Hugging Face's logo

yesj1234
/
en_xlsr_100p_run1

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
./train_dataset.py
Generated from Trainer
Inference Endpoints
Model card Files Community
yesj1234 commited on
Commit
b84724e
•
1 Parent(s): 31d88dd

Upload folder using huggingface_hub

Browse files
Files changed (42) hide show
  1. README.md +60 -0
  2. added_tokens.json +4 -0
  3. all_results.json +8 -0
  4. checkpoint-388669/added_tokens.json +4 -0
  5. checkpoint-388669/config.json +117 -0
  6. checkpoint-388669/model.safetensors +3 -0
  7. checkpoint-388669/optimizer.pt +3 -0
  8. checkpoint-388669/preprocessor_config.json +10 -0
  9. checkpoint-388669/rng_state_0.pth +3 -0
  10. checkpoint-388669/rng_state_1.pth +3 -0
  11. checkpoint-388669/rng_state_2.pth +3 -0
  12. checkpoint-388669/rng_state_3.pth +3 -0
  13. checkpoint-388669/scheduler.pt +3 -0
  14. checkpoint-388669/special_tokens_map.json +30 -0
  15. checkpoint-388669/tokenizer_config.json +51 -0
  16. checkpoint-388669/trainer_state.json +1183 -0
  17. checkpoint-388669/training_args.bin +3 -0
  18. checkpoint-388669/vocab.json +200 -0
  19. checkpoint-431850/added_tokens.json +4 -0
  20. checkpoint-431850/config.json +117 -0
  21. checkpoint-431850/model.safetensors +3 -0
  22. checkpoint-431850/optimizer.pt +3 -0
  23. checkpoint-431850/preprocessor_config.json +10 -0
  24. checkpoint-431850/rng_state_0.pth +3 -0
  25. checkpoint-431850/rng_state_1.pth +3 -0
  26. checkpoint-431850/rng_state_2.pth +3 -0
  27. checkpoint-431850/rng_state_3.pth +3 -0
  28. checkpoint-431850/scheduler.pt +3 -0
  29. checkpoint-431850/special_tokens_map.json +30 -0
  30. checkpoint-431850/tokenizer_config.json +51 -0
  31. checkpoint-431850/trainer_state.json +1309 -0
  32. checkpoint-431850/training_args.bin +3 -0
  33. checkpoint-431850/vocab.json +200 -0
  34. config.json +117 -0
  35. model.safetensors +3 -0
  36. preprocessor_config.json +10 -0
  37. special_tokens_map.json +30 -0
  38. tokenizer_config.json +51 -0
  39. train_results.json +8 -0
  40. trainer_state.json +1318 -0
  41. training_args.bin +3 -0
  42. vocab.json +200 -0
README.md ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: facebook/wav2vec2-large-xlsr-53
4
+ tags:
5
+ - automatic-speech-recognition
6
+ - ./train_dataset.py
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: en_xlsr_100p_run1
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # en_xlsr_100p_run1
17
+
18
+ This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the ./TRAIN_DATASET.PY - NA dataset.
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 0.0003
38
+ - train_batch_size: 2
39
+ - eval_batch_size: 8
40
+ - seed: 42
41
+ - distributed_type: multi-GPU
42
+ - num_devices: 4
43
+ - gradient_accumulation_steps: 2
44
+ - total_train_batch_size: 16
45
+ - total_eval_batch_size: 32
46
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
+ - lr_scheduler_type: linear
48
+ - lr_scheduler_warmup_steps: 2000
49
+ - num_epochs: 10
50
+
51
+ ### Training results
52
+
53
+
54
+
55
+ ### Framework versions
56
+
57
+ - Transformers 4.35.0
58
+ - Pytorch 2.1.0+cu121
59
+ - Datasets 2.14.6
60
+ - Tokenizers 0.14.1
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "</s>": 199,
3
+ "<s>": 198
4
+ }
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "train_loss": 0.806209775733219,
4
+ "train_runtime": 212917.0057,
5
+ "train_samples": 690961,
6
+ "train_samples_per_second": 32.452,
7
+ "train_steps_per_second": 2.028
8
+ }
checkpoint-388669/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "</s>": 199,
3
+ "<s>": 198
4
+ }
checkpoint-388669/config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
3
+ "activation_dropout": 0.0,
4
+ "adapter_attn_dim": null,
5
+ "adapter_kernel_size": 3,
6
+ "adapter_stride": 2,
7
+ "add_adapter": false,
8
+ "apply_spec_augment": true,
9
+ "architectures": [
10
+ "Wav2Vec2ForCTC"
11
+ ],
12
+ "attention_dropout": 0.05,
13
+ "bos_token_id": 1,
14
+ "classifier_proj_size": 256,
15
+ "codevector_dim": 768,
16
+ "contrastive_logits_temperature": 0.1,
17
+ "conv_bias": true,
18
+ "conv_dim": [
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512,
25
+ 512
26
+ ],
27
+ "conv_kernel": [
28
+ 10,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 3,
33
+ 2,
34
+ 2
35
+ ],
36
+ "conv_stride": [
37
+ 5,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2,
43
+ 2
44
+ ],
45
+ "ctc_loss_reduction": "mean",
46
+ "ctc_zero_infinity": true,
47
+ "diversity_loss_weight": 0.1,
48
+ "do_stable_layer_norm": true,
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.05,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.0,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.05,
59
+ "hidden_size": 1024,
60
+ "initializer_range": 0.02,
61
+ "intermediate_size": 4096,
62
+ "layer_norm_eps": 1e-05,
63
+ "layerdrop": 0.05,
64
+ "mask_channel_length": 10,
65
+ "mask_channel_min_space": 1,
66
+ "mask_channel_other": 0.0,
67
+ "mask_channel_prob": 0.0,
68
+ "mask_channel_selection": "static",
69
+ "mask_feature_length": 10,
70
+ "mask_feature_min_masks": 0,
71
+ "mask_feature_prob": 0.0,
72
+ "mask_time_length": 10,
73
+ "mask_time_min_masks": 2,
74
+ "mask_time_min_space": 1,
75
+ "mask_time_other": 0.0,
76
+ "mask_time_prob": 0.05,
77
+ "mask_time_selection": "static",
78
+ "model_type": "wav2vec2",
79
+ "num_adapter_layers": 3,
80
+ "num_attention_heads": 16,
81
+ "num_codevector_groups": 2,
82
+ "num_codevectors_per_group": 320,
83
+ "num_conv_pos_embedding_groups": 16,
84
+ "num_conv_pos_embeddings": 128,
85
+ "num_feat_extract_layers": 7,
86
+ "num_hidden_layers": 24,
87
+ "num_negatives": 100,
88
+ "output_hidden_size": 1024,
89
+ "pad_token_id": 197,
90
+ "proj_codevector_dim": 768,
91
+ "tdnn_dilation": [
92
+ 1,
93
+ 2,
94
+ 3,
95
+ 1,
96
+ 1
97
+ ],
98
+ "tdnn_dim": [
99
+ 512,
100
+ 512,
101
+ 512,
102
+ 512,
103
+ 1500
104
+ ],
105
+ "tdnn_kernel": [
106
+ 5,
107
+ 3,
108
+ 3,
109
+ 1,
110
+ 1
111
+ ],
112
+ "torch_dtype": "float32",
113
+ "transformers_version": "4.35.0",
114
+ "use_weighted_layer_sum": false,
115
+ "vocab_size": 200,
116
+ "xvector_output_dim": 512
117
+ }
checkpoint-388669/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d8cc756fc8a1089df0e7b4f8d3b6aff9fdb6ad74fe4ded82dbc2cf3889f89dc
3
+ size 1262627480
checkpoint-388669/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61a8fec816d6d71ddfc0261a7bb92d082ded19e88b6a9c38dd83273f64445fab
3
+ size 2491799734
checkpoint-388669/preprocessor_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "processor_class": "Wav2Vec2Processor",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 16000
10
+ }
checkpoint-388669/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc938da77fbb2f3c230fc440a01f7d27edefc2331544ba5d990f7ff6e7860453
3
+ size 15024
checkpoint-388669/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20eee81b7164a33cb2ce66e8f8252f8dd3f047a3cecba1c4241e11653d3eeec1
3
+ size 15024
checkpoint-388669/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d266ac9deadb673db37aa9d0d394adc3051fcae4c991b6731b08beaa4519d800
3
+ size 15024
checkpoint-388669/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:321b02fc28d9e8f436f4893ae049bcbce48be1f261bf7897c6c4a45204d41750
3
+ size 15024
checkpoint-388669/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9438b6fbaf26db6eea5eb488f51137765b0d0d1f081deb3bbf39d55cb2140922
3
+ size 1064
checkpoint-388669/special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": true,
19
+ "normalized": false,
20
+ "rstrip": true,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "[UNK]",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": true,
28
+ "single_word": false
29
+ }
30
+ }
checkpoint-388669/tokenizer_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "196": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "197": {
12
+ "content": "[PAD]",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "198": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": true,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "199": {
28
+ "content": "</s>",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ }
35
+ },
36
+ "bos_token": "<s>",
37
+ "clean_up_tokenization_spaces": true,
38
+ "config": null,
39
+ "do_lower_case": false,
40
+ "eos_token": "</s>",
41
+ "model_max_length": 1000000000000000019884624838656,
42
+ "pad_token": "[PAD]",
43
+ "processor_class": "Wav2Vec2Processor",
44
+ "replace_word_delimiter_char": " ",
45
+ "target_lang": null,
46
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
47
+ "tokenizer_type": "wav2vec2",
48
+ "trust_remote_code": false,
49
+ "unk_token": "[UNK]",
50
+ "word_delimiter_token": "|"
51
+ }
checkpoint-388669/trainer_state.json ADDED
@@ -0,0 +1,1183 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 8.999988422039806,
5
+ "eval_steps": 500,
6
+ "global_step": 388669,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05,
13
+ "learning_rate": 0.0003,
14
+ "loss": 3.5282,
15
+ "step": 2000
16
+ },
17
+ {
18
+ "epoch": 0.09,
19
+ "learning_rate": 0.00029860416424334066,
20
+ "loss": 1.3191,
21
+ "step": 4000
22
+ },
23
+ {
24
+ "epoch": 0.14,
25
+ "learning_rate": 0.0002972083284866814,
26
+ "loss": 1.2226,
27
+ "step": 6000
28
+ },
29
+ {
30
+ "epoch": 0.19,
31
+ "learning_rate": 0.00029581249273002205,
32
+ "loss": 1.1832,
33
+ "step": 8000
34
+ },
35
+ {
36
+ "epoch": 0.23,
37
+ "learning_rate": 0.0002944166569733628,
38
+ "loss": 1.1595,
39
+ "step": 10000
40
+ },
41
+ {
42
+ "epoch": 0.28,
43
+ "learning_rate": 0.0002930208212167035,
44
+ "loss": 1.1368,
45
+ "step": 12000
46
+ },
47
+ {
48
+ "epoch": 0.32,
49
+ "learning_rate": 0.00029162498546004417,
50
+ "loss": 1.1176,
51
+ "step": 14000
52
+ },
53
+ {
54
+ "epoch": 0.37,
55
+ "learning_rate": 0.00029022914970338486,
56
+ "loss": 1.1044,
57
+ "step": 16000
58
+ },
59
+ {
60
+ "epoch": 0.42,
61
+ "learning_rate": 0.00028883331394672555,
62
+ "loss": 1.0954,
63
+ "step": 18000
64
+ },
65
+ {
66
+ "epoch": 0.46,
67
+ "learning_rate": 0.0002874374781900663,
68
+ "loss": 1.088,
69
+ "step": 20000
70
+ },
71
+ {
72
+ "epoch": 0.51,
73
+ "learning_rate": 0.000286041642433407,
74
+ "loss": 1.0765,
75
+ "step": 22000
76
+ },
77
+ {
78
+ "epoch": 0.56,
79
+ "learning_rate": 0.0002846458066767477,
80
+ "loss": 1.069,
81
+ "step": 24000
82
+ },
83
+ {
84
+ "epoch": 0.6,
85
+ "learning_rate": 0.00028324997092008837,
86
+ "loss": 1.0623,
87
+ "step": 26000
88
+ },
89
+ {
90
+ "epoch": 0.65,
91
+ "learning_rate": 0.00028185413516342906,
92
+ "loss": 1.0653,
93
+ "step": 28000
94
+ },
95
+ {
96
+ "epoch": 0.69,
97
+ "learning_rate": 0.0002804582994067698,
98
+ "loss": 1.0467,
99
+ "step": 30000
100
+ },
101
+ {
102
+ "epoch": 0.74,
103
+ "learning_rate": 0.00027906246365011044,
104
+ "loss": 1.0395,
105
+ "step": 32000
106
+ },
107
+ {
108
+ "epoch": 0.79,
109
+ "learning_rate": 0.0002776666278934512,
110
+ "loss": 1.0419,
111
+ "step": 34000
112
+ },
113
+ {
114
+ "epoch": 0.83,
115
+ "learning_rate": 0.0002762707921367919,
116
+ "loss": 1.0385,
117
+ "step": 36000
118
+ },
119
+ {
120
+ "epoch": 0.88,
121
+ "learning_rate": 0.00027487495638013257,
122
+ "loss": 1.0375,
123
+ "step": 38000
124
+ },
125
+ {
126
+ "epoch": 0.93,
127
+ "learning_rate": 0.00027347912062347326,
128
+ "loss": 1.0211,
129
+ "step": 40000
130
+ },
131
+ {
132
+ "epoch": 0.97,
133
+ "learning_rate": 0.00027208328486681395,
134
+ "loss": 1.0248,
135
+ "step": 42000
136
+ },
137
+ {
138
+ "epoch": 1.02,
139
+ "learning_rate": 0.0002706874491101547,
140
+ "loss": 1.0071,
141
+ "step": 44000
142
+ },
143
+ {
144
+ "epoch": 1.07,
145
+ "learning_rate": 0.0002692916133534954,
146
+ "loss": 0.9825,
147
+ "step": 46000
148
+ },
149
+ {
150
+ "epoch": 1.11,
151
+ "learning_rate": 0.0002678957775968361,
152
+ "loss": 0.983,
153
+ "step": 48000
154
+ },
155
+ {
156
+ "epoch": 1.16,
157
+ "learning_rate": 0.00026649994184017677,
158
+ "loss": 0.9831,
159
+ "step": 50000
160
+ },
161
+ {
162
+ "epoch": 1.2,
163
+ "learning_rate": 0.00026510410608351746,
164
+ "loss": 0.9792,
165
+ "step": 52000
166
+ },
167
+ {
168
+ "epoch": 1.25,
169
+ "learning_rate": 0.0002637082703268582,
170
+ "loss": 0.9744,
171
+ "step": 54000
172
+ },
173
+ {
174
+ "epoch": 1.3,
175
+ "learning_rate": 0.0002623124345701989,
176
+ "loss": 0.9668,
177
+ "step": 56000
178
+ },
179
+ {
180
+ "epoch": 1.34,
181
+ "learning_rate": 0.0002609165988135396,
182
+ "loss": 0.971,
183
+ "step": 58000
184
+ },
185
+ {
186
+ "epoch": 1.39,
187
+ "learning_rate": 0.0002595207630568803,
188
+ "loss": 0.9789,
189
+ "step": 60000
190
+ },
191
+ {
192
+ "epoch": 1.44,
193
+ "learning_rate": 0.000258124927300221,
194
+ "loss": 0.9794,
195
+ "step": 62000
196
+ },
197
+ {
198
+ "epoch": 1.48,
199
+ "learning_rate": 0.00025672909154356166,
200
+ "loss": 0.9615,
201
+ "step": 64000
202
+ },
203
+ {
204
+ "epoch": 1.53,
205
+ "learning_rate": 0.0002553332557869024,
206
+ "loss": 0.9577,
207
+ "step": 66000
208
+ },
209
+ {
210
+ "epoch": 1.57,
211
+ "learning_rate": 0.0002539374200302431,
212
+ "loss": 0.9573,
213
+ "step": 68000
214
+ },
215
+ {
216
+ "epoch": 1.62,
217
+ "learning_rate": 0.0002525415842735838,
218
+ "loss": 0.9633,
219
+ "step": 70000
220
+ },
221
+ {
222
+ "epoch": 1.67,
223
+ "learning_rate": 0.0002511457485169245,
224
+ "loss": 0.946,
225
+ "step": 72000
226
+ },
227
+ {
228
+ "epoch": 1.71,
229
+ "learning_rate": 0.00024974991276026517,
230
+ "loss": 0.9528,
231
+ "step": 74000
232
+ },
233
+ {
234
+ "epoch": 1.76,
235
+ "learning_rate": 0.0002483540770036059,
236
+ "loss": 0.9474,
237
+ "step": 76000
238
+ },
239
+ {
240
+ "epoch": 1.81,
241
+ "learning_rate": 0.0002469582412469466,
242
+ "loss": 0.9431,
243
+ "step": 78000
244
+ },
245
+ {
246
+ "epoch": 1.85,
247
+ "learning_rate": 0.0002455624054902873,
248
+ "loss": 0.9469,
249
+ "step": 80000
250
+ },
251
+ {
252
+ "epoch": 1.9,
253
+ "learning_rate": 0.000244166569733628,
254
+ "loss": 0.9442,
255
+ "step": 82000
256
+ },
257
+ {
258
+ "epoch": 1.95,
259
+ "learning_rate": 0.0002427707339769687,
260
+ "loss": 0.9387,
261
+ "step": 84000
262
+ },
263
+ {
264
+ "epoch": 1.99,
265
+ "learning_rate": 0.0002413748982203094,
266
+ "loss": 0.9341,
267
+ "step": 86000
268
+ },
269
+ {
270
+ "epoch": 2.04,
271
+ "learning_rate": 0.00023997906246365008,
272
+ "loss": 0.9124,
273
+ "step": 88000
274
+ },
275
+ {
276
+ "epoch": 2.08,
277
+ "learning_rate": 0.00023858322670699077,
278
+ "loss": 0.8962,
279
+ "step": 90000
280
+ },
281
+ {
282
+ "epoch": 2.13,
283
+ "learning_rate": 0.0002371873909503315,
284
+ "loss": 0.8974,
285
+ "step": 92000
286
+ },
287
+ {
288
+ "epoch": 2.18,
289
+ "learning_rate": 0.0002357915551936722,
290
+ "loss": 0.9122,
291
+ "step": 94000
292
+ },
293
+ {
294
+ "epoch": 2.22,
295
+ "learning_rate": 0.00023439571943701287,
296
+ "loss": 0.8977,
297
+ "step": 96000
298
+ },
299
+ {
300
+ "epoch": 2.27,
301
+ "learning_rate": 0.0002329998836803536,
302
+ "loss": 0.8972,
303
+ "step": 98000
304
+ },
305
+ {
306
+ "epoch": 2.32,
307
+ "learning_rate": 0.00023160404792369428,
308
+ "loss": 0.8969,
309
+ "step": 100000
310
+ },
311
+ {
312
+ "epoch": 2.36,
313
+ "learning_rate": 0.000230208212167035,
314
+ "loss": 0.9021,
315
+ "step": 102000
316
+ },
317
+ {
318
+ "epoch": 2.41,
319
+ "learning_rate": 0.00022881237641037572,
320
+ "loss": 0.9083,
321
+ "step": 104000
322
+ },
323
+ {
324
+ "epoch": 2.45,
325
+ "learning_rate": 0.00022741654065371638,
326
+ "loss": 0.8928,
327
+ "step": 106000
328
+ },
329
+ {
330
+ "epoch": 2.5,
331
+ "learning_rate": 0.0002260207048970571,
332
+ "loss": 0.8871,
333
+ "step": 108000
334
+ },
335
+ {
336
+ "epoch": 2.55,
337
+ "learning_rate": 0.0002246248691403978,
338
+ "loss": 0.9017,
339
+ "step": 110000
340
+ },
341
+ {
342
+ "epoch": 2.59,
343
+ "learning_rate": 0.0002232290333837385,
344
+ "loss": 0.8968,
345
+ "step": 112000
346
+ },
347
+ {
348
+ "epoch": 2.64,
349
+ "learning_rate": 0.00022183319762707917,
350
+ "loss": 0.8813,
351
+ "step": 114000
352
+ },
353
+ {
354
+ "epoch": 2.69,
355
+ "learning_rate": 0.0002204373618704199,
356
+ "loss": 0.8807,
357
+ "step": 116000
358
+ },
359
+ {
360
+ "epoch": 2.73,
361
+ "learning_rate": 0.0002190415261137606,
362
+ "loss": 0.8816,
363
+ "step": 118000
364
+ },
365
+ {
366
+ "epoch": 2.78,
367
+ "learning_rate": 0.0002176456903571013,
368
+ "loss": 0.8776,
369
+ "step": 120000
370
+ },
371
+ {
372
+ "epoch": 2.83,
373
+ "learning_rate": 0.000216249854600442,
374
+ "loss": 0.8738,
375
+ "step": 122000
376
+ },
377
+ {
378
+ "epoch": 2.87,
379
+ "learning_rate": 0.00021485401884378268,
380
+ "loss": 0.8826,
381
+ "step": 124000
382
+ },
383
+ {
384
+ "epoch": 2.92,
385
+ "learning_rate": 0.0002134581830871234,
386
+ "loss": 0.893,
387
+ "step": 126000
388
+ },
389
+ {
390
+ "epoch": 2.96,
391
+ "learning_rate": 0.00021206234733046412,
392
+ "loss": 0.8753,
393
+ "step": 128000
394
+ },
395
+ {
396
+ "epoch": 3.01,
397
+ "learning_rate": 0.00021066651157380478,
398
+ "loss": 0.8671,
399
+ "step": 130000
400
+ },
401
+ {
402
+ "epoch": 3.06,
403
+ "learning_rate": 0.0002092706758171455,
404
+ "loss": 0.8497,
405
+ "step": 132000
406
+ },
407
+ {
408
+ "epoch": 3.1,
409
+ "learning_rate": 0.00020787484006048622,
410
+ "loss": 0.8443,
411
+ "step": 134000
412
+ },
413
+ {
414
+ "epoch": 3.15,
415
+ "learning_rate": 0.0002064790043038269,
416
+ "loss": 0.8477,
417
+ "step": 136000
418
+ },
419
+ {
420
+ "epoch": 3.2,
421
+ "learning_rate": 0.0002050831685471676,
422
+ "loss": 0.8465,
423
+ "step": 138000
424
+ },
425
+ {
426
+ "epoch": 3.24,
427
+ "learning_rate": 0.0002036873327905083,
428
+ "loss": 0.8501,
429
+ "step": 140000
430
+ },
431
+ {
432
+ "epoch": 3.29,
433
+ "learning_rate": 0.000202291497033849,
434
+ "loss": 0.8451,
435
+ "step": 142000
436
+ },
437
+ {
438
+ "epoch": 3.33,
439
+ "learning_rate": 0.00020089566127718972,
440
+ "loss": 0.84,
441
+ "step": 144000
442
+ },
443
+ {
444
+ "epoch": 3.38,
445
+ "learning_rate": 0.0001994998255205304,
446
+ "loss": 0.8482,
447
+ "step": 146000
448
+ },
449
+ {
450
+ "epoch": 3.43,
451
+ "learning_rate": 0.0001981039897638711,
452
+ "loss": 0.8383,
453
+ "step": 148000
454
+ },
455
+ {
456
+ "epoch": 3.47,
457
+ "learning_rate": 0.0001967081540072118,
458
+ "loss": 0.8353,
459
+ "step": 150000
460
+ },
461
+ {
462
+ "epoch": 3.52,
463
+ "learning_rate": 0.00019531231825055251,
464
+ "loss": 0.8334,
465
+ "step": 152000
466
+ },
467
+ {
468
+ "epoch": 3.57,
469
+ "learning_rate": 0.00019391648249389318,
470
+ "loss": 0.8446,
471
+ "step": 154000
472
+ },
473
+ {
474
+ "epoch": 3.61,
475
+ "learning_rate": 0.0001925206467372339,
476
+ "loss": 0.8373,
477
+ "step": 156000
478
+ },
479
+ {
480
+ "epoch": 3.66,
481
+ "learning_rate": 0.00019112481098057461,
482
+ "loss": 0.8412,
483
+ "step": 158000
484
+ },
485
+ {
486
+ "epoch": 3.7,
487
+ "learning_rate": 0.0001897289752239153,
488
+ "loss": 0.8286,
489
+ "step": 160000
490
+ },
491
+ {
492
+ "epoch": 3.75,
493
+ "learning_rate": 0.000188333139467256,
494
+ "loss": 0.8327,
495
+ "step": 162000
496
+ },
497
+ {
498
+ "epoch": 3.8,
499
+ "learning_rate": 0.0001869373037105967,
500
+ "loss": 0.8426,
501
+ "step": 164000
502
+ },
503
+ {
504
+ "epoch": 3.84,
505
+ "learning_rate": 0.0001855414679539374,
506
+ "loss": 0.8291,
507
+ "step": 166000
508
+ },
509
+ {
510
+ "epoch": 3.89,
511
+ "learning_rate": 0.00018414563219727812,
512
+ "loss": 0.8271,
513
+ "step": 168000
514
+ },
515
+ {
516
+ "epoch": 3.94,
517
+ "learning_rate": 0.00018274979644061879,
518
+ "loss": 0.8278,
519
+ "step": 170000
520
+ },
521
+ {
522
+ "epoch": 3.98,
523
+ "learning_rate": 0.0001813539606839595,
524
+ "loss": 0.8307,
525
+ "step": 172000
526
+ },
527
+ {
528
+ "epoch": 4.03,
529
+ "learning_rate": 0.0001799581249273002,
530
+ "loss": 0.8109,
531
+ "step": 174000
532
+ },
533
+ {
534
+ "epoch": 4.08,
535
+ "learning_rate": 0.0001785622891706409,
536
+ "loss": 0.8027,
537
+ "step": 176000
538
+ },
539
+ {
540
+ "epoch": 4.12,
541
+ "learning_rate": 0.0001771664534139816,
542
+ "loss": 0.7955,
543
+ "step": 178000
544
+ },
545
+ {
546
+ "epoch": 4.17,
547
+ "learning_rate": 0.0001757706176573223,
548
+ "loss": 0.7888,
549
+ "step": 180000
550
+ },
551
+ {
552
+ "epoch": 4.21,
553
+ "learning_rate": 0.000174374781900663,
554
+ "loss": 0.799,
555
+ "step": 182000
556
+ },
557
+ {
558
+ "epoch": 4.26,
559
+ "learning_rate": 0.0001729789461440037,
560
+ "loss": 0.7949,
561
+ "step": 184000
562
+ },
563
+ {
564
+ "epoch": 4.31,
565
+ "learning_rate": 0.0001715831103873444,
566
+ "loss": 0.7913,
567
+ "step": 186000
568
+ },
569
+ {
570
+ "epoch": 4.35,
571
+ "learning_rate": 0.0001701872746306851,
572
+ "loss": 0.7862,
573
+ "step": 188000
574
+ },
575
+ {
576
+ "epoch": 4.4,
577
+ "learning_rate": 0.0001687914388740258,
578
+ "loss": 0.7882,
579
+ "step": 190000
580
+ },
581
+ {
582
+ "epoch": 4.45,
583
+ "learning_rate": 0.00016739560311736652,
584
+ "loss": 0.8017,
585
+ "step": 192000
586
+ },
587
+ {
588
+ "epoch": 4.49,
589
+ "learning_rate": 0.00016599976736070724,
590
+ "loss": 0.787,
591
+ "step": 194000
592
+ },
593
+ {
594
+ "epoch": 4.54,
595
+ "learning_rate": 0.0001646039316040479,
596
+ "loss": 0.7975,
597
+ "step": 196000
598
+ },
599
+ {
600
+ "epoch": 4.58,
601
+ "learning_rate": 0.00016320809584738862,
602
+ "loss": 0.7938,
603
+ "step": 198000
604
+ },
605
+ {
606
+ "epoch": 4.63,
607
+ "learning_rate": 0.0001618122600907293,
608
+ "loss": 0.799,
609
+ "step": 200000
610
+ },
611
+ {
612
+ "epoch": 4.68,
613
+ "learning_rate": 0.00016041642433407003,
614
+ "loss": 0.7811,
615
+ "step": 202000
616
+ },
617
+ {
618
+ "epoch": 4.72,
619
+ "learning_rate": 0.0001590205885774107,
620
+ "loss": 0.7941,
621
+ "step": 204000
622
+ },
623
+ {
624
+ "epoch": 4.77,
625
+ "learning_rate": 0.0001576247528207514,
626
+ "loss": 0.7918,
627
+ "step": 206000
628
+ },
629
+ {
630
+ "epoch": 4.82,
631
+ "learning_rate": 0.00015622891706409213,
632
+ "loss": 0.8006,
633
+ "step": 208000
634
+ },
635
+ {
636
+ "epoch": 4.86,
637
+ "learning_rate": 0.00015483308130743282,
638
+ "loss": 0.7896,
639
+ "step": 210000
640
+ },
641
+ {
642
+ "epoch": 4.91,
643
+ "learning_rate": 0.0001534372455507735,
644
+ "loss": 0.7944,
645
+ "step": 212000
646
+ },
647
+ {
648
+ "epoch": 4.96,
649
+ "learning_rate": 0.0001520414097941142,
650
+ "loss": 0.7869,
651
+ "step": 214000
652
+ },
653
+ {
654
+ "epoch": 5.0,
655
+ "learning_rate": 0.00015064557403745492,
656
+ "loss": 0.7847,
657
+ "step": 216000
658
+ },
659
+ {
660
+ "epoch": 5.05,
661
+ "learning_rate": 0.0001492497382807956,
662
+ "loss": 0.7577,
663
+ "step": 218000
664
+ },
665
+ {
666
+ "epoch": 5.09,
667
+ "learning_rate": 0.00014785390252413633,
668
+ "loss": 0.7527,
669
+ "step": 220000
670
+ },
671
+ {
672
+ "epoch": 5.14,
673
+ "learning_rate": 0.00014645806676747702,
674
+ "loss": 0.7554,
675
+ "step": 222000
676
+ },
677
+ {
678
+ "epoch": 5.19,
679
+ "learning_rate": 0.0001450622310108177,
680
+ "loss": 0.7633,
681
+ "step": 224000
682
+ },
683
+ {
684
+ "epoch": 5.23,
685
+ "learning_rate": 0.0001436663952541584,
686
+ "loss": 0.754,
687
+ "step": 226000
688
+ },
689
+ {
690
+ "epoch": 5.28,
691
+ "learning_rate": 0.00014227055949749912,
692
+ "loss": 0.7438,
693
+ "step": 228000
694
+ },
695
+ {
696
+ "epoch": 5.33,
697
+ "learning_rate": 0.0001408747237408398,
698
+ "loss": 0.7514,
699
+ "step": 230000
700
+ },
701
+ {
702
+ "epoch": 5.37,
703
+ "learning_rate": 0.00013947888798418053,
704
+ "loss": 0.7522,
705
+ "step": 232000
706
+ },
707
+ {
708
+ "epoch": 5.42,
709
+ "learning_rate": 0.00013808305222752122,
710
+ "loss": 0.7525,
711
+ "step": 234000
712
+ },
713
+ {
714
+ "epoch": 5.46,
715
+ "learning_rate": 0.00013668721647086193,
716
+ "loss": 0.7476,
717
+ "step": 236000
718
+ },
719
+ {
720
+ "epoch": 5.51,
721
+ "learning_rate": 0.00013529138071420263,
722
+ "loss": 0.747,
723
+ "step": 238000
724
+ },
725
+ {
726
+ "epoch": 5.56,
727
+ "learning_rate": 0.00013389554495754332,
728
+ "loss": 0.7505,
729
+ "step": 240000
730
+ },
731
+ {
732
+ "epoch": 5.6,
733
+ "learning_rate": 0.000132499709200884,
734
+ "loss": 0.7501,
735
+ "step": 242000
736
+ },
737
+ {
738
+ "epoch": 5.65,
739
+ "learning_rate": 0.00013110387344422472,
740
+ "loss": 0.7568,
741
+ "step": 244000
742
+ },
743
+ {
744
+ "epoch": 5.7,
745
+ "learning_rate": 0.00012970803768756542,
746
+ "loss": 0.7383,
747
+ "step": 246000
748
+ },
749
+ {
750
+ "epoch": 5.74,
751
+ "learning_rate": 0.00012831220193090613,
752
+ "loss": 0.7535,
753
+ "step": 248000
754
+ },
755
+ {
756
+ "epoch": 5.79,
757
+ "learning_rate": 0.00012691636617424682,
758
+ "loss": 0.751,
759
+ "step": 250000
760
+ },
761
+ {
762
+ "epoch": 5.84,
763
+ "learning_rate": 0.00012552053041758751,
764
+ "loss": 0.7396,
765
+ "step": 252000
766
+ },
767
+ {
768
+ "epoch": 5.88,
769
+ "learning_rate": 0.0001241246946609282,
770
+ "loss": 0.7448,
771
+ "step": 254000
772
+ },
773
+ {
774
+ "epoch": 5.93,
775
+ "learning_rate": 0.00012272885890426892,
776
+ "loss": 0.7521,
777
+ "step": 256000
778
+ },
779
+ {
780
+ "epoch": 5.97,
781
+ "learning_rate": 0.00012133302314760961,
782
+ "loss": 0.7472,
783
+ "step": 258000
784
+ },
785
+ {
786
+ "epoch": 6.02,
787
+ "learning_rate": 0.00011993718739095032,
788
+ "loss": 0.7344,
789
+ "step": 260000
790
+ },
791
+ {
792
+ "epoch": 6.07,
793
+ "learning_rate": 0.00011854135163429101,
794
+ "loss": 0.7143,
795
+ "step": 262000
796
+ },
797
+ {
798
+ "epoch": 6.11,
799
+ "learning_rate": 0.00011714551587763173,
800
+ "loss": 0.717,
801
+ "step": 264000
802
+ },
803
+ {
804
+ "epoch": 6.16,
805
+ "learning_rate": 0.00011574968012097242,
806
+ "loss": 0.7194,
807
+ "step": 266000
808
+ },
809
+ {
810
+ "epoch": 6.21,
811
+ "learning_rate": 0.00011435384436431312,
812
+ "loss": 0.7206,
813
+ "step": 268000
814
+ },
815
+ {
816
+ "epoch": 6.25,
817
+ "learning_rate": 0.00011295800860765381,
818
+ "loss": 0.7148,
819
+ "step": 270000
820
+ },
821
+ {
822
+ "epoch": 6.3,
823
+ "learning_rate": 0.00011156217285099453,
824
+ "loss": 0.7137,
825
+ "step": 272000
826
+ },
827
+ {
828
+ "epoch": 6.34,
829
+ "learning_rate": 0.00011016633709433522,
830
+ "loss": 0.7168,
831
+ "step": 274000
832
+ },
833
+ {
834
+ "epoch": 6.39,
835
+ "learning_rate": 0.00010877050133767593,
836
+ "loss": 0.7097,
837
+ "step": 276000
838
+ },
839
+ {
840
+ "epoch": 6.44,
841
+ "learning_rate": 0.00010737466558101662,
842
+ "loss": 0.7084,
843
+ "step": 278000
844
+ },
845
+ {
846
+ "epoch": 6.48,
847
+ "learning_rate": 0.00010597882982435732,
848
+ "loss": 0.7101,
849
+ "step": 280000
850
+ },
851
+ {
852
+ "epoch": 6.53,
853
+ "learning_rate": 0.00010458299406769801,
854
+ "loss": 0.7117,
855
+ "step": 282000
856
+ },
857
+ {
858
+ "epoch": 6.58,
859
+ "learning_rate": 0.00010318715831103873,
860
+ "loss": 0.7067,
861
+ "step": 284000
862
+ },
863
+ {
864
+ "epoch": 6.62,
865
+ "learning_rate": 0.00010179132255437942,
866
+ "loss": 0.7026,
867
+ "step": 286000
868
+ },
869
+ {
870
+ "epoch": 6.67,
871
+ "learning_rate": 0.00010039548679772013,
872
+ "loss": 0.7144,
873
+ "step": 288000
874
+ },
875
+ {
876
+ "epoch": 6.72,
877
+ "learning_rate": 9.899965104106083e-05,
878
+ "loss": 0.7118,
879
+ "step": 290000
880
+ },
881
+ {
882
+ "epoch": 6.76,
883
+ "learning_rate": 9.760381528440152e-05,
884
+ "loss": 0.7028,
885
+ "step": 292000
886
+ },
887
+ {
888
+ "epoch": 6.81,
889
+ "learning_rate": 9.620797952774224e-05,
890
+ "loss": 0.7044,
891
+ "step": 294000
892
+ },
893
+ {
894
+ "epoch": 6.85,
895
+ "learning_rate": 9.481214377108293e-05,
896
+ "loss": 0.704,
897
+ "step": 296000
898
+ },
899
+ {
900
+ "epoch": 6.9,
901
+ "learning_rate": 9.341630801442363e-05,
902
+ "loss": 0.7049,
903
+ "step": 298000
904
+ },
905
+ {
906
+ "epoch": 6.95,
907
+ "learning_rate": 9.202047225776432e-05,
908
+ "loss": 0.6937,
909
+ "step": 300000
910
+ },
911
+ {
912
+ "epoch": 6.99,
913
+ "learning_rate": 9.062463650110504e-05,
914
+ "loss": 0.7096,
915
+ "step": 302000
916
+ },
917
+ {
918
+ "epoch": 7.04,
919
+ "learning_rate": 8.922880074444573e-05,
920
+ "loss": 0.6749,
921
+ "step": 304000
922
+ },
923
+ {
924
+ "epoch": 7.09,
925
+ "learning_rate": 8.783296498778644e-05,
926
+ "loss": 0.6718,
927
+ "step": 306000
928
+ },
929
+ {
930
+ "epoch": 7.13,
931
+ "learning_rate": 8.643712923112713e-05,
932
+ "loss": 0.6789,
933
+ "step": 308000
934
+ },
935
+ {
936
+ "epoch": 7.18,
937
+ "learning_rate": 8.504129347446783e-05,
938
+ "loss": 0.6805,
939
+ "step": 310000
940
+ },
941
+ {
942
+ "epoch": 7.22,
943
+ "learning_rate": 8.364545771780852e-05,
944
+ "loss": 0.6776,
945
+ "step": 312000
946
+ },
947
+ {
948
+ "epoch": 7.27,
949
+ "learning_rate": 8.224962196114924e-05,
950
+ "loss": 0.6734,
951
+ "step": 314000
952
+ },
953
+ {
954
+ "epoch": 7.32,
955
+ "learning_rate": 8.085378620448993e-05,
956
+ "loss": 0.6765,
957
+ "step": 316000
958
+ },
959
+ {
960
+ "epoch": 7.36,
961
+ "learning_rate": 7.945795044783064e-05,
962
+ "loss": 0.6709,
963
+ "step": 318000
964
+ },
965
+ {
966
+ "epoch": 7.41,
967
+ "learning_rate": 7.806211469117133e-05,
968
+ "loss": 0.6804,
969
+ "step": 320000
970
+ },
971
+ {
972
+ "epoch": 7.46,
973
+ "learning_rate": 7.666627893451203e-05,
974
+ "loss": 0.6688,
975
+ "step": 322000
976
+ },
977
+ {
978
+ "epoch": 7.5,
979
+ "learning_rate": 7.527044317785274e-05,
980
+ "loss": 0.6789,
981
+ "step": 324000
982
+ },
983
+ {
984
+ "epoch": 7.55,
985
+ "learning_rate": 7.387460742119343e-05,
986
+ "loss": 0.6699,
987
+ "step": 326000
988
+ },
989
+ {
990
+ "epoch": 7.6,
991
+ "learning_rate": 7.247877166453413e-05,
992
+ "loss": 0.662,
993
+ "step": 328000
994
+ },
995
+ {
996
+ "epoch": 7.64,
997
+ "learning_rate": 7.108293590787484e-05,
998
+ "loss": 0.6657,
999
+ "step": 330000
1000
+ },
1001
+ {
1002
+ "epoch": 7.69,
1003
+ "learning_rate": 6.968710015121553e-05,
1004
+ "loss": 0.6676,
1005
+ "step": 332000
1006
+ },
1007
+ {
1008
+ "epoch": 7.73,
1009
+ "learning_rate": 6.829126439455623e-05,
1010
+ "loss": 0.6698,
1011
+ "step": 334000
1012
+ },
1013
+ {
1014
+ "epoch": 7.78,
1015
+ "learning_rate": 6.689542863789693e-05,
1016
+ "loss": 0.6755,
1017
+ "step": 336000
1018
+ },
1019
+ {
1020
+ "epoch": 7.83,
1021
+ "learning_rate": 6.549959288123764e-05,
1022
+ "loss": 0.672,
1023
+ "step": 338000
1024
+ },
1025
+ {
1026
+ "epoch": 7.87,
1027
+ "learning_rate": 6.410375712457834e-05,
1028
+ "loss": 0.6616,
1029
+ "step": 340000
1030
+ },
1031
+ {
1032
+ "epoch": 7.92,
1033
+ "learning_rate": 6.270792136791903e-05,
1034
+ "loss": 0.6654,
1035
+ "step": 342000
1036
+ },
1037
+ {
1038
+ "epoch": 7.97,
1039
+ "learning_rate": 6.131208561125974e-05,
1040
+ "loss": 0.6606,
1041
+ "step": 344000
1042
+ },
1043
+ {
1044
+ "epoch": 8.01,
1045
+ "learning_rate": 5.991624985460044e-05,
1046
+ "loss": 0.6642,
1047
+ "step": 346000
1048
+ },
1049
+ {
1050
+ "epoch": 8.06,
1051
+ "learning_rate": 5.852041409794114e-05,
1052
+ "loss": 0.639,
1053
+ "step": 348000
1054
+ },
1055
+ {
1056
+ "epoch": 8.1,
1057
+ "learning_rate": 5.712457834128184e-05,
1058
+ "loss": 0.6463,
1059
+ "step": 350000
1060
+ },
1061
+ {
1062
+ "epoch": 8.15,
1063
+ "learning_rate": 5.572874258462254e-05,
1064
+ "loss": 0.6426,
1065
+ "step": 352000
1066
+ },
1067
+ {
1068
+ "epoch": 8.2,
1069
+ "learning_rate": 5.433290682796324e-05,
1070
+ "loss": 0.6356,
1071
+ "step": 354000
1072
+ },
1073
+ {
1074
+ "epoch": 8.24,
1075
+ "learning_rate": 5.2937071071303944e-05,
1076
+ "loss": 0.6336,
1077
+ "step": 356000
1078
+ },
1079
+ {
1080
+ "epoch": 8.29,
1081
+ "learning_rate": 5.154123531464464e-05,
1082
+ "loss": 0.631,
1083
+ "step": 358000
1084
+ },
1085
+ {
1086
+ "epoch": 8.34,
1087
+ "learning_rate": 5.014539955798534e-05,
1088
+ "loss": 0.6425,
1089
+ "step": 360000
1090
+ },
1091
+ {
1092
+ "epoch": 8.38,
1093
+ "learning_rate": 4.8749563801326044e-05,
1094
+ "loss": 0.6432,
1095
+ "step": 362000
1096
+ },
1097
+ {
1098
+ "epoch": 8.43,
1099
+ "learning_rate": 4.735372804466674e-05,
1100
+ "loss": 0.6449,
1101
+ "step": 364000
1102
+ },
1103
+ {
1104
+ "epoch": 8.48,
1105
+ "learning_rate": 4.595789228800744e-05,
1106
+ "loss": 0.635,
1107
+ "step": 366000
1108
+ },
1109
+ {
1110
+ "epoch": 8.52,
1111
+ "learning_rate": 4.456205653134814e-05,
1112
+ "loss": 0.6352,
1113
+ "step": 368000
1114
+ },
1115
+ {
1116
+ "epoch": 8.57,
1117
+ "learning_rate": 4.316622077468884e-05,
1118
+ "loss": 0.6281,
1119
+ "step": 370000
1120
+ },
1121
+ {
1122
+ "epoch": 8.61,
1123
+ "learning_rate": 4.1770385018029545e-05,
1124
+ "loss": 0.6291,
1125
+ "step": 372000
1126
+ },
1127
+ {
1128
+ "epoch": 8.66,
1129
+ "learning_rate": 4.037454926137024e-05,
1130
+ "loss": 0.6317,
1131
+ "step": 374000
1132
+ },
1133
+ {
1134
+ "epoch": 8.71,
1135
+ "learning_rate": 3.897871350471094e-05,
1136
+ "loss": 0.6296,
1137
+ "step": 376000
1138
+ },
1139
+ {
1140
+ "epoch": 8.75,
1141
+ "learning_rate": 3.7582877748051644e-05,
1142
+ "loss": 0.6279,
1143
+ "step": 378000
1144
+ },
1145
+ {
1146
+ "epoch": 8.8,
1147
+ "learning_rate": 3.618704199139234e-05,
1148
+ "loss": 0.6379,
1149
+ "step": 380000
1150
+ },
1151
+ {
1152
+ "epoch": 8.85,
1153
+ "learning_rate": 3.4791206234733046e-05,
1154
+ "loss": 0.6292,
1155
+ "step": 382000
1156
+ },
1157
+ {
1158
+ "epoch": 8.89,
1159
+ "learning_rate": 3.3395370478073744e-05,
1160
+ "loss": 0.6248,
1161
+ "step": 384000
1162
+ },
1163
+ {
1164
+ "epoch": 8.94,
1165
+ "learning_rate": 3.199953472141444e-05,
1166
+ "loss": 0.6199,
1167
+ "step": 386000
1168
+ },
1169
+ {
1170
+ "epoch": 8.98,
1171
+ "learning_rate": 3.0603698964755146e-05,
1172
+ "loss": 0.6173,
1173
+ "step": 388000
1174
+ }
1175
+ ],
1176
+ "logging_steps": 2000,
1177
+ "max_steps": 431850,
1178
+ "num_train_epochs": 10,
1179
+ "save_steps": 500,
1180
+ "total_flos": 7.94798748892796e+20,
1181
+ "trial_name": null,
1182
+ "trial_params": null
1183
+ }
checkpoint-388669/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:783723dff03848f1deb1a19f34bdbddf1b8e878418b808096c108f32e10ad612
3
+ size 4600
checkpoint-388669/vocab.json ADDED
@@ -0,0 +1,200 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "\b": 0,
3
+ "\"": 2,
4
+ "#": 3,
5
+ ";": 4,
6
+ "=": 5,
7
+ "[": 6,
8
+ "[PAD]": 197,
9
+ "[UNK]": 196,
10
+ "\\": 7,
11
+ "]": 8,
12
+ "_": 9,
13
+ "`": 10,
14
+ "a": 11,
15
+ "b": 12,
16
+ "c": 13,
17
+ "d": 14,
18
+ "e": 15,
19
+ "f": 16,
20
+ "g": 17,
21
+ "h": 18,
22
+ "i": 19,
23
+ "j": 20,
24
+ "k": 21,
25
+ "l": 22,
26
+ "m": 23,
27
+ "n": 24,
28
+ "o": 25,
29
+ "p": 26,
30
+ "q": 27,
31
+ "r": 28,
32
+ "s": 29,
33
+ "t": 30,
34
+ "u": 31,
35
+ "v": 32,
36
+ "w": 33,
37
+ "x": 34,
38
+ "y": 35,
39
+ "z": 36,
40
+ "|": 1,
41
+ "£": 37,
42
+ "°": 38,
43
+ "½": 39,
44
+ "ß": 40,
45
+ "à": 41,
46
+ "â": 42,
47
+ "ã": 43,
48
+ "ä": 44,
49
+ "ç": 45,
50
+ "è": 46,
51
+ "é": 47,
52
+ "ê": 48,
53
+ "ë": 49,
54
+ "ì": 50,
55
+ "í": 51,
56
+ "ò": 52,
57
+ "ô": 53,
58
+ "ö": 54,
59
+ "ø": 55,
60
+ "ú": 56,
61
+ "ü": 57,
62
+ "ą": 58,
63
+ "č": 59,
64
+ "đ": 60,
65
+ "ę": 61,
66
+ "ğ": 62,
67
+ "ī": 63,
68
+ "ō": 64,
69
+ "œ": 65,
70
+ "ş": 66,
71
+ "ż": 67,
72
+ "ž": 68,
73
+ "ǒ": 69,
74
+ "ǔ": 70,
75
+ "˚": 71,
76
+ "υ": 72,
77
+ "а": 73,
78
+ "б": 74,
79
+ "в": 75,
80
+ "г": 76,
81
+ "д": 77,
82
+ "е": 78,
83
+ "ж": 79,
84
+ "и": 80,
85
+ "й": 81,
86
+ "к": 82,
87
+ "л": 83,
88
+ "м": 84,
89
+ "н": 85,
90
+ "о": 86,
91
+ "п": 87,
92
+ "р": 88,
93
+ "с": 89,
94
+ "т": 90,
95
+ "у": 91,
96
+ "ф": 92,
97
+ "х": 93,
98
+ "ц": 94,
99
+ "ч": 95,
100
+ "ш": 96,
101
+ "щ": 97,
102
+ "ы": 98,
103
+ "ь": 99,
104
+ "э": 100,
105
+ "ю": 101,
106
+ "я": 102,
107
+ "ا": 103,
108
+ "ر": 104,
109
+ "ص": 105,
110
+ "َ": 106,
111
+ "ّ": 107,
112
+ "น": 108,
113
+ "อ": 109,
114
+ "ế": 110,
115
+ "ố": 111,
116
+ "​": 112,
117
+ "‑": 113,
118
+ "–": 114,
119
+ "—": 115,
120
+ "…": 116,
121
+ "⁄": 117,
122
+ "℃": 118,
123
+ "℉": 119,
124
+ "™": 120,
125
+ "≈": 121,
126
+ "♫": 122,
127
+ "【": 123,
128
+ "あ": 124,
129
+ "い": 125,
130
+ "う": 126,
131
+ "が": 127,
132
+ "じ": 128,
133
+ "す": 129,
134
+ "た": 130,
135
+ "っ": 131,
136
+ "つ": 132,
137
+ "と": 133,
138
+ "な": 134,
139
+ "は": 135,
140
+ "ま": 136,
141
+ "み": 137,
142
+ "り": 138,
143
+ "ん": 139,
144
+ "シ": 140,
145
+ "ッ": 141,
146
+ "デ": 142,
147
+ "ニ": 143,
148
+ "ュ": 144,
149
+ "ㄴ": 145,
150
+ "ㅒ": 146,
151
+ "ㅜ": 147,
152
+ "ㅠ": 148,
153
+ "ㅣ": 149,
154
+ "不": 150,
155
+ "京": 151,
156
+ "張": 152,
157
+ "按": 153,
158
+ "最": 154,
159
+ "理": 155,
160
+ "要": 156,
161
+ "讚": 157,
162
+ "近": 158,
163
+ "頑": 159,
164
+ "가": 160,
165
+ "간": 161,
166
+ "고": 162,
167
+ "구": 163,
168
+ "금": 164,
169
+ "꼭": 165,
170
+ "나": 166,
171
+ "녕": 167,
172
+ "님": 168,
173
+ "다": 169,
174
+ "든": 170,
175
+ "발": 171,
176
+ "배": 172,
177
+ "뿌": 173,
178
+ "상": 174,
179
+ "설": 175,
180
+ "아": 176,
181
+ "안": 177,
182
+ "어": 178,
183
+ "영": 179,
184
+ "왜": 180,
185
+ "요": 181,
186
+ "유": 182,
187
+ "은": 183,
188
+ "을": 184,
189
+ "음": 185,
190
+ "이": 186,
191
+ "인": 187,
192
+ "재": 188,
193
+ "절": 189,
194
+ "정": 190,
195
+ "쨌": 191,
196
+ "파": 192,
197
+ "하": 193,
198
+ "화": 194,
199
+ "�": 195
200
+ }
checkpoint-431850/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "</s>": 199,
3
+ "<s>": 198
4
+ }
checkpoint-431850/config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
3
+ "activation_dropout": 0.0,
4
+ "adapter_attn_dim": null,
5
+ "adapter_kernel_size": 3,
6
+ "adapter_stride": 2,
7
+ "add_adapter": false,
8
+ "apply_spec_augment": true,
9
+ "architectures": [
10
+ "Wav2Vec2ForCTC"
11
+ ],
12
+ "attention_dropout": 0.05,
13
+ "bos_token_id": 1,
14
+ "classifier_proj_size": 256,
15
+ "codevector_dim": 768,
16
+ "contrastive_logits_temperature": 0.1,
17
+ "conv_bias": true,
18
+ "conv_dim": [
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512,
25
+ 512
26
+ ],
27
+ "conv_kernel": [
28
+ 10,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 3,
33
+ 2,
34
+ 2
35
+ ],
36
+ "conv_stride": [
37
+ 5,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2,
43
+ 2
44
+ ],
45
+ "ctc_loss_reduction": "mean",
46
+ "ctc_zero_infinity": true,
47
+ "diversity_loss_weight": 0.1,
48
+ "do_stable_layer_norm": true,
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.05,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.0,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.05,
59
+ "hidden_size": 1024,
60
+ "initializer_range": 0.02,
61
+ "intermediate_size": 4096,
62
+ "layer_norm_eps": 1e-05,
63
+ "layerdrop": 0.05,
64
+ "mask_channel_length": 10,
65
+ "mask_channel_min_space": 1,
66
+ "mask_channel_other": 0.0,
67
+ "mask_channel_prob": 0.0,
68
+ "mask_channel_selection": "static",
69
+ "mask_feature_length": 10,
70
+ "mask_feature_min_masks": 0,
71
+ "mask_feature_prob": 0.0,
72
+ "mask_time_length": 10,
73
+ "mask_time_min_masks": 2,
74
+ "mask_time_min_space": 1,
75
+ "mask_time_other": 0.0,
76
+ "mask_time_prob": 0.05,
77
+ "mask_time_selection": "static",
78
+ "model_type": "wav2vec2",
79
+ "num_adapter_layers": 3,
80
+ "num_attention_heads": 16,
81
+ "num_codevector_groups": 2,
82
+ "num_codevectors_per_group": 320,
83
+ "num_conv_pos_embedding_groups": 16,
84
+ "num_conv_pos_embeddings": 128,
85
+ "num_feat_extract_layers": 7,
86
+ "num_hidden_layers": 24,
87
+ "num_negatives": 100,
88
+ "output_hidden_size": 1024,
89
+ "pad_token_id": 197,
90
+ "proj_codevector_dim": 768,
91
+ "tdnn_dilation": [
92
+ 1,
93
+ 2,
94
+ 3,
95
+ 1,
96
+ 1
97
+ ],
98
+ "tdnn_dim": [
99
+ 512,
100
+ 512,
101
+ 512,
102
+ 512,
103
+ 1500
104
+ ],
105
+ "tdnn_kernel": [
106
+ 5,
107
+ 3,
108
+ 3,
109
+ 1,
110
+ 1
111
+ ],
112
+ "torch_dtype": "float32",
113
+ "transformers_version": "4.35.0",
114
+ "use_weighted_layer_sum": false,
115
+ "vocab_size": 200,
116
+ "xvector_output_dim": 512
117
+ }
checkpoint-431850/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:724e00446fba78a52abbddd258cfc4e543ca0c2d96145e6ddf9d911cf63d42f9
3
+ size 1262627480
checkpoint-431850/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc327eeb61db3b4c9722ecbda77ef136b769107e7b32946cfad2563c48854b86
3
+ size 2491799734
checkpoint-431850/preprocessor_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "processor_class": "Wav2Vec2Processor",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 16000
10
+ }
checkpoint-431850/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3261a8225a0dca17952bcca03d33ee5ef9019985a54211a12fe8ccf280deaab4
3
+ size 15024
checkpoint-431850/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c8a530fcd348d29af66aca115d09bafe1d547fc6e1b31593d9d8515a70a090a
3
+ size 15024
checkpoint-431850/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce7d81a5aba94e736972ae26a754b51015c17d9a3ef4a53f375eae1a3ca5f1f1
3
+ size 15088
checkpoint-431850/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57519a36b8b2eb51a399677cd5d8b70200eaca94ae24c95acc3e22d1ae20d992
3
+ size 15024
checkpoint-431850/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf7918d7e5736351d58de8a54e5187fb5812af472e95cbacf8d8b84611326131
3
+ size 1064
checkpoint-431850/special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": true,
19
+ "normalized": false,
20
+ "rstrip": true,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "[UNK]",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": true,
28
+ "single_word": false
29
+ }
30
+ }
checkpoint-431850/tokenizer_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "196": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "197": {
12
+ "content": "[PAD]",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "198": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": true,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "199": {
28
+ "content": "</s>",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ }
35
+ },
36
+ "bos_token": "<s>",
37
+ "clean_up_tokenization_spaces": true,
38
+ "config": null,
39
+ "do_lower_case": false,
40
+ "eos_token": "</s>",
41
+ "model_max_length": 1000000000000000019884624838656,
42
+ "pad_token": "[PAD]",
43
+ "processor_class": "Wav2Vec2Processor",
44
+ "replace_word_delimiter_char": " ",
45
+ "target_lang": null,
46
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
47
+ "tokenizer_type": "wav2vec2",
48
+ "trust_remote_code": false,
49
+ "unk_token": "[UNK]",
50
+ "word_delimiter_token": "|"
51
+ }
checkpoint-431850/trainer_state.json ADDED
@@ -0,0 +1,1309 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.99988422039805,
5
+ "eval_steps": 500,
6
+ "global_step": 431850,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05,
13
+ "learning_rate": 0.0003,
14
+ "loss": 3.5282,
15
+ "step": 2000
16
+ },
17
+ {
18
+ "epoch": 0.09,
19
+ "learning_rate": 0.00029860416424334066,
20
+ "loss": 1.3191,
21
+ "step": 4000
22
+ },
23
+ {
24
+ "epoch": 0.14,
25
+ "learning_rate": 0.0002972083284866814,
26
+ "loss": 1.2226,
27
+ "step": 6000
28
+ },
29
+ {
30
+ "epoch": 0.19,
31
+ "learning_rate": 0.00029581249273002205,
32
+ "loss": 1.1832,
33
+ "step": 8000
34
+ },
35
+ {
36
+ "epoch": 0.23,
37
+ "learning_rate": 0.0002944166569733628,
38
+ "loss": 1.1595,
39
+ "step": 10000
40
+ },
41
+ {
42
+ "epoch": 0.28,
43
+ "learning_rate": 0.0002930208212167035,
44
+ "loss": 1.1368,
45
+ "step": 12000
46
+ },
47
+ {
48
+ "epoch": 0.32,
49
+ "learning_rate": 0.00029162498546004417,
50
+ "loss": 1.1176,
51
+ "step": 14000
52
+ },
53
+ {
54
+ "epoch": 0.37,
55
+ "learning_rate": 0.00029022914970338486,
56
+ "loss": 1.1044,
57
+ "step": 16000
58
+ },
59
+ {
60
+ "epoch": 0.42,
61
+ "learning_rate": 0.00028883331394672555,
62
+ "loss": 1.0954,
63
+ "step": 18000
64
+ },
65
+ {
66
+ "epoch": 0.46,
67
+ "learning_rate": 0.0002874374781900663,
68
+ "loss": 1.088,
69
+ "step": 20000
70
+ },
71
+ {
72
+ "epoch": 0.51,
73
+ "learning_rate": 0.000286041642433407,
74
+ "loss": 1.0765,
75
+ "step": 22000
76
+ },
77
+ {
78
+ "epoch": 0.56,
79
+ "learning_rate": 0.0002846458066767477,
80
+ "loss": 1.069,
81
+ "step": 24000
82
+ },
83
+ {
84
+ "epoch": 0.6,
85
+ "learning_rate": 0.00028324997092008837,
86
+ "loss": 1.0623,
87
+ "step": 26000
88
+ },
89
+ {
90
+ "epoch": 0.65,
91
+ "learning_rate": 0.00028185413516342906,
92
+ "loss": 1.0653,
93
+ "step": 28000
94
+ },
95
+ {
96
+ "epoch": 0.69,
97
+ "learning_rate": 0.0002804582994067698,
98
+ "loss": 1.0467,
99
+ "step": 30000
100
+ },
101
+ {
102
+ "epoch": 0.74,
103
+ "learning_rate": 0.00027906246365011044,
104
+ "loss": 1.0395,
105
+ "step": 32000
106
+ },
107
+ {
108
+ "epoch": 0.79,
109
+ "learning_rate": 0.0002776666278934512,
110
+ "loss": 1.0419,
111
+ "step": 34000
112
+ },
113
+ {
114
+ "epoch": 0.83,
115
+ "learning_rate": 0.0002762707921367919,
116
+ "loss": 1.0385,
117
+ "step": 36000
118
+ },
119
+ {
120
+ "epoch": 0.88,
121
+ "learning_rate": 0.00027487495638013257,
122
+ "loss": 1.0375,
123
+ "step": 38000
124
+ },
125
+ {
126
+ "epoch": 0.93,
127
+ "learning_rate": 0.00027347912062347326,
128
+ "loss": 1.0211,
129
+ "step": 40000
130
+ },
131
+ {
132
+ "epoch": 0.97,
133
+ "learning_rate": 0.00027208328486681395,
134
+ "loss": 1.0248,
135
+ "step": 42000
136
+ },
137
+ {
138
+ "epoch": 1.02,
139
+ "learning_rate": 0.0002706874491101547,
140
+ "loss": 1.0071,
141
+ "step": 44000
142
+ },
143
+ {
144
+ "epoch": 1.07,
145
+ "learning_rate": 0.0002692916133534954,
146
+ "loss": 0.9825,
147
+ "step": 46000
148
+ },
149
+ {
150
+ "epoch": 1.11,
151
+ "learning_rate": 0.0002678957775968361,
152
+ "loss": 0.983,
153
+ "step": 48000
154
+ },
155
+ {
156
+ "epoch": 1.16,
157
+ "learning_rate": 0.00026649994184017677,
158
+ "loss": 0.9831,
159
+ "step": 50000
160
+ },
161
+ {
162
+ "epoch": 1.2,
163
+ "learning_rate": 0.00026510410608351746,
164
+ "loss": 0.9792,
165
+ "step": 52000
166
+ },
167
+ {
168
+ "epoch": 1.25,
169
+ "learning_rate": 0.0002637082703268582,
170
+ "loss": 0.9744,
171
+ "step": 54000
172
+ },
173
+ {
174
+ "epoch": 1.3,
175
+ "learning_rate": 0.0002623124345701989,
176
+ "loss": 0.9668,
177
+ "step": 56000
178
+ },
179
+ {
180
+ "epoch": 1.34,
181
+ "learning_rate": 0.0002609165988135396,
182
+ "loss": 0.971,
183
+ "step": 58000
184
+ },
185
+ {
186
+ "epoch": 1.39,
187
+ "learning_rate": 0.0002595207630568803,
188
+ "loss": 0.9789,
189
+ "step": 60000
190
+ },
191
+ {
192
+ "epoch": 1.44,
193
+ "learning_rate": 0.000258124927300221,
194
+ "loss": 0.9794,
195
+ "step": 62000
196
+ },
197
+ {
198
+ "epoch": 1.48,
199
+ "learning_rate": 0.00025672909154356166,
200
+ "loss": 0.9615,
201
+ "step": 64000
202
+ },
203
+ {
204
+ "epoch": 1.53,
205
+ "learning_rate": 0.0002553332557869024,
206
+ "loss": 0.9577,
207
+ "step": 66000
208
+ },
209
+ {
210
+ "epoch": 1.57,
211
+ "learning_rate": 0.0002539374200302431,
212
+ "loss": 0.9573,
213
+ "step": 68000
214
+ },
215
+ {
216
+ "epoch": 1.62,
217
+ "learning_rate": 0.0002525415842735838,
218
+ "loss": 0.9633,
219
+ "step": 70000
220
+ },
221
+ {
222
+ "epoch": 1.67,
223
+ "learning_rate": 0.0002511457485169245,
224
+ "loss": 0.946,
225
+ "step": 72000
226
+ },
227
+ {
228
+ "epoch": 1.71,
229
+ "learning_rate": 0.00024974991276026517,
230
+ "loss": 0.9528,
231
+ "step": 74000
232
+ },
233
+ {
234
+ "epoch": 1.76,
235
+ "learning_rate": 0.0002483540770036059,
236
+ "loss": 0.9474,
237
+ "step": 76000
238
+ },
239
+ {
240
+ "epoch": 1.81,
241
+ "learning_rate": 0.0002469582412469466,
242
+ "loss": 0.9431,
243
+ "step": 78000
244
+ },
245
+ {
246
+ "epoch": 1.85,
247
+ "learning_rate": 0.0002455624054902873,
248
+ "loss": 0.9469,
249
+ "step": 80000
250
+ },
251
+ {
252
+ "epoch": 1.9,
253
+ "learning_rate": 0.000244166569733628,
254
+ "loss": 0.9442,
255
+ "step": 82000
256
+ },
257
+ {
258
+ "epoch": 1.95,
259
+ "learning_rate": 0.0002427707339769687,
260
+ "loss": 0.9387,
261
+ "step": 84000
262
+ },
263
+ {
264
+ "epoch": 1.99,
265
+ "learning_rate": 0.0002413748982203094,
266
+ "loss": 0.9341,
267
+ "step": 86000
268
+ },
269
+ {
270
+ "epoch": 2.04,
271
+ "learning_rate": 0.00023997906246365008,
272
+ "loss": 0.9124,
273
+ "step": 88000
274
+ },
275
+ {
276
+ "epoch": 2.08,
277
+ "learning_rate": 0.00023858322670699077,
278
+ "loss": 0.8962,
279
+ "step": 90000
280
+ },
281
+ {
282
+ "epoch": 2.13,
283
+ "learning_rate": 0.0002371873909503315,
284
+ "loss": 0.8974,
285
+ "step": 92000
286
+ },
287
+ {
288
+ "epoch": 2.18,
289
+ "learning_rate": 0.0002357915551936722,
290
+ "loss": 0.9122,
291
+ "step": 94000
292
+ },
293
+ {
294
+ "epoch": 2.22,
295
+ "learning_rate": 0.00023439571943701287,
296
+ "loss": 0.8977,
297
+ "step": 96000
298
+ },
299
+ {
300
+ "epoch": 2.27,
301
+ "learning_rate": 0.0002329998836803536,
302
+ "loss": 0.8972,
303
+ "step": 98000
304
+ },
305
+ {
306
+ "epoch": 2.32,
307
+ "learning_rate": 0.00023160404792369428,
308
+ "loss": 0.8969,
309
+ "step": 100000
310
+ },
311
+ {
312
+ "epoch": 2.36,
313
+ "learning_rate": 0.000230208212167035,
314
+ "loss": 0.9021,
315
+ "step": 102000
316
+ },
317
+ {
318
+ "epoch": 2.41,
319
+ "learning_rate": 0.00022881237641037572,
320
+ "loss": 0.9083,
321
+ "step": 104000
322
+ },
323
+ {
324
+ "epoch": 2.45,
325
+ "learning_rate": 0.00022741654065371638,
326
+ "loss": 0.8928,
327
+ "step": 106000
328
+ },
329
+ {
330
+ "epoch": 2.5,
331
+ "learning_rate": 0.0002260207048970571,
332
+ "loss": 0.8871,
333
+ "step": 108000
334
+ },
335
+ {
336
+ "epoch": 2.55,
337
+ "learning_rate": 0.0002246248691403978,
338
+ "loss": 0.9017,
339
+ "step": 110000
340
+ },
341
+ {
342
+ "epoch": 2.59,
343
+ "learning_rate": 0.0002232290333837385,
344
+ "loss": 0.8968,
345
+ "step": 112000
346
+ },
347
+ {
348
+ "epoch": 2.64,
349
+ "learning_rate": 0.00022183319762707917,
350
+ "loss": 0.8813,
351
+ "step": 114000
352
+ },
353
+ {
354
+ "epoch": 2.69,
355
+ "learning_rate": 0.0002204373618704199,
356
+ "loss": 0.8807,
357
+ "step": 116000
358
+ },
359
+ {
360
+ "epoch": 2.73,
361
+ "learning_rate": 0.0002190415261137606,
362
+ "loss": 0.8816,
363
+ "step": 118000
364
+ },
365
+ {
366
+ "epoch": 2.78,
367
+ "learning_rate": 0.0002176456903571013,
368
+ "loss": 0.8776,
369
+ "step": 120000
370
+ },
371
+ {
372
+ "epoch": 2.83,
373
+ "learning_rate": 0.000216249854600442,
374
+ "loss": 0.8738,
375
+ "step": 122000
376
+ },
377
+ {
378
+ "epoch": 2.87,
379
+ "learning_rate": 0.00021485401884378268,
380
+ "loss": 0.8826,
381
+ "step": 124000
382
+ },
383
+ {
384
+ "epoch": 2.92,
385
+ "learning_rate": 0.0002134581830871234,
386
+ "loss": 0.893,
387
+ "step": 126000
388
+ },
389
+ {
390
+ "epoch": 2.96,
391
+ "learning_rate": 0.00021206234733046412,
392
+ "loss": 0.8753,
393
+ "step": 128000
394
+ },
395
+ {
396
+ "epoch": 3.01,
397
+ "learning_rate": 0.00021066651157380478,
398
+ "loss": 0.8671,
399
+ "step": 130000
400
+ },
401
+ {
402
+ "epoch": 3.06,
403
+ "learning_rate": 0.0002092706758171455,
404
+ "loss": 0.8497,
405
+ "step": 132000
406
+ },
407
+ {
408
+ "epoch": 3.1,
409
+ "learning_rate": 0.00020787484006048622,
410
+ "loss": 0.8443,
411
+ "step": 134000
412
+ },
413
+ {
414
+ "epoch": 3.15,
415
+ "learning_rate": 0.0002064790043038269,
416
+ "loss": 0.8477,
417
+ "step": 136000
418
+ },
419
+ {
420
+ "epoch": 3.2,
421
+ "learning_rate": 0.0002050831685471676,
422
+ "loss": 0.8465,
423
+ "step": 138000
424
+ },
425
+ {
426
+ "epoch": 3.24,
427
+ "learning_rate": 0.0002036873327905083,
428
+ "loss": 0.8501,
429
+ "step": 140000
430
+ },
431
+ {
432
+ "epoch": 3.29,
433
+ "learning_rate": 0.000202291497033849,
434
+ "loss": 0.8451,
435
+ "step": 142000
436
+ },
437
+ {
438
+ "epoch": 3.33,
439
+ "learning_rate": 0.00020089566127718972,
440
+ "loss": 0.84,
441
+ "step": 144000
442
+ },
443
+ {
444
+ "epoch": 3.38,
445
+ "learning_rate": 0.0001994998255205304,
446
+ "loss": 0.8482,
447
+ "step": 146000
448
+ },
449
+ {
450
+ "epoch": 3.43,
451
+ "learning_rate": 0.0001981039897638711,
452
+ "loss": 0.8383,
453
+ "step": 148000
454
+ },
455
+ {
456
+ "epoch": 3.47,
457
+ "learning_rate": 0.0001967081540072118,
458
+ "loss": 0.8353,
459
+ "step": 150000
460
+ },
461
+ {
462
+ "epoch": 3.52,
463
+ "learning_rate": 0.00019531231825055251,
464
+ "loss": 0.8334,
465
+ "step": 152000
466
+ },
467
+ {
468
+ "epoch": 3.57,
469
+ "learning_rate": 0.00019391648249389318,
470
+ "loss": 0.8446,
471
+ "step": 154000
472
+ },
473
+ {
474
+ "epoch": 3.61,
475
+ "learning_rate": 0.0001925206467372339,
476
+ "loss": 0.8373,
477
+ "step": 156000
478
+ },
479
+ {
480
+ "epoch": 3.66,
481
+ "learning_rate": 0.00019112481098057461,
482
+ "loss": 0.8412,
483
+ "step": 158000
484
+ },
485
+ {
486
+ "epoch": 3.7,
487
+ "learning_rate": 0.0001897289752239153,
488
+ "loss": 0.8286,
489
+ "step": 160000
490
+ },
491
+ {
492
+ "epoch": 3.75,
493
+ "learning_rate": 0.000188333139467256,
494
+ "loss": 0.8327,
495
+ "step": 162000
496
+ },
497
+ {
498
+ "epoch": 3.8,
499
+ "learning_rate": 0.0001869373037105967,
500
+ "loss": 0.8426,
501
+ "step": 164000
502
+ },
503
+ {
504
+ "epoch": 3.84,
505
+ "learning_rate": 0.0001855414679539374,
506
+ "loss": 0.8291,
507
+ "step": 166000
508
+ },
509
+ {
510
+ "epoch": 3.89,
511
+ "learning_rate": 0.00018414563219727812,
512
+ "loss": 0.8271,
513
+ "step": 168000
514
+ },
515
+ {
516
+ "epoch": 3.94,
517
+ "learning_rate": 0.00018274979644061879,
518
+ "loss": 0.8278,
519
+ "step": 170000
520
+ },
521
+ {
522
+ "epoch": 3.98,
523
+ "learning_rate": 0.0001813539606839595,
524
+ "loss": 0.8307,
525
+ "step": 172000
526
+ },
527
+ {
528
+ "epoch": 4.03,
529
+ "learning_rate": 0.0001799581249273002,
530
+ "loss": 0.8109,
531
+ "step": 174000
532
+ },
533
+ {
534
+ "epoch": 4.08,
535
+ "learning_rate": 0.0001785622891706409,
536
+ "loss": 0.8027,
537
+ "step": 176000
538
+ },
539
+ {
540
+ "epoch": 4.12,
541
+ "learning_rate": 0.0001771664534139816,
542
+ "loss": 0.7955,
543
+ "step": 178000
544
+ },
545
+ {
546
+ "epoch": 4.17,
547
+ "learning_rate": 0.0001757706176573223,
548
+ "loss": 0.7888,
549
+ "step": 180000
550
+ },
551
+ {
552
+ "epoch": 4.21,
553
+ "learning_rate": 0.000174374781900663,
554
+ "loss": 0.799,
555
+ "step": 182000
556
+ },
557
+ {
558
+ "epoch": 4.26,
559
+ "learning_rate": 0.0001729789461440037,
560
+ "loss": 0.7949,
561
+ "step": 184000
562
+ },
563
+ {
564
+ "epoch": 4.31,
565
+ "learning_rate": 0.0001715831103873444,
566
+ "loss": 0.7913,
567
+ "step": 186000
568
+ },
569
+ {
570
+ "epoch": 4.35,
571
+ "learning_rate": 0.0001701872746306851,
572
+ "loss": 0.7862,
573
+ "step": 188000
574
+ },
575
+ {
576
+ "epoch": 4.4,
577
+ "learning_rate": 0.0001687914388740258,
578
+ "loss": 0.7882,
579
+ "step": 190000
580
+ },
581
+ {
582
+ "epoch": 4.45,
583
+ "learning_rate": 0.00016739560311736652,
584
+ "loss": 0.8017,
585
+ "step": 192000
586
+ },
587
+ {
588
+ "epoch": 4.49,
589
+ "learning_rate": 0.00016599976736070724,
590
+ "loss": 0.787,
591
+ "step": 194000
592
+ },
593
+ {
594
+ "epoch": 4.54,
595
+ "learning_rate": 0.0001646039316040479,
596
+ "loss": 0.7975,
597
+ "step": 196000
598
+ },
599
+ {
600
+ "epoch": 4.58,
601
+ "learning_rate": 0.00016320809584738862,
602
+ "loss": 0.7938,
603
+ "step": 198000
604
+ },
605
+ {
606
+ "epoch": 4.63,
607
+ "learning_rate": 0.0001618122600907293,
608
+ "loss": 0.799,
609
+ "step": 200000
610
+ },
611
+ {
612
+ "epoch": 4.68,
613
+ "learning_rate": 0.00016041642433407003,
614
+ "loss": 0.7811,
615
+ "step": 202000
616
+ },
617
+ {
618
+ "epoch": 4.72,
619
+ "learning_rate": 0.0001590205885774107,
620
+ "loss": 0.7941,
621
+ "step": 204000
622
+ },
623
+ {
624
+ "epoch": 4.77,
625
+ "learning_rate": 0.0001576247528207514,
626
+ "loss": 0.7918,
627
+ "step": 206000
628
+ },
629
+ {
630
+ "epoch": 4.82,
631
+ "learning_rate": 0.00015622891706409213,
632
+ "loss": 0.8006,
633
+ "step": 208000
634
+ },
635
+ {
636
+ "epoch": 4.86,
637
+ "learning_rate": 0.00015483308130743282,
638
+ "loss": 0.7896,
639
+ "step": 210000
640
+ },
641
+ {
642
+ "epoch": 4.91,
643
+ "learning_rate": 0.0001534372455507735,
644
+ "loss": 0.7944,
645
+ "step": 212000
646
+ },
647
+ {
648
+ "epoch": 4.96,
649
+ "learning_rate": 0.0001520414097941142,
650
+ "loss": 0.7869,
651
+ "step": 214000
652
+ },
653
+ {
654
+ "epoch": 5.0,
655
+ "learning_rate": 0.00015064557403745492,
656
+ "loss": 0.7847,
657
+ "step": 216000
658
+ },
659
+ {
660
+ "epoch": 5.05,
661
+ "learning_rate": 0.0001492497382807956,
662
+ "loss": 0.7577,
663
+ "step": 218000
664
+ },
665
+ {
666
+ "epoch": 5.09,
667
+ "learning_rate": 0.00014785390252413633,
668
+ "loss": 0.7527,
669
+ "step": 220000
670
+ },
671
+ {
672
+ "epoch": 5.14,
673
+ "learning_rate": 0.00014645806676747702,
674
+ "loss": 0.7554,
675
+ "step": 222000
676
+ },
677
+ {
678
+ "epoch": 5.19,
679
+ "learning_rate": 0.0001450622310108177,
680
+ "loss": 0.7633,
681
+ "step": 224000
682
+ },
683
+ {
684
+ "epoch": 5.23,
685
+ "learning_rate": 0.0001436663952541584,
686
+ "loss": 0.754,
687
+ "step": 226000
688
+ },
689
+ {
690
+ "epoch": 5.28,
691
+ "learning_rate": 0.00014227055949749912,
692
+ "loss": 0.7438,
693
+ "step": 228000
694
+ },
695
+ {
696
+ "epoch": 5.33,
697
+ "learning_rate": 0.0001408747237408398,
698
+ "loss": 0.7514,
699
+ "step": 230000
700
+ },
701
+ {
702
+ "epoch": 5.37,
703
+ "learning_rate": 0.00013947888798418053,
704
+ "loss": 0.7522,
705
+ "step": 232000
706
+ },
707
+ {
708
+ "epoch": 5.42,
709
+ "learning_rate": 0.00013808305222752122,
710
+ "loss": 0.7525,
711
+ "step": 234000
712
+ },
713
+ {
714
+ "epoch": 5.46,
715
+ "learning_rate": 0.00013668721647086193,
716
+ "loss": 0.7476,
717
+ "step": 236000
718
+ },
719
+ {
720
+ "epoch": 5.51,
721
+ "learning_rate": 0.00013529138071420263,
722
+ "loss": 0.747,
723
+ "step": 238000
724
+ },
725
+ {
726
+ "epoch": 5.56,
727
+ "learning_rate": 0.00013389554495754332,
728
+ "loss": 0.7505,
729
+ "step": 240000
730
+ },
731
+ {
732
+ "epoch": 5.6,
733
+ "learning_rate": 0.000132499709200884,
734
+ "loss": 0.7501,
735
+ "step": 242000
736
+ },
737
+ {
738
+ "epoch": 5.65,
739
+ "learning_rate": 0.00013110387344422472,
740
+ "loss": 0.7568,
741
+ "step": 244000
742
+ },
743
+ {
744
+ "epoch": 5.7,
745
+ "learning_rate": 0.00012970803768756542,
746
+ "loss": 0.7383,
747
+ "step": 246000
748
+ },
749
+ {
750
+ "epoch": 5.74,
751
+ "learning_rate": 0.00012831220193090613,
752
+ "loss": 0.7535,
753
+ "step": 248000
754
+ },
755
+ {
756
+ "epoch": 5.79,
757
+ "learning_rate": 0.00012691636617424682,
758
+ "loss": 0.751,
759
+ "step": 250000
760
+ },
761
+ {
762
+ "epoch": 5.84,
763
+ "learning_rate": 0.00012552053041758751,
764
+ "loss": 0.7396,
765
+ "step": 252000
766
+ },
767
+ {
768
+ "epoch": 5.88,
769
+ "learning_rate": 0.0001241246946609282,
770
+ "loss": 0.7448,
771
+ "step": 254000
772
+ },
773
+ {
774
+ "epoch": 5.93,
775
+ "learning_rate": 0.00012272885890426892,
776
+ "loss": 0.7521,
777
+ "step": 256000
778
+ },
779
+ {
780
+ "epoch": 5.97,
781
+ "learning_rate": 0.00012133302314760961,
782
+ "loss": 0.7472,
783
+ "step": 258000
784
+ },
785
+ {
786
+ "epoch": 6.02,
787
+ "learning_rate": 0.00011993718739095032,
788
+ "loss": 0.7344,
789
+ "step": 260000
790
+ },
791
+ {
792
+ "epoch": 6.07,
793
+ "learning_rate": 0.00011854135163429101,
794
+ "loss": 0.7143,
795
+ "step": 262000
796
+ },
797
+ {
798
+ "epoch": 6.11,
799
+ "learning_rate": 0.00011714551587763173,
800
+ "loss": 0.717,
801
+ "step": 264000
802
+ },
803
+ {
804
+ "epoch": 6.16,
805
+ "learning_rate": 0.00011574968012097242,
806
+ "loss": 0.7194,
807
+ "step": 266000
808
+ },
809
+ {
810
+ "epoch": 6.21,
811
+ "learning_rate": 0.00011435384436431312,
812
+ "loss": 0.7206,
813
+ "step": 268000
814
+ },
815
+ {
816
+ "epoch": 6.25,
817
+ "learning_rate": 0.00011295800860765381,
818
+ "loss": 0.7148,
819
+ "step": 270000
820
+ },
821
+ {
822
+ "epoch": 6.3,
823
+ "learning_rate": 0.00011156217285099453,
824
+ "loss": 0.7137,
825
+ "step": 272000
826
+ },
827
+ {
828
+ "epoch": 6.34,
829
+ "learning_rate": 0.00011016633709433522,
830
+ "loss": 0.7168,
831
+ "step": 274000
832
+ },
833
+ {
834
+ "epoch": 6.39,
835
+ "learning_rate": 0.00010877050133767593,
836
+ "loss": 0.7097,
837
+ "step": 276000
838
+ },
839
+ {
840
+ "epoch": 6.44,
841
+ "learning_rate": 0.00010737466558101662,
842
+ "loss": 0.7084,
843
+ "step": 278000
844
+ },
845
+ {
846
+ "epoch": 6.48,
847
+ "learning_rate": 0.00010597882982435732,
848
+ "loss": 0.7101,
849
+ "step": 280000
850
+ },
851
+ {
852
+ "epoch": 6.53,
853
+ "learning_rate": 0.00010458299406769801,
854
+ "loss": 0.7117,
855
+ "step": 282000
856
+ },
857
+ {
858
+ "epoch": 6.58,
859
+ "learning_rate": 0.00010318715831103873,
860
+ "loss": 0.7067,
861
+ "step": 284000
862
+ },
863
+ {
864
+ "epoch": 6.62,
865
+ "learning_rate": 0.00010179132255437942,
866
+ "loss": 0.7026,
867
+ "step": 286000
868
+ },
869
+ {
870
+ "epoch": 6.67,
871
+ "learning_rate": 0.00010039548679772013,
872
+ "loss": 0.7144,
873
+ "step": 288000
874
+ },
875
+ {
876
+ "epoch": 6.72,
877
+ "learning_rate": 9.899965104106083e-05,
878
+ "loss": 0.7118,
879
+ "step": 290000
880
+ },
881
+ {
882
+ "epoch": 6.76,
883
+ "learning_rate": 9.760381528440152e-05,
884
+ "loss": 0.7028,
885
+ "step": 292000
886
+ },
887
+ {
888
+ "epoch": 6.81,
889
+ "learning_rate": 9.620797952774224e-05,
890
+ "loss": 0.7044,
891
+ "step": 294000
892
+ },
893
+ {
894
+ "epoch": 6.85,
895
+ "learning_rate": 9.481214377108293e-05,
896
+ "loss": 0.704,
897
+ "step": 296000
898
+ },
899
+ {
900
+ "epoch": 6.9,
901
+ "learning_rate": 9.341630801442363e-05,
902
+ "loss": 0.7049,
903
+ "step": 298000
904
+ },
905
+ {
906
+ "epoch": 6.95,
907
+ "learning_rate": 9.202047225776432e-05,
908
+ "loss": 0.6937,
909
+ "step": 300000
910
+ },
911
+ {
912
+ "epoch": 6.99,
913
+ "learning_rate": 9.062463650110504e-05,
914
+ "loss": 0.7096,
915
+ "step": 302000
916
+ },
917
+ {
918
+ "epoch": 7.04,
919
+ "learning_rate": 8.922880074444573e-05,
920
+ "loss": 0.6749,
921
+ "step": 304000
922
+ },
923
+ {
924
+ "epoch": 7.09,
925
+ "learning_rate": 8.783296498778644e-05,
926
+ "loss": 0.6718,
927
+ "step": 306000
928
+ },
929
+ {
930
+ "epoch": 7.13,
931
+ "learning_rate": 8.643712923112713e-05,
932
+ "loss": 0.6789,
933
+ "step": 308000
934
+ },
935
+ {
936
+ "epoch": 7.18,
937
+ "learning_rate": 8.504129347446783e-05,
938
+ "loss": 0.6805,
939
+ "step": 310000
940
+ },
941
+ {
942
+ "epoch": 7.22,
943
+ "learning_rate": 8.364545771780852e-05,
944
+ "loss": 0.6776,
945
+ "step": 312000
946
+ },
947
+ {
948
+ "epoch": 7.27,
949
+ "learning_rate": 8.224962196114924e-05,
950
+ "loss": 0.6734,
951
+ "step": 314000
952
+ },
953
+ {
954
+ "epoch": 7.32,
955
+ "learning_rate": 8.085378620448993e-05,
956
+ "loss": 0.6765,
957
+ "step": 316000
958
+ },
959
+ {
960
+ "epoch": 7.36,
961
+ "learning_rate": 7.945795044783064e-05,
962
+ "loss": 0.6709,
963
+ "step": 318000
964
+ },
965
+ {
966
+ "epoch": 7.41,
967
+ "learning_rate": 7.806211469117133e-05,
968
+ "loss": 0.6804,
969
+ "step": 320000
970
+ },
971
+ {
972
+ "epoch": 7.46,
973
+ "learning_rate": 7.666627893451203e-05,
974
+ "loss": 0.6688,
975
+ "step": 322000
976
+ },
977
+ {
978
+ "epoch": 7.5,
979
+ "learning_rate": 7.527044317785274e-05,
980
+ "loss": 0.6789,
981
+ "step": 324000
982
+ },
983
+ {
984
+ "epoch": 7.55,
985
+ "learning_rate": 7.387460742119343e-05,
986
+ "loss": 0.6699,
987
+ "step": 326000
988
+ },
989
+ {
990
+ "epoch": 7.6,
991
+ "learning_rate": 7.247877166453413e-05,
992
+ "loss": 0.662,
993
+ "step": 328000
994
+ },
995
+ {
996
+ "epoch": 7.64,
997
+ "learning_rate": 7.108293590787484e-05,
998
+ "loss": 0.6657,
999
+ "step": 330000
1000
+ },
1001
+ {
1002
+ "epoch": 7.69,
1003
+ "learning_rate": 6.968710015121553e-05,
1004
+ "loss": 0.6676,
1005
+ "step": 332000
1006
+ },
1007
+ {
1008
+ "epoch": 7.73,
1009
+ "learning_rate": 6.829126439455623e-05,
1010
+ "loss": 0.6698,
1011
+ "step": 334000
1012
+ },
1013
+ {
1014
+ "epoch": 7.78,
1015
+ "learning_rate": 6.689542863789693e-05,
1016
+ "loss": 0.6755,
1017
+ "step": 336000
1018
+ },
1019
+ {
1020
+ "epoch": 7.83,
1021
+ "learning_rate": 6.549959288123764e-05,
1022
+ "loss": 0.672,
1023
+ "step": 338000
1024
+ },
1025
+ {
1026
+ "epoch": 7.87,
1027
+ "learning_rate": 6.410375712457834e-05,
1028
+ "loss": 0.6616,
1029
+ "step": 340000
1030
+ },
1031
+ {
1032
+ "epoch": 7.92,
1033
+ "learning_rate": 6.270792136791903e-05,
1034
+ "loss": 0.6654,
1035
+ "step": 342000
1036
+ },
1037
+ {
1038
+ "epoch": 7.97,
1039
+ "learning_rate": 6.131208561125974e-05,
1040
+ "loss": 0.6606,
1041
+ "step": 344000
1042
+ },
1043
+ {
1044
+ "epoch": 8.01,
1045
+ "learning_rate": 5.991624985460044e-05,
1046
+ "loss": 0.6642,
1047
+ "step": 346000
1048
+ },
1049
+ {
1050
+ "epoch": 8.06,
1051
+ "learning_rate": 5.852041409794114e-05,
1052
+ "loss": 0.639,
1053
+ "step": 348000
1054
+ },
1055
+ {
1056
+ "epoch": 8.1,
1057
+ "learning_rate": 5.712457834128184e-05,
1058
+ "loss": 0.6463,
1059
+ "step": 350000
1060
+ },
1061
+ {
1062
+ "epoch": 8.15,
1063
+ "learning_rate": 5.572874258462254e-05,
1064
+ "loss": 0.6426,
1065
+ "step": 352000
1066
+ },
1067
+ {
1068
+ "epoch": 8.2,
1069
+ "learning_rate": 5.433290682796324e-05,
1070
+ "loss": 0.6356,
1071
+ "step": 354000
1072
+ },
1073
+ {
1074
+ "epoch": 8.24,
1075
+ "learning_rate": 5.2937071071303944e-05,
1076
+ "loss": 0.6336,
1077
+ "step": 356000
1078
+ },
1079
+ {
1080
+ "epoch": 8.29,
1081
+ "learning_rate": 5.154123531464464e-05,
1082
+ "loss": 0.631,
1083
+ "step": 358000
1084
+ },
1085
+ {
1086
+ "epoch": 8.34,
1087
+ "learning_rate": 5.014539955798534e-05,
1088
+ "loss": 0.6425,
1089
+ "step": 360000
1090
+ },
1091
+ {
1092
+ "epoch": 8.38,
1093
+ "learning_rate": 4.8749563801326044e-05,
1094
+ "loss": 0.6432,
1095
+ "step": 362000
1096
+ },
1097
+ {
1098
+ "epoch": 8.43,
1099
+ "learning_rate": 4.735372804466674e-05,
1100
+ "loss": 0.6449,
1101
+ "step": 364000
1102
+ },
1103
+ {
1104
+ "epoch": 8.48,
1105
+ "learning_rate": 4.595789228800744e-05,
1106
+ "loss": 0.635,
1107
+ "step": 366000
1108
+ },
1109
+ {
1110
+ "epoch": 8.52,
1111
+ "learning_rate": 4.456205653134814e-05,
1112
+ "loss": 0.6352,
1113
+ "step": 368000
1114
+ },
1115
+ {
1116
+ "epoch": 8.57,
1117
+ "learning_rate": 4.316622077468884e-05,
1118
+ "loss": 0.6281,
1119
+ "step": 370000
1120
+ },
1121
+ {
1122
+ "epoch": 8.61,
1123
+ "learning_rate": 4.1770385018029545e-05,
1124
+ "loss": 0.6291,
1125
+ "step": 372000
1126
+ },
1127
+ {
1128
+ "epoch": 8.66,
1129
+ "learning_rate": 4.037454926137024e-05,
1130
+ "loss": 0.6317,
1131
+ "step": 374000
1132
+ },
1133
+ {
1134
+ "epoch": 8.71,
1135
+ "learning_rate": 3.897871350471094e-05,
1136
+ "loss": 0.6296,
1137
+ "step": 376000
1138
+ },
1139
+ {
1140
+ "epoch": 8.75,
1141
+ "learning_rate": 3.7582877748051644e-05,
1142
+ "loss": 0.6279,
1143
+ "step": 378000
1144
+ },
1145
+ {
1146
+ "epoch": 8.8,
1147
+ "learning_rate": 3.618704199139234e-05,
1148
+ "loss": 0.6379,
1149
+ "step": 380000
1150
+ },
1151
+ {
1152
+ "epoch": 8.85,
1153
+ "learning_rate": 3.4791206234733046e-05,
1154
+ "loss": 0.6292,
1155
+ "step": 382000
1156
+ },
1157
+ {
1158
+ "epoch": 8.89,
1159
+ "learning_rate": 3.3395370478073744e-05,
1160
+ "loss": 0.6248,
1161
+ "step": 384000
1162
+ },
1163
+ {
1164
+ "epoch": 8.94,
1165
+ "learning_rate": 3.199953472141444e-05,
1166
+ "loss": 0.6199,
1167
+ "step": 386000
1168
+ },
1169
+ {
1170
+ "epoch": 8.98,
1171
+ "learning_rate": 3.0603698964755146e-05,
1172
+ "loss": 0.6173,
1173
+ "step": 388000
1174
+ },
1175
+ {
1176
+ "epoch": 9.03,
1177
+ "learning_rate": 2.9207863208095843e-05,
1178
+ "loss": 0.6129,
1179
+ "step": 390000
1180
+ },
1181
+ {
1182
+ "epoch": 9.08,
1183
+ "learning_rate": 2.7812027451436544e-05,
1184
+ "loss": 0.6013,
1185
+ "step": 392000
1186
+ },
1187
+ {
1188
+ "epoch": 9.12,
1189
+ "learning_rate": 2.6416191694777245e-05,
1190
+ "loss": 0.6045,
1191
+ "step": 394000
1192
+ },
1193
+ {
1194
+ "epoch": 9.17,
1195
+ "learning_rate": 2.5020355938117946e-05,
1196
+ "loss": 0.6134,
1197
+ "step": 396000
1198
+ },
1199
+ {
1200
+ "epoch": 9.22,
1201
+ "learning_rate": 2.3624520181458644e-05,
1202
+ "loss": 0.6056,
1203
+ "step": 398000
1204
+ },
1205
+ {
1206
+ "epoch": 9.26,
1207
+ "learning_rate": 2.2228684424799345e-05,
1208
+ "loss": 0.6112,
1209
+ "step": 400000
1210
+ },
1211
+ {
1212
+ "epoch": 9.31,
1213
+ "learning_rate": 2.0832848668140046e-05,
1214
+ "loss": 0.6086,
1215
+ "step": 402000
1216
+ },
1217
+ {
1218
+ "epoch": 9.35,
1219
+ "learning_rate": 1.9437012911480747e-05,
1220
+ "loss": 0.6079,
1221
+ "step": 404000
1222
+ },
1223
+ {
1224
+ "epoch": 9.4,
1225
+ "learning_rate": 1.8041177154821448e-05,
1226
+ "loss": 0.6012,
1227
+ "step": 406000
1228
+ },
1229
+ {
1230
+ "epoch": 9.45,
1231
+ "learning_rate": 1.664534139816215e-05,
1232
+ "loss": 0.6029,
1233
+ "step": 408000
1234
+ },
1235
+ {
1236
+ "epoch": 9.49,
1237
+ "learning_rate": 1.5249505641502848e-05,
1238
+ "loss": 0.5948,
1239
+ "step": 410000
1240
+ },
1241
+ {
1242
+ "epoch": 9.54,
1243
+ "learning_rate": 1.3853669884843549e-05,
1244
+ "loss": 0.6018,
1245
+ "step": 412000
1246
+ },
1247
+ {
1248
+ "epoch": 9.59,
1249
+ "learning_rate": 1.2457834128184248e-05,
1250
+ "loss": 0.5983,
1251
+ "step": 414000
1252
+ },
1253
+ {
1254
+ "epoch": 9.63,
1255
+ "learning_rate": 1.1061998371524949e-05,
1256
+ "loss": 0.6046,
1257
+ "step": 416000
1258
+ },
1259
+ {
1260
+ "epoch": 9.68,
1261
+ "learning_rate": 9.666162614865652e-06,
1262
+ "loss": 0.5971,
1263
+ "step": 418000
1264
+ },
1265
+ {
1266
+ "epoch": 9.73,
1267
+ "learning_rate": 8.27032685820635e-06,
1268
+ "loss": 0.6002,
1269
+ "step": 420000
1270
+ },
1271
+ {
1272
+ "epoch": 9.77,
1273
+ "learning_rate": 6.87449110154705e-06,
1274
+ "loss": 0.598,
1275
+ "step": 422000
1276
+ },
1277
+ {
1278
+ "epoch": 9.82,
1279
+ "learning_rate": 5.478655344887752e-06,
1280
+ "loss": 0.6042,
1281
+ "step": 424000
1282
+ },
1283
+ {
1284
+ "epoch": 9.86,
1285
+ "learning_rate": 4.082819588228451e-06,
1286
+ "loss": 0.5917,
1287
+ "step": 426000
1288
+ },
1289
+ {
1290
+ "epoch": 9.91,
1291
+ "learning_rate": 2.686983831569152e-06,
1292
+ "loss": 0.5985,
1293
+ "step": 428000
1294
+ },
1295
+ {
1296
+ "epoch": 9.96,
1297
+ "learning_rate": 1.2911480749098522e-06,
1298
+ "loss": 0.6036,
1299
+ "step": 430000
1300
+ }
1301
+ ],
1302
+ "logging_steps": 2000,
1303
+ "max_steps": 431850,
1304
+ "num_train_epochs": 10,
1305
+ "save_steps": 500,
1306
+ "total_flos": 8.831009191456261e+20,
1307
+ "trial_name": null,
1308
+ "trial_params": null
1309
+ }
checkpoint-431850/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:783723dff03848f1deb1a19f34bdbddf1b8e878418b808096c108f32e10ad612
3
+ size 4600
checkpoint-431850/vocab.json ADDED
@@ -0,0 +1,200 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "\b": 0,
3
+ "\"": 2,
4
+ "#": 3,
5
+ ";": 4,
6
+ "=": 5,
7
+ "[": 6,
8
+ "[PAD]": 197,
9
+ "[UNK]": 196,
10
+ "\\": 7,
11
+ "]": 8,
12
+ "_": 9,
13
+ "`": 10,
14
+ "a": 11,
15
+ "b": 12,
16
+ "c": 13,
17
+ "d": 14,
18
+ "e": 15,
19
+ "f": 16,
20
+ "g": 17,
21
+ "h": 18,
22
+ "i": 19,
23
+ "j": 20,
24
+ "k": 21,
25
+ "l": 22,
26
+ "m": 23,
27
+ "n": 24,
28
+ "o": 25,
29
+ "p": 26,
30
+ "q": 27,
31
+ "r": 28,
32
+ "s": 29,
33
+ "t": 30,
34
+ "u": 31,
35
+ "v": 32,
36
+ "w": 33,
37
+ "x": 34,
38
+ "y": 35,
39
+ "z": 36,
40
+ "|": 1,
41
+ "£": 37,
42
+ "°": 38,
43
+ "½": 39,
44
+ "ß": 40,
45
+ "à": 41,
46
+ "â": 42,
47
+ "ã": 43,
48
+ "ä": 44,
49
+ "ç": 45,
50
+ "è": 46,
51
+ "é": 47,
52
+ "ê": 48,
53
+ "ë": 49,
54
+ "ì": 50,
55
+ "í": 51,
56
+ "ò": 52,
57
+ "ô": 53,
58
+ "ö": 54,
59
+ "ø": 55,
60
+ "ú": 56,
61
+ "ü": 57,
62
+ "ą": 58,
63
+ "č": 59,
64
+ "đ": 60,
65
+ "ę": 61,
66
+ "ğ": 62,
67
+ "ī": 63,
68
+ "ō": 64,
69
+ "œ": 65,
70
+ "ş": 66,
71
+ "ż": 67,
72
+ "ž": 68,
73
+ "ǒ": 69,
74
+ "ǔ": 70,
75
+ "˚": 71,
76
+ "υ": 72,
77
+ "а": 73,
78
+ "б": 74,
79
+ "в": 75,
80
+ "г": 76,
81
+ "д": 77,
82
+ "е": 78,
83
+ "ж": 79,
84
+ "и": 80,
85
+ "й": 81,
86
+ "к": 82,
87
+ "л": 83,
88
+ "м": 84,
89
+ "н": 85,
90
+ "о": 86,
91
+ "п": 87,
92
+ "р": 88,
93
+ "с": 89,
94
+ "т": 90,
95
+ "у": 91,
96
+ "ф": 92,
97
+ "х": 93,
98
+ "ц": 94,
99
+ "ч": 95,
100
+ "ш": 96,
101
+ "щ": 97,
102
+ "ы": 98,
103
+ "ь": 99,
104
+ "э": 100,
105
+ "ю": 101,
106
+ "я": 102,
107
+ "ا": 103,
108
+ "ر": 104,
109
+ "ص": 105,
110
+ "َ": 106,
111
+ "ّ": 107,
112
+ "น": 108,
113
+ "อ": 109,
114
+ "ế": 110,
115
+ "ố": 111,
116
+ "​": 112,
117
+ "‑": 113,
118
+ "–": 114,
119
+ "—": 115,
120
+ "…": 116,
121
+ "⁄": 117,
122
+ "℃": 118,
123
+ "℉": 119,
124
+ "™": 120,
125
+ "≈": 121,
126
+ "♫": 122,
127
+ "【": 123,
128
+ "あ": 124,
129
+ "い": 125,
130
+ "う": 126,
131
+ "が": 127,
132
+ "じ": 128,
133
+ "す": 129,
134
+ "た": 130,
135
+ "っ": 131,
136
+ "つ": 132,
137
+ "と": 133,
138
+ "な": 134,
139
+ "は": 135,
140
+ "ま": 136,
141
+ "み": 137,
142
+ "り": 138,
143
+ "ん": 139,
144
+ "シ": 140,
145
+ "ッ": 141,
146
+ "デ": 142,
147
+ "ニ": 143,
148
+ "ュ": 144,
149
+ "ㄴ": 145,
150
+ "ㅒ": 146,
151
+ "ㅜ": 147,
152
+ "ㅠ": 148,
153
+ "ㅣ": 149,
154
+ "不": 150,
155
+ "京": 151,
156
+ "張": 152,
157
+ "按": 153,
158
+ "最": 154,
159
+ "理": 155,
160
+ "要": 156,
161
+ "讚": 157,
162
+ "近": 158,
163
+ "頑": 159,
164
+ "가": 160,
165
+ "간": 161,
166
+ "고": 162,
167
+ "구": 163,
168
+ "금": 164,
169
+ "꼭": 165,
170
+ "나": 166,
171
+ "녕": 167,
172
+ "님": 168,
173
+ "다": 169,
174
+ "든": 170,
175
+ "발": 171,
176
+ "배": 172,
177
+ "뿌": 173,
178
+ "상": 174,
179
+ "설": 175,
180
+ "아": 176,
181
+ "안": 177,
182
+ "어": 178,
183
+ "영": 179,
184
+ "왜": 180,
185
+ "요": 181,
186
+ "유": 182,
187
+ "은": 183,
188
+ "을": 184,
189
+ "음": 185,
190
+ "이": 186,
191
+ "인": 187,
192
+ "재": 188,
193
+ "절": 189,
194
+ "정": 190,
195
+ "쨌": 191,
196
+ "파": 192,
197
+ "하": 193,
198
+ "화": 194,
199
+ "�": 195
200
+ }
config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
3
+ "activation_dropout": 0.0,
4
+ "adapter_attn_dim": null,
5
+ "adapter_kernel_size": 3,
6
+ "adapter_stride": 2,
7
+ "add_adapter": false,
8
+ "apply_spec_augment": true,
9
+ "architectures": [
10
+ "Wav2Vec2ForCTC"
11
+ ],
12
+ "attention_dropout": 0.05,
13
+ "bos_token_id": 1,
14
+ "classifier_proj_size": 256,
15
+ "codevector_dim": 768,
16
+ "contrastive_logits_temperature": 0.1,
17
+ "conv_bias": true,
18
+ "conv_dim": [
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512,
25
+ 512
26
+ ],
27
+ "conv_kernel": [
28
+ 10,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 3,
33
+ 2,
34
+ 2
35
+ ],
36
+ "conv_stride": [
37
+ 5,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2,
43
+ 2
44
+ ],
45
+ "ctc_loss_reduction": "mean",
46
+ "ctc_zero_infinity": true,
47
+ "diversity_loss_weight": 0.1,
48
+ "do_stable_layer_norm": true,
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.05,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.0,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.05,
59
+ "hidden_size": 1024,
60
+ "initializer_range": 0.02,
61
+ "intermediate_size": 4096,
62
+ "layer_norm_eps": 1e-05,
63
+ "layerdrop": 0.05,
64
+ "mask_channel_length": 10,
65
+ "mask_channel_min_space": 1,
66
+ "mask_channel_other": 0.0,
67
+ "mask_channel_prob": 0.0,
68
+ "mask_channel_selection": "static",
69
+ "mask_feature_length": 10,
70
+ "mask_feature_min_masks": 0,
71
+ "mask_feature_prob": 0.0,
72
+ "mask_time_length": 10,
73
+ "mask_time_min_masks": 2,
74
+ "mask_time_min_space": 1,
75
+ "mask_time_other": 0.0,
76
+ "mask_time_prob": 0.05,
77
+ "mask_time_selection": "static",
78
+ "model_type": "wav2vec2",
79
+ "num_adapter_layers": 3,
80
+ "num_attention_heads": 16,
81
+ "num_codevector_groups": 2,
82
+ "num_codevectors_per_group": 320,
83
+ "num_conv_pos_embedding_groups": 16,
84
+ "num_conv_pos_embeddings": 128,
85
+ "num_feat_extract_layers": 7,
86
+ "num_hidden_layers": 24,
87
+ "num_negatives": 100,
88
+ "output_hidden_size": 1024,
89
+ "pad_token_id": 197,
90
+ "proj_codevector_dim": 768,
91
+ "tdnn_dilation": [
92
+ 1,
93
+ 2,
94
+ 3,
95
+ 1,
96
+ 1
97
+ ],
98
+ "tdnn_dim": [
99
+ 512,
100
+ 512,
101
+ 512,
102
+ 512,
103
+ 1500
104
+ ],
105
+ "tdnn_kernel": [
106
+ 5,
107
+ 3,
108
+ 3,
109
+ 1,
110
+ 1
111
+ ],
112
+ "torch_dtype": "float32",
113
+ "transformers_version": "4.35.0",
114
+ "use_weighted_layer_sum": false,
115
+ "vocab_size": 200,
116
+ "xvector_output_dim": 512
117
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:724e00446fba78a52abbddd258cfc4e543ca0c2d96145e6ddf9d911cf63d42f9
3
+ size 1262627480
preprocessor_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "processor_class": "Wav2Vec2Processor",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 16000
10
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": true,
19
+ "normalized": false,
20
+ "rstrip": true,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "[UNK]",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": true,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "196": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "197": {
12
+ "content": "[PAD]",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "198": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": true,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "199": {
28
+ "content": "</s>",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ }
35
+ },
36
+ "bos_token": "<s>",
37
+ "clean_up_tokenization_spaces": true,
38
+ "config": null,
39
+ "do_lower_case": false,
40
+ "eos_token": "</s>",
41
+ "model_max_length": 1000000000000000019884624838656,
42
+ "pad_token": "[PAD]",
43
+ "processor_class": "Wav2Vec2Processor",
44
+ "replace_word_delimiter_char": " ",
45
+ "target_lang": null,
46
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
47
+ "tokenizer_type": "wav2vec2",
48
+ "trust_remote_code": false,
49
+ "unk_token": "[UNK]",
50
+ "word_delimiter_token": "|"
51
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "train_loss": 0.806209775733219,
4
+ "train_runtime": 212917.0057,
5
+ "train_samples": 690961,
6
+ "train_samples_per_second": 32.452,
7
+ "train_steps_per_second": 2.028
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1318 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.99988422039805,
5
+ "eval_steps": 500,
6
+ "global_step": 431850,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05,
13
+ "learning_rate": 0.0003,
14
+ "loss": 3.5282,
15
+ "step": 2000
16
+ },
17
+ {
18
+ "epoch": 0.09,
19
+ "learning_rate": 0.00029860416424334066,
20
+ "loss": 1.3191,
21
+ "step": 4000
22
+ },
23
+ {
24
+ "epoch": 0.14,
25
+ "learning_rate": 0.0002972083284866814,
26
+ "loss": 1.2226,
27
+ "step": 6000
28
+ },
29
+ {
30
+ "epoch": 0.19,
31
+ "learning_rate": 0.00029581249273002205,
32
+ "loss": 1.1832,
33
+ "step": 8000
34
+ },
35
+ {
36
+ "epoch": 0.23,
37
+ "learning_rate": 0.0002944166569733628,
38
+ "loss": 1.1595,
39
+ "step": 10000
40
+ },
41
+ {
42
+ "epoch": 0.28,
43
+ "learning_rate": 0.0002930208212167035,
44
+ "loss": 1.1368,
45
+ "step": 12000
46
+ },
47
+ {
48
+ "epoch": 0.32,
49
+ "learning_rate": 0.00029162498546004417,
50
+ "loss": 1.1176,
51
+ "step": 14000
52
+ },
53
+ {
54
+ "epoch": 0.37,
55
+ "learning_rate": 0.00029022914970338486,
56
+ "loss": 1.1044,
57
+ "step": 16000
58
+ },
59
+ {
60
+ "epoch": 0.42,
61
+ "learning_rate": 0.00028883331394672555,
62
+ "loss": 1.0954,
63
+ "step": 18000
64
+ },
65
+ {
66
+ "epoch": 0.46,
67
+ "learning_rate": 0.0002874374781900663,
68
+ "loss": 1.088,
69
+ "step": 20000
70
+ },
71
+ {
72
+ "epoch": 0.51,
73
+ "learning_rate": 0.000286041642433407,
74
+ "loss": 1.0765,
75
+ "step": 22000
76
+ },
77
+ {
78
+ "epoch": 0.56,
79
+ "learning_rate": 0.0002846458066767477,
80
+ "loss": 1.069,
81
+ "step": 24000
82
+ },
83
+ {
84
+ "epoch": 0.6,
85
+ "learning_rate": 0.00028324997092008837,
86
+ "loss": 1.0623,
87
+ "step": 26000
88
+ },
89
+ {
90
+ "epoch": 0.65,
91
+ "learning_rate": 0.00028185413516342906,
92
+ "loss": 1.0653,
93
+ "step": 28000
94
+ },
95
+ {
96
+ "epoch": 0.69,
97
+ "learning_rate": 0.0002804582994067698,
98
+ "loss": 1.0467,
99
+ "step": 30000
100
+ },
101
+ {
102
+ "epoch": 0.74,
103
+ "learning_rate": 0.00027906246365011044,
104
+ "loss": 1.0395,
105
+ "step": 32000
106
+ },
107
+ {
108
+ "epoch": 0.79,
109
+ "learning_rate": 0.0002776666278934512,
110
+ "loss": 1.0419,
111
+ "step": 34000
112
+ },
113
+ {
114
+ "epoch": 0.83,
115
+ "learning_rate": 0.0002762707921367919,
116
+ "loss": 1.0385,
117
+ "step": 36000
118
+ },
119
+ {
120
+ "epoch": 0.88,
121
+ "learning_rate": 0.00027487495638013257,
122
+ "loss": 1.0375,
123
+ "step": 38000
124
+ },
125
+ {
126
+ "epoch": 0.93,
127
+ "learning_rate": 0.00027347912062347326,
128
+ "loss": 1.0211,
129
+ "step": 40000
130
+ },
131
+ {
132
+ "epoch": 0.97,
133
+ "learning_rate": 0.00027208328486681395,
134
+ "loss": 1.0248,
135
+ "step": 42000
136
+ },
137
+ {
138
+ "epoch": 1.02,
139
+ "learning_rate": 0.0002706874491101547,
140
+ "loss": 1.0071,
141
+ "step": 44000
142
+ },
143
+ {
144
+ "epoch": 1.07,
145
+ "learning_rate": 0.0002692916133534954,
146
+ "loss": 0.9825,
147
+ "step": 46000
148
+ },
149
+ {
150
+ "epoch": 1.11,
151
+ "learning_rate": 0.0002678957775968361,
152
+ "loss": 0.983,
153
+ "step": 48000
154
+ },
155
+ {
156
+ "epoch": 1.16,
157
+ "learning_rate": 0.00026649994184017677,
158
+ "loss": 0.9831,
159
+ "step": 50000
160
+ },
161
+ {
162
+ "epoch": 1.2,
163
+ "learning_rate": 0.00026510410608351746,
164
+ "loss": 0.9792,
165
+ "step": 52000
166
+ },
167
+ {
168
+ "epoch": 1.25,
169
+ "learning_rate": 0.0002637082703268582,
170
+ "loss": 0.9744,
171
+ "step": 54000
172
+ },
173
+ {
174
+ "epoch": 1.3,
175
+ "learning_rate": 0.0002623124345701989,
176
+ "loss": 0.9668,
177
+ "step": 56000
178
+ },
179
+ {
180
+ "epoch": 1.34,
181
+ "learning_rate": 0.0002609165988135396,
182
+ "loss": 0.971,
183
+ "step": 58000
184
+ },
185
+ {
186
+ "epoch": 1.39,
187
+ "learning_rate": 0.0002595207630568803,
188
+ "loss": 0.9789,
189
+ "step": 60000
190
+ },
191
+ {
192
+ "epoch": 1.44,
193
+ "learning_rate": 0.000258124927300221,
194
+ "loss": 0.9794,
195
+ "step": 62000
196
+ },
197
+ {
198
+ "epoch": 1.48,
199
+ "learning_rate": 0.00025672909154356166,
200
+ "loss": 0.9615,
201
+ "step": 64000
202
+ },
203
+ {
204
+ "epoch": 1.53,
205
+ "learning_rate": 0.0002553332557869024,
206
+ "loss": 0.9577,
207
+ "step": 66000
208
+ },
209
+ {
210
+ "epoch": 1.57,
211
+ "learning_rate": 0.0002539374200302431,
212
+ "loss": 0.9573,
213
+ "step": 68000
214
+ },
215
+ {
216
+ "epoch": 1.62,
217
+ "learning_rate": 0.0002525415842735838,
218
+ "loss": 0.9633,
219
+ "step": 70000
220
+ },
221
+ {
222
+ "epoch": 1.67,
223
+ "learning_rate": 0.0002511457485169245,
224
+ "loss": 0.946,
225
+ "step": 72000
226
+ },
227
+ {
228
+ "epoch": 1.71,
229
+ "learning_rate": 0.00024974991276026517,
230
+ "loss": 0.9528,
231
+ "step": 74000
232
+ },
233
+ {
234
+ "epoch": 1.76,
235
+ "learning_rate": 0.0002483540770036059,
236
+ "loss": 0.9474,
237
+ "step": 76000
238
+ },
239
+ {
240
+ "epoch": 1.81,
241
+ "learning_rate": 0.0002469582412469466,
242
+ "loss": 0.9431,
243
+ "step": 78000
244
+ },
245
+ {
246
+ "epoch": 1.85,
247
+ "learning_rate": 0.0002455624054902873,
248
+ "loss": 0.9469,
249
+ "step": 80000
250
+ },
251
+ {
252
+ "epoch": 1.9,
253
+ "learning_rate": 0.000244166569733628,
254
+ "loss": 0.9442,
255
+ "step": 82000
256
+ },
257
+ {
258
+ "epoch": 1.95,
259
+ "learning_rate": 0.0002427707339769687,
260
+ "loss": 0.9387,
261
+ "step": 84000
262
+ },
263
+ {
264
+ "epoch": 1.99,
265
+ "learning_rate": 0.0002413748982203094,
266
+ "loss": 0.9341,
267
+ "step": 86000
268
+ },
269
+ {
270
+ "epoch": 2.04,
271
+ "learning_rate": 0.00023997906246365008,
272
+ "loss": 0.9124,
273
+ "step": 88000
274
+ },
275
+ {
276
+ "epoch": 2.08,
277
+ "learning_rate": 0.00023858322670699077,
278
+ "loss": 0.8962,
279
+ "step": 90000
280
+ },
281
+ {
282
+ "epoch": 2.13,
283
+ "learning_rate": 0.0002371873909503315,
284
+ "loss": 0.8974,
285
+ "step": 92000
286
+ },
287
+ {
288
+ "epoch": 2.18,
289
+ "learning_rate": 0.0002357915551936722,
290
+ "loss": 0.9122,
291
+ "step": 94000
292
+ },
293
+ {
294
+ "epoch": 2.22,
295
+ "learning_rate": 0.00023439571943701287,
296
+ "loss": 0.8977,
297
+ "step": 96000
298
+ },
299
+ {
300
+ "epoch": 2.27,
301
+ "learning_rate": 0.0002329998836803536,
302
+ "loss": 0.8972,
303
+ "step": 98000
304
+ },
305
+ {
306
+ "epoch": 2.32,
307
+ "learning_rate": 0.00023160404792369428,
308
+ "loss": 0.8969,
309
+ "step": 100000
310
+ },
311
+ {
312
+ "epoch": 2.36,
313
+ "learning_rate": 0.000230208212167035,
314
+ "loss": 0.9021,
315
+ "step": 102000
316
+ },
317
+ {
318
+ "epoch": 2.41,
319
+ "learning_rate": 0.00022881237641037572,
320
+ "loss": 0.9083,
321
+ "step": 104000
322
+ },
323
+ {
324
+ "epoch": 2.45,
325
+ "learning_rate": 0.00022741654065371638,
326
+ "loss": 0.8928,
327
+ "step": 106000
328
+ },
329
+ {
330
+ "epoch": 2.5,
331
+ "learning_rate": 0.0002260207048970571,
332
+ "loss": 0.8871,
333
+ "step": 108000
334
+ },
335
+ {
336
+ "epoch": 2.55,
337
+ "learning_rate": 0.0002246248691403978,
338
+ "loss": 0.9017,
339
+ "step": 110000
340
+ },
341
+ {
342
+ "epoch": 2.59,
343
+ "learning_rate": 0.0002232290333837385,
344
+ "loss": 0.8968,
345
+ "step": 112000
346
+ },
347
+ {
348
+ "epoch": 2.64,
349
+ "learning_rate": 0.00022183319762707917,
350
+ "loss": 0.8813,
351
+ "step": 114000
352
+ },
353
+ {
354
+ "epoch": 2.69,
355
+ "learning_rate": 0.0002204373618704199,
356
+ "loss": 0.8807,
357
+ "step": 116000
358
+ },
359
+ {
360
+ "epoch": 2.73,
361
+ "learning_rate": 0.0002190415261137606,
362
+ "loss": 0.8816,
363
+ "step": 118000
364
+ },
365
+ {
366
+ "epoch": 2.78,
367
+ "learning_rate": 0.0002176456903571013,
368
+ "loss": 0.8776,
369
+ "step": 120000
370
+ },
371
+ {
372
+ "epoch": 2.83,
373
+ "learning_rate": 0.000216249854600442,
374
+ "loss": 0.8738,
375
+ "step": 122000
376
+ },
377
+ {
378
+ "epoch": 2.87,
379
+ "learning_rate": 0.00021485401884378268,
380
+ "loss": 0.8826,
381
+ "step": 124000
382
+ },
383
+ {
384
+ "epoch": 2.92,
385
+ "learning_rate": 0.0002134581830871234,
386
+ "loss": 0.893,
387
+ "step": 126000
388
+ },
389
+ {
390
+ "epoch": 2.96,
391
+ "learning_rate": 0.00021206234733046412,
392
+ "loss": 0.8753,
393
+ "step": 128000
394
+ },
395
+ {
396
+ "epoch": 3.01,
397
+ "learning_rate": 0.00021066651157380478,
398
+ "loss": 0.8671,
399
+ "step": 130000
400
+ },
401
+ {
402
+ "epoch": 3.06,
403
+ "learning_rate": 0.0002092706758171455,
404
+ "loss": 0.8497,
405
+ "step": 132000
406
+ },
407
+ {
408
+ "epoch": 3.1,
409
+ "learning_rate": 0.00020787484006048622,
410
+ "loss": 0.8443,
411
+ "step": 134000
412
+ },
413
+ {
414
+ "epoch": 3.15,
415
+ "learning_rate": 0.0002064790043038269,
416
+ "loss": 0.8477,
417
+ "step": 136000
418
+ },
419
+ {
420
+ "epoch": 3.2,
421
+ "learning_rate": 0.0002050831685471676,
422
+ "loss": 0.8465,
423
+ "step": 138000
424
+ },
425
+ {
426
+ "epoch": 3.24,
427
+ "learning_rate": 0.0002036873327905083,
428
+ "loss": 0.8501,
429
+ "step": 140000
430
+ },
431
+ {
432
+ "epoch": 3.29,
433
+ "learning_rate": 0.000202291497033849,
434
+ "loss": 0.8451,
435
+ "step": 142000
436
+ },
437
+ {
438
+ "epoch": 3.33,
439
+ "learning_rate": 0.00020089566127718972,
440
+ "loss": 0.84,
441
+ "step": 144000
442
+ },
443
+ {
444
+ "epoch": 3.38,
445
+ "learning_rate": 0.0001994998255205304,
446
+ "loss": 0.8482,
447
+ "step": 146000
448
+ },
449
+ {
450
+ "epoch": 3.43,
451
+ "learning_rate": 0.0001981039897638711,
452
+ "loss": 0.8383,
453
+ "step": 148000
454
+ },
455
+ {
456
+ "epoch": 3.47,
457
+ "learning_rate": 0.0001967081540072118,
458
+ "loss": 0.8353,
459
+ "step": 150000
460
+ },
461
+ {
462
+ "epoch": 3.52,
463
+ "learning_rate": 0.00019531231825055251,
464
+ "loss": 0.8334,
465
+ "step": 152000
466
+ },
467
+ {
468
+ "epoch": 3.57,
469
+ "learning_rate": 0.00019391648249389318,
470
+ "loss": 0.8446,
471
+ "step": 154000
472
+ },
473
+ {
474
+ "epoch": 3.61,
475
+ "learning_rate": 0.0001925206467372339,
476
+ "loss": 0.8373,
477
+ "step": 156000
478
+ },
479
+ {
480
+ "epoch": 3.66,
481
+ "learning_rate": 0.00019112481098057461,
482
+ "loss": 0.8412,
483
+ "step": 158000
484
+ },
485
+ {
486
+ "epoch": 3.7,
487
+ "learning_rate": 0.0001897289752239153,
488
+ "loss": 0.8286,
489
+ "step": 160000
490
+ },
491
+ {
492
+ "epoch": 3.75,
493
+ "learning_rate": 0.000188333139467256,
494
+ "loss": 0.8327,
495
+ "step": 162000
496
+ },
497
+ {
498
+ "epoch": 3.8,
499
+ "learning_rate": 0.0001869373037105967,
500
+ "loss": 0.8426,
501
+ "step": 164000
502
+ },
503
+ {
504
+ "epoch": 3.84,
505
+ "learning_rate": 0.0001855414679539374,
506
+ "loss": 0.8291,
507
+ "step": 166000
508
+ },
509
+ {
510
+ "epoch": 3.89,
511
+ "learning_rate": 0.00018414563219727812,
512
+ "loss": 0.8271,
513
+ "step": 168000
514
+ },
515
+ {
516
+ "epoch": 3.94,
517
+ "learning_rate": 0.00018274979644061879,
518
+ "loss": 0.8278,
519
+ "step": 170000
520
+ },
521
+ {
522
+ "epoch": 3.98,
523
+ "learning_rate": 0.0001813539606839595,
524
+ "loss": 0.8307,
525
+ "step": 172000
526
+ },
527
+ {
528
+ "epoch": 4.03,
529
+ "learning_rate": 0.0001799581249273002,
530
+ "loss": 0.8109,
531
+ "step": 174000
532
+ },
533
+ {
534
+ "epoch": 4.08,
535
+ "learning_rate": 0.0001785622891706409,
536
+ "loss": 0.8027,
537
+ "step": 176000
538
+ },
539
+ {
540
+ "epoch": 4.12,
541
+ "learning_rate": 0.0001771664534139816,
542
+ "loss": 0.7955,
543
+ "step": 178000
544
+ },
545
+ {
546
+ "epoch": 4.17,
547
+ "learning_rate": 0.0001757706176573223,
548
+ "loss": 0.7888,
549
+ "step": 180000
550
+ },
551
+ {
552
+ "epoch": 4.21,
553
+ "learning_rate": 0.000174374781900663,
554
+ "loss": 0.799,
555
+ "step": 182000
556
+ },
557
+ {
558
+ "epoch": 4.26,
559
+ "learning_rate": 0.0001729789461440037,
560
+ "loss": 0.7949,
561
+ "step": 184000
562
+ },
563
+ {
564
+ "epoch": 4.31,
565
+ "learning_rate": 0.0001715831103873444,
566
+ "loss": 0.7913,
567
+ "step": 186000
568
+ },
569
+ {
570
+ "epoch": 4.35,
571
+ "learning_rate": 0.0001701872746306851,
572
+ "loss": 0.7862,
573
+ "step": 188000
574
+ },
575
+ {
576
+ "epoch": 4.4,
577
+ "learning_rate": 0.0001687914388740258,
578
+ "loss": 0.7882,
579
+ "step": 190000
580
+ },
581
+ {
582
+ "epoch": 4.45,
583
+ "learning_rate": 0.00016739560311736652,
584
+ "loss": 0.8017,
585
+ "step": 192000
586
+ },
587
+ {
588
+ "epoch": 4.49,
589
+ "learning_rate": 0.00016599976736070724,
590
+ "loss": 0.787,
591
+ "step": 194000
592
+ },
593
+ {
594
+ "epoch": 4.54,
595
+ "learning_rate": 0.0001646039316040479,
596
+ "loss": 0.7975,
597
+ "step": 196000
598
+ },
599
+ {
600
+ "epoch": 4.58,
601
+ "learning_rate": 0.00016320809584738862,
602
+ "loss": 0.7938,
603
+ "step": 198000
604
+ },
605
+ {
606
+ "epoch": 4.63,
607
+ "learning_rate": 0.0001618122600907293,
608
+ "loss": 0.799,
609
+ "step": 200000
610
+ },
611
+ {
612
+ "epoch": 4.68,
613
+ "learning_rate": 0.00016041642433407003,
614
+ "loss": 0.7811,
615
+ "step": 202000
616
+ },
617
+ {
618
+ "epoch": 4.72,
619
+ "learning_rate": 0.0001590205885774107,
620
+ "loss": 0.7941,
621
+ "step": 204000
622
+ },
623
+ {
624
+ "epoch": 4.77,
625
+ "learning_rate": 0.0001576247528207514,
626
+ "loss": 0.7918,
627
+ "step": 206000
628
+ },
629
+ {
630
+ "epoch": 4.82,
631
+ "learning_rate": 0.00015622891706409213,
632
+ "loss": 0.8006,
633
+ "step": 208000
634
+ },
635
+ {
636
+ "epoch": 4.86,
637
+ "learning_rate": 0.00015483308130743282,
638
+ "loss": 0.7896,
639
+ "step": 210000
640
+ },
641
+ {
642
+ "epoch": 4.91,
643
+ "learning_rate": 0.0001534372455507735,
644
+ "loss": 0.7944,
645
+ "step": 212000
646
+ },
647
+ {
648
+ "epoch": 4.96,
649
+ "learning_rate": 0.0001520414097941142,
650
+ "loss": 0.7869,
651
+ "step": 214000
652
+ },
653
+ {
654
+ "epoch": 5.0,
655
+ "learning_rate": 0.00015064557403745492,
656
+ "loss": 0.7847,
657
+ "step": 216000
658
+ },
659
+ {
660
+ "epoch": 5.05,
661
+ "learning_rate": 0.0001492497382807956,
662
+ "loss": 0.7577,
663
+ "step": 218000
664
+ },
665
+ {
666
+ "epoch": 5.09,
667
+ "learning_rate": 0.00014785390252413633,
668
+ "loss": 0.7527,
669
+ "step": 220000
670
+ },
671
+ {
672
+ "epoch": 5.14,
673
+ "learning_rate": 0.00014645806676747702,
674
+ "loss": 0.7554,
675
+ "step": 222000
676
+ },
677
+ {
678
+ "epoch": 5.19,
679
+ "learning_rate": 0.0001450622310108177,
680
+ "loss": 0.7633,
681
+ "step": 224000
682
+ },
683
+ {
684
+ "epoch": 5.23,
685
+ "learning_rate": 0.0001436663952541584,
686
+ "loss": 0.754,
687
+ "step": 226000
688
+ },
689
+ {
690
+ "epoch": 5.28,
691
+ "learning_rate": 0.00014227055949749912,
692
+ "loss": 0.7438,
693
+ "step": 228000
694
+ },
695
+ {
696
+ "epoch": 5.33,
697
+ "learning_rate": 0.0001408747237408398,
698
+ "loss": 0.7514,
699
+ "step": 230000
700
+ },
701
+ {
702
+ "epoch": 5.37,
703
+ "learning_rate": 0.00013947888798418053,
704
+ "loss": 0.7522,
705
+ "step": 232000
706
+ },
707
+ {
708
+ "epoch": 5.42,
709
+ "learning_rate": 0.00013808305222752122,
710
+ "loss": 0.7525,
711
+ "step": 234000
712
+ },
713
+ {
714
+ "epoch": 5.46,
715
+ "learning_rate": 0.00013668721647086193,
716
+ "loss": 0.7476,
717
+ "step": 236000
718
+ },
719
+ {
720
+ "epoch": 5.51,
721
+ "learning_rate": 0.00013529138071420263,
722
+ "loss": 0.747,
723
+ "step": 238000
724
+ },
725
+ {
726
+ "epoch": 5.56,
727
+ "learning_rate": 0.00013389554495754332,
728
+ "loss": 0.7505,
729
+ "step": 240000
730
+ },
731
+ {
732
+ "epoch": 5.6,
733
+ "learning_rate": 0.000132499709200884,
734
+ "loss": 0.7501,
735
+ "step": 242000
736
+ },
737
+ {
738
+ "epoch": 5.65,
739
+ "learning_rate": 0.00013110387344422472,
740
+ "loss": 0.7568,
741
+ "step": 244000
742
+ },
743
+ {
744
+ "epoch": 5.7,
745
+ "learning_rate": 0.00012970803768756542,
746
+ "loss": 0.7383,
747
+ "step": 246000
748
+ },
749
+ {
750
+ "epoch": 5.74,
751
+ "learning_rate": 0.00012831220193090613,
752
+ "loss": 0.7535,
753
+ "step": 248000
754
+ },
755
+ {
756
+ "epoch": 5.79,
757
+ "learning_rate": 0.00012691636617424682,
758
+ "loss": 0.751,
759
+ "step": 250000
760
+ },
761
+ {
762
+ "epoch": 5.84,
763
+ "learning_rate": 0.00012552053041758751,
764
+ "loss": 0.7396,
765
+ "step": 252000
766
+ },
767
+ {
768
+ "epoch": 5.88,
769
+ "learning_rate": 0.0001241246946609282,
770
+ "loss": 0.7448,
771
+ "step": 254000
772
+ },
773
+ {
774
+ "epoch": 5.93,
775
+ "learning_rate": 0.00012272885890426892,
776
+ "loss": 0.7521,
777
+ "step": 256000
778
+ },
779
+ {
780
+ "epoch": 5.97,
781
+ "learning_rate": 0.00012133302314760961,
782
+ "loss": 0.7472,
783
+ "step": 258000
784
+ },
785
+ {
786
+ "epoch": 6.02,
787
+ "learning_rate": 0.00011993718739095032,
788
+ "loss": 0.7344,
789
+ "step": 260000
790
+ },
791
+ {
792
+ "epoch": 6.07,
793
+ "learning_rate": 0.00011854135163429101,
794
+ "loss": 0.7143,
795
+ "step": 262000
796
+ },
797
+ {
798
+ "epoch": 6.11,
799
+ "learning_rate": 0.00011714551587763173,
800
+ "loss": 0.717,
801
+ "step": 264000
802
+ },
803
+ {
804
+ "epoch": 6.16,
805
+ "learning_rate": 0.00011574968012097242,
806
+ "loss": 0.7194,
807
+ "step": 266000
808
+ },
809
+ {
810
+ "epoch": 6.21,
811
+ "learning_rate": 0.00011435384436431312,
812
+ "loss": 0.7206,
813
+ "step": 268000
814
+ },
815
+ {
816
+ "epoch": 6.25,
817
+ "learning_rate": 0.00011295800860765381,
818
+ "loss": 0.7148,
819
+ "step": 270000
820
+ },
821
+ {
822
+ "epoch": 6.3,
823
+ "learning_rate": 0.00011156217285099453,
824
+ "loss": 0.7137,
825
+ "step": 272000
826
+ },
827
+ {
828
+ "epoch": 6.34,
829
+ "learning_rate": 0.00011016633709433522,
830
+ "loss": 0.7168,
831
+ "step": 274000
832
+ },
833
+ {
834
+ "epoch": 6.39,
835
+ "learning_rate": 0.00010877050133767593,
836
+ "loss": 0.7097,
837
+ "step": 276000
838
+ },
839
+ {
840
+ "epoch": 6.44,
841
+ "learning_rate": 0.00010737466558101662,
842
+ "loss": 0.7084,
843
+ "step": 278000
844
+ },
845
+ {
846
+ "epoch": 6.48,
847
+ "learning_rate": 0.00010597882982435732,
848
+ "loss": 0.7101,
849
+ "step": 280000
850
+ },
851
+ {
852
+ "epoch": 6.53,
853
+ "learning_rate": 0.00010458299406769801,
854
+ "loss": 0.7117,
855
+ "step": 282000
856
+ },
857
+ {
858
+ "epoch": 6.58,
859
+ "learning_rate": 0.00010318715831103873,
860
+ "loss": 0.7067,
861
+ "step": 284000
862
+ },
863
+ {
864
+ "epoch": 6.62,
865
+ "learning_rate": 0.00010179132255437942,
866
+ "loss": 0.7026,
867
+ "step": 286000
868
+ },
869
+ {
870
+ "epoch": 6.67,
871
+ "learning_rate": 0.00010039548679772013,
872
+ "loss": 0.7144,
873
+ "step": 288000
874
+ },
875
+ {
876
+ "epoch": 6.72,
877
+ "learning_rate": 9.899965104106083e-05,
878
+ "loss": 0.7118,
879
+ "step": 290000
880
+ },
881
+ {
882
+ "epoch": 6.76,
883
+ "learning_rate": 9.760381528440152e-05,
884
+ "loss": 0.7028,
885
+ "step": 292000
886
+ },
887
+ {
888
+ "epoch": 6.81,
889
+ "learning_rate": 9.620797952774224e-05,
890
+ "loss": 0.7044,
891
+ "step": 294000
892
+ },
893
+ {
894
+ "epoch": 6.85,
895
+ "learning_rate": 9.481214377108293e-05,
896
+ "loss": 0.704,
897
+ "step": 296000
898
+ },
899
+ {
900
+ "epoch": 6.9,
901
+ "learning_rate": 9.341630801442363e-05,
902
+ "loss": 0.7049,
903
+ "step": 298000
904
+ },
905
+ {
906
+ "epoch": 6.95,
907
+ "learning_rate": 9.202047225776432e-05,
908
+ "loss": 0.6937,
909
+ "step": 300000
910
+ },
911
+ {
912
+ "epoch": 6.99,
913
+ "learning_rate": 9.062463650110504e-05,
914
+ "loss": 0.7096,
915
+ "step": 302000
916
+ },
917
+ {
918
+ "epoch": 7.04,
919
+ "learning_rate": 8.922880074444573e-05,
920
+ "loss": 0.6749,
921
+ "step": 304000
922
+ },
923
+ {
924
+ "epoch": 7.09,
925
+ "learning_rate": 8.783296498778644e-05,
926
+ "loss": 0.6718,
927
+ "step": 306000
928
+ },
929
+ {
930
+ "epoch": 7.13,
931
+ "learning_rate": 8.643712923112713e-05,
932
+ "loss": 0.6789,
933
+ "step": 308000
934
+ },
935
+ {
936
+ "epoch": 7.18,
937
+ "learning_rate": 8.504129347446783e-05,
938
+ "loss": 0.6805,
939
+ "step": 310000
940
+ },
941
+ {
942
+ "epoch": 7.22,
943
+ "learning_rate": 8.364545771780852e-05,
944
+ "loss": 0.6776,
945
+ "step": 312000
946
+ },
947
+ {
948
+ "epoch": 7.27,
949
+ "learning_rate": 8.224962196114924e-05,
950
+ "loss": 0.6734,
951
+ "step": 314000
952
+ },
953
+ {
954
+ "epoch": 7.32,
955
+ "learning_rate": 8.085378620448993e-05,
956
+ "loss": 0.6765,
957
+ "step": 316000
958
+ },
959
+ {
960
+ "epoch": 7.36,
961
+ "learning_rate": 7.945795044783064e-05,
962
+ "loss": 0.6709,
963
+ "step": 318000
964
+ },
965
+ {
966
+ "epoch": 7.41,
967
+ "learning_rate": 7.806211469117133e-05,
968
+ "loss": 0.6804,
969
+ "step": 320000
970
+ },
971
+ {
972
+ "epoch": 7.46,
973
+ "learning_rate": 7.666627893451203e-05,
974
+ "loss": 0.6688,
975
+ "step": 322000
976
+ },
977
+ {
978
+ "epoch": 7.5,
979
+ "learning_rate": 7.527044317785274e-05,
980
+ "loss": 0.6789,
981
+ "step": 324000
982
+ },
983
+ {
984
+ "epoch": 7.55,
985
+ "learning_rate": 7.387460742119343e-05,
986
+ "loss": 0.6699,
987
+ "step": 326000
988
+ },
989
+ {
990
+ "epoch": 7.6,
991
+ "learning_rate": 7.247877166453413e-05,
992
+ "loss": 0.662,
993
+ "step": 328000
994
+ },
995
+ {
996
+ "epoch": 7.64,
997
+ "learning_rate": 7.108293590787484e-05,
998
+ "loss": 0.6657,
999
+ "step": 330000
1000
+ },
1001
+ {
1002
+ "epoch": 7.69,
1003
+ "learning_rate": 6.968710015121553e-05,
1004
+ "loss": 0.6676,
1005
+ "step": 332000
1006
+ },
1007
+ {
1008
+ "epoch": 7.73,
1009
+ "learning_rate": 6.829126439455623e-05,
1010
+ "loss": 0.6698,
1011
+ "step": 334000
1012
+ },
1013
+ {
1014
+ "epoch": 7.78,
1015
+ "learning_rate": 6.689542863789693e-05,
1016
+ "loss": 0.6755,
1017
+ "step": 336000
1018
+ },
1019
+ {
1020
+ "epoch": 7.83,
1021
+ "learning_rate": 6.549959288123764e-05,
1022
+ "loss": 0.672,
1023
+ "step": 338000
1024
+ },
1025
+ {
1026
+ "epoch": 7.87,
1027
+ "learning_rate": 6.410375712457834e-05,
1028
+ "loss": 0.6616,
1029
+ "step": 340000
1030
+ },
1031
+ {
1032
+ "epoch": 7.92,
1033
+ "learning_rate": 6.270792136791903e-05,
1034
+ "loss": 0.6654,
1035
+ "step": 342000
1036
+ },
1037
+ {
1038
+ "epoch": 7.97,
1039
+ "learning_rate": 6.131208561125974e-05,
1040
+ "loss": 0.6606,
1041
+ "step": 344000
1042
+ },
1043
+ {
1044
+ "epoch": 8.01,
1045
+ "learning_rate": 5.991624985460044e-05,
1046
+ "loss": 0.6642,
1047
+ "step": 346000
1048
+ },
1049
+ {
1050
+ "epoch": 8.06,
1051
+ "learning_rate": 5.852041409794114e-05,
1052
+ "loss": 0.639,
1053
+ "step": 348000
1054
+ },
1055
+ {
1056
+ "epoch": 8.1,
1057
+ "learning_rate": 5.712457834128184e-05,
1058
+ "loss": 0.6463,
1059
+ "step": 350000
1060
+ },
1061
+ {
1062
+ "epoch": 8.15,
1063
+ "learning_rate": 5.572874258462254e-05,
1064
+ "loss": 0.6426,
1065
+ "step": 352000
1066
+ },
1067
+ {
1068
+ "epoch": 8.2,
1069
+ "learning_rate": 5.433290682796324e-05,
1070
+ "loss": 0.6356,
1071
+ "step": 354000
1072
+ },
1073
+ {
1074
+ "epoch": 8.24,
1075
+ "learning_rate": 5.2937071071303944e-05,
1076
+ "loss": 0.6336,
1077
+ "step": 356000
1078
+ },
1079
+ {
1080
+ "epoch": 8.29,
1081
+ "learning_rate": 5.154123531464464e-05,
1082
+ "loss": 0.631,
1083
+ "step": 358000
1084
+ },
1085
+ {
1086
+ "epoch": 8.34,
1087
+ "learning_rate": 5.014539955798534e-05,
1088
+ "loss": 0.6425,
1089
+ "step": 360000
1090
+ },
1091
+ {
1092
+ "epoch": 8.38,
1093
+ "learning_rate": 4.8749563801326044e-05,
1094
+ "loss": 0.6432,
1095
+ "step": 362000
1096
+ },
1097
+ {
1098
+ "epoch": 8.43,
1099
+ "learning_rate": 4.735372804466674e-05,
1100
+ "loss": 0.6449,
1101
+ "step": 364000
1102
+ },
1103
+ {
1104
+ "epoch": 8.48,
1105
+ "learning_rate": 4.595789228800744e-05,
1106
+ "loss": 0.635,
1107
+ "step": 366000
1108
+ },
1109
+ {
1110
+ "epoch": 8.52,
1111
+ "learning_rate": 4.456205653134814e-05,
1112
+ "loss": 0.6352,
1113
+ "step": 368000
1114
+ },
1115
+ {
1116
+ "epoch": 8.57,
1117
+ "learning_rate": 4.316622077468884e-05,
1118
+ "loss": 0.6281,
1119
+ "step": 370000
1120
+ },
1121
+ {
1122
+ "epoch": 8.61,
1123
+ "learning_rate": 4.1770385018029545e-05,
1124
+ "loss": 0.6291,
1125
+ "step": 372000
1126
+ },
1127
+ {
1128
+ "epoch": 8.66,
1129
+ "learning_rate": 4.037454926137024e-05,
1130
+ "loss": 0.6317,
1131
+ "step": 374000
1132
+ },
1133
+ {
1134
+ "epoch": 8.71,
1135
+ "learning_rate": 3.897871350471094e-05,
1136
+ "loss": 0.6296,
1137
+ "step": 376000
1138
+ },
1139
+ {
1140
+ "epoch": 8.75,
1141
+ "learning_rate": 3.7582877748051644e-05,
1142
+ "loss": 0.6279,
1143
+ "step": 378000
1144
+ },
1145
+ {
1146
+ "epoch": 8.8,
1147
+ "learning_rate": 3.618704199139234e-05,
1148
+ "loss": 0.6379,
1149
+ "step": 380000
1150
+ },
1151
+ {
1152
+ "epoch": 8.85,
1153
+ "learning_rate": 3.4791206234733046e-05,
1154
+ "loss": 0.6292,
1155
+ "step": 382000
1156
+ },
1157
+ {
1158
+ "epoch": 8.89,
1159
+ "learning_rate": 3.3395370478073744e-05,
1160
+ "loss": 0.6248,
1161
+ "step": 384000
1162
+ },
1163
+ {
1164
+ "epoch": 8.94,
1165
+ "learning_rate": 3.199953472141444e-05,
1166
+ "loss": 0.6199,
1167
+ "step": 386000
1168
+ },
1169
+ {
1170
+ "epoch": 8.98,
1171
+ "learning_rate": 3.0603698964755146e-05,
1172
+ "loss": 0.6173,
1173
+ "step": 388000
1174
+ },
1175
+ {
1176
+ "epoch": 9.03,
1177
+ "learning_rate": 2.9207863208095843e-05,
1178
+ "loss": 0.6129,
1179
+ "step": 390000
1180
+ },
1181
+ {
1182
+ "epoch": 9.08,
1183
+ "learning_rate": 2.7812027451436544e-05,
1184
+ "loss": 0.6013,
1185
+ "step": 392000
1186
+ },
1187
+ {
1188
+ "epoch": 9.12,
1189
+ "learning_rate": 2.6416191694777245e-05,
1190
+ "loss": 0.6045,
1191
+ "step": 394000
1192
+ },
1193
+ {
1194
+ "epoch": 9.17,
1195
+ "learning_rate": 2.5020355938117946e-05,
1196
+ "loss": 0.6134,
1197
+ "step": 396000
1198
+ },
1199
+ {
1200
+ "epoch": 9.22,
1201
+ "learning_rate": 2.3624520181458644e-05,
1202
+ "loss": 0.6056,
1203
+ "step": 398000
1204
+ },
1205
+ {
1206
+ "epoch": 9.26,
1207
+ "learning_rate": 2.2228684424799345e-05,
1208
+ "loss": 0.6112,
1209
+ "step": 400000
1210
+ },
1211
+ {
1212
+ "epoch": 9.31,
1213
+ "learning_rate": 2.0832848668140046e-05,
1214
+ "loss": 0.6086,
1215
+ "step": 402000
1216
+ },
1217
+ {
1218
+ "epoch": 9.35,
1219
+ "learning_rate": 1.9437012911480747e-05,
1220
+ "loss": 0.6079,
1221
+ "step": 404000
1222
+ },
1223
+ {
1224
+ "epoch": 9.4,
1225
+ "learning_rate": 1.8041177154821448e-05,
1226
+ "loss": 0.6012,
1227
+ "step": 406000
1228
+ },
1229
+ {
1230
+ "epoch": 9.45,
1231
+ "learning_rate": 1.664534139816215e-05,
1232
+ "loss": 0.6029,
1233
+ "step": 408000
1234
+ },
1235
+ {
1236
+ "epoch": 9.49,
1237
+ "learning_rate": 1.5249505641502848e-05,
1238
+ "loss": 0.5948,
1239
+ "step": 410000
1240
+ },
1241
+ {
1242
+ "epoch": 9.54,
1243
+ "learning_rate": 1.3853669884843549e-05,
1244
+ "loss": 0.6018,
1245
+ "step": 412000
1246
+ },
1247
+ {
1248
+ "epoch": 9.59,
1249
+ "learning_rate": 1.2457834128184248e-05,
1250
+ "loss": 0.5983,
1251
+ "step": 414000
1252
+ },
1253
+ {
1254
+ "epoch": 9.63,
1255
+ "learning_rate": 1.1061998371524949e-05,
1256
+ "loss": 0.6046,
1257
+ "step": 416000
1258
+ },
1259
+ {
1260
+ "epoch": 9.68,
1261
+ "learning_rate": 9.666162614865652e-06,
1262
+ "loss": 0.5971,
1263
+ "step": 418000
1264
+ },
1265
+ {
1266
+ "epoch": 9.73,
1267
+ "learning_rate": 8.27032685820635e-06,
1268
+ "loss": 0.6002,
1269
+ "step": 420000
1270
+ },
1271
+ {
1272
+ "epoch": 9.77,
1273
+ "learning_rate": 6.87449110154705e-06,
1274
+ "loss": 0.598,
1275
+ "step": 422000
1276
+ },
1277
+ {
1278
+ "epoch": 9.82,
1279
+ "learning_rate": 5.478655344887752e-06,
1280
+ "loss": 0.6042,
1281
+ "step": 424000
1282
+ },
1283
+ {
1284
+ "epoch": 9.86,
1285
+ "learning_rate": 4.082819588228451e-06,
1286
+ "loss": 0.5917,
1287
+ "step": 426000
1288
+ },
1289
+ {
1290
+ "epoch": 9.91,
1291
+ "learning_rate": 2.686983831569152e-06,
1292
+ "loss": 0.5985,
1293
+ "step": 428000
1294
+ },
1295
+ {
1296
+ "epoch": 9.96,
1297
+ "learning_rate": 1.2911480749098522e-06,
1298
+ "loss": 0.6036,
1299
+ "step": 430000
1300
+ },
1301
+ {
1302
+ "epoch": 10.0,
1303
+ "step": 431850,
1304
+ "total_flos": 8.831009191456261e+20,
1305
+ "train_loss": 0.806209775733219,
1306
+ "train_runtime": 212917.0057,
1307
+ "train_samples_per_second": 32.452,
1308
+ "train_steps_per_second": 2.028
1309
+ }
1310
+ ],
1311
+ "logging_steps": 2000,
1312
+ "max_steps": 431850,
1313
+ "num_train_epochs": 10,
1314
+ "save_steps": 500,
1315
+ "total_flos": 8.831009191456261e+20,
1316
+ "trial_name": null,
1317
+ "trial_params": null
1318
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:783723dff03848f1deb1a19f34bdbddf1b8e878418b808096c108f32e10ad612
3
+ size 4600
vocab.json ADDED
@@ -0,0 +1,200 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "\b": 0,
3
+ "\"": 2,
4
+ "#": 3,
5
+ ";": 4,
6
+ "=": 5,
7
+ "[": 6,
8
+ "[PAD]": 197,
9
+ "[UNK]": 196,
10
+ "\\": 7,
11
+ "]": 8,
12
+ "_": 9,
13
+ "`": 10,
14
+ "a": 11,
15
+ "b": 12,
16
+ "c": 13,
17
+ "d": 14,
18
+ "e": 15,
19
+ "f": 16,
20
+ "g": 17,
21
+ "h": 18,
22
+ "i": 19,
23
+ "j": 20,
24
+ "k": 21,
25
+ "l": 22,
26
+ "m": 23,
27
+ "n": 24,
28
+ "o": 25,
29
+ "p": 26,
30
+ "q": 27,
31
+ "r": 28,
32
+ "s": 29,
33
+ "t": 30,
34
+ "u": 31,
35
+ "v": 32,
36
+ "w": 33,
37
+ "x": 34,
38
+ "y": 35,
39
+ "z": 36,
40
+ "|": 1,
41
+ "£": 37,
42
+ "°": 38,
43
+ "½": 39,
44
+ "ß": 40,
45
+ "à": 41,
46
+ "â": 42,
47
+ "ã": 43,
48
+ "ä": 44,
49
+ "ç": 45,
50
+ "è": 46,
51
+ "é": 47,
52
+ "ê": 48,
53
+ "ë": 49,
54
+ "ì": 50,
55
+ "í": 51,
56
+ "ò": 52,
57
+ "ô": 53,
58
+ "ö": 54,
59
+ "ø": 55,
60
+ "ú": 56,
61
+ "ü": 57,
62
+ "ą": 58,
63
+ "č": 59,
64
+ "đ": 60,
65
+ "ę": 61,
66
+ "ğ": 62,
67
+ "ī": 63,
68
+ "ō": 64,
69
+ "œ": 65,
70
+ "ş": 66,
71
+ "ż": 67,
72
+ "ž": 68,
73
+ "ǒ": 69,
74
+ "ǔ": 70,
75
+ "˚": 71,
76
+ "υ": 72,
77
+ "а": 73,
78
+ "б": 74,
79
+ "в": 75,
80
+ "г": 76,
81
+ "д": 77,
82
+ "е": 78,
83
+ "ж": 79,
84
+ "и": 80,
85
+ "й": 81,
86
+ "к": 82,
87
+ "л": 83,
88
+ "м": 84,
89
+ "н": 85,
90
+ "о": 86,
91
+ "п": 87,
92
+ "р": 88,
93
+ "с": 89,
94
+ "т": 90,
95
+ "у": 91,
96
+ "ф": 92,
97
+ "х": 93,
98
+ "ц": 94,
99
+ "ч": 95,
100
+ "ш": 96,
101
+ "щ": 97,
102
+ "ы": 98,
103
+ "ь": 99,
104
+ "э": 100,
105
+ "ю": 101,
106
+ "я": 102,
107
+ "ا": 103,
108
+ "ر": 104,
109
+ "ص": 105,
110
+ "َ": 106,
111
+ "ّ": 107,
112
+ "น": 108,
113
+ "อ": 109,
114
+ "ế": 110,
115
+ "ố": 111,
116
+ "​": 112,
117
+ "‑": 113,
118
+ "–": 114,
119
+ "—": 115,
120
+ "…": 116,
121
+ "⁄": 117,
122
+ "℃": 118,
123
+ "℉": 119,
124
+ "™": 120,
125
+ "≈": 121,
126
+ "♫": 122,
127
+ "【": 123,
128
+ "あ": 124,
129
+ "い": 125,
130
+ "う": 126,
131
+ "が": 127,
132
+ "じ": 128,
133
+ "す": 129,
134
+ "た": 130,
135
+ "っ": 131,
136
+ "つ": 132,
137
+ "と": 133,
138
+ "な": 134,
139
+ "は": 135,
140
+ "ま": 136,
141
+ "み": 137,
142
+ "り": 138,
143
+ "ん": 139,
144
+ "シ": 140,
145
+ "ッ": 141,
146
+ "デ": 142,
147
+ "ニ": 143,
148
+ "ュ": 144,
149
+ "ㄴ": 145,
150
+ "ㅒ": 146,
151
+ "ㅜ": 147,
152
+ "ㅠ": 148,
153
+ "ㅣ": 149,
154
+ "不": 150,
155
+ "京": 151,
156
+ "張": 152,
157
+ "按": 153,
158
+ "最": 154,
159
+ "理": 155,
160
+ "要": 156,
161
+ "讚": 157,
162
+ "近": 158,
163
+ "頑": 159,
164
+ "가": 160,
165
+ "간": 161,
166
+ "고": 162,
167
+ "구": 163,
168
+ "금": 164,
169
+ "꼭": 165,
170
+ "나": 166,
171
+ "녕": 167,
172
+ "님": 168,
173
+ "다": 169,
174
+ "든": 170,
175
+ "발": 171,
176
+ "배": 172,
177
+ "뿌": 173,
178
+ "상": 174,
179
+ "설": 175,
180
+ "아": 176,
181
+ "안": 177,
182
+ "어": 178,
183
+ "영": 179,
184
+ "왜": 180,
185
+ "요": 181,
186
+ "유": 182,
187
+ "은": 183,
188
+ "을": 184,
189
+ "음": 185,
190
+ "이": 186,
191
+ "인": 187,
192
+ "재": 188,
193
+ "절": 189,
194
+ "정": 190,
195
+ "쨌": 191,
196
+ "파": 192,
197
+ "하": 193,
198
+ "화": 194,
199
+ "�": 195
200
+ }