Hugging Face's logo

yesj1234
/
enko_mbartLarge_sup_run1

Transformers
Safetensors
English
Korean
mbart
text2text-generation
Generated from Trainer
Model card Files
xet
 Community
yesj1234 commited on
Commit
da68fa1
·
verified ·
1 Parent(s): 4510ac6

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +5 -0
  2. README.md +75 -0
  3. all_results.json +22 -0
  4. checkpoint-190759/config.json +59 -0
  5. checkpoint-190759/generation_config.json +12 -0
  6. checkpoint-190759/model.safetensors +3 -0
  7. checkpoint-190759/optimizer.pt +3 -0
  8. checkpoint-190759/rng_state_0.pth +3 -0
  9. checkpoint-190759/rng_state_1.pth +3 -0
  10. checkpoint-190759/rng_state_2.pth +3 -0
  11. checkpoint-190759/rng_state_3.pth +3 -0
  12. checkpoint-190759/scheduler.pt +3 -0
  13. checkpoint-190759/sentencepiece.bpe.model +3 -0
  14. checkpoint-190759/special_tokens_map.json +69 -0
  15. checkpoint-190759/tokenizer.json +3 -0
  16. checkpoint-190759/tokenizer_config.json +528 -0
  17. checkpoint-190759/trainer_state.json +2337 -0
  18. checkpoint-190759/training_args.bin +3 -0
  19. checkpoint-317932/config.json +59 -0
  20. checkpoint-317932/generation_config.json +12 -0
  21. checkpoint-317932/model.safetensors +3 -0
  22. checkpoint-317932/optimizer.pt +3 -0
  23. checkpoint-317932/rng_state_0.pth +3 -0
  24. checkpoint-317932/rng_state_1.pth +3 -0
  25. checkpoint-317932/rng_state_2.pth +3 -0
  26. checkpoint-317932/rng_state_3.pth +3 -0
  27. checkpoint-317932/scheduler.pt +3 -0
  28. checkpoint-317932/sentencepiece.bpe.model +3 -0
  29. checkpoint-317932/special_tokens_map.json +69 -0
  30. checkpoint-317932/tokenizer.json +3 -0
  31. checkpoint-317932/tokenizer_config.json +528 -0
  32. checkpoint-317932/trainer_state.json +3881 -0
  33. checkpoint-317932/training_args.bin +3 -0
  34. checkpoint-381519/config.json +59 -0
  35. checkpoint-381519/generation_config.json +12 -0
  36. checkpoint-381519/model.safetensors +3 -0
  37. checkpoint-381519/optimizer.pt +3 -0
  38. checkpoint-381519/rng_state_0.pth +3 -0
  39. checkpoint-381519/rng_state_1.pth +3 -0
  40. checkpoint-381519/rng_state_2.pth +3 -0
  41. checkpoint-381519/rng_state_3.pth +3 -0
  42. checkpoint-381519/scheduler.pt +3 -0
  43. checkpoint-381519/sentencepiece.bpe.model +3 -0
  44. checkpoint-381519/special_tokens_map.json +69 -0
  45. checkpoint-381519/tokenizer.json +3 -0
  46. checkpoint-381519/tokenizer_config.json +528 -0
  47. checkpoint-381519/trainer_state.json +0 -0
  48. checkpoint-381519/training_args.bin +3 -0
  49. config.json +59 -0
  50. eval_results.json +10 -0
.gitattributes CHANGED
@@ -33,3 +33,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ checkpoint-190759/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ checkpoint-317932/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
+ checkpoint-381519/tokenizer.json filter=lfs diff=lfs merge=lfs -text
39
+ generated_predictions.txt filter=lfs diff=lfs merge=lfs -text
40
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ - ko
5
+ base_model: facebook/mbart-large-50-many-to-many-mmt
6
+ tags:
7
+ - generated_from_trainer
8
+ metrics:
9
+ - bleu
10
+ model-index:
11
+ - name: enko_mbartLarge_100p_run1
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ # enko_mbartLarge_100p_run1
19
+
20
+ This model is a fine-tuned version of [facebook/mbart-large-50-many-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-many-mmt) on an unknown dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 0.6636
23
+ - Bleu: 55.7536
24
+ - Gen Len: 15.2192
25
+
26
+ ## Model description
27
+
28
+ More information needed
29
+
30
+ ## Intended uses & limitations
31
+
32
+ More information needed
33
+
34
+ ## Training and evaluation data
35
+
36
+ More information needed
37
+
38
+ ## Training procedure
39
+
40
+ ### Training hyperparameters
41
+
42
+ The following hyperparameters were used during training:
43
+ - learning_rate: 5e-05
44
+ - train_batch_size: 2
45
+ - eval_batch_size: 2
46
+ - seed: 42
47
+ - distributed_type: multi-GPU
48
+ - num_devices: 4
49
+ - gradient_accumulation_steps: 2
50
+ - total_train_batch_size: 16
51
+ - total_eval_batch_size: 8
52
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
53
+ - lr_scheduler_type: linear
54
+ - lr_scheduler_warmup_steps: 2500
55
+ - num_epochs: 15
56
+ - mixed_precision_training: Native AMP
57
+
58
+ ### Training results
59
+
60
+ | Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
61
+ |:-------------:|:-----:|:------:|:---------------:|:-------:|:-------:|
62
+ | 0.8673 | 1.0 | 63586 | 0.7307 | 52.6771 | 15.2057 |
63
+ | 0.7184 | 2.0 | 127173 | 0.6737 | 55.1744 | 15.3621 |
64
+ | 0.6294 | 3.0 | 190759 | 0.6636 | 55.7536 | 15.2192 |
65
+ | 0.5313 | 4.0 | 254346 | 0.6860 | 55.9422 | 15.2674 |
66
+ | 0.4325 | 5.0 | 317932 | 0.7183 | 56.2388 | 15.235 |
67
+ | 0.3525 | 6.0 | 381519 | 0.7727 | 55.8867 | 15.2207 |
68
+
69
+
70
+ ### Framework versions
71
+
72
+ - Transformers 4.36.2
73
+ - Pytorch 2.1.2+cu121
74
+ - Datasets 2.16.1
75
+ - Tokenizers 0.15.0
all_results.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 6.0,
3
+ "eval_bleu": 55.7536,
4
+ "eval_gen_len": 15.2192,
5
+ "eval_loss": 0.6636257171630859,
6
+ "eval_runtime": 9612.8661,
7
+ "eval_samples": 129095,
8
+ "eval_samples_per_second": 13.429,
9
+ "eval_steps_per_second": 1.679,
10
+ "predict_bleu": 55.3196,
11
+ "predict_gen_len": 15.0949,
12
+ "predict_loss": 0.6450165510177612,
13
+ "predict_runtime": 10044.7438,
14
+ "predict_samples": 136501,
15
+ "predict_samples_per_second": 13.589,
16
+ "predict_steps_per_second": 1.699,
17
+ "train_loss": 0.59605707192661,
18
+ "train_runtime": 353473.1733,
19
+ "train_samples": 1017377,
20
+ "train_samples_per_second": 43.173,
21
+ "train_steps_per_second": 2.698
22
+ }
checkpoint-190759/config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/mbart-large-50-many-to-many-mmt",
3
+ "_num_labels": 3,
4
+ "activation_dropout": 0.0,
5
+ "activation_function": "relu",
6
+ "add_bias_logits": false,
7
+ "add_final_layer_norm": true,
8
+ "architectures": [
9
+ "MBartForConditionalGeneration"
10
+ ],
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 0,
13
+ "classif_dropout": 0.0,
14
+ "classifier_dropout": 0.0,
15
+ "d_model": 1024,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_ffn_dim": 4096,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 12,
20
+ "decoder_start_token_id": 2,
21
+ "dropout": 0.1,
22
+ "early_stopping": true,
23
+ "encoder_attention_heads": 16,
24
+ "encoder_ffn_dim": 4096,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 2,
28
+ "forced_bos_token_id": 250014,
29
+ "forced_eos_token_id": 2,
30
+ "gradient_checkpointing": false,
31
+ "id2label": {
32
+ "0": "LABEL_0",
33
+ "1": "LABEL_1",
34
+ "2": "LABEL_2"
35
+ },
36
+ "init_std": 0.02,
37
+ "is_encoder_decoder": true,
38
+ "label2id": {
39
+ "LABEL_0": 0,
40
+ "LABEL_1": 1,
41
+ "LABEL_2": 2
42
+ },
43
+ "max_length": 200,
44
+ "max_position_embeddings": 1024,
45
+ "model_type": "mbart",
46
+ "normalize_before": true,
47
+ "normalize_embedding": true,
48
+ "num_beams": 5,
49
+ "num_hidden_layers": 12,
50
+ "output_past": true,
51
+ "pad_token_id": 1,
52
+ "scale_embedding": true,
53
+ "static_position_embeddings": false,
54
+ "tokenizer_class": "MBart50Tokenizer",
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.36.2",
57
+ "use_cache": true,
58
+ "vocab_size": 250054
59
+ }
checkpoint-190759/generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 0,
3
+ "decoder_start_token_id": 2,
4
+ "early_stopping": true,
5
+ "eos_token_id": 2,
6
+ "forced_bos_token_id": 250014,
7
+ "forced_eos_token_id": 2,
8
+ "max_length": 200,
9
+ "num_beams": 5,
10
+ "pad_token_id": 1,
11
+ "transformers_version": "4.36.2"
12
+ }
checkpoint-190759/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08f14ab01afbc18086a9f02dc385597e4f0482a79b4d2f209f0f44f939633029
3
+ size 2444578688
checkpoint-190759/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9298b33a2c34fe4fdea68eb0dcfdba1679dde945514c58eb3f8f8e092041adf
3
+ size 4887473903
checkpoint-190759/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e138f5e10cf32e8bd3cab286fe69e1c7685e2982d1923dd1736e27339bdd4ae9
3
+ size 15024
checkpoint-190759/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1475b7e7a0d5a3ce1eea714fa9ecf4a42b43e402a8c1d97a39e807a79aa1484
3
+ size 15024
checkpoint-190759/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5fdc5a3b32715fc09ae70cfaae2c360c9ee910de149c9569ed720dff1b6595b3
3
+ size 15024
checkpoint-190759/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5abe19adf7fc15b641894f37a40c807f002c656532c9427fa63b25c942694582
3
+ size 15024
checkpoint-190759/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:724fc615e3db34b0073a9aaff75acb754d55937f3448bca1567ef60c6e8334b8
3
+ size 1064
checkpoint-190759/sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
checkpoint-190759/special_tokens_map.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "ar_AR",
4
+ "cs_CZ",
5
+ "de_DE",
6
+ "en_XX",
7
+ "es_XX",
8
+ "et_EE",
9
+ "fi_FI",
10
+ "fr_XX",
11
+ "gu_IN",
12
+ "hi_IN",
13
+ "it_IT",
14
+ "ja_XX",
15
+ "kk_KZ",
16
+ "ko_KR",
17
+ "lt_LT",
18
+ "lv_LV",
19
+ "my_MM",
20
+ "ne_NP",
21
+ "nl_XX",
22
+ "ro_RO",
23
+ "ru_RU",
24
+ "si_LK",
25
+ "tr_TR",
26
+ "vi_VN",
27
+ "zh_CN",
28
+ "af_ZA",
29
+ "az_AZ",
30
+ "bn_IN",
31
+ "fa_IR",
32
+ "he_IL",
33
+ "hr_HR",
34
+ "id_ID",
35
+ "ka_GE",
36
+ "km_KH",
37
+ "mk_MK",
38
+ "ml_IN",
39
+ "mn_MN",
40
+ "mr_IN",
41
+ "pl_PL",
42
+ "ps_AF",
43
+ "pt_XX",
44
+ "sv_SE",
45
+ "sw_KE",
46
+ "ta_IN",
47
+ "te_IN",
48
+ "th_TH",
49
+ "tl_XX",
50
+ "uk_UA",
51
+ "ur_PK",
52
+ "xh_ZA",
53
+ "gl_ES",
54
+ "sl_SI"
55
+ ],
56
+ "bos_token": "<s>",
57
+ "cls_token": "<s>",
58
+ "eos_token": "</s>",
59
+ "mask_token": {
60
+ "content": "<mask>",
61
+ "lstrip": true,
62
+ "normalized": true,
63
+ "rstrip": false,
64
+ "single_word": false
65
+ },
66
+ "pad_token": "<pad>",
67
+ "sep_token": "</s>",
68
+ "unk_token": "<unk>"
69
+ }
checkpoint-190759/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd7eee7b2b672fc40148ee5d62e7dcc8473f2ac9630ead56f0791e9eb15d1f01
3
+ size 17110107
checkpoint-190759/tokenizer_config.json ADDED
@@ -0,0 +1,528 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "ar_AR",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "250002": {
44
+ "content": "cs_CZ",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "250003": {
52
+ "content": "de_DE",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "250004": {
60
+ "content": "en_XX",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "250005": {
68
+ "content": "es_XX",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ },
75
+ "250006": {
76
+ "content": "et_EE",
77
+ "lstrip": false,
78
+ "normalized": false,
79
+ "rstrip": false,
80
+ "single_word": false,
81
+ "special": true
82
+ },
83
+ "250007": {
84
+ "content": "fi_FI",
85
+ "lstrip": false,
86
+ "normalized": false,
87
+ "rstrip": false,
88
+ "single_word": false,
89
+ "special": true
90
+ },
91
+ "250008": {
92
+ "content": "fr_XX",
93
+ "lstrip": false,
94
+ "normalized": false,
95
+ "rstrip": false,
96
+ "single_word": false,
97
+ "special": true
98
+ },
99
+ "250009": {
100
+ "content": "gu_IN",
101
+ "lstrip": false,
102
+ "normalized": false,
103
+ "rstrip": false,
104
+ "single_word": false,
105
+ "special": true
106
+ },
107
+ "250010": {
108
+ "content": "hi_IN",
109
+ "lstrip": false,
110
+ "normalized": false,
111
+ "rstrip": false,
112
+ "single_word": false,
113
+ "special": true
114
+ },
115
+ "250011": {
116
+ "content": "it_IT",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false,
121
+ "special": true
122
+ },
123
+ "250012": {
124
+ "content": "ja_XX",
125
+ "lstrip": false,
126
+ "normalized": false,
127
+ "rstrip": false,
128
+ "single_word": false,
129
+ "special": true
130
+ },
131
+ "250013": {
132
+ "content": "kk_KZ",
133
+ "lstrip": false,
134
+ "normalized": false,
135
+ "rstrip": false,
136
+ "single_word": false,
137
+ "special": true
138
+ },
139
+ "250014": {
140
+ "content": "ko_KR",
141
+ "lstrip": false,
142
+ "normalized": false,
143
+ "rstrip": false,
144
+ "single_word": false,
145
+ "special": true
146
+ },
147
+ "250015": {
148
+ "content": "lt_LT",
149
+ "lstrip": false,
150
+ "normalized": false,
151
+ "rstrip": false,
152
+ "single_word": false,
153
+ "special": true
154
+ },
155
+ "250016": {
156
+ "content": "lv_LV",
157
+ "lstrip": false,
158
+ "normalized": false,
159
+ "rstrip": false,
160
+ "single_word": false,
161
+ "special": true
162
+ },
163
+ "250017": {
164
+ "content": "my_MM",
165
+ "lstrip": false,
166
+ "normalized": false,
167
+ "rstrip": false,
168
+ "single_word": false,
169
+ "special": true
170
+ },
171
+ "250018": {
172
+ "content": "ne_NP",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false,
177
+ "special": true
178
+ },
179
+ "250019": {
180
+ "content": "nl_XX",
181
+ "lstrip": false,
182
+ "normalized": false,
183
+ "rstrip": false,
184
+ "single_word": false,
185
+ "special": true
186
+ },
187
+ "250020": {
188
+ "content": "ro_RO",
189
+ "lstrip": false,
190
+ "normalized": false,
191
+ "rstrip": false,
192
+ "single_word": false,
193
+ "special": true
194
+ },
195
+ "250021": {
196
+ "content": "ru_RU",
197
+ "lstrip": false,
198
+ "normalized": false,
199
+ "rstrip": false,
200
+ "single_word": false,
201
+ "special": true
202
+ },
203
+ "250022": {
204
+ "content": "si_LK",
205
+ "lstrip": false,
206
+ "normalized": false,
207
+ "rstrip": false,
208
+ "single_word": false,
209
+ "special": true
210
+ },
211
+ "250023": {
212
+ "content": "tr_TR",
213
+ "lstrip": false,
214
+ "normalized": false,
215
+ "rstrip": false,
216
+ "single_word": false,
217
+ "special": true
218
+ },
219
+ "250024": {
220
+ "content": "vi_VN",
221
+ "lstrip": false,
222
+ "normalized": false,
223
+ "rstrip": false,
224
+ "single_word": false,
225
+ "special": true
226
+ },
227
+ "250025": {
228
+ "content": "zh_CN",
229
+ "lstrip": false,
230
+ "normalized": false,
231
+ "rstrip": false,
232
+ "single_word": false,
233
+ "special": true
234
+ },
235
+ "250026": {
236
+ "content": "af_ZA",
237
+ "lstrip": false,
238
+ "normalized": false,
239
+ "rstrip": false,
240
+ "single_word": false,
241
+ "special": true
242
+ },
243
+ "250027": {
244
+ "content": "az_AZ",
245
+ "lstrip": false,
246
+ "normalized": false,
247
+ "rstrip": false,
248
+ "single_word": false,
249
+ "special": true
250
+ },
251
+ "250028": {
252
+ "content": "bn_IN",
253
+ "lstrip": false,
254
+ "normalized": false,
255
+ "rstrip": false,
256
+ "single_word": false,
257
+ "special": true
258
+ },
259
+ "250029": {
260
+ "content": "fa_IR",
261
+ "lstrip": false,
262
+ "normalized": false,
263
+ "rstrip": false,
264
+ "single_word": false,
265
+ "special": true
266
+ },
267
+ "250030": {
268
+ "content": "he_IL",
269
+ "lstrip": false,
270
+ "normalized": false,
271
+ "rstrip": false,
272
+ "single_word": false,
273
+ "special": true
274
+ },
275
+ "250031": {
276
+ "content": "hr_HR",
277
+ "lstrip": false,
278
+ "normalized": false,
279
+ "rstrip": false,
280
+ "single_word": false,
281
+ "special": true
282
+ },
283
+ "250032": {
284
+ "content": "id_ID",
285
+ "lstrip": false,
286
+ "normalized": false,
287
+ "rstrip": false,
288
+ "single_word": false,
289
+ "special": true
290
+ },
291
+ "250033": {
292
+ "content": "ka_GE",
293
+ "lstrip": false,
294
+ "normalized": false,
295
+ "rstrip": false,
296
+ "single_word": false,
297
+ "special": true
298
+ },
299
+ "250034": {
300
+ "content": "km_KH",
301
+ "lstrip": false,
302
+ "normalized": false,
303
+ "rstrip": false,
304
+ "single_word": false,
305
+ "special": true
306
+ },
307
+ "250035": {
308
+ "content": "mk_MK",
309
+ "lstrip": false,
310
+ "normalized": false,
311
+ "rstrip": false,
312
+ "single_word": false,
313
+ "special": true
314
+ },
315
+ "250036": {
316
+ "content": "ml_IN",
317
+ "lstrip": false,
318
+ "normalized": false,
319
+ "rstrip": false,
320
+ "single_word": false,
321
+ "special": true
322
+ },
323
+ "250037": {
324
+ "content": "mn_MN",
325
+ "lstrip": false,
326
+ "normalized": false,
327
+ "rstrip": false,
328
+ "single_word": false,
329
+ "special": true
330
+ },
331
+ "250038": {
332
+ "content": "mr_IN",
333
+ "lstrip": false,
334
+ "normalized": false,
335
+ "rstrip": false,
336
+ "single_word": false,
337
+ "special": true
338
+ },
339
+ "250039": {
340
+ "content": "pl_PL",
341
+ "lstrip": false,
342
+ "normalized": false,
343
+ "rstrip": false,
344
+ "single_word": false,
345
+ "special": true
346
+ },
347
+ "250040": {
348
+ "content": "ps_AF",
349
+ "lstrip": false,
350
+ "normalized": false,
351
+ "rstrip": false,
352
+ "single_word": false,
353
+ "special": true
354
+ },
355
+ "250041": {
356
+ "content": "pt_XX",
357
+ "lstrip": false,
358
+ "normalized": false,
359
+ "rstrip": false,
360
+ "single_word": false,
361
+ "special": true
362
+ },
363
+ "250042": {
364
+ "content": "sv_SE",
365
+ "lstrip": false,
366
+ "normalized": false,
367
+ "rstrip": false,
368
+ "single_word": false,
369
+ "special": true
370
+ },
371
+ "250043": {
372
+ "content": "sw_KE",
373
+ "lstrip": false,
374
+ "normalized": false,
375
+ "rstrip": false,
376
+ "single_word": false,
377
+ "special": true
378
+ },
379
+ "250044": {
380
+ "content": "ta_IN",
381
+ "lstrip": false,
382
+ "normalized": false,
383
+ "rstrip": false,
384
+ "single_word": false,
385
+ "special": true
386
+ },
387
+ "250045": {
388
+ "content": "te_IN",
389
+ "lstrip": false,
390
+ "normalized": false,
391
+ "rstrip": false,
392
+ "single_word": false,
393
+ "special": true
394
+ },
395
+ "250046": {
396
+ "content": "th_TH",
397
+ "lstrip": false,
398
+ "normalized": false,
399
+ "rstrip": false,
400
+ "single_word": false,
401
+ "special": true
402
+ },
403
+ "250047": {
404
+ "content": "tl_XX",
405
+ "lstrip": false,
406
+ "normalized": false,
407
+ "rstrip": false,
408
+ "single_word": false,
409
+ "special": true
410
+ },
411
+ "250048": {
412
+ "content": "uk_UA",
413
+ "lstrip": false,
414
+ "normalized": false,
415
+ "rstrip": false,
416
+ "single_word": false,
417
+ "special": true
418
+ },
419
+ "250049": {
420
+ "content": "ur_PK",
421
+ "lstrip": false,
422
+ "normalized": false,
423
+ "rstrip": false,
424
+ "single_word": false,
425
+ "special": true
426
+ },
427
+ "250050": {
428
+ "content": "xh_ZA",
429
+ "lstrip": false,
430
+ "normalized": false,
431
+ "rstrip": false,
432
+ "single_word": false,
433
+ "special": true
434
+ },
435
+ "250051": {
436
+ "content": "gl_ES",
437
+ "lstrip": false,
438
+ "normalized": false,
439
+ "rstrip": false,
440
+ "single_word": false,
441
+ "special": true
442
+ },
443
+ "250052": {
444
+ "content": "sl_SI",
445
+ "lstrip": false,
446
+ "normalized": false,
447
+ "rstrip": false,
448
+ "single_word": false,
449
+ "special": true
450
+ },
451
+ "250053": {
452
+ "content": "<mask>",
453
+ "lstrip": true,
454
+ "normalized": true,
455
+ "rstrip": false,
456
+ "single_word": false,
457
+ "special": true
458
+ }
459
+ },
460
+ "additional_special_tokens": [
461
+ "ar_AR",
462
+ "cs_CZ",
463
+ "de_DE",
464
+ "en_XX",
465
+ "es_XX",
466
+ "et_EE",
467
+ "fi_FI",
468
+ "fr_XX",
469
+ "gu_IN",
470
+ "hi_IN",
471
+ "it_IT",
472
+ "ja_XX",
473
+ "kk_KZ",
474
+ "ko_KR",
475
+ "lt_LT",
476
+ "lv_LV",
477
+ "my_MM",
478
+ "ne_NP",
479
+ "nl_XX",
480
+ "ro_RO",
481
+ "ru_RU",
482
+ "si_LK",
483
+ "tr_TR",
484
+ "vi_VN",
485
+ "zh_CN",
486
+ "af_ZA",
487
+ "az_AZ",
488
+ "bn_IN",
489
+ "fa_IR",
490
+ "he_IL",
491
+ "hr_HR",
492
+ "id_ID",
493
+ "ka_GE",
494
+ "km_KH",
495
+ "mk_MK",
496
+ "ml_IN",
497
+ "mn_MN",
498
+ "mr_IN",
499
+ "pl_PL",
500
+ "ps_AF",
501
+ "pt_XX",
502
+ "sv_SE",
503
+ "sw_KE",
504
+ "ta_IN",
505
+ "te_IN",
506
+ "th_TH",
507
+ "tl_XX",
508
+ "uk_UA",
509
+ "ur_PK",
510
+ "xh_ZA",
511
+ "gl_ES",
512
+ "sl_SI"
513
+ ],
514
+ "bos_token": "<s>",
515
+ "clean_up_tokenization_spaces": true,
516
+ "cls_token": "<s>",
517
+ "eos_token": "</s>",
518
+ "language_codes": "ML50",
519
+ "mask_token": "<mask>",
520
+ "model_max_length": 1000000000000000019884624838656,
521
+ "pad_token": "<pad>",
522
+ "sep_token": "</s>",
523
+ "sp_model_kwargs": {},
524
+ "src_lang": "en_XX",
525
+ "tgt_lang": "ko_KR",
526
+ "tokenizer_class": "MBart50Tokenizer",
527
+ "unk_token": "<unk>"
528
+ }
checkpoint-190759/trainer_state.json ADDED
@@ -0,0 +1,2337 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.6636257171630859,
3
+ "best_model_checkpoint": "./enko_mbartLarge_100p_run1/checkpoint-190759",
4
+ "epoch": 2.999992136695682,
5
+ "eval_steps": 500,
6
+ "global_step": 190759,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 9.940000000000001e-06,
14
+ "loss": 1.6525,
15
+ "step": 500
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.994e-05,
20
+ "loss": 1.3306,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 2.994e-05,
26
+ "loss": 1.3005,
27
+ "step": 1500
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 3.994e-05,
32
+ "loss": 1.2871,
33
+ "step": 2000
34
+ },
35
+ {
36
+ "epoch": 0.04,
37
+ "learning_rate": 4.9940000000000006e-05,
38
+ "loss": 1.306,
39
+ "step": 2500
40
+ },
41
+ {
42
+ "epoch": 0.05,
43
+ "learning_rate": 4.9973930136971904e-05,
44
+ "loss": 1.2876,
45
+ "step": 3000
46
+ },
47
+ {
48
+ "epoch": 0.06,
49
+ "learning_rate": 4.994770259332065e-05,
50
+ "loss": 1.2609,
51
+ "step": 3500
52
+ },
53
+ {
54
+ "epoch": 0.06,
55
+ "learning_rate": 4.992142248946168e-05,
56
+ "loss": 1.244,
57
+ "step": 4000
58
+ },
59
+ {
60
+ "epoch": 0.07,
61
+ "learning_rate": 4.9895142385602714e-05,
62
+ "loss": 1.2051,
63
+ "step": 4500
64
+ },
65
+ {
66
+ "epoch": 0.08,
67
+ "learning_rate": 4.986886228174374e-05,
68
+ "loss": 1.1997,
69
+ "step": 5000
70
+ },
71
+ {
72
+ "epoch": 0.09,
73
+ "learning_rate": 4.9842582177884766e-05,
74
+ "loss": 1.1854,
75
+ "step": 5500
76
+ },
77
+ {
78
+ "epoch": 0.09,
79
+ "learning_rate": 4.98163020740258e-05,
80
+ "loss": 1.1666,
81
+ "step": 6000
82
+ },
83
+ {
84
+ "epoch": 0.1,
85
+ "learning_rate": 4.9790021970166825e-05,
86
+ "loss": 1.1596,
87
+ "step": 6500
88
+ },
89
+ {
90
+ "epoch": 0.11,
91
+ "learning_rate": 4.976374186630786e-05,
92
+ "loss": 1.1559,
93
+ "step": 7000
94
+ },
95
+ {
96
+ "epoch": 0.12,
97
+ "learning_rate": 4.973746176244889e-05,
98
+ "loss": 1.1555,
99
+ "step": 7500
100
+ },
101
+ {
102
+ "epoch": 0.13,
103
+ "learning_rate": 4.9711181658589916e-05,
104
+ "loss": 1.1458,
105
+ "step": 8000
106
+ },
107
+ {
108
+ "epoch": 0.13,
109
+ "learning_rate": 4.968495411493866e-05,
110
+ "loss": 1.1098,
111
+ "step": 8500
112
+ },
113
+ {
114
+ "epoch": 0.14,
115
+ "learning_rate": 4.9658674011079694e-05,
116
+ "loss": 1.1169,
117
+ "step": 9000
118
+ },
119
+ {
120
+ "epoch": 0.15,
121
+ "learning_rate": 4.963239390722073e-05,
122
+ "loss": 1.114,
123
+ "step": 9500
124
+ },
125
+ {
126
+ "epoch": 0.16,
127
+ "learning_rate": 4.960616636356947e-05,
128
+ "loss": 1.1099,
129
+ "step": 10000
130
+ },
131
+ {
132
+ "epoch": 0.17,
133
+ "learning_rate": 4.95798862597105e-05,
134
+ "loss": 1.0846,
135
+ "step": 10500
136
+ },
137
+ {
138
+ "epoch": 0.17,
139
+ "learning_rate": 4.955360615585153e-05,
140
+ "loss": 1.1021,
141
+ "step": 11000
142
+ },
143
+ {
144
+ "epoch": 0.18,
145
+ "learning_rate": 4.952732605199256e-05,
146
+ "loss": 1.0739,
147
+ "step": 11500
148
+ },
149
+ {
150
+ "epoch": 0.19,
151
+ "learning_rate": 4.950104594813359e-05,
152
+ "loss": 1.0657,
153
+ "step": 12000
154
+ },
155
+ {
156
+ "epoch": 0.2,
157
+ "learning_rate": 4.9474765844274615e-05,
158
+ "loss": 1.0641,
159
+ "step": 12500
160
+ },
161
+ {
162
+ "epoch": 0.2,
163
+ "learning_rate": 4.944848574041565e-05,
164
+ "loss": 1.0696,
165
+ "step": 13000
166
+ },
167
+ {
168
+ "epoch": 0.21,
169
+ "learning_rate": 4.94222581967644e-05,
170
+ "loss": 1.0457,
171
+ "step": 13500
172
+ },
173
+ {
174
+ "epoch": 0.22,
175
+ "learning_rate": 4.9396030653113144e-05,
176
+ "loss": 1.0573,
177
+ "step": 14000
178
+ },
179
+ {
180
+ "epoch": 0.23,
181
+ "learning_rate": 4.936975054925417e-05,
182
+ "loss": 1.0302,
183
+ "step": 14500
184
+ },
185
+ {
186
+ "epoch": 0.24,
187
+ "learning_rate": 4.93434704453952e-05,
188
+ "loss": 1.0347,
189
+ "step": 15000
190
+ },
191
+ {
192
+ "epoch": 0.24,
193
+ "learning_rate": 4.9317190341536236e-05,
194
+ "loss": 1.0304,
195
+ "step": 15500
196
+ },
197
+ {
198
+ "epoch": 0.25,
199
+ "learning_rate": 4.929096279788498e-05,
200
+ "loss": 1.0454,
201
+ "step": 16000
202
+ },
203
+ {
204
+ "epoch": 0.26,
205
+ "learning_rate": 4.926468269402601e-05,
206
+ "loss": 1.0406,
207
+ "step": 16500
208
+ },
209
+ {
210
+ "epoch": 0.27,
211
+ "learning_rate": 4.923840259016704e-05,
212
+ "loss": 1.0306,
213
+ "step": 17000
214
+ },
215
+ {
216
+ "epoch": 0.28,
217
+ "learning_rate": 4.921212248630807e-05,
218
+ "loss": 1.0306,
219
+ "step": 17500
220
+ },
221
+ {
222
+ "epoch": 0.28,
223
+ "learning_rate": 4.91858423824491e-05,
224
+ "loss": 1.0237,
225
+ "step": 18000
226
+ },
227
+ {
228
+ "epoch": 0.29,
229
+ "learning_rate": 4.9159562278590124e-05,
230
+ "loss": 1.0113,
231
+ "step": 18500
232
+ },
233
+ {
234
+ "epoch": 0.3,
235
+ "learning_rate": 4.913328217473116e-05,
236
+ "loss": 1.0101,
237
+ "step": 19000
238
+ },
239
+ {
240
+ "epoch": 0.31,
241
+ "learning_rate": 4.910700207087219e-05,
242
+ "loss": 0.9928,
243
+ "step": 19500
244
+ },
245
+ {
246
+ "epoch": 0.31,
247
+ "learning_rate": 4.9080721967013216e-05,
248
+ "loss": 0.9942,
249
+ "step": 20000
250
+ },
251
+ {
252
+ "epoch": 0.32,
253
+ "learning_rate": 4.905444186315424e-05,
254
+ "loss": 1.0043,
255
+ "step": 20500
256
+ },
257
+ {
258
+ "epoch": 0.33,
259
+ "learning_rate": 4.9028161759295275e-05,
260
+ "loss": 1.0028,
261
+ "step": 21000
262
+ },
263
+ {
264
+ "epoch": 0.34,
265
+ "learning_rate": 4.900188165543631e-05,
266
+ "loss": 0.9883,
267
+ "step": 21500
268
+ },
269
+ {
270
+ "epoch": 0.35,
271
+ "learning_rate": 4.897565411178505e-05,
272
+ "loss": 1.0023,
273
+ "step": 22000
274
+ },
275
+ {
276
+ "epoch": 0.35,
277
+ "learning_rate": 4.89494265681338e-05,
278
+ "loss": 0.9935,
279
+ "step": 22500
280
+ },
281
+ {
282
+ "epoch": 0.36,
283
+ "learning_rate": 4.892314646427483e-05,
284
+ "loss": 0.9928,
285
+ "step": 23000
286
+ },
287
+ {
288
+ "epoch": 0.37,
289
+ "learning_rate": 4.889686636041586e-05,
290
+ "loss": 0.9948,
291
+ "step": 23500
292
+ },
293
+ {
294
+ "epoch": 0.38,
295
+ "learning_rate": 4.887058625655689e-05,
296
+ "loss": 0.974,
297
+ "step": 24000
298
+ },
299
+ {
300
+ "epoch": 0.39,
301
+ "learning_rate": 4.8844306152697915e-05,
302
+ "loss": 0.9837,
303
+ "step": 24500
304
+ },
305
+ {
306
+ "epoch": 0.39,
307
+ "learning_rate": 4.881802604883895e-05,
308
+ "loss": 0.9751,
309
+ "step": 25000
310
+ },
311
+ {
312
+ "epoch": 0.4,
313
+ "learning_rate": 4.879174594497998e-05,
314
+ "loss": 0.9702,
315
+ "step": 25500
316
+ },
317
+ {
318
+ "epoch": 0.41,
319
+ "learning_rate": 4.8765465841121006e-05,
320
+ "loss": 0.9717,
321
+ "step": 26000
322
+ },
323
+ {
324
+ "epoch": 0.42,
325
+ "learning_rate": 4.873918573726204e-05,
326
+ "loss": 0.9724,
327
+ "step": 26500
328
+ },
329
+ {
330
+ "epoch": 0.42,
331
+ "learning_rate": 4.8712905633403065e-05,
332
+ "loss": 0.9692,
333
+ "step": 27000
334
+ },
335
+ {
336
+ "epoch": 0.43,
337
+ "learning_rate": 4.868667808975182e-05,
338
+ "loss": 0.9535,
339
+ "step": 27500
340
+ },
341
+ {
342
+ "epoch": 0.44,
343
+ "learning_rate": 4.866039798589284e-05,
344
+ "loss": 0.9641,
345
+ "step": 28000
346
+ },
347
+ {
348
+ "epoch": 0.45,
349
+ "learning_rate": 4.863417044224159e-05,
350
+ "loss": 0.9774,
351
+ "step": 28500
352
+ },
353
+ {
354
+ "epoch": 0.46,
355
+ "learning_rate": 4.860789033838262e-05,
356
+ "loss": 0.9671,
357
+ "step": 29000
358
+ },
359
+ {
360
+ "epoch": 0.46,
361
+ "learning_rate": 4.858161023452365e-05,
362
+ "loss": 0.9458,
363
+ "step": 29500
364
+ },
365
+ {
366
+ "epoch": 0.47,
367
+ "learning_rate": 4.855533013066468e-05,
368
+ "loss": 0.9606,
369
+ "step": 30000
370
+ },
371
+ {
372
+ "epoch": 0.48,
373
+ "learning_rate": 4.8529050026805705e-05,
374
+ "loss": 0.9557,
375
+ "step": 30500
376
+ },
377
+ {
378
+ "epoch": 0.49,
379
+ "learning_rate": 4.850276992294674e-05,
380
+ "loss": 0.9664,
381
+ "step": 31000
382
+ },
383
+ {
384
+ "epoch": 0.5,
385
+ "learning_rate": 4.8476489819087764e-05,
386
+ "loss": 0.9427,
387
+ "step": 31500
388
+ },
389
+ {
390
+ "epoch": 0.5,
391
+ "learning_rate": 4.84502097152288e-05,
392
+ "loss": 0.957,
393
+ "step": 32000
394
+ },
395
+ {
396
+ "epoch": 0.51,
397
+ "learning_rate": 4.842392961136983e-05,
398
+ "loss": 0.9466,
399
+ "step": 32500
400
+ },
401
+ {
402
+ "epoch": 0.52,
403
+ "learning_rate": 4.8397649507510856e-05,
404
+ "loss": 0.9262,
405
+ "step": 33000
406
+ },
407
+ {
408
+ "epoch": 0.53,
409
+ "learning_rate": 4.837136940365188e-05,
410
+ "loss": 0.9524,
411
+ "step": 33500
412
+ },
413
+ {
414
+ "epoch": 0.53,
415
+ "learning_rate": 4.8345089299792914e-05,
416
+ "loss": 0.9303,
417
+ "step": 34000
418
+ },
419
+ {
420
+ "epoch": 0.54,
421
+ "learning_rate": 4.831880919593395e-05,
422
+ "loss": 0.9355,
423
+ "step": 34500
424
+ },
425
+ {
426
+ "epoch": 0.55,
427
+ "learning_rate": 4.829252909207497e-05,
428
+ "loss": 0.9463,
429
+ "step": 35000
430
+ },
431
+ {
432
+ "epoch": 0.56,
433
+ "learning_rate": 4.8266248988216006e-05,
434
+ "loss": 0.9333,
435
+ "step": 35500
436
+ },
437
+ {
438
+ "epoch": 0.57,
439
+ "learning_rate": 4.823996888435703e-05,
440
+ "loss": 0.9293,
441
+ "step": 36000
442
+ },
443
+ {
444
+ "epoch": 0.57,
445
+ "learning_rate": 4.8213688780498065e-05,
446
+ "loss": 0.9272,
447
+ "step": 36500
448
+ },
449
+ {
450
+ "epoch": 0.58,
451
+ "learning_rate": 4.818740867663909e-05,
452
+ "loss": 0.928,
453
+ "step": 37000
454
+ },
455
+ {
456
+ "epoch": 0.59,
457
+ "learning_rate": 4.8161128572780124e-05,
458
+ "loss": 0.9346,
459
+ "step": 37500
460
+ },
461
+ {
462
+ "epoch": 0.6,
463
+ "learning_rate": 4.813484846892115e-05,
464
+ "loss": 0.9361,
465
+ "step": 38000
466
+ },
467
+ {
468
+ "epoch": 0.61,
469
+ "learning_rate": 4.81086209252699e-05,
470
+ "loss": 0.9197,
471
+ "step": 38500
472
+ },
473
+ {
474
+ "epoch": 0.61,
475
+ "learning_rate": 4.8082393381618646e-05,
476
+ "loss": 0.9289,
477
+ "step": 39000
478
+ },
479
+ {
480
+ "epoch": 0.62,
481
+ "learning_rate": 4.805611327775967e-05,
482
+ "loss": 0.9232,
483
+ "step": 39500
484
+ },
485
+ {
486
+ "epoch": 0.63,
487
+ "learning_rate": 4.8029833173900705e-05,
488
+ "loss": 0.9134,
489
+ "step": 40000
490
+ },
491
+ {
492
+ "epoch": 0.64,
493
+ "learning_rate": 4.800355307004174e-05,
494
+ "loss": 0.9232,
495
+ "step": 40500
496
+ },
497
+ {
498
+ "epoch": 0.64,
499
+ "learning_rate": 4.7977272966182764e-05,
500
+ "loss": 0.9315,
501
+ "step": 41000
502
+ },
503
+ {
504
+ "epoch": 0.65,
505
+ "learning_rate": 4.7950992862323796e-05,
506
+ "loss": 0.9152,
507
+ "step": 41500
508
+ },
509
+ {
510
+ "epoch": 0.66,
511
+ "learning_rate": 4.792471275846482e-05,
512
+ "loss": 0.9179,
513
+ "step": 42000
514
+ },
515
+ {
516
+ "epoch": 0.67,
517
+ "learning_rate": 4.7898485214813574e-05,
518
+ "loss": 0.914,
519
+ "step": 42500
520
+ },
521
+ {
522
+ "epoch": 0.68,
523
+ "learning_rate": 4.787225767116232e-05,
524
+ "loss": 0.9078,
525
+ "step": 43000
526
+ },
527
+ {
528
+ "epoch": 0.68,
529
+ "learning_rate": 4.7845977567303345e-05,
530
+ "loss": 0.9129,
531
+ "step": 43500
532
+ },
533
+ {
534
+ "epoch": 0.69,
535
+ "learning_rate": 4.781969746344438e-05,
536
+ "loss": 0.899,
537
+ "step": 44000
538
+ },
539
+ {
540
+ "epoch": 0.7,
541
+ "learning_rate": 4.779341735958541e-05,
542
+ "loss": 0.9003,
543
+ "step": 44500
544
+ },
545
+ {
546
+ "epoch": 0.71,
547
+ "learning_rate": 4.7767137255726436e-05,
548
+ "loss": 0.8976,
549
+ "step": 45000
550
+ },
551
+ {
552
+ "epoch": 0.72,
553
+ "learning_rate": 4.774085715186747e-05,
554
+ "loss": 0.9007,
555
+ "step": 45500
556
+ },
557
+ {
558
+ "epoch": 0.72,
559
+ "learning_rate": 4.7714577048008495e-05,
560
+ "loss": 0.9053,
561
+ "step": 46000
562
+ },
563
+ {
564
+ "epoch": 0.73,
565
+ "learning_rate": 4.768829694414952e-05,
566
+ "loss": 0.892,
567
+ "step": 46500
568
+ },
569
+ {
570
+ "epoch": 0.74,
571
+ "learning_rate": 4.766201684029056e-05,
572
+ "loss": 0.9049,
573
+ "step": 47000
574
+ },
575
+ {
576
+ "epoch": 0.75,
577
+ "learning_rate": 4.763573673643159e-05,
578
+ "loss": 0.8988,
579
+ "step": 47500
580
+ },
581
+ {
582
+ "epoch": 0.75,
583
+ "learning_rate": 4.760956175298805e-05,
584
+ "loss": 0.9086,
585
+ "step": 48000
586
+ },
587
+ {
588
+ "epoch": 0.76,
589
+ "learning_rate": 4.758328164912908e-05,
590
+ "loss": 0.903,
591
+ "step": 48500
592
+ },
593
+ {
594
+ "epoch": 0.77,
595
+ "learning_rate": 4.755700154527011e-05,
596
+ "loss": 0.895,
597
+ "step": 49000
598
+ },
599
+ {
600
+ "epoch": 0.78,
601
+ "learning_rate": 4.753072144141114e-05,
602
+ "loss": 0.8982,
603
+ "step": 49500
604
+ },
605
+ {
606
+ "epoch": 0.79,
607
+ "learning_rate": 4.750444133755217e-05,
608
+ "loss": 0.9047,
609
+ "step": 50000
610
+ },
611
+ {
612
+ "epoch": 0.79,
613
+ "learning_rate": 4.7478161233693194e-05,
614
+ "loss": 0.9088,
615
+ "step": 50500
616
+ },
617
+ {
618
+ "epoch": 0.8,
619
+ "learning_rate": 4.745188112983423e-05,
620
+ "loss": 0.892,
621
+ "step": 51000
622
+ },
623
+ {
624
+ "epoch": 0.81,
625
+ "learning_rate": 4.742560102597526e-05,
626
+ "loss": 0.89,
627
+ "step": 51500
628
+ },
629
+ {
630
+ "epoch": 0.82,
631
+ "learning_rate": 4.7399320922116286e-05,
632
+ "loss": 0.8909,
633
+ "step": 52000
634
+ },
635
+ {
636
+ "epoch": 0.83,
637
+ "learning_rate": 4.737309337846503e-05,
638
+ "loss": 0.8792,
639
+ "step": 52500
640
+ },
641
+ {
642
+ "epoch": 0.83,
643
+ "learning_rate": 4.734681327460606e-05,
644
+ "loss": 0.8783,
645
+ "step": 53000
646
+ },
647
+ {
648
+ "epoch": 0.84,
649
+ "learning_rate": 4.7320533170747096e-05,
650
+ "loss": 0.8869,
651
+ "step": 53500
652
+ },
653
+ {
654
+ "epoch": 0.85,
655
+ "learning_rate": 4.729425306688812e-05,
656
+ "loss": 0.8873,
657
+ "step": 54000
658
+ },
659
+ {
660
+ "epoch": 0.86,
661
+ "learning_rate": 4.726797296302915e-05,
662
+ "loss": 0.8927,
663
+ "step": 54500
664
+ },
665
+ {
666
+ "epoch": 0.86,
667
+ "learning_rate": 4.72417454193779e-05,
668
+ "loss": 0.8817,
669
+ "step": 55000
670
+ },
671
+ {
672
+ "epoch": 0.87,
673
+ "learning_rate": 4.721546531551893e-05,
674
+ "loss": 0.8746,
675
+ "step": 55500
676
+ },
677
+ {
678
+ "epoch": 0.88,
679
+ "learning_rate": 4.718918521165996e-05,
680
+ "loss": 0.8845,
681
+ "step": 56000
682
+ },
683
+ {
684
+ "epoch": 0.89,
685
+ "learning_rate": 4.7162905107800984e-05,
686
+ "loss": 0.8768,
687
+ "step": 56500
688
+ },
689
+ {
690
+ "epoch": 0.9,
691
+ "learning_rate": 4.713662500394202e-05,
692
+ "loss": 0.8834,
693
+ "step": 57000
694
+ },
695
+ {
696
+ "epoch": 0.9,
697
+ "learning_rate": 4.711034490008305e-05,
698
+ "loss": 0.8769,
699
+ "step": 57500
700
+ },
701
+ {
702
+ "epoch": 0.91,
703
+ "learning_rate": 4.7084064796224076e-05,
704
+ "loss": 0.865,
705
+ "step": 58000
706
+ },
707
+ {
708
+ "epoch": 0.92,
709
+ "learning_rate": 4.705783725257282e-05,
710
+ "loss": 0.8786,
711
+ "step": 58500
712
+ },
713
+ {
714
+ "epoch": 0.93,
715
+ "learning_rate": 4.7031557148713854e-05,
716
+ "loss": 0.885,
717
+ "step": 59000
718
+ },
719
+ {
720
+ "epoch": 0.94,
721
+ "learning_rate": 4.7005277044854886e-05,
722
+ "loss": 0.8804,
723
+ "step": 59500
724
+ },
725
+ {
726
+ "epoch": 0.94,
727
+ "learning_rate": 4.697899694099591e-05,
728
+ "loss": 0.8834,
729
+ "step": 60000
730
+ },
731
+ {
732
+ "epoch": 0.95,
733
+ "learning_rate": 4.6952716837136945e-05,
734
+ "loss": 0.8698,
735
+ "step": 60500
736
+ },
737
+ {
738
+ "epoch": 0.96,
739
+ "learning_rate": 4.692643673327797e-05,
740
+ "loss": 0.8605,
741
+ "step": 61000
742
+ },
743
+ {
744
+ "epoch": 0.97,
745
+ "learning_rate": 4.6900156629419004e-05,
746
+ "loss": 0.8721,
747
+ "step": 61500
748
+ },
749
+ {
750
+ "epoch": 0.98,
751
+ "learning_rate": 4.687387652556003e-05,
752
+ "loss": 0.8914,
753
+ "step": 62000
754
+ },
755
+ {
756
+ "epoch": 0.98,
757
+ "learning_rate": 4.684759642170106e-05,
758
+ "loss": 0.8743,
759
+ "step": 62500
760
+ },
761
+ {
762
+ "epoch": 0.99,
763
+ "learning_rate": 4.682131631784209e-05,
764
+ "loss": 0.8677,
765
+ "step": 63000
766
+ },
767
+ {
768
+ "epoch": 1.0,
769
+ "learning_rate": 4.679514133439856e-05,
770
+ "loss": 0.8673,
771
+ "step": 63500
772
+ },
773
+ {
774
+ "epoch": 1.0,
775
+ "eval_bleu": 52.6771,
776
+ "eval_gen_len": 15.2057,
777
+ "eval_loss": 0.7307217121124268,
778
+ "eval_runtime": 9605.9098,
779
+ "eval_samples_per_second": 13.439,
780
+ "eval_steps_per_second": 1.68,
781
+ "step": 63586
782
+ },
783
+ {
784
+ "epoch": 1.01,
785
+ "learning_rate": 4.6768913790747304e-05,
786
+ "loss": 0.7651,
787
+ "step": 64000
788
+ },
789
+ {
790
+ "epoch": 1.01,
791
+ "learning_rate": 4.674263368688833e-05,
792
+ "loss": 0.7382,
793
+ "step": 64500
794
+ },
795
+ {
796
+ "epoch": 1.02,
797
+ "learning_rate": 4.671635358302936e-05,
798
+ "loss": 0.7193,
799
+ "step": 65000
800
+ },
801
+ {
802
+ "epoch": 1.03,
803
+ "learning_rate": 4.6690073479170396e-05,
804
+ "loss": 0.7344,
805
+ "step": 65500
806
+ },
807
+ {
808
+ "epoch": 1.04,
809
+ "learning_rate": 4.666384593551914e-05,
810
+ "loss": 0.7191,
811
+ "step": 66000
812
+ },
813
+ {
814
+ "epoch": 1.05,
815
+ "learning_rate": 4.6637565831660166e-05,
816
+ "loss": 0.7228,
817
+ "step": 66500
818
+ },
819
+ {
820
+ "epoch": 1.05,
821
+ "learning_rate": 4.66112857278012e-05,
822
+ "loss": 0.7285,
823
+ "step": 67000
824
+ },
825
+ {
826
+ "epoch": 1.06,
827
+ "learning_rate": 4.658500562394223e-05,
828
+ "loss": 0.7373,
829
+ "step": 67500
830
+ },
831
+ {
832
+ "epoch": 1.07,
833
+ "learning_rate": 4.655872552008326e-05,
834
+ "loss": 0.7209,
835
+ "step": 68000
836
+ },
837
+ {
838
+ "epoch": 1.08,
839
+ "learning_rate": 4.6532445416224284e-05,
840
+ "loss": 0.7345,
841
+ "step": 68500
842
+ },
843
+ {
844
+ "epoch": 1.09,
845
+ "learning_rate": 4.6506217872573036e-05,
846
+ "loss": 0.7348,
847
+ "step": 69000
848
+ },
849
+ {
850
+ "epoch": 1.09,
851
+ "learning_rate": 4.647993776871407e-05,
852
+ "loss": 0.7391,
853
+ "step": 69500
854
+ },
855
+ {
856
+ "epoch": 1.1,
857
+ "learning_rate": 4.6453657664855094e-05,
858
+ "loss": 0.7367,
859
+ "step": 70000
860
+ },
861
+ {
862
+ "epoch": 1.11,
863
+ "learning_rate": 4.642737756099612e-05,
864
+ "loss": 0.7282,
865
+ "step": 70500
866
+ },
867
+ {
868
+ "epoch": 1.12,
869
+ "learning_rate": 4.640109745713715e-05,
870
+ "loss": 0.7379,
871
+ "step": 71000
872
+ },
873
+ {
874
+ "epoch": 1.12,
875
+ "learning_rate": 4.6374817353278186e-05,
876
+ "loss": 0.7313,
877
+ "step": 71500
878
+ },
879
+ {
880
+ "epoch": 1.13,
881
+ "learning_rate": 4.634853724941921e-05,
882
+ "loss": 0.7399,
883
+ "step": 72000
884
+ },
885
+ {
886
+ "epoch": 1.14,
887
+ "learning_rate": 4.6322257145560245e-05,
888
+ "loss": 0.731,
889
+ "step": 72500
890
+ },
891
+ {
892
+ "epoch": 1.15,
893
+ "learning_rate": 4.629597704170127e-05,
894
+ "loss": 0.7435,
895
+ "step": 73000
896
+ },
897
+ {
898
+ "epoch": 1.16,
899
+ "learning_rate": 4.62696969378423e-05,
900
+ "loss": 0.7361,
901
+ "step": 73500
902
+ },
903
+ {
904
+ "epoch": 1.16,
905
+ "learning_rate": 4.624341683398333e-05,
906
+ "loss": 0.7315,
907
+ "step": 74000
908
+ },
909
+ {
910
+ "epoch": 1.17,
911
+ "learning_rate": 4.621713673012436e-05,
912
+ "loss": 0.7404,
913
+ "step": 74500
914
+ },
915
+ {
916
+ "epoch": 1.18,
917
+ "learning_rate": 4.619090918647311e-05,
918
+ "loss": 0.7399,
919
+ "step": 75000
920
+ },
921
+ {
922
+ "epoch": 1.19,
923
+ "learning_rate": 4.616468164282186e-05,
924
+ "loss": 0.7378,
925
+ "step": 75500
926
+ },
927
+ {
928
+ "epoch": 1.2,
929
+ "learning_rate": 4.6138401538962885e-05,
930
+ "loss": 0.7391,
931
+ "step": 76000
932
+ },
933
+ {
934
+ "epoch": 1.2,
935
+ "learning_rate": 4.611212143510391e-05,
936
+ "loss": 0.7499,
937
+ "step": 76500
938
+ },
939
+ {
940
+ "epoch": 1.21,
941
+ "learning_rate": 4.6085841331244944e-05,
942
+ "loss": 0.7537,
943
+ "step": 77000
944
+ },
945
+ {
946
+ "epoch": 1.22,
947
+ "learning_rate": 4.605956122738597e-05,
948
+ "loss": 0.7423,
949
+ "step": 77500
950
+ },
951
+ {
952
+ "epoch": 1.23,
953
+ "learning_rate": 4.6033281123527e-05,
954
+ "loss": 0.7453,
955
+ "step": 78000
956
+ },
957
+ {
958
+ "epoch": 1.23,
959
+ "learning_rate": 4.6007001019668035e-05,
960
+ "loss": 0.7439,
961
+ "step": 78500
962
+ },
963
+ {
964
+ "epoch": 1.24,
965
+ "learning_rate": 4.598072091580906e-05,
966
+ "loss": 0.7446,
967
+ "step": 79000
968
+ },
969
+ {
970
+ "epoch": 1.25,
971
+ "learning_rate": 4.595444081195009e-05,
972
+ "loss": 0.7412,
973
+ "step": 79500
974
+ },
975
+ {
976
+ "epoch": 1.26,
977
+ "learning_rate": 4.592821326829884e-05,
978
+ "loss": 0.7399,
979
+ "step": 80000
980
+ },
981
+ {
982
+ "epoch": 1.27,
983
+ "learning_rate": 4.590193316443987e-05,
984
+ "loss": 0.7302,
985
+ "step": 80500
986
+ },
987
+ {
988
+ "epoch": 1.27,
989
+ "learning_rate": 4.58756530605809e-05,
990
+ "loss": 0.7431,
991
+ "step": 81000
992
+ },
993
+ {
994
+ "epoch": 1.28,
995
+ "learning_rate": 4.5849372956721924e-05,
996
+ "loss": 0.7437,
997
+ "step": 81500
998
+ },
999
+ {
1000
+ "epoch": 1.29,
1001
+ "learning_rate": 4.5823145413070675e-05,
1002
+ "loss": 0.7443,
1003
+ "step": 82000
1004
+ },
1005
+ {
1006
+ "epoch": 1.3,
1007
+ "learning_rate": 4.579686530921171e-05,
1008
+ "loss": 0.7496,
1009
+ "step": 82500
1010
+ },
1011
+ {
1012
+ "epoch": 1.31,
1013
+ "learning_rate": 4.577063776556045e-05,
1014
+ "loss": 0.7395,
1015
+ "step": 83000
1016
+ },
1017
+ {
1018
+ "epoch": 1.31,
1019
+ "learning_rate": 4.574435766170148e-05,
1020
+ "loss": 0.7307,
1021
+ "step": 83500
1022
+ },
1023
+ {
1024
+ "epoch": 1.32,
1025
+ "learning_rate": 4.571807755784251e-05,
1026
+ "loss": 0.7467,
1027
+ "step": 84000
1028
+ },
1029
+ {
1030
+ "epoch": 1.33,
1031
+ "learning_rate": 4.5691797453983544e-05,
1032
+ "loss": 0.7474,
1033
+ "step": 84500
1034
+ },
1035
+ {
1036
+ "epoch": 1.34,
1037
+ "learning_rate": 4.566551735012457e-05,
1038
+ "loss": 0.7352,
1039
+ "step": 85000
1040
+ },
1041
+ {
1042
+ "epoch": 1.34,
1043
+ "learning_rate": 4.5639237246265596e-05,
1044
+ "loss": 0.7377,
1045
+ "step": 85500
1046
+ },
1047
+ {
1048
+ "epoch": 1.35,
1049
+ "learning_rate": 4.561295714240663e-05,
1050
+ "loss": 0.7389,
1051
+ "step": 86000
1052
+ },
1053
+ {
1054
+ "epoch": 1.36,
1055
+ "learning_rate": 4.558667703854766e-05,
1056
+ "loss": 0.7372,
1057
+ "step": 86500
1058
+ },
1059
+ {
1060
+ "epoch": 1.37,
1061
+ "learning_rate": 4.556039693468869e-05,
1062
+ "loss": 0.7488,
1063
+ "step": 87000
1064
+ },
1065
+ {
1066
+ "epoch": 1.38,
1067
+ "learning_rate": 4.553416939103743e-05,
1068
+ "loss": 0.7303,
1069
+ "step": 87500
1070
+ },
1071
+ {
1072
+ "epoch": 1.38,
1073
+ "learning_rate": 4.5507889287178466e-05,
1074
+ "loss": 0.7243,
1075
+ "step": 88000
1076
+ },
1077
+ {
1078
+ "epoch": 1.39,
1079
+ "learning_rate": 4.54816091833195e-05,
1080
+ "loss": 0.7318,
1081
+ "step": 88500
1082
+ },
1083
+ {
1084
+ "epoch": 1.4,
1085
+ "learning_rate": 4.5455329079460524e-05,
1086
+ "loss": 0.7421,
1087
+ "step": 89000
1088
+ },
1089
+ {
1090
+ "epoch": 1.41,
1091
+ "learning_rate": 4.542904897560155e-05,
1092
+ "loss": 0.7441,
1093
+ "step": 89500
1094
+ },
1095
+ {
1096
+ "epoch": 1.42,
1097
+ "learning_rate": 4.540276887174258e-05,
1098
+ "loss": 0.7388,
1099
+ "step": 90000
1100
+ },
1101
+ {
1102
+ "epoch": 1.42,
1103
+ "learning_rate": 4.537659388829905e-05,
1104
+ "loss": 0.7381,
1105
+ "step": 90500
1106
+ },
1107
+ {
1108
+ "epoch": 1.43,
1109
+ "learning_rate": 4.535031378444008e-05,
1110
+ "loss": 0.7402,
1111
+ "step": 91000
1112
+ },
1113
+ {
1114
+ "epoch": 1.44,
1115
+ "learning_rate": 4.5324033680581106e-05,
1116
+ "loss": 0.7377,
1117
+ "step": 91500
1118
+ },
1119
+ {
1120
+ "epoch": 1.45,
1121
+ "learning_rate": 4.529775357672214e-05,
1122
+ "loss": 0.7474,
1123
+ "step": 92000
1124
+ },
1125
+ {
1126
+ "epoch": 1.45,
1127
+ "learning_rate": 4.527152603307088e-05,
1128
+ "loss": 0.7473,
1129
+ "step": 92500
1130
+ },
1131
+ {
1132
+ "epoch": 1.46,
1133
+ "learning_rate": 4.5245245929211916e-05,
1134
+ "loss": 0.7345,
1135
+ "step": 93000
1136
+ },
1137
+ {
1138
+ "epoch": 1.47,
1139
+ "learning_rate": 4.521896582535294e-05,
1140
+ "loss": 0.7366,
1141
+ "step": 93500
1142
+ },
1143
+ {
1144
+ "epoch": 1.48,
1145
+ "learning_rate": 4.5192685721493975e-05,
1146
+ "loss": 0.7518,
1147
+ "step": 94000
1148
+ },
1149
+ {
1150
+ "epoch": 1.49,
1151
+ "learning_rate": 4.516640561763501e-05,
1152
+ "loss": 0.7454,
1153
+ "step": 94500
1154
+ },
1155
+ {
1156
+ "epoch": 1.49,
1157
+ "learning_rate": 4.514017807398375e-05,
1158
+ "loss": 0.7432,
1159
+ "step": 95000
1160
+ },
1161
+ {
1162
+ "epoch": 1.5,
1163
+ "learning_rate": 4.511389797012478e-05,
1164
+ "loss": 0.7448,
1165
+ "step": 95500
1166
+ },
1167
+ {
1168
+ "epoch": 1.51,
1169
+ "learning_rate": 4.508761786626581e-05,
1170
+ "loss": 0.7415,
1171
+ "step": 96000
1172
+ },
1173
+ {
1174
+ "epoch": 1.52,
1175
+ "learning_rate": 4.5061337762406844e-05,
1176
+ "loss": 0.7358,
1177
+ "step": 96500
1178
+ },
1179
+ {
1180
+ "epoch": 1.53,
1181
+ "learning_rate": 4.503505765854787e-05,
1182
+ "loss": 0.7491,
1183
+ "step": 97000
1184
+ },
1185
+ {
1186
+ "epoch": 1.53,
1187
+ "learning_rate": 4.5008777554688896e-05,
1188
+ "loss": 0.7429,
1189
+ "step": 97500
1190
+ },
1191
+ {
1192
+ "epoch": 1.54,
1193
+ "learning_rate": 4.498249745082993e-05,
1194
+ "loss": 0.748,
1195
+ "step": 98000
1196
+ },
1197
+ {
1198
+ "epoch": 1.55,
1199
+ "learning_rate": 4.4956217346970955e-05,
1200
+ "loss": 0.7467,
1201
+ "step": 98500
1202
+ },
1203
+ {
1204
+ "epoch": 1.56,
1205
+ "learning_rate": 4.492993724311199e-05,
1206
+ "loss": 0.7459,
1207
+ "step": 99000
1208
+ },
1209
+ {
1210
+ "epoch": 1.56,
1211
+ "learning_rate": 4.490365713925302e-05,
1212
+ "loss": 0.7402,
1213
+ "step": 99500
1214
+ },
1215
+ {
1216
+ "epoch": 1.57,
1217
+ "learning_rate": 4.4877377035394046e-05,
1218
+ "loss": 0.7532,
1219
+ "step": 100000
1220
+ },
1221
+ {
1222
+ "epoch": 1.58,
1223
+ "learning_rate": 4.485120205195052e-05,
1224
+ "loss": 0.7464,
1225
+ "step": 100500
1226
+ },
1227
+ {
1228
+ "epoch": 1.59,
1229
+ "learning_rate": 4.482492194809154e-05,
1230
+ "loss": 0.7441,
1231
+ "step": 101000
1232
+ },
1233
+ {
1234
+ "epoch": 1.6,
1235
+ "learning_rate": 4.479869440444029e-05,
1236
+ "loss": 0.7445,
1237
+ "step": 101500
1238
+ },
1239
+ {
1240
+ "epoch": 1.6,
1241
+ "learning_rate": 4.4772414300581314e-05,
1242
+ "loss": 0.7316,
1243
+ "step": 102000
1244
+ },
1245
+ {
1246
+ "epoch": 1.61,
1247
+ "learning_rate": 4.474613419672235e-05,
1248
+ "loss": 0.7437,
1249
+ "step": 102500
1250
+ },
1251
+ {
1252
+ "epoch": 1.62,
1253
+ "learning_rate": 4.471985409286338e-05,
1254
+ "loss": 0.7456,
1255
+ "step": 103000
1256
+ },
1257
+ {
1258
+ "epoch": 1.63,
1259
+ "learning_rate": 4.4693573989004405e-05,
1260
+ "loss": 0.7508,
1261
+ "step": 103500
1262
+ },
1263
+ {
1264
+ "epoch": 1.64,
1265
+ "learning_rate": 4.466729388514543e-05,
1266
+ "loss": 0.7436,
1267
+ "step": 104000
1268
+ },
1269
+ {
1270
+ "epoch": 1.64,
1271
+ "learning_rate": 4.4641013781286464e-05,
1272
+ "loss": 0.7413,
1273
+ "step": 104500
1274
+ },
1275
+ {
1276
+ "epoch": 1.65,
1277
+ "learning_rate": 4.46147336774275e-05,
1278
+ "loss": 0.7305,
1279
+ "step": 105000
1280
+ },
1281
+ {
1282
+ "epoch": 1.66,
1283
+ "learning_rate": 4.458845357356852e-05,
1284
+ "loss": 0.7496,
1285
+ "step": 105500
1286
+ },
1287
+ {
1288
+ "epoch": 1.67,
1289
+ "learning_rate": 4.4562173469709555e-05,
1290
+ "loss": 0.7467,
1291
+ "step": 106000
1292
+ },
1293
+ {
1294
+ "epoch": 1.67,
1295
+ "learning_rate": 4.453589336585058e-05,
1296
+ "loss": 0.7404,
1297
+ "step": 106500
1298
+ },
1299
+ {
1300
+ "epoch": 1.68,
1301
+ "learning_rate": 4.4509613261991614e-05,
1302
+ "loss": 0.74,
1303
+ "step": 107000
1304
+ },
1305
+ {
1306
+ "epoch": 1.69,
1307
+ "learning_rate": 4.448333315813265e-05,
1308
+ "loss": 0.7432,
1309
+ "step": 107500
1310
+ },
1311
+ {
1312
+ "epoch": 1.7,
1313
+ "learning_rate": 4.445705305427367e-05,
1314
+ "loss": 0.7401,
1315
+ "step": 108000
1316
+ },
1317
+ {
1318
+ "epoch": 1.71,
1319
+ "learning_rate": 4.44307729504147e-05,
1320
+ "loss": 0.7382,
1321
+ "step": 108500
1322
+ },
1323
+ {
1324
+ "epoch": 1.71,
1325
+ "learning_rate": 4.440449284655573e-05,
1326
+ "loss": 0.7352,
1327
+ "step": 109000
1328
+ },
1329
+ {
1330
+ "epoch": 1.72,
1331
+ "learning_rate": 4.4378212742696765e-05,
1332
+ "loss": 0.7293,
1333
+ "step": 109500
1334
+ },
1335
+ {
1336
+ "epoch": 1.73,
1337
+ "learning_rate": 4.435193263883779e-05,
1338
+ "loss": 0.742,
1339
+ "step": 110000
1340
+ },
1341
+ {
1342
+ "epoch": 1.74,
1343
+ "learning_rate": 4.432565253497882e-05,
1344
+ "loss": 0.7426,
1345
+ "step": 110500
1346
+ },
1347
+ {
1348
+ "epoch": 1.75,
1349
+ "learning_rate": 4.429942499132757e-05,
1350
+ "loss": 0.7428,
1351
+ "step": 111000
1352
+ },
1353
+ {
1354
+ "epoch": 1.75,
1355
+ "learning_rate": 4.42731448874686e-05,
1356
+ "loss": 0.7391,
1357
+ "step": 111500
1358
+ },
1359
+ {
1360
+ "epoch": 1.76,
1361
+ "learning_rate": 4.4246917343817346e-05,
1362
+ "loss": 0.7449,
1363
+ "step": 112000
1364
+ },
1365
+ {
1366
+ "epoch": 1.77,
1367
+ "learning_rate": 4.422063723995837e-05,
1368
+ "loss": 0.7377,
1369
+ "step": 112500
1370
+ },
1371
+ {
1372
+ "epoch": 1.78,
1373
+ "learning_rate": 4.4194357136099405e-05,
1374
+ "loss": 0.7468,
1375
+ "step": 113000
1376
+ },
1377
+ {
1378
+ "epoch": 1.78,
1379
+ "learning_rate": 4.416807703224044e-05,
1380
+ "loss": 0.7398,
1381
+ "step": 113500
1382
+ },
1383
+ {
1384
+ "epoch": 1.79,
1385
+ "learning_rate": 4.4141796928381463e-05,
1386
+ "loss": 0.7371,
1387
+ "step": 114000
1388
+ },
1389
+ {
1390
+ "epoch": 1.8,
1391
+ "learning_rate": 4.411556938473021e-05,
1392
+ "loss": 0.7371,
1393
+ "step": 114500
1394
+ },
1395
+ {
1396
+ "epoch": 1.81,
1397
+ "learning_rate": 4.4089289280871234e-05,
1398
+ "loss": 0.7443,
1399
+ "step": 115000
1400
+ },
1401
+ {
1402
+ "epoch": 1.82,
1403
+ "learning_rate": 4.4063009177012274e-05,
1404
+ "loss": 0.7375,
1405
+ "step": 115500
1406
+ },
1407
+ {
1408
+ "epoch": 1.82,
1409
+ "learning_rate": 4.40367290731533e-05,
1410
+ "loss": 0.7339,
1411
+ "step": 116000
1412
+ },
1413
+ {
1414
+ "epoch": 1.83,
1415
+ "learning_rate": 4.4010448969294326e-05,
1416
+ "loss": 0.7267,
1417
+ "step": 116500
1418
+ },
1419
+ {
1420
+ "epoch": 1.84,
1421
+ "learning_rate": 4.398416886543536e-05,
1422
+ "loss": 0.7378,
1423
+ "step": 117000
1424
+ },
1425
+ {
1426
+ "epoch": 1.85,
1427
+ "learning_rate": 4.3957888761576385e-05,
1428
+ "loss": 0.7313,
1429
+ "step": 117500
1430
+ },
1431
+ {
1432
+ "epoch": 1.86,
1433
+ "learning_rate": 4.393160865771742e-05,
1434
+ "loss": 0.739,
1435
+ "step": 118000
1436
+ },
1437
+ {
1438
+ "epoch": 1.86,
1439
+ "learning_rate": 4.390532855385845e-05,
1440
+ "loss": 0.7314,
1441
+ "step": 118500
1442
+ },
1443
+ {
1444
+ "epoch": 1.87,
1445
+ "learning_rate": 4.3879048449999476e-05,
1446
+ "loss": 0.7328,
1447
+ "step": 119000
1448
+ },
1449
+ {
1450
+ "epoch": 1.88,
1451
+ "learning_rate": 4.385282090634822e-05,
1452
+ "loss": 0.741,
1453
+ "step": 119500
1454
+ },
1455
+ {
1456
+ "epoch": 1.89,
1457
+ "learning_rate": 4.382659336269697e-05,
1458
+ "loss": 0.7419,
1459
+ "step": 120000
1460
+ },
1461
+ {
1462
+ "epoch": 1.9,
1463
+ "learning_rate": 4.3800313258838e-05,
1464
+ "loss": 0.7353,
1465
+ "step": 120500
1466
+ },
1467
+ {
1468
+ "epoch": 1.9,
1469
+ "learning_rate": 4.377403315497903e-05,
1470
+ "loss": 0.7315,
1471
+ "step": 121000
1472
+ },
1473
+ {
1474
+ "epoch": 1.91,
1475
+ "learning_rate": 4.374775305112006e-05,
1476
+ "loss": 0.7457,
1477
+ "step": 121500
1478
+ },
1479
+ {
1480
+ "epoch": 1.92,
1481
+ "learning_rate": 4.372147294726109e-05,
1482
+ "loss": 0.7442,
1483
+ "step": 122000
1484
+ },
1485
+ {
1486
+ "epoch": 1.93,
1487
+ "learning_rate": 4.369519284340212e-05,
1488
+ "loss": 0.7421,
1489
+ "step": 122500
1490
+ },
1491
+ {
1492
+ "epoch": 1.93,
1493
+ "learning_rate": 4.366891273954315e-05,
1494
+ "loss": 0.7298,
1495
+ "step": 123000
1496
+ },
1497
+ {
1498
+ "epoch": 1.94,
1499
+ "learning_rate": 4.3642632635684175e-05,
1500
+ "loss": 0.7387,
1501
+ "step": 123500
1502
+ },
1503
+ {
1504
+ "epoch": 1.95,
1505
+ "learning_rate": 4.361635253182521e-05,
1506
+ "loss": 0.7322,
1507
+ "step": 124000
1508
+ },
1509
+ {
1510
+ "epoch": 1.96,
1511
+ "learning_rate": 4.359012498817395e-05,
1512
+ "loss": 0.7308,
1513
+ "step": 124500
1514
+ },
1515
+ {
1516
+ "epoch": 1.97,
1517
+ "learning_rate": 4.3563844884314985e-05,
1518
+ "loss": 0.7491,
1519
+ "step": 125000
1520
+ },
1521
+ {
1522
+ "epoch": 1.97,
1523
+ "learning_rate": 4.353756478045601e-05,
1524
+ "loss": 0.7292,
1525
+ "step": 125500
1526
+ },
1527
+ {
1528
+ "epoch": 1.98,
1529
+ "learning_rate": 4.3511284676597044e-05,
1530
+ "loss": 0.7307,
1531
+ "step": 126000
1532
+ },
1533
+ {
1534
+ "epoch": 1.99,
1535
+ "learning_rate": 4.348500457273808e-05,
1536
+ "loss": 0.7399,
1537
+ "step": 126500
1538
+ },
1539
+ {
1540
+ "epoch": 2.0,
1541
+ "learning_rate": 4.345877702908682e-05,
1542
+ "loss": 0.7184,
1543
+ "step": 127000
1544
+ },
1545
+ {
1546
+ "epoch": 2.0,
1547
+ "eval_bleu": 55.1744,
1548
+ "eval_gen_len": 15.3621,
1549
+ "eval_loss": 0.6736727952957153,
1550
+ "eval_runtime": 9675.9055,
1551
+ "eval_samples_per_second": 13.342,
1552
+ "eval_steps_per_second": 1.668,
1553
+ "step": 127173
1554
+ },
1555
+ {
1556
+ "epoch": 2.01,
1557
+ "learning_rate": 4.343249692522785e-05,
1558
+ "loss": 0.6388,
1559
+ "step": 127500
1560
+ },
1561
+ {
1562
+ "epoch": 2.01,
1563
+ "learning_rate": 4.340621682136888e-05,
1564
+ "loss": 0.5765,
1565
+ "step": 128000
1566
+ },
1567
+ {
1568
+ "epoch": 2.02,
1569
+ "learning_rate": 4.3379936717509913e-05,
1570
+ "loss": 0.5839,
1571
+ "step": 128500
1572
+ },
1573
+ {
1574
+ "epoch": 2.03,
1575
+ "learning_rate": 4.335365661365094e-05,
1576
+ "loss": 0.5815,
1577
+ "step": 129000
1578
+ },
1579
+ {
1580
+ "epoch": 2.04,
1581
+ "learning_rate": 4.3327376509791965e-05,
1582
+ "loss": 0.5716,
1583
+ "step": 129500
1584
+ },
1585
+ {
1586
+ "epoch": 2.04,
1587
+ "learning_rate": 4.3301096405933e-05,
1588
+ "loss": 0.5826,
1589
+ "step": 130000
1590
+ },
1591
+ {
1592
+ "epoch": 2.05,
1593
+ "learning_rate": 4.327481630207403e-05,
1594
+ "loss": 0.5783,
1595
+ "step": 130500
1596
+ },
1597
+ {
1598
+ "epoch": 2.06,
1599
+ "learning_rate": 4.3248588758422776e-05,
1600
+ "loss": 0.5879,
1601
+ "step": 131000
1602
+ },
1603
+ {
1604
+ "epoch": 2.07,
1605
+ "learning_rate": 4.32223086545638e-05,
1606
+ "loss": 0.5824,
1607
+ "step": 131500
1608
+ },
1609
+ {
1610
+ "epoch": 2.08,
1611
+ "learning_rate": 4.3196028550704835e-05,
1612
+ "loss": 0.583,
1613
+ "step": 132000
1614
+ },
1615
+ {
1616
+ "epoch": 2.08,
1617
+ "learning_rate": 4.316974844684587e-05,
1618
+ "loss": 0.5797,
1619
+ "step": 132500
1620
+ },
1621
+ {
1622
+ "epoch": 2.09,
1623
+ "learning_rate": 4.3143468342986893e-05,
1624
+ "loss": 0.5963,
1625
+ "step": 133000
1626
+ },
1627
+ {
1628
+ "epoch": 2.1,
1629
+ "learning_rate": 4.311724079933564e-05,
1630
+ "loss": 0.5841,
1631
+ "step": 133500
1632
+ },
1633
+ {
1634
+ "epoch": 2.11,
1635
+ "learning_rate": 4.309096069547667e-05,
1636
+ "loss": 0.5863,
1637
+ "step": 134000
1638
+ },
1639
+ {
1640
+ "epoch": 2.12,
1641
+ "learning_rate": 4.3064680591617704e-05,
1642
+ "loss": 0.5947,
1643
+ "step": 134500
1644
+ },
1645
+ {
1646
+ "epoch": 2.12,
1647
+ "learning_rate": 4.303840048775873e-05,
1648
+ "loss": 0.5977,
1649
+ "step": 135000
1650
+ },
1651
+ {
1652
+ "epoch": 2.13,
1653
+ "learning_rate": 4.3012172944107475e-05,
1654
+ "loss": 0.5848,
1655
+ "step": 135500
1656
+ },
1657
+ {
1658
+ "epoch": 2.14,
1659
+ "learning_rate": 4.298589284024851e-05,
1660
+ "loss": 0.5972,
1661
+ "step": 136000
1662
+ },
1663
+ {
1664
+ "epoch": 2.15,
1665
+ "learning_rate": 4.295961273638954e-05,
1666
+ "loss": 0.5979,
1667
+ "step": 136500
1668
+ },
1669
+ {
1670
+ "epoch": 2.15,
1671
+ "learning_rate": 4.2933385192738285e-05,
1672
+ "loss": 0.5879,
1673
+ "step": 137000
1674
+ },
1675
+ {
1676
+ "epoch": 2.16,
1677
+ "learning_rate": 4.290710508887931e-05,
1678
+ "loss": 0.6024,
1679
+ "step": 137500
1680
+ },
1681
+ {
1682
+ "epoch": 2.17,
1683
+ "learning_rate": 4.288082498502034e-05,
1684
+ "loss": 0.5893,
1685
+ "step": 138000
1686
+ },
1687
+ {
1688
+ "epoch": 2.18,
1689
+ "learning_rate": 4.285459744136909e-05,
1690
+ "loss": 0.6006,
1691
+ "step": 138500
1692
+ },
1693
+ {
1694
+ "epoch": 2.19,
1695
+ "learning_rate": 4.282831733751012e-05,
1696
+ "loss": 0.5983,
1697
+ "step": 139000
1698
+ },
1699
+ {
1700
+ "epoch": 2.19,
1701
+ "learning_rate": 4.280203723365115e-05,
1702
+ "loss": 0.6009,
1703
+ "step": 139500
1704
+ },
1705
+ {
1706
+ "epoch": 2.2,
1707
+ "learning_rate": 4.2775757129792173e-05,
1708
+ "loss": 0.6056,
1709
+ "step": 140000
1710
+ },
1711
+ {
1712
+ "epoch": 2.21,
1713
+ "learning_rate": 4.274947702593321e-05,
1714
+ "loss": 0.5942,
1715
+ "step": 140500
1716
+ },
1717
+ {
1718
+ "epoch": 2.22,
1719
+ "learning_rate": 4.272330204248967e-05,
1720
+ "loss": 0.5949,
1721
+ "step": 141000
1722
+ },
1723
+ {
1724
+ "epoch": 2.23,
1725
+ "learning_rate": 4.26970219386307e-05,
1726
+ "loss": 0.5972,
1727
+ "step": 141500
1728
+ },
1729
+ {
1730
+ "epoch": 2.23,
1731
+ "learning_rate": 4.2670741834771735e-05,
1732
+ "loss": 0.6073,
1733
+ "step": 142000
1734
+ },
1735
+ {
1736
+ "epoch": 2.24,
1737
+ "learning_rate": 4.264446173091276e-05,
1738
+ "loss": 0.6031,
1739
+ "step": 142500
1740
+ },
1741
+ {
1742
+ "epoch": 2.25,
1743
+ "learning_rate": 4.2618181627053794e-05,
1744
+ "loss": 0.5975,
1745
+ "step": 143000
1746
+ },
1747
+ {
1748
+ "epoch": 2.26,
1749
+ "learning_rate": 4.259190152319482e-05,
1750
+ "loss": 0.5965,
1751
+ "step": 143500
1752
+ },
1753
+ {
1754
+ "epoch": 2.26,
1755
+ "learning_rate": 4.2565621419335846e-05,
1756
+ "loss": 0.5991,
1757
+ "step": 144000
1758
+ },
1759
+ {
1760
+ "epoch": 2.27,
1761
+ "learning_rate": 4.2539341315476886e-05,
1762
+ "loss": 0.597,
1763
+ "step": 144500
1764
+ },
1765
+ {
1766
+ "epoch": 2.28,
1767
+ "learning_rate": 4.251306121161791e-05,
1768
+ "loss": 0.5909,
1769
+ "step": 145000
1770
+ },
1771
+ {
1772
+ "epoch": 2.29,
1773
+ "learning_rate": 4.248678110775894e-05,
1774
+ "loss": 0.6014,
1775
+ "step": 145500
1776
+ },
1777
+ {
1778
+ "epoch": 2.3,
1779
+ "learning_rate": 4.246050100389997e-05,
1780
+ "loss": 0.6016,
1781
+ "step": 146000
1782
+ },
1783
+ {
1784
+ "epoch": 2.3,
1785
+ "learning_rate": 4.2434220900041e-05,
1786
+ "loss": 0.6034,
1787
+ "step": 146500
1788
+ },
1789
+ {
1790
+ "epoch": 2.31,
1791
+ "learning_rate": 4.240794079618203e-05,
1792
+ "loss": 0.6033,
1793
+ "step": 147000
1794
+ },
1795
+ {
1796
+ "epoch": 2.32,
1797
+ "learning_rate": 4.238166069232306e-05,
1798
+ "loss": 0.6034,
1799
+ "step": 147500
1800
+ },
1801
+ {
1802
+ "epoch": 2.33,
1803
+ "learning_rate": 4.235538058846409e-05,
1804
+ "loss": 0.6123,
1805
+ "step": 148000
1806
+ },
1807
+ {
1808
+ "epoch": 2.34,
1809
+ "learning_rate": 4.2329100484605114e-05,
1810
+ "loss": 0.6087,
1811
+ "step": 148500
1812
+ },
1813
+ {
1814
+ "epoch": 2.34,
1815
+ "learning_rate": 4.230282038074615e-05,
1816
+ "loss": 0.6092,
1817
+ "step": 149000
1818
+ },
1819
+ {
1820
+ "epoch": 2.35,
1821
+ "learning_rate": 4.227654027688718e-05,
1822
+ "loss": 0.6045,
1823
+ "step": 149500
1824
+ },
1825
+ {
1826
+ "epoch": 2.36,
1827
+ "learning_rate": 4.2250312733235925e-05,
1828
+ "loss": 0.5975,
1829
+ "step": 150000
1830
+ },
1831
+ {
1832
+ "epoch": 2.37,
1833
+ "learning_rate": 4.2224137749792395e-05,
1834
+ "loss": 0.6023,
1835
+ "step": 150500
1836
+ },
1837
+ {
1838
+ "epoch": 2.37,
1839
+ "learning_rate": 4.219785764593342e-05,
1840
+ "loss": 0.6035,
1841
+ "step": 151000
1842
+ },
1843
+ {
1844
+ "epoch": 2.38,
1845
+ "learning_rate": 4.2171630102282166e-05,
1846
+ "loss": 0.5928,
1847
+ "step": 151500
1848
+ },
1849
+ {
1850
+ "epoch": 2.39,
1851
+ "learning_rate": 4.214534999842319e-05,
1852
+ "loss": 0.6116,
1853
+ "step": 152000
1854
+ },
1855
+ {
1856
+ "epoch": 2.4,
1857
+ "learning_rate": 4.2119069894564225e-05,
1858
+ "loss": 0.6186,
1859
+ "step": 152500
1860
+ },
1861
+ {
1862
+ "epoch": 2.41,
1863
+ "learning_rate": 4.209278979070526e-05,
1864
+ "loss": 0.6025,
1865
+ "step": 153000
1866
+ },
1867
+ {
1868
+ "epoch": 2.41,
1869
+ "learning_rate": 4.2066509686846283e-05,
1870
+ "loss": 0.5997,
1871
+ "step": 153500
1872
+ },
1873
+ {
1874
+ "epoch": 2.42,
1875
+ "learning_rate": 4.204022958298731e-05,
1876
+ "loss": 0.6094,
1877
+ "step": 154000
1878
+ },
1879
+ {
1880
+ "epoch": 2.43,
1881
+ "learning_rate": 4.201394947912834e-05,
1882
+ "loss": 0.6109,
1883
+ "step": 154500
1884
+ },
1885
+ {
1886
+ "epoch": 2.44,
1887
+ "learning_rate": 4.1987669375269375e-05,
1888
+ "loss": 0.5984,
1889
+ "step": 155000
1890
+ },
1891
+ {
1892
+ "epoch": 2.45,
1893
+ "learning_rate": 4.19613892714104e-05,
1894
+ "loss": 0.6177,
1895
+ "step": 155500
1896
+ },
1897
+ {
1898
+ "epoch": 2.45,
1899
+ "learning_rate": 4.1935109167551434e-05,
1900
+ "loss": 0.6136,
1901
+ "step": 156000
1902
+ },
1903
+ {
1904
+ "epoch": 2.46,
1905
+ "learning_rate": 4.190882906369246e-05,
1906
+ "loss": 0.6128,
1907
+ "step": 156500
1908
+ },
1909
+ {
1910
+ "epoch": 2.47,
1911
+ "learning_rate": 4.188260152004121e-05,
1912
+ "loss": 0.6172,
1913
+ "step": 157000
1914
+ },
1915
+ {
1916
+ "epoch": 2.48,
1917
+ "learning_rate": 4.185632141618224e-05,
1918
+ "loss": 0.6041,
1919
+ "step": 157500
1920
+ },
1921
+ {
1922
+ "epoch": 2.48,
1923
+ "learning_rate": 4.183004131232327e-05,
1924
+ "loss": 0.6154,
1925
+ "step": 158000
1926
+ },
1927
+ {
1928
+ "epoch": 2.49,
1929
+ "learning_rate": 4.1803761208464296e-05,
1930
+ "loss": 0.6094,
1931
+ "step": 158500
1932
+ },
1933
+ {
1934
+ "epoch": 2.5,
1935
+ "learning_rate": 4.177748110460533e-05,
1936
+ "loss": 0.6059,
1937
+ "step": 159000
1938
+ },
1939
+ {
1940
+ "epoch": 2.51,
1941
+ "learning_rate": 4.175120100074636e-05,
1942
+ "loss": 0.6089,
1943
+ "step": 159500
1944
+ },
1945
+ {
1946
+ "epoch": 2.52,
1947
+ "learning_rate": 4.1724973457095107e-05,
1948
+ "loss": 0.6122,
1949
+ "step": 160000
1950
+ },
1951
+ {
1952
+ "epoch": 2.52,
1953
+ "learning_rate": 4.169874591344385e-05,
1954
+ "loss": 0.6093,
1955
+ "step": 160500
1956
+ },
1957
+ {
1958
+ "epoch": 2.53,
1959
+ "learning_rate": 4.1672465809584884e-05,
1960
+ "loss": 0.6174,
1961
+ "step": 161000
1962
+ },
1963
+ {
1964
+ "epoch": 2.54,
1965
+ "learning_rate": 4.164618570572591e-05,
1966
+ "loss": 0.6301,
1967
+ "step": 161500
1968
+ },
1969
+ {
1970
+ "epoch": 2.55,
1971
+ "learning_rate": 4.161990560186694e-05,
1972
+ "loss": 0.6204,
1973
+ "step": 162000
1974
+ },
1975
+ {
1976
+ "epoch": 2.56,
1977
+ "learning_rate": 4.159362549800797e-05,
1978
+ "loss": 0.6207,
1979
+ "step": 162500
1980
+ },
1981
+ {
1982
+ "epoch": 2.56,
1983
+ "learning_rate": 4.1567345394149e-05,
1984
+ "loss": 0.6102,
1985
+ "step": 163000
1986
+ },
1987
+ {
1988
+ "epoch": 2.57,
1989
+ "learning_rate": 4.154106529029003e-05,
1990
+ "loss": 0.6088,
1991
+ "step": 163500
1992
+ },
1993
+ {
1994
+ "epoch": 2.58,
1995
+ "learning_rate": 4.151478518643106e-05,
1996
+ "loss": 0.6231,
1997
+ "step": 164000
1998
+ },
1999
+ {
2000
+ "epoch": 2.59,
2001
+ "learning_rate": 4.1488505082572087e-05,
2002
+ "loss": 0.6195,
2003
+ "step": 164500
2004
+ },
2005
+ {
2006
+ "epoch": 2.59,
2007
+ "learning_rate": 4.146222497871311e-05,
2008
+ "loss": 0.6214,
2009
+ "step": 165000
2010
+ },
2011
+ {
2012
+ "epoch": 2.6,
2013
+ "learning_rate": 4.1435997435061864e-05,
2014
+ "loss": 0.624,
2015
+ "step": 165500
2016
+ },
2017
+ {
2018
+ "epoch": 2.61,
2019
+ "learning_rate": 4.14097173312029e-05,
2020
+ "loss": 0.6218,
2021
+ "step": 166000
2022
+ },
2023
+ {
2024
+ "epoch": 2.62,
2025
+ "learning_rate": 4.138343722734392e-05,
2026
+ "loss": 0.6084,
2027
+ "step": 166500
2028
+ },
2029
+ {
2030
+ "epoch": 2.63,
2031
+ "learning_rate": 4.135715712348495e-05,
2032
+ "loss": 0.6155,
2033
+ "step": 167000
2034
+ },
2035
+ {
2036
+ "epoch": 2.63,
2037
+ "learning_rate": 4.133087701962599e-05,
2038
+ "loss": 0.6153,
2039
+ "step": 167500
2040
+ },
2041
+ {
2042
+ "epoch": 2.64,
2043
+ "learning_rate": 4.1304596915767015e-05,
2044
+ "loss": 0.6224,
2045
+ "step": 168000
2046
+ },
2047
+ {
2048
+ "epoch": 2.65,
2049
+ "learning_rate": 4.127831681190804e-05,
2050
+ "loss": 0.6132,
2051
+ "step": 168500
2052
+ },
2053
+ {
2054
+ "epoch": 2.66,
2055
+ "learning_rate": 4.125203670804907e-05,
2056
+ "loss": 0.6136,
2057
+ "step": 169000
2058
+ },
2059
+ {
2060
+ "epoch": 2.67,
2061
+ "learning_rate": 4.12257566041901e-05,
2062
+ "loss": 0.6097,
2063
+ "step": 169500
2064
+ },
2065
+ {
2066
+ "epoch": 2.67,
2067
+ "learning_rate": 4.119952906053885e-05,
2068
+ "loss": 0.6226,
2069
+ "step": 170000
2070
+ },
2071
+ {
2072
+ "epoch": 2.68,
2073
+ "learning_rate": 4.117324895667988e-05,
2074
+ "loss": 0.6152,
2075
+ "step": 170500
2076
+ },
2077
+ {
2078
+ "epoch": 2.69,
2079
+ "learning_rate": 4.114696885282091e-05,
2080
+ "loss": 0.6188,
2081
+ "step": 171000
2082
+ },
2083
+ {
2084
+ "epoch": 2.7,
2085
+ "learning_rate": 4.112074130916966e-05,
2086
+ "loss": 0.6199,
2087
+ "step": 171500
2088
+ },
2089
+ {
2090
+ "epoch": 2.7,
2091
+ "learning_rate": 4.109446120531069e-05,
2092
+ "loss": 0.626,
2093
+ "step": 172000
2094
+ },
2095
+ {
2096
+ "epoch": 2.71,
2097
+ "learning_rate": 4.106823366165943e-05,
2098
+ "loss": 0.6256,
2099
+ "step": 172500
2100
+ },
2101
+ {
2102
+ "epoch": 2.72,
2103
+ "learning_rate": 4.104195355780046e-05,
2104
+ "loss": 0.6169,
2105
+ "step": 173000
2106
+ },
2107
+ {
2108
+ "epoch": 2.73,
2109
+ "learning_rate": 4.10156734539415e-05,
2110
+ "loss": 0.6167,
2111
+ "step": 173500
2112
+ },
2113
+ {
2114
+ "epoch": 2.74,
2115
+ "learning_rate": 4.0989393350082524e-05,
2116
+ "loss": 0.6147,
2117
+ "step": 174000
2118
+ },
2119
+ {
2120
+ "epoch": 2.74,
2121
+ "learning_rate": 4.096311324622355e-05,
2122
+ "loss": 0.621,
2123
+ "step": 174500
2124
+ },
2125
+ {
2126
+ "epoch": 2.75,
2127
+ "learning_rate": 4.0936833142364576e-05,
2128
+ "loss": 0.6063,
2129
+ "step": 175000
2130
+ },
2131
+ {
2132
+ "epoch": 2.76,
2133
+ "learning_rate": 4.091055303850561e-05,
2134
+ "loss": 0.6216,
2135
+ "step": 175500
2136
+ },
2137
+ {
2138
+ "epoch": 2.77,
2139
+ "learning_rate": 4.088427293464664e-05,
2140
+ "loss": 0.6307,
2141
+ "step": 176000
2142
+ },
2143
+ {
2144
+ "epoch": 2.78,
2145
+ "learning_rate": 4.085799283078767e-05,
2146
+ "loss": 0.6313,
2147
+ "step": 176500
2148
+ },
2149
+ {
2150
+ "epoch": 2.78,
2151
+ "learning_rate": 4.08317127269287e-05,
2152
+ "loss": 0.6286,
2153
+ "step": 177000
2154
+ },
2155
+ {
2156
+ "epoch": 2.79,
2157
+ "learning_rate": 4.0805432623069726e-05,
2158
+ "loss": 0.6109,
2159
+ "step": 177500
2160
+ },
2161
+ {
2162
+ "epoch": 2.8,
2163
+ "learning_rate": 4.077915251921076e-05,
2164
+ "loss": 0.6175,
2165
+ "step": 178000
2166
+ },
2167
+ {
2168
+ "epoch": 2.81,
2169
+ "learning_rate": 4.075287241535179e-05,
2170
+ "loss": 0.6211,
2171
+ "step": 178500
2172
+ },
2173
+ {
2174
+ "epoch": 2.82,
2175
+ "learning_rate": 4.072659231149282e-05,
2176
+ "loss": 0.6227,
2177
+ "step": 179000
2178
+ },
2179
+ {
2180
+ "epoch": 2.82,
2181
+ "learning_rate": 4.0700312207633844e-05,
2182
+ "loss": 0.6254,
2183
+ "step": 179500
2184
+ },
2185
+ {
2186
+ "epoch": 2.83,
2187
+ "learning_rate": 4.0674032103774877e-05,
2188
+ "loss": 0.6212,
2189
+ "step": 180000
2190
+ },
2191
+ {
2192
+ "epoch": 2.84,
2193
+ "learning_rate": 4.064775199991591e-05,
2194
+ "loss": 0.6257,
2195
+ "step": 180500
2196
+ },
2197
+ {
2198
+ "epoch": 2.85,
2199
+ "learning_rate": 4.0621471896056935e-05,
2200
+ "loss": 0.6255,
2201
+ "step": 181000
2202
+ },
2203
+ {
2204
+ "epoch": 2.85,
2205
+ "learning_rate": 4.059519179219797e-05,
2206
+ "loss": 0.622,
2207
+ "step": 181500
2208
+ },
2209
+ {
2210
+ "epoch": 2.86,
2211
+ "learning_rate": 4.0568911688338994e-05,
2212
+ "loss": 0.6278,
2213
+ "step": 182000
2214
+ },
2215
+ {
2216
+ "epoch": 2.87,
2217
+ "learning_rate": 4.0542684144687746e-05,
2218
+ "loss": 0.6298,
2219
+ "step": 182500
2220
+ },
2221
+ {
2222
+ "epoch": 2.88,
2223
+ "learning_rate": 4.051640404082877e-05,
2224
+ "loss": 0.6197,
2225
+ "step": 183000
2226
+ },
2227
+ {
2228
+ "epoch": 2.89,
2229
+ "learning_rate": 4.0490176497177517e-05,
2230
+ "loss": 0.6169,
2231
+ "step": 183500
2232
+ },
2233
+ {
2234
+ "epoch": 2.89,
2235
+ "learning_rate": 4.046389639331855e-05,
2236
+ "loss": 0.6148,
2237
+ "step": 184000
2238
+ },
2239
+ {
2240
+ "epoch": 2.9,
2241
+ "learning_rate": 4.043761628945958e-05,
2242
+ "loss": 0.6153,
2243
+ "step": 184500
2244
+ },
2245
+ {
2246
+ "epoch": 2.91,
2247
+ "learning_rate": 4.041133618560061e-05,
2248
+ "loss": 0.622,
2249
+ "step": 185000
2250
+ },
2251
+ {
2252
+ "epoch": 2.92,
2253
+ "learning_rate": 4.0385056081741634e-05,
2254
+ "loss": 0.6291,
2255
+ "step": 185500
2256
+ },
2257
+ {
2258
+ "epoch": 2.93,
2259
+ "learning_rate": 4.035882853809038e-05,
2260
+ "loss": 0.6172,
2261
+ "step": 186000
2262
+ },
2263
+ {
2264
+ "epoch": 2.93,
2265
+ "learning_rate": 4.033254843423142e-05,
2266
+ "loss": 0.6156,
2267
+ "step": 186500
2268
+ },
2269
+ {
2270
+ "epoch": 2.94,
2271
+ "learning_rate": 4.0306268330372445e-05,
2272
+ "loss": 0.6274,
2273
+ "step": 187000
2274
+ },
2275
+ {
2276
+ "epoch": 2.95,
2277
+ "learning_rate": 4.028004078672119e-05,
2278
+ "loss": 0.6319,
2279
+ "step": 187500
2280
+ },
2281
+ {
2282
+ "epoch": 2.96,
2283
+ "learning_rate": 4.025381324306994e-05,
2284
+ "loss": 0.6218,
2285
+ "step": 188000
2286
+ },
2287
+ {
2288
+ "epoch": 2.96,
2289
+ "learning_rate": 4.022753313921097e-05,
2290
+ "loss": 0.6184,
2291
+ "step": 188500
2292
+ },
2293
+ {
2294
+ "epoch": 2.97,
2295
+ "learning_rate": 4.0201253035352e-05,
2296
+ "loss": 0.6275,
2297
+ "step": 189000
2298
+ },
2299
+ {
2300
+ "epoch": 2.98,
2301
+ "learning_rate": 4.0174972931493026e-05,
2302
+ "loss": 0.6218,
2303
+ "step": 189500
2304
+ },
2305
+ {
2306
+ "epoch": 2.99,
2307
+ "learning_rate": 4.014869282763405e-05,
2308
+ "loss": 0.6179,
2309
+ "step": 190000
2310
+ },
2311
+ {
2312
+ "epoch": 3.0,
2313
+ "learning_rate": 4.01224652839828e-05,
2314
+ "loss": 0.6294,
2315
+ "step": 190500
2316
+ },
2317
+ {
2318
+ "epoch": 3.0,
2319
+ "eval_bleu": 55.7536,
2320
+ "eval_gen_len": 15.2192,
2321
+ "eval_loss": 0.6636257171630859,
2322
+ "eval_runtime": 9606.0302,
2323
+ "eval_samples_per_second": 13.439,
2324
+ "eval_steps_per_second": 1.68,
2325
+ "step": 190759
2326
+ }
2327
+ ],
2328
+ "logging_steps": 500,
2329
+ "max_steps": 953790,
2330
+ "num_input_tokens_seen": 0,
2331
+ "num_train_epochs": 15,
2332
+ "save_steps": 500,
2333
+ "total_flos": 6.614409455331181e+18,
2334
+ "train_batch_size": 2,
2335
+ "trial_name": null,
2336
+ "trial_params": null
2337
+ }
checkpoint-190759/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:348f748cc2c1841db000c69cf5b90e94af304038cbe239207bfe033a7791efcc
3
+ size 4856
checkpoint-317932/config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/mbart-large-50-many-to-many-mmt",
3
+ "_num_labels": 3,
4
+ "activation_dropout": 0.0,
5
+ "activation_function": "relu",
6
+ "add_bias_logits": false,
7
+ "add_final_layer_norm": true,
8
+ "architectures": [
9
+ "MBartForConditionalGeneration"
10
+ ],
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 0,
13
+ "classif_dropout": 0.0,
14
+ "classifier_dropout": 0.0,
15
+ "d_model": 1024,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_ffn_dim": 4096,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 12,
20
+ "decoder_start_token_id": 2,
21
+ "dropout": 0.1,
22
+ "early_stopping": true,
23
+ "encoder_attention_heads": 16,
24
+ "encoder_ffn_dim": 4096,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 2,
28
+ "forced_bos_token_id": 250014,
29
+ "forced_eos_token_id": 2,
30
+ "gradient_checkpointing": false,
31
+ "id2label": {
32
+ "0": "LABEL_0",
33
+ "1": "LABEL_1",
34
+ "2": "LABEL_2"
35
+ },
36
+ "init_std": 0.02,
37
+ "is_encoder_decoder": true,
38
+ "label2id": {
39
+ "LABEL_0": 0,
40
+ "LABEL_1": 1,
41
+ "LABEL_2": 2
42
+ },
43
+ "max_length": 200,
44
+ "max_position_embeddings": 1024,
45
+ "model_type": "mbart",
46
+ "normalize_before": true,
47
+ "normalize_embedding": true,
48
+ "num_beams": 5,
49
+ "num_hidden_layers": 12,
50
+ "output_past": true,
51
+ "pad_token_id": 1,
52
+ "scale_embedding": true,
53
+ "static_position_embeddings": false,
54
+ "tokenizer_class": "MBart50Tokenizer",
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.36.2",
57
+ "use_cache": true,
58
+ "vocab_size": 250054
59
+ }
checkpoint-317932/generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 0,
3
+ "decoder_start_token_id": 2,
4
+ "early_stopping": true,
5
+ "eos_token_id": 2,
6
+ "forced_bos_token_id": 250014,
7
+ "forced_eos_token_id": 2,
8
+ "max_length": 200,
9
+ "num_beams": 5,
10
+ "pad_token_id": 1,
11
+ "transformers_version": "4.36.2"
12
+ }
checkpoint-317932/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6237dae17b7dfd125338db2d4df9b963281af2e9c74854e3fb77a2842a99437b
3
+ size 2444578688
checkpoint-317932/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ff59e705a7d6dc9ac818173bdd66be89c6d108e4fbed389a382b163ac49b2a3
3
+ size 4887473903
checkpoint-317932/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2eda4fd9b68dfcf1bda1ce9e5a19170f3b7e8f41802814523904baa824d3b7d1
3
+ size 15024
checkpoint-317932/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:047266d450adf33dc201bcdba87c7a4916969f2566b9791ce97897d8fdc50604
3
+ size 15024
checkpoint-317932/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0e9cc1a47d078153206e10df224efead7f4c671da92d18eab50b53498b3709f
3
+ size 15024
checkpoint-317932/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ad3b839e2c54d6c7b6f531e54b10932030f07ea12b5f9872ca6aed6a5ede2b4
3
+ size 15024
checkpoint-317932/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9de3a6f311eb999b2f36380fe9f09d8ba53ddb5792881ae1ab696964fed580b0
3
+ size 1064
checkpoint-317932/sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
checkpoint-317932/special_tokens_map.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "ar_AR",
4
+ "cs_CZ",
5
+ "de_DE",
6
+ "en_XX",
7
+ "es_XX",
8
+ "et_EE",
9
+ "fi_FI",
10
+ "fr_XX",
11
+ "gu_IN",
12
+ "hi_IN",
13
+ "it_IT",
14
+ "ja_XX",
15
+ "kk_KZ",
16
+ "ko_KR",
17
+ "lt_LT",
18
+ "lv_LV",
19
+ "my_MM",
20
+ "ne_NP",
21
+ "nl_XX",
22
+ "ro_RO",
23
+ "ru_RU",
24
+ "si_LK",
25
+ "tr_TR",
26
+ "vi_VN",
27
+ "zh_CN",
28
+ "af_ZA",
29
+ "az_AZ",
30
+ "bn_IN",
31
+ "fa_IR",
32
+ "he_IL",
33
+ "hr_HR",
34
+ "id_ID",
35
+ "ka_GE",
36
+ "km_KH",
37
+ "mk_MK",
38
+ "ml_IN",
39
+ "mn_MN",
40
+ "mr_IN",
41
+ "pl_PL",
42
+ "ps_AF",
43
+ "pt_XX",
44
+ "sv_SE",
45
+ "sw_KE",
46
+ "ta_IN",
47
+ "te_IN",
48
+ "th_TH",
49
+ "tl_XX",
50
+ "uk_UA",
51
+ "ur_PK",
52
+ "xh_ZA",
53
+ "gl_ES",
54
+ "sl_SI"
55
+ ],
56
+ "bos_token": "<s>",
57
+ "cls_token": "<s>",
58
+ "eos_token": "</s>",
59
+ "mask_token": {
60
+ "content": "<mask>",
61
+ "lstrip": true,
62
+ "normalized": true,
63
+ "rstrip": false,
64
+ "single_word": false
65
+ },
66
+ "pad_token": "<pad>",
67
+ "sep_token": "</s>",
68
+ "unk_token": "<unk>"
69
+ }
checkpoint-317932/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd7eee7b2b672fc40148ee5d62e7dcc8473f2ac9630ead56f0791e9eb15d1f01
3
+ size 17110107
checkpoint-317932/tokenizer_config.json ADDED
@@ -0,0 +1,528 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "ar_AR",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "250002": {
44
+ "content": "cs_CZ",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "250003": {
52
+ "content": "de_DE",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "250004": {
60
+ "content": "en_XX",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "250005": {
68
+ "content": "es_XX",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ },
75
+ "250006": {
76
+ "content": "et_EE",
77
+ "lstrip": false,
78
+ "normalized": false,
79
+ "rstrip": false,
80
+ "single_word": false,
81
+ "special": true
82
+ },
83
+ "250007": {
84
+ "content": "fi_FI",
85
+ "lstrip": false,
86
+ "normalized": false,
87
+ "rstrip": false,
88
+ "single_word": false,
89
+ "special": true
90
+ },
91
+ "250008": {
92
+ "content": "fr_XX",
93
+ "lstrip": false,
94
+ "normalized": false,
95
+ "rstrip": false,
96
+ "single_word": false,
97
+ "special": true
98
+ },
99
+ "250009": {
100
+ "content": "gu_IN",
101
+ "lstrip": false,
102
+ "normalized": false,
103
+ "rstrip": false,
104
+ "single_word": false,
105
+ "special": true
106
+ },
107
+ "250010": {
108
+ "content": "hi_IN",
109
+ "lstrip": false,
110
+ "normalized": false,
111
+ "rstrip": false,
112
+ "single_word": false,
113
+ "special": true
114
+ },
115
+ "250011": {
116
+ "content": "it_IT",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false,
121
+ "special": true
122
+ },
123
+ "250012": {
124
+ "content": "ja_XX",
125
+ "lstrip": false,
126
+ "normalized": false,
127
+ "rstrip": false,
128
+ "single_word": false,
129
+ "special": true
130
+ },
131
+ "250013": {
132
+ "content": "kk_KZ",
133
+ "lstrip": false,
134
+ "normalized": false,
135
+ "rstrip": false,
136
+ "single_word": false,
137
+ "special": true
138
+ },
139
+ "250014": {
140
+ "content": "ko_KR",
141
+ "lstrip": false,
142
+ "normalized": false,
143
+ "rstrip": false,
144
+ "single_word": false,
145
+ "special": true
146
+ },
147
+ "250015": {
148
+ "content": "lt_LT",
149
+ "lstrip": false,
150
+ "normalized": false,
151
+ "rstrip": false,
152
+ "single_word": false,
153
+ "special": true
154
+ },
155
+ "250016": {
156
+ "content": "lv_LV",
157
+ "lstrip": false,
158
+ "normalized": false,
159
+ "rstrip": false,
160
+ "single_word": false,
161
+ "special": true
162
+ },
163
+ "250017": {
164
+ "content": "my_MM",
165
+ "lstrip": false,
166
+ "normalized": false,
167
+ "rstrip": false,
168
+ "single_word": false,
169
+ "special": true
170
+ },
171
+ "250018": {
172
+ "content": "ne_NP",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false,
177
+ "special": true
178
+ },
179
+ "250019": {
180
+ "content": "nl_XX",
181
+ "lstrip": false,
182
+ "normalized": false,
183
+ "rstrip": false,
184
+ "single_word": false,
185
+ "special": true
186
+ },
187
+ "250020": {
188
+ "content": "ro_RO",
189
+ "lstrip": false,
190
+ "normalized": false,
191
+ "rstrip": false,
192
+ "single_word": false,
193
+ "special": true
194
+ },
195
+ "250021": {
196
+ "content": "ru_RU",
197
+ "lstrip": false,
198
+ "normalized": false,
199
+ "rstrip": false,
200
+ "single_word": false,
201
+ "special": true
202
+ },
203
+ "250022": {
204
+ "content": "si_LK",
205
+ "lstrip": false,
206
+ "normalized": false,
207
+ "rstrip": false,
208
+ "single_word": false,
209
+ "special": true
210
+ },
211
+ "250023": {
212
+ "content": "tr_TR",
213
+ "lstrip": false,
214
+ "normalized": false,
215
+ "rstrip": false,
216
+ "single_word": false,
217
+ "special": true
218
+ },
219
+ "250024": {
220
+ "content": "vi_VN",
221
+ "lstrip": false,
222
+ "normalized": false,
223
+ "rstrip": false,
224
+ "single_word": false,
225
+ "special": true
226
+ },
227
+ "250025": {
228
+ "content": "zh_CN",
229
+ "lstrip": false,
230
+ "normalized": false,
231
+ "rstrip": false,
232
+ "single_word": false,
233
+ "special": true
234
+ },
235
+ "250026": {
236
+ "content": "af_ZA",
237
+ "lstrip": false,
238
+ "normalized": false,
239
+ "rstrip": false,
240
+ "single_word": false,
241
+ "special": true
242
+ },
243
+ "250027": {
244
+ "content": "az_AZ",
245
+ "lstrip": false,
246
+ "normalized": false,
247
+ "rstrip": false,
248
+ "single_word": false,
249
+ "special": true
250
+ },
251
+ "250028": {
252
+ "content": "bn_IN",
253
+ "lstrip": false,
254
+ "normalized": false,
255
+ "rstrip": false,
256
+ "single_word": false,
257
+ "special": true
258
+ },
259
+ "250029": {
260
+ "content": "fa_IR",
261
+ "lstrip": false,
262
+ "normalized": false,
263
+ "rstrip": false,
264
+ "single_word": false,
265
+ "special": true
266
+ },
267
+ "250030": {
268
+ "content": "he_IL",
269
+ "lstrip": false,
270
+ "normalized": false,
271
+ "rstrip": false,
272
+ "single_word": false,
273
+ "special": true
274
+ },
275
+ "250031": {
276
+ "content": "hr_HR",
277
+ "lstrip": false,
278
+ "normalized": false,
279
+ "rstrip": false,
280
+ "single_word": false,
281
+ "special": true
282
+ },
283
+ "250032": {
284
+ "content": "id_ID",
285
+ "lstrip": false,
286
+ "normalized": false,
287
+ "rstrip": false,
288
+ "single_word": false,
289
+ "special": true
290
+ },
291
+ "250033": {
292
+ "content": "ka_GE",
293
+ "lstrip": false,
294
+ "normalized": false,
295
+ "rstrip": false,
296
+ "single_word": false,
297
+ "special": true
298
+ },
299
+ "250034": {
300
+ "content": "km_KH",
301
+ "lstrip": false,
302
+ "normalized": false,
303
+ "rstrip": false,
304
+ "single_word": false,
305
+ "special": true
306
+ },
307
+ "250035": {
308
+ "content": "mk_MK",
309
+ "lstrip": false,
310
+ "normalized": false,
311
+ "rstrip": false,
312
+ "single_word": false,
313
+ "special": true
314
+ },
315
+ "250036": {
316
+ "content": "ml_IN",
317
+ "lstrip": false,
318
+ "normalized": false,
319
+ "rstrip": false,
320
+ "single_word": false,
321
+ "special": true
322
+ },
323
+ "250037": {
324
+ "content": "mn_MN",
325
+ "lstrip": false,
326
+ "normalized": false,
327
+ "rstrip": false,
328
+ "single_word": false,
329
+ "special": true
330
+ },
331
+ "250038": {
332
+ "content": "mr_IN",
333
+ "lstrip": false,
334
+ "normalized": false,
335
+ "rstrip": false,
336
+ "single_word": false,
337
+ "special": true
338
+ },
339
+ "250039": {
340
+ "content": "pl_PL",
341
+ "lstrip": false,
342
+ "normalized": false,
343
+ "rstrip": false,
344
+ "single_word": false,
345
+ "special": true
346
+ },
347
+ "250040": {
348
+ "content": "ps_AF",
349
+ "lstrip": false,
350
+ "normalized": false,
351
+ "rstrip": false,
352
+ "single_word": false,
353
+ "special": true
354
+ },
355
+ "250041": {
356
+ "content": "pt_XX",
357
+ "lstrip": false,
358
+ "normalized": false,
359
+ "rstrip": false,
360
+ "single_word": false,
361
+ "special": true
362
+ },
363
+ "250042": {
364
+ "content": "sv_SE",
365
+ "lstrip": false,
366
+ "normalized": false,
367
+ "rstrip": false,
368
+ "single_word": false,
369
+ "special": true
370
+ },
371
+ "250043": {
372
+ "content": "sw_KE",
373
+ "lstrip": false,
374
+ "normalized": false,
375
+ "rstrip": false,
376
+ "single_word": false,
377
+ "special": true
378
+ },
379
+ "250044": {
380
+ "content": "ta_IN",
381
+ "lstrip": false,
382
+ "normalized": false,
383
+ "rstrip": false,
384
+ "single_word": false,
385
+ "special": true
386
+ },
387
+ "250045": {
388
+ "content": "te_IN",
389
+ "lstrip": false,
390
+ "normalized": false,
391
+ "rstrip": false,
392
+ "single_word": false,
393
+ "special": true
394
+ },
395
+ "250046": {
396
+ "content": "th_TH",
397
+ "lstrip": false,
398
+ "normalized": false,
399
+ "rstrip": false,
400
+ "single_word": false,
401
+ "special": true
402
+ },
403
+ "250047": {
404
+ "content": "tl_XX",
405
+ "lstrip": false,
406
+ "normalized": false,
407
+ "rstrip": false,
408
+ "single_word": false,
409
+ "special": true
410
+ },
411
+ "250048": {
412
+ "content": "uk_UA",
413
+ "lstrip": false,
414
+ "normalized": false,
415
+ "rstrip": false,
416
+ "single_word": false,
417
+ "special": true
418
+ },
419
+ "250049": {
420
+ "content": "ur_PK",
421
+ "lstrip": false,
422
+ "normalized": false,
423
+ "rstrip": false,
424
+ "single_word": false,
425
+ "special": true
426
+ },
427
+ "250050": {
428
+ "content": "xh_ZA",
429
+ "lstrip": false,
430
+ "normalized": false,
431
+ "rstrip": false,
432
+ "single_word": false,
433
+ "special": true
434
+ },
435
+ "250051": {
436
+ "content": "gl_ES",
437
+ "lstrip": false,
438
+ "normalized": false,
439
+ "rstrip": false,
440
+ "single_word": false,
441
+ "special": true
442
+ },
443
+ "250052": {
444
+ "content": "sl_SI",
445
+ "lstrip": false,
446
+ "normalized": false,
447
+ "rstrip": false,
448
+ "single_word": false,
449
+ "special": true
450
+ },
451
+ "250053": {
452
+ "content": "<mask>",
453
+ "lstrip": true,
454
+ "normalized": true,
455
+ "rstrip": false,
456
+ "single_word": false,
457
+ "special": true
458
+ }
459
+ },
460
+ "additional_special_tokens": [
461
+ "ar_AR",
462
+ "cs_CZ",
463
+ "de_DE",
464
+ "en_XX",
465
+ "es_XX",
466
+ "et_EE",
467
+ "fi_FI",
468
+ "fr_XX",
469
+ "gu_IN",
470
+ "hi_IN",
471
+ "it_IT",
472
+ "ja_XX",
473
+ "kk_KZ",
474
+ "ko_KR",
475
+ "lt_LT",
476
+ "lv_LV",
477
+ "my_MM",
478
+ "ne_NP",
479
+ "nl_XX",
480
+ "ro_RO",
481
+ "ru_RU",
482
+ "si_LK",
483
+ "tr_TR",
484
+ "vi_VN",
485
+ "zh_CN",
486
+ "af_ZA",
487
+ "az_AZ",
488
+ "bn_IN",
489
+ "fa_IR",
490
+ "he_IL",
491
+ "hr_HR",
492
+ "id_ID",
493
+ "ka_GE",
494
+ "km_KH",
495
+ "mk_MK",
496
+ "ml_IN",
497
+ "mn_MN",
498
+ "mr_IN",
499
+ "pl_PL",
500
+ "ps_AF",
501
+ "pt_XX",
502
+ "sv_SE",
503
+ "sw_KE",
504
+ "ta_IN",
505
+ "te_IN",
506
+ "th_TH",
507
+ "tl_XX",
508
+ "uk_UA",
509
+ "ur_PK",
510
+ "xh_ZA",
511
+ "gl_ES",
512
+ "sl_SI"
513
+ ],
514
+ "bos_token": "<s>",
515
+ "clean_up_tokenization_spaces": true,
516
+ "cls_token": "<s>",
517
+ "eos_token": "</s>",
518
+ "language_codes": "ML50",
519
+ "mask_token": "<mask>",
520
+ "model_max_length": 1000000000000000019884624838656,
521
+ "pad_token": "<pad>",
522
+ "sep_token": "</s>",
523
+ "sp_model_kwargs": {},
524
+ "src_lang": "en_XX",
525
+ "tgt_lang": "ko_KR",
526
+ "tokenizer_class": "MBart50Tokenizer",
527
+ "unk_token": "<unk>"
528
+ }
checkpoint-317932/trainer_state.json ADDED
@@ -0,0 +1,3881 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.6636257171630859,
3
+ "best_model_checkpoint": "./enko_mbartLarge_100p_run1/checkpoint-190759",
4
+ "epoch": 4.999992136695682,
5
+ "eval_steps": 500,
6
+ "global_step": 317932,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 9.940000000000001e-06,
14
+ "loss": 1.6525,
15
+ "step": 500
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.994e-05,
20
+ "loss": 1.3306,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 2.994e-05,
26
+ "loss": 1.3005,
27
+ "step": 1500
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 3.994e-05,
32
+ "loss": 1.2871,
33
+ "step": 2000
34
+ },
35
+ {
36
+ "epoch": 0.04,
37
+ "learning_rate": 4.9940000000000006e-05,
38
+ "loss": 1.306,
39
+ "step": 2500
40
+ },
41
+ {
42
+ "epoch": 0.05,
43
+ "learning_rate": 4.9973930136971904e-05,
44
+ "loss": 1.2876,
45
+ "step": 3000
46
+ },
47
+ {
48
+ "epoch": 0.06,
49
+ "learning_rate": 4.994770259332065e-05,
50
+ "loss": 1.2609,
51
+ "step": 3500
52
+ },
53
+ {
54
+ "epoch": 0.06,
55
+ "learning_rate": 4.992142248946168e-05,
56
+ "loss": 1.244,
57
+ "step": 4000
58
+ },
59
+ {
60
+ "epoch": 0.07,
61
+ "learning_rate": 4.9895142385602714e-05,
62
+ "loss": 1.2051,
63
+ "step": 4500
64
+ },
65
+ {
66
+ "epoch": 0.08,
67
+ "learning_rate": 4.986886228174374e-05,
68
+ "loss": 1.1997,
69
+ "step": 5000
70
+ },
71
+ {
72
+ "epoch": 0.09,
73
+ "learning_rate": 4.9842582177884766e-05,
74
+ "loss": 1.1854,
75
+ "step": 5500
76
+ },
77
+ {
78
+ "epoch": 0.09,
79
+ "learning_rate": 4.98163020740258e-05,
80
+ "loss": 1.1666,
81
+ "step": 6000
82
+ },
83
+ {
84
+ "epoch": 0.1,
85
+ "learning_rate": 4.9790021970166825e-05,
86
+ "loss": 1.1596,
87
+ "step": 6500
88
+ },
89
+ {
90
+ "epoch": 0.11,
91
+ "learning_rate": 4.976374186630786e-05,
92
+ "loss": 1.1559,
93
+ "step": 7000
94
+ },
95
+ {
96
+ "epoch": 0.12,
97
+ "learning_rate": 4.973746176244889e-05,
98
+ "loss": 1.1555,
99
+ "step": 7500
100
+ },
101
+ {
102
+ "epoch": 0.13,
103
+ "learning_rate": 4.9711181658589916e-05,
104
+ "loss": 1.1458,
105
+ "step": 8000
106
+ },
107
+ {
108
+ "epoch": 0.13,
109
+ "learning_rate": 4.968495411493866e-05,
110
+ "loss": 1.1098,
111
+ "step": 8500
112
+ },
113
+ {
114
+ "epoch": 0.14,
115
+ "learning_rate": 4.9658674011079694e-05,
116
+ "loss": 1.1169,
117
+ "step": 9000
118
+ },
119
+ {
120
+ "epoch": 0.15,
121
+ "learning_rate": 4.963239390722073e-05,
122
+ "loss": 1.114,
123
+ "step": 9500
124
+ },
125
+ {
126
+ "epoch": 0.16,
127
+ "learning_rate": 4.960616636356947e-05,
128
+ "loss": 1.1099,
129
+ "step": 10000
130
+ },
131
+ {
132
+ "epoch": 0.17,
133
+ "learning_rate": 4.95798862597105e-05,
134
+ "loss": 1.0846,
135
+ "step": 10500
136
+ },
137
+ {
138
+ "epoch": 0.17,
139
+ "learning_rate": 4.955360615585153e-05,
140
+ "loss": 1.1021,
141
+ "step": 11000
142
+ },
143
+ {
144
+ "epoch": 0.18,
145
+ "learning_rate": 4.952732605199256e-05,
146
+ "loss": 1.0739,
147
+ "step": 11500
148
+ },
149
+ {
150
+ "epoch": 0.19,
151
+ "learning_rate": 4.950104594813359e-05,
152
+ "loss": 1.0657,
153
+ "step": 12000
154
+ },
155
+ {
156
+ "epoch": 0.2,
157
+ "learning_rate": 4.9474765844274615e-05,
158
+ "loss": 1.0641,
159
+ "step": 12500
160
+ },
161
+ {
162
+ "epoch": 0.2,
163
+ "learning_rate": 4.944848574041565e-05,
164
+ "loss": 1.0696,
165
+ "step": 13000
166
+ },
167
+ {
168
+ "epoch": 0.21,
169
+ "learning_rate": 4.94222581967644e-05,
170
+ "loss": 1.0457,
171
+ "step": 13500
172
+ },
173
+ {
174
+ "epoch": 0.22,
175
+ "learning_rate": 4.9396030653113144e-05,
176
+ "loss": 1.0573,
177
+ "step": 14000
178
+ },
179
+ {
180
+ "epoch": 0.23,
181
+ "learning_rate": 4.936975054925417e-05,
182
+ "loss": 1.0302,
183
+ "step": 14500
184
+ },
185
+ {
186
+ "epoch": 0.24,
187
+ "learning_rate": 4.93434704453952e-05,
188
+ "loss": 1.0347,
189
+ "step": 15000
190
+ },
191
+ {
192
+ "epoch": 0.24,
193
+ "learning_rate": 4.9317190341536236e-05,
194
+ "loss": 1.0304,
195
+ "step": 15500
196
+ },
197
+ {
198
+ "epoch": 0.25,
199
+ "learning_rate": 4.929096279788498e-05,
200
+ "loss": 1.0454,
201
+ "step": 16000
202
+ },
203
+ {
204
+ "epoch": 0.26,
205
+ "learning_rate": 4.926468269402601e-05,
206
+ "loss": 1.0406,
207
+ "step": 16500
208
+ },
209
+ {
210
+ "epoch": 0.27,
211
+ "learning_rate": 4.923840259016704e-05,
212
+ "loss": 1.0306,
213
+ "step": 17000
214
+ },
215
+ {
216
+ "epoch": 0.28,
217
+ "learning_rate": 4.921212248630807e-05,
218
+ "loss": 1.0306,
219
+ "step": 17500
220
+ },
221
+ {
222
+ "epoch": 0.28,
223
+ "learning_rate": 4.91858423824491e-05,
224
+ "loss": 1.0237,
225
+ "step": 18000
226
+ },
227
+ {
228
+ "epoch": 0.29,
229
+ "learning_rate": 4.9159562278590124e-05,
230
+ "loss": 1.0113,
231
+ "step": 18500
232
+ },
233
+ {
234
+ "epoch": 0.3,
235
+ "learning_rate": 4.913328217473116e-05,
236
+ "loss": 1.0101,
237
+ "step": 19000
238
+ },
239
+ {
240
+ "epoch": 0.31,
241
+ "learning_rate": 4.910700207087219e-05,
242
+ "loss": 0.9928,
243
+ "step": 19500
244
+ },
245
+ {
246
+ "epoch": 0.31,
247
+ "learning_rate": 4.9080721967013216e-05,
248
+ "loss": 0.9942,
249
+ "step": 20000
250
+ },
251
+ {
252
+ "epoch": 0.32,
253
+ "learning_rate": 4.905444186315424e-05,
254
+ "loss": 1.0043,
255
+ "step": 20500
256
+ },
257
+ {
258
+ "epoch": 0.33,
259
+ "learning_rate": 4.9028161759295275e-05,
260
+ "loss": 1.0028,
261
+ "step": 21000
262
+ },
263
+ {
264
+ "epoch": 0.34,
265
+ "learning_rate": 4.900188165543631e-05,
266
+ "loss": 0.9883,
267
+ "step": 21500
268
+ },
269
+ {
270
+ "epoch": 0.35,
271
+ "learning_rate": 4.897565411178505e-05,
272
+ "loss": 1.0023,
273
+ "step": 22000
274
+ },
275
+ {
276
+ "epoch": 0.35,
277
+ "learning_rate": 4.89494265681338e-05,
278
+ "loss": 0.9935,
279
+ "step": 22500
280
+ },
281
+ {
282
+ "epoch": 0.36,
283
+ "learning_rate": 4.892314646427483e-05,
284
+ "loss": 0.9928,
285
+ "step": 23000
286
+ },
287
+ {
288
+ "epoch": 0.37,
289
+ "learning_rate": 4.889686636041586e-05,
290
+ "loss": 0.9948,
291
+ "step": 23500
292
+ },
293
+ {
294
+ "epoch": 0.38,
295
+ "learning_rate": 4.887058625655689e-05,
296
+ "loss": 0.974,
297
+ "step": 24000
298
+ },
299
+ {
300
+ "epoch": 0.39,
301
+ "learning_rate": 4.8844306152697915e-05,
302
+ "loss": 0.9837,
303
+ "step": 24500
304
+ },
305
+ {
306
+ "epoch": 0.39,
307
+ "learning_rate": 4.881802604883895e-05,
308
+ "loss": 0.9751,
309
+ "step": 25000
310
+ },
311
+ {
312
+ "epoch": 0.4,
313
+ "learning_rate": 4.879174594497998e-05,
314
+ "loss": 0.9702,
315
+ "step": 25500
316
+ },
317
+ {
318
+ "epoch": 0.41,
319
+ "learning_rate": 4.8765465841121006e-05,
320
+ "loss": 0.9717,
321
+ "step": 26000
322
+ },
323
+ {
324
+ "epoch": 0.42,
325
+ "learning_rate": 4.873918573726204e-05,
326
+ "loss": 0.9724,
327
+ "step": 26500
328
+ },
329
+ {
330
+ "epoch": 0.42,
331
+ "learning_rate": 4.8712905633403065e-05,
332
+ "loss": 0.9692,
333
+ "step": 27000
334
+ },
335
+ {
336
+ "epoch": 0.43,
337
+ "learning_rate": 4.868667808975182e-05,
338
+ "loss": 0.9535,
339
+ "step": 27500
340
+ },
341
+ {
342
+ "epoch": 0.44,
343
+ "learning_rate": 4.866039798589284e-05,
344
+ "loss": 0.9641,
345
+ "step": 28000
346
+ },
347
+ {
348
+ "epoch": 0.45,
349
+ "learning_rate": 4.863417044224159e-05,
350
+ "loss": 0.9774,
351
+ "step": 28500
352
+ },
353
+ {
354
+ "epoch": 0.46,
355
+ "learning_rate": 4.860789033838262e-05,
356
+ "loss": 0.9671,
357
+ "step": 29000
358
+ },
359
+ {
360
+ "epoch": 0.46,
361
+ "learning_rate": 4.858161023452365e-05,
362
+ "loss": 0.9458,
363
+ "step": 29500
364
+ },
365
+ {
366
+ "epoch": 0.47,
367
+ "learning_rate": 4.855533013066468e-05,
368
+ "loss": 0.9606,
369
+ "step": 30000
370
+ },
371
+ {
372
+ "epoch": 0.48,
373
+ "learning_rate": 4.8529050026805705e-05,
374
+ "loss": 0.9557,
375
+ "step": 30500
376
+ },
377
+ {
378
+ "epoch": 0.49,
379
+ "learning_rate": 4.850276992294674e-05,
380
+ "loss": 0.9664,
381
+ "step": 31000
382
+ },
383
+ {
384
+ "epoch": 0.5,
385
+ "learning_rate": 4.8476489819087764e-05,
386
+ "loss": 0.9427,
387
+ "step": 31500
388
+ },
389
+ {
390
+ "epoch": 0.5,
391
+ "learning_rate": 4.84502097152288e-05,
392
+ "loss": 0.957,
393
+ "step": 32000
394
+ },
395
+ {
396
+ "epoch": 0.51,
397
+ "learning_rate": 4.842392961136983e-05,
398
+ "loss": 0.9466,
399
+ "step": 32500
400
+ },
401
+ {
402
+ "epoch": 0.52,
403
+ "learning_rate": 4.8397649507510856e-05,
404
+ "loss": 0.9262,
405
+ "step": 33000
406
+ },
407
+ {
408
+ "epoch": 0.53,
409
+ "learning_rate": 4.837136940365188e-05,
410
+ "loss": 0.9524,
411
+ "step": 33500
412
+ },
413
+ {
414
+ "epoch": 0.53,
415
+ "learning_rate": 4.8345089299792914e-05,
416
+ "loss": 0.9303,
417
+ "step": 34000
418
+ },
419
+ {
420
+ "epoch": 0.54,
421
+ "learning_rate": 4.831880919593395e-05,
422
+ "loss": 0.9355,
423
+ "step": 34500
424
+ },
425
+ {
426
+ "epoch": 0.55,
427
+ "learning_rate": 4.829252909207497e-05,
428
+ "loss": 0.9463,
429
+ "step": 35000
430
+ },
431
+ {
432
+ "epoch": 0.56,
433
+ "learning_rate": 4.8266248988216006e-05,
434
+ "loss": 0.9333,
435
+ "step": 35500
436
+ },
437
+ {
438
+ "epoch": 0.57,
439
+ "learning_rate": 4.823996888435703e-05,
440
+ "loss": 0.9293,
441
+ "step": 36000
442
+ },
443
+ {
444
+ "epoch": 0.57,
445
+ "learning_rate": 4.8213688780498065e-05,
446
+ "loss": 0.9272,
447
+ "step": 36500
448
+ },
449
+ {
450
+ "epoch": 0.58,
451
+ "learning_rate": 4.818740867663909e-05,
452
+ "loss": 0.928,
453
+ "step": 37000
454
+ },
455
+ {
456
+ "epoch": 0.59,
457
+ "learning_rate": 4.8161128572780124e-05,
458
+ "loss": 0.9346,
459
+ "step": 37500
460
+ },
461
+ {
462
+ "epoch": 0.6,
463
+ "learning_rate": 4.813484846892115e-05,
464
+ "loss": 0.9361,
465
+ "step": 38000
466
+ },
467
+ {
468
+ "epoch": 0.61,
469
+ "learning_rate": 4.81086209252699e-05,
470
+ "loss": 0.9197,
471
+ "step": 38500
472
+ },
473
+ {
474
+ "epoch": 0.61,
475
+ "learning_rate": 4.8082393381618646e-05,
476
+ "loss": 0.9289,
477
+ "step": 39000
478
+ },
479
+ {
480
+ "epoch": 0.62,
481
+ "learning_rate": 4.805611327775967e-05,
482
+ "loss": 0.9232,
483
+ "step": 39500
484
+ },
485
+ {
486
+ "epoch": 0.63,
487
+ "learning_rate": 4.8029833173900705e-05,
488
+ "loss": 0.9134,
489
+ "step": 40000
490
+ },
491
+ {
492
+ "epoch": 0.64,
493
+ "learning_rate": 4.800355307004174e-05,
494
+ "loss": 0.9232,
495
+ "step": 40500
496
+ },
497
+ {
498
+ "epoch": 0.64,
499
+ "learning_rate": 4.7977272966182764e-05,
500
+ "loss": 0.9315,
501
+ "step": 41000
502
+ },
503
+ {
504
+ "epoch": 0.65,
505
+ "learning_rate": 4.7950992862323796e-05,
506
+ "loss": 0.9152,
507
+ "step": 41500
508
+ },
509
+ {
510
+ "epoch": 0.66,
511
+ "learning_rate": 4.792471275846482e-05,
512
+ "loss": 0.9179,
513
+ "step": 42000
514
+ },
515
+ {
516
+ "epoch": 0.67,
517
+ "learning_rate": 4.7898485214813574e-05,
518
+ "loss": 0.914,
519
+ "step": 42500
520
+ },
521
+ {
522
+ "epoch": 0.68,
523
+ "learning_rate": 4.787225767116232e-05,
524
+ "loss": 0.9078,
525
+ "step": 43000
526
+ },
527
+ {
528
+ "epoch": 0.68,
529
+ "learning_rate": 4.7845977567303345e-05,
530
+ "loss": 0.9129,
531
+ "step": 43500
532
+ },
533
+ {
534
+ "epoch": 0.69,
535
+ "learning_rate": 4.781969746344438e-05,
536
+ "loss": 0.899,
537
+ "step": 44000
538
+ },
539
+ {
540
+ "epoch": 0.7,
541
+ "learning_rate": 4.779341735958541e-05,
542
+ "loss": 0.9003,
543
+ "step": 44500
544
+ },
545
+ {
546
+ "epoch": 0.71,
547
+ "learning_rate": 4.7767137255726436e-05,
548
+ "loss": 0.8976,
549
+ "step": 45000
550
+ },
551
+ {
552
+ "epoch": 0.72,
553
+ "learning_rate": 4.774085715186747e-05,
554
+ "loss": 0.9007,
555
+ "step": 45500
556
+ },
557
+ {
558
+ "epoch": 0.72,
559
+ "learning_rate": 4.7714577048008495e-05,
560
+ "loss": 0.9053,
561
+ "step": 46000
562
+ },
563
+ {
564
+ "epoch": 0.73,
565
+ "learning_rate": 4.768829694414952e-05,
566
+ "loss": 0.892,
567
+ "step": 46500
568
+ },
569
+ {
570
+ "epoch": 0.74,
571
+ "learning_rate": 4.766201684029056e-05,
572
+ "loss": 0.9049,
573
+ "step": 47000
574
+ },
575
+ {
576
+ "epoch": 0.75,
577
+ "learning_rate": 4.763573673643159e-05,
578
+ "loss": 0.8988,
579
+ "step": 47500
580
+ },
581
+ {
582
+ "epoch": 0.75,
583
+ "learning_rate": 4.760956175298805e-05,
584
+ "loss": 0.9086,
585
+ "step": 48000
586
+ },
587
+ {
588
+ "epoch": 0.76,
589
+ "learning_rate": 4.758328164912908e-05,
590
+ "loss": 0.903,
591
+ "step": 48500
592
+ },
593
+ {
594
+ "epoch": 0.77,
595
+ "learning_rate": 4.755700154527011e-05,
596
+ "loss": 0.895,
597
+ "step": 49000
598
+ },
599
+ {
600
+ "epoch": 0.78,
601
+ "learning_rate": 4.753072144141114e-05,
602
+ "loss": 0.8982,
603
+ "step": 49500
604
+ },
605
+ {
606
+ "epoch": 0.79,
607
+ "learning_rate": 4.750444133755217e-05,
608
+ "loss": 0.9047,
609
+ "step": 50000
610
+ },
611
+ {
612
+ "epoch": 0.79,
613
+ "learning_rate": 4.7478161233693194e-05,
614
+ "loss": 0.9088,
615
+ "step": 50500
616
+ },
617
+ {
618
+ "epoch": 0.8,
619
+ "learning_rate": 4.745188112983423e-05,
620
+ "loss": 0.892,
621
+ "step": 51000
622
+ },
623
+ {
624
+ "epoch": 0.81,
625
+ "learning_rate": 4.742560102597526e-05,
626
+ "loss": 0.89,
627
+ "step": 51500
628
+ },
629
+ {
630
+ "epoch": 0.82,
631
+ "learning_rate": 4.7399320922116286e-05,
632
+ "loss": 0.8909,
633
+ "step": 52000
634
+ },
635
+ {
636
+ "epoch": 0.83,
637
+ "learning_rate": 4.737309337846503e-05,
638
+ "loss": 0.8792,
639
+ "step": 52500
640
+ },
641
+ {
642
+ "epoch": 0.83,
643
+ "learning_rate": 4.734681327460606e-05,
644
+ "loss": 0.8783,
645
+ "step": 53000
646
+ },
647
+ {
648
+ "epoch": 0.84,
649
+ "learning_rate": 4.7320533170747096e-05,
650
+ "loss": 0.8869,
651
+ "step": 53500
652
+ },
653
+ {
654
+ "epoch": 0.85,
655
+ "learning_rate": 4.729425306688812e-05,
656
+ "loss": 0.8873,
657
+ "step": 54000
658
+ },
659
+ {
660
+ "epoch": 0.86,
661
+ "learning_rate": 4.726797296302915e-05,
662
+ "loss": 0.8927,
663
+ "step": 54500
664
+ },
665
+ {
666
+ "epoch": 0.86,
667
+ "learning_rate": 4.72417454193779e-05,
668
+ "loss": 0.8817,
669
+ "step": 55000
670
+ },
671
+ {
672
+ "epoch": 0.87,
673
+ "learning_rate": 4.721546531551893e-05,
674
+ "loss": 0.8746,
675
+ "step": 55500
676
+ },
677
+ {
678
+ "epoch": 0.88,
679
+ "learning_rate": 4.718918521165996e-05,
680
+ "loss": 0.8845,
681
+ "step": 56000
682
+ },
683
+ {
684
+ "epoch": 0.89,
685
+ "learning_rate": 4.7162905107800984e-05,
686
+ "loss": 0.8768,
687
+ "step": 56500
688
+ },
689
+ {
690
+ "epoch": 0.9,
691
+ "learning_rate": 4.713662500394202e-05,
692
+ "loss": 0.8834,
693
+ "step": 57000
694
+ },
695
+ {
696
+ "epoch": 0.9,
697
+ "learning_rate": 4.711034490008305e-05,
698
+ "loss": 0.8769,
699
+ "step": 57500
700
+ },
701
+ {
702
+ "epoch": 0.91,
703
+ "learning_rate": 4.7084064796224076e-05,
704
+ "loss": 0.865,
705
+ "step": 58000
706
+ },
707
+ {
708
+ "epoch": 0.92,
709
+ "learning_rate": 4.705783725257282e-05,
710
+ "loss": 0.8786,
711
+ "step": 58500
712
+ },
713
+ {
714
+ "epoch": 0.93,
715
+ "learning_rate": 4.7031557148713854e-05,
716
+ "loss": 0.885,
717
+ "step": 59000
718
+ },
719
+ {
720
+ "epoch": 0.94,
721
+ "learning_rate": 4.7005277044854886e-05,
722
+ "loss": 0.8804,
723
+ "step": 59500
724
+ },
725
+ {
726
+ "epoch": 0.94,
727
+ "learning_rate": 4.697899694099591e-05,
728
+ "loss": 0.8834,
729
+ "step": 60000
730
+ },
731
+ {
732
+ "epoch": 0.95,
733
+ "learning_rate": 4.6952716837136945e-05,
734
+ "loss": 0.8698,
735
+ "step": 60500
736
+ },
737
+ {
738
+ "epoch": 0.96,
739
+ "learning_rate": 4.692643673327797e-05,
740
+ "loss": 0.8605,
741
+ "step": 61000
742
+ },
743
+ {
744
+ "epoch": 0.97,
745
+ "learning_rate": 4.6900156629419004e-05,
746
+ "loss": 0.8721,
747
+ "step": 61500
748
+ },
749
+ {
750
+ "epoch": 0.98,
751
+ "learning_rate": 4.687387652556003e-05,
752
+ "loss": 0.8914,
753
+ "step": 62000
754
+ },
755
+ {
756
+ "epoch": 0.98,
757
+ "learning_rate": 4.684759642170106e-05,
758
+ "loss": 0.8743,
759
+ "step": 62500
760
+ },
761
+ {
762
+ "epoch": 0.99,
763
+ "learning_rate": 4.682131631784209e-05,
764
+ "loss": 0.8677,
765
+ "step": 63000
766
+ },
767
+ {
768
+ "epoch": 1.0,
769
+ "learning_rate": 4.679514133439856e-05,
770
+ "loss": 0.8673,
771
+ "step": 63500
772
+ },
773
+ {
774
+ "epoch": 1.0,
775
+ "eval_bleu": 52.6771,
776
+ "eval_gen_len": 15.2057,
777
+ "eval_loss": 0.7307217121124268,
778
+ "eval_runtime": 9605.9098,
779
+ "eval_samples_per_second": 13.439,
780
+ "eval_steps_per_second": 1.68,
781
+ "step": 63586
782
+ },
783
+ {
784
+ "epoch": 1.01,
785
+ "learning_rate": 4.6768913790747304e-05,
786
+ "loss": 0.7651,
787
+ "step": 64000
788
+ },
789
+ {
790
+ "epoch": 1.01,
791
+ "learning_rate": 4.674263368688833e-05,
792
+ "loss": 0.7382,
793
+ "step": 64500
794
+ },
795
+ {
796
+ "epoch": 1.02,
797
+ "learning_rate": 4.671635358302936e-05,
798
+ "loss": 0.7193,
799
+ "step": 65000
800
+ },
801
+ {
802
+ "epoch": 1.03,
803
+ "learning_rate": 4.6690073479170396e-05,
804
+ "loss": 0.7344,
805
+ "step": 65500
806
+ },
807
+ {
808
+ "epoch": 1.04,
809
+ "learning_rate": 4.666384593551914e-05,
810
+ "loss": 0.7191,
811
+ "step": 66000
812
+ },
813
+ {
814
+ "epoch": 1.05,
815
+ "learning_rate": 4.6637565831660166e-05,
816
+ "loss": 0.7228,
817
+ "step": 66500
818
+ },
819
+ {
820
+ "epoch": 1.05,
821
+ "learning_rate": 4.66112857278012e-05,
822
+ "loss": 0.7285,
823
+ "step": 67000
824
+ },
825
+ {
826
+ "epoch": 1.06,
827
+ "learning_rate": 4.658500562394223e-05,
828
+ "loss": 0.7373,
829
+ "step": 67500
830
+ },
831
+ {
832
+ "epoch": 1.07,
833
+ "learning_rate": 4.655872552008326e-05,
834
+ "loss": 0.7209,
835
+ "step": 68000
836
+ },
837
+ {
838
+ "epoch": 1.08,
839
+ "learning_rate": 4.6532445416224284e-05,
840
+ "loss": 0.7345,
841
+ "step": 68500
842
+ },
843
+ {
844
+ "epoch": 1.09,
845
+ "learning_rate": 4.6506217872573036e-05,
846
+ "loss": 0.7348,
847
+ "step": 69000
848
+ },
849
+ {
850
+ "epoch": 1.09,
851
+ "learning_rate": 4.647993776871407e-05,
852
+ "loss": 0.7391,
853
+ "step": 69500
854
+ },
855
+ {
856
+ "epoch": 1.1,
857
+ "learning_rate": 4.6453657664855094e-05,
858
+ "loss": 0.7367,
859
+ "step": 70000
860
+ },
861
+ {
862
+ "epoch": 1.11,
863
+ "learning_rate": 4.642737756099612e-05,
864
+ "loss": 0.7282,
865
+ "step": 70500
866
+ },
867
+ {
868
+ "epoch": 1.12,
869
+ "learning_rate": 4.640109745713715e-05,
870
+ "loss": 0.7379,
871
+ "step": 71000
872
+ },
873
+ {
874
+ "epoch": 1.12,
875
+ "learning_rate": 4.6374817353278186e-05,
876
+ "loss": 0.7313,
877
+ "step": 71500
878
+ },
879
+ {
880
+ "epoch": 1.13,
881
+ "learning_rate": 4.634853724941921e-05,
882
+ "loss": 0.7399,
883
+ "step": 72000
884
+ },
885
+ {
886
+ "epoch": 1.14,
887
+ "learning_rate": 4.6322257145560245e-05,
888
+ "loss": 0.731,
889
+ "step": 72500
890
+ },
891
+ {
892
+ "epoch": 1.15,
893
+ "learning_rate": 4.629597704170127e-05,
894
+ "loss": 0.7435,
895
+ "step": 73000
896
+ },
897
+ {
898
+ "epoch": 1.16,
899
+ "learning_rate": 4.62696969378423e-05,
900
+ "loss": 0.7361,
901
+ "step": 73500
902
+ },
903
+ {
904
+ "epoch": 1.16,
905
+ "learning_rate": 4.624341683398333e-05,
906
+ "loss": 0.7315,
907
+ "step": 74000
908
+ },
909
+ {
910
+ "epoch": 1.17,
911
+ "learning_rate": 4.621713673012436e-05,
912
+ "loss": 0.7404,
913
+ "step": 74500
914
+ },
915
+ {
916
+ "epoch": 1.18,
917
+ "learning_rate": 4.619090918647311e-05,
918
+ "loss": 0.7399,
919
+ "step": 75000
920
+ },
921
+ {
922
+ "epoch": 1.19,
923
+ "learning_rate": 4.616468164282186e-05,
924
+ "loss": 0.7378,
925
+ "step": 75500
926
+ },
927
+ {
928
+ "epoch": 1.2,
929
+ "learning_rate": 4.6138401538962885e-05,
930
+ "loss": 0.7391,
931
+ "step": 76000
932
+ },
933
+ {
934
+ "epoch": 1.2,
935
+ "learning_rate": 4.611212143510391e-05,
936
+ "loss": 0.7499,
937
+ "step": 76500
938
+ },
939
+ {
940
+ "epoch": 1.21,
941
+ "learning_rate": 4.6085841331244944e-05,
942
+ "loss": 0.7537,
943
+ "step": 77000
944
+ },
945
+ {
946
+ "epoch": 1.22,
947
+ "learning_rate": 4.605956122738597e-05,
948
+ "loss": 0.7423,
949
+ "step": 77500
950
+ },
951
+ {
952
+ "epoch": 1.23,
953
+ "learning_rate": 4.6033281123527e-05,
954
+ "loss": 0.7453,
955
+ "step": 78000
956
+ },
957
+ {
958
+ "epoch": 1.23,
959
+ "learning_rate": 4.6007001019668035e-05,
960
+ "loss": 0.7439,
961
+ "step": 78500
962
+ },
963
+ {
964
+ "epoch": 1.24,
965
+ "learning_rate": 4.598072091580906e-05,
966
+ "loss": 0.7446,
967
+ "step": 79000
968
+ },
969
+ {
970
+ "epoch": 1.25,
971
+ "learning_rate": 4.595444081195009e-05,
972
+ "loss": 0.7412,
973
+ "step": 79500
974
+ },
975
+ {
976
+ "epoch": 1.26,
977
+ "learning_rate": 4.592821326829884e-05,
978
+ "loss": 0.7399,
979
+ "step": 80000
980
+ },
981
+ {
982
+ "epoch": 1.27,
983
+ "learning_rate": 4.590193316443987e-05,
984
+ "loss": 0.7302,
985
+ "step": 80500
986
+ },
987
+ {
988
+ "epoch": 1.27,
989
+ "learning_rate": 4.58756530605809e-05,
990
+ "loss": 0.7431,
991
+ "step": 81000
992
+ },
993
+ {
994
+ "epoch": 1.28,
995
+ "learning_rate": 4.5849372956721924e-05,
996
+ "loss": 0.7437,
997
+ "step": 81500
998
+ },
999
+ {
1000
+ "epoch": 1.29,
1001
+ "learning_rate": 4.5823145413070675e-05,
1002
+ "loss": 0.7443,
1003
+ "step": 82000
1004
+ },
1005
+ {
1006
+ "epoch": 1.3,
1007
+ "learning_rate": 4.579686530921171e-05,
1008
+ "loss": 0.7496,
1009
+ "step": 82500
1010
+ },
1011
+ {
1012
+ "epoch": 1.31,
1013
+ "learning_rate": 4.577063776556045e-05,
1014
+ "loss": 0.7395,
1015
+ "step": 83000
1016
+ },
1017
+ {
1018
+ "epoch": 1.31,
1019
+ "learning_rate": 4.574435766170148e-05,
1020
+ "loss": 0.7307,
1021
+ "step": 83500
1022
+ },
1023
+ {
1024
+ "epoch": 1.32,
1025
+ "learning_rate": 4.571807755784251e-05,
1026
+ "loss": 0.7467,
1027
+ "step": 84000
1028
+ },
1029
+ {
1030
+ "epoch": 1.33,
1031
+ "learning_rate": 4.5691797453983544e-05,
1032
+ "loss": 0.7474,
1033
+ "step": 84500
1034
+ },
1035
+ {
1036
+ "epoch": 1.34,
1037
+ "learning_rate": 4.566551735012457e-05,
1038
+ "loss": 0.7352,
1039
+ "step": 85000
1040
+ },
1041
+ {
1042
+ "epoch": 1.34,
1043
+ "learning_rate": 4.5639237246265596e-05,
1044
+ "loss": 0.7377,
1045
+ "step": 85500
1046
+ },
1047
+ {
1048
+ "epoch": 1.35,
1049
+ "learning_rate": 4.561295714240663e-05,
1050
+ "loss": 0.7389,
1051
+ "step": 86000
1052
+ },
1053
+ {
1054
+ "epoch": 1.36,
1055
+ "learning_rate": 4.558667703854766e-05,
1056
+ "loss": 0.7372,
1057
+ "step": 86500
1058
+ },
1059
+ {
1060
+ "epoch": 1.37,
1061
+ "learning_rate": 4.556039693468869e-05,
1062
+ "loss": 0.7488,
1063
+ "step": 87000
1064
+ },
1065
+ {
1066
+ "epoch": 1.38,
1067
+ "learning_rate": 4.553416939103743e-05,
1068
+ "loss": 0.7303,
1069
+ "step": 87500
1070
+ },
1071
+ {
1072
+ "epoch": 1.38,
1073
+ "learning_rate": 4.5507889287178466e-05,
1074
+ "loss": 0.7243,
1075
+ "step": 88000
1076
+ },
1077
+ {
1078
+ "epoch": 1.39,
1079
+ "learning_rate": 4.54816091833195e-05,
1080
+ "loss": 0.7318,
1081
+ "step": 88500
1082
+ },
1083
+ {
1084
+ "epoch": 1.4,
1085
+ "learning_rate": 4.5455329079460524e-05,
1086
+ "loss": 0.7421,
1087
+ "step": 89000
1088
+ },
1089
+ {
1090
+ "epoch": 1.41,
1091
+ "learning_rate": 4.542904897560155e-05,
1092
+ "loss": 0.7441,
1093
+ "step": 89500
1094
+ },
1095
+ {
1096
+ "epoch": 1.42,
1097
+ "learning_rate": 4.540276887174258e-05,
1098
+ "loss": 0.7388,
1099
+ "step": 90000
1100
+ },
1101
+ {
1102
+ "epoch": 1.42,
1103
+ "learning_rate": 4.537659388829905e-05,
1104
+ "loss": 0.7381,
1105
+ "step": 90500
1106
+ },
1107
+ {
1108
+ "epoch": 1.43,
1109
+ "learning_rate": 4.535031378444008e-05,
1110
+ "loss": 0.7402,
1111
+ "step": 91000
1112
+ },
1113
+ {
1114
+ "epoch": 1.44,
1115
+ "learning_rate": 4.5324033680581106e-05,
1116
+ "loss": 0.7377,
1117
+ "step": 91500
1118
+ },
1119
+ {
1120
+ "epoch": 1.45,
1121
+ "learning_rate": 4.529775357672214e-05,
1122
+ "loss": 0.7474,
1123
+ "step": 92000
1124
+ },
1125
+ {
1126
+ "epoch": 1.45,
1127
+ "learning_rate": 4.527152603307088e-05,
1128
+ "loss": 0.7473,
1129
+ "step": 92500
1130
+ },
1131
+ {
1132
+ "epoch": 1.46,
1133
+ "learning_rate": 4.5245245929211916e-05,
1134
+ "loss": 0.7345,
1135
+ "step": 93000
1136
+ },
1137
+ {
1138
+ "epoch": 1.47,
1139
+ "learning_rate": 4.521896582535294e-05,
1140
+ "loss": 0.7366,
1141
+ "step": 93500
1142
+ },
1143
+ {
1144
+ "epoch": 1.48,
1145
+ "learning_rate": 4.5192685721493975e-05,
1146
+ "loss": 0.7518,
1147
+ "step": 94000
1148
+ },
1149
+ {
1150
+ "epoch": 1.49,
1151
+ "learning_rate": 4.516640561763501e-05,
1152
+ "loss": 0.7454,
1153
+ "step": 94500
1154
+ },
1155
+ {
1156
+ "epoch": 1.49,
1157
+ "learning_rate": 4.514017807398375e-05,
1158
+ "loss": 0.7432,
1159
+ "step": 95000
1160
+ },
1161
+ {
1162
+ "epoch": 1.5,
1163
+ "learning_rate": 4.511389797012478e-05,
1164
+ "loss": 0.7448,
1165
+ "step": 95500
1166
+ },
1167
+ {
1168
+ "epoch": 1.51,
1169
+ "learning_rate": 4.508761786626581e-05,
1170
+ "loss": 0.7415,
1171
+ "step": 96000
1172
+ },
1173
+ {
1174
+ "epoch": 1.52,
1175
+ "learning_rate": 4.5061337762406844e-05,
1176
+ "loss": 0.7358,
1177
+ "step": 96500
1178
+ },
1179
+ {
1180
+ "epoch": 1.53,
1181
+ "learning_rate": 4.503505765854787e-05,
1182
+ "loss": 0.7491,
1183
+ "step": 97000
1184
+ },
1185
+ {
1186
+ "epoch": 1.53,
1187
+ "learning_rate": 4.5008777554688896e-05,
1188
+ "loss": 0.7429,
1189
+ "step": 97500
1190
+ },
1191
+ {
1192
+ "epoch": 1.54,
1193
+ "learning_rate": 4.498249745082993e-05,
1194
+ "loss": 0.748,
1195
+ "step": 98000
1196
+ },
1197
+ {
1198
+ "epoch": 1.55,
1199
+ "learning_rate": 4.4956217346970955e-05,
1200
+ "loss": 0.7467,
1201
+ "step": 98500
1202
+ },
1203
+ {
1204
+ "epoch": 1.56,
1205
+ "learning_rate": 4.492993724311199e-05,
1206
+ "loss": 0.7459,
1207
+ "step": 99000
1208
+ },
1209
+ {
1210
+ "epoch": 1.56,
1211
+ "learning_rate": 4.490365713925302e-05,
1212
+ "loss": 0.7402,
1213
+ "step": 99500
1214
+ },
1215
+ {
1216
+ "epoch": 1.57,
1217
+ "learning_rate": 4.4877377035394046e-05,
1218
+ "loss": 0.7532,
1219
+ "step": 100000
1220
+ },
1221
+ {
1222
+ "epoch": 1.58,
1223
+ "learning_rate": 4.485120205195052e-05,
1224
+ "loss": 0.7464,
1225
+ "step": 100500
1226
+ },
1227
+ {
1228
+ "epoch": 1.59,
1229
+ "learning_rate": 4.482492194809154e-05,
1230
+ "loss": 0.7441,
1231
+ "step": 101000
1232
+ },
1233
+ {
1234
+ "epoch": 1.6,
1235
+ "learning_rate": 4.479869440444029e-05,
1236
+ "loss": 0.7445,
1237
+ "step": 101500
1238
+ },
1239
+ {
1240
+ "epoch": 1.6,
1241
+ "learning_rate": 4.4772414300581314e-05,
1242
+ "loss": 0.7316,
1243
+ "step": 102000
1244
+ },
1245
+ {
1246
+ "epoch": 1.61,
1247
+ "learning_rate": 4.474613419672235e-05,
1248
+ "loss": 0.7437,
1249
+ "step": 102500
1250
+ },
1251
+ {
1252
+ "epoch": 1.62,
1253
+ "learning_rate": 4.471985409286338e-05,
1254
+ "loss": 0.7456,
1255
+ "step": 103000
1256
+ },
1257
+ {
1258
+ "epoch": 1.63,
1259
+ "learning_rate": 4.4693573989004405e-05,
1260
+ "loss": 0.7508,
1261
+ "step": 103500
1262
+ },
1263
+ {
1264
+ "epoch": 1.64,
1265
+ "learning_rate": 4.466729388514543e-05,
1266
+ "loss": 0.7436,
1267
+ "step": 104000
1268
+ },
1269
+ {
1270
+ "epoch": 1.64,
1271
+ "learning_rate": 4.4641013781286464e-05,
1272
+ "loss": 0.7413,
1273
+ "step": 104500
1274
+ },
1275
+ {
1276
+ "epoch": 1.65,
1277
+ "learning_rate": 4.46147336774275e-05,
1278
+ "loss": 0.7305,
1279
+ "step": 105000
1280
+ },
1281
+ {
1282
+ "epoch": 1.66,
1283
+ "learning_rate": 4.458845357356852e-05,
1284
+ "loss": 0.7496,
1285
+ "step": 105500
1286
+ },
1287
+ {
1288
+ "epoch": 1.67,
1289
+ "learning_rate": 4.4562173469709555e-05,
1290
+ "loss": 0.7467,
1291
+ "step": 106000
1292
+ },
1293
+ {
1294
+ "epoch": 1.67,
1295
+ "learning_rate": 4.453589336585058e-05,
1296
+ "loss": 0.7404,
1297
+ "step": 106500
1298
+ },
1299
+ {
1300
+ "epoch": 1.68,
1301
+ "learning_rate": 4.4509613261991614e-05,
1302
+ "loss": 0.74,
1303
+ "step": 107000
1304
+ },
1305
+ {
1306
+ "epoch": 1.69,
1307
+ "learning_rate": 4.448333315813265e-05,
1308
+ "loss": 0.7432,
1309
+ "step": 107500
1310
+ },
1311
+ {
1312
+ "epoch": 1.7,
1313
+ "learning_rate": 4.445705305427367e-05,
1314
+ "loss": 0.7401,
1315
+ "step": 108000
1316
+ },
1317
+ {
1318
+ "epoch": 1.71,
1319
+ "learning_rate": 4.44307729504147e-05,
1320
+ "loss": 0.7382,
1321
+ "step": 108500
1322
+ },
1323
+ {
1324
+ "epoch": 1.71,
1325
+ "learning_rate": 4.440449284655573e-05,
1326
+ "loss": 0.7352,
1327
+ "step": 109000
1328
+ },
1329
+ {
1330
+ "epoch": 1.72,
1331
+ "learning_rate": 4.4378212742696765e-05,
1332
+ "loss": 0.7293,
1333
+ "step": 109500
1334
+ },
1335
+ {
1336
+ "epoch": 1.73,
1337
+ "learning_rate": 4.435193263883779e-05,
1338
+ "loss": 0.742,
1339
+ "step": 110000
1340
+ },
1341
+ {
1342
+ "epoch": 1.74,
1343
+ "learning_rate": 4.432565253497882e-05,
1344
+ "loss": 0.7426,
1345
+ "step": 110500
1346
+ },
1347
+ {
1348
+ "epoch": 1.75,
1349
+ "learning_rate": 4.429942499132757e-05,
1350
+ "loss": 0.7428,
1351
+ "step": 111000
1352
+ },
1353
+ {
1354
+ "epoch": 1.75,
1355
+ "learning_rate": 4.42731448874686e-05,
1356
+ "loss": 0.7391,
1357
+ "step": 111500
1358
+ },
1359
+ {
1360
+ "epoch": 1.76,
1361
+ "learning_rate": 4.4246917343817346e-05,
1362
+ "loss": 0.7449,
1363
+ "step": 112000
1364
+ },
1365
+ {
1366
+ "epoch": 1.77,
1367
+ "learning_rate": 4.422063723995837e-05,
1368
+ "loss": 0.7377,
1369
+ "step": 112500
1370
+ },
1371
+ {
1372
+ "epoch": 1.78,
1373
+ "learning_rate": 4.4194357136099405e-05,
1374
+ "loss": 0.7468,
1375
+ "step": 113000
1376
+ },
1377
+ {
1378
+ "epoch": 1.78,
1379
+ "learning_rate": 4.416807703224044e-05,
1380
+ "loss": 0.7398,
1381
+ "step": 113500
1382
+ },
1383
+ {
1384
+ "epoch": 1.79,
1385
+ "learning_rate": 4.4141796928381463e-05,
1386
+ "loss": 0.7371,
1387
+ "step": 114000
1388
+ },
1389
+ {
1390
+ "epoch": 1.8,
1391
+ "learning_rate": 4.411556938473021e-05,
1392
+ "loss": 0.7371,
1393
+ "step": 114500
1394
+ },
1395
+ {
1396
+ "epoch": 1.81,
1397
+ "learning_rate": 4.4089289280871234e-05,
1398
+ "loss": 0.7443,
1399
+ "step": 115000
1400
+ },
1401
+ {
1402
+ "epoch": 1.82,
1403
+ "learning_rate": 4.4063009177012274e-05,
1404
+ "loss": 0.7375,
1405
+ "step": 115500
1406
+ },
1407
+ {
1408
+ "epoch": 1.82,
1409
+ "learning_rate": 4.40367290731533e-05,
1410
+ "loss": 0.7339,
1411
+ "step": 116000
1412
+ },
1413
+ {
1414
+ "epoch": 1.83,
1415
+ "learning_rate": 4.4010448969294326e-05,
1416
+ "loss": 0.7267,
1417
+ "step": 116500
1418
+ },
1419
+ {
1420
+ "epoch": 1.84,
1421
+ "learning_rate": 4.398416886543536e-05,
1422
+ "loss": 0.7378,
1423
+ "step": 117000
1424
+ },
1425
+ {
1426
+ "epoch": 1.85,
1427
+ "learning_rate": 4.3957888761576385e-05,
1428
+ "loss": 0.7313,
1429
+ "step": 117500
1430
+ },
1431
+ {
1432
+ "epoch": 1.86,
1433
+ "learning_rate": 4.393160865771742e-05,
1434
+ "loss": 0.739,
1435
+ "step": 118000
1436
+ },
1437
+ {
1438
+ "epoch": 1.86,
1439
+ "learning_rate": 4.390532855385845e-05,
1440
+ "loss": 0.7314,
1441
+ "step": 118500
1442
+ },
1443
+ {
1444
+ "epoch": 1.87,
1445
+ "learning_rate": 4.3879048449999476e-05,
1446
+ "loss": 0.7328,
1447
+ "step": 119000
1448
+ },
1449
+ {
1450
+ "epoch": 1.88,
1451
+ "learning_rate": 4.385282090634822e-05,
1452
+ "loss": 0.741,
1453
+ "step": 119500
1454
+ },
1455
+ {
1456
+ "epoch": 1.89,
1457
+ "learning_rate": 4.382659336269697e-05,
1458
+ "loss": 0.7419,
1459
+ "step": 120000
1460
+ },
1461
+ {
1462
+ "epoch": 1.9,
1463
+ "learning_rate": 4.3800313258838e-05,
1464
+ "loss": 0.7353,
1465
+ "step": 120500
1466
+ },
1467
+ {
1468
+ "epoch": 1.9,
1469
+ "learning_rate": 4.377403315497903e-05,
1470
+ "loss": 0.7315,
1471
+ "step": 121000
1472
+ },
1473
+ {
1474
+ "epoch": 1.91,
1475
+ "learning_rate": 4.374775305112006e-05,
1476
+ "loss": 0.7457,
1477
+ "step": 121500
1478
+ },
1479
+ {
1480
+ "epoch": 1.92,
1481
+ "learning_rate": 4.372147294726109e-05,
1482
+ "loss": 0.7442,
1483
+ "step": 122000
1484
+ },
1485
+ {
1486
+ "epoch": 1.93,
1487
+ "learning_rate": 4.369519284340212e-05,
1488
+ "loss": 0.7421,
1489
+ "step": 122500
1490
+ },
1491
+ {
1492
+ "epoch": 1.93,
1493
+ "learning_rate": 4.366891273954315e-05,
1494
+ "loss": 0.7298,
1495
+ "step": 123000
1496
+ },
1497
+ {
1498
+ "epoch": 1.94,
1499
+ "learning_rate": 4.3642632635684175e-05,
1500
+ "loss": 0.7387,
1501
+ "step": 123500
1502
+ },
1503
+ {
1504
+ "epoch": 1.95,
1505
+ "learning_rate": 4.361635253182521e-05,
1506
+ "loss": 0.7322,
1507
+ "step": 124000
1508
+ },
1509
+ {
1510
+ "epoch": 1.96,
1511
+ "learning_rate": 4.359012498817395e-05,
1512
+ "loss": 0.7308,
1513
+ "step": 124500
1514
+ },
1515
+ {
1516
+ "epoch": 1.97,
1517
+ "learning_rate": 4.3563844884314985e-05,
1518
+ "loss": 0.7491,
1519
+ "step": 125000
1520
+ },
1521
+ {
1522
+ "epoch": 1.97,
1523
+ "learning_rate": 4.353756478045601e-05,
1524
+ "loss": 0.7292,
1525
+ "step": 125500
1526
+ },
1527
+ {
1528
+ "epoch": 1.98,
1529
+ "learning_rate": 4.3511284676597044e-05,
1530
+ "loss": 0.7307,
1531
+ "step": 126000
1532
+ },
1533
+ {
1534
+ "epoch": 1.99,
1535
+ "learning_rate": 4.348500457273808e-05,
1536
+ "loss": 0.7399,
1537
+ "step": 126500
1538
+ },
1539
+ {
1540
+ "epoch": 2.0,
1541
+ "learning_rate": 4.345877702908682e-05,
1542
+ "loss": 0.7184,
1543
+ "step": 127000
1544
+ },
1545
+ {
1546
+ "epoch": 2.0,
1547
+ "eval_bleu": 55.1744,
1548
+ "eval_gen_len": 15.3621,
1549
+ "eval_loss": 0.6736727952957153,
1550
+ "eval_runtime": 9675.9055,
1551
+ "eval_samples_per_second": 13.342,
1552
+ "eval_steps_per_second": 1.668,
1553
+ "step": 127173
1554
+ },
1555
+ {
1556
+ "epoch": 2.01,
1557
+ "learning_rate": 4.343249692522785e-05,
1558
+ "loss": 0.6388,
1559
+ "step": 127500
1560
+ },
1561
+ {
1562
+ "epoch": 2.01,
1563
+ "learning_rate": 4.340621682136888e-05,
1564
+ "loss": 0.5765,
1565
+ "step": 128000
1566
+ },
1567
+ {
1568
+ "epoch": 2.02,
1569
+ "learning_rate": 4.3379936717509913e-05,
1570
+ "loss": 0.5839,
1571
+ "step": 128500
1572
+ },
1573
+ {
1574
+ "epoch": 2.03,
1575
+ "learning_rate": 4.335365661365094e-05,
1576
+ "loss": 0.5815,
1577
+ "step": 129000
1578
+ },
1579
+ {
1580
+ "epoch": 2.04,
1581
+ "learning_rate": 4.3327376509791965e-05,
1582
+ "loss": 0.5716,
1583
+ "step": 129500
1584
+ },
1585
+ {
1586
+ "epoch": 2.04,
1587
+ "learning_rate": 4.3301096405933e-05,
1588
+ "loss": 0.5826,
1589
+ "step": 130000
1590
+ },
1591
+ {
1592
+ "epoch": 2.05,
1593
+ "learning_rate": 4.327481630207403e-05,
1594
+ "loss": 0.5783,
1595
+ "step": 130500
1596
+ },
1597
+ {
1598
+ "epoch": 2.06,
1599
+ "learning_rate": 4.3248588758422776e-05,
1600
+ "loss": 0.5879,
1601
+ "step": 131000
1602
+ },
1603
+ {
1604
+ "epoch": 2.07,
1605
+ "learning_rate": 4.32223086545638e-05,
1606
+ "loss": 0.5824,
1607
+ "step": 131500
1608
+ },
1609
+ {
1610
+ "epoch": 2.08,
1611
+ "learning_rate": 4.3196028550704835e-05,
1612
+ "loss": 0.583,
1613
+ "step": 132000
1614
+ },
1615
+ {
1616
+ "epoch": 2.08,
1617
+ "learning_rate": 4.316974844684587e-05,
1618
+ "loss": 0.5797,
1619
+ "step": 132500
1620
+ },
1621
+ {
1622
+ "epoch": 2.09,
1623
+ "learning_rate": 4.3143468342986893e-05,
1624
+ "loss": 0.5963,
1625
+ "step": 133000
1626
+ },
1627
+ {
1628
+ "epoch": 2.1,
1629
+ "learning_rate": 4.311724079933564e-05,
1630
+ "loss": 0.5841,
1631
+ "step": 133500
1632
+ },
1633
+ {
1634
+ "epoch": 2.11,
1635
+ "learning_rate": 4.309096069547667e-05,
1636
+ "loss": 0.5863,
1637
+ "step": 134000
1638
+ },
1639
+ {
1640
+ "epoch": 2.12,
1641
+ "learning_rate": 4.3064680591617704e-05,
1642
+ "loss": 0.5947,
1643
+ "step": 134500
1644
+ },
1645
+ {
1646
+ "epoch": 2.12,
1647
+ "learning_rate": 4.303840048775873e-05,
1648
+ "loss": 0.5977,
1649
+ "step": 135000
1650
+ },
1651
+ {
1652
+ "epoch": 2.13,
1653
+ "learning_rate": 4.3012172944107475e-05,
1654
+ "loss": 0.5848,
1655
+ "step": 135500
1656
+ },
1657
+ {
1658
+ "epoch": 2.14,
1659
+ "learning_rate": 4.298589284024851e-05,
1660
+ "loss": 0.5972,
1661
+ "step": 136000
1662
+ },
1663
+ {
1664
+ "epoch": 2.15,
1665
+ "learning_rate": 4.295961273638954e-05,
1666
+ "loss": 0.5979,
1667
+ "step": 136500
1668
+ },
1669
+ {
1670
+ "epoch": 2.15,
1671
+ "learning_rate": 4.2933385192738285e-05,
1672
+ "loss": 0.5879,
1673
+ "step": 137000
1674
+ },
1675
+ {
1676
+ "epoch": 2.16,
1677
+ "learning_rate": 4.290710508887931e-05,
1678
+ "loss": 0.6024,
1679
+ "step": 137500
1680
+ },
1681
+ {
1682
+ "epoch": 2.17,
1683
+ "learning_rate": 4.288082498502034e-05,
1684
+ "loss": 0.5893,
1685
+ "step": 138000
1686
+ },
1687
+ {
1688
+ "epoch": 2.18,
1689
+ "learning_rate": 4.285459744136909e-05,
1690
+ "loss": 0.6006,
1691
+ "step": 138500
1692
+ },
1693
+ {
1694
+ "epoch": 2.19,
1695
+ "learning_rate": 4.282831733751012e-05,
1696
+ "loss": 0.5983,
1697
+ "step": 139000
1698
+ },
1699
+ {
1700
+ "epoch": 2.19,
1701
+ "learning_rate": 4.280203723365115e-05,
1702
+ "loss": 0.6009,
1703
+ "step": 139500
1704
+ },
1705
+ {
1706
+ "epoch": 2.2,
1707
+ "learning_rate": 4.2775757129792173e-05,
1708
+ "loss": 0.6056,
1709
+ "step": 140000
1710
+ },
1711
+ {
1712
+ "epoch": 2.21,
1713
+ "learning_rate": 4.274947702593321e-05,
1714
+ "loss": 0.5942,
1715
+ "step": 140500
1716
+ },
1717
+ {
1718
+ "epoch": 2.22,
1719
+ "learning_rate": 4.272330204248967e-05,
1720
+ "loss": 0.5949,
1721
+ "step": 141000
1722
+ },
1723
+ {
1724
+ "epoch": 2.23,
1725
+ "learning_rate": 4.26970219386307e-05,
1726
+ "loss": 0.5972,
1727
+ "step": 141500
1728
+ },
1729
+ {
1730
+ "epoch": 2.23,
1731
+ "learning_rate": 4.2670741834771735e-05,
1732
+ "loss": 0.6073,
1733
+ "step": 142000
1734
+ },
1735
+ {
1736
+ "epoch": 2.24,
1737
+ "learning_rate": 4.264446173091276e-05,
1738
+ "loss": 0.6031,
1739
+ "step": 142500
1740
+ },
1741
+ {
1742
+ "epoch": 2.25,
1743
+ "learning_rate": 4.2618181627053794e-05,
1744
+ "loss": 0.5975,
1745
+ "step": 143000
1746
+ },
1747
+ {
1748
+ "epoch": 2.26,
1749
+ "learning_rate": 4.259190152319482e-05,
1750
+ "loss": 0.5965,
1751
+ "step": 143500
1752
+ },
1753
+ {
1754
+ "epoch": 2.26,
1755
+ "learning_rate": 4.2565621419335846e-05,
1756
+ "loss": 0.5991,
1757
+ "step": 144000
1758
+ },
1759
+ {
1760
+ "epoch": 2.27,
1761
+ "learning_rate": 4.2539341315476886e-05,
1762
+ "loss": 0.597,
1763
+ "step": 144500
1764
+ },
1765
+ {
1766
+ "epoch": 2.28,
1767
+ "learning_rate": 4.251306121161791e-05,
1768
+ "loss": 0.5909,
1769
+ "step": 145000
1770
+ },
1771
+ {
1772
+ "epoch": 2.29,
1773
+ "learning_rate": 4.248678110775894e-05,
1774
+ "loss": 0.6014,
1775
+ "step": 145500
1776
+ },
1777
+ {
1778
+ "epoch": 2.3,
1779
+ "learning_rate": 4.246050100389997e-05,
1780
+ "loss": 0.6016,
1781
+ "step": 146000
1782
+ },
1783
+ {
1784
+ "epoch": 2.3,
1785
+ "learning_rate": 4.2434220900041e-05,
1786
+ "loss": 0.6034,
1787
+ "step": 146500
1788
+ },
1789
+ {
1790
+ "epoch": 2.31,
1791
+ "learning_rate": 4.240794079618203e-05,
1792
+ "loss": 0.6033,
1793
+ "step": 147000
1794
+ },
1795
+ {
1796
+ "epoch": 2.32,
1797
+ "learning_rate": 4.238166069232306e-05,
1798
+ "loss": 0.6034,
1799
+ "step": 147500
1800
+ },
1801
+ {
1802
+ "epoch": 2.33,
1803
+ "learning_rate": 4.235538058846409e-05,
1804
+ "loss": 0.6123,
1805
+ "step": 148000
1806
+ },
1807
+ {
1808
+ "epoch": 2.34,
1809
+ "learning_rate": 4.2329100484605114e-05,
1810
+ "loss": 0.6087,
1811
+ "step": 148500
1812
+ },
1813
+ {
1814
+ "epoch": 2.34,
1815
+ "learning_rate": 4.230282038074615e-05,
1816
+ "loss": 0.6092,
1817
+ "step": 149000
1818
+ },
1819
+ {
1820
+ "epoch": 2.35,
1821
+ "learning_rate": 4.227654027688718e-05,
1822
+ "loss": 0.6045,
1823
+ "step": 149500
1824
+ },
1825
+ {
1826
+ "epoch": 2.36,
1827
+ "learning_rate": 4.2250312733235925e-05,
1828
+ "loss": 0.5975,
1829
+ "step": 150000
1830
+ },
1831
+ {
1832
+ "epoch": 2.37,
1833
+ "learning_rate": 4.2224137749792395e-05,
1834
+ "loss": 0.6023,
1835
+ "step": 150500
1836
+ },
1837
+ {
1838
+ "epoch": 2.37,
1839
+ "learning_rate": 4.219785764593342e-05,
1840
+ "loss": 0.6035,
1841
+ "step": 151000
1842
+ },
1843
+ {
1844
+ "epoch": 2.38,
1845
+ "learning_rate": 4.2171630102282166e-05,
1846
+ "loss": 0.5928,
1847
+ "step": 151500
1848
+ },
1849
+ {
1850
+ "epoch": 2.39,
1851
+ "learning_rate": 4.214534999842319e-05,
1852
+ "loss": 0.6116,
1853
+ "step": 152000
1854
+ },
1855
+ {
1856
+ "epoch": 2.4,
1857
+ "learning_rate": 4.2119069894564225e-05,
1858
+ "loss": 0.6186,
1859
+ "step": 152500
1860
+ },
1861
+ {
1862
+ "epoch": 2.41,
1863
+ "learning_rate": 4.209278979070526e-05,
1864
+ "loss": 0.6025,
1865
+ "step": 153000
1866
+ },
1867
+ {
1868
+ "epoch": 2.41,
1869
+ "learning_rate": 4.2066509686846283e-05,
1870
+ "loss": 0.5997,
1871
+ "step": 153500
1872
+ },
1873
+ {
1874
+ "epoch": 2.42,
1875
+ "learning_rate": 4.204022958298731e-05,
1876
+ "loss": 0.6094,
1877
+ "step": 154000
1878
+ },
1879
+ {
1880
+ "epoch": 2.43,
1881
+ "learning_rate": 4.201394947912834e-05,
1882
+ "loss": 0.6109,
1883
+ "step": 154500
1884
+ },
1885
+ {
1886
+ "epoch": 2.44,
1887
+ "learning_rate": 4.1987669375269375e-05,
1888
+ "loss": 0.5984,
1889
+ "step": 155000
1890
+ },
1891
+ {
1892
+ "epoch": 2.45,
1893
+ "learning_rate": 4.19613892714104e-05,
1894
+ "loss": 0.6177,
1895
+ "step": 155500
1896
+ },
1897
+ {
1898
+ "epoch": 2.45,
1899
+ "learning_rate": 4.1935109167551434e-05,
1900
+ "loss": 0.6136,
1901
+ "step": 156000
1902
+ },
1903
+ {
1904
+ "epoch": 2.46,
1905
+ "learning_rate": 4.190882906369246e-05,
1906
+ "loss": 0.6128,
1907
+ "step": 156500
1908
+ },
1909
+ {
1910
+ "epoch": 2.47,
1911
+ "learning_rate": 4.188260152004121e-05,
1912
+ "loss": 0.6172,
1913
+ "step": 157000
1914
+ },
1915
+ {
1916
+ "epoch": 2.48,
1917
+ "learning_rate": 4.185632141618224e-05,
1918
+ "loss": 0.6041,
1919
+ "step": 157500
1920
+ },
1921
+ {
1922
+ "epoch": 2.48,
1923
+ "learning_rate": 4.183004131232327e-05,
1924
+ "loss": 0.6154,
1925
+ "step": 158000
1926
+ },
1927
+ {
1928
+ "epoch": 2.49,
1929
+ "learning_rate": 4.1803761208464296e-05,
1930
+ "loss": 0.6094,
1931
+ "step": 158500
1932
+ },
1933
+ {
1934
+ "epoch": 2.5,
1935
+ "learning_rate": 4.177748110460533e-05,
1936
+ "loss": 0.6059,
1937
+ "step": 159000
1938
+ },
1939
+ {
1940
+ "epoch": 2.51,
1941
+ "learning_rate": 4.175120100074636e-05,
1942
+ "loss": 0.6089,
1943
+ "step": 159500
1944
+ },
1945
+ {
1946
+ "epoch": 2.52,
1947
+ "learning_rate": 4.1724973457095107e-05,
1948
+ "loss": 0.6122,
1949
+ "step": 160000
1950
+ },
1951
+ {
1952
+ "epoch": 2.52,
1953
+ "learning_rate": 4.169874591344385e-05,
1954
+ "loss": 0.6093,
1955
+ "step": 160500
1956
+ },
1957
+ {
1958
+ "epoch": 2.53,
1959
+ "learning_rate": 4.1672465809584884e-05,
1960
+ "loss": 0.6174,
1961
+ "step": 161000
1962
+ },
1963
+ {
1964
+ "epoch": 2.54,
1965
+ "learning_rate": 4.164618570572591e-05,
1966
+ "loss": 0.6301,
1967
+ "step": 161500
1968
+ },
1969
+ {
1970
+ "epoch": 2.55,
1971
+ "learning_rate": 4.161990560186694e-05,
1972
+ "loss": 0.6204,
1973
+ "step": 162000
1974
+ },
1975
+ {
1976
+ "epoch": 2.56,
1977
+ "learning_rate": 4.159362549800797e-05,
1978
+ "loss": 0.6207,
1979
+ "step": 162500
1980
+ },
1981
+ {
1982
+ "epoch": 2.56,
1983
+ "learning_rate": 4.1567345394149e-05,
1984
+ "loss": 0.6102,
1985
+ "step": 163000
1986
+ },
1987
+ {
1988
+ "epoch": 2.57,
1989
+ "learning_rate": 4.154106529029003e-05,
1990
+ "loss": 0.6088,
1991
+ "step": 163500
1992
+ },
1993
+ {
1994
+ "epoch": 2.58,
1995
+ "learning_rate": 4.151478518643106e-05,
1996
+ "loss": 0.6231,
1997
+ "step": 164000
1998
+ },
1999
+ {
2000
+ "epoch": 2.59,
2001
+ "learning_rate": 4.1488505082572087e-05,
2002
+ "loss": 0.6195,
2003
+ "step": 164500
2004
+ },
2005
+ {
2006
+ "epoch": 2.59,
2007
+ "learning_rate": 4.146222497871311e-05,
2008
+ "loss": 0.6214,
2009
+ "step": 165000
2010
+ },
2011
+ {
2012
+ "epoch": 2.6,
2013
+ "learning_rate": 4.1435997435061864e-05,
2014
+ "loss": 0.624,
2015
+ "step": 165500
2016
+ },
2017
+ {
2018
+ "epoch": 2.61,
2019
+ "learning_rate": 4.14097173312029e-05,
2020
+ "loss": 0.6218,
2021
+ "step": 166000
2022
+ },
2023
+ {
2024
+ "epoch": 2.62,
2025
+ "learning_rate": 4.138343722734392e-05,
2026
+ "loss": 0.6084,
2027
+ "step": 166500
2028
+ },
2029
+ {
2030
+ "epoch": 2.63,
2031
+ "learning_rate": 4.135715712348495e-05,
2032
+ "loss": 0.6155,
2033
+ "step": 167000
2034
+ },
2035
+ {
2036
+ "epoch": 2.63,
2037
+ "learning_rate": 4.133087701962599e-05,
2038
+ "loss": 0.6153,
2039
+ "step": 167500
2040
+ },
2041
+ {
2042
+ "epoch": 2.64,
2043
+ "learning_rate": 4.1304596915767015e-05,
2044
+ "loss": 0.6224,
2045
+ "step": 168000
2046
+ },
2047
+ {
2048
+ "epoch": 2.65,
2049
+ "learning_rate": 4.127831681190804e-05,
2050
+ "loss": 0.6132,
2051
+ "step": 168500
2052
+ },
2053
+ {
2054
+ "epoch": 2.66,
2055
+ "learning_rate": 4.125203670804907e-05,
2056
+ "loss": 0.6136,
2057
+ "step": 169000
2058
+ },
2059
+ {
2060
+ "epoch": 2.67,
2061
+ "learning_rate": 4.12257566041901e-05,
2062
+ "loss": 0.6097,
2063
+ "step": 169500
2064
+ },
2065
+ {
2066
+ "epoch": 2.67,
2067
+ "learning_rate": 4.119952906053885e-05,
2068
+ "loss": 0.6226,
2069
+ "step": 170000
2070
+ },
2071
+ {
2072
+ "epoch": 2.68,
2073
+ "learning_rate": 4.117324895667988e-05,
2074
+ "loss": 0.6152,
2075
+ "step": 170500
2076
+ },
2077
+ {
2078
+ "epoch": 2.69,
2079
+ "learning_rate": 4.114696885282091e-05,
2080
+ "loss": 0.6188,
2081
+ "step": 171000
2082
+ },
2083
+ {
2084
+ "epoch": 2.7,
2085
+ "learning_rate": 4.112074130916966e-05,
2086
+ "loss": 0.6199,
2087
+ "step": 171500
2088
+ },
2089
+ {
2090
+ "epoch": 2.7,
2091
+ "learning_rate": 4.109446120531069e-05,
2092
+ "loss": 0.626,
2093
+ "step": 172000
2094
+ },
2095
+ {
2096
+ "epoch": 2.71,
2097
+ "learning_rate": 4.106823366165943e-05,
2098
+ "loss": 0.6256,
2099
+ "step": 172500
2100
+ },
2101
+ {
2102
+ "epoch": 2.72,
2103
+ "learning_rate": 4.104195355780046e-05,
2104
+ "loss": 0.6169,
2105
+ "step": 173000
2106
+ },
2107
+ {
2108
+ "epoch": 2.73,
2109
+ "learning_rate": 4.10156734539415e-05,
2110
+ "loss": 0.6167,
2111
+ "step": 173500
2112
+ },
2113
+ {
2114
+ "epoch": 2.74,
2115
+ "learning_rate": 4.0989393350082524e-05,
2116
+ "loss": 0.6147,
2117
+ "step": 174000
2118
+ },
2119
+ {
2120
+ "epoch": 2.74,
2121
+ "learning_rate": 4.096311324622355e-05,
2122
+ "loss": 0.621,
2123
+ "step": 174500
2124
+ },
2125
+ {
2126
+ "epoch": 2.75,
2127
+ "learning_rate": 4.0936833142364576e-05,
2128
+ "loss": 0.6063,
2129
+ "step": 175000
2130
+ },
2131
+ {
2132
+ "epoch": 2.76,
2133
+ "learning_rate": 4.091055303850561e-05,
2134
+ "loss": 0.6216,
2135
+ "step": 175500
2136
+ },
2137
+ {
2138
+ "epoch": 2.77,
2139
+ "learning_rate": 4.088427293464664e-05,
2140
+ "loss": 0.6307,
2141
+ "step": 176000
2142
+ },
2143
+ {
2144
+ "epoch": 2.78,
2145
+ "learning_rate": 4.085799283078767e-05,
2146
+ "loss": 0.6313,
2147
+ "step": 176500
2148
+ },
2149
+ {
2150
+ "epoch": 2.78,
2151
+ "learning_rate": 4.08317127269287e-05,
2152
+ "loss": 0.6286,
2153
+ "step": 177000
2154
+ },
2155
+ {
2156
+ "epoch": 2.79,
2157
+ "learning_rate": 4.0805432623069726e-05,
2158
+ "loss": 0.6109,
2159
+ "step": 177500
2160
+ },
2161
+ {
2162
+ "epoch": 2.8,
2163
+ "learning_rate": 4.077915251921076e-05,
2164
+ "loss": 0.6175,
2165
+ "step": 178000
2166
+ },
2167
+ {
2168
+ "epoch": 2.81,
2169
+ "learning_rate": 4.075287241535179e-05,
2170
+ "loss": 0.6211,
2171
+ "step": 178500
2172
+ },
2173
+ {
2174
+ "epoch": 2.82,
2175
+ "learning_rate": 4.072659231149282e-05,
2176
+ "loss": 0.6227,
2177
+ "step": 179000
2178
+ },
2179
+ {
2180
+ "epoch": 2.82,
2181
+ "learning_rate": 4.0700312207633844e-05,
2182
+ "loss": 0.6254,
2183
+ "step": 179500
2184
+ },
2185
+ {
2186
+ "epoch": 2.83,
2187
+ "learning_rate": 4.0674032103774877e-05,
2188
+ "loss": 0.6212,
2189
+ "step": 180000
2190
+ },
2191
+ {
2192
+ "epoch": 2.84,
2193
+ "learning_rate": 4.064775199991591e-05,
2194
+ "loss": 0.6257,
2195
+ "step": 180500
2196
+ },
2197
+ {
2198
+ "epoch": 2.85,
2199
+ "learning_rate": 4.0621471896056935e-05,
2200
+ "loss": 0.6255,
2201
+ "step": 181000
2202
+ },
2203
+ {
2204
+ "epoch": 2.85,
2205
+ "learning_rate": 4.059519179219797e-05,
2206
+ "loss": 0.622,
2207
+ "step": 181500
2208
+ },
2209
+ {
2210
+ "epoch": 2.86,
2211
+ "learning_rate": 4.0568911688338994e-05,
2212
+ "loss": 0.6278,
2213
+ "step": 182000
2214
+ },
2215
+ {
2216
+ "epoch": 2.87,
2217
+ "learning_rate": 4.0542684144687746e-05,
2218
+ "loss": 0.6298,
2219
+ "step": 182500
2220
+ },
2221
+ {
2222
+ "epoch": 2.88,
2223
+ "learning_rate": 4.051640404082877e-05,
2224
+ "loss": 0.6197,
2225
+ "step": 183000
2226
+ },
2227
+ {
2228
+ "epoch": 2.89,
2229
+ "learning_rate": 4.0490176497177517e-05,
2230
+ "loss": 0.6169,
2231
+ "step": 183500
2232
+ },
2233
+ {
2234
+ "epoch": 2.89,
2235
+ "learning_rate": 4.046389639331855e-05,
2236
+ "loss": 0.6148,
2237
+ "step": 184000
2238
+ },
2239
+ {
2240
+ "epoch": 2.9,
2241
+ "learning_rate": 4.043761628945958e-05,
2242
+ "loss": 0.6153,
2243
+ "step": 184500
2244
+ },
2245
+ {
2246
+ "epoch": 2.91,
2247
+ "learning_rate": 4.041133618560061e-05,
2248
+ "loss": 0.622,
2249
+ "step": 185000
2250
+ },
2251
+ {
2252
+ "epoch": 2.92,
2253
+ "learning_rate": 4.0385056081741634e-05,
2254
+ "loss": 0.6291,
2255
+ "step": 185500
2256
+ },
2257
+ {
2258
+ "epoch": 2.93,
2259
+ "learning_rate": 4.035882853809038e-05,
2260
+ "loss": 0.6172,
2261
+ "step": 186000
2262
+ },
2263
+ {
2264
+ "epoch": 2.93,
2265
+ "learning_rate": 4.033254843423142e-05,
2266
+ "loss": 0.6156,
2267
+ "step": 186500
2268
+ },
2269
+ {
2270
+ "epoch": 2.94,
2271
+ "learning_rate": 4.0306268330372445e-05,
2272
+ "loss": 0.6274,
2273
+ "step": 187000
2274
+ },
2275
+ {
2276
+ "epoch": 2.95,
2277
+ "learning_rate": 4.028004078672119e-05,
2278
+ "loss": 0.6319,
2279
+ "step": 187500
2280
+ },
2281
+ {
2282
+ "epoch": 2.96,
2283
+ "learning_rate": 4.025381324306994e-05,
2284
+ "loss": 0.6218,
2285
+ "step": 188000
2286
+ },
2287
+ {
2288
+ "epoch": 2.96,
2289
+ "learning_rate": 4.022753313921097e-05,
2290
+ "loss": 0.6184,
2291
+ "step": 188500
2292
+ },
2293
+ {
2294
+ "epoch": 2.97,
2295
+ "learning_rate": 4.0201253035352e-05,
2296
+ "loss": 0.6275,
2297
+ "step": 189000
2298
+ },
2299
+ {
2300
+ "epoch": 2.98,
2301
+ "learning_rate": 4.0174972931493026e-05,
2302
+ "loss": 0.6218,
2303
+ "step": 189500
2304
+ },
2305
+ {
2306
+ "epoch": 2.99,
2307
+ "learning_rate": 4.014869282763405e-05,
2308
+ "loss": 0.6179,
2309
+ "step": 190000
2310
+ },
2311
+ {
2312
+ "epoch": 3.0,
2313
+ "learning_rate": 4.01224652839828e-05,
2314
+ "loss": 0.6294,
2315
+ "step": 190500
2316
+ },
2317
+ {
2318
+ "epoch": 3.0,
2319
+ "eval_bleu": 55.7536,
2320
+ "eval_gen_len": 15.2192,
2321
+ "eval_loss": 0.6636257171630859,
2322
+ "eval_runtime": 9606.0302,
2323
+ "eval_samples_per_second": 13.439,
2324
+ "eval_steps_per_second": 1.68,
2325
+ "step": 190759
2326
+ },
2327
+ {
2328
+ "epoch": 3.0,
2329
+ "learning_rate": 4.0096185180123836e-05,
2330
+ "loss": 0.551,
2331
+ "step": 191000
2332
+ },
2333
+ {
2334
+ "epoch": 3.01,
2335
+ "learning_rate": 4.006990507626486e-05,
2336
+ "loss": 0.4652,
2337
+ "step": 191500
2338
+ },
2339
+ {
2340
+ "epoch": 3.02,
2341
+ "learning_rate": 4.004362497240589e-05,
2342
+ "loss": 0.4579,
2343
+ "step": 192000
2344
+ },
2345
+ {
2346
+ "epoch": 3.03,
2347
+ "learning_rate": 4.001734486854692e-05,
2348
+ "loss": 0.4586,
2349
+ "step": 192500
2350
+ },
2351
+ {
2352
+ "epoch": 3.04,
2353
+ "learning_rate": 3.9991064764687954e-05,
2354
+ "loss": 0.4651,
2355
+ "step": 193000
2356
+ },
2357
+ {
2358
+ "epoch": 3.04,
2359
+ "learning_rate": 3.99648372210367e-05,
2360
+ "loss": 0.4688,
2361
+ "step": 193500
2362
+ },
2363
+ {
2364
+ "epoch": 3.05,
2365
+ "learning_rate": 3.9938557117177725e-05,
2366
+ "loss": 0.4685,
2367
+ "step": 194000
2368
+ },
2369
+ {
2370
+ "epoch": 3.06,
2371
+ "learning_rate": 3.991227701331876e-05,
2372
+ "loss": 0.4687,
2373
+ "step": 194500
2374
+ },
2375
+ {
2376
+ "epoch": 3.07,
2377
+ "learning_rate": 3.988599690945979e-05,
2378
+ "loss": 0.4672,
2379
+ "step": 195000
2380
+ },
2381
+ {
2382
+ "epoch": 3.07,
2383
+ "learning_rate": 3.9859716805600816e-05,
2384
+ "loss": 0.4693,
2385
+ "step": 195500
2386
+ },
2387
+ {
2388
+ "epoch": 3.08,
2389
+ "learning_rate": 3.983348926194956e-05,
2390
+ "loss": 0.4741,
2391
+ "step": 196000
2392
+ },
2393
+ {
2394
+ "epoch": 3.09,
2395
+ "learning_rate": 3.9807209158090594e-05,
2396
+ "loss": 0.4708,
2397
+ "step": 196500
2398
+ },
2399
+ {
2400
+ "epoch": 3.1,
2401
+ "learning_rate": 3.9780929054231627e-05,
2402
+ "loss": 0.4696,
2403
+ "step": 197000
2404
+ },
2405
+ {
2406
+ "epoch": 3.11,
2407
+ "learning_rate": 3.975464895037265e-05,
2408
+ "loss": 0.469,
2409
+ "step": 197500
2410
+ },
2411
+ {
2412
+ "epoch": 3.11,
2413
+ "learning_rate": 3.9728368846513685e-05,
2414
+ "loss": 0.4665,
2415
+ "step": 198000
2416
+ },
2417
+ {
2418
+ "epoch": 3.12,
2419
+ "learning_rate": 3.970208874265471e-05,
2420
+ "loss": 0.4742,
2421
+ "step": 198500
2422
+ },
2423
+ {
2424
+ "epoch": 3.13,
2425
+ "learning_rate": 3.9675808638795744e-05,
2426
+ "loss": 0.4854,
2427
+ "step": 199000
2428
+ },
2429
+ {
2430
+ "epoch": 3.14,
2431
+ "learning_rate": 3.964952853493677e-05,
2432
+ "loss": 0.4816,
2433
+ "step": 199500
2434
+ },
2435
+ {
2436
+ "epoch": 3.15,
2437
+ "learning_rate": 3.96232484310778e-05,
2438
+ "loss": 0.4851,
2439
+ "step": 200000
2440
+ },
2441
+ {
2442
+ "epoch": 3.15,
2443
+ "learning_rate": 3.959702088742655e-05,
2444
+ "loss": 0.4814,
2445
+ "step": 200500
2446
+ },
2447
+ {
2448
+ "epoch": 3.16,
2449
+ "learning_rate": 3.957074078356758e-05,
2450
+ "loss": 0.4766,
2451
+ "step": 201000
2452
+ },
2453
+ {
2454
+ "epoch": 3.17,
2455
+ "learning_rate": 3.9544513239916325e-05,
2456
+ "loss": 0.4869,
2457
+ "step": 201500
2458
+ },
2459
+ {
2460
+ "epoch": 3.18,
2461
+ "learning_rate": 3.951823313605735e-05,
2462
+ "loss": 0.4754,
2463
+ "step": 202000
2464
+ },
2465
+ {
2466
+ "epoch": 3.18,
2467
+ "learning_rate": 3.9491953032198384e-05,
2468
+ "loss": 0.4805,
2469
+ "step": 202500
2470
+ },
2471
+ {
2472
+ "epoch": 3.19,
2473
+ "learning_rate": 3.946567292833942e-05,
2474
+ "loss": 0.4876,
2475
+ "step": 203000
2476
+ },
2477
+ {
2478
+ "epoch": 3.2,
2479
+ "learning_rate": 3.943939282448044e-05,
2480
+ "loss": 0.4936,
2481
+ "step": 203500
2482
+ },
2483
+ {
2484
+ "epoch": 3.21,
2485
+ "learning_rate": 3.9413112720621476e-05,
2486
+ "loss": 0.4869,
2487
+ "step": 204000
2488
+ },
2489
+ {
2490
+ "epoch": 3.22,
2491
+ "learning_rate": 3.93868326167625e-05,
2492
+ "loss": 0.4943,
2493
+ "step": 204500
2494
+ },
2495
+ {
2496
+ "epoch": 3.22,
2497
+ "learning_rate": 3.9360552512903535e-05,
2498
+ "loss": 0.4947,
2499
+ "step": 205000
2500
+ },
2501
+ {
2502
+ "epoch": 3.23,
2503
+ "learning_rate": 3.933427240904457e-05,
2504
+ "loss": 0.4947,
2505
+ "step": 205500
2506
+ },
2507
+ {
2508
+ "epoch": 3.24,
2509
+ "learning_rate": 3.930799230518559e-05,
2510
+ "loss": 0.4849,
2511
+ "step": 206000
2512
+ },
2513
+ {
2514
+ "epoch": 3.25,
2515
+ "learning_rate": 3.928171220132662e-05,
2516
+ "loss": 0.4994,
2517
+ "step": 206500
2518
+ },
2519
+ {
2520
+ "epoch": 3.26,
2521
+ "learning_rate": 3.925548465767537e-05,
2522
+ "loss": 0.4892,
2523
+ "step": 207000
2524
+ },
2525
+ {
2526
+ "epoch": 3.26,
2527
+ "learning_rate": 3.9229204553816404e-05,
2528
+ "loss": 0.4856,
2529
+ "step": 207500
2530
+ },
2531
+ {
2532
+ "epoch": 3.27,
2533
+ "learning_rate": 3.920292444995743e-05,
2534
+ "loss": 0.4945,
2535
+ "step": 208000
2536
+ },
2537
+ {
2538
+ "epoch": 3.28,
2539
+ "learning_rate": 3.9176696906306175e-05,
2540
+ "loss": 0.4906,
2541
+ "step": 208500
2542
+ },
2543
+ {
2544
+ "epoch": 3.29,
2545
+ "learning_rate": 3.9150469362654926e-05,
2546
+ "loss": 0.4948,
2547
+ "step": 209000
2548
+ },
2549
+ {
2550
+ "epoch": 3.29,
2551
+ "learning_rate": 3.912418925879595e-05,
2552
+ "loss": 0.486,
2553
+ "step": 209500
2554
+ },
2555
+ {
2556
+ "epoch": 3.3,
2557
+ "learning_rate": 3.9097909154936985e-05,
2558
+ "loss": 0.4957,
2559
+ "step": 210000
2560
+ },
2561
+ {
2562
+ "epoch": 3.31,
2563
+ "learning_rate": 3.907162905107801e-05,
2564
+ "loss": 0.4912,
2565
+ "step": 210500
2566
+ },
2567
+ {
2568
+ "epoch": 3.32,
2569
+ "learning_rate": 3.904540150742676e-05,
2570
+ "loss": 0.4905,
2571
+ "step": 211000
2572
+ },
2573
+ {
2574
+ "epoch": 3.33,
2575
+ "learning_rate": 3.901912140356779e-05,
2576
+ "loss": 0.5008,
2577
+ "step": 211500
2578
+ },
2579
+ {
2580
+ "epoch": 3.33,
2581
+ "learning_rate": 3.899284129970882e-05,
2582
+ "loss": 0.4983,
2583
+ "step": 212000
2584
+ },
2585
+ {
2586
+ "epoch": 3.34,
2587
+ "learning_rate": 3.896656119584985e-05,
2588
+ "loss": 0.4988,
2589
+ "step": 212500
2590
+ },
2591
+ {
2592
+ "epoch": 3.35,
2593
+ "learning_rate": 3.894028109199087e-05,
2594
+ "loss": 0.4952,
2595
+ "step": 213000
2596
+ },
2597
+ {
2598
+ "epoch": 3.36,
2599
+ "learning_rate": 3.8914000988131906e-05,
2600
+ "loss": 0.4925,
2601
+ "step": 213500
2602
+ },
2603
+ {
2604
+ "epoch": 3.37,
2605
+ "learning_rate": 3.888772088427294e-05,
2606
+ "loss": 0.4992,
2607
+ "step": 214000
2608
+ },
2609
+ {
2610
+ "epoch": 3.37,
2611
+ "learning_rate": 3.8861440780413965e-05,
2612
+ "loss": 0.5013,
2613
+ "step": 214500
2614
+ },
2615
+ {
2616
+ "epoch": 3.38,
2617
+ "learning_rate": 3.883521323676271e-05,
2618
+ "loss": 0.5053,
2619
+ "step": 215000
2620
+ },
2621
+ {
2622
+ "epoch": 3.39,
2623
+ "learning_rate": 3.880893313290374e-05,
2624
+ "loss": 0.5039,
2625
+ "step": 215500
2626
+ },
2627
+ {
2628
+ "epoch": 3.4,
2629
+ "learning_rate": 3.8782653029044775e-05,
2630
+ "loss": 0.4962,
2631
+ "step": 216000
2632
+ },
2633
+ {
2634
+ "epoch": 3.4,
2635
+ "learning_rate": 3.87563729251858e-05,
2636
+ "loss": 0.4977,
2637
+ "step": 216500
2638
+ },
2639
+ {
2640
+ "epoch": 3.41,
2641
+ "learning_rate": 3.873009282132683e-05,
2642
+ "loss": 0.495,
2643
+ "step": 217000
2644
+ },
2645
+ {
2646
+ "epoch": 3.42,
2647
+ "learning_rate": 3.870381271746786e-05,
2648
+ "loss": 0.5101,
2649
+ "step": 217500
2650
+ },
2651
+ {
2652
+ "epoch": 3.43,
2653
+ "learning_rate": 3.867753261360889e-05,
2654
+ "loss": 0.4998,
2655
+ "step": 218000
2656
+ },
2657
+ {
2658
+ "epoch": 3.44,
2659
+ "learning_rate": 3.865125250974992e-05,
2660
+ "loss": 0.5069,
2661
+ "step": 218500
2662
+ },
2663
+ {
2664
+ "epoch": 3.44,
2665
+ "learning_rate": 3.862497240589095e-05,
2666
+ "loss": 0.5002,
2667
+ "step": 219000
2668
+ },
2669
+ {
2670
+ "epoch": 3.45,
2671
+ "learning_rate": 3.859869230203198e-05,
2672
+ "loss": 0.5035,
2673
+ "step": 219500
2674
+ },
2675
+ {
2676
+ "epoch": 3.46,
2677
+ "learning_rate": 3.857241219817301e-05,
2678
+ "loss": 0.5015,
2679
+ "step": 220000
2680
+ },
2681
+ {
2682
+ "epoch": 3.47,
2683
+ "learning_rate": 3.8546132094314036e-05,
2684
+ "loss": 0.5082,
2685
+ "step": 220500
2686
+ },
2687
+ {
2688
+ "epoch": 3.48,
2689
+ "learning_rate": 3.851985199045507e-05,
2690
+ "loss": 0.5069,
2691
+ "step": 221000
2692
+ },
2693
+ {
2694
+ "epoch": 3.48,
2695
+ "learning_rate": 3.8493571886596095e-05,
2696
+ "loss": 0.4975,
2697
+ "step": 221500
2698
+ },
2699
+ {
2700
+ "epoch": 3.49,
2701
+ "learning_rate": 3.846729178273713e-05,
2702
+ "loss": 0.512,
2703
+ "step": 222000
2704
+ },
2705
+ {
2706
+ "epoch": 3.5,
2707
+ "learning_rate": 3.844111679929359e-05,
2708
+ "loss": 0.5093,
2709
+ "step": 222500
2710
+ },
2711
+ {
2712
+ "epoch": 3.51,
2713
+ "learning_rate": 3.8414889255642337e-05,
2714
+ "loss": 0.5142,
2715
+ "step": 223000
2716
+ },
2717
+ {
2718
+ "epoch": 3.51,
2719
+ "learning_rate": 3.838860915178337e-05,
2720
+ "loss": 0.5048,
2721
+ "step": 223500
2722
+ },
2723
+ {
2724
+ "epoch": 3.52,
2725
+ "learning_rate": 3.83623290479244e-05,
2726
+ "loss": 0.5055,
2727
+ "step": 224000
2728
+ },
2729
+ {
2730
+ "epoch": 3.53,
2731
+ "learning_rate": 3.833604894406543e-05,
2732
+ "loss": 0.5102,
2733
+ "step": 224500
2734
+ },
2735
+ {
2736
+ "epoch": 3.54,
2737
+ "learning_rate": 3.8309768840206454e-05,
2738
+ "loss": 0.5213,
2739
+ "step": 225000
2740
+ },
2741
+ {
2742
+ "epoch": 3.55,
2743
+ "learning_rate": 3.8283541296555206e-05,
2744
+ "loss": 0.5063,
2745
+ "step": 225500
2746
+ },
2747
+ {
2748
+ "epoch": 3.55,
2749
+ "learning_rate": 3.825726119269624e-05,
2750
+ "loss": 0.5071,
2751
+ "step": 226000
2752
+ },
2753
+ {
2754
+ "epoch": 3.56,
2755
+ "learning_rate": 3.823103364904498e-05,
2756
+ "loss": 0.5042,
2757
+ "step": 226500
2758
+ },
2759
+ {
2760
+ "epoch": 3.57,
2761
+ "learning_rate": 3.820475354518601e-05,
2762
+ "loss": 0.5094,
2763
+ "step": 227000
2764
+ },
2765
+ {
2766
+ "epoch": 3.58,
2767
+ "learning_rate": 3.817847344132704e-05,
2768
+ "loss": 0.5141,
2769
+ "step": 227500
2770
+ },
2771
+ {
2772
+ "epoch": 3.59,
2773
+ "learning_rate": 3.8152193337468075e-05,
2774
+ "loss": 0.5031,
2775
+ "step": 228000
2776
+ },
2777
+ {
2778
+ "epoch": 3.59,
2779
+ "learning_rate": 3.81259132336091e-05,
2780
+ "loss": 0.5116,
2781
+ "step": 228500
2782
+ },
2783
+ {
2784
+ "epoch": 3.6,
2785
+ "learning_rate": 3.809963312975013e-05,
2786
+ "loss": 0.5178,
2787
+ "step": 229000
2788
+ },
2789
+ {
2790
+ "epoch": 3.61,
2791
+ "learning_rate": 3.807335302589116e-05,
2792
+ "loss": 0.5109,
2793
+ "step": 229500
2794
+ },
2795
+ {
2796
+ "epoch": 3.62,
2797
+ "learning_rate": 3.804707292203219e-05,
2798
+ "loss": 0.5091,
2799
+ "step": 230000
2800
+ },
2801
+ {
2802
+ "epoch": 3.62,
2803
+ "learning_rate": 3.802084537838094e-05,
2804
+ "loss": 0.507,
2805
+ "step": 230500
2806
+ },
2807
+ {
2808
+ "epoch": 3.63,
2809
+ "learning_rate": 3.799456527452196e-05,
2810
+ "loss": 0.508,
2811
+ "step": 231000
2812
+ },
2813
+ {
2814
+ "epoch": 3.64,
2815
+ "learning_rate": 3.7968285170662996e-05,
2816
+ "loss": 0.515,
2817
+ "step": 231500
2818
+ },
2819
+ {
2820
+ "epoch": 3.65,
2821
+ "learning_rate": 3.794200506680403e-05,
2822
+ "loss": 0.5169,
2823
+ "step": 232000
2824
+ },
2825
+ {
2826
+ "epoch": 3.66,
2827
+ "learning_rate": 3.7915724962945055e-05,
2828
+ "loss": 0.5249,
2829
+ "step": 232500
2830
+ },
2831
+ {
2832
+ "epoch": 3.66,
2833
+ "learning_rate": 3.788944485908609e-05,
2834
+ "loss": 0.5171,
2835
+ "step": 233000
2836
+ },
2837
+ {
2838
+ "epoch": 3.67,
2839
+ "learning_rate": 3.7863164755227114e-05,
2840
+ "loss": 0.5122,
2841
+ "step": 233500
2842
+ },
2843
+ {
2844
+ "epoch": 3.68,
2845
+ "learning_rate": 3.783688465136814e-05,
2846
+ "loss": 0.5101,
2847
+ "step": 234000
2848
+ },
2849
+ {
2850
+ "epoch": 3.69,
2851
+ "learning_rate": 3.781060454750917e-05,
2852
+ "loss": 0.5136,
2853
+ "step": 234500
2854
+ },
2855
+ {
2856
+ "epoch": 3.7,
2857
+ "learning_rate": 3.7784324443650205e-05,
2858
+ "loss": 0.5287,
2859
+ "step": 235000
2860
+ },
2861
+ {
2862
+ "epoch": 3.7,
2863
+ "learning_rate": 3.775804433979123e-05,
2864
+ "loss": 0.5192,
2865
+ "step": 235500
2866
+ },
2867
+ {
2868
+ "epoch": 3.71,
2869
+ "learning_rate": 3.773176423593226e-05,
2870
+ "loss": 0.5051,
2871
+ "step": 236000
2872
+ },
2873
+ {
2874
+ "epoch": 3.72,
2875
+ "learning_rate": 3.770548413207329e-05,
2876
+ "loss": 0.5229,
2877
+ "step": 236500
2878
+ },
2879
+ {
2880
+ "epoch": 3.73,
2881
+ "learning_rate": 3.767920402821432e-05,
2882
+ "loss": 0.5212,
2883
+ "step": 237000
2884
+ },
2885
+ {
2886
+ "epoch": 3.74,
2887
+ "learning_rate": 3.765292392435535e-05,
2888
+ "loss": 0.5236,
2889
+ "step": 237500
2890
+ },
2891
+ {
2892
+ "epoch": 3.74,
2893
+ "learning_rate": 3.7626696380704094e-05,
2894
+ "loss": 0.5146,
2895
+ "step": 238000
2896
+ },
2897
+ {
2898
+ "epoch": 3.75,
2899
+ "learning_rate": 3.7600416276845126e-05,
2900
+ "loss": 0.5245,
2901
+ "step": 238500
2902
+ },
2903
+ {
2904
+ "epoch": 3.76,
2905
+ "learning_rate": 3.757418873319388e-05,
2906
+ "loss": 0.5239,
2907
+ "step": 239000
2908
+ },
2909
+ {
2910
+ "epoch": 3.77,
2911
+ "learning_rate": 3.7547908629334904e-05,
2912
+ "loss": 0.5203,
2913
+ "step": 239500
2914
+ },
2915
+ {
2916
+ "epoch": 3.77,
2917
+ "learning_rate": 3.752162852547593e-05,
2918
+ "loss": 0.5051,
2919
+ "step": 240000
2920
+ },
2921
+ {
2922
+ "epoch": 3.78,
2923
+ "learning_rate": 3.749534842161696e-05,
2924
+ "loss": 0.519,
2925
+ "step": 240500
2926
+ },
2927
+ {
2928
+ "epoch": 3.79,
2929
+ "learning_rate": 3.7469120877965714e-05,
2930
+ "loss": 0.5311,
2931
+ "step": 241000
2932
+ },
2933
+ {
2934
+ "epoch": 3.8,
2935
+ "learning_rate": 3.744284077410674e-05,
2936
+ "loss": 0.5146,
2937
+ "step": 241500
2938
+ },
2939
+ {
2940
+ "epoch": 3.81,
2941
+ "learning_rate": 3.7416560670247766e-05,
2942
+ "loss": 0.523,
2943
+ "step": 242000
2944
+ },
2945
+ {
2946
+ "epoch": 3.81,
2947
+ "learning_rate": 3.739033312659652e-05,
2948
+ "loss": 0.5274,
2949
+ "step": 242500
2950
+ },
2951
+ {
2952
+ "epoch": 3.82,
2953
+ "learning_rate": 3.736405302273755e-05,
2954
+ "loss": 0.5089,
2955
+ "step": 243000
2956
+ },
2957
+ {
2958
+ "epoch": 3.83,
2959
+ "learning_rate": 3.733777291887858e-05,
2960
+ "loss": 0.5125,
2961
+ "step": 243500
2962
+ },
2963
+ {
2964
+ "epoch": 3.84,
2965
+ "learning_rate": 3.73114928150196e-05,
2966
+ "loss": 0.526,
2967
+ "step": 244000
2968
+ },
2969
+ {
2970
+ "epoch": 3.85,
2971
+ "learning_rate": 3.7285212711160636e-05,
2972
+ "loss": 0.5223,
2973
+ "step": 244500
2974
+ },
2975
+ {
2976
+ "epoch": 3.85,
2977
+ "learning_rate": 3.725893260730167e-05,
2978
+ "loss": 0.5305,
2979
+ "step": 245000
2980
+ },
2981
+ {
2982
+ "epoch": 3.86,
2983
+ "learning_rate": 3.7232652503442694e-05,
2984
+ "loss": 0.5268,
2985
+ "step": 245500
2986
+ },
2987
+ {
2988
+ "epoch": 3.87,
2989
+ "learning_rate": 3.720637239958373e-05,
2990
+ "loss": 0.5213,
2991
+ "step": 246000
2992
+ },
2993
+ {
2994
+ "epoch": 3.88,
2995
+ "learning_rate": 3.718009229572475e-05,
2996
+ "loss": 0.5203,
2997
+ "step": 246500
2998
+ },
2999
+ {
3000
+ "epoch": 3.88,
3001
+ "learning_rate": 3.7153812191865786e-05,
3002
+ "loss": 0.5339,
3003
+ "step": 247000
3004
+ },
3005
+ {
3006
+ "epoch": 3.89,
3007
+ "learning_rate": 3.712753208800681e-05,
3008
+ "loss": 0.5215,
3009
+ "step": 247500
3010
+ },
3011
+ {
3012
+ "epoch": 3.9,
3013
+ "learning_rate": 3.7101251984147845e-05,
3014
+ "loss": 0.5203,
3015
+ "step": 248000
3016
+ },
3017
+ {
3018
+ "epoch": 3.91,
3019
+ "learning_rate": 3.707497188028887e-05,
3020
+ "loss": 0.5162,
3021
+ "step": 248500
3022
+ },
3023
+ {
3024
+ "epoch": 3.92,
3025
+ "learning_rate": 3.70486917764299e-05,
3026
+ "loss": 0.5214,
3027
+ "step": 249000
3028
+ },
3029
+ {
3030
+ "epoch": 3.92,
3031
+ "learning_rate": 3.7022411672570936e-05,
3032
+ "loss": 0.5278,
3033
+ "step": 249500
3034
+ },
3035
+ {
3036
+ "epoch": 3.93,
3037
+ "learning_rate": 3.699613156871196e-05,
3038
+ "loss": 0.5251,
3039
+ "step": 250000
3040
+ },
3041
+ {
3042
+ "epoch": 3.94,
3043
+ "learning_rate": 3.696985146485299e-05,
3044
+ "loss": 0.5265,
3045
+ "step": 250500
3046
+ },
3047
+ {
3048
+ "epoch": 3.95,
3049
+ "learning_rate": 3.694357136099402e-05,
3050
+ "loss": 0.5222,
3051
+ "step": 251000
3052
+ },
3053
+ {
3054
+ "epoch": 3.96,
3055
+ "learning_rate": 3.691729125713505e-05,
3056
+ "loss": 0.518,
3057
+ "step": 251500
3058
+ },
3059
+ {
3060
+ "epoch": 3.96,
3061
+ "learning_rate": 3.68910637134838e-05,
3062
+ "loss": 0.5299,
3063
+ "step": 252000
3064
+ },
3065
+ {
3066
+ "epoch": 3.97,
3067
+ "learning_rate": 3.6864783609624825e-05,
3068
+ "loss": 0.5255,
3069
+ "step": 252500
3070
+ },
3071
+ {
3072
+ "epoch": 3.98,
3073
+ "learning_rate": 3.683855606597357e-05,
3074
+ "loss": 0.513,
3075
+ "step": 253000
3076
+ },
3077
+ {
3078
+ "epoch": 3.99,
3079
+ "learning_rate": 3.681232852232232e-05,
3080
+ "loss": 0.5178,
3081
+ "step": 253500
3082
+ },
3083
+ {
3084
+ "epoch": 3.99,
3085
+ "learning_rate": 3.6786048418463354e-05,
3086
+ "loss": 0.5313,
3087
+ "step": 254000
3088
+ },
3089
+ {
3090
+ "epoch": 4.0,
3091
+ "eval_bleu": 55.9422,
3092
+ "eval_gen_len": 15.2674,
3093
+ "eval_loss": 0.6860370635986328,
3094
+ "eval_runtime": 9607.1426,
3095
+ "eval_samples_per_second": 13.437,
3096
+ "eval_steps_per_second": 1.68,
3097
+ "step": 254346
3098
+ },
3099
+ {
3100
+ "epoch": 4.0,
3101
+ "learning_rate": 3.675976831460438e-05,
3102
+ "loss": 0.4774,
3103
+ "step": 254500
3104
+ },
3105
+ {
3106
+ "epoch": 4.01,
3107
+ "learning_rate": 3.673354077095313e-05,
3108
+ "loss": 0.3734,
3109
+ "step": 255000
3110
+ },
3111
+ {
3112
+ "epoch": 4.02,
3113
+ "learning_rate": 3.6707313227301876e-05,
3114
+ "loss": 0.3715,
3115
+ "step": 255500
3116
+ },
3117
+ {
3118
+ "epoch": 4.03,
3119
+ "learning_rate": 3.66810331234429e-05,
3120
+ "loss": 0.3741,
3121
+ "step": 256000
3122
+ },
3123
+ {
3124
+ "epoch": 4.03,
3125
+ "learning_rate": 3.6654753019583935e-05,
3126
+ "loss": 0.3715,
3127
+ "step": 256500
3128
+ },
3129
+ {
3130
+ "epoch": 4.04,
3131
+ "learning_rate": 3.662847291572497e-05,
3132
+ "loss": 0.3737,
3133
+ "step": 257000
3134
+ },
3135
+ {
3136
+ "epoch": 4.05,
3137
+ "learning_rate": 3.6602192811865994e-05,
3138
+ "loss": 0.3778,
3139
+ "step": 257500
3140
+ },
3141
+ {
3142
+ "epoch": 4.06,
3143
+ "learning_rate": 3.657591270800703e-05,
3144
+ "loss": 0.3739,
3145
+ "step": 258000
3146
+ },
3147
+ {
3148
+ "epoch": 4.07,
3149
+ "learning_rate": 3.654963260414805e-05,
3150
+ "loss": 0.3824,
3151
+ "step": 258500
3152
+ },
3153
+ {
3154
+ "epoch": 4.07,
3155
+ "learning_rate": 3.652335250028908e-05,
3156
+ "loss": 0.3834,
3157
+ "step": 259000
3158
+ },
3159
+ {
3160
+ "epoch": 4.08,
3161
+ "learning_rate": 3.649707239643011e-05,
3162
+ "loss": 0.3714,
3163
+ "step": 259500
3164
+ },
3165
+ {
3166
+ "epoch": 4.09,
3167
+ "learning_rate": 3.6470792292571144e-05,
3168
+ "loss": 0.3817,
3169
+ "step": 260000
3170
+ },
3171
+ {
3172
+ "epoch": 4.1,
3173
+ "learning_rate": 3.644451218871217e-05,
3174
+ "loss": 0.3781,
3175
+ "step": 260500
3176
+ },
3177
+ {
3178
+ "epoch": 4.1,
3179
+ "learning_rate": 3.6418284645060915e-05,
3180
+ "loss": 0.3836,
3181
+ "step": 261000
3182
+ },
3183
+ {
3184
+ "epoch": 4.11,
3185
+ "learning_rate": 3.639200454120195e-05,
3186
+ "loss": 0.3865,
3187
+ "step": 261500
3188
+ },
3189
+ {
3190
+ "epoch": 4.12,
3191
+ "learning_rate": 3.636572443734298e-05,
3192
+ "loss": 0.3837,
3193
+ "step": 262000
3194
+ },
3195
+ {
3196
+ "epoch": 4.13,
3197
+ "learning_rate": 3.633944433348401e-05,
3198
+ "loss": 0.384,
3199
+ "step": 262500
3200
+ },
3201
+ {
3202
+ "epoch": 4.14,
3203
+ "learning_rate": 3.631316422962503e-05,
3204
+ "loss": 0.3911,
3205
+ "step": 263000
3206
+ },
3207
+ {
3208
+ "epoch": 4.14,
3209
+ "learning_rate": 3.6286884125766066e-05,
3210
+ "loss": 0.3785,
3211
+ "step": 263500
3212
+ },
3213
+ {
3214
+ "epoch": 4.15,
3215
+ "learning_rate": 3.62606040219071e-05,
3216
+ "loss": 0.3826,
3217
+ "step": 264000
3218
+ },
3219
+ {
3220
+ "epoch": 4.16,
3221
+ "learning_rate": 3.6234323918048124e-05,
3222
+ "loss": 0.3827,
3223
+ "step": 264500
3224
+ },
3225
+ {
3226
+ "epoch": 4.17,
3227
+ "learning_rate": 3.620809637439687e-05,
3228
+ "loss": 0.393,
3229
+ "step": 265000
3230
+ },
3231
+ {
3232
+ "epoch": 4.18,
3233
+ "learning_rate": 3.61818162705379e-05,
3234
+ "loss": 0.3783,
3235
+ "step": 265500
3236
+ },
3237
+ {
3238
+ "epoch": 4.18,
3239
+ "learning_rate": 3.6155536166678935e-05,
3240
+ "loss": 0.3854,
3241
+ "step": 266000
3242
+ },
3243
+ {
3244
+ "epoch": 4.19,
3245
+ "learning_rate": 3.612925606281996e-05,
3246
+ "loss": 0.3824,
3247
+ "step": 266500
3248
+ },
3249
+ {
3250
+ "epoch": 4.2,
3251
+ "learning_rate": 3.6103028519168706e-05,
3252
+ "loss": 0.3858,
3253
+ "step": 267000
3254
+ },
3255
+ {
3256
+ "epoch": 4.21,
3257
+ "learning_rate": 3.607674841530974e-05,
3258
+ "loss": 0.3866,
3259
+ "step": 267500
3260
+ },
3261
+ {
3262
+ "epoch": 4.21,
3263
+ "learning_rate": 3.605046831145077e-05,
3264
+ "loss": 0.3862,
3265
+ "step": 268000
3266
+ },
3267
+ {
3268
+ "epoch": 4.22,
3269
+ "learning_rate": 3.60241882075918e-05,
3270
+ "loss": 0.3897,
3271
+ "step": 268500
3272
+ },
3273
+ {
3274
+ "epoch": 4.23,
3275
+ "learning_rate": 3.599790810373283e-05,
3276
+ "loss": 0.3894,
3277
+ "step": 269000
3278
+ },
3279
+ {
3280
+ "epoch": 4.24,
3281
+ "learning_rate": 3.5971627999873856e-05,
3282
+ "loss": 0.3895,
3283
+ "step": 269500
3284
+ },
3285
+ {
3286
+ "epoch": 4.25,
3287
+ "learning_rate": 3.594534789601489e-05,
3288
+ "loss": 0.3927,
3289
+ "step": 270000
3290
+ },
3291
+ {
3292
+ "epoch": 4.25,
3293
+ "learning_rate": 3.5919067792155915e-05,
3294
+ "loss": 0.4029,
3295
+ "step": 270500
3296
+ },
3297
+ {
3298
+ "epoch": 4.26,
3299
+ "learning_rate": 3.589278768829695e-05,
3300
+ "loss": 0.397,
3301
+ "step": 271000
3302
+ },
3303
+ {
3304
+ "epoch": 4.27,
3305
+ "learning_rate": 3.5866507584437974e-05,
3306
+ "loss": 0.3946,
3307
+ "step": 271500
3308
+ },
3309
+ {
3310
+ "epoch": 4.28,
3311
+ "learning_rate": 3.5840227480579e-05,
3312
+ "loss": 0.393,
3313
+ "step": 272000
3314
+ },
3315
+ {
3316
+ "epoch": 4.29,
3317
+ "learning_rate": 3.581399993692775e-05,
3318
+ "loss": 0.4006,
3319
+ "step": 272500
3320
+ },
3321
+ {
3322
+ "epoch": 4.29,
3323
+ "learning_rate": 3.5787719833068784e-05,
3324
+ "loss": 0.3956,
3325
+ "step": 273000
3326
+ },
3327
+ {
3328
+ "epoch": 4.3,
3329
+ "learning_rate": 3.576143972920981e-05,
3330
+ "loss": 0.3998,
3331
+ "step": 273500
3332
+ },
3333
+ {
3334
+ "epoch": 4.31,
3335
+ "learning_rate": 3.5735159625350836e-05,
3336
+ "loss": 0.3982,
3337
+ "step": 274000
3338
+ },
3339
+ {
3340
+ "epoch": 4.32,
3341
+ "learning_rate": 3.5708879521491876e-05,
3342
+ "loss": 0.3974,
3343
+ "step": 274500
3344
+ },
3345
+ {
3346
+ "epoch": 4.32,
3347
+ "learning_rate": 3.56825994176329e-05,
3348
+ "loss": 0.4028,
3349
+ "step": 275000
3350
+ },
3351
+ {
3352
+ "epoch": 4.33,
3353
+ "learning_rate": 3.5656371873981646e-05,
3354
+ "loss": 0.4118,
3355
+ "step": 275500
3356
+ },
3357
+ {
3358
+ "epoch": 4.34,
3359
+ "learning_rate": 3.563009177012267e-05,
3360
+ "loss": 0.4067,
3361
+ "step": 276000
3362
+ },
3363
+ {
3364
+ "epoch": 4.35,
3365
+ "learning_rate": 3.560381166626371e-05,
3366
+ "loss": 0.4008,
3367
+ "step": 276500
3368
+ },
3369
+ {
3370
+ "epoch": 4.36,
3371
+ "learning_rate": 3.557753156240474e-05,
3372
+ "loss": 0.4045,
3373
+ "step": 277000
3374
+ },
3375
+ {
3376
+ "epoch": 4.36,
3377
+ "learning_rate": 3.5551251458545764e-05,
3378
+ "loss": 0.4092,
3379
+ "step": 277500
3380
+ },
3381
+ {
3382
+ "epoch": 4.37,
3383
+ "learning_rate": 3.55249713546868e-05,
3384
+ "loss": 0.4109,
3385
+ "step": 278000
3386
+ },
3387
+ {
3388
+ "epoch": 4.38,
3389
+ "learning_rate": 3.549869125082782e-05,
3390
+ "loss": 0.409,
3391
+ "step": 278500
3392
+ },
3393
+ {
3394
+ "epoch": 4.39,
3395
+ "learning_rate": 3.5472411146968856e-05,
3396
+ "loss": 0.4024,
3397
+ "step": 279000
3398
+ },
3399
+ {
3400
+ "epoch": 4.4,
3401
+ "learning_rate": 3.544613104310988e-05,
3402
+ "loss": 0.4075,
3403
+ "step": 279500
3404
+ },
3405
+ {
3406
+ "epoch": 4.4,
3407
+ "learning_rate": 3.5419850939250914e-05,
3408
+ "loss": 0.4087,
3409
+ "step": 280000
3410
+ },
3411
+ {
3412
+ "epoch": 4.41,
3413
+ "learning_rate": 3.539357083539194e-05,
3414
+ "loss": 0.4039,
3415
+ "step": 280500
3416
+ },
3417
+ {
3418
+ "epoch": 4.42,
3419
+ "learning_rate": 3.536729073153297e-05,
3420
+ "loss": 0.4098,
3421
+ "step": 281000
3422
+ },
3423
+ {
3424
+ "epoch": 4.43,
3425
+ "learning_rate": 3.5341010627674006e-05,
3426
+ "loss": 0.4019,
3427
+ "step": 281500
3428
+ },
3429
+ {
3430
+ "epoch": 4.43,
3431
+ "learning_rate": 3.531473052381503e-05,
3432
+ "loss": 0.406,
3433
+ "step": 282000
3434
+ },
3435
+ {
3436
+ "epoch": 4.44,
3437
+ "learning_rate": 3.528845041995606e-05,
3438
+ "loss": 0.4078,
3439
+ "step": 282500
3440
+ },
3441
+ {
3442
+ "epoch": 4.45,
3443
+ "learning_rate": 3.526217031609709e-05,
3444
+ "loss": 0.409,
3445
+ "step": 283000
3446
+ },
3447
+ {
3448
+ "epoch": 4.46,
3449
+ "learning_rate": 3.523594277244584e-05,
3450
+ "loss": 0.4122,
3451
+ "step": 283500
3452
+ },
3453
+ {
3454
+ "epoch": 4.47,
3455
+ "learning_rate": 3.520966266858687e-05,
3456
+ "loss": 0.405,
3457
+ "step": 284000
3458
+ },
3459
+ {
3460
+ "epoch": 4.47,
3461
+ "learning_rate": 3.5183382564727894e-05,
3462
+ "loss": 0.4101,
3463
+ "step": 284500
3464
+ },
3465
+ {
3466
+ "epoch": 4.48,
3467
+ "learning_rate": 3.515710246086893e-05,
3468
+ "loss": 0.4218,
3469
+ "step": 285000
3470
+ },
3471
+ {
3472
+ "epoch": 4.49,
3473
+ "learning_rate": 3.513082235700996e-05,
3474
+ "loss": 0.4186,
3475
+ "step": 285500
3476
+ },
3477
+ {
3478
+ "epoch": 4.5,
3479
+ "learning_rate": 3.5104542253150986e-05,
3480
+ "loss": 0.4088,
3481
+ "step": 286000
3482
+ },
3483
+ {
3484
+ "epoch": 4.51,
3485
+ "learning_rate": 3.507826214929202e-05,
3486
+ "loss": 0.416,
3487
+ "step": 286500
3488
+ },
3489
+ {
3490
+ "epoch": 4.51,
3491
+ "learning_rate": 3.5051982045433045e-05,
3492
+ "loss": 0.4186,
3493
+ "step": 287000
3494
+ },
3495
+ {
3496
+ "epoch": 4.52,
3497
+ "learning_rate": 3.502570194157407e-05,
3498
+ "loss": 0.4146,
3499
+ "step": 287500
3500
+ },
3501
+ {
3502
+ "epoch": 4.53,
3503
+ "learning_rate": 3.499947439792282e-05,
3504
+ "loss": 0.4118,
3505
+ "step": 288000
3506
+ },
3507
+ {
3508
+ "epoch": 4.54,
3509
+ "learning_rate": 3.497319429406385e-05,
3510
+ "loss": 0.4141,
3511
+ "step": 288500
3512
+ },
3513
+ {
3514
+ "epoch": 4.54,
3515
+ "learning_rate": 3.494691419020488e-05,
3516
+ "loss": 0.4115,
3517
+ "step": 289000
3518
+ },
3519
+ {
3520
+ "epoch": 4.55,
3521
+ "learning_rate": 3.492068664655363e-05,
3522
+ "loss": 0.4136,
3523
+ "step": 289500
3524
+ },
3525
+ {
3526
+ "epoch": 4.56,
3527
+ "learning_rate": 3.489440654269466e-05,
3528
+ "loss": 0.4198,
3529
+ "step": 290000
3530
+ },
3531
+ {
3532
+ "epoch": 4.57,
3533
+ "learning_rate": 3.4868126438835685e-05,
3534
+ "loss": 0.4176,
3535
+ "step": 290500
3536
+ },
3537
+ {
3538
+ "epoch": 4.58,
3539
+ "learning_rate": 3.484189889518443e-05,
3540
+ "loss": 0.4212,
3541
+ "step": 291000
3542
+ },
3543
+ {
3544
+ "epoch": 4.58,
3545
+ "learning_rate": 3.481561879132547e-05,
3546
+ "loss": 0.4169,
3547
+ "step": 291500
3548
+ },
3549
+ {
3550
+ "epoch": 4.59,
3551
+ "learning_rate": 3.4789391247674214e-05,
3552
+ "loss": 0.407,
3553
+ "step": 292000
3554
+ },
3555
+ {
3556
+ "epoch": 4.6,
3557
+ "learning_rate": 3.476311114381524e-05,
3558
+ "loss": 0.4112,
3559
+ "step": 292500
3560
+ },
3561
+ {
3562
+ "epoch": 4.61,
3563
+ "learning_rate": 3.4736831039956266e-05,
3564
+ "loss": 0.4177,
3565
+ "step": 293000
3566
+ },
3567
+ {
3568
+ "epoch": 4.62,
3569
+ "learning_rate": 3.4710550936097306e-05,
3570
+ "loss": 0.4272,
3571
+ "step": 293500
3572
+ },
3573
+ {
3574
+ "epoch": 4.62,
3575
+ "learning_rate": 3.468427083223833e-05,
3576
+ "loss": 0.4221,
3577
+ "step": 294000
3578
+ },
3579
+ {
3580
+ "epoch": 4.63,
3581
+ "learning_rate": 3.465799072837936e-05,
3582
+ "loss": 0.4131,
3583
+ "step": 294500
3584
+ },
3585
+ {
3586
+ "epoch": 4.64,
3587
+ "learning_rate": 3.463171062452039e-05,
3588
+ "loss": 0.4238,
3589
+ "step": 295000
3590
+ },
3591
+ {
3592
+ "epoch": 4.65,
3593
+ "learning_rate": 3.4605430520661416e-05,
3594
+ "loss": 0.4213,
3595
+ "step": 295500
3596
+ },
3597
+ {
3598
+ "epoch": 4.66,
3599
+ "learning_rate": 3.457915041680245e-05,
3600
+ "loss": 0.4252,
3601
+ "step": 296000
3602
+ },
3603
+ {
3604
+ "epoch": 4.66,
3605
+ "learning_rate": 3.455287031294348e-05,
3606
+ "loss": 0.4222,
3607
+ "step": 296500
3608
+ },
3609
+ {
3610
+ "epoch": 4.67,
3611
+ "learning_rate": 3.452659020908451e-05,
3612
+ "loss": 0.4206,
3613
+ "step": 297000
3614
+ },
3615
+ {
3616
+ "epoch": 4.68,
3617
+ "learning_rate": 3.4500310105225534e-05,
3618
+ "loss": 0.415,
3619
+ "step": 297500
3620
+ },
3621
+ {
3622
+ "epoch": 4.69,
3623
+ "learning_rate": 3.447403000136657e-05,
3624
+ "loss": 0.4165,
3625
+ "step": 298000
3626
+ },
3627
+ {
3628
+ "epoch": 4.69,
3629
+ "learning_rate": 3.44477498975076e-05,
3630
+ "loss": 0.4266,
3631
+ "step": 298500
3632
+ },
3633
+ {
3634
+ "epoch": 4.7,
3635
+ "learning_rate": 3.4421522353856344e-05,
3636
+ "loss": 0.425,
3637
+ "step": 299000
3638
+ },
3639
+ {
3640
+ "epoch": 4.71,
3641
+ "learning_rate": 3.439524224999737e-05,
3642
+ "loss": 0.42,
3643
+ "step": 299500
3644
+ },
3645
+ {
3646
+ "epoch": 4.72,
3647
+ "learning_rate": 3.43689621461384e-05,
3648
+ "loss": 0.4271,
3649
+ "step": 300000
3650
+ },
3651
+ {
3652
+ "epoch": 4.73,
3653
+ "learning_rate": 3.4342734602487155e-05,
3654
+ "loss": 0.4308,
3655
+ "step": 300500
3656
+ },
3657
+ {
3658
+ "epoch": 4.73,
3659
+ "learning_rate": 3.431645449862818e-05,
3660
+ "loss": 0.4283,
3661
+ "step": 301000
3662
+ },
3663
+ {
3664
+ "epoch": 4.74,
3665
+ "learning_rate": 3.429017439476921e-05,
3666
+ "loss": 0.4296,
3667
+ "step": 301500
3668
+ },
3669
+ {
3670
+ "epoch": 4.75,
3671
+ "learning_rate": 3.426389429091024e-05,
3672
+ "loss": 0.4219,
3673
+ "step": 302000
3674
+ },
3675
+ {
3676
+ "epoch": 4.76,
3677
+ "learning_rate": 3.423761418705127e-05,
3678
+ "loss": 0.4176,
3679
+ "step": 302500
3680
+ },
3681
+ {
3682
+ "epoch": 4.77,
3683
+ "learning_rate": 3.42113340831923e-05,
3684
+ "loss": 0.43,
3685
+ "step": 303000
3686
+ },
3687
+ {
3688
+ "epoch": 4.77,
3689
+ "learning_rate": 3.418510653954104e-05,
3690
+ "loss": 0.4241,
3691
+ "step": 303500
3692
+ },
3693
+ {
3694
+ "epoch": 4.78,
3695
+ "learning_rate": 3.4158826435682076e-05,
3696
+ "loss": 0.4314,
3697
+ "step": 304000
3698
+ },
3699
+ {
3700
+ "epoch": 4.79,
3701
+ "learning_rate": 3.413254633182311e-05,
3702
+ "loss": 0.4217,
3703
+ "step": 304500
3704
+ },
3705
+ {
3706
+ "epoch": 4.8,
3707
+ "learning_rate": 3.4106266227964135e-05,
3708
+ "loss": 0.4378,
3709
+ "step": 305000
3710
+ },
3711
+ {
3712
+ "epoch": 4.8,
3713
+ "learning_rate": 3.407998612410516e-05,
3714
+ "loss": 0.4243,
3715
+ "step": 305500
3716
+ },
3717
+ {
3718
+ "epoch": 4.81,
3719
+ "learning_rate": 3.4053706020246193e-05,
3720
+ "loss": 0.4247,
3721
+ "step": 306000
3722
+ },
3723
+ {
3724
+ "epoch": 4.82,
3725
+ "learning_rate": 3.4027425916387226e-05,
3726
+ "loss": 0.4255,
3727
+ "step": 306500
3728
+ },
3729
+ {
3730
+ "epoch": 4.83,
3731
+ "learning_rate": 3.400114581252825e-05,
3732
+ "loss": 0.4317,
3733
+ "step": 307000
3734
+ },
3735
+ {
3736
+ "epoch": 4.84,
3737
+ "learning_rate": 3.3974865708669285e-05,
3738
+ "loss": 0.4201,
3739
+ "step": 307500
3740
+ },
3741
+ {
3742
+ "epoch": 4.84,
3743
+ "learning_rate": 3.394858560481031e-05,
3744
+ "loss": 0.4273,
3745
+ "step": 308000
3746
+ },
3747
+ {
3748
+ "epoch": 4.85,
3749
+ "learning_rate": 3.392230550095134e-05,
3750
+ "loss": 0.4322,
3751
+ "step": 308500
3752
+ },
3753
+ {
3754
+ "epoch": 4.86,
3755
+ "learning_rate": 3.389602539709238e-05,
3756
+ "loss": 0.4273,
3757
+ "step": 309000
3758
+ },
3759
+ {
3760
+ "epoch": 4.87,
3761
+ "learning_rate": 3.386979785344112e-05,
3762
+ "loss": 0.4291,
3763
+ "step": 309500
3764
+ },
3765
+ {
3766
+ "epoch": 4.88,
3767
+ "learning_rate": 3.3843570309789866e-05,
3768
+ "loss": 0.4234,
3769
+ "step": 310000
3770
+ },
3771
+ {
3772
+ "epoch": 4.88,
3773
+ "learning_rate": 3.38172902059309e-05,
3774
+ "loss": 0.4283,
3775
+ "step": 310500
3776
+ },
3777
+ {
3778
+ "epoch": 4.89,
3779
+ "learning_rate": 3.3791010102071925e-05,
3780
+ "loss": 0.4297,
3781
+ "step": 311000
3782
+ },
3783
+ {
3784
+ "epoch": 4.9,
3785
+ "learning_rate": 3.376472999821296e-05,
3786
+ "loss": 0.4347,
3787
+ "step": 311500
3788
+ },
3789
+ {
3790
+ "epoch": 4.91,
3791
+ "learning_rate": 3.37385024545617e-05,
3792
+ "loss": 0.4325,
3793
+ "step": 312000
3794
+ },
3795
+ {
3796
+ "epoch": 4.91,
3797
+ "learning_rate": 3.3712222350702735e-05,
3798
+ "loss": 0.4215,
3799
+ "step": 312500
3800
+ },
3801
+ {
3802
+ "epoch": 4.92,
3803
+ "learning_rate": 3.368594224684376e-05,
3804
+ "loss": 0.4354,
3805
+ "step": 313000
3806
+ },
3807
+ {
3808
+ "epoch": 4.93,
3809
+ "learning_rate": 3.365966214298479e-05,
3810
+ "loss": 0.4367,
3811
+ "step": 313500
3812
+ },
3813
+ {
3814
+ "epoch": 4.94,
3815
+ "learning_rate": 3.363343459933354e-05,
3816
+ "loss": 0.4402,
3817
+ "step": 314000
3818
+ },
3819
+ {
3820
+ "epoch": 4.95,
3821
+ "learning_rate": 3.360715449547457e-05,
3822
+ "loss": 0.4269,
3823
+ "step": 314500
3824
+ },
3825
+ {
3826
+ "epoch": 4.95,
3827
+ "learning_rate": 3.35808743916156e-05,
3828
+ "loss": 0.433,
3829
+ "step": 315000
3830
+ },
3831
+ {
3832
+ "epoch": 4.96,
3833
+ "learning_rate": 3.3554594287756624e-05,
3834
+ "loss": 0.4356,
3835
+ "step": 315500
3836
+ },
3837
+ {
3838
+ "epoch": 4.97,
3839
+ "learning_rate": 3.352831418389766e-05,
3840
+ "loss": 0.4309,
3841
+ "step": 316000
3842
+ },
3843
+ {
3844
+ "epoch": 4.98,
3845
+ "learning_rate": 3.350203408003868e-05,
3846
+ "loss": 0.4304,
3847
+ "step": 316500
3848
+ },
3849
+ {
3850
+ "epoch": 4.99,
3851
+ "learning_rate": 3.3475753976179715e-05,
3852
+ "loss": 0.433,
3853
+ "step": 317000
3854
+ },
3855
+ {
3856
+ "epoch": 4.99,
3857
+ "learning_rate": 3.344947387232075e-05,
3858
+ "loss": 0.4325,
3859
+ "step": 317500
3860
+ },
3861
+ {
3862
+ "epoch": 5.0,
3863
+ "eval_bleu": 56.2388,
3864
+ "eval_gen_len": 15.235,
3865
+ "eval_loss": 0.7183424234390259,
3866
+ "eval_runtime": 9602.8422,
3867
+ "eval_samples_per_second": 13.443,
3868
+ "eval_steps_per_second": 1.68,
3869
+ "step": 317932
3870
+ }
3871
+ ],
3872
+ "logging_steps": 500,
3873
+ "max_steps": 953790,
3874
+ "num_input_tokens_seen": 0,
3875
+ "num_train_epochs": 15,
3876
+ "save_steps": 500,
3877
+ "total_flos": 1.1024015759243215e+19,
3878
+ "train_batch_size": 2,
3879
+ "trial_name": null,
3880
+ "trial_params": null
3881
+ }
checkpoint-317932/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:348f748cc2c1841db000c69cf5b90e94af304038cbe239207bfe033a7791efcc
3
+ size 4856
checkpoint-381519/config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/mbart-large-50-many-to-many-mmt",
3
+ "_num_labels": 3,
4
+ "activation_dropout": 0.0,
5
+ "activation_function": "relu",
6
+ "add_bias_logits": false,
7
+ "add_final_layer_norm": true,
8
+ "architectures": [
9
+ "MBartForConditionalGeneration"
10
+ ],
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 0,
13
+ "classif_dropout": 0.0,
14
+ "classifier_dropout": 0.0,
15
+ "d_model": 1024,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_ffn_dim": 4096,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 12,
20
+ "decoder_start_token_id": 2,
21
+ "dropout": 0.1,
22
+ "early_stopping": true,
23
+ "encoder_attention_heads": 16,
24
+ "encoder_ffn_dim": 4096,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 2,
28
+ "forced_bos_token_id": 250014,
29
+ "forced_eos_token_id": 2,
30
+ "gradient_checkpointing": false,
31
+ "id2label": {
32
+ "0": "LABEL_0",
33
+ "1": "LABEL_1",
34
+ "2": "LABEL_2"
35
+ },
36
+ "init_std": 0.02,
37
+ "is_encoder_decoder": true,
38
+ "label2id": {
39
+ "LABEL_0": 0,
40
+ "LABEL_1": 1,
41
+ "LABEL_2": 2
42
+ },
43
+ "max_length": 200,
44
+ "max_position_embeddings": 1024,
45
+ "model_type": "mbart",
46
+ "normalize_before": true,
47
+ "normalize_embedding": true,
48
+ "num_beams": 5,
49
+ "num_hidden_layers": 12,
50
+ "output_past": true,
51
+ "pad_token_id": 1,
52
+ "scale_embedding": true,
53
+ "static_position_embeddings": false,
54
+ "tokenizer_class": "MBart50Tokenizer",
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.36.2",
57
+ "use_cache": true,
58
+ "vocab_size": 250054
59
+ }
checkpoint-381519/generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 0,
3
+ "decoder_start_token_id": 2,
4
+ "early_stopping": true,
5
+ "eos_token_id": 2,
6
+ "forced_bos_token_id": 250014,
7
+ "forced_eos_token_id": 2,
8
+ "max_length": 200,
9
+ "num_beams": 5,
10
+ "pad_token_id": 1,
11
+ "transformers_version": "4.36.2"
12
+ }
checkpoint-381519/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bbdd251c2d06d0ba7871e0757bf594d3a6feb418a4ba8b6eaebf1a256ce45c8f
3
+ size 2444578688
checkpoint-381519/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce448f423ef757e44aaba3514e8b8fd6b4c11b99433149864bece233b11c1ce0
3
+ size 4887473903
checkpoint-381519/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c884ae396e721537af6a54b5466411d6cc746579b42c11a36bc7f2e3e0d8d42
3
+ size 15024
checkpoint-381519/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:964da0db1a17bc1317edf9b20c889fe0a9a7e8d400fca202c77e0e0dea7919c7
3
+ size 15024
checkpoint-381519/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41c2257d5f2bc49c84420af06fad8fe87b619ad61f20d4a6ee57fc5ec13c243e
3
+ size 15024
checkpoint-381519/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ddf86bbde2c9e220402c199e70206c54ab4067d7475538c4fb83cd593ccd1265
3
+ size 15024
checkpoint-381519/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd159fba43de4380dbfec6a8dbdfb79407dfd90edce01a7b3e3bbac8cd332756
3
+ size 1064
checkpoint-381519/sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
checkpoint-381519/special_tokens_map.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "ar_AR",
4
+ "cs_CZ",
5
+ "de_DE",
6
+ "en_XX",
7
+ "es_XX",
8
+ "et_EE",
9
+ "fi_FI",
10
+ "fr_XX",
11
+ "gu_IN",
12
+ "hi_IN",
13
+ "it_IT",
14
+ "ja_XX",
15
+ "kk_KZ",
16
+ "ko_KR",
17
+ "lt_LT",
18
+ "lv_LV",
19
+ "my_MM",
20
+ "ne_NP",
21
+ "nl_XX",
22
+ "ro_RO",
23
+ "ru_RU",
24
+ "si_LK",
25
+ "tr_TR",
26
+ "vi_VN",
27
+ "zh_CN",
28
+ "af_ZA",
29
+ "az_AZ",
30
+ "bn_IN",
31
+ "fa_IR",
32
+ "he_IL",
33
+ "hr_HR",
34
+ "id_ID",
35
+ "ka_GE",
36
+ "km_KH",
37
+ "mk_MK",
38
+ "ml_IN",
39
+ "mn_MN",
40
+ "mr_IN",
41
+ "pl_PL",
42
+ "ps_AF",
43
+ "pt_XX",
44
+ "sv_SE",
45
+ "sw_KE",
46
+ "ta_IN",
47
+ "te_IN",
48
+ "th_TH",
49
+ "tl_XX",
50
+ "uk_UA",
51
+ "ur_PK",
52
+ "xh_ZA",
53
+ "gl_ES",
54
+ "sl_SI"
55
+ ],
56
+ "bos_token": "<s>",
57
+ "cls_token": "<s>",
58
+ "eos_token": "</s>",
59
+ "mask_token": {
60
+ "content": "<mask>",
61
+ "lstrip": true,
62
+ "normalized": true,
63
+ "rstrip": false,
64
+ "single_word": false
65
+ },
66
+ "pad_token": "<pad>",
67
+ "sep_token": "</s>",
68
+ "unk_token": "<unk>"
69
+ }
checkpoint-381519/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd7eee7b2b672fc40148ee5d62e7dcc8473f2ac9630ead56f0791e9eb15d1f01
3
+ size 17110107
checkpoint-381519/tokenizer_config.json ADDED
@@ -0,0 +1,528 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "ar_AR",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "250002": {
44
+ "content": "cs_CZ",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "250003": {
52
+ "content": "de_DE",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "250004": {
60
+ "content": "en_XX",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "250005": {
68
+ "content": "es_XX",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ },
75
+ "250006": {
76
+ "content": "et_EE",
77
+ "lstrip": false,
78
+ "normalized": false,
79
+ "rstrip": false,
80
+ "single_word": false,
81
+ "special": true
82
+ },
83
+ "250007": {
84
+ "content": "fi_FI",
85
+ "lstrip": false,
86
+ "normalized": false,
87
+ "rstrip": false,
88
+ "single_word": false,
89
+ "special": true
90
+ },
91
+ "250008": {
92
+ "content": "fr_XX",
93
+ "lstrip": false,
94
+ "normalized": false,
95
+ "rstrip": false,
96
+ "single_word": false,
97
+ "special": true
98
+ },
99
+ "250009": {
100
+ "content": "gu_IN",
101
+ "lstrip": false,
102
+ "normalized": false,
103
+ "rstrip": false,
104
+ "single_word": false,
105
+ "special": true
106
+ },
107
+ "250010": {
108
+ "content": "hi_IN",
109
+ "lstrip": false,
110
+ "normalized": false,
111
+ "rstrip": false,
112
+ "single_word": false,
113
+ "special": true
114
+ },
115
+ "250011": {
116
+ "content": "it_IT",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false,
121
+ "special": true
122
+ },
123
+ "250012": {
124
+ "content": "ja_XX",
125
+ "lstrip": false,
126
+ "normalized": false,
127
+ "rstrip": false,
128
+ "single_word": false,
129
+ "special": true
130
+ },
131
+ "250013": {
132
+ "content": "kk_KZ",
133
+ "lstrip": false,
134
+ "normalized": false,
135
+ "rstrip": false,
136
+ "single_word": false,
137
+ "special": true
138
+ },
139
+ "250014": {
140
+ "content": "ko_KR",
141
+ "lstrip": false,
142
+ "normalized": false,
143
+ "rstrip": false,
144
+ "single_word": false,
145
+ "special": true
146
+ },
147
+ "250015": {
148
+ "content": "lt_LT",
149
+ "lstrip": false,
150
+ "normalized": false,
151
+ "rstrip": false,
152
+ "single_word": false,
153
+ "special": true
154
+ },
155
+ "250016": {
156
+ "content": "lv_LV",
157
+ "lstrip": false,
158
+ "normalized": false,
159
+ "rstrip": false,
160
+ "single_word": false,
161
+ "special": true
162
+ },
163
+ "250017": {
164
+ "content": "my_MM",
165
+ "lstrip": false,
166
+ "normalized": false,
167
+ "rstrip": false,
168
+ "single_word": false,
169
+ "special": true
170
+ },
171
+ "250018": {
172
+ "content": "ne_NP",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false,
177
+ "special": true
178
+ },
179
+ "250019": {
180
+ "content": "nl_XX",
181
+ "lstrip": false,
182
+ "normalized": false,
183
+ "rstrip": false,
184
+ "single_word": false,
185
+ "special": true
186
+ },
187
+ "250020": {
188
+ "content": "ro_RO",
189
+ "lstrip": false,
190
+ "normalized": false,
191
+ "rstrip": false,
192
+ "single_word": false,
193
+ "special": true
194
+ },
195
+ "250021": {
196
+ "content": "ru_RU",
197
+ "lstrip": false,
198
+ "normalized": false,
199
+ "rstrip": false,
200
+ "single_word": false,
201
+ "special": true
202
+ },
203
+ "250022": {
204
+ "content": "si_LK",
205
+ "lstrip": false,
206
+ "normalized": false,
207
+ "rstrip": false,
208
+ "single_word": false,
209
+ "special": true
210
+ },
211
+ "250023": {
212
+ "content": "tr_TR",
213
+ "lstrip": false,
214
+ "normalized": false,
215
+ "rstrip": false,
216
+ "single_word": false,
217
+ "special": true
218
+ },
219
+ "250024": {
220
+ "content": "vi_VN",
221
+ "lstrip": false,
222
+ "normalized": false,
223
+ "rstrip": false,
224
+ "single_word": false,
225
+ "special": true
226
+ },
227
+ "250025": {
228
+ "content": "zh_CN",
229
+ "lstrip": false,
230
+ "normalized": false,
231
+ "rstrip": false,
232
+ "single_word": false,
233
+ "special": true
234
+ },
235
+ "250026": {
236
+ "content": "af_ZA",
237
+ "lstrip": false,
238
+ "normalized": false,
239
+ "rstrip": false,
240
+ "single_word": false,
241
+ "special": true
242
+ },
243
+ "250027": {
244
+ "content": "az_AZ",
245
+ "lstrip": false,
246
+ "normalized": false,
247
+ "rstrip": false,
248
+ "single_word": false,
249
+ "special": true
250
+ },
251
+ "250028": {
252
+ "content": "bn_IN",
253
+ "lstrip": false,
254
+ "normalized": false,
255
+ "rstrip": false,
256
+ "single_word": false,
257
+ "special": true
258
+ },
259
+ "250029": {
260
+ "content": "fa_IR",
261
+ "lstrip": false,
262
+ "normalized": false,
263
+ "rstrip": false,
264
+ "single_word": false,
265
+ "special": true
266
+ },
267
+ "250030": {
268
+ "content": "he_IL",
269
+ "lstrip": false,
270
+ "normalized": false,
271
+ "rstrip": false,
272
+ "single_word": false,
273
+ "special": true
274
+ },
275
+ "250031": {
276
+ "content": "hr_HR",
277
+ "lstrip": false,
278
+ "normalized": false,
279
+ "rstrip": false,
280
+ "single_word": false,
281
+ "special": true
282
+ },
283
+ "250032": {
284
+ "content": "id_ID",
285
+ "lstrip": false,
286
+ "normalized": false,
287
+ "rstrip": false,
288
+ "single_word": false,
289
+ "special": true
290
+ },
291
+ "250033": {
292
+ "content": "ka_GE",
293
+ "lstrip": false,
294
+ "normalized": false,
295
+ "rstrip": false,
296
+ "single_word": false,
297
+ "special": true
298
+ },
299
+ "250034": {
300
+ "content": "km_KH",
301
+ "lstrip": false,
302
+ "normalized": false,
303
+ "rstrip": false,
304
+ "single_word": false,
305
+ "special": true
306
+ },
307
+ "250035": {
308
+ "content": "mk_MK",
309
+ "lstrip": false,
310
+ "normalized": false,
311
+ "rstrip": false,
312
+ "single_word": false,
313
+ "special": true
314
+ },
315
+ "250036": {
316
+ "content": "ml_IN",
317
+ "lstrip": false,
318
+ "normalized": false,
319
+ "rstrip": false,
320
+ "single_word": false,
321
+ "special": true
322
+ },
323
+ "250037": {
324
+ "content": "mn_MN",
325
+ "lstrip": false,
326
+ "normalized": false,
327
+ "rstrip": false,
328
+ "single_word": false,
329
+ "special": true
330
+ },
331
+ "250038": {
332
+ "content": "mr_IN",
333
+ "lstrip": false,
334
+ "normalized": false,
335
+ "rstrip": false,
336
+ "single_word": false,
337
+ "special": true
338
+ },
339
+ "250039": {
340
+ "content": "pl_PL",
341
+ "lstrip": false,
342
+ "normalized": false,
343
+ "rstrip": false,
344
+ "single_word": false,
345
+ "special": true
346
+ },
347
+ "250040": {
348
+ "content": "ps_AF",
349
+ "lstrip": false,
350
+ "normalized": false,
351
+ "rstrip": false,
352
+ "single_word": false,
353
+ "special": true
354
+ },
355
+ "250041": {
356
+ "content": "pt_XX",
357
+ "lstrip": false,
358
+ "normalized": false,
359
+ "rstrip": false,
360
+ "single_word": false,
361
+ "special": true
362
+ },
363
+ "250042": {
364
+ "content": "sv_SE",
365
+ "lstrip": false,
366
+ "normalized": false,
367
+ "rstrip": false,
368
+ "single_word": false,
369
+ "special": true
370
+ },
371
+ "250043": {
372
+ "content": "sw_KE",
373
+ "lstrip": false,
374
+ "normalized": false,
375
+ "rstrip": false,
376
+ "single_word": false,
377
+ "special": true
378
+ },
379
+ "250044": {
380
+ "content": "ta_IN",
381
+ "lstrip": false,
382
+ "normalized": false,
383
+ "rstrip": false,
384
+ "single_word": false,
385
+ "special": true
386
+ },
387
+ "250045": {
388
+ "content": "te_IN",
389
+ "lstrip": false,
390
+ "normalized": false,
391
+ "rstrip": false,
392
+ "single_word": false,
393
+ "special": true
394
+ },
395
+ "250046": {
396
+ "content": "th_TH",
397
+ "lstrip": false,
398
+ "normalized": false,
399
+ "rstrip": false,
400
+ "single_word": false,
401
+ "special": true
402
+ },
403
+ "250047": {
404
+ "content": "tl_XX",
405
+ "lstrip": false,
406
+ "normalized": false,
407
+ "rstrip": false,
408
+ "single_word": false,
409
+ "special": true
410
+ },
411
+ "250048": {
412
+ "content": "uk_UA",
413
+ "lstrip": false,
414
+ "normalized": false,
415
+ "rstrip": false,
416
+ "single_word": false,
417
+ "special": true
418
+ },
419
+ "250049": {
420
+ "content": "ur_PK",
421
+ "lstrip": false,
422
+ "normalized": false,
423
+ "rstrip": false,
424
+ "single_word": false,
425
+ "special": true
426
+ },
427
+ "250050": {
428
+ "content": "xh_ZA",
429
+ "lstrip": false,
430
+ "normalized": false,
431
+ "rstrip": false,
432
+ "single_word": false,
433
+ "special": true
434
+ },
435
+ "250051": {
436
+ "content": "gl_ES",
437
+ "lstrip": false,
438
+ "normalized": false,
439
+ "rstrip": false,
440
+ "single_word": false,
441
+ "special": true
442
+ },
443
+ "250052": {
444
+ "content": "sl_SI",
445
+ "lstrip": false,
446
+ "normalized": false,
447
+ "rstrip": false,
448
+ "single_word": false,
449
+ "special": true
450
+ },
451
+ "250053": {
452
+ "content": "<mask>",
453
+ "lstrip": true,
454
+ "normalized": true,
455
+ "rstrip": false,
456
+ "single_word": false,
457
+ "special": true
458
+ }
459
+ },
460
+ "additional_special_tokens": [
461
+ "ar_AR",
462
+ "cs_CZ",
463
+ "de_DE",
464
+ "en_XX",
465
+ "es_XX",
466
+ "et_EE",
467
+ "fi_FI",
468
+ "fr_XX",
469
+ "gu_IN",
470
+ "hi_IN",
471
+ "it_IT",
472
+ "ja_XX",
473
+ "kk_KZ",
474
+ "ko_KR",
475
+ "lt_LT",
476
+ "lv_LV",
477
+ "my_MM",
478
+ "ne_NP",
479
+ "nl_XX",
480
+ "ro_RO",
481
+ "ru_RU",
482
+ "si_LK",
483
+ "tr_TR",
484
+ "vi_VN",
485
+ "zh_CN",
486
+ "af_ZA",
487
+ "az_AZ",
488
+ "bn_IN",
489
+ "fa_IR",
490
+ "he_IL",
491
+ "hr_HR",
492
+ "id_ID",
493
+ "ka_GE",
494
+ "km_KH",
495
+ "mk_MK",
496
+ "ml_IN",
497
+ "mn_MN",
498
+ "mr_IN",
499
+ "pl_PL",
500
+ "ps_AF",
501
+ "pt_XX",
502
+ "sv_SE",
503
+ "sw_KE",
504
+ "ta_IN",
505
+ "te_IN",
506
+ "th_TH",
507
+ "tl_XX",
508
+ "uk_UA",
509
+ "ur_PK",
510
+ "xh_ZA",
511
+ "gl_ES",
512
+ "sl_SI"
513
+ ],
514
+ "bos_token": "<s>",
515
+ "clean_up_tokenization_spaces": true,
516
+ "cls_token": "<s>",
517
+ "eos_token": "</s>",
518
+ "language_codes": "ML50",
519
+ "mask_token": "<mask>",
520
+ "model_max_length": 1000000000000000019884624838656,
521
+ "pad_token": "<pad>",
522
+ "sep_token": "</s>",
523
+ "sp_model_kwargs": {},
524
+ "src_lang": "en_XX",
525
+ "tgt_lang": "ko_KR",
526
+ "tokenizer_class": "MBart50Tokenizer",
527
+ "unk_token": "<unk>"
528
+ }
checkpoint-381519/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-381519/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:348f748cc2c1841db000c69cf5b90e94af304038cbe239207bfe033a7791efcc
3
+ size 4856
config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/mbart-large-50-many-to-many-mmt",
3
+ "_num_labels": 3,
4
+ "activation_dropout": 0.0,
5
+ "activation_function": "relu",
6
+ "add_bias_logits": false,
7
+ "add_final_layer_norm": true,
8
+ "architectures": [
9
+ "MBartForConditionalGeneration"
10
+ ],
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 0,
13
+ "classif_dropout": 0.0,
14
+ "classifier_dropout": 0.0,
15
+ "d_model": 1024,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_ffn_dim": 4096,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 12,
20
+ "decoder_start_token_id": 2,
21
+ "dropout": 0.1,
22
+ "early_stopping": true,
23
+ "encoder_attention_heads": 16,
24
+ "encoder_ffn_dim": 4096,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 2,
28
+ "forced_bos_token_id": 250014,
29
+ "forced_eos_token_id": 2,
30
+ "gradient_checkpointing": false,
31
+ "id2label": {
32
+ "0": "LABEL_0",
33
+ "1": "LABEL_1",
34
+ "2": "LABEL_2"
35
+ },
36
+ "init_std": 0.02,
37
+ "is_encoder_decoder": true,
38
+ "label2id": {
39
+ "LABEL_0": 0,
40
+ "LABEL_1": 1,
41
+ "LABEL_2": 2
42
+ },
43
+ "max_length": 200,
44
+ "max_position_embeddings": 1024,
45
+ "model_type": "mbart",
46
+ "normalize_before": true,
47
+ "normalize_embedding": true,
48
+ "num_beams": 5,
49
+ "num_hidden_layers": 12,
50
+ "output_past": true,
51
+ "pad_token_id": 1,
52
+ "scale_embedding": true,
53
+ "static_position_embeddings": false,
54
+ "tokenizer_class": "MBart50Tokenizer",
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.36.2",
57
+ "use_cache": true,
58
+ "vocab_size": 250054
59
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 6.0,
3
+ "eval_bleu": 55.7536,
4
+ "eval_gen_len": 15.2192,
5
+ "eval_loss": 0.6636257171630859,
6
+ "eval_runtime": 9612.8661,
7
+ "eval_samples": 129095,
8
+ "eval_samples_per_second": 13.429,
9
+ "eval_steps_per_second": 1.679
10
+ }