Hugging Face's logo

Ifyokoh
/
m2m100_dialect_ibo_en_mt

Translation
PyTorch
Igbo
m2m_100
Model card Files
xet
Community
1
Ifyokoh commited on
Commit
4e13d07
·
verified ·
1 Parent(s): a880637

Upload 14 files

Browse files
Files changed (14) hide show
  1. all_results.json +23 -0
  2. config.json +38 -0
  3. eval_results.json +10 -0
  4. generated_predictions.txt +0 -0
  5. generation_config.json +12 -0
  6. predict_results.json +9 -0
  7. pytorch_model.bin +3 -0
  8. sentencepiece.bpe.model +3 -0
  9. special_tokens_map.json +109 -0
  10. tokenizer_config.json +118 -0
  11. train_results.json +9 -0
  12. trainer_state.json +2899 -0
  13. training_args.bin +3 -0
  14. vocab.json +0 -0
all_results.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_bleu": 91.1198,
4
+ "eval_gen_len": 12.4142,
5
+ "eval_loss": 0.12630991637706757,
6
+ "eval_runtime": 3089.8664,
7
+ "eval_samples": 23975,
8
+ "eval_samples_per_second": 7.759,
9
+ "eval_steps_per_second": 1.94,
10
+ "predict_bleu": 90.8075,
11
+ "predict_gen_len": 12.3944,
12
+ "predict_loss": 0.13137167692184448,
13
+ "predict_runtime": 3196.0717,
14
+ "predict_samples": 23970,
15
+ "predict_samples_per_second": 7.5,
16
+ "predict_steps_per_second": 1.875,
17
+ "total_flos": 3.624187009725235e+16,
18
+ "train_loss": 0.02392078927181249,
19
+ "train_runtime": 8161.1668,
20
+ "train_samples": 191802,
21
+ "train_samples_per_second": 117.509,
22
+ "train_steps_per_second": 29.378
23
+ }
config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "masakhane/m2m100_418M_ibo_en_rel_news",
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": 128042,
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": "4.31.0",
36
+ "use_cache": true,
37
+ "vocab_size": 128104
38
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_bleu": 91.1198,
4
+ "eval_gen_len": 12.4142,
5
+ "eval_loss": 0.12630991637706757,
6
+ "eval_runtime": 3089.8664,
7
+ "eval_samples": 23975,
8
+ "eval_samples_per_second": 7.759,
9
+ "eval_steps_per_second": 1.94
10
+ }
generated_predictions.txt ADDED
The diff for this file is too large to render. See raw diff
 
generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "decoder_start_token_id": 2,
5
+ "early_stopping": true,
6
+ "eos_token_id": 2,
7
+ "forced_bos_token_id": 128022,
8
+ "max_length": 200,
9
+ "num_beams": 5,
10
+ "pad_token_id": 1,
11
+ "transformers_version": "4.31.0"
12
+ }
predict_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_bleu": 90.8075,
3
+ "predict_gen_len": 12.3944,
4
+ "predict_loss": 0.13137167692184448,
5
+ "predict_runtime": 3196.0717,
6
+ "predict_samples": 23970,
7
+ "predict_samples_per_second": 7.5,
8
+ "predict_steps_per_second": 1.875
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c67ef9e159ea95390fcb13eb00108443b1295176285819811511732353b8d1e
3
+ size 1935759303
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 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "__af__",
4
+ "__am__",
5
+ "__ar__",
6
+ "__ast__",
7
+ "__az__",
8
+ "__ba__",
9
+ "__be__",
10
+ "__bg__",
11
+ "__bn__",
12
+ "__br__",
13
+ "__bs__",
14
+ "__ca__",
15
+ "__ceb__",
16
+ "__cs__",
17
+ "__cy__",
18
+ "__da__",
19
+ "__de__",
20
+ "__el__",
21
+ "__en__",
22
+ "__es__",
23
+ "__et__",
24
+ "__fa__",
25
+ "__ff__",
26
+ "__fi__",
27
+ "__fr__",
28
+ "__fy__",
29
+ "__ga__",
30
+ "__gd__",
31
+ "__gl__",
32
+ "__gu__",
33
+ "__ha__",
34
+ "__he__",
35
+ "__hi__",
36
+ "__hr__",
37
+ "__ht__",
38
+ "__hu__",
39
+ "__hy__",
40
+ "__id__",
41
+ "__ig__",
42
+ "__ilo__",
43
+ "__is__",
44
+ "__it__",
45
+ "__ja__",
46
+ "__jv__",
47
+ "__ka__",
48
+ "__kk__",
49
+ "__km__",
50
+ "__kn__",
51
+ "__ko__",
52
+ "__lb__",
53
+ "__lg__",
54
+ "__ln__",
55
+ "__lo__",
56
+ "__lt__",
57
+ "__lv__",
58
+ "__mg__",
59
+ "__mk__",
60
+ "__ml__",
61
+ "__mn__",
62
+ "__mr__",
63
+ "__ms__",
64
+ "__my__",
65
+ "__ne__",
66
+ "__nl__",
67
+ "__no__",
68
+ "__ns__",
69
+ "__oc__",
70
+ "__or__",
71
+ "__pa__",
72
+ "__pl__",
73
+ "__ps__",
74
+ "__pt__",
75
+ "__ro__",
76
+ "__ru__",
77
+ "__sd__",
78
+ "__si__",
79
+ "__sk__",
80
+ "__sl__",
81
+ "__so__",
82
+ "__sq__",
83
+ "__sr__",
84
+ "__ss__",
85
+ "__su__",
86
+ "__sv__",
87
+ "__sw__",
88
+ "__ta__",
89
+ "__th__",
90
+ "__tl__",
91
+ "__tn__",
92
+ "__tr__",
93
+ "__uk__",
94
+ "__ur__",
95
+ "__uz__",
96
+ "__vi__",
97
+ "__wo__",
98
+ "__xh__",
99
+ "__yi__",
100
+ "__yo__",
101
+ "__zh__",
102
+ "__zu__"
103
+ ],
104
+ "bos_token": "<s>",
105
+ "eos_token": "</s>",
106
+ "pad_token": "<pad>",
107
+ "sep_token": "</s>",
108
+ "unk_token": "<unk>"
109
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "__af__",
4
+ "__am__",
5
+ "__ar__",
6
+ "__ast__",
7
+ "__az__",
8
+ "__ba__",
9
+ "__be__",
10
+ "__bg__",
11
+ "__bn__",
12
+ "__br__",
13
+ "__bs__",
14
+ "__ca__",
15
+ "__ceb__",
16
+ "__cs__",
17
+ "__cy__",
18
+ "__da__",
19
+ "__de__",
20
+ "__el__",
21
+ "__en__",
22
+ "__es__",
23
+ "__et__",
24
+ "__fa__",
25
+ "__ff__",
26
+ "__fi__",
27
+ "__fr__",
28
+ "__fy__",
29
+ "__ga__",
30
+ "__gd__",
31
+ "__gl__",
32
+ "__gu__",
33
+ "__ha__",
34
+ "__he__",
35
+ "__hi__",
36
+ "__hr__",
37
+ "__ht__",
38
+ "__hu__",
39
+ "__hy__",
40
+ "__id__",
41
+ "__ig__",
42
+ "__ilo__",
43
+ "__is__",
44
+ "__it__",
45
+ "__ja__",
46
+ "__jv__",
47
+ "__ka__",
48
+ "__kk__",
49
+ "__km__",
50
+ "__kn__",
51
+ "__ko__",
52
+ "__lb__",
53
+ "__lg__",
54
+ "__ln__",
55
+ "__lo__",
56
+ "__lt__",
57
+ "__lv__",
58
+ "__mg__",
59
+ "__mk__",
60
+ "__ml__",
61
+ "__mn__",
62
+ "__mr__",
63
+ "__ms__",
64
+ "__my__",
65
+ "__ne__",
66
+ "__nl__",
67
+ "__no__",
68
+ "__ns__",
69
+ "__oc__",
70
+ "__or__",
71
+ "__pa__",
72
+ "__pl__",
73
+ "__ps__",
74
+ "__pt__",
75
+ "__ro__",
76
+ "__ru__",
77
+ "__sd__",
78
+ "__si__",
79
+ "__sk__",
80
+ "__sl__",
81
+ "__so__",
82
+ "__sq__",
83
+ "__sr__",
84
+ "__ss__",
85
+ "__su__",
86
+ "__sv__",
87
+ "__sw__",
88
+ "__ta__",
89
+ "__th__",
90
+ "__tl__",
91
+ "__tn__",
92
+ "__tr__",
93
+ "__uk__",
94
+ "__ur__",
95
+ "__uz__",
96
+ "__vi__",
97
+ "__wo__",
98
+ "__xh__",
99
+ "__yi__",
100
+ "__yo__",
101
+ "__zh__",
102
+ "__zu__"
103
+ ],
104
+ "bos_token": "<s>",
105
+ "clean_up_tokenization_spaces": true,
106
+ "eos_token": "</s>",
107
+ "language_codes": "m2m100",
108
+ "model_max_length": 1024,
109
+ "num_madeup_words": 8,
110
+ "pad_token": "<pad>",
111
+ "sep_token": "</s>",
112
+ "sp_model_kwargs": {},
113
+ "src_lang": null,
114
+ "tgt_lang": null,
115
+ "tokenizer_class": "M2M100Tokenizer",
116
+ "tokenizer_file": null,
117
+ "unk_token": "<unk>"
118
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "total_flos": 3.624187009725235e+16,
4
+ "train_loss": 0.02392078927181249,
5
+ "train_runtime": 8161.1668,
6
+ "train_samples": 191802,
7
+ "train_samples_per_second": 117.509,
8
+ "train_steps_per_second": 29.378
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,2899 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "global_step": 239755,
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": 4.9895726887864694e-05,
13
+ "loss": 2.0097,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.979145377572939e-05,
19
+ "loss": 1.8204,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 4.968718066359409e-05,
25
+ "loss": 1.6177,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 4.958290755145878e-05,
31
+ "loss": 1.543,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 4.947863443932348e-05,
37
+ "loss": 1.5073,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.06,
42
+ "learning_rate": 4.937436132718817e-05,
43
+ "loss": 1.4273,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.07,
48
+ "learning_rate": 4.9270088215052865e-05,
49
+ "loss": 1.3995,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.08,
54
+ "learning_rate": 4.916581510291756e-05,
55
+ "loss": 1.2902,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.09,
60
+ "learning_rate": 4.906154199078226e-05,
61
+ "loss": 1.2541,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.1,
66
+ "learning_rate": 4.895726887864695e-05,
67
+ "loss": 1.1981,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.11,
72
+ "learning_rate": 4.885299576651165e-05,
73
+ "loss": 1.1933,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.13,
78
+ "learning_rate": 4.8748722654376343e-05,
79
+ "loss": 1.1827,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.14,
84
+ "learning_rate": 4.8644449542241035e-05,
85
+ "loss": 1.1112,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.15,
90
+ "learning_rate": 4.8540176430105734e-05,
91
+ "loss": 1.1385,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.16,
96
+ "learning_rate": 4.843590331797043e-05,
97
+ "loss": 1.0471,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.17,
102
+ "learning_rate": 4.8331630205835124e-05,
103
+ "loss": 1.0287,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.18,
108
+ "learning_rate": 4.822735709369982e-05,
109
+ "loss": 1.0343,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.19,
114
+ "learning_rate": 4.8123083981564514e-05,
115
+ "loss": 1.0191,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.2,
120
+ "learning_rate": 4.801881086942921e-05,
121
+ "loss": 1.0007,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.21,
126
+ "learning_rate": 4.791453775729391e-05,
127
+ "loss": 0.9207,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.22,
132
+ "learning_rate": 4.78102646451586e-05,
133
+ "loss": 0.9356,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.23,
138
+ "learning_rate": 4.7705991533023294e-05,
139
+ "loss": 0.8999,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.24,
144
+ "learning_rate": 4.760171842088799e-05,
145
+ "loss": 0.9197,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.25,
150
+ "learning_rate": 4.7497445308752684e-05,
151
+ "loss": 0.8851,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.26,
156
+ "learning_rate": 4.739317219661738e-05,
157
+ "loss": 0.8467,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.27,
162
+ "learning_rate": 4.728889908448208e-05,
163
+ "loss": 0.8545,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.28,
168
+ "learning_rate": 4.718462597234677e-05,
169
+ "loss": 0.844,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.29,
174
+ "learning_rate": 4.708035286021147e-05,
175
+ "loss": 0.8036,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.3,
180
+ "learning_rate": 4.697607974807616e-05,
181
+ "loss": 0.7726,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.31,
186
+ "learning_rate": 4.6871806635940855e-05,
187
+ "loss": 0.8192,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.32,
192
+ "learning_rate": 4.676753352380555e-05,
193
+ "loss": 0.786,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.33,
198
+ "learning_rate": 4.666326041167025e-05,
199
+ "loss": 0.7673,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.34,
204
+ "learning_rate": 4.6558987299534944e-05,
205
+ "loss": 0.7822,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.35,
210
+ "learning_rate": 4.645471418739964e-05,
211
+ "loss": 0.7499,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.36,
216
+ "learning_rate": 4.6350441075264334e-05,
217
+ "loss": 0.7388,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.38,
222
+ "learning_rate": 4.6246167963129025e-05,
223
+ "loss": 0.7096,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.39,
228
+ "learning_rate": 4.614189485099373e-05,
229
+ "loss": 0.6983,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.4,
234
+ "learning_rate": 4.603762173885842e-05,
235
+ "loss": 0.712,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.41,
240
+ "learning_rate": 4.5933348626723114e-05,
241
+ "loss": 0.6714,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.42,
246
+ "learning_rate": 4.582907551458781e-05,
247
+ "loss": 0.6959,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.43,
252
+ "learning_rate": 4.5724802402452504e-05,
253
+ "loss": 0.6654,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.44,
258
+ "learning_rate": 4.5620529290317196e-05,
259
+ "loss": 0.659,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.45,
264
+ "learning_rate": 4.55162561781819e-05,
265
+ "loss": 0.6296,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.46,
270
+ "learning_rate": 4.541198306604659e-05,
271
+ "loss": 0.6371,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.47,
276
+ "learning_rate": 4.5307709953911285e-05,
277
+ "loss": 0.6505,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.48,
282
+ "learning_rate": 4.520343684177598e-05,
283
+ "loss": 0.621,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.49,
288
+ "learning_rate": 4.5099163729640675e-05,
289
+ "loss": 0.6013,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.5,
294
+ "learning_rate": 4.4994890617505366e-05,
295
+ "loss": 0.6206,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.51,
300
+ "learning_rate": 4.489061750537007e-05,
301
+ "loss": 0.5886,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.52,
306
+ "learning_rate": 4.478634439323476e-05,
307
+ "loss": 0.5938,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.53,
312
+ "learning_rate": 4.4682071281099455e-05,
313
+ "loss": 0.5512,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.54,
318
+ "learning_rate": 4.4577798168964154e-05,
319
+ "loss": 0.5524,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.55,
324
+ "learning_rate": 4.4473525056828845e-05,
325
+ "loss": 0.5555,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.56,
330
+ "learning_rate": 4.4369251944693544e-05,
331
+ "loss": 0.5736,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.57,
336
+ "learning_rate": 4.426497883255824e-05,
337
+ "loss": 0.5437,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.58,
342
+ "learning_rate": 4.4160705720422934e-05,
343
+ "loss": 0.5492,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.59,
348
+ "learning_rate": 4.4056432608287626e-05,
349
+ "loss": 0.5497,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.6,
354
+ "learning_rate": 4.3952159496152324e-05,
355
+ "loss": 0.538,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.62,
360
+ "learning_rate": 4.3847886384017016e-05,
361
+ "loss": 0.5215,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.63,
366
+ "learning_rate": 4.3743613271881714e-05,
367
+ "loss": 0.4956,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.64,
372
+ "learning_rate": 4.363934015974641e-05,
373
+ "loss": 0.5114,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.65,
378
+ "learning_rate": 4.3535067047611104e-05,
379
+ "loss": 0.51,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.66,
384
+ "learning_rate": 4.34307939354758e-05,
385
+ "loss": 0.5159,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.67,
390
+ "learning_rate": 4.3326520823340495e-05,
391
+ "loss": 0.4825,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.68,
396
+ "learning_rate": 4.3222247711205186e-05,
397
+ "loss": 0.495,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.69,
402
+ "learning_rate": 4.3117974599069885e-05,
403
+ "loss": 0.4683,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.7,
408
+ "learning_rate": 4.301370148693458e-05,
409
+ "loss": 0.4663,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.71,
414
+ "learning_rate": 4.2909428374799275e-05,
415
+ "loss": 0.4545,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.72,
420
+ "learning_rate": 4.280515526266397e-05,
421
+ "loss": 0.4684,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.73,
426
+ "learning_rate": 4.2700882150528665e-05,
427
+ "loss": 0.4444,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.74,
432
+ "learning_rate": 4.2596609038393363e-05,
433
+ "loss": 0.4467,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.75,
438
+ "learning_rate": 4.249233592625806e-05,
439
+ "loss": 0.4538,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.76,
444
+ "learning_rate": 4.2388062814122754e-05,
445
+ "loss": 0.4192,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.77,
450
+ "learning_rate": 4.2283789701987445e-05,
451
+ "loss": 0.4444,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.78,
456
+ "learning_rate": 4.2179516589852144e-05,
457
+ "loss": 0.4396,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.79,
462
+ "learning_rate": 4.2075243477716836e-05,
463
+ "loss": 0.4456,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.8,
468
+ "learning_rate": 4.1970970365581534e-05,
469
+ "loss": 0.4337,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.81,
474
+ "learning_rate": 4.186669725344623e-05,
475
+ "loss": 0.4108,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.82,
480
+ "learning_rate": 4.1762424141310924e-05,
481
+ "loss": 0.4416,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.83,
486
+ "learning_rate": 4.1658151029175616e-05,
487
+ "loss": 0.4061,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.84,
492
+ "learning_rate": 4.1553877917040314e-05,
493
+ "loss": 0.3816,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.86,
498
+ "learning_rate": 4.1449604804905006e-05,
499
+ "loss": 0.3916,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.87,
504
+ "learning_rate": 4.1345331692769704e-05,
505
+ "loss": 0.3914,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.88,
510
+ "learning_rate": 4.12410585806344e-05,
511
+ "loss": 0.4327,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.89,
516
+ "learning_rate": 4.1136785468499095e-05,
517
+ "loss": 0.3784,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.9,
522
+ "learning_rate": 4.1032512356363786e-05,
523
+ "loss": 0.4018,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.91,
528
+ "learning_rate": 4.0928239244228485e-05,
529
+ "loss": 0.3416,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.92,
534
+ "learning_rate": 4.0823966132093177e-05,
535
+ "loss": 0.3602,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.93,
540
+ "learning_rate": 4.0719693019957875e-05,
541
+ "loss": 0.3911,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.94,
546
+ "learning_rate": 4.0615419907822573e-05,
547
+ "loss": 0.4008,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.95,
552
+ "learning_rate": 4.0511146795687265e-05,
553
+ "loss": 0.3659,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.96,
558
+ "learning_rate": 4.040687368355196e-05,
559
+ "loss": 0.3536,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.97,
564
+ "learning_rate": 4.0302600571416655e-05,
565
+ "loss": 0.3782,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.98,
570
+ "learning_rate": 4.0198327459281354e-05,
571
+ "loss": 0.3624,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.99,
576
+ "learning_rate": 4.0094054347146045e-05,
577
+ "loss": 0.3284,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 1.0,
582
+ "learning_rate": 3.9989781235010744e-05,
583
+ "loss": 0.1845,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 1.01,
588
+ "learning_rate": 3.9885508122875436e-05,
589
+ "loss": 0.2659,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 1.02,
594
+ "learning_rate": 3.9781235010740134e-05,
595
+ "loss": 0.2861,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 1.03,
600
+ "learning_rate": 3.9676961898604826e-05,
601
+ "loss": 0.2826,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 1.04,
606
+ "learning_rate": 3.9572688786469524e-05,
607
+ "loss": 0.2703,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 1.05,
612
+ "learning_rate": 3.9468415674334216e-05,
613
+ "loss": 0.2455,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 1.06,
618
+ "learning_rate": 3.9364142562198914e-05,
619
+ "loss": 0.2555,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 1.07,
624
+ "learning_rate": 3.9259869450063606e-05,
625
+ "loss": 0.262,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 1.08,
630
+ "learning_rate": 3.9155596337928305e-05,
631
+ "loss": 0.2483,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 1.09,
636
+ "learning_rate": 3.9051323225792996e-05,
637
+ "loss": 0.2604,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 1.11,
642
+ "learning_rate": 3.8947050113657695e-05,
643
+ "loss": 0.2376,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 1.12,
648
+ "learning_rate": 3.884277700152239e-05,
649
+ "loss": 0.2499,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 1.13,
654
+ "learning_rate": 3.8738503889387085e-05,
655
+ "loss": 0.2632,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 1.14,
660
+ "learning_rate": 3.863423077725178e-05,
661
+ "loss": 0.2506,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 1.15,
666
+ "learning_rate": 3.8529957665116475e-05,
667
+ "loss": 0.2541,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 1.16,
672
+ "learning_rate": 3.8425684552981174e-05,
673
+ "loss": 0.2272,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 1.17,
678
+ "learning_rate": 3.8321411440845865e-05,
679
+ "loss": 0.2688,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 1.18,
684
+ "learning_rate": 3.8217138328710564e-05,
685
+ "loss": 0.235,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 1.19,
690
+ "learning_rate": 3.8112865216575255e-05,
691
+ "loss": 0.2608,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 1.2,
696
+ "learning_rate": 3.800859210443995e-05,
697
+ "loss": 0.2506,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 1.21,
702
+ "learning_rate": 3.7904318992304646e-05,
703
+ "loss": 0.2437,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 1.22,
708
+ "learning_rate": 3.7800045880169344e-05,
709
+ "loss": 0.2361,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 1.23,
714
+ "learning_rate": 3.7695772768034036e-05,
715
+ "loss": 0.2512,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 1.24,
720
+ "learning_rate": 3.7591499655898734e-05,
721
+ "loss": 0.2152,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 1.25,
726
+ "learning_rate": 3.7487226543763426e-05,
727
+ "loss": 0.2545,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 1.26,
732
+ "learning_rate": 3.738295343162812e-05,
733
+ "loss": 0.2286,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 1.27,
738
+ "learning_rate": 3.7278680319492816e-05,
739
+ "loss": 0.2297,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 1.28,
744
+ "learning_rate": 3.7174407207357515e-05,
745
+ "loss": 0.2478,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 1.29,
750
+ "learning_rate": 3.7070134095222206e-05,
751
+ "loss": 0.2344,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 1.3,
756
+ "learning_rate": 3.6965860983086905e-05,
757
+ "loss": 0.2134,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 1.31,
762
+ "learning_rate": 3.6861587870951596e-05,
763
+ "loss": 0.241,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 1.32,
768
+ "learning_rate": 3.675731475881629e-05,
769
+ "loss": 0.2374,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 1.33,
774
+ "learning_rate": 3.6653041646680993e-05,
775
+ "loss": 0.216,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 1.35,
780
+ "learning_rate": 3.6548768534545685e-05,
781
+ "loss": 0.2258,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 1.36,
786
+ "learning_rate": 3.644449542241038e-05,
787
+ "loss": 0.2325,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 1.37,
792
+ "learning_rate": 3.6340222310275075e-05,
793
+ "loss": 0.2286,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 1.38,
798
+ "learning_rate": 3.623594919813977e-05,
799
+ "loss": 0.2417,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 1.39,
804
+ "learning_rate": 3.6131676086004465e-05,
805
+ "loss": 0.2324,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 1.4,
810
+ "learning_rate": 3.6027402973869164e-05,
811
+ "loss": 0.2056,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 1.41,
816
+ "learning_rate": 3.5923129861733856e-05,
817
+ "loss": 0.231,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 1.42,
822
+ "learning_rate": 3.581885674959855e-05,
823
+ "loss": 0.2391,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 1.43,
828
+ "learning_rate": 3.5714583637463246e-05,
829
+ "loss": 0.2188,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 1.44,
834
+ "learning_rate": 3.561031052532794e-05,
835
+ "loss": 0.213,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 1.45,
840
+ "learning_rate": 3.5506037413192636e-05,
841
+ "loss": 0.2096,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 1.46,
846
+ "learning_rate": 3.5401764301057334e-05,
847
+ "loss": 0.24,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 1.47,
852
+ "learning_rate": 3.5297491188922026e-05,
853
+ "loss": 0.2246,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 1.48,
858
+ "learning_rate": 3.5193218076786725e-05,
859
+ "loss": 0.2104,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 1.49,
864
+ "learning_rate": 3.5088944964651416e-05,
865
+ "loss": 0.2397,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 1.5,
870
+ "learning_rate": 3.498467185251611e-05,
871
+ "loss": 0.2076,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 1.51,
876
+ "learning_rate": 3.4880398740380806e-05,
877
+ "loss": 0.1854,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 1.52,
882
+ "learning_rate": 3.4776125628245505e-05,
883
+ "loss": 0.2053,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 1.53,
888
+ "learning_rate": 3.4671852516110197e-05,
889
+ "loss": 0.1901,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 1.54,
894
+ "learning_rate": 3.4567579403974895e-05,
895
+ "loss": 0.2008,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 1.55,
900
+ "learning_rate": 3.446330629183959e-05,
901
+ "loss": 0.2115,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 1.56,
906
+ "learning_rate": 3.435903317970428e-05,
907
+ "loss": 0.2057,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 1.57,
912
+ "learning_rate": 3.4254760067568984e-05,
913
+ "loss": 0.1976,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 1.58,
918
+ "learning_rate": 3.4150486955433675e-05,
919
+ "loss": 0.1856,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 1.6,
924
+ "learning_rate": 3.404621384329837e-05,
925
+ "loss": 0.2248,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 1.61,
930
+ "learning_rate": 3.3941940731163066e-05,
931
+ "loss": 0.1992,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 1.62,
936
+ "learning_rate": 3.383766761902776e-05,
937
+ "loss": 0.2046,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 1.63,
942
+ "learning_rate": 3.373339450689245e-05,
943
+ "loss": 0.1814,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 1.64,
948
+ "learning_rate": 3.3629121394757154e-05,
949
+ "loss": 0.193,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 1.65,
954
+ "learning_rate": 3.3524848282621846e-05,
955
+ "loss": 0.1805,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 1.66,
960
+ "learning_rate": 3.342057517048654e-05,
961
+ "loss": 0.2065,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 1.67,
966
+ "learning_rate": 3.3316302058351236e-05,
967
+ "loss": 0.1871,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 1.68,
972
+ "learning_rate": 3.321202894621593e-05,
973
+ "loss": 0.191,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 1.69,
978
+ "learning_rate": 3.3107755834080626e-05,
979
+ "loss": 0.1948,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 1.7,
984
+ "learning_rate": 3.3003482721945325e-05,
985
+ "loss": 0.1803,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 1.71,
990
+ "learning_rate": 3.2899209609810016e-05,
991
+ "loss": 0.1796,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 1.72,
996
+ "learning_rate": 3.279493649767471e-05,
997
+ "loss": 0.1847,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 1.73,
1002
+ "learning_rate": 3.2690663385539407e-05,
1003
+ "loss": 0.1512,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 1.74,
1008
+ "learning_rate": 3.25863902734041e-05,
1009
+ "loss": 0.2074,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 1.75,
1014
+ "learning_rate": 3.24821171612688e-05,
1015
+ "loss": 0.1718,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 1.76,
1020
+ "learning_rate": 3.2377844049133495e-05,
1021
+ "loss": 0.1735,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 1.77,
1026
+ "learning_rate": 3.227357093699819e-05,
1027
+ "loss": 0.1928,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 1.78,
1032
+ "learning_rate": 3.216929782486288e-05,
1033
+ "loss": 0.1916,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 1.79,
1038
+ "learning_rate": 3.206502471272758e-05,
1039
+ "loss": 0.1708,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 1.8,
1044
+ "learning_rate": 3.196075160059227e-05,
1045
+ "loss": 0.1976,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 1.81,
1050
+ "learning_rate": 3.185647848845697e-05,
1051
+ "loss": 0.1646,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 1.82,
1056
+ "learning_rate": 3.1752205376321666e-05,
1057
+ "loss": 0.1641,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 1.84,
1062
+ "learning_rate": 3.164793226418636e-05,
1063
+ "loss": 0.1692,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 1.85,
1068
+ "learning_rate": 3.1543659152051056e-05,
1069
+ "loss": 0.1572,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 1.86,
1074
+ "learning_rate": 3.143938603991575e-05,
1075
+ "loss": 0.1892,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 1.87,
1080
+ "learning_rate": 3.1335112927780446e-05,
1081
+ "loss": 0.1725,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 1.88,
1086
+ "learning_rate": 3.123083981564514e-05,
1087
+ "loss": 0.1531,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 1.89,
1092
+ "learning_rate": 3.1126566703509836e-05,
1093
+ "loss": 0.1645,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 1.9,
1098
+ "learning_rate": 3.102229359137453e-05,
1099
+ "loss": 0.1838,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 1.91,
1104
+ "learning_rate": 3.0918020479239226e-05,
1105
+ "loss": 0.187,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 1.92,
1110
+ "learning_rate": 3.081374736710392e-05,
1111
+ "loss": 0.1606,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 1.93,
1116
+ "learning_rate": 3.0709474254968617e-05,
1117
+ "loss": 0.1669,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 1.94,
1122
+ "learning_rate": 3.0605201142833315e-05,
1123
+ "loss": 0.1454,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 1.95,
1128
+ "learning_rate": 3.0500928030698007e-05,
1129
+ "loss": 0.1528,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 1.96,
1134
+ "learning_rate": 3.0396654918562702e-05,
1135
+ "loss": 0.1522,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 1.97,
1140
+ "learning_rate": 3.0292381806427393e-05,
1141
+ "loss": 0.1398,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 1.98,
1146
+ "learning_rate": 3.018810869429209e-05,
1147
+ "loss": 0.1848,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 1.99,
1152
+ "learning_rate": 3.0083835582156787e-05,
1153
+ "loss": 0.1568,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 2.0,
1158
+ "learning_rate": 2.9979562470021482e-05,
1159
+ "loss": 0.0976,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 2.01,
1164
+ "learning_rate": 2.9875289357886177e-05,
1165
+ "loss": 0.0937,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 2.02,
1170
+ "learning_rate": 2.9771016245750872e-05,
1171
+ "loss": 0.1108,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 2.03,
1176
+ "learning_rate": 2.9666743133615564e-05,
1177
+ "loss": 0.1167,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 2.04,
1182
+ "learning_rate": 2.9562470021480266e-05,
1183
+ "loss": 0.0964,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 2.05,
1188
+ "learning_rate": 2.945819690934496e-05,
1189
+ "loss": 0.1095,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 2.06,
1194
+ "learning_rate": 2.9353923797209653e-05,
1195
+ "loss": 0.1088,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 2.08,
1200
+ "learning_rate": 2.9249650685074348e-05,
1201
+ "loss": 0.1049,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 2.09,
1206
+ "learning_rate": 2.9145377572939043e-05,
1207
+ "loss": 0.1072,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 2.1,
1212
+ "learning_rate": 2.9041104460803738e-05,
1213
+ "loss": 0.1173,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 2.11,
1218
+ "learning_rate": 2.8936831348668436e-05,
1219
+ "loss": 0.1058,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 2.12,
1224
+ "learning_rate": 2.883255823653313e-05,
1225
+ "loss": 0.1017,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 2.13,
1230
+ "learning_rate": 2.8728285124397823e-05,
1231
+ "loss": 0.1008,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 2.14,
1236
+ "learning_rate": 2.8624012012262518e-05,
1237
+ "loss": 0.1167,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 2.15,
1242
+ "learning_rate": 2.8519738900127213e-05,
1243
+ "loss": 0.0925,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 2.16,
1248
+ "learning_rate": 2.841546578799191e-05,
1249
+ "loss": 0.1107,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 2.17,
1254
+ "learning_rate": 2.8311192675856607e-05,
1255
+ "loss": 0.1131,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 2.18,
1260
+ "learning_rate": 2.8206919563721302e-05,
1261
+ "loss": 0.1059,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 2.19,
1266
+ "learning_rate": 2.8102646451585997e-05,
1267
+ "loss": 0.1073,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 2.2,
1272
+ "learning_rate": 2.799837333945069e-05,
1273
+ "loss": 0.112,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 2.21,
1278
+ "learning_rate": 2.7894100227315384e-05,
1279
+ "loss": 0.1114,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 2.22,
1284
+ "learning_rate": 2.7789827115180082e-05,
1285
+ "loss": 0.1333,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 2.23,
1290
+ "learning_rate": 2.7685554003044777e-05,
1291
+ "loss": 0.1007,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 2.24,
1296
+ "learning_rate": 2.7581280890909472e-05,
1297
+ "loss": 0.1127,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 2.25,
1302
+ "learning_rate": 2.7477007778774168e-05,
1303
+ "loss": 0.1023,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 2.26,
1308
+ "learning_rate": 2.737273466663886e-05,
1309
+ "loss": 0.0971,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 2.27,
1314
+ "learning_rate": 2.7268461554503554e-05,
1315
+ "loss": 0.1018,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 2.28,
1320
+ "learning_rate": 2.7164188442368256e-05,
1321
+ "loss": 0.1137,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 2.29,
1326
+ "learning_rate": 2.7059915330232948e-05,
1327
+ "loss": 0.106,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 2.3,
1332
+ "learning_rate": 2.6955642218097643e-05,
1333
+ "loss": 0.0998,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 2.31,
1338
+ "learning_rate": 2.6851369105962338e-05,
1339
+ "loss": 0.0934,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 2.33,
1344
+ "learning_rate": 2.6747095993827033e-05,
1345
+ "loss": 0.1166,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 2.34,
1350
+ "learning_rate": 2.6642822881691725e-05,
1351
+ "loss": 0.1028,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 2.35,
1356
+ "learning_rate": 2.6538549769556427e-05,
1357
+ "loss": 0.1005,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 2.36,
1362
+ "learning_rate": 2.643427665742112e-05,
1363
+ "loss": 0.1105,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 2.37,
1368
+ "learning_rate": 2.6330003545285813e-05,
1369
+ "loss": 0.1114,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 2.38,
1374
+ "learning_rate": 2.622573043315051e-05,
1375
+ "loss": 0.0951,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 2.39,
1380
+ "learning_rate": 2.6121457321015204e-05,
1381
+ "loss": 0.1141,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 2.4,
1386
+ "learning_rate": 2.6017184208879895e-05,
1387
+ "loss": 0.0885,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 2.41,
1392
+ "learning_rate": 2.5912911096744597e-05,
1393
+ "loss": 0.112,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 2.42,
1398
+ "learning_rate": 2.5808637984609292e-05,
1399
+ "loss": 0.1167,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 2.43,
1404
+ "learning_rate": 2.5704364872473984e-05,
1405
+ "loss": 0.0975,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 2.44,
1410
+ "learning_rate": 2.560009176033868e-05,
1411
+ "loss": 0.0982,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 2.45,
1416
+ "learning_rate": 2.5495818648203374e-05,
1417
+ "loss": 0.1061,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 2.46,
1422
+ "learning_rate": 2.5391545536068073e-05,
1423
+ "loss": 0.1037,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 2.47,
1428
+ "learning_rate": 2.5287272423932768e-05,
1429
+ "loss": 0.0899,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 2.48,
1434
+ "learning_rate": 2.5182999311797463e-05,
1435
+ "loss": 0.0965,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 2.49,
1440
+ "learning_rate": 2.5078726199662154e-05,
1441
+ "loss": 0.1122,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 2.5,
1446
+ "learning_rate": 2.4974453087526853e-05,
1447
+ "loss": 0.1096,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 2.51,
1452
+ "learning_rate": 2.4870179975391548e-05,
1453
+ "loss": 0.0948,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 2.52,
1458
+ "learning_rate": 2.476590686325624e-05,
1459
+ "loss": 0.0827,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 2.53,
1464
+ "learning_rate": 2.4661633751120938e-05,
1465
+ "loss": 0.0865,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 2.54,
1470
+ "learning_rate": 2.4557360638985633e-05,
1471
+ "loss": 0.0936,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 2.55,
1476
+ "learning_rate": 2.445308752685033e-05,
1477
+ "loss": 0.1097,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 2.57,
1482
+ "learning_rate": 2.4348814414715023e-05,
1483
+ "loss": 0.1022,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 2.58,
1488
+ "learning_rate": 2.424454130257972e-05,
1489
+ "loss": 0.089,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 2.59,
1494
+ "learning_rate": 2.4140268190444414e-05,
1495
+ "loss": 0.0932,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 2.6,
1500
+ "learning_rate": 2.403599507830911e-05,
1501
+ "loss": 0.0871,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 2.61,
1506
+ "learning_rate": 2.3931721966173804e-05,
1507
+ "loss": 0.1136,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 2.62,
1512
+ "learning_rate": 2.38274488540385e-05,
1513
+ "loss": 0.0951,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 2.63,
1518
+ "learning_rate": 2.3723175741903194e-05,
1519
+ "loss": 0.098,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 2.64,
1524
+ "learning_rate": 2.361890262976789e-05,
1525
+ "loss": 0.0828,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 2.65,
1530
+ "learning_rate": 2.3514629517632584e-05,
1531
+ "loss": 0.1078,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 2.66,
1536
+ "learning_rate": 2.341035640549728e-05,
1537
+ "loss": 0.0896,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 2.67,
1542
+ "learning_rate": 2.3306083293361974e-05,
1543
+ "loss": 0.0901,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 2.68,
1548
+ "learning_rate": 2.3201810181226673e-05,
1549
+ "loss": 0.0887,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 2.69,
1554
+ "learning_rate": 2.3097537069091364e-05,
1555
+ "loss": 0.0915,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 2.7,
1560
+ "learning_rate": 2.299326395695606e-05,
1561
+ "loss": 0.0857,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 2.71,
1566
+ "learning_rate": 2.2888990844820758e-05,
1567
+ "loss": 0.1049,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 2.72,
1572
+ "learning_rate": 2.278471773268545e-05,
1573
+ "loss": 0.1008,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 2.73,
1578
+ "learning_rate": 2.2680444620550145e-05,
1579
+ "loss": 0.0904,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 2.74,
1584
+ "learning_rate": 2.2576171508414843e-05,
1585
+ "loss": 0.0968,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 2.75,
1590
+ "learning_rate": 2.2471898396279535e-05,
1591
+ "loss": 0.0824,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 2.76,
1596
+ "learning_rate": 2.236762528414423e-05,
1597
+ "loss": 0.0866,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 2.77,
1602
+ "learning_rate": 2.226335217200893e-05,
1603
+ "loss": 0.0836,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 2.78,
1608
+ "learning_rate": 2.2159079059873624e-05,
1609
+ "loss": 0.1039,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 2.79,
1614
+ "learning_rate": 2.2054805947738315e-05,
1615
+ "loss": 0.0776,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 2.8,
1620
+ "learning_rate": 2.1950532835603014e-05,
1621
+ "loss": 0.0774,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 2.82,
1626
+ "learning_rate": 2.184625972346771e-05,
1627
+ "loss": 0.0937,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 2.83,
1632
+ "learning_rate": 2.17419866113324e-05,
1633
+ "loss": 0.0965,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 2.84,
1638
+ "learning_rate": 2.16377134991971e-05,
1639
+ "loss": 0.0917,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 2.85,
1644
+ "learning_rate": 2.1533440387061794e-05,
1645
+ "loss": 0.0817,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 2.86,
1650
+ "learning_rate": 2.1429167274926486e-05,
1651
+ "loss": 0.0769,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 2.87,
1656
+ "learning_rate": 2.1324894162791184e-05,
1657
+ "loss": 0.1112,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 2.88,
1662
+ "learning_rate": 2.122062105065588e-05,
1663
+ "loss": 0.0763,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 2.89,
1668
+ "learning_rate": 2.1116347938520574e-05,
1669
+ "loss": 0.0891,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 2.9,
1674
+ "learning_rate": 2.101207482638527e-05,
1675
+ "loss": 0.0885,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 2.91,
1680
+ "learning_rate": 2.0907801714249965e-05,
1681
+ "loss": 0.0709,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 2.92,
1686
+ "learning_rate": 2.080352860211466e-05,
1687
+ "loss": 0.0844,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 2.93,
1692
+ "learning_rate": 2.0699255489979355e-05,
1693
+ "loss": 0.0868,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 2.94,
1698
+ "learning_rate": 2.059498237784405e-05,
1699
+ "loss": 0.0867,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 2.95,
1704
+ "learning_rate": 2.0490709265708745e-05,
1705
+ "loss": 0.0902,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 2.96,
1710
+ "learning_rate": 2.038643615357344e-05,
1711
+ "loss": 0.0807,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 2.97,
1716
+ "learning_rate": 2.0282163041438135e-05,
1717
+ "loss": 0.0819,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 2.98,
1722
+ "learning_rate": 2.017788992930283e-05,
1723
+ "loss": 0.0979,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 2.99,
1728
+ "learning_rate": 2.0073616817167525e-05,
1729
+ "loss": 0.0796,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 3.0,
1734
+ "learning_rate": 4.9969343705032224e-05,
1735
+ "loss": 0.0773,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 3.01,
1740
+ "learning_rate": 4.9865070592896916e-05,
1741
+ "loss": 0.0796,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 3.02,
1746
+ "learning_rate": 4.9760797480761615e-05,
1747
+ "loss": 0.0831,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 3.03,
1752
+ "learning_rate": 4.9656524368626306e-05,
1753
+ "loss": 0.0935,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 3.04,
1758
+ "learning_rate": 4.9552251256491005e-05,
1759
+ "loss": 0.0973,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 3.06,
1764
+ "learning_rate": 4.94479781443557e-05,
1765
+ "loss": 0.1146,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 3.07,
1770
+ "learning_rate": 4.9343705032220395e-05,
1771
+ "loss": 0.1104,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 3.08,
1776
+ "learning_rate": 4.923943192008509e-05,
1777
+ "loss": 0.109,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 3.09,
1782
+ "learning_rate": 4.9135158807949785e-05,
1783
+ "loss": 0.1118,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 3.1,
1788
+ "learning_rate": 4.903088569581448e-05,
1789
+ "loss": 0.1169,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 3.11,
1794
+ "learning_rate": 4.8926612583679175e-05,
1795
+ "loss": 0.1463,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 3.12,
1800
+ "learning_rate": 4.8822339471543874e-05,
1801
+ "loss": 0.1227,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 3.13,
1806
+ "learning_rate": 4.8718066359408565e-05,
1807
+ "loss": 0.1233,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 3.14,
1812
+ "learning_rate": 4.861379324727326e-05,
1813
+ "loss": 0.1258,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 3.15,
1818
+ "learning_rate": 4.8509520135137956e-05,
1819
+ "loss": 0.1439,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 3.16,
1824
+ "learning_rate": 4.8405247023002654e-05,
1825
+ "loss": 0.1225,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 3.17,
1830
+ "learning_rate": 4.8300973910867346e-05,
1831
+ "loss": 0.1204,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 3.18,
1836
+ "learning_rate": 4.8196700798732044e-05,
1837
+ "loss": 0.1256,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 3.19,
1842
+ "learning_rate": 4.8092427686596736e-05,
1843
+ "loss": 0.1498,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 3.2,
1848
+ "learning_rate": 4.798815457446143e-05,
1849
+ "loss": 0.1416,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 3.21,
1854
+ "learning_rate": 4.7883881462326126e-05,
1855
+ "loss": 0.1263,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 3.22,
1860
+ "learning_rate": 4.7779608350190825e-05,
1861
+ "loss": 0.1285,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 3.23,
1866
+ "learning_rate": 4.7675335238055516e-05,
1867
+ "loss": 0.1294,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 3.24,
1872
+ "learning_rate": 4.7571062125920215e-05,
1873
+ "loss": 0.1491,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 3.25,
1878
+ "learning_rate": 4.7466789013784906e-05,
1879
+ "loss": 0.1495,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 3.26,
1884
+ "learning_rate": 4.73625159016496e-05,
1885
+ "loss": 0.1414,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 3.27,
1890
+ "learning_rate": 4.7258242789514297e-05,
1891
+ "loss": 0.1384,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 3.28,
1896
+ "learning_rate": 4.7153969677378995e-05,
1897
+ "loss": 0.1376,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 3.3,
1902
+ "learning_rate": 4.704969656524369e-05,
1903
+ "loss": 0.1456,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 3.31,
1908
+ "learning_rate": 4.6945423453108385e-05,
1909
+ "loss": 0.1326,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 3.32,
1914
+ "learning_rate": 4.684115034097308e-05,
1915
+ "loss": 0.1676,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 3.33,
1920
+ "learning_rate": 4.6736877228837775e-05,
1921
+ "loss": 0.1485,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 3.34,
1926
+ "learning_rate": 4.6632604116702474e-05,
1927
+ "loss": 0.1319,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 3.35,
1932
+ "learning_rate": 4.6528331004567166e-05,
1933
+ "loss": 0.1349,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 3.36,
1938
+ "learning_rate": 4.642405789243186e-05,
1939
+ "loss": 0.1386,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 3.37,
1944
+ "learning_rate": 4.6319784780296556e-05,
1945
+ "loss": 0.1562,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 3.38,
1950
+ "learning_rate": 4.621551166816125e-05,
1951
+ "loss": 0.1441,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 3.39,
1956
+ "learning_rate": 4.6111238556025946e-05,
1957
+ "loss": 0.1413,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 3.4,
1962
+ "learning_rate": 4.6006965443890644e-05,
1963
+ "loss": 0.1614,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 3.41,
1968
+ "learning_rate": 4.5902692331755336e-05,
1969
+ "loss": 0.1516,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 3.42,
1974
+ "learning_rate": 4.5798419219620035e-05,
1975
+ "loss": 0.1448,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 3.43,
1980
+ "learning_rate": 4.5694146107484726e-05,
1981
+ "loss": 0.1514,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 3.44,
1986
+ "learning_rate": 4.558987299534942e-05,
1987
+ "loss": 0.1558,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 3.45,
1992
+ "learning_rate": 4.5485599883214116e-05,
1993
+ "loss": 0.1487,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 3.46,
1998
+ "learning_rate": 4.5381326771078815e-05,
1999
+ "loss": 0.1487,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 3.47,
2004
+ "learning_rate": 4.5277053658943507e-05,
2005
+ "loss": 0.1447,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 3.48,
2010
+ "learning_rate": 4.5172780546808205e-05,
2011
+ "loss": 0.1729,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 3.49,
2016
+ "learning_rate": 4.50685074346729e-05,
2017
+ "loss": 0.1425,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 3.5,
2022
+ "learning_rate": 4.496423432253759e-05,
2023
+ "loss": 0.1562,
2024
+ "step": 168000
2025
+ },
2026
+ {
2027
+ "epoch": 3.51,
2028
+ "learning_rate": 4.4859961210402294e-05,
2029
+ "loss": 0.1481,
2030
+ "step": 168500
2031
+ },
2032
+ {
2033
+ "epoch": 3.52,
2034
+ "learning_rate": 4.4755688098266985e-05,
2035
+ "loss": 0.1466,
2036
+ "step": 169000
2037
+ },
2038
+ {
2039
+ "epoch": 3.53,
2040
+ "learning_rate": 4.465141498613168e-05,
2041
+ "loss": 0.133,
2042
+ "step": 169500
2043
+ },
2044
+ {
2045
+ "epoch": 3.55,
2046
+ "learning_rate": 4.4547141873996376e-05,
2047
+ "loss": 0.1485,
2048
+ "step": 170000
2049
+ },
2050
+ {
2051
+ "epoch": 3.56,
2052
+ "learning_rate": 4.444286876186107e-05,
2053
+ "loss": 0.1332,
2054
+ "step": 170500
2055
+ },
2056
+ {
2057
+ "epoch": 3.57,
2058
+ "learning_rate": 4.433859564972576e-05,
2059
+ "loss": 0.1463,
2060
+ "step": 171000
2061
+ },
2062
+ {
2063
+ "epoch": 3.58,
2064
+ "learning_rate": 4.4234322537590464e-05,
2065
+ "loss": 0.1279,
2066
+ "step": 171500
2067
+ },
2068
+ {
2069
+ "epoch": 3.59,
2070
+ "learning_rate": 4.4130049425455156e-05,
2071
+ "loss": 0.1312,
2072
+ "step": 172000
2073
+ },
2074
+ {
2075
+ "epoch": 3.6,
2076
+ "learning_rate": 4.402577631331985e-05,
2077
+ "loss": 0.1387,
2078
+ "step": 172500
2079
+ },
2080
+ {
2081
+ "epoch": 3.61,
2082
+ "learning_rate": 4.3921503201184546e-05,
2083
+ "loss": 0.1305,
2084
+ "step": 173000
2085
+ },
2086
+ {
2087
+ "epoch": 3.62,
2088
+ "learning_rate": 4.381723008904924e-05,
2089
+ "loss": 0.1403,
2090
+ "step": 173500
2091
+ },
2092
+ {
2093
+ "epoch": 3.63,
2094
+ "learning_rate": 4.371295697691393e-05,
2095
+ "loss": 0.1462,
2096
+ "step": 174000
2097
+ },
2098
+ {
2099
+ "epoch": 3.64,
2100
+ "learning_rate": 4.3608683864778635e-05,
2101
+ "loss": 0.1404,
2102
+ "step": 174500
2103
+ },
2104
+ {
2105
+ "epoch": 3.65,
2106
+ "learning_rate": 4.3504410752643326e-05,
2107
+ "loss": 0.1384,
2108
+ "step": 175000
2109
+ },
2110
+ {
2111
+ "epoch": 3.66,
2112
+ "learning_rate": 4.340013764050802e-05,
2113
+ "loss": 0.1396,
2114
+ "step": 175500
2115
+ },
2116
+ {
2117
+ "epoch": 3.67,
2118
+ "learning_rate": 4.3295864528372717e-05,
2119
+ "loss": 0.1235,
2120
+ "step": 176000
2121
+ },
2122
+ {
2123
+ "epoch": 3.68,
2124
+ "learning_rate": 4.319159141623741e-05,
2125
+ "loss": 0.1337,
2126
+ "step": 176500
2127
+ },
2128
+ {
2129
+ "epoch": 3.69,
2130
+ "learning_rate": 4.308731830410211e-05,
2131
+ "loss": 0.1382,
2132
+ "step": 177000
2133
+ },
2134
+ {
2135
+ "epoch": 3.7,
2136
+ "learning_rate": 4.2983045191966805e-05,
2137
+ "loss": 0.1376,
2138
+ "step": 177500
2139
+ },
2140
+ {
2141
+ "epoch": 3.71,
2142
+ "learning_rate": 4.28787720798315e-05,
2143
+ "loss": 0.1458,
2144
+ "step": 178000
2145
+ },
2146
+ {
2147
+ "epoch": 3.72,
2148
+ "learning_rate": 4.277449896769619e-05,
2149
+ "loss": 0.1541,
2150
+ "step": 178500
2151
+ },
2152
+ {
2153
+ "epoch": 3.73,
2154
+ "learning_rate": 4.267022585556089e-05,
2155
+ "loss": 0.1327,
2156
+ "step": 179000
2157
+ },
2158
+ {
2159
+ "epoch": 3.74,
2160
+ "learning_rate": 4.256595274342558e-05,
2161
+ "loss": 0.1395,
2162
+ "step": 179500
2163
+ },
2164
+ {
2165
+ "epoch": 3.75,
2166
+ "learning_rate": 4.246167963129028e-05,
2167
+ "loss": 0.1272,
2168
+ "step": 180000
2169
+ },
2170
+ {
2171
+ "epoch": 3.76,
2172
+ "learning_rate": 4.2357406519154976e-05,
2173
+ "loss": 0.1276,
2174
+ "step": 180500
2175
+ },
2176
+ {
2177
+ "epoch": 3.77,
2178
+ "learning_rate": 4.225313340701967e-05,
2179
+ "loss": 0.1519,
2180
+ "step": 181000
2181
+ },
2182
+ {
2183
+ "epoch": 3.79,
2184
+ "learning_rate": 4.2148860294884366e-05,
2185
+ "loss": 0.1236,
2186
+ "step": 181500
2187
+ },
2188
+ {
2189
+ "epoch": 3.8,
2190
+ "learning_rate": 4.204458718274906e-05,
2191
+ "loss": 0.1173,
2192
+ "step": 182000
2193
+ },
2194
+ {
2195
+ "epoch": 3.81,
2196
+ "learning_rate": 4.194031407061375e-05,
2197
+ "loss": 0.144,
2198
+ "step": 182500
2199
+ },
2200
+ {
2201
+ "epoch": 3.82,
2202
+ "learning_rate": 4.183604095847845e-05,
2203
+ "loss": 0.1489,
2204
+ "step": 183000
2205
+ },
2206
+ {
2207
+ "epoch": 3.83,
2208
+ "learning_rate": 4.1731767846343146e-05,
2209
+ "loss": 0.1405,
2210
+ "step": 183500
2211
+ },
2212
+ {
2213
+ "epoch": 3.84,
2214
+ "learning_rate": 4.162749473420784e-05,
2215
+ "loss": 0.12,
2216
+ "step": 184000
2217
+ },
2218
+ {
2219
+ "epoch": 3.85,
2220
+ "learning_rate": 4.1523221622072536e-05,
2221
+ "loss": 0.1278,
2222
+ "step": 184500
2223
+ },
2224
+ {
2225
+ "epoch": 3.86,
2226
+ "learning_rate": 4.141894850993723e-05,
2227
+ "loss": 0.1298,
2228
+ "step": 185000
2229
+ },
2230
+ {
2231
+ "epoch": 3.87,
2232
+ "learning_rate": 4.1314675397801927e-05,
2233
+ "loss": 0.1269,
2234
+ "step": 185500
2235
+ },
2236
+ {
2237
+ "epoch": 3.88,
2238
+ "learning_rate": 4.1210402285666625e-05,
2239
+ "loss": 0.1278,
2240
+ "step": 186000
2241
+ },
2242
+ {
2243
+ "epoch": 3.89,
2244
+ "learning_rate": 4.110612917353132e-05,
2245
+ "loss": 0.129,
2246
+ "step": 186500
2247
+ },
2248
+ {
2249
+ "epoch": 3.9,
2250
+ "learning_rate": 4.100185606139601e-05,
2251
+ "loss": 0.1369,
2252
+ "step": 187000
2253
+ },
2254
+ {
2255
+ "epoch": 3.91,
2256
+ "learning_rate": 4.089758294926071e-05,
2257
+ "loss": 0.1337,
2258
+ "step": 187500
2259
+ },
2260
+ {
2261
+ "epoch": 3.92,
2262
+ "learning_rate": 4.07933098371254e-05,
2263
+ "loss": 0.1352,
2264
+ "step": 188000
2265
+ },
2266
+ {
2267
+ "epoch": 3.93,
2268
+ "learning_rate": 4.06890367249901e-05,
2269
+ "loss": 0.1133,
2270
+ "step": 188500
2271
+ },
2272
+ {
2273
+ "epoch": 3.94,
2274
+ "learning_rate": 4.0584763612854795e-05,
2275
+ "loss": 0.1498,
2276
+ "step": 189000
2277
+ },
2278
+ {
2279
+ "epoch": 3.95,
2280
+ "learning_rate": 4.048049050071949e-05,
2281
+ "loss": 0.1344,
2282
+ "step": 189500
2283
+ },
2284
+ {
2285
+ "epoch": 3.96,
2286
+ "learning_rate": 4.037621738858418e-05,
2287
+ "loss": 0.1467,
2288
+ "step": 190000
2289
+ },
2290
+ {
2291
+ "epoch": 3.97,
2292
+ "learning_rate": 4.027194427644888e-05,
2293
+ "loss": 0.1235,
2294
+ "step": 190500
2295
+ },
2296
+ {
2297
+ "epoch": 3.98,
2298
+ "learning_rate": 4.016767116431357e-05,
2299
+ "loss": 0.1333,
2300
+ "step": 191000
2301
+ },
2302
+ {
2303
+ "epoch": 3.99,
2304
+ "learning_rate": 4.006339805217827e-05,
2305
+ "loss": 0.1198,
2306
+ "step": 191500
2307
+ },
2308
+ {
2309
+ "epoch": 4.0,
2310
+ "learning_rate": 4.9959124940042965e-05,
2311
+ "loss": 0.0767,
2312
+ "step": 192000
2313
+ },
2314
+ {
2315
+ "epoch": 4.01,
2316
+ "learning_rate": 4.985485182790766e-05,
2317
+ "loss": 0.0963,
2318
+ "step": 192500
2319
+ },
2320
+ {
2321
+ "epoch": 4.02,
2322
+ "learning_rate": 4.9750578715772355e-05,
2323
+ "loss": 0.1067,
2324
+ "step": 193000
2325
+ },
2326
+ {
2327
+ "epoch": 4.04,
2328
+ "learning_rate": 4.964630560363705e-05,
2329
+ "loss": 0.1075,
2330
+ "step": 193500
2331
+ },
2332
+ {
2333
+ "epoch": 4.05,
2334
+ "learning_rate": 4.9542032491501745e-05,
2335
+ "loss": 0.1101,
2336
+ "step": 194000
2337
+ },
2338
+ {
2339
+ "epoch": 4.06,
2340
+ "learning_rate": 4.943775937936644e-05,
2341
+ "loss": 0.1055,
2342
+ "step": 194500
2343
+ },
2344
+ {
2345
+ "epoch": 4.07,
2346
+ "learning_rate": 4.9333486267231136e-05,
2347
+ "loss": 0.103,
2348
+ "step": 195000
2349
+ },
2350
+ {
2351
+ "epoch": 4.08,
2352
+ "learning_rate": 4.922921315509583e-05,
2353
+ "loss": 0.114,
2354
+ "step": 195500
2355
+ },
2356
+ {
2357
+ "epoch": 4.09,
2358
+ "learning_rate": 4.9124940042960526e-05,
2359
+ "loss": 0.1182,
2360
+ "step": 196000
2361
+ },
2362
+ {
2363
+ "epoch": 4.1,
2364
+ "learning_rate": 4.9020666930825224e-05,
2365
+ "loss": 0.1224,
2366
+ "step": 196500
2367
+ },
2368
+ {
2369
+ "epoch": 4.11,
2370
+ "learning_rate": 4.8916393818689916e-05,
2371
+ "loss": 0.1168,
2372
+ "step": 197000
2373
+ },
2374
+ {
2375
+ "epoch": 4.12,
2376
+ "learning_rate": 4.881212070655461e-05,
2377
+ "loss": 0.1395,
2378
+ "step": 197500
2379
+ },
2380
+ {
2381
+ "epoch": 4.13,
2382
+ "learning_rate": 4.8707847594419306e-05,
2383
+ "loss": 0.1203,
2384
+ "step": 198000
2385
+ },
2386
+ {
2387
+ "epoch": 4.14,
2388
+ "learning_rate": 4.8603574482284005e-05,
2389
+ "loss": 0.1176,
2390
+ "step": 198500
2391
+ },
2392
+ {
2393
+ "epoch": 4.15,
2394
+ "learning_rate": 4.8499301370148696e-05,
2395
+ "loss": 0.1254,
2396
+ "step": 199000
2397
+ },
2398
+ {
2399
+ "epoch": 4.16,
2400
+ "learning_rate": 4.8395028258013395e-05,
2401
+ "loss": 0.1148,
2402
+ "step": 199500
2403
+ },
2404
+ {
2405
+ "epoch": 4.17,
2406
+ "learning_rate": 4.8290755145878086e-05,
2407
+ "loss": 0.1202,
2408
+ "step": 200000
2409
+ },
2410
+ {
2411
+ "epoch": 4.18,
2412
+ "learning_rate": 4.818648203374278e-05,
2413
+ "loss": 0.1228,
2414
+ "step": 200500
2415
+ },
2416
+ {
2417
+ "epoch": 4.19,
2418
+ "learning_rate": 4.808220892160748e-05,
2419
+ "loss": 0.1249,
2420
+ "step": 201000
2421
+ },
2422
+ {
2423
+ "epoch": 4.2,
2424
+ "learning_rate": 4.7977935809472175e-05,
2425
+ "loss": 0.1413,
2426
+ "step": 201500
2427
+ },
2428
+ {
2429
+ "epoch": 4.21,
2430
+ "learning_rate": 4.787366269733687e-05,
2431
+ "loss": 0.151,
2432
+ "step": 202000
2433
+ },
2434
+ {
2435
+ "epoch": 4.22,
2436
+ "learning_rate": 4.7769389585201565e-05,
2437
+ "loss": 0.1217,
2438
+ "step": 202500
2439
+ },
2440
+ {
2441
+ "epoch": 4.23,
2442
+ "learning_rate": 4.766511647306626e-05,
2443
+ "loss": 0.1251,
2444
+ "step": 203000
2445
+ },
2446
+ {
2447
+ "epoch": 4.24,
2448
+ "learning_rate": 4.756084336093095e-05,
2449
+ "loss": 0.1299,
2450
+ "step": 203500
2451
+ },
2452
+ {
2453
+ "epoch": 4.25,
2454
+ "learning_rate": 4.745657024879565e-05,
2455
+ "loss": 0.1337,
2456
+ "step": 204000
2457
+ },
2458
+ {
2459
+ "epoch": 4.26,
2460
+ "learning_rate": 4.7352297136660346e-05,
2461
+ "loss": 0.1249,
2462
+ "step": 204500
2463
+ },
2464
+ {
2465
+ "epoch": 4.28,
2466
+ "learning_rate": 4.724802402452504e-05,
2467
+ "loss": 0.1273,
2468
+ "step": 205000
2469
+ },
2470
+ {
2471
+ "epoch": 4.29,
2472
+ "learning_rate": 4.7143750912389736e-05,
2473
+ "loss": 0.123,
2474
+ "step": 205500
2475
+ },
2476
+ {
2477
+ "epoch": 4.3,
2478
+ "learning_rate": 4.703947780025443e-05,
2479
+ "loss": 0.1142,
2480
+ "step": 206000
2481
+ },
2482
+ {
2483
+ "epoch": 4.31,
2484
+ "learning_rate": 4.693520468811912e-05,
2485
+ "loss": 0.1485,
2486
+ "step": 206500
2487
+ },
2488
+ {
2489
+ "epoch": 4.32,
2490
+ "learning_rate": 4.683093157598382e-05,
2491
+ "loss": 0.1247,
2492
+ "step": 207000
2493
+ },
2494
+ {
2495
+ "epoch": 4.33,
2496
+ "learning_rate": 4.6726658463848516e-05,
2497
+ "loss": 0.1238,
2498
+ "step": 207500
2499
+ },
2500
+ {
2501
+ "epoch": 4.34,
2502
+ "learning_rate": 4.662238535171321e-05,
2503
+ "loss": 0.1242,
2504
+ "step": 208000
2505
+ },
2506
+ {
2507
+ "epoch": 4.35,
2508
+ "learning_rate": 4.6518112239577906e-05,
2509
+ "loss": 0.1253,
2510
+ "step": 208500
2511
+ },
2512
+ {
2513
+ "epoch": 4.36,
2514
+ "learning_rate": 4.64138391274426e-05,
2515
+ "loss": 0.1325,
2516
+ "step": 209000
2517
+ },
2518
+ {
2519
+ "epoch": 4.37,
2520
+ "learning_rate": 4.630956601530729e-05,
2521
+ "loss": 0.1258,
2522
+ "step": 209500
2523
+ },
2524
+ {
2525
+ "epoch": 4.38,
2526
+ "learning_rate": 4.6205292903171995e-05,
2527
+ "loss": 0.1184,
2528
+ "step": 210000
2529
+ },
2530
+ {
2531
+ "epoch": 4.39,
2532
+ "learning_rate": 4.6101019791036687e-05,
2533
+ "loss": 0.1195,
2534
+ "step": 210500
2535
+ },
2536
+ {
2537
+ "epoch": 4.4,
2538
+ "learning_rate": 4.599674667890138e-05,
2539
+ "loss": 0.1382,
2540
+ "step": 211000
2541
+ },
2542
+ {
2543
+ "epoch": 4.41,
2544
+ "learning_rate": 4.589247356676608e-05,
2545
+ "loss": 0.138,
2546
+ "step": 211500
2547
+ },
2548
+ {
2549
+ "epoch": 4.42,
2550
+ "learning_rate": 4.578820045463077e-05,
2551
+ "loss": 0.1297,
2552
+ "step": 212000
2553
+ },
2554
+ {
2555
+ "epoch": 4.43,
2556
+ "learning_rate": 4.568392734249547e-05,
2557
+ "loss": 0.1161,
2558
+ "step": 212500
2559
+ },
2560
+ {
2561
+ "epoch": 4.44,
2562
+ "learning_rate": 4.5579654230360165e-05,
2563
+ "loss": 0.1273,
2564
+ "step": 213000
2565
+ },
2566
+ {
2567
+ "epoch": 4.45,
2568
+ "learning_rate": 4.547538111822486e-05,
2569
+ "loss": 0.1513,
2570
+ "step": 213500
2571
+ },
2572
+ {
2573
+ "epoch": 4.46,
2574
+ "learning_rate": 4.537110800608955e-05,
2575
+ "loss": 0.1043,
2576
+ "step": 214000
2577
+ },
2578
+ {
2579
+ "epoch": 4.47,
2580
+ "learning_rate": 4.526683489395425e-05,
2581
+ "loss": 0.1333,
2582
+ "step": 214500
2583
+ },
2584
+ {
2585
+ "epoch": 4.48,
2586
+ "learning_rate": 4.516256178181894e-05,
2587
+ "loss": 0.1228,
2588
+ "step": 215000
2589
+ },
2590
+ {
2591
+ "epoch": 4.49,
2592
+ "learning_rate": 4.505828866968364e-05,
2593
+ "loss": 0.1255,
2594
+ "step": 215500
2595
+ },
2596
+ {
2597
+ "epoch": 4.5,
2598
+ "learning_rate": 4.4954015557548336e-05,
2599
+ "loss": 0.1047,
2600
+ "step": 216000
2601
+ },
2602
+ {
2603
+ "epoch": 4.52,
2604
+ "learning_rate": 4.484974244541303e-05,
2605
+ "loss": 0.1412,
2606
+ "step": 216500
2607
+ },
2608
+ {
2609
+ "epoch": 4.53,
2610
+ "learning_rate": 4.4745469333277726e-05,
2611
+ "loss": 0.1399,
2612
+ "step": 217000
2613
+ },
2614
+ {
2615
+ "epoch": 4.54,
2616
+ "learning_rate": 4.464119622114242e-05,
2617
+ "loss": 0.12,
2618
+ "step": 217500
2619
+ },
2620
+ {
2621
+ "epoch": 4.55,
2622
+ "learning_rate": 4.453692310900711e-05,
2623
+ "loss": 0.1273,
2624
+ "step": 218000
2625
+ },
2626
+ {
2627
+ "epoch": 4.56,
2628
+ "learning_rate": 4.443264999687181e-05,
2629
+ "loss": 0.1154,
2630
+ "step": 218500
2631
+ },
2632
+ {
2633
+ "epoch": 4.57,
2634
+ "learning_rate": 4.4328376884736506e-05,
2635
+ "loss": 0.1364,
2636
+ "step": 219000
2637
+ },
2638
+ {
2639
+ "epoch": 4.58,
2640
+ "learning_rate": 4.42241037726012e-05,
2641
+ "loss": 0.1231,
2642
+ "step": 219500
2643
+ },
2644
+ {
2645
+ "epoch": 4.59,
2646
+ "learning_rate": 4.4119830660465897e-05,
2647
+ "loss": 0.1127,
2648
+ "step": 220000
2649
+ },
2650
+ {
2651
+ "epoch": 4.6,
2652
+ "learning_rate": 4.401555754833059e-05,
2653
+ "loss": 0.1183,
2654
+ "step": 220500
2655
+ },
2656
+ {
2657
+ "epoch": 4.61,
2658
+ "learning_rate": 4.391128443619528e-05,
2659
+ "loss": 0.1202,
2660
+ "step": 221000
2661
+ },
2662
+ {
2663
+ "epoch": 4.62,
2664
+ "learning_rate": 4.3807011324059985e-05,
2665
+ "loss": 0.1313,
2666
+ "step": 221500
2667
+ },
2668
+ {
2669
+ "epoch": 4.63,
2670
+ "learning_rate": 4.370273821192468e-05,
2671
+ "loss": 0.1213,
2672
+ "step": 222000
2673
+ },
2674
+ {
2675
+ "epoch": 4.64,
2676
+ "learning_rate": 4.359846509978937e-05,
2677
+ "loss": 0.1199,
2678
+ "step": 222500
2679
+ },
2680
+ {
2681
+ "epoch": 4.65,
2682
+ "learning_rate": 4.349419198765407e-05,
2683
+ "loss": 0.1051,
2684
+ "step": 223000
2685
+ },
2686
+ {
2687
+ "epoch": 4.66,
2688
+ "learning_rate": 4.338991887551876e-05,
2689
+ "loss": 0.1251,
2690
+ "step": 223500
2691
+ },
2692
+ {
2693
+ "epoch": 4.67,
2694
+ "learning_rate": 4.328564576338345e-05,
2695
+ "loss": 0.1292,
2696
+ "step": 224000
2697
+ },
2698
+ {
2699
+ "epoch": 4.68,
2700
+ "learning_rate": 4.3181372651248156e-05,
2701
+ "loss": 0.1228,
2702
+ "step": 224500
2703
+ },
2704
+ {
2705
+ "epoch": 4.69,
2706
+ "learning_rate": 4.307709953911285e-05,
2707
+ "loss": 0.1158,
2708
+ "step": 225000
2709
+ },
2710
+ {
2711
+ "epoch": 4.7,
2712
+ "learning_rate": 4.297282642697754e-05,
2713
+ "loss": 0.1084,
2714
+ "step": 225500
2715
+ },
2716
+ {
2717
+ "epoch": 4.71,
2718
+ "learning_rate": 4.286855331484224e-05,
2719
+ "loss": 0.1183,
2720
+ "step": 226000
2721
+ },
2722
+ {
2723
+ "epoch": 4.72,
2724
+ "learning_rate": 4.276428020270693e-05,
2725
+ "loss": 0.1103,
2726
+ "step": 226500
2727
+ },
2728
+ {
2729
+ "epoch": 4.73,
2730
+ "learning_rate": 4.266000709057163e-05,
2731
+ "loss": 0.1037,
2732
+ "step": 227000
2733
+ },
2734
+ {
2735
+ "epoch": 4.74,
2736
+ "learning_rate": 4.2555733978436326e-05,
2737
+ "loss": 0.1093,
2738
+ "step": 227500
2739
+ },
2740
+ {
2741
+ "epoch": 4.75,
2742
+ "learning_rate": 4.245146086630102e-05,
2743
+ "loss": 0.1005,
2744
+ "step": 228000
2745
+ },
2746
+ {
2747
+ "epoch": 4.77,
2748
+ "learning_rate": 4.234718775416571e-05,
2749
+ "loss": 0.1307,
2750
+ "step": 228500
2751
+ },
2752
+ {
2753
+ "epoch": 4.78,
2754
+ "learning_rate": 4.224291464203041e-05,
2755
+ "loss": 0.125,
2756
+ "step": 229000
2757
+ },
2758
+ {
2759
+ "epoch": 4.79,
2760
+ "learning_rate": 4.21386415298951e-05,
2761
+ "loss": 0.113,
2762
+ "step": 229500
2763
+ },
2764
+ {
2765
+ "epoch": 4.8,
2766
+ "learning_rate": 4.20343684177598e-05,
2767
+ "loss": 0.1093,
2768
+ "step": 230000
2769
+ },
2770
+ {
2771
+ "epoch": 4.81,
2772
+ "learning_rate": 4.19300953056245e-05,
2773
+ "loss": 0.1135,
2774
+ "step": 230500
2775
+ },
2776
+ {
2777
+ "epoch": 4.82,
2778
+ "learning_rate": 4.182582219348919e-05,
2779
+ "loss": 0.1086,
2780
+ "step": 231000
2781
+ },
2782
+ {
2783
+ "epoch": 4.83,
2784
+ "learning_rate": 4.172154908135388e-05,
2785
+ "loss": 0.1282,
2786
+ "step": 231500
2787
+ },
2788
+ {
2789
+ "epoch": 4.84,
2790
+ "learning_rate": 4.161727596921858e-05,
2791
+ "loss": 0.0988,
2792
+ "step": 232000
2793
+ },
2794
+ {
2795
+ "epoch": 4.85,
2796
+ "learning_rate": 4.151300285708327e-05,
2797
+ "loss": 0.1118,
2798
+ "step": 232500
2799
+ },
2800
+ {
2801
+ "epoch": 4.86,
2802
+ "learning_rate": 4.140872974494797e-05,
2803
+ "loss": 0.1233,
2804
+ "step": 233000
2805
+ },
2806
+ {
2807
+ "epoch": 4.87,
2808
+ "learning_rate": 4.130445663281267e-05,
2809
+ "loss": 0.1327,
2810
+ "step": 233500
2811
+ },
2812
+ {
2813
+ "epoch": 4.88,
2814
+ "learning_rate": 4.120018352067736e-05,
2815
+ "loss": 0.1061,
2816
+ "step": 234000
2817
+ },
2818
+ {
2819
+ "epoch": 4.89,
2820
+ "learning_rate": 4.109591040854206e-05,
2821
+ "loss": 0.1085,
2822
+ "step": 234500
2823
+ },
2824
+ {
2825
+ "epoch": 4.9,
2826
+ "learning_rate": 4.099163729640675e-05,
2827
+ "loss": 0.0944,
2828
+ "step": 235000
2829
+ },
2830
+ {
2831
+ "epoch": 4.91,
2832
+ "learning_rate": 4.088736418427145e-05,
2833
+ "loss": 0.1009,
2834
+ "step": 235500
2835
+ },
2836
+ {
2837
+ "epoch": 4.92,
2838
+ "learning_rate": 4.078309107213614e-05,
2839
+ "loss": 0.1104,
2840
+ "step": 236000
2841
+ },
2842
+ {
2843
+ "epoch": 4.93,
2844
+ "learning_rate": 4.067881796000084e-05,
2845
+ "loss": 0.1201,
2846
+ "step": 236500
2847
+ },
2848
+ {
2849
+ "epoch": 4.94,
2850
+ "learning_rate": 4.057454484786553e-05,
2851
+ "loss": 0.1067,
2852
+ "step": 237000
2853
+ },
2854
+ {
2855
+ "epoch": 4.95,
2856
+ "learning_rate": 4.047027173573023e-05,
2857
+ "loss": 0.117,
2858
+ "step": 237500
2859
+ },
2860
+ {
2861
+ "epoch": 4.96,
2862
+ "learning_rate": 4.036599862359492e-05,
2863
+ "loss": 0.1221,
2864
+ "step": 238000
2865
+ },
2866
+ {
2867
+ "epoch": 4.97,
2868
+ "learning_rate": 4.026172551145962e-05,
2869
+ "loss": 0.1082,
2870
+ "step": 238500
2871
+ },
2872
+ {
2873
+ "epoch": 4.98,
2874
+ "learning_rate": 4.0157452399324317e-05,
2875
+ "loss": 0.1087,
2876
+ "step": 239000
2877
+ },
2878
+ {
2879
+ "epoch": 4.99,
2880
+ "learning_rate": 4.005317928718901e-05,
2881
+ "loss": 0.1024,
2882
+ "step": 239500
2883
+ },
2884
+ {
2885
+ "epoch": 5.0,
2886
+ "step": 239755,
2887
+ "total_flos": 3.624187009725235e+16,
2888
+ "train_loss": 0.02392078927181249,
2889
+ "train_runtime": 8161.1668,
2890
+ "train_samples_per_second": 117.509,
2891
+ "train_steps_per_second": 29.378
2892
+ }
2893
+ ],
2894
+ "max_steps": 239755,
2895
+ "num_train_epochs": 5,
2896
+ "total_flos": 3.624187009725235e+16,
2897
+ "trial_name": null,
2898
+ "trial_params": null
2899
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cbba0fade586e575f79d7e64aa0f2c1946cca763cddb931c1c565c6449f4039
3
+ size 4207
vocab.json ADDED
The diff for this file is too large to render. See raw diff