Davlan commited on
Commit
bb96aae
1 Parent(s): 3a255ad

add MT model

Browse files
all_results.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_bleu": 7.2053,
4
+ "eval_gen_len": 52.5325,
5
+ "eval_loss": 2.363677978515625,
6
+ "eval_runtime": 639.4154,
7
+ "eval_samples": 1414,
8
+ "eval_samples_per_second": 2.211,
9
+ "eval_steps_per_second": 0.222,
10
+ "predict_bleu": 7.3259,
11
+ "predict_gen_len": 52.499,
12
+ "predict_loss": 2.3529598712921143,
13
+ "predict_runtime": 673.7839,
14
+ "predict_samples": 1563,
15
+ "predict_samples_per_second": 2.32,
16
+ "predict_steps_per_second": 0.233,
17
+ "train_loss": 1.0171057777132477,
18
+ "train_runtime": 64358.9209,
19
+ "train_samples": 620585,
20
+ "train_samples_per_second": 28.928,
21
+ "train_steps_per_second": 2.893
22
+ }
config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/m2m100_418M",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "relu",
5
+ "architectures": [
6
+ "M2M100ForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 0,
10
+ "d_model": 1024,
11
+ "decoder_attention_heads": 16,
12
+ "decoder_ffn_dim": 4096,
13
+ "decoder_layerdrop": 0.05,
14
+ "decoder_layers": 12,
15
+ "decoder_start_token_id": 2,
16
+ "dropout": 0.1,
17
+ "early_stopping": true,
18
+ "encoder_attention_heads": 16,
19
+ "encoder_ffn_dim": 4096,
20
+ "encoder_layerdrop": 0.05,
21
+ "encoder_layers": 12,
22
+ "eos_token_id": 2,
23
+ "forced_bos_token_id": 128088,
24
+ "gradient_checkpointing": false,
25
+ "init_std": 0.02,
26
+ "is_encoder_decoder": true,
27
+ "max_length": 200,
28
+ "max_position_embeddings": 1024,
29
+ "model_type": "m2m_100",
30
+ "num_beams": 5,
31
+ "num_hidden_layers": 12,
32
+ "pad_token_id": 1,
33
+ "scale_embedding": true,
34
+ "torch_dtype": "float32",
35
+ "transformers_version": "4.10.3",
36
+ "use_cache": true,
37
+ "vocab_size": 128112
38
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_bleu": 7.2053,
4
+ "eval_gen_len": 52.5325,
5
+ "eval_loss": 2.363677978515625,
6
+ "eval_runtime": 639.4154,
7
+ "eval_samples": 1414,
8
+ "eval_samples_per_second": 2.211,
9
+ "eval_steps_per_second": 0.222
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": 7.3259,
3
+ "predict_gen_len": 52.499,
4
+ "predict_loss": 2.3529598712921143,
5
+ "predict_runtime": 673.7839,
6
+ "predict_samples": 1563,
7
+ "predict_samples_per_second": 2.32,
8
+ "predict_steps_per_second": 0.233
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7b49c7e664b2e645194893bc51a45c92c74a92600a9f09ee4b1d34ad1ee77a5
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": 1.0171057777132477,
4
+ "train_runtime": 64358.9209,
5
+ "train_samples": 620585,
6
+ "train_samples_per_second": 28.928,
7
+ "train_steps_per_second": 2.893
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2257 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 186177,
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": 4.986571918120928e-05,
13
+ "loss": 3.0976,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.973143836241856e-05,
19
+ "loss": 2.3443,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 4.959715754362784e-05,
25
+ "loss": 2.1104,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.03,
30
+ "learning_rate": 4.946287672483712e-05,
31
+ "loss": 1.9899,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.04,
36
+ "learning_rate": 4.93285959060464e-05,
37
+ "loss": 1.8868,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.05,
42
+ "learning_rate": 4.9194315087255675e-05,
43
+ "loss": 1.813,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.06,
48
+ "learning_rate": 4.906003426846495e-05,
49
+ "loss": 1.7437,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.06,
54
+ "learning_rate": 4.8925753449674237e-05,
55
+ "loss": 1.6987,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.07,
60
+ "learning_rate": 4.8791472630883514e-05,
61
+ "loss": 1.6737,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.08,
66
+ "learning_rate": 4.865719181209279e-05,
67
+ "loss": 1.6303,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.09,
72
+ "learning_rate": 4.8522910993302076e-05,
73
+ "loss": 1.6134,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.1,
78
+ "learning_rate": 4.8388630174511354e-05,
79
+ "loss": 1.5786,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.1,
84
+ "learning_rate": 4.825434935572064e-05,
85
+ "loss": 1.5773,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.11,
90
+ "learning_rate": 4.8120068536929916e-05,
91
+ "loss": 1.5281,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.12,
96
+ "learning_rate": 4.798578771813919e-05,
97
+ "loss": 1.5081,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.13,
102
+ "learning_rate": 4.785150689934847e-05,
103
+ "loss": 1.4943,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.14,
108
+ "learning_rate": 4.7717226080557755e-05,
109
+ "loss": 1.488,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.15,
114
+ "learning_rate": 4.758294526176703e-05,
115
+ "loss": 1.4594,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.15,
120
+ "learning_rate": 4.744866444297631e-05,
121
+ "loss": 1.4463,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.16,
126
+ "learning_rate": 4.731438362418559e-05,
127
+ "loss": 1.4442,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.17,
132
+ "learning_rate": 4.7180102805394865e-05,
133
+ "loss": 1.434,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.18,
138
+ "learning_rate": 4.704582198660415e-05,
139
+ "loss": 1.4108,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.19,
144
+ "learning_rate": 4.691154116781343e-05,
145
+ "loss": 1.4195,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.19,
150
+ "learning_rate": 4.6777260349022705e-05,
151
+ "loss": 1.3936,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.2,
156
+ "learning_rate": 4.664297953023198e-05,
157
+ "loss": 1.3723,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.21,
162
+ "learning_rate": 4.650869871144127e-05,
163
+ "loss": 1.3742,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.22,
168
+ "learning_rate": 4.6374417892650544e-05,
169
+ "loss": 1.3785,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.23,
174
+ "learning_rate": 4.624013707385982e-05,
175
+ "loss": 1.346,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.23,
180
+ "learning_rate": 4.61058562550691e-05,
181
+ "loss": 1.3453,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.24,
186
+ "learning_rate": 4.597157543627838e-05,
187
+ "loss": 1.3503,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.25,
192
+ "learning_rate": 4.583729461748766e-05,
193
+ "loss": 1.3297,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.26,
198
+ "learning_rate": 4.570301379869694e-05,
199
+ "loss": 1.3313,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.27,
204
+ "learning_rate": 4.5568732979906224e-05,
205
+ "loss": 1.3174,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.27,
210
+ "learning_rate": 4.54344521611155e-05,
211
+ "loss": 1.307,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.28,
216
+ "learning_rate": 4.530017134232478e-05,
217
+ "loss": 1.3019,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.29,
222
+ "learning_rate": 4.516589052353406e-05,
223
+ "loss": 1.3,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.3,
228
+ "learning_rate": 4.503160970474334e-05,
229
+ "loss": 1.3041,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.31,
234
+ "learning_rate": 4.489732888595262e-05,
235
+ "loss": 1.2835,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.31,
240
+ "learning_rate": 4.4763048067161896e-05,
241
+ "loss": 1.2905,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.32,
246
+ "learning_rate": 4.462876724837118e-05,
247
+ "loss": 1.2791,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.33,
252
+ "learning_rate": 4.449448642958046e-05,
253
+ "loss": 1.2863,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.34,
258
+ "learning_rate": 4.4360205610789735e-05,
259
+ "loss": 1.2746,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.35,
264
+ "learning_rate": 4.422592479199901e-05,
265
+ "loss": 1.2679,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.35,
270
+ "learning_rate": 4.409164397320829e-05,
271
+ "loss": 1.2545,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.36,
276
+ "learning_rate": 4.3957363154417575e-05,
277
+ "loss": 1.2722,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.37,
282
+ "learning_rate": 4.382308233562685e-05,
283
+ "loss": 1.2607,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.38,
288
+ "learning_rate": 4.368880151683613e-05,
289
+ "loss": 1.2486,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.39,
294
+ "learning_rate": 4.355452069804541e-05,
295
+ "loss": 1.2391,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.39,
300
+ "learning_rate": 4.3420239879254685e-05,
301
+ "loss": 1.2433,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.4,
306
+ "learning_rate": 4.328595906046397e-05,
307
+ "loss": 1.2383,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.41,
312
+ "learning_rate": 4.315167824167325e-05,
313
+ "loss": 1.2562,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.42,
318
+ "learning_rate": 4.3017397422882525e-05,
319
+ "loss": 1.2276,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.43,
324
+ "learning_rate": 4.28831166040918e-05,
325
+ "loss": 1.2264,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.44,
330
+ "learning_rate": 4.2748835785301087e-05,
331
+ "loss": 1.204,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.44,
336
+ "learning_rate": 4.2614554966510364e-05,
337
+ "loss": 1.2207,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.45,
342
+ "learning_rate": 4.248027414771965e-05,
343
+ "loss": 1.2016,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.46,
348
+ "learning_rate": 4.2345993328928926e-05,
349
+ "loss": 1.2169,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.47,
354
+ "learning_rate": 4.2211712510138204e-05,
355
+ "loss": 1.2196,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.48,
360
+ "learning_rate": 4.207743169134749e-05,
361
+ "loss": 1.2089,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.48,
366
+ "learning_rate": 4.1943150872556766e-05,
367
+ "loss": 1.2052,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.49,
372
+ "learning_rate": 4.180887005376604e-05,
373
+ "loss": 1.2013,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.5,
378
+ "learning_rate": 4.167458923497532e-05,
379
+ "loss": 1.2111,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.51,
384
+ "learning_rate": 4.15403084161846e-05,
385
+ "loss": 1.1811,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.52,
390
+ "learning_rate": 4.140602759739388e-05,
391
+ "loss": 1.1944,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.52,
396
+ "learning_rate": 4.127174677860316e-05,
397
+ "loss": 1.193,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.53,
402
+ "learning_rate": 4.113746595981244e-05,
403
+ "loss": 1.1803,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.54,
408
+ "learning_rate": 4.1003185141021715e-05,
409
+ "loss": 1.1855,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.55,
414
+ "learning_rate": 4.0868904322231e-05,
415
+ "loss": 1.1893,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.56,
420
+ "learning_rate": 4.073462350344028e-05,
421
+ "loss": 1.1678,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.56,
426
+ "learning_rate": 4.0600342684649555e-05,
427
+ "loss": 1.1765,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.57,
432
+ "learning_rate": 4.046606186585883e-05,
433
+ "loss": 1.1572,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.58,
438
+ "learning_rate": 4.033178104706811e-05,
439
+ "loss": 1.1713,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.59,
444
+ "learning_rate": 4.0197500228277394e-05,
445
+ "loss": 1.1736,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.6,
450
+ "learning_rate": 4.006321940948667e-05,
451
+ "loss": 1.1525,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.6,
456
+ "learning_rate": 3.992893859069595e-05,
457
+ "loss": 1.1498,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.61,
462
+ "learning_rate": 3.979465777190523e-05,
463
+ "loss": 1.1521,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.62,
468
+ "learning_rate": 3.966037695311451e-05,
469
+ "loss": 1.1467,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.63,
474
+ "learning_rate": 3.952609613432379e-05,
475
+ "loss": 1.1678,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.64,
480
+ "learning_rate": 3.9391815315533074e-05,
481
+ "loss": 1.145,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.64,
486
+ "learning_rate": 3.925753449674235e-05,
487
+ "loss": 1.1444,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.65,
492
+ "learning_rate": 3.912325367795163e-05,
493
+ "loss": 1.1433,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.66,
498
+ "learning_rate": 3.898897285916091e-05,
499
+ "loss": 1.13,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.67,
504
+ "learning_rate": 3.885469204037019e-05,
505
+ "loss": 1.133,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.68,
510
+ "learning_rate": 3.872041122157947e-05,
511
+ "loss": 1.1509,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.68,
516
+ "learning_rate": 3.8586130402788746e-05,
517
+ "loss": 1.1204,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.69,
522
+ "learning_rate": 3.845184958399802e-05,
523
+ "loss": 1.1411,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.7,
528
+ "learning_rate": 3.831756876520731e-05,
529
+ "loss": 1.1342,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.71,
534
+ "learning_rate": 3.8183287946416585e-05,
535
+ "loss": 1.1341,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.72,
540
+ "learning_rate": 3.804900712762586e-05,
541
+ "loss": 1.118,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.73,
546
+ "learning_rate": 3.791472630883514e-05,
547
+ "loss": 1.1068,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.73,
552
+ "learning_rate": 3.778044549004442e-05,
553
+ "loss": 1.1106,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.74,
558
+ "learning_rate": 3.76461646712537e-05,
559
+ "loss": 1.1376,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.75,
564
+ "learning_rate": 3.751188385246298e-05,
565
+ "loss": 1.1356,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.76,
570
+ "learning_rate": 3.737760303367226e-05,
571
+ "loss": 1.1204,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.77,
576
+ "learning_rate": 3.7243322214881535e-05,
577
+ "loss": 1.1179,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.77,
582
+ "learning_rate": 3.710904139609082e-05,
583
+ "loss": 1.0864,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.78,
588
+ "learning_rate": 3.69747605773001e-05,
589
+ "loss": 1.1158,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.79,
594
+ "learning_rate": 3.6840479758509375e-05,
595
+ "loss": 1.1112,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.8,
600
+ "learning_rate": 3.670619893971866e-05,
601
+ "loss": 1.1064,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.81,
606
+ "learning_rate": 3.6571918120927937e-05,
607
+ "loss": 1.107,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.81,
612
+ "learning_rate": 3.6437637302137214e-05,
613
+ "loss": 1.0997,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.82,
618
+ "learning_rate": 3.63033564833465e-05,
619
+ "loss": 1.0821,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.83,
624
+ "learning_rate": 3.6169075664555776e-05,
625
+ "loss": 1.1076,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.84,
630
+ "learning_rate": 3.6034794845765054e-05,
631
+ "loss": 1.0918,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.85,
636
+ "learning_rate": 3.590051402697433e-05,
637
+ "loss": 1.0955,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.85,
642
+ "learning_rate": 3.5766233208183616e-05,
643
+ "loss": 1.0878,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.86,
648
+ "learning_rate": 3.563195238939289e-05,
649
+ "loss": 1.0931,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.87,
654
+ "learning_rate": 3.549767157060217e-05,
655
+ "loss": 1.0888,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.88,
660
+ "learning_rate": 3.536339075181145e-05,
661
+ "loss": 1.0972,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.89,
666
+ "learning_rate": 3.522910993302073e-05,
667
+ "loss": 1.0944,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.89,
672
+ "learning_rate": 3.509482911423001e-05,
673
+ "loss": 1.088,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.9,
678
+ "learning_rate": 3.496054829543929e-05,
679
+ "loss": 1.0848,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.91,
684
+ "learning_rate": 3.4826267476648565e-05,
685
+ "loss": 1.0822,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.92,
690
+ "learning_rate": 3.469198665785784e-05,
691
+ "loss": 1.0767,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.93,
696
+ "learning_rate": 3.455770583906713e-05,
697
+ "loss": 1.0767,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.93,
702
+ "learning_rate": 3.4423425020276405e-05,
703
+ "loss": 1.08,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.94,
708
+ "learning_rate": 3.428914420148568e-05,
709
+ "loss": 1.0785,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.95,
714
+ "learning_rate": 3.415486338269496e-05,
715
+ "loss": 1.0734,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.96,
720
+ "learning_rate": 3.402058256390424e-05,
721
+ "loss": 1.0726,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.97,
726
+ "learning_rate": 3.388630174511352e-05,
727
+ "loss": 1.0752,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.97,
732
+ "learning_rate": 3.37520209263228e-05,
733
+ "loss": 1.0855,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.98,
738
+ "learning_rate": 3.3617740107532084e-05,
739
+ "loss": 1.0524,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.99,
744
+ "learning_rate": 3.348345928874136e-05,
745
+ "loss": 1.0454,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 1.0,
750
+ "learning_rate": 3.3349178469950646e-05,
751
+ "loss": 1.064,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 1.01,
756
+ "learning_rate": 3.3214897651159924e-05,
757
+ "loss": 1.0134,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 1.02,
762
+ "learning_rate": 3.30806168323692e-05,
763
+ "loss": 0.9915,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 1.02,
768
+ "learning_rate": 3.294633601357848e-05,
769
+ "loss": 0.992,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 1.03,
774
+ "learning_rate": 3.2812055194787756e-05,
775
+ "loss": 0.9785,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 1.04,
780
+ "learning_rate": 3.267777437599704e-05,
781
+ "loss": 0.9876,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 1.05,
786
+ "learning_rate": 3.254349355720632e-05,
787
+ "loss": 1.0002,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 1.06,
792
+ "learning_rate": 3.2409212738415596e-05,
793
+ "loss": 0.9987,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 1.06,
798
+ "learning_rate": 3.227493191962487e-05,
799
+ "loss": 0.9855,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 1.07,
804
+ "learning_rate": 3.214065110083415e-05,
805
+ "loss": 0.9993,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 1.08,
810
+ "learning_rate": 3.2006370282043435e-05,
811
+ "loss": 0.9843,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 1.09,
816
+ "learning_rate": 3.187208946325271e-05,
817
+ "loss": 0.9876,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 1.1,
822
+ "learning_rate": 3.173780864446199e-05,
823
+ "loss": 0.9772,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 1.1,
828
+ "learning_rate": 3.160352782567127e-05,
829
+ "loss": 0.9881,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 1.11,
834
+ "learning_rate": 3.146924700688055e-05,
835
+ "loss": 0.9769,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 1.12,
840
+ "learning_rate": 3.133496618808983e-05,
841
+ "loss": 1.0013,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 1.13,
846
+ "learning_rate": 3.120068536929911e-05,
847
+ "loss": 0.9821,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 1.14,
852
+ "learning_rate": 3.1066404550508385e-05,
853
+ "loss": 0.9781,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 1.14,
858
+ "learning_rate": 3.093212373171766e-05,
859
+ "loss": 0.9898,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 1.15,
864
+ "learning_rate": 3.079784291292695e-05,
865
+ "loss": 0.98,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 1.16,
870
+ "learning_rate": 3.0663562094136225e-05,
871
+ "loss": 0.984,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 1.17,
876
+ "learning_rate": 3.052928127534551e-05,
877
+ "loss": 0.9762,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 1.18,
882
+ "learning_rate": 3.0395000456554783e-05,
883
+ "loss": 0.9903,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 1.18,
888
+ "learning_rate": 3.026071963776406e-05,
889
+ "loss": 0.9871,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 1.19,
894
+ "learning_rate": 3.0126438818973345e-05,
895
+ "loss": 0.9855,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 1.2,
900
+ "learning_rate": 2.9992158000182623e-05,
901
+ "loss": 0.9887,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 1.21,
906
+ "learning_rate": 2.9857877181391904e-05,
907
+ "loss": 0.9652,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 1.22,
912
+ "learning_rate": 2.972359636260118e-05,
913
+ "loss": 0.9842,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 1.22,
918
+ "learning_rate": 2.9589315543810466e-05,
919
+ "loss": 0.9783,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 1.23,
924
+ "learning_rate": 2.9455034725019743e-05,
925
+ "loss": 0.9884,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 1.24,
930
+ "learning_rate": 2.932075390622902e-05,
931
+ "loss": 0.9659,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 1.25,
936
+ "learning_rate": 2.91864730874383e-05,
937
+ "loss": 0.9714,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 1.26,
942
+ "learning_rate": 2.9052192268647576e-05,
943
+ "loss": 0.9771,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 1.26,
948
+ "learning_rate": 2.891791144985686e-05,
949
+ "loss": 0.9608,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 1.27,
954
+ "learning_rate": 2.8783630631066138e-05,
955
+ "loss": 0.9529,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 1.28,
960
+ "learning_rate": 2.8649349812275415e-05,
961
+ "loss": 0.969,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 1.29,
966
+ "learning_rate": 2.8515068993484696e-05,
967
+ "loss": 0.9829,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 1.3,
972
+ "learning_rate": 2.8380788174693974e-05,
973
+ "loss": 0.973,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 1.31,
978
+ "learning_rate": 2.824650735590326e-05,
979
+ "loss": 0.9665,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 1.31,
984
+ "learning_rate": 2.8112226537112536e-05,
985
+ "loss": 0.9726,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 1.32,
990
+ "learning_rate": 2.7977945718321814e-05,
991
+ "loss": 0.9702,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 1.33,
996
+ "learning_rate": 2.784366489953109e-05,
997
+ "loss": 0.9725,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 1.34,
1002
+ "learning_rate": 2.770938408074037e-05,
1003
+ "loss": 0.9613,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 1.35,
1008
+ "learning_rate": 2.7575103261949653e-05,
1009
+ "loss": 0.9664,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 1.35,
1014
+ "learning_rate": 2.744082244315893e-05,
1015
+ "loss": 0.9563,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 1.36,
1020
+ "learning_rate": 2.7306541624368208e-05,
1021
+ "loss": 0.9591,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 1.37,
1026
+ "learning_rate": 2.7172260805577486e-05,
1027
+ "loss": 0.9638,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 1.38,
1032
+ "learning_rate": 2.703797998678677e-05,
1033
+ "loss": 0.9619,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 1.39,
1038
+ "learning_rate": 2.6903699167996048e-05,
1039
+ "loss": 0.9727,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 1.39,
1044
+ "learning_rate": 2.676941834920533e-05,
1045
+ "loss": 0.9677,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 1.4,
1050
+ "learning_rate": 2.6635137530414606e-05,
1051
+ "loss": 0.9619,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 1.41,
1056
+ "learning_rate": 2.6500856711623884e-05,
1057
+ "loss": 0.9803,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 1.42,
1062
+ "learning_rate": 2.6366575892833168e-05,
1063
+ "loss": 0.9497,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 1.43,
1068
+ "learning_rate": 2.6232295074042446e-05,
1069
+ "loss": 0.9605,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 1.43,
1074
+ "learning_rate": 2.6098014255251723e-05,
1075
+ "loss": 0.9565,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 1.44,
1080
+ "learning_rate": 2.5963733436461e-05,
1081
+ "loss": 0.9654,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 1.45,
1086
+ "learning_rate": 2.582945261767028e-05,
1087
+ "loss": 0.9638,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 1.46,
1092
+ "learning_rate": 2.5695171798879563e-05,
1093
+ "loss": 0.9748,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 1.47,
1098
+ "learning_rate": 2.556089098008884e-05,
1099
+ "loss": 0.9542,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 1.47,
1104
+ "learning_rate": 2.542661016129812e-05,
1105
+ "loss": 0.9514,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 1.48,
1110
+ "learning_rate": 2.52923293425074e-05,
1111
+ "loss": 0.9713,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 1.49,
1116
+ "learning_rate": 2.5158048523716683e-05,
1117
+ "loss": 0.9559,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 1.5,
1122
+ "learning_rate": 2.502376770492596e-05,
1123
+ "loss": 0.9683,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 1.51,
1128
+ "learning_rate": 2.488948688613524e-05,
1129
+ "loss": 0.9669,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 1.51,
1134
+ "learning_rate": 2.4755206067344516e-05,
1135
+ "loss": 0.9465,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 1.52,
1140
+ "learning_rate": 2.4620925248553797e-05,
1141
+ "loss": 0.9582,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 1.53,
1146
+ "learning_rate": 2.4486644429763075e-05,
1147
+ "loss": 0.9459,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 1.54,
1152
+ "learning_rate": 2.4352363610972356e-05,
1153
+ "loss": 0.9454,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 1.55,
1158
+ "learning_rate": 2.4218082792181633e-05,
1159
+ "loss": 0.9545,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 1.55,
1164
+ "learning_rate": 2.4083801973390914e-05,
1165
+ "loss": 0.9517,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 1.56,
1170
+ "learning_rate": 2.3949521154600195e-05,
1171
+ "loss": 0.9605,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 1.57,
1176
+ "learning_rate": 2.3815240335809473e-05,
1177
+ "loss": 0.9642,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 1.58,
1182
+ "learning_rate": 2.3680959517018754e-05,
1183
+ "loss": 0.9507,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 1.59,
1188
+ "learning_rate": 2.354667869822803e-05,
1189
+ "loss": 0.9531,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 1.6,
1194
+ "learning_rate": 2.3412397879437312e-05,
1195
+ "loss": 0.9504,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 1.6,
1200
+ "learning_rate": 2.327811706064659e-05,
1201
+ "loss": 0.958,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 1.61,
1206
+ "learning_rate": 2.314383624185587e-05,
1207
+ "loss": 0.9257,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 1.62,
1212
+ "learning_rate": 2.300955542306515e-05,
1213
+ "loss": 0.9429,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 1.63,
1218
+ "learning_rate": 2.2875274604274426e-05,
1219
+ "loss": 0.947,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 1.64,
1224
+ "learning_rate": 2.2740993785483707e-05,
1225
+ "loss": 0.9439,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 1.64,
1230
+ "learning_rate": 2.2606712966692984e-05,
1231
+ "loss": 0.9408,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 1.65,
1236
+ "learning_rate": 2.2472432147902265e-05,
1237
+ "loss": 0.9349,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 1.66,
1242
+ "learning_rate": 2.2338151329111546e-05,
1243
+ "loss": 0.9384,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 1.67,
1248
+ "learning_rate": 2.2203870510320827e-05,
1249
+ "loss": 0.9231,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 1.68,
1254
+ "learning_rate": 2.2069589691530105e-05,
1255
+ "loss": 0.9487,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 1.68,
1260
+ "learning_rate": 2.1935308872739383e-05,
1261
+ "loss": 0.9394,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 1.69,
1266
+ "learning_rate": 2.1801028053948664e-05,
1267
+ "loss": 0.9295,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 1.7,
1272
+ "learning_rate": 2.166674723515794e-05,
1273
+ "loss": 0.9449,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 1.71,
1278
+ "learning_rate": 2.1532466416367222e-05,
1279
+ "loss": 0.933,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 1.72,
1284
+ "learning_rate": 2.13981855975765e-05,
1285
+ "loss": 0.9224,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 1.72,
1290
+ "learning_rate": 2.126390477878578e-05,
1291
+ "loss": 0.9286,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 1.73,
1296
+ "learning_rate": 2.1129623959995058e-05,
1297
+ "loss": 0.9281,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 1.74,
1302
+ "learning_rate": 2.099534314120434e-05,
1303
+ "loss": 0.9378,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 1.75,
1308
+ "learning_rate": 2.086106232241362e-05,
1309
+ "loss": 0.9395,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 1.76,
1314
+ "learning_rate": 2.0726781503622898e-05,
1315
+ "loss": 0.9429,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 1.76,
1320
+ "learning_rate": 2.059250068483218e-05,
1321
+ "loss": 0.9353,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 1.77,
1326
+ "learning_rate": 2.0458219866041456e-05,
1327
+ "loss": 0.9298,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 1.78,
1332
+ "learning_rate": 2.0323939047250737e-05,
1333
+ "loss": 0.9213,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 1.79,
1338
+ "learning_rate": 2.0189658228460015e-05,
1339
+ "loss": 0.9411,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 1.8,
1344
+ "learning_rate": 2.0055377409669292e-05,
1345
+ "loss": 0.9181,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 1.8,
1350
+ "learning_rate": 1.9921096590878573e-05,
1351
+ "loss": 0.9239,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 1.81,
1356
+ "learning_rate": 1.978681577208785e-05,
1357
+ "loss": 0.936,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 1.82,
1362
+ "learning_rate": 1.9652534953297132e-05,
1363
+ "loss": 0.9262,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 1.83,
1368
+ "learning_rate": 1.9518254134506413e-05,
1369
+ "loss": 0.9138,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 1.84,
1374
+ "learning_rate": 1.9383973315715694e-05,
1375
+ "loss": 0.9285,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 1.85,
1380
+ "learning_rate": 1.924969249692497e-05,
1381
+ "loss": 0.9122,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 1.85,
1386
+ "learning_rate": 1.911541167813425e-05,
1387
+ "loss": 0.9323,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 1.86,
1392
+ "learning_rate": 1.898113085934353e-05,
1393
+ "loss": 0.9014,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 1.87,
1398
+ "learning_rate": 1.8846850040552808e-05,
1399
+ "loss": 0.929,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 1.88,
1404
+ "learning_rate": 1.871256922176209e-05,
1405
+ "loss": 0.9196,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 1.89,
1410
+ "learning_rate": 1.8578288402971366e-05,
1411
+ "loss": 0.9293,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 1.89,
1416
+ "learning_rate": 1.8444007584180647e-05,
1417
+ "loss": 0.9272,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 1.9,
1422
+ "learning_rate": 1.8309726765389925e-05,
1423
+ "loss": 0.925,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 1.91,
1428
+ "learning_rate": 1.8175445946599202e-05,
1429
+ "loss": 0.9263,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 1.92,
1434
+ "learning_rate": 1.8041165127808483e-05,
1435
+ "loss": 0.9116,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 1.93,
1440
+ "learning_rate": 1.7906884309017764e-05,
1441
+ "loss": 0.92,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 1.93,
1446
+ "learning_rate": 1.7772603490227045e-05,
1447
+ "loss": 0.9166,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 1.94,
1452
+ "learning_rate": 1.7638322671436323e-05,
1453
+ "loss": 0.9122,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 1.95,
1458
+ "learning_rate": 1.75040418526456e-05,
1459
+ "loss": 0.9195,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 1.96,
1464
+ "learning_rate": 1.736976103385488e-05,
1465
+ "loss": 0.9175,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 1.97,
1470
+ "learning_rate": 1.723548021506416e-05,
1471
+ "loss": 0.9181,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 1.97,
1476
+ "learning_rate": 1.710119939627344e-05,
1477
+ "loss": 0.9189,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 1.98,
1482
+ "learning_rate": 1.6966918577482717e-05,
1483
+ "loss": 0.9044,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 1.99,
1488
+ "learning_rate": 1.6832637758692e-05,
1489
+ "loss": 0.9123,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 2.0,
1494
+ "learning_rate": 1.6698356939901276e-05,
1495
+ "loss": 0.9144,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 2.01,
1500
+ "learning_rate": 1.6564076121110557e-05,
1501
+ "loss": 0.864,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 2.01,
1506
+ "learning_rate": 1.6429795302319838e-05,
1507
+ "loss": 0.8406,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 2.02,
1512
+ "learning_rate": 1.6295514483529115e-05,
1513
+ "loss": 0.8351,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 2.03,
1518
+ "learning_rate": 1.6161233664738396e-05,
1519
+ "loss": 0.8429,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 2.04,
1524
+ "learning_rate": 1.6026952845947674e-05,
1525
+ "loss": 0.8383,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 2.05,
1530
+ "learning_rate": 1.5892672027156955e-05,
1531
+ "loss": 0.8479,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 2.05,
1536
+ "learning_rate": 1.5758391208366233e-05,
1537
+ "loss": 0.8453,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 2.06,
1542
+ "learning_rate": 1.562411038957551e-05,
1543
+ "loss": 0.8251,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 2.07,
1548
+ "learning_rate": 1.548982957078479e-05,
1549
+ "loss": 0.8412,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 2.08,
1554
+ "learning_rate": 1.535554875199407e-05,
1555
+ "loss": 0.8415,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 2.09,
1560
+ "learning_rate": 1.5221267933203351e-05,
1561
+ "loss": 0.8252,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 2.09,
1566
+ "learning_rate": 1.5086987114412629e-05,
1567
+ "loss": 0.8376,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 2.1,
1572
+ "learning_rate": 1.495270629562191e-05,
1573
+ "loss": 0.8352,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 2.11,
1578
+ "learning_rate": 1.4818425476831187e-05,
1579
+ "loss": 0.8302,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 2.12,
1584
+ "learning_rate": 1.4684144658040467e-05,
1585
+ "loss": 0.8369,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 2.13,
1590
+ "learning_rate": 1.4549863839249748e-05,
1591
+ "loss": 0.8247,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 2.14,
1596
+ "learning_rate": 1.4415583020459025e-05,
1597
+ "loss": 0.8256,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 2.14,
1602
+ "learning_rate": 1.4281302201668306e-05,
1603
+ "loss": 0.8398,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 2.15,
1608
+ "learning_rate": 1.4147021382877584e-05,
1609
+ "loss": 0.8537,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 2.16,
1614
+ "learning_rate": 1.4012740564086865e-05,
1615
+ "loss": 0.811,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 2.17,
1620
+ "learning_rate": 1.3878459745296144e-05,
1621
+ "loss": 0.8252,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 2.18,
1626
+ "learning_rate": 1.3744178926505422e-05,
1627
+ "loss": 0.8281,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 2.18,
1632
+ "learning_rate": 1.3609898107714703e-05,
1633
+ "loss": 0.8347,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 2.19,
1638
+ "learning_rate": 1.347561728892398e-05,
1639
+ "loss": 0.8341,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 2.2,
1644
+ "learning_rate": 1.3341336470133261e-05,
1645
+ "loss": 0.8269,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 2.21,
1650
+ "learning_rate": 1.320705565134254e-05,
1651
+ "loss": 0.8336,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 2.22,
1656
+ "learning_rate": 1.3072774832551821e-05,
1657
+ "loss": 0.8297,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 2.22,
1662
+ "learning_rate": 1.2938494013761099e-05,
1663
+ "loss": 0.8384,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 2.23,
1668
+ "learning_rate": 1.2804213194970377e-05,
1669
+ "loss": 0.8339,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 2.24,
1674
+ "learning_rate": 1.2669932376179658e-05,
1675
+ "loss": 0.8311,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 2.25,
1680
+ "learning_rate": 1.2535651557388937e-05,
1681
+ "loss": 0.8215,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 2.26,
1686
+ "learning_rate": 1.2401370738598218e-05,
1687
+ "loss": 0.8382,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 2.26,
1692
+ "learning_rate": 1.2267089919807495e-05,
1693
+ "loss": 0.8266,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 2.27,
1698
+ "learning_rate": 1.2132809101016775e-05,
1699
+ "loss": 0.8302,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 2.28,
1704
+ "learning_rate": 1.1998528282226054e-05,
1705
+ "loss": 0.8256,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 2.29,
1710
+ "learning_rate": 1.1864247463435333e-05,
1711
+ "loss": 0.8348,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 2.3,
1716
+ "learning_rate": 1.1729966644644612e-05,
1717
+ "loss": 0.8316,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 2.3,
1722
+ "learning_rate": 1.1595685825853893e-05,
1723
+ "loss": 0.8313,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 2.31,
1728
+ "learning_rate": 1.1461405007063171e-05,
1729
+ "loss": 0.822,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 2.32,
1734
+ "learning_rate": 1.132712418827245e-05,
1735
+ "loss": 0.8347,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 2.33,
1740
+ "learning_rate": 1.119284336948173e-05,
1741
+ "loss": 0.8186,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 2.34,
1746
+ "learning_rate": 1.1058562550691009e-05,
1747
+ "loss": 0.8177,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 2.34,
1752
+ "learning_rate": 1.092428173190029e-05,
1753
+ "loss": 0.8221,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 2.35,
1758
+ "learning_rate": 1.0790000913109569e-05,
1759
+ "loss": 0.8208,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 2.36,
1764
+ "learning_rate": 1.0655720094318848e-05,
1765
+ "loss": 0.8267,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 2.37,
1770
+ "learning_rate": 1.0521439275528126e-05,
1771
+ "loss": 0.8322,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 2.38,
1776
+ "learning_rate": 1.0387158456737405e-05,
1777
+ "loss": 0.8276,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 2.38,
1782
+ "learning_rate": 1.0252877637946686e-05,
1783
+ "loss": 0.8148,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 2.39,
1788
+ "learning_rate": 1.0118596819155965e-05,
1789
+ "loss": 0.8249,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 2.4,
1794
+ "learning_rate": 9.984316000365245e-06,
1795
+ "loss": 0.8347,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 2.41,
1800
+ "learning_rate": 9.850035181574524e-06,
1801
+ "loss": 0.8113,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 2.42,
1806
+ "learning_rate": 9.715754362783803e-06,
1807
+ "loss": 0.8207,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 2.43,
1812
+ "learning_rate": 9.58147354399308e-06,
1813
+ "loss": 0.8338,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 2.43,
1818
+ "learning_rate": 9.447192725202362e-06,
1819
+ "loss": 0.8126,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 2.44,
1824
+ "learning_rate": 9.312911906411641e-06,
1825
+ "loss": 0.8014,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 2.45,
1830
+ "learning_rate": 9.17863108762092e-06,
1831
+ "loss": 0.8084,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 2.46,
1836
+ "learning_rate": 9.0443502688302e-06,
1837
+ "loss": 0.8142,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 2.47,
1842
+ "learning_rate": 8.910069450039479e-06,
1843
+ "loss": 0.8236,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 2.47,
1848
+ "learning_rate": 8.775788631248758e-06,
1849
+ "loss": 0.8219,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 2.48,
1854
+ "learning_rate": 8.641507812458037e-06,
1855
+ "loss": 0.8155,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 2.49,
1860
+ "learning_rate": 8.507226993667317e-06,
1861
+ "loss": 0.8175,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 2.5,
1866
+ "learning_rate": 8.372946174876596e-06,
1867
+ "loss": 0.8047,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 2.51,
1872
+ "learning_rate": 8.238665356085875e-06,
1873
+ "loss": 0.8201,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 2.51,
1878
+ "learning_rate": 8.104384537295155e-06,
1879
+ "loss": 0.8181,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 2.52,
1884
+ "learning_rate": 7.970103718504436e-06,
1885
+ "loss": 0.823,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 2.53,
1890
+ "learning_rate": 7.835822899713715e-06,
1891
+ "loss": 0.8173,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 2.54,
1896
+ "learning_rate": 7.701542080922992e-06,
1897
+ "loss": 0.8122,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 2.55,
1902
+ "learning_rate": 7.567261262132272e-06,
1903
+ "loss": 0.8293,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 2.55,
1908
+ "learning_rate": 7.432980443341552e-06,
1909
+ "loss": 0.816,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 2.56,
1914
+ "learning_rate": 7.298699624550831e-06,
1915
+ "loss": 0.8054,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 2.57,
1920
+ "learning_rate": 7.16441880576011e-06,
1921
+ "loss": 0.8179,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 2.58,
1926
+ "learning_rate": 7.0301379869693904e-06,
1927
+ "loss": 0.812,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 2.59,
1932
+ "learning_rate": 6.89585716817867e-06,
1933
+ "loss": 0.8185,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 2.59,
1938
+ "learning_rate": 6.761576349387948e-06,
1939
+ "loss": 0.8036,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 2.6,
1944
+ "learning_rate": 6.627295530597227e-06,
1945
+ "loss": 0.8133,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 2.61,
1950
+ "learning_rate": 6.493014711806507e-06,
1951
+ "loss": 0.7993,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 2.62,
1956
+ "learning_rate": 6.358733893015786e-06,
1957
+ "loss": 0.8073,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 2.63,
1962
+ "learning_rate": 6.224453074225066e-06,
1963
+ "loss": 0.8145,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 2.63,
1968
+ "learning_rate": 6.0901722554343445e-06,
1969
+ "loss": 0.821,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 2.64,
1974
+ "learning_rate": 5.955891436643625e-06,
1975
+ "loss": 0.8106,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 2.65,
1980
+ "learning_rate": 5.821610617852904e-06,
1981
+ "loss": 0.8131,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 2.66,
1986
+ "learning_rate": 5.687329799062183e-06,
1987
+ "loss": 0.8109,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 2.67,
1992
+ "learning_rate": 5.5530489802714624e-06,
1993
+ "loss": 0.8107,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 2.67,
1998
+ "learning_rate": 5.418768161480742e-06,
1999
+ "loss": 0.8099,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 2.68,
2004
+ "learning_rate": 5.284487342690021e-06,
2005
+ "loss": 0.7997,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 2.69,
2010
+ "learning_rate": 5.1502065238993e-06,
2011
+ "loss": 0.8312,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 2.7,
2016
+ "learning_rate": 5.0159257051085795e-06,
2017
+ "loss": 0.8136,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 2.71,
2022
+ "learning_rate": 4.881644886317859e-06,
2023
+ "loss": 0.8117,
2024
+ "step": 168000
2025
+ },
2026
+ {
2027
+ "epoch": 2.72,
2028
+ "learning_rate": 4.747364067527139e-06,
2029
+ "loss": 0.8043,
2030
+ "step": 168500
2031
+ },
2032
+ {
2033
+ "epoch": 2.72,
2034
+ "learning_rate": 4.613083248736417e-06,
2035
+ "loss": 0.7999,
2036
+ "step": 169000
2037
+ },
2038
+ {
2039
+ "epoch": 2.73,
2040
+ "learning_rate": 4.4788024299456975e-06,
2041
+ "loss": 0.7979,
2042
+ "step": 169500
2043
+ },
2044
+ {
2045
+ "epoch": 2.74,
2046
+ "learning_rate": 4.344521611154977e-06,
2047
+ "loss": 0.8133,
2048
+ "step": 170000
2049
+ },
2050
+ {
2051
+ "epoch": 2.75,
2052
+ "learning_rate": 4.210240792364255e-06,
2053
+ "loss": 0.8057,
2054
+ "step": 170500
2055
+ },
2056
+ {
2057
+ "epoch": 2.76,
2058
+ "learning_rate": 4.075959973573535e-06,
2059
+ "loss": 0.8122,
2060
+ "step": 171000
2061
+ },
2062
+ {
2063
+ "epoch": 2.76,
2064
+ "learning_rate": 3.941679154782815e-06,
2065
+ "loss": 0.8173,
2066
+ "step": 171500
2067
+ },
2068
+ {
2069
+ "epoch": 2.77,
2070
+ "learning_rate": 3.8073983359920934e-06,
2071
+ "loss": 0.8082,
2072
+ "step": 172000
2073
+ },
2074
+ {
2075
+ "epoch": 2.78,
2076
+ "learning_rate": 3.673117517201373e-06,
2077
+ "loss": 0.8142,
2078
+ "step": 172500
2079
+ },
2080
+ {
2081
+ "epoch": 2.79,
2082
+ "learning_rate": 3.5388366984106524e-06,
2083
+ "loss": 0.7987,
2084
+ "step": 173000
2085
+ },
2086
+ {
2087
+ "epoch": 2.8,
2088
+ "learning_rate": 3.404555879619932e-06,
2089
+ "loss": 0.8019,
2090
+ "step": 173500
2091
+ },
2092
+ {
2093
+ "epoch": 2.8,
2094
+ "learning_rate": 3.270275060829211e-06,
2095
+ "loss": 0.8086,
2096
+ "step": 174000
2097
+ },
2098
+ {
2099
+ "epoch": 2.81,
2100
+ "learning_rate": 3.1359942420384902e-06,
2101
+ "loss": 0.8075,
2102
+ "step": 174500
2103
+ },
2104
+ {
2105
+ "epoch": 2.82,
2106
+ "learning_rate": 3.00171342324777e-06,
2107
+ "loss": 0.823,
2108
+ "step": 175000
2109
+ },
2110
+ {
2111
+ "epoch": 2.83,
2112
+ "learning_rate": 2.867432604457049e-06,
2113
+ "loss": 0.8049,
2114
+ "step": 175500
2115
+ },
2116
+ {
2117
+ "epoch": 2.84,
2118
+ "learning_rate": 2.7331517856663285e-06,
2119
+ "loss": 0.8073,
2120
+ "step": 176000
2121
+ },
2122
+ {
2123
+ "epoch": 2.84,
2124
+ "learning_rate": 2.5988709668756078e-06,
2125
+ "loss": 0.8055,
2126
+ "step": 176500
2127
+ },
2128
+ {
2129
+ "epoch": 2.85,
2130
+ "learning_rate": 2.464590148084887e-06,
2131
+ "loss": 0.8171,
2132
+ "step": 177000
2133
+ },
2134
+ {
2135
+ "epoch": 2.86,
2136
+ "learning_rate": 2.3303093292941667e-06,
2137
+ "loss": 0.7918,
2138
+ "step": 177500
2139
+ },
2140
+ {
2141
+ "epoch": 2.87,
2142
+ "learning_rate": 2.1960285105034456e-06,
2143
+ "loss": 0.8077,
2144
+ "step": 178000
2145
+ },
2146
+ {
2147
+ "epoch": 2.88,
2148
+ "learning_rate": 2.061747691712725e-06,
2149
+ "loss": 0.8148,
2150
+ "step": 178500
2151
+ },
2152
+ {
2153
+ "epoch": 2.88,
2154
+ "learning_rate": 1.9274668729220045e-06,
2155
+ "loss": 0.8054,
2156
+ "step": 179000
2157
+ },
2158
+ {
2159
+ "epoch": 2.89,
2160
+ "learning_rate": 1.7931860541312836e-06,
2161
+ "loss": 0.803,
2162
+ "step": 179500
2163
+ },
2164
+ {
2165
+ "epoch": 2.9,
2166
+ "learning_rate": 1.658905235340563e-06,
2167
+ "loss": 0.7945,
2168
+ "step": 180000
2169
+ },
2170
+ {
2171
+ "epoch": 2.91,
2172
+ "learning_rate": 1.5246244165498424e-06,
2173
+ "loss": 0.8051,
2174
+ "step": 180500
2175
+ },
2176
+ {
2177
+ "epoch": 2.92,
2178
+ "learning_rate": 1.3903435977591216e-06,
2179
+ "loss": 0.7973,
2180
+ "step": 181000
2181
+ },
2182
+ {
2183
+ "epoch": 2.92,
2184
+ "learning_rate": 1.2560627789684011e-06,
2185
+ "loss": 0.7985,
2186
+ "step": 181500
2187
+ },
2188
+ {
2189
+ "epoch": 2.93,
2190
+ "learning_rate": 1.1217819601776804e-06,
2191
+ "loss": 0.7967,
2192
+ "step": 182000
2193
+ },
2194
+ {
2195
+ "epoch": 2.94,
2196
+ "learning_rate": 9.875011413869599e-07,
2197
+ "loss": 0.7941,
2198
+ "step": 182500
2199
+ },
2200
+ {
2201
+ "epoch": 2.95,
2202
+ "learning_rate": 8.532203225962391e-07,
2203
+ "loss": 0.8017,
2204
+ "step": 183000
2205
+ },
2206
+ {
2207
+ "epoch": 2.96,
2208
+ "learning_rate": 7.189395038055184e-07,
2209
+ "loss": 0.8099,
2210
+ "step": 183500
2211
+ },
2212
+ {
2213
+ "epoch": 2.96,
2214
+ "learning_rate": 5.846586850147977e-07,
2215
+ "loss": 0.802,
2216
+ "step": 184000
2217
+ },
2218
+ {
2219
+ "epoch": 2.97,
2220
+ "learning_rate": 4.503778662240771e-07,
2221
+ "loss": 0.794,
2222
+ "step": 184500
2223
+ },
2224
+ {
2225
+ "epoch": 2.98,
2226
+ "learning_rate": 3.160970474333564e-07,
2227
+ "loss": 0.7971,
2228
+ "step": 185000
2229
+ },
2230
+ {
2231
+ "epoch": 2.99,
2232
+ "learning_rate": 1.818162286426358e-07,
2233
+ "loss": 0.7976,
2234
+ "step": 185500
2235
+ },
2236
+ {
2237
+ "epoch": 3.0,
2238
+ "learning_rate": 4.7535409851915116e-08,
2239
+ "loss": 0.7943,
2240
+ "step": 186000
2241
+ },
2242
+ {
2243
+ "epoch": 3.0,
2244
+ "step": 186177,
2245
+ "total_flos": 2.7863550937792512e+17,
2246
+ "train_loss": 1.0171057777132477,
2247
+ "train_runtime": 64358.9209,
2248
+ "train_samples_per_second": 28.928,
2249
+ "train_steps_per_second": 2.893
2250
+ }
2251
+ ],
2252
+ "max_steps": 186177,
2253
+ "num_train_epochs": 3,
2254
+ "total_flos": 2.7863550937792512e+17,
2255
+ "trial_name": null,
2256
+ "trial_params": null
2257
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c42f43a16cedd729244a418c3c1a298787ed525ce6478a11952ad48c1154091a
3
+ size 2799
vocab.json ADDED
The diff for this file is too large to render. See raw diff