3033
+ },
3034
+ {
3035
+ "epoch": 2.9,
3036
+ "learning_rate": 1.7005241210451818e-06,
3037
+ "loss": 0.4904,
3038
+ "step": 252500
3039
+ },
3040
+ {
3041
+ "epoch": 2.9,
3042
+ "learning_rate": 1.6048815945522018e-06,
3043
+ "loss": 0.4939,
3044
+ "step": 253000
3045
+ },
3046
+ {
3047
+ "epoch": 2.91,
3048
+ "learning_rate": 1.509239068059222e-06,
3049
+ "loss": 0.4875,
3050
+ "step": 253500
3051
+ },
3052
+ {
3053
+ "epoch": 2.92,
3054
+ "learning_rate": 1.4135965415662421e-06,
3055
+ "loss": 0.4834,
3056
+ "step": 254000
3057
+ },
3058
+ {
3059
+ "epoch": 2.92,
3060
+ "learning_rate": 1.3179540150732623e-06,
3061
+ "loss": 0.4835,
3062
+ "step": 254500
3063
+ },
3064
+ {
3065
+ "epoch": 2.93,
3066
+ "learning_rate": 1.2223114885802825e-06,
3067
+ "loss": 0.4854,
3068
+ "step": 255000
3069
+ },
3070
+ {
3071
+ "epoch": 2.93,
3072
+ "learning_rate": 1.1266689620873024e-06,
3073
+ "loss": 0.4905,
3074
+ "step": 255500
3075
+ },
3076
+ {
3077
+ "epoch": 2.94,
3078
+ "learning_rate": 1.0310264355943226e-06,
3079
+ "loss": 0.4839,
3080
+ "step": 256000
3081
+ },
3082
+ {
3083
+ "epoch": 2.94,
3084
+ "learning_rate": 9.353839091013429e-07,
3085
+ "loss": 0.4851,
3086
+ "step": 256500
3087
+ },
3088
+ {
3089
+ "epoch": 2.95,
3090
+ "learning_rate": 8.397413826083631e-07,
3091
+ "loss": 0.4846,
3092
+ "step": 257000
3093
+ },
3094
+ {
3095
+ "epoch": 2.96,
3096
+ "learning_rate": 7.440988561153832e-07,
3097
+ "loss": 0.4837,
3098
+ "step": 257500
3099
+ },
3100
+ {
3101
+ "epoch": 2.96,
3102
+ "learning_rate": 6.484563296224033e-07,
3103
+ "loss": 0.4912,
3104
+ "step": 258000
3105
+ },
3106
+ {
3107
+ "epoch": 2.97,
3108
+ "learning_rate": 5.528138031294235e-07,
3109
+ "loss": 0.4962,
3110
+ "step": 258500
3111
+ },
3112
+ {
3113
+ "epoch": 2.97,
3114
+ "learning_rate": 4.571712766364437e-07,
3115
+ "loss": 0.4901,
3116
+ "step": 259000
3117
+ },
3118
+ {
3119
+ "epoch": 2.98,
3120
+ "learning_rate": 3.6152875014346377e-07,
3121
+ "loss": 0.495,
3122
+ "step": 259500
3123
+ },
3124
+ {
3125
+ "epoch": 2.98,
3126
+ "learning_rate": 2.65886223650484e-07,
3127
+ "loss": 0.4968,
3128
+ "step": 260000
3129
+ },
3130
+ {
3131
+ "epoch": 2.99,
3132
+ "learning_rate": 1.7024369715750412e-07,
3133
+ "loss": 0.4846,
3134
+ "step": 260500
3135
+ },
3136
+ {
3137
+ "epoch": 3.0,
3138
+ "learning_rate": 7.460117066452428e-08,
3139
+ "loss": 0.4792,
3140
+ "step": 261000
3141
+ },
3142
+ {
3143
+ "epoch": 3.0,
3144
+ "step": 261390,
3145
+ "total_flos": 2.999406650108805e+17,
3146
+ "train_loss": 0.6627257539563876,
3147
+ "train_runtime": 78487.0606,
3148
+ "train_samples_per_second": 33.304,
3149
+ "train_steps_per_second": 3.33
3150
+ }
3151
+ ],
3152
+ "max_steps": 261390,
3153
+ "num_train_epochs": 3,
3154
+ "total_flos": 2.999406650108805e+17,
3155
+ "trial_name": null,
3156
+ "trial_params": null
3157
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:598b4ef848e93d3fa2f8a806689d90cd07051d60f491e3ac1a7c8aab5db84141
3
+ size 2799
vocab.json ADDED
The diff for this file is too large to render. See raw diff