yesj1234 commited on
Commit
80c01f1
1 Parent(s): 2ec47ba

Upload folder using huggingface_hub

Browse files
Files changed (43) hide show
  1. README.md +99 -0
  2. added_tokens.json +4 -0
  3. all_results.json +15 -0
  4. checkpoint-60000/added_tokens.json +4 -0
  5. checkpoint-60000/config.json +117 -0
  6. checkpoint-60000/optimizer.pt +3 -0
  7. checkpoint-60000/preprocessor_config.json +10 -0
  8. checkpoint-60000/pytorch_model.bin +3 -0
  9. checkpoint-60000/rng_state_0.pth +3 -0
  10. checkpoint-60000/rng_state_1.pth +3 -0
  11. checkpoint-60000/rng_state_2.pth +3 -0
  12. checkpoint-60000/rng_state_3.pth +3 -0
  13. checkpoint-60000/scheduler.pt +3 -0
  14. checkpoint-60000/special_tokens_map.json +30 -0
  15. checkpoint-60000/tokenizer_config.json +51 -0
  16. checkpoint-60000/trainer_state.json +1039 -0
  17. checkpoint-60000/training_args.bin +3 -0
  18. checkpoint-60000/vocab.json +1792 -0
  19. checkpoint-62000/added_tokens.json +4 -0
  20. checkpoint-62000/config.json +117 -0
  21. checkpoint-62000/optimizer.pt +3 -0
  22. checkpoint-62000/preprocessor_config.json +10 -0
  23. checkpoint-62000/pytorch_model.bin +3 -0
  24. checkpoint-62000/rng_state_0.pth +3 -0
  25. checkpoint-62000/rng_state_1.pth +3 -0
  26. checkpoint-62000/rng_state_2.pth +3 -0
  27. checkpoint-62000/rng_state_3.pth +3 -0
  28. checkpoint-62000/scheduler.pt +3 -0
  29. checkpoint-62000/special_tokens_map.json +30 -0
  30. checkpoint-62000/tokenizer_config.json +51 -0
  31. checkpoint-62000/trainer_state.json +1073 -0
  32. checkpoint-62000/training_args.bin +3 -0
  33. checkpoint-62000/vocab.json +1792 -0
  34. config.json +117 -0
  35. eval_results.json +10 -0
  36. preprocessor_config.json +10 -0
  37. pytorch_model.bin +3 -0
  38. special_tokens_map.json +30 -0
  39. tokenizer_config.json +51 -0
  40. train_results.json +8 -0
  41. trainer_state.json +1100 -0
  42. training_args.bin +3 -0
  43. vocab.json +1792 -0
README.md ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: facebook/wav2vec2-large-xlsr-53
4
+ tags:
5
+ - automatic-speech-recognition
6
+ - ./sample_speech.py
7
+ - generated_from_trainer
8
+ metrics:
9
+ - wer
10
+ model-index:
11
+ - name: ko-xlsr2
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ # ko-xlsr2
19
+
20
+ This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the ./SAMPLE_SPEECH.PY - NA dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 0.4239
23
+ - Cer: 0.1113
24
+ - Wer: 0.3038
25
+
26
+ ## Model description
27
+
28
+ More information needed
29
+
30
+ ## Intended uses & limitations
31
+
32
+ More information needed
33
+
34
+ ## Training and evaluation data
35
+
36
+ More information needed
37
+
38
+ ## Training procedure
39
+
40
+ ### Training hyperparameters
41
+
42
+ The following hyperparameters were used during training:
43
+ - learning_rate: 0.0003
44
+ - train_batch_size: 4
45
+ - eval_batch_size: 4
46
+ - seed: 42
47
+ - distributed_type: multi-GPU
48
+ - num_devices: 4
49
+ - gradient_accumulation_steps: 4
50
+ - total_train_batch_size: 64
51
+ - total_eval_batch_size: 16
52
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
53
+ - lr_scheduler_type: linear
54
+ - lr_scheduler_warmup_steps: 1000
55
+ - num_epochs: 30
56
+
57
+ ### Training results
58
+
59
+ | Training Loss | Epoch | Step | Validation Loss | Cer | Wer |
60
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|
61
+ | 1.7721 | 0.94 | 2000 | 1.1368 | 0.2903 | 0.6589 |
62
+ | 1.3501 | 1.89 | 4000 | 0.8561 | 0.2240 | 0.5451 |
63
+ | 1.2133 | 2.83 | 6000 | 0.7505 | 0.2003 | 0.4974 |
64
+ | 1.0981 | 3.77 | 8000 | 0.6768 | 0.1842 | 0.4686 |
65
+ | 1.0375 | 4.72 | 10000 | 0.6413 | 0.1707 | 0.4404 |
66
+ | 0.9927 | 5.66 | 12000 | 0.6106 | 0.1634 | 0.4246 |
67
+ | 0.9439 | 6.6 | 14000 | 0.5999 | 0.1613 | 0.4159 |
68
+ | 0.9059 | 7.55 | 16000 | 0.5740 | 0.1535 | 0.3985 |
69
+ | 0.8772 | 8.49 | 18000 | 0.5569 | 0.1478 | 0.3954 |
70
+ | 0.8483 | 9.43 | 20000 | 0.5407 | 0.1427 | 0.3784 |
71
+ | 0.81 | 10.37 | 22000 | 0.5283 | 0.1415 | 0.3744 |
72
+ | 0.793 | 11.32 | 24000 | 0.5179 | 0.1366 | 0.3663 |
73
+ | 0.7577 | 12.26 | 26000 | 0.5059 | 0.1359 | 0.3595 |
74
+ | 0.7379 | 13.2 | 28000 | 0.4969 | 0.1333 | 0.3532 |
75
+ | 0.7328 | 14.15 | 30000 | 0.4908 | 0.1308 | 0.3475 |
76
+ | 0.7119 | 15.09 | 32000 | 0.4887 | 0.1286 | 0.3478 |
77
+ | 0.7572 | 16.03 | 34000 | 0.5170 | 0.1327 | 0.3577 |
78
+ | 0.8198 | 16.98 | 36000 | 0.5839 | 0.1432 | 0.3825 |
79
+ | 0.8008 | 17.92 | 38000 | 0.5447 | 0.1376 | 0.3661 |
80
+ | 0.759 | 18.86 | 40000 | 0.4998 | 0.1337 | 0.3534 |
81
+ | 0.6907 | 19.81 | 42000 | 0.4710 | 0.1288 | 0.3412 |
82
+ | 0.659 | 20.75 | 44000 | 0.4578 | 0.1242 | 0.3325 |
83
+ | 0.6345 | 21.69 | 46000 | 0.4531 | 0.1221 | 0.3257 |
84
+ | 0.6242 | 22.64 | 48000 | 0.4498 | 0.1209 | 0.3218 |
85
+ | 0.6163 | 23.58 | 50000 | 0.4552 | 0.1194 | 0.3188 |
86
+ | 0.6121 | 24.52 | 52000 | 0.4633 | 0.1154 | 0.3137 |
87
+ | 0.6054 | 25.47 | 54000 | 0.4623 | 0.1176 | 0.3171 |
88
+ | 0.591 | 26.41 | 56000 | 0.4413 | 0.1146 | 0.3116 |
89
+ | 0.5713 | 27.35 | 58000 | 0.4338 | 0.1135 | 0.3093 |
90
+ | 0.5703 | 28.3 | 60000 | 0.4280 | 0.1121 | 0.3061 |
91
+ | 0.5576 | 29.24 | 62000 | 0.4248 | 0.1119 | 0.3047 |
92
+
93
+
94
+ ### Framework versions
95
+
96
+ - Transformers 4.34.1
97
+ - Pytorch 2.1.0+cu121
98
+ - Datasets 2.14.5
99
+ - Tokenizers 0.14.1
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "</s>": 1791,
3
+ "<s>": 1790
4
+ }
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 29.99,
3
+ "eval_cer": 0.11131710849470199,
4
+ "eval_loss": 0.4239382743835449,
5
+ "eval_runtime": 2675.296,
6
+ "eval_samples": 16914,
7
+ "eval_samples_per_second": 6.322,
8
+ "eval_steps_per_second": 0.395,
9
+ "eval_wer": 0.30376043574435696,
10
+ "train_loss": 0.9622837933354408,
11
+ "train_runtime": 156348.4692,
12
+ "train_samples": 135710,
13
+ "train_samples_per_second": 26.04,
14
+ "train_steps_per_second": 0.407
15
+ }
checkpoint-60000/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "</s>": 1791,
3
+ "<s>": 1790
4
+ }
checkpoint-60000/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": 0.1,
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": false,
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": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.0,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_size": 1024,
60
+ "initializer_range": 0.02,
61
+ "intermediate_size": 4096,
62
+ "layer_norm_eps": 1e-05,
63
+ "layerdrop": 0.1,
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": 0.1,
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": 1789,
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": "4.34.1",
114
+ "use_weighted_layer_sum": false,
115
+ "vocab_size": 1792,
116
+ "xvector_output_dim": 512
117
+ }
checkpoint-60000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c475809cd860e7c7b84ce823d7686af30b310040de19115e4b91548b568af936
3
+ size 2504854198
checkpoint-60000/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-60000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f5f70c093f9766fb8280a610ba836a4a3547c733f2bb487597e8974ed515e04
3
+ size 1269249578
checkpoint-60000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4dc858df1f07dc02d42752a206ad2957ac5569768a94b4bffd7dad33dd690d04
3
+ size 15024
checkpoint-60000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad321d52e149a9d1b2518d4ee35e9c05ebc51bd70db82988b1a11b9f102b5cb0
3
+ size 15024
checkpoint-60000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc162f1c3e4d4c40ff3a9b59fc79b09e933da21a9d09578b3c75aff92143aa5d
3
+ size 15024
checkpoint-60000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e703f842452820c443e0da6f98f859f76e39044d5bc48346461deefdc43c8295
3
+ size 15024
checkpoint-60000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37ec4599eeb0e61ab118cbe6a7610a4fe42cf896feb6c39e886b8ebaa27e8e57
3
+ size 1064
checkpoint-60000/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-60000/tokenizer_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "1788": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "1789": {
12
+ "content": "[PAD]",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "1790": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": true,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "1791": {
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-60000/trainer_state.json ADDED
@@ -0,0 +1,1039 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 28.295213393067673,
5
+ "eval_steps": 2000,
6
+ "global_step": 60000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.24,
13
+ "learning_rate": 0.00014849999999999998,
14
+ "loss": 15.6695,
15
+ "step": 500
16
+ },
17
+ {
18
+ "epoch": 0.47,
19
+ "learning_rate": 0.0002982,
20
+ "loss": 4.4987,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.71,
25
+ "learning_rate": 0.00029763258785942486,
26
+ "loss": 2.3147,
27
+ "step": 1500
28
+ },
29
+ {
30
+ "epoch": 0.94,
31
+ "learning_rate": 0.0002952364217252396,
32
+ "loss": 1.7721,
33
+ "step": 2000
34
+ },
35
+ {
36
+ "epoch": 0.94,
37
+ "eval_cer": 0.29026945911872865,
38
+ "eval_loss": 1.1368173360824585,
39
+ "eval_runtime": 2718.6813,
40
+ "eval_samples_per_second": 6.221,
41
+ "eval_steps_per_second": 0.389,
42
+ "eval_wer": 0.6589364856615053,
43
+ "step": 2000
44
+ },
45
+ {
46
+ "epoch": 1.18,
47
+ "learning_rate": 0.0002928402555910543,
48
+ "loss": 1.5735,
49
+ "step": 2500
50
+ },
51
+ {
52
+ "epoch": 1.41,
53
+ "learning_rate": 0.000290444089456869,
54
+ "loss": 1.4611,
55
+ "step": 3000
56
+ },
57
+ {
58
+ "epoch": 1.65,
59
+ "learning_rate": 0.00028805271565495203,
60
+ "loss": 1.4056,
61
+ "step": 3500
62
+ },
63
+ {
64
+ "epoch": 1.89,
65
+ "learning_rate": 0.00028565654952076674,
66
+ "loss": 1.3501,
67
+ "step": 4000
68
+ },
69
+ {
70
+ "epoch": 1.89,
71
+ "eval_cer": 0.2240432390931656,
72
+ "eval_loss": 0.8561204671859741,
73
+ "eval_runtime": 2698.4922,
74
+ "eval_samples_per_second": 6.268,
75
+ "eval_steps_per_second": 0.392,
76
+ "eval_wer": 0.5451323644839807,
77
+ "step": 4000
78
+ },
79
+ {
80
+ "epoch": 2.12,
81
+ "learning_rate": 0.00028326038338658146,
82
+ "loss": 1.2967,
83
+ "step": 4500
84
+ },
85
+ {
86
+ "epoch": 2.36,
87
+ "learning_rate": 0.00028086421725239617,
88
+ "loss": 1.244,
89
+ "step": 5000
90
+ },
91
+ {
92
+ "epoch": 2.59,
93
+ "learning_rate": 0.0002784728434504792,
94
+ "loss": 1.2361,
95
+ "step": 5500
96
+ },
97
+ {
98
+ "epoch": 2.83,
99
+ "learning_rate": 0.0002760766773162939,
100
+ "loss": 1.2133,
101
+ "step": 6000
102
+ },
103
+ {
104
+ "epoch": 2.83,
105
+ "eval_cer": 0.2002927101645895,
106
+ "eval_loss": 0.7505359053611755,
107
+ "eval_runtime": 2721.6856,
108
+ "eval_samples_per_second": 6.215,
109
+ "eval_steps_per_second": 0.389,
110
+ "eval_wer": 0.49743516558444134,
111
+ "step": 6000
112
+ },
113
+ {
114
+ "epoch": 3.07,
115
+ "learning_rate": 0.0002736805111821086,
116
+ "loss": 1.1755,
117
+ "step": 6500
118
+ },
119
+ {
120
+ "epoch": 3.3,
121
+ "learning_rate": 0.0002712843450479233,
122
+ "loss": 1.1403,
123
+ "step": 7000
124
+ },
125
+ {
126
+ "epoch": 3.54,
127
+ "learning_rate": 0.000268888178913738,
128
+ "loss": 1.1253,
129
+ "step": 7500
130
+ },
131
+ {
132
+ "epoch": 3.77,
133
+ "learning_rate": 0.00026649680511182103,
134
+ "loss": 1.0981,
135
+ "step": 8000
136
+ },
137
+ {
138
+ "epoch": 3.77,
139
+ "eval_cer": 0.18418492651709803,
140
+ "eval_loss": 0.6767656803131104,
141
+ "eval_runtime": 2740.3957,
142
+ "eval_samples_per_second": 6.172,
143
+ "eval_steps_per_second": 0.386,
144
+ "eval_wer": 0.46858375155594667,
145
+ "step": 8000
146
+ },
147
+ {
148
+ "epoch": 4.01,
149
+ "learning_rate": 0.00026410063897763575,
150
+ "loss": 1.0984,
151
+ "step": 8500
152
+ },
153
+ {
154
+ "epoch": 4.24,
155
+ "learning_rate": 0.00026170447284345046,
156
+ "loss": 1.0614,
157
+ "step": 9000
158
+ },
159
+ {
160
+ "epoch": 4.48,
161
+ "learning_rate": 0.00025930830670926517,
162
+ "loss": 1.0553,
163
+ "step": 9500
164
+ },
165
+ {
166
+ "epoch": 4.72,
167
+ "learning_rate": 0.0002569169329073482,
168
+ "loss": 1.0375,
169
+ "step": 10000
170
+ },
171
+ {
172
+ "epoch": 4.72,
173
+ "eval_cer": 0.17071197058066542,
174
+ "eval_loss": 0.64134681224823,
175
+ "eval_runtime": 2703.0312,
176
+ "eval_samples_per_second": 6.257,
177
+ "eval_steps_per_second": 0.391,
178
+ "eval_wer": 0.44041417913406117,
179
+ "step": 10000
180
+ },
181
+ {
182
+ "epoch": 4.95,
183
+ "learning_rate": 0.0002545207667731629,
184
+ "loss": 1.0351,
185
+ "step": 10500
186
+ },
187
+ {
188
+ "epoch": 5.19,
189
+ "learning_rate": 0.00025212460063897763,
190
+ "loss": 1.0087,
191
+ "step": 11000
192
+ },
193
+ {
194
+ "epoch": 5.42,
195
+ "learning_rate": 0.00024973322683706067,
196
+ "loss": 0.9934,
197
+ "step": 11500
198
+ },
199
+ {
200
+ "epoch": 5.66,
201
+ "learning_rate": 0.0002473370607028754,
202
+ "loss": 0.9927,
203
+ "step": 12000
204
+ },
205
+ {
206
+ "epoch": 5.66,
207
+ "eval_cer": 0.16338767301962598,
208
+ "eval_loss": 0.6106029152870178,
209
+ "eval_runtime": 2826.5991,
210
+ "eval_samples_per_second": 5.984,
211
+ "eval_steps_per_second": 0.374,
212
+ "eval_wer": 0.42461289632043386,
213
+ "step": 12000
214
+ },
215
+ {
216
+ "epoch": 5.89,
217
+ "learning_rate": 0.0002449408945686901,
218
+ "loss": 0.9821,
219
+ "step": 12500
220
+ },
221
+ {
222
+ "epoch": 6.13,
223
+ "learning_rate": 0.0002425447284345048,
224
+ "loss": 0.9637,
225
+ "step": 13000
226
+ },
227
+ {
228
+ "epoch": 6.37,
229
+ "learning_rate": 0.00024014856230031946,
230
+ "loss": 0.9497,
231
+ "step": 13500
232
+ },
233
+ {
234
+ "epoch": 6.6,
235
+ "learning_rate": 0.00023775239616613414,
236
+ "loss": 0.9439,
237
+ "step": 14000
238
+ },
239
+ {
240
+ "epoch": 6.6,
241
+ "eval_cer": 0.1613024947979602,
242
+ "eval_loss": 0.5999171733856201,
243
+ "eval_runtime": 2741.759,
244
+ "eval_samples_per_second": 6.169,
245
+ "eval_steps_per_second": 0.386,
246
+ "eval_wer": 0.41588373807768236,
247
+ "step": 14000
248
+ },
249
+ {
250
+ "epoch": 6.84,
251
+ "learning_rate": 0.00023536102236421723,
252
+ "loss": 0.948,
253
+ "step": 14500
254
+ },
255
+ {
256
+ "epoch": 7.07,
257
+ "learning_rate": 0.0002329696485623003,
258
+ "loss": 0.9367,
259
+ "step": 15000
260
+ },
261
+ {
262
+ "epoch": 7.31,
263
+ "learning_rate": 0.000230573482428115,
264
+ "loss": 0.9087,
265
+ "step": 15500
266
+ },
267
+ {
268
+ "epoch": 7.55,
269
+ "learning_rate": 0.0002281773162939297,
270
+ "loss": 0.9059,
271
+ "step": 16000
272
+ },
273
+ {
274
+ "epoch": 7.55,
275
+ "eval_cer": 0.15351797484699242,
276
+ "eval_loss": 0.5740103721618652,
277
+ "eval_runtime": 2702.3695,
278
+ "eval_samples_per_second": 6.259,
279
+ "eval_steps_per_second": 0.392,
280
+ "eval_wer": 0.39852056228147376,
281
+ "step": 16000
282
+ },
283
+ {
284
+ "epoch": 7.78,
285
+ "learning_rate": 0.00022578115015974438,
286
+ "loss": 0.9075,
287
+ "step": 16500
288
+ },
289
+ {
290
+ "epoch": 8.02,
291
+ "learning_rate": 0.0002233849840255591,
292
+ "loss": 0.8999,
293
+ "step": 17000
294
+ },
295
+ {
296
+ "epoch": 8.25,
297
+ "learning_rate": 0.00022098881789137377,
298
+ "loss": 0.8597,
299
+ "step": 17500
300
+ },
301
+ {
302
+ "epoch": 8.49,
303
+ "learning_rate": 0.00021859265175718849,
304
+ "loss": 0.8772,
305
+ "step": 18000
306
+ },
307
+ {
308
+ "epoch": 8.49,
309
+ "eval_cer": 0.14781427082015555,
310
+ "eval_loss": 0.5568912625312805,
311
+ "eval_runtime": 2727.4118,
312
+ "eval_samples_per_second": 6.201,
313
+ "eval_steps_per_second": 0.388,
314
+ "eval_wer": 0.3954364182701837,
315
+ "step": 18000
316
+ },
317
+ {
318
+ "epoch": 8.72,
319
+ "learning_rate": 0.00021619648562300317,
320
+ "loss": 0.8785,
321
+ "step": 18500
322
+ },
323
+ {
324
+ "epoch": 8.96,
325
+ "learning_rate": 0.00021380511182108623,
326
+ "loss": 0.8702,
327
+ "step": 19000
328
+ },
329
+ {
330
+ "epoch": 9.2,
331
+ "learning_rate": 0.00021140894568690095,
332
+ "loss": 0.8447,
333
+ "step": 19500
334
+ },
335
+ {
336
+ "epoch": 9.43,
337
+ "learning_rate": 0.00020901277955271563,
338
+ "loss": 0.8483,
339
+ "step": 20000
340
+ },
341
+ {
342
+ "epoch": 9.43,
343
+ "eval_cer": 0.14274528108464166,
344
+ "eval_loss": 0.5406663417816162,
345
+ "eval_runtime": 2824.1098,
346
+ "eval_samples_per_second": 5.989,
347
+ "eval_steps_per_second": 0.375,
348
+ "eval_wer": 0.3784141632772796,
349
+ "step": 20000
350
+ },
351
+ {
352
+ "epoch": 9.67,
353
+ "learning_rate": 0.00020661661341853031,
354
+ "loss": 0.8358,
355
+ "step": 20500
356
+ },
357
+ {
358
+ "epoch": 9.9,
359
+ "learning_rate": 0.00020422523961661338,
360
+ "loss": 0.8391,
361
+ "step": 21000
362
+ },
363
+ {
364
+ "epoch": 10.14,
365
+ "learning_rate": 0.00020182907348242812,
366
+ "loss": 0.8215,
367
+ "step": 21500
368
+ },
369
+ {
370
+ "epoch": 10.37,
371
+ "learning_rate": 0.00019943769968051118,
372
+ "loss": 0.81,
373
+ "step": 22000
374
+ },
375
+ {
376
+ "epoch": 10.37,
377
+ "eval_cer": 0.1415456492625536,
378
+ "eval_loss": 0.5282983779907227,
379
+ "eval_runtime": 2742.622,
380
+ "eval_samples_per_second": 6.167,
381
+ "eval_steps_per_second": 0.386,
382
+ "eval_wer": 0.37441032593614476,
383
+ "step": 22000
384
+ },
385
+ {
386
+ "epoch": 10.61,
387
+ "learning_rate": 0.00019704153354632587,
388
+ "loss": 0.8108,
389
+ "step": 22500
390
+ },
391
+ {
392
+ "epoch": 10.85,
393
+ "learning_rate": 0.00019464536741214058,
394
+ "loss": 0.8175,
395
+ "step": 23000
396
+ },
397
+ {
398
+ "epoch": 11.08,
399
+ "learning_rate": 0.00019224920127795526,
400
+ "loss": 0.8001,
401
+ "step": 23500
402
+ },
403
+ {
404
+ "epoch": 11.32,
405
+ "learning_rate": 0.00018985782747603833,
406
+ "loss": 0.793,
407
+ "step": 24000
408
+ },
409
+ {
410
+ "epoch": 11.32,
411
+ "eval_cer": 0.13662715879199255,
412
+ "eval_loss": 0.517921507358551,
413
+ "eval_runtime": 2729.8979,
414
+ "eval_samples_per_second": 6.196,
415
+ "eval_steps_per_second": 0.388,
416
+ "eval_wer": 0.36633129573690426,
417
+ "step": 24000
418
+ },
419
+ {
420
+ "epoch": 11.55,
421
+ "learning_rate": 0.000187461661341853,
422
+ "loss": 0.7827,
423
+ "step": 24500
424
+ },
425
+ {
426
+ "epoch": 11.79,
427
+ "learning_rate": 0.00018506549520766772,
428
+ "loss": 0.7899,
429
+ "step": 25000
430
+ },
431
+ {
432
+ "epoch": 12.03,
433
+ "learning_rate": 0.0001826693290734824,
434
+ "loss": 0.7806,
435
+ "step": 25500
436
+ },
437
+ {
438
+ "epoch": 12.26,
439
+ "learning_rate": 0.00018027316293929712,
440
+ "loss": 0.7577,
441
+ "step": 26000
442
+ },
443
+ {
444
+ "epoch": 12.26,
445
+ "eval_cer": 0.1359182854425769,
446
+ "eval_loss": 0.5058821439743042,
447
+ "eval_runtime": 2722.7634,
448
+ "eval_samples_per_second": 6.212,
449
+ "eval_steps_per_second": 0.389,
450
+ "eval_wer": 0.35946530932616605,
451
+ "step": 26000
452
+ },
453
+ {
454
+ "epoch": 12.5,
455
+ "learning_rate": 0.00017788178913738016,
456
+ "loss": 0.762,
457
+ "step": 26500
458
+ },
459
+ {
460
+ "epoch": 12.73,
461
+ "learning_rate": 0.00017548562300319487,
462
+ "loss": 0.7595,
463
+ "step": 27000
464
+ },
465
+ {
466
+ "epoch": 12.97,
467
+ "learning_rate": 0.00017308945686900955,
468
+ "loss": 0.7629,
469
+ "step": 27500
470
+ },
471
+ {
472
+ "epoch": 13.2,
473
+ "learning_rate": 0.00017069329073482426,
474
+ "loss": 0.7379,
475
+ "step": 28000
476
+ },
477
+ {
478
+ "epoch": 13.2,
479
+ "eval_cer": 0.13330526921919236,
480
+ "eval_loss": 0.4969228506088257,
481
+ "eval_runtime": 2824.8712,
482
+ "eval_samples_per_second": 5.988,
483
+ "eval_steps_per_second": 0.375,
484
+ "eval_wer": 0.35324945095893884,
485
+ "step": 28000
486
+ },
487
+ {
488
+ "epoch": 13.44,
489
+ "learning_rate": 0.00016830191693290736,
490
+ "loss": 0.737,
491
+ "step": 28500
492
+ },
493
+ {
494
+ "epoch": 13.68,
495
+ "learning_rate": 0.000165905750798722,
496
+ "loss": 0.7444,
497
+ "step": 29000
498
+ },
499
+ {
500
+ "epoch": 13.91,
501
+ "learning_rate": 0.00016350958466453675,
502
+ "loss": 0.7372,
503
+ "step": 29500
504
+ },
505
+ {
506
+ "epoch": 14.15,
507
+ "learning_rate": 0.00016111341853035144,
508
+ "loss": 0.7328,
509
+ "step": 30000
510
+ },
511
+ {
512
+ "epoch": 14.15,
513
+ "eval_cer": 0.13079476698787718,
514
+ "eval_loss": 0.4908413589000702,
515
+ "eval_runtime": 2825.8542,
516
+ "eval_samples_per_second": 5.985,
517
+ "eval_steps_per_second": 0.374,
518
+ "eval_wer": 0.3475251528197322,
519
+ "step": 30000
520
+ },
521
+ {
522
+ "epoch": 14.38,
523
+ "learning_rate": 0.0001587172523961661,
524
+ "loss": 0.7184,
525
+ "step": 30500
526
+ },
527
+ {
528
+ "epoch": 14.62,
529
+ "learning_rate": 0.00015632108626198083,
530
+ "loss": 0.7216,
531
+ "step": 31000
532
+ },
533
+ {
534
+ "epoch": 14.85,
535
+ "learning_rate": 0.0001539297124600639,
536
+ "loss": 0.7238,
537
+ "step": 31500
538
+ },
539
+ {
540
+ "epoch": 15.09,
541
+ "learning_rate": 0.00015153354632587858,
542
+ "loss": 0.7119,
543
+ "step": 32000
544
+ },
545
+ {
546
+ "epoch": 15.09,
547
+ "eval_cer": 0.12864851660072327,
548
+ "eval_loss": 0.4887321889400482,
549
+ "eval_runtime": 2731.0336,
550
+ "eval_samples_per_second": 6.193,
551
+ "eval_steps_per_second": 0.387,
552
+ "eval_wer": 0.34784228845071313,
553
+ "step": 32000
554
+ },
555
+ {
556
+ "epoch": 15.33,
557
+ "learning_rate": 0.00014914217252396165,
558
+ "loss": 0.7124,
559
+ "step": 32500
560
+ },
561
+ {
562
+ "epoch": 15.56,
563
+ "learning_rate": 0.00014674600638977636,
564
+ "loss": 0.7294,
565
+ "step": 33000
566
+ },
567
+ {
568
+ "epoch": 15.8,
569
+ "learning_rate": 0.00014434984025559104,
570
+ "loss": 0.7545,
571
+ "step": 33500
572
+ },
573
+ {
574
+ "epoch": 16.03,
575
+ "learning_rate": 0.00014195367412140575,
576
+ "loss": 0.7572,
577
+ "step": 34000
578
+ },
579
+ {
580
+ "epoch": 16.03,
581
+ "eval_cer": 0.13271854020075294,
582
+ "eval_loss": 0.5169993042945862,
583
+ "eval_runtime": 2729.0002,
584
+ "eval_samples_per_second": 6.198,
585
+ "eval_steps_per_second": 0.388,
586
+ "eval_wer": 0.3576893497926726,
587
+ "step": 34000
588
+ },
589
+ {
590
+ "epoch": 16.27,
591
+ "learning_rate": 0.00013956230031948882,
592
+ "loss": 0.7687,
593
+ "step": 34500
594
+ },
595
+ {
596
+ "epoch": 16.51,
597
+ "learning_rate": 0.0001371661341853035,
598
+ "loss": 0.7884,
599
+ "step": 35000
600
+ },
601
+ {
602
+ "epoch": 16.74,
603
+ "learning_rate": 0.00013476996805111819,
604
+ "loss": 0.8156,
605
+ "step": 35500
606
+ },
607
+ {
608
+ "epoch": 16.98,
609
+ "learning_rate": 0.0001323738019169329,
610
+ "loss": 0.8198,
611
+ "step": 36000
612
+ },
613
+ {
614
+ "epoch": 16.98,
615
+ "eval_cer": 0.1431662427967562,
616
+ "eval_loss": 0.5838645696640015,
617
+ "eval_runtime": 2730.4526,
618
+ "eval_samples_per_second": 6.195,
619
+ "eval_steps_per_second": 0.387,
620
+ "eval_wer": 0.38254485487080686,
621
+ "step": 36000
622
+ },
623
+ {
624
+ "epoch": 17.21,
625
+ "learning_rate": 0.00012997763578274758,
626
+ "loss": 0.819,
627
+ "step": 36500
628
+ },
629
+ {
630
+ "epoch": 17.45,
631
+ "learning_rate": 0.00012758626198083067,
632
+ "loss": 0.8411,
633
+ "step": 37000
634
+ },
635
+ {
636
+ "epoch": 17.68,
637
+ "learning_rate": 0.00012519009584664536,
638
+ "loss": 0.8366,
639
+ "step": 37500
640
+ },
641
+ {
642
+ "epoch": 17.92,
643
+ "learning_rate": 0.00012279392971246007,
644
+ "loss": 0.8008,
645
+ "step": 38000
646
+ },
647
+ {
648
+ "epoch": 17.92,
649
+ "eval_cer": 0.13762394377870937,
650
+ "eval_loss": 0.5447062253952026,
651
+ "eval_runtime": 2738.2931,
652
+ "eval_samples_per_second": 6.177,
653
+ "eval_steps_per_second": 0.386,
654
+ "eval_wer": 0.36609344401366856,
655
+ "step": 38000
656
+ },
657
+ {
658
+ "epoch": 18.16,
659
+ "learning_rate": 0.00012039776357827474,
660
+ "loss": 0.8032,
661
+ "step": 38500
662
+ },
663
+ {
664
+ "epoch": 18.39,
665
+ "learning_rate": 0.00011800159744408944,
666
+ "loss": 0.7753,
667
+ "step": 39000
668
+ },
669
+ {
670
+ "epoch": 18.63,
671
+ "learning_rate": 0.00011560543130990414,
672
+ "loss": 0.7608,
673
+ "step": 39500
674
+ },
675
+ {
676
+ "epoch": 18.86,
677
+ "learning_rate": 0.00011321405750798721,
678
+ "loss": 0.759,
679
+ "step": 40000
680
+ },
681
+ {
682
+ "epoch": 18.86,
683
+ "eval_cer": 0.1336804268071908,
684
+ "eval_loss": 0.49982598423957825,
685
+ "eval_runtime": 2725.5181,
686
+ "eval_samples_per_second": 6.206,
687
+ "eval_steps_per_second": 0.388,
688
+ "eval_wer": 0.3533921619928803,
689
+ "step": 40000
690
+ },
691
+ {
692
+ "epoch": 19.1,
693
+ "learning_rate": 0.00011081789137380191,
694
+ "loss": 0.7285,
695
+ "step": 40500
696
+ },
697
+ {
698
+ "epoch": 19.34,
699
+ "learning_rate": 0.00010842172523961661,
700
+ "loss": 0.7036,
701
+ "step": 41000
702
+ },
703
+ {
704
+ "epoch": 19.57,
705
+ "learning_rate": 0.00010602555910543131,
706
+ "loss": 0.6953,
707
+ "step": 41500
708
+ },
709
+ {
710
+ "epoch": 19.81,
711
+ "learning_rate": 0.00010363418530351436,
712
+ "loss": 0.6907,
713
+ "step": 42000
714
+ },
715
+ {
716
+ "epoch": 19.81,
717
+ "eval_cer": 0.12877502322923437,
718
+ "eval_loss": 0.47100237011909485,
719
+ "eval_runtime": 2667.1801,
720
+ "eval_samples_per_second": 6.342,
721
+ "eval_steps_per_second": 0.397,
722
+ "eval_wer": 0.34119829698166165,
723
+ "step": 42000
724
+ },
725
+ {
726
+ "epoch": 20.04,
727
+ "learning_rate": 0.00010123801916932906,
728
+ "loss": 0.6858,
729
+ "step": 42500
730
+ },
731
+ {
732
+ "epoch": 20.28,
733
+ "learning_rate": 9.884664536741213e-05,
734
+ "loss": 0.6603,
735
+ "step": 43000
736
+ },
737
+ {
738
+ "epoch": 20.51,
739
+ "learning_rate": 9.645047923322683e-05,
740
+ "loss": 0.6609,
741
+ "step": 43500
742
+ },
743
+ {
744
+ "epoch": 20.75,
745
+ "learning_rate": 9.405431309904153e-05,
746
+ "loss": 0.659,
747
+ "step": 44000
748
+ },
749
+ {
750
+ "epoch": 20.75,
751
+ "eval_cer": 0.12423387149543921,
752
+ "eval_loss": 0.4578304886817932,
753
+ "eval_runtime": 2665.8908,
754
+ "eval_samples_per_second": 6.345,
755
+ "eval_steps_per_second": 0.397,
756
+ "eval_wer": 0.3324532819573611,
757
+ "step": 44000
758
+ },
759
+ {
760
+ "epoch": 20.99,
761
+ "learning_rate": 9.165814696485623e-05,
762
+ "loss": 0.6567,
763
+ "step": 44500
764
+ },
765
+ {
766
+ "epoch": 21.22,
767
+ "learning_rate": 8.926198083067093e-05,
768
+ "loss": 0.6437,
769
+ "step": 45000
770
+ },
771
+ {
772
+ "epoch": 21.46,
773
+ "learning_rate": 8.686581469648561e-05,
774
+ "loss": 0.6371,
775
+ "step": 45500
776
+ },
777
+ {
778
+ "epoch": 21.69,
779
+ "learning_rate": 8.447444089456868e-05,
780
+ "loss": 0.6345,
781
+ "step": 46000
782
+ },
783
+ {
784
+ "epoch": 21.69,
785
+ "eval_cer": 0.12205708502554125,
786
+ "eval_loss": 0.45305466651916504,
787
+ "eval_runtime": 2667.1819,
788
+ "eval_samples_per_second": 6.342,
789
+ "eval_steps_per_second": 0.397,
790
+ "eval_wer": 0.3256982930174662,
791
+ "step": 46000
792
+ },
793
+ {
794
+ "epoch": 21.93,
795
+ "learning_rate": 8.207827476038337e-05,
796
+ "loss": 0.6418,
797
+ "step": 46500
798
+ },
799
+ {
800
+ "epoch": 22.16,
801
+ "learning_rate": 7.968210862619807e-05,
802
+ "loss": 0.6306,
803
+ "step": 47000
804
+ },
805
+ {
806
+ "epoch": 22.4,
807
+ "learning_rate": 7.728594249201278e-05,
808
+ "loss": 0.6213,
809
+ "step": 47500
810
+ },
811
+ {
812
+ "epoch": 22.64,
813
+ "learning_rate": 7.489456869009583e-05,
814
+ "loss": 0.6242,
815
+ "step": 48000
816
+ },
817
+ {
818
+ "epoch": 22.64,
819
+ "eval_cer": 0.12094251800538308,
820
+ "eval_loss": 0.4497627019882202,
821
+ "eval_runtime": 2727.8154,
822
+ "eval_samples_per_second": 6.201,
823
+ "eval_steps_per_second": 0.388,
824
+ "eval_wer": 0.32180545314717474,
825
+ "step": 48000
826
+ },
827
+ {
828
+ "epoch": 22.87,
829
+ "learning_rate": 7.249840255591053e-05,
830
+ "loss": 0.6294,
831
+ "step": 48500
832
+ },
833
+ {
834
+ "epoch": 23.11,
835
+ "learning_rate": 7.010223642172524e-05,
836
+ "loss": 0.6141,
837
+ "step": 49000
838
+ },
839
+ {
840
+ "epoch": 23.34,
841
+ "learning_rate": 6.770607028753993e-05,
842
+ "loss": 0.6155,
843
+ "step": 49500
844
+ },
845
+ {
846
+ "epoch": 23.58,
847
+ "learning_rate": 6.530990415335462e-05,
848
+ "loss": 0.6163,
849
+ "step": 50000
850
+ },
851
+ {
852
+ "epoch": 23.58,
853
+ "eval_cer": 0.11941571386818009,
854
+ "eval_loss": 0.45521289110183716,
855
+ "eval_runtime": 2664.8843,
856
+ "eval_samples_per_second": 6.347,
857
+ "eval_steps_per_second": 0.397,
858
+ "eval_wer": 0.3188402349975026,
859
+ "step": 50000
860
+ },
861
+ {
862
+ "epoch": 23.82,
863
+ "learning_rate": 6.291373801916932e-05,
864
+ "loss": 0.6167,
865
+ "step": 50500
866
+ },
867
+ {
868
+ "epoch": 24.05,
869
+ "learning_rate": 6.0522364217252394e-05,
870
+ "loss": 0.6179,
871
+ "step": 51000
872
+ },
873
+ {
874
+ "epoch": 24.29,
875
+ "learning_rate": 5.8126198083067085e-05,
876
+ "loss": 0.6154,
877
+ "step": 51500
878
+ },
879
+ {
880
+ "epoch": 24.52,
881
+ "learning_rate": 5.573482428115016e-05,
882
+ "loss": 0.6121,
883
+ "step": 52000
884
+ },
885
+ {
886
+ "epoch": 24.52,
887
+ "eval_cer": 0.1153500525656853,
888
+ "eval_loss": 0.46334853768348694,
889
+ "eval_runtime": 2666.3375,
890
+ "eval_samples_per_second": 6.344,
891
+ "eval_steps_per_second": 0.397,
892
+ "eval_wer": 0.3136947093848362,
893
+ "step": 52000
894
+ },
895
+ {
896
+ "epoch": 24.76,
897
+ "learning_rate": 5.3338658146964855e-05,
898
+ "loss": 0.6227,
899
+ "step": 52500
900
+ },
901
+ {
902
+ "epoch": 24.99,
903
+ "learning_rate": 5.0942492012779546e-05,
904
+ "loss": 0.6156,
905
+ "step": 53000
906
+ },
907
+ {
908
+ "epoch": 25.23,
909
+ "learning_rate": 4.854632587859424e-05,
910
+ "loss": 0.6159,
911
+ "step": 53500
912
+ },
913
+ {
914
+ "epoch": 25.47,
915
+ "learning_rate": 4.615015974440894e-05,
916
+ "loss": 0.6054,
917
+ "step": 54000
918
+ },
919
+ {
920
+ "epoch": 25.47,
921
+ "eval_cer": 0.11759009234983882,
922
+ "eval_loss": 0.46227386593818665,
923
+ "eval_runtime": 2666.9277,
924
+ "eval_samples_per_second": 6.342,
925
+ "eval_steps_per_second": 0.397,
926
+ "eval_wer": 0.3171356309809798,
927
+ "step": 54000
928
+ },
929
+ {
930
+ "epoch": 25.7,
931
+ "learning_rate": 4.375399361022364e-05,
932
+ "loss": 0.6051,
933
+ "step": 54500
934
+ },
935
+ {
936
+ "epoch": 25.94,
937
+ "learning_rate": 4.1362619808306704e-05,
938
+ "loss": 0.5986,
939
+ "step": 55000
940
+ },
941
+ {
942
+ "epoch": 26.17,
943
+ "learning_rate": 3.89664536741214e-05,
944
+ "loss": 0.5916,
945
+ "step": 55500
946
+ },
947
+ {
948
+ "epoch": 26.41,
949
+ "learning_rate": 3.65702875399361e-05,
950
+ "loss": 0.591,
951
+ "step": 56000
952
+ },
953
+ {
954
+ "epoch": 26.41,
955
+ "eval_cer": 0.11455829556310718,
956
+ "eval_loss": 0.4413212835788727,
957
+ "eval_runtime": 2669.4551,
958
+ "eval_samples_per_second": 6.336,
959
+ "eval_steps_per_second": 0.396,
960
+ "eval_wer": 0.31158575743881267,
961
+ "step": 56000
962
+ },
963
+ {
964
+ "epoch": 26.64,
965
+ "learning_rate": 3.41741214057508e-05,
966
+ "loss": 0.5904,
967
+ "step": 56500
968
+ },
969
+ {
970
+ "epoch": 26.88,
971
+ "learning_rate": 3.178274760383386e-05,
972
+ "loss": 0.5887,
973
+ "step": 57000
974
+ },
975
+ {
976
+ "epoch": 27.12,
977
+ "learning_rate": 2.9386581469648557e-05,
978
+ "loss": 0.5768,
979
+ "step": 57500
980
+ },
981
+ {
982
+ "epoch": 27.35,
983
+ "learning_rate": 2.6990415335463258e-05,
984
+ "loss": 0.5713,
985
+ "step": 58000
986
+ },
987
+ {
988
+ "epoch": 27.35,
989
+ "eval_cer": 0.11345245313801873,
990
+ "eval_loss": 0.4338010549545288,
991
+ "eval_runtime": 2668.5066,
992
+ "eval_samples_per_second": 6.338,
993
+ "eval_steps_per_second": 0.396,
994
+ "eval_wer": 0.3092706673326515,
995
+ "step": 58000
996
+ },
997
+ {
998
+ "epoch": 27.59,
999
+ "learning_rate": 2.4594249201277952e-05,
1000
+ "loss": 0.5653,
1001
+ "step": 58500
1002
+ },
1003
+ {
1004
+ "epoch": 27.82,
1005
+ "learning_rate": 2.219808306709265e-05,
1006
+ "loss": 0.569,
1007
+ "step": 59000
1008
+ },
1009
+ {
1010
+ "epoch": 28.06,
1011
+ "learning_rate": 1.980670926517572e-05,
1012
+ "loss": 0.5748,
1013
+ "step": 59500
1014
+ },
1015
+ {
1016
+ "epoch": 28.3,
1017
+ "learning_rate": 1.7410543130990413e-05,
1018
+ "loss": 0.5703,
1019
+ "step": 60000
1020
+ },
1021
+ {
1022
+ "epoch": 28.3,
1023
+ "eval_cer": 0.11209795975344294,
1024
+ "eval_loss": 0.42797738313674927,
1025
+ "eval_runtime": 2667.5384,
1026
+ "eval_samples_per_second": 6.341,
1027
+ "eval_steps_per_second": 0.397,
1028
+ "eval_wer": 0.30612309619516526,
1029
+ "step": 60000
1030
+ }
1031
+ ],
1032
+ "logging_steps": 500,
1033
+ "max_steps": 63600,
1034
+ "num_train_epochs": 30,
1035
+ "save_steps": 2000,
1036
+ "total_flos": 4.029225126717586e+20,
1037
+ "trial_name": null,
1038
+ "trial_params": null
1039
+ }
checkpoint-60000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dffbeeea609cd56167ef7580a93c1e52573b1cb0b44833565d5736f74bf636b
3
+ size 4536
checkpoint-60000/vocab.json ADDED
@@ -0,0 +1,1792 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "0": 1,
3
+ "1": 2,
4
+ "2": 3,
5
+ "3": 4,
6
+ "4": 5,
7
+ "5": 6,
8
+ "6": 7,
9
+ "7": 8,
10
+ "8": 9,
11
+ "9": 10,
12
+ "=": 11,
13
+ "[PAD]": 1789,
14
+ "[UNK]": 1788,
15
+ "_": 12,
16
+ "a": 13,
17
+ "b": 14,
18
+ "c": 15,
19
+ "d": 16,
20
+ "e": 17,
21
+ "f": 18,
22
+ "g": 19,
23
+ "h": 20,
24
+ "i": 21,
25
+ "j": 22,
26
+ "k": 23,
27
+ "l": 24,
28
+ "m": 25,
29
+ "n": 26,
30
+ "o": 27,
31
+ "p": 28,
32
+ "q": 29,
33
+ "r": 30,
34
+ "s": 31,
35
+ "t": 32,
36
+ "u": 33,
37
+ "v": 34,
38
+ "w": 35,
39
+ "x": 36,
40
+ "y": 37,
41
+ "z": 38,
42
+ "|": 0,
43
+ "ㄱ": 39,
44
+ "ㄴ": 40,
45
+ "ㄹ": 41,
46
+ "ㅁ": 42,
47
+ "ㅅ": 43,
48
+ "ㅇ": 44,
49
+ "ㅈ": 45,
50
+ "ㅋ": 46,
51
+ "ㅍ": 47,
52
+ "ㅠ": 48,
53
+ "ㅡ": 49,
54
+ "㎛": 50,
55
+ "가": 51,
56
+ "각": 52,
57
+ "간": 53,
58
+ "갇": 54,
59
+ "갈": 55,
60
+ "갉": 56,
61
+ "감": 57,
62
+ "갑": 58,
63
+ "값": 59,
64
+ "갓": 60,
65
+ "갔": 61,
66
+ "강": 62,
67
+ "갖": 63,
68
+ "같": 64,
69
+ "갚": 65,
70
+ "갛": 66,
71
+ "개": 67,
72
+ "객": 68,
73
+ "갠": 69,
74
+ "갤": 70,
75
+ "갬": 71,
76
+ "갭": 72,
77
+ "갯": 73,
78
+ "갰": 74,
79
+ "갱": 75,
80
+ "갸": 76,
81
+ "걀": 77,
82
+ "걍": 78,
83
+ "걔": 79,
84
+ "걘": 80,
85
+ "거": 81,
86
+ "걱": 82,
87
+ "건": 83,
88
+ "걷": 84,
89
+ "걸": 85,
90
+ "검": 86,
91
+ "겁": 87,
92
+ "것": 88,
93
+ "겄": 89,
94
+ "겅": 90,
95
+ "겉": 91,
96
+ "겊": 92,
97
+ "겋": 93,
98
+ "게": 94,
99
+ "겍": 95,
100
+ "겐": 96,
101
+ "겔": 97,
102
+ "겜": 98,
103
+ "겟": 99,
104
+ "겠": 100,
105
+ "겨": 101,
106
+ "격": 102,
107
+ "겪": 103,
108
+ "견": 104,
109
+ "결": 105,
110
+ "겸": 106,
111
+ "겹": 107,
112
+ "겼": 108,
113
+ "경": 109,
114
+ "곁": 110,
115
+ "계": 111,
116
+ "곗": 112,
117
+ "곘": 113,
118
+ "고": 114,
119
+ "곡": 115,
120
+ "곤": 116,
121
+ "곧": 117,
122
+ "골": 118,
123
+ "곪": 119,
124
+ "곯": 120,
125
+ "곰": 121,
126
+ "곱": 122,
127
+ "곳": 123,
128
+ "공": 124,
129
+ "과": 125,
130
+ "곽": 126,
131
+ "관": 127,
132
+ "괄": 128,
133
+ "괌": 129,
134
+ "광": 130,
135
+ "괘": 131,
136
+ "괜": 132,
137
+ "괴": 133,
138
+ "굉": 134,
139
+ "교": 135,
140
+ "굘": 136,
141
+ "굥": 137,
142
+ "구": 138,
143
+ "국": 139,
144
+ "군": 140,
145
+ "굳": 141,
146
+ "굴": 142,
147
+ "굵": 143,
148
+ "굶": 144,
149
+ "굷": 145,
150
+ "굼": 146,
151
+ "굽": 147,
152
+ "굿": 148,
153
+ "궁": 149,
154
+ "궈": 150,
155
+ "권": 151,
156
+ "궐": 152,
157
+ "궜": 153,
158
+ "궤": 154,
159
+ "귀": 155,
160
+ "귄": 156,
161
+ "귓": 157,
162
+ "규": 158,
163
+ "균": 159,
164
+ "귤": 160,
165
+ "그": 161,
166
+ "극": 162,
167
+ "근": 163,
168
+ "귿": 164,
169
+ "글": 165,
170
+ "긁": 166,
171
+ "금": 167,
172
+ "급": 168,
173
+ "긋": 169,
174
+ "긍": 170,
175
+ "긎": 171,
176
+ "긓": 172,
177
+ "기": 173,
178
+ "긱": 174,
179
+ "긴": 175,
180
+ "길": 176,
181
+ "김": 177,
182
+ "깁": 178,
183
+ "깃": 179,
184
+ "깄": 180,
185
+ "깅": 181,
186
+ "깊": 182,
187
+ "까": 183,
188
+ "깍": 184,
189
+ "깎": 185,
190
+ "깐": 186,
191
+ "깔": 187,
192
+ "깜": 188,
193
+ "깝": 189,
194
+ "깠": 190,
195
+ "깡": 191,
196
+ "깥": 192,
197
+ "깨": 193,
198
+ "깬": 194,
199
+ "깰": 195,
200
+ "깸": 196,
201
+ "깹": 197,
202
+ "깻": 198,
203
+ "깼": 199,
204
+ "깽": 200,
205
+ "꺄": 201,
206
+ "꺠": 202,
207
+ "꺵": 203,
208
+ "꺼": 204,
209
+ "꺽": 205,
210
+ "꺾": 206,
211
+ "껀": 207,
212
+ "껄": 208,
213
+ "껌": 209,
214
+ "껍": 210,
215
+ "껏": 211,
216
+ "껐": 212,
217
+ "껑": 213,
218
+ "께": 214,
219
+ "껬": 215,
220
+ "껴": 216,
221
+ "꼈": 217,
222
+ "꼐": 218,
223
+ "꼬": 219,
224
+ "꼭": 220,
225
+ "꼰": 221,
226
+ "꼴": 222,
227
+ "꼼": 223,
228
+ "꼽": 224,
229
+ "꼿": 225,
230
+ "꽁": 226,
231
+ "꽂": 227,
232
+ "꽃": 228,
233
+ "꽈": 229,
234
+ "꽉": 230,
235
+ "꽝": 231,
236
+ "꽤": 232,
237
+ "꽥": 233,
238
+ "꽹": 234,
239
+ "꾀": 235,
240
+ "꾸": 236,
241
+ "꾹": 237,
242
+ "꾼": 238,
243
+ "꿀": 239,
244
+ "꿇": 240,
245
+ "꿈": 241,
246
+ "꿉": 242,
247
+ "꿋": 243,
248
+ "꿍": 244,
249
+ "꿔": 245,
250
+ "꿨": 246,
251
+ "꿩": 247,
252
+ "꿰": 248,
253
+ "뀌": 249,
254
+ "뀐": 250,
255
+ "뀔": 251,
256
+ "뀝": 252,
257
+ "뀨": 253,
258
+ "끄": 254,
259
+ "끅": 255,
260
+ "끈": 256,
261
+ "끊": 257,
262
+ "끌": 258,
263
+ "끍": 259,
264
+ "끓": 260,
265
+ "끔": 261,
266
+ "끕": 262,
267
+ "끗": 263,
268
+ "끙": 264,
269
+ "끝": 265,
270
+ "끼": 266,
271
+ "끽": 267,
272
+ "낀": 268,
273
+ "낄": 269,
274
+ "낌": 270,
275
+ "낍": 271,
276
+ "낑": 272,
277
+ "나": 273,
278
+ "낙": 274,
279
+ "낚": 275,
280
+ "난": 276,
281
+ "날": 277,
282
+ "낡": 278,
283
+ "남": 279,
284
+ "납": 280,
285
+ "낫": 281,
286
+ "났": 282,
287
+ "낭": 283,
288
+ "낮": 284,
289
+ "낯": 285,
290
+ "낱": 286,
291
+ "낳": 287,
292
+ "내": 288,
293
+ "낸": 289,
294
+ "낼": 290,
295
+ "냄": 291,
296
+ "냅": 292,
297
+ "냈": 293,
298
+ "냉": 294,
299
+ "냐": 295,
300
+ "냠": 296,
301
+ "냥": 297,
302
+ "너": 298,
303
+ "넉": 299,
304
+ "넋": 300,
305
+ "넌": 301,
306
+ "널": 302,
307
+ "넒": 303,
308
+ "넓": 304,
309
+ "넘": 305,
310
+ "넙": 306,
311
+ "넛": 307,
312
+ "넜": 308,
313
+ "넝": 309,
314
+ "넣": 310,
315
+ "네": 311,
316
+ "넥": 312,
317
+ "넬": 313,
318
+ "넴": 314,
319
+ "넵": 315,
320
+ "넷": 316,
321
+ "녀": 317,
322
+ "녁": 318,
323
+ "년": 319,
324
+ "념": 320,
325
+ "녔": 321,
326
+ "녕": 322,
327
+ "녜": 323,
328
+ "노": 324,
329
+ "녹": 325,
330
+ "논": 326,
331
+ "놀": 327,
332
+ "놈": 328,
333
+ "놉": 329,
334
+ "농": 330,
335
+ "높": 331,
336
+ "놓": 332,
337
+ "놔": 333,
338
+ "놨": 334,
339
+ "뇌": 335,
340
+ "뇨": 336,
341
+ "뇰": 337,
342
+ "뇽": 338,
343
+ "누": 339,
344
+ "눅": 340,
345
+ "눈": 341,
346
+ "눌": 342,
347
+ "눔": 343,
348
+ "눕": 344,
349
+ "눗": 345,
350
+ "눙": 346,
351
+ "눠": 347,
352
+ "눴": 348,
353
+ "뉘": 349,
354
+ "뉜": 350,
355
+ "뉴": 351,
356
+ "늄": 352,
357
+ "느": 353,
358
+ "늑": 354,
359
+ "는": 355,
360
+ "늘": 356,
361
+ "늙": 357,
362
+ "늠": 358,
363
+ "능": 359,
364
+ "늦": 360,
365
+ "늬": 361,
366
+ "니": 362,
367
+ "닉": 363,
368
+ "닌": 364,
369
+ "닐": 365,
370
+ "님": 366,
371
+ "닙": 367,
372
+ "닛": 368,
373
+ "닝": 369,
374
+ "닢": 370,
375
+ "다": 371,
376
+ "닥": 372,
377
+ "닦": 373,
378
+ "단": 374,
379
+ "닫": 375,
380
+ "달": 376,
381
+ "닭": 377,
382
+ "닮": 378,
383
+ "닳": 379,
384
+ "담": 380,
385
+ "답": 381,
386
+ "닷": 382,
387
+ "당": 383,
388
+ "닻": 384,
389
+ "닿": 385,
390
+ "대": 386,
391
+ "댁": 387,
392
+ "댄": 388,
393
+ "댈": 389,
394
+ "댐": 390,
395
+ "댑": 391,
396
+ "댓": 392,
397
+ "댔": 393,
398
+ "댕": 394,
399
+ "댜": 395,
400
+ "더": 396,
401
+ "덕": 397,
402
+ "던": 398,
403
+ "덜": 399,
404
+ "덞": 400,
405
+ "덟": 401,
406
+ "덤": 402,
407
+ "덥": 403,
408
+ "덧": 404,
409
+ "덩": 405,
410
+ "덫": 406,
411
+ "덮": 407,
412
+ "데": 408,
413
+ "덱": 409,
414
+ "덴": 410,
415
+ "델": 411,
416
+ "뎀": 412,
417
+ "뎁": 413,
418
+ "뎃": 414,
419
+ "뎅": 415,
420
+ "뎌": 416,
421
+ "도": 417,
422
+ "독": 418,
423
+ "돈": 419,
424
+ "돋": 420,
425
+ "돌": 421,
426
+ "돔": 422,
427
+ "돕": 423,
428
+ "돗": 424,
429
+ "동": 425,
430
+ "돠": 426,
431
+ "돼": 427,
432
+ "됄": 428,
433
+ "됐": 429,
434
+ "되": 430,
435
+ "된": 431,
436
+ "될": 432,
437
+ "됨": 433,
438
+ "됩": 434,
439
+ "됬": 435,
440
+ "두": 436,
441
+ "둑": 437,
442
+ "둔": 438,
443
+ "둘": 439,
444
+ "둠": 440,
445
+ "둡": 441,
446
+ "둥": 442,
447
+ "둬": 443,
448
+ "뒀": 444,
449
+ "뒤": 445,
450
+ "뒨": 446,
451
+ "뒷": 447,
452
+ "뒹": 448,
453
+ "듀": 449,
454
+ "듄": 450,
455
+ "드": 451,
456
+ "득": 452,
457
+ "든": 453,
458
+ "듣": 454,
459
+ "들": 455,
460
+ "듦": 456,
461
+ "듬": 457,
462
+ "듭": 458,
463
+ "듯": 459,
464
+ "등": 460,
465
+ "디": 461,
466
+ "딕": 462,
467
+ "딘": 463,
468
+ "딛": 464,
469
+ "딜": 465,
470
+ "딤": 466,
471
+ "딧": 467,
472
+ "딨": 468,
473
+ "딩": 469,
474
+ "딪": 470,
475
+ "따": 471,
476
+ "딱": 472,
477
+ "딲": 473,
478
+ "딴": 474,
479
+ "딸": 475,
480
+ "땀": 476,
481
+ "땄": 477,
482
+ "땅": 478,
483
+ "때": 479,
484
+ "땐": 480,
485
+ "땜": 481,
486
+ "땠": 482,
487
+ "땡": 483,
488
+ "떄": 484,
489
+ "떙": 485,
490
+ "떠": 486,
491
+ "떡": 487,
492
+ "떤": 488,
493
+ "떨": 489,
494
+ "떫": 490,
495
+ "떱": 491,
496
+ "떳": 492,
497
+ "떴": 493,
498
+ "떻": 494,
499
+ "떼": 495,
500
+ "뗀": 496,
501
+ "뗏": 497,
502
+ "뗐": 498,
503
+ "또": 499,
504
+ "똑": 500,
505
+ "똘": 501,
506
+ "똠": 502,
507
+ "똥": 503,
508
+ "뙤": 504,
509
+ "뚜": 505,
510
+ "뚝": 506,
511
+ "뚤": 507,
512
+ "뚫": 508,
513
+ "뚱": 509,
514
+ "뛰": 510,
515
+ "뛴": 511,
516
+ "뛸": 512,
517
+ "뜀": 513,
518
+ "뜁": 514,
519
+ "뜨": 515,
520
+ "뜩": 516,
521
+ "뜬": 517,
522
+ "뜯": 518,
523
+ "뜰": 519,
524
+ "뜸": 520,
525
+ "뜹": 521,
526
+ "뜻": 522,
527
+ "뜽": 523,
528
+ "띄": 524,
529
+ "띈": 525,
530
+ "띌": 526,
531
+ "띕": 527,
532
+ "띠": 528,
533
+ "띡": 529,
534
+ "띤": 530,
535
+ "띵": 531,
536
+ "라": 532,
537
+ "락": 533,
538
+ "란": 534,
539
+ "랄": 535,
540
+ "람": 536,
541
+ "랍": 537,
542
+ "랏": 538,
543
+ "랐": 539,
544
+ "랑": 540,
545
+ "랖": 541,
546
+ "랗": 542,
547
+ "래": 543,
548
+ "랙": 544,
549
+ "랜": 545,
550
+ "랠": 546,
551
+ "램": 547,
552
+ "랩": 548,
553
+ "랫": 549,
554
+ "랬": 550,
555
+ "랭": 551,
556
+ "랴": 552,
557
+ "략": 553,
558
+ "럈": 554,
559
+ "량": 555,
560
+ "러": 556,
561
+ "럭": 557,
562
+ "런": 558,
563
+ "럴": 559,
564
+ "럼": 560,
565
+ "럽": 561,
566
+ "럿": 562,
567
+ "렀": 563,
568
+ "렁": 564,
569
+ "렇": 565,
570
+ "레": 566,
571
+ "렉": 567,
572
+ "렌": 568,
573
+ "렐": 569,
574
+ "렘": 570,
575
+ "렙": 571,
576
+ "렛": 572,
577
+ "렝": 573,
578
+ "려": 574,
579
+ "력": 575,
580
+ "련": 576,
581
+ "렬": 577,
582
+ "렴": 578,
583
+ "렵": 579,
584
+ "렷": 580,
585
+ "렸": 581,
586
+ "령": 582,
587
+ "렼": 583,
588
+ "례": 584,
589
+ "로": 585,
590
+ "록": 586,
591
+ "론": 587,
592
+ "롤": 588,
593
+ "롬": 589,
594
+ "롭": 590,
595
+ "롯": 591,
596
+ "롱": 592,
597
+ "롷": 593,
598
+ "롸": 594,
599
+ "뢰": 595,
600
+ "료": 596,
601
+ "룐": 597,
602
+ "룟": 598,
603
+ "룡": 599,
604
+ "루": 600,
605
+ "룩": 601,
606
+ "룬": 602,
607
+ "룰": 603,
608
+ "룸": 604,
609
+ "룹": 605,
610
+ "룻": 606,
611
+ "룽": 607,
612
+ "뤄": 608,
613
+ "뤘": 609,
614
+ "뤼": 610,
615
+ "류": 611,
616
+ "륙": 612,
617
+ "륜": 613,
618
+ "률": 614,
619
+ "륨": 615,
620
+ "륭": 616,
621
+ "르": 617,
622
+ "륵": 618,
623
+ "른": 619,
624
+ "를": 620,
625
+ "름": 621,
626
+ "릅": 622,
627
+ "릇": 623,
628
+ "릉": 624,
629
+ "릎": 625,
630
+ "릏": 626,
631
+ "리": 627,
632
+ "릭": 628,
633
+ "린": 629,
634
+ "릴": 630,
635
+ "림": 631,
636
+ "립": 632,
637
+ "릿": 633,
638
+ "링": 634,
639
+ "마": 635,
640
+ "막": 636,
641
+ "만": 637,
642
+ "많": 638,
643
+ "맏": 639,
644
+ "말": 640,
645
+ "맑": 641,
646
+ "맘": 642,
647
+ "맙": 643,
648
+ "맛": 644,
649
+ "망": 645,
650
+ "맞": 646,
651
+ "맡": 647,
652
+ "맣": 648,
653
+ "매": 649,
654
+ "맥": 650,
655
+ "맨": 651,
656
+ "맬": 652,
657
+ "맴": 653,
658
+ "맵": 654,
659
+ "맷": 655,
660
+ "맸": 656,
661
+ "맹": 657,
662
+ "맺": 658,
663
+ "먀": 659,
664
+ "머": 660,
665
+ "먹": 661,
666
+ "먼": 662,
667
+ "멀": 663,
668
+ "멈": 664,
669
+ "멉": 665,
670
+ "멋": 666,
671
+ "멍": 667,
672
+ "멎": 668,
673
+ "멓": 669,
674
+ "메": 670,
675
+ "멕": 671,
676
+ "멘": 672,
677
+ "멜": 673,
678
+ "멤": 674,
679
+ "멧": 675,
680
+ "멩": 676,
681
+ "며": 677,
682
+ "멱": 678,
683
+ "면": 679,
684
+ "멸": 680,
685
+ "명": 681,
686
+ "몇": 682,
687
+ "모": 683,
688
+ "목": 684,
689
+ "몫": 685,
690
+ "몬": 686,
691
+ "몰": 687,
692
+ "몸": 688,
693
+ "몹": 689,
694
+ "못": 690,
695
+ "���": 691,
696
+ "묘": 692,
697
+ "무": 693,
698
+ "묵": 694,
699
+ "묶": 695,
700
+ "문": 696,
701
+ "묻": 697,
702
+ "물": 698,
703
+ "묽": 699,
704
+ "뭅": 700,
705
+ "뭇": 701,
706
+ "뭉": 702,
707
+ "뭍": 703,
708
+ "뭐": 704,
709
+ "뭔": 705,
710
+ "뭘": 706,
711
+ "뭡": 707,
712
+ "뭣": 708,
713
+ "뮈": 709,
714
+ "뮤": 710,
715
+ "뮬": 711,
716
+ "뮹": 712,
717
+ "므": 713,
718
+ "믄": 714,
719
+ "믈": 715,
720
+ "미": 716,
721
+ "믹": 717,
722
+ "민": 718,
723
+ "믿": 719,
724
+ "밀": 720,
725
+ "밈": 721,
726
+ "밉": 722,
727
+ "밋": 723,
728
+ "밌": 724,
729
+ "밍": 725,
730
+ "및": 726,
731
+ "밑": 727,
732
+ "바": 728,
733
+ "박": 729,
734
+ "밖": 730,
735
+ "반": 731,
736
+ "받": 732,
737
+ "발": 733,
738
+ "밝": 734,
739
+ "밟": 735,
740
+ "밤": 736,
741
+ "밥": 737,
742
+ "밧": 738,
743
+ "방": 739,
744
+ "밭": 740,
745
+ "배": 741,
746
+ "백": 742,
747
+ "밴": 743,
748
+ "밸": 744,
749
+ "뱀": 745,
750
+ "뱁": 746,
751
+ "뱃": 747,
752
+ "뱅": 748,
753
+ "뱉": 749,
754
+ "버": 750,
755
+ "벅": 751,
756
+ "번": 752,
757
+ "벌": 753,
758
+ "범": 754,
759
+ "법": 755,
760
+ "벗": 756,
761
+ "벙": 757,
762
+ "벚": 758,
763
+ "베": 759,
764
+ "벡": 760,
765
+ "벤": 761,
766
+ "벨": 762,
767
+ "벳": 763,
768
+ "벵": 764,
769
+ "벼": 765,
770
+ "벽": 766,
771
+ "변": 767,
772
+ "별": 768,
773
+ "볍": 769,
774
+ "볐": 770,
775
+ "병": 771,
776
+ "볕": 772,
777
+ "보": 773,
778
+ "복": 774,
779
+ "볶": 775,
780
+ "본": 776,
781
+ "볼": 777,
782
+ "봄": 778,
783
+ "봅": 779,
784
+ "봇": 780,
785
+ "봈": 781,
786
+ "봉": 782,
787
+ "봐": 783,
788
+ "봤": 784,
789
+ "봬": 785,
790
+ "뵀": 786,
791
+ "뵈": 787,
792
+ "뵌": 788,
793
+ "뵐": 789,
794
+ "뵙": 790,
795
+ "뵜": 791,
796
+ "부": 792,
797
+ "북": 793,
798
+ "분": 794,
799
+ "불": 795,
800
+ "붉": 796,
801
+ "붐": 797,
802
+ "붑": 798,
803
+ "붓": 799,
804
+ "붕": 800,
805
+ "붙": 801,
806
+ "뷔": 802,
807
+ "뷰": 803,
808
+ "뷸": 804,
809
+ "브": 805,
810
+ "븐": 806,
811
+ "블": 807,
812
+ "비": 808,
813
+ "빅": 809,
814
+ "빈": 810,
815
+ "빌": 811,
816
+ "빔": 812,
817
+ "빕": 813,
818
+ "빗": 814,
819
+ "빙": 815,
820
+ "빚": 816,
821
+ "빛": 817,
822
+ "빠": 818,
823
+ "빡": 819,
824
+ "빤": 820,
825
+ "빨": 821,
826
+ "빰": 822,
827
+ "빳": 823,
828
+ "빴": 824,
829
+ "빵": 825,
830
+ "빼": 826,
831
+ "빽": 827,
832
+ "뺀": 828,
833
+ "뺄": 829,
834
+ "뺍": 830,
835
+ "뺏": 831,
836
+ "뺐": 832,
837
+ "뺑": 833,
838
+ "뺨": 834,
839
+ "뺭": 835,
840
+ "뺴": 836,
841
+ "뺼": 837,
842
+ "뻇": 838,
843
+ "뻐": 839,
844
+ "뻑": 840,
845
+ "뻔": 841,
846
+ "뻗": 842,
847
+ "뻘": 843,
848
+ "뻣": 844,
849
+ "뻤": 845,
850
+ "뻥": 846,
851
+ "뼈": 847,
852
+ "뼘": 848,
853
+ "뼝": 849,
854
+ "뽀": 850,
855
+ "뽁": 851,
856
+ "뽄": 852,
857
+ "뽑": 853,
858
+ "뽕": 854,
859
+ "뾰": 855,
860
+ "뾱": 856,
861
+ "뿅": 857,
862
+ "뿌": 858,
863
+ "뿍": 859,
864
+ "뿐": 860,
865
+ "뿔": 861,
866
+ "뿜": 862,
867
+ "뿡": 863,
868
+ "쁘": 864,
869
+ "쁜": 865,
870
+ "쁠": 866,
871
+ "쁨": 867,
872
+ "쁩": 868,
873
+ "삐": 869,
874
+ "삔": 870,
875
+ "삘": 871,
876
+ "삥": 872,
877
+ "사": 873,
878
+ "삭": 874,
879
+ "산": 875,
880
+ "살": 876,
881
+ "삶": 877,
882
+ "삼": 878,
883
+ "삽": 879,
884
+ "삿": 880,
885
+ "샀": 881,
886
+ "상": 882,
887
+ "새": 883,
888
+ "색": 884,
889
+ "샊": 885,
890
+ "샌": 886,
891
+ "샐": 887,
892
+ "샘": 888,
893
+ "샙": 889,
894
+ "샛": 890,
895
+ "샜": 891,
896
+ "생": 892,
897
+ "샤": 893,
898
+ "샥": 894,
899
+ "샨": 895,
900
+ "샬": 896,
901
+ "샴": 897,
902
+ "샵": 898,
903
+ "샷": 899,
904
+ "샹": 900,
905
+ "섀": 901,
906
+ "서": 902,
907
+ "석": 903,
908
+ "섞": 904,
909
+ "선": 905,
910
+ "섣": 906,
911
+ "설": 907,
912
+ "섬": 908,
913
+ "섭": 909,
914
+ "섯": 910,
915
+ "섰": 911,
916
+ "성": 912,
917
+ "세": 913,
918
+ "섹": 914,
919
+ "센": 915,
920
+ "셀": 916,
921
+ "셈": 917,
922
+ "셉": 918,
923
+ "셋": 919,
924
+ "셌": 920,
925
+ "셔": 921,
926
+ "션": 922,
927
+ "셜": 923,
928
+ "셥": 924,
929
+ "셧": 925,
930
+ "셨": 926,
931
+ "셰": 927,
932
+ "소": 928,
933
+ "속": 929,
934
+ "솎": 930,
935
+ "손": 931,
936
+ "솔": 932,
937
+ "솜": 933,
938
+ "솟": 934,
939
+ "송": 935,
940
+ "솥": 936,
941
+ "솨": 937,
942
+ "쇄": 938,
943
+ "쇠": 939,
944
+ "쇼": 940,
945
+ "쇽": 941,
946
+ "숀": 942,
947
+ "숌": 943,
948
+ "숍": 944,
949
+ "숏": 945,
950
+ "숑": 946,
951
+ "수": 947,
952
+ "숙": 948,
953
+ "순": 949,
954
+ "숟": 950,
955
+ "술": 951,
956
+ "숨": 952,
957
+ "숩": 953,
958
+ "숫": 954,
959
+ "숭": 955,
960
+ "숯": 956,
961
+ "숱": 957,
962
+ "숲": 958,
963
+ "숴": 959,
964
+ "쉐": 960,
965
+ "쉘": 961,
966
+ "쉣": 962,
967
+ "쉬": 963,
968
+ "쉭": 964,
969
+ "쉰": 965,
970
+ "쉴": 966,
971
+ "쉼": 967,
972
+ "쉽": 968,
973
+ "쉿": 969,
974
+ "슁": 970,
975
+ "슈": 971,
976
+ "슉": 972,
977
+ "슐": 973,
978
+ "슘": 974,
979
+ "슛": 975,
980
+ "슝": 976,
981
+ "스": 977,
982
+ "슥": 978,
983
+ "슨": 979,
984
+ "슬": 980,
985
+ "슴": 981,
986
+ "습": 982,
987
+ "슷": 983,
988
+ "승": 984,
989
+ "시": 985,
990
+ "식": 986,
991
+ "신": 987,
992
+ "싣": 988,
993
+ "실": 989,
994
+ "싫": 990,
995
+ "심": 991,
996
+ "십": 992,
997
+ "싯": 993,
998
+ "싱": 994,
999
+ "싴": 995,
1000
+ "싶": 996,
1001
+ "싸": 997,
1002
+ "싹": 998,
1003
+ "싼": 999,
1004
+ "쌀": 1000,
1005
+ "쌈": 1001,
1006
+ "쌉": 1002,
1007
+ "쌌": 1003,
1008
+ "쌍": 1004,
1009
+ "쌓": 1005,
1010
+ "쌔": 1006,
1011
+ "쌘": 1007,
1012
+ "쌤": 1008,
1013
+ "쌩": 1009,
1014
+ "쌰": 1010,
1015
+ "썜": 1011,
1016
+ "써": 1012,
1017
+ "썩": 1013,
1018
+ "썬": 1014,
1019
+ "썰": 1015,
1020
+ "썸": 1016,
1021
+ "썹": 1017,
1022
+ "썻": 1018,
1023
+ "썼": 1019,
1024
+ "썽": 1020,
1025
+ "쎄": 1021,
1026
+ "쎈": 1022,
1027
+ "쎌": 1023,
1028
+ "쎘": 1024,
1029
+ "쎼": 1025,
1030
+ "쏘": 1026,
1031
+ "쏙": 1027,
1032
+ "쏜": 1028,
1033
+ "쏟": 1029,
1034
+ "쏠": 1030,
1035
+ "쏩": 1031,
1036
+ "쏭": 1032,
1037
+ "쏴": 1033,
1038
+ "쏵": 1034,
1039
+ "쐈": 1035,
1040
+ "쐬": 1036,
1041
+ "쑈": 1037,
1042
+ "쑤": 1038,
1043
+ "쑥": 1039,
1044
+ "쒸": 1040,
1045
+ "쓰": 1041,
1046
+ "쓱": 1042,
1047
+ "쓴": 1043,
1048
+ "쓸": 1044,
1049
+ "씀": 1045,
1050
+ "씁": 1046,
1051
+ "씌": 1047,
1052
+ "씨": 1048,
1053
+ "씩": 1049,
1054
+ "씬": 1050,
1055
+ "씸": 1051,
1056
+ "씹": 1052,
1057
+ "씻": 1053,
1058
+ "씽": 1054,
1059
+ "아": 1055,
1060
+ "악": 1056,
1061
+ "안": 1057,
1062
+ "앉": 1058,
1063
+ "않": 1059,
1064
+ "알": 1060,
1065
+ "앓": 1061,
1066
+ "암": 1062,
1067
+ "압": 1063,
1068
+ "앗": 1064,
1069
+ "았": 1065,
1070
+ "앙": 1066,
1071
+ "앞": 1067,
1072
+ "애": 1068,
1073
+ "액": 1069,
1074
+ "앤": 1070,
1075
+ "앨": 1071,
1076
+ "앰": 1072,
1077
+ "앱": 1073,
1078
+ "앳": 1074,
1079
+ "앴": 1075,
1080
+ "앵": 1076,
1081
+ "야": 1077,
1082
+ "약": 1078,
1083
+ "얀": 1079,
1084
+ "얄": 1080,
1085
+ "얇": 1081,
1086
+ "얌": 1082,
1087
+ "얍": 1083,
1088
+ "양": 1084,
1089
+ "얕": 1085,
1090
+ "얗": 1086,
1091
+ "얘": 1087,
1092
+ "얜": 1088,
1093
+ "어": 1089,
1094
+ "억": 1090,
1095
+ "언": 1091,
1096
+ "얹": 1092,
1097
+ "얻": 1093,
1098
+ "얼": 1094,
1099
+ "얽": 1095,
1100
+ "엄": 1096,
1101
+ "업": 1097,
1102
+ "없": 1098,
1103
+ "엇": 1099,
1104
+ "었": 1100,
1105
+ "엉": 1101,
1106
+ "엊": 1102,
1107
+ "엌": 1103,
1108
+ "엎": 1104,
1109
+ "에": 1105,
1110
+ "엑": 1106,
1111
+ "엔": 1107,
1112
+ "엘": 1108,
1113
+ "엠": 1109,
1114
+ "엣": 1110,
1115
+ "엥": 1111,
1116
+ "여": 1112,
1117
+ "역": 1113,
1118
+ "엮": 1114,
1119
+ "연": 1115,
1120
+ "열": 1116,
1121
+ "염": 1117,
1122
+ "엽": 1118,
1123
+ "엿": 1119,
1124
+ "였": 1120,
1125
+ "영": 1121,
1126
+ "옅": 1122,
1127
+ "옆": 1123,
1128
+ "옇": 1124,
1129
+ "예": 1125,
1130
+ "옌": 1126,
1131
+ "옐": 1127,
1132
+ "옛": 1128,
1133
+ "오": 1129,
1134
+ "옥": 1130,
1135
+ "온": 1131,
1136
+ "올": 1132,
1137
+ "옮": 1133,
1138
+ "옳": 1134,
1139
+ "옴": 1135,
1140
+ "옵": 1136,
1141
+ "옷": 1137,
1142
+ "옹": 1138,
1143
+ "옻": 1139,
1144
+ "와": 1140,
1145
+ "왁": 1141,
1146
+ "완": 1142,
1147
+ "왈": 1143,
1148
+ "왓": 1144,
1149
+ "왔": 1145,
1150
+ "왕": 1146,
1151
+ "왜": 1147,
1152
+ "왠": 1148,
1153
+ "왤": 1149,
1154
+ "외": 1150,
1155
+ "왼": 1151,
1156
+ "욀": 1152,
1157
+ "요": 1153,
1158
+ "욕": 1154,
1159
+ "욘": 1155,
1160
+ "욜": 1156,
1161
+ "욤": 1157,
1162
+ "욥": 1158,
1163
+ "용": 1159,
1164
+ "우": 1160,
1165
+ "욱": 1161,
1166
+ "운": 1162,
1167
+ "울": 1163,
1168
+ "움": 1164,
1169
+ "웁": 1165,
1170
+ "웃": 1166,
1171
+ "웅": 1167,
1172
+ "워": 1168,
1173
+ "웍": 1169,
1174
+ "원": 1170,
1175
+ "월": 1171,
1176
+ "웜": 1172,
1177
+ "웠": 1173,
1178
+ "웨": 1174,
1179
+ "웩": 1175,
1180
+ "웬": 1176,
1181
+ "웰": 1177,
1182
+ "웹": 1178,
1183
+ "웻": 1179,
1184
+ "웽": 1180,
1185
+ "위": 1181,
1186
+ "윅": 1182,
1187
+ "윈": 1183,
1188
+ "윌": 1184,
1189
+ "윔": 1185,
1190
+ "윗": 1186,
1191
+ "윙": 1187,
1192
+ "유": 1188,
1193
+ "육": 1189,
1194
+ "윤": 1190,
1195
+ "율": 1191,
1196
+ "윱": 1192,
1197
+ "윳": 1193,
1198
+ "융": 1194,
1199
+ "윶": 1195,
1200
+ "윷": 1196,
1201
+ "으": 1197,
1202
+ "윽": 1198,
1203
+ "은": 1199,
1204
+ "읃": 1200,
1205
+ "을": 1201,
1206
+ "읊": 1202,
1207
+ "음": 1203,
1208
+ "읍": 1204,
1209
+ "읏": 1205,
1210
+ "응": 1206,
1211
+ "읒": 1207,
1212
+ "의": 1208,
1213
+ "이": 1209,
1214
+ "익": 1210,
1215
+ "인": 1211,
1216
+ "일": 1212,
1217
+ "읽": 1213,
1218
+ "잃": 1214,
1219
+ "임": 1215,
1220
+ "입": 1216,
1221
+ "잇": 1217,
1222
+ "있": 1218,
1223
+ "잉": 1219,
1224
+ "잊": 1220,
1225
+ "잌": 1221,
1226
+ "잎": 1222,
1227
+ "자": 1223,
1228
+ "작": 1224,
1229
+ "잔": 1225,
1230
+ "잖": 1226,
1231
+ "잘": 1227,
1232
+ "잠": 1228,
1233
+ "잡": 1229,
1234
+ "잣": 1230,
1235
+ "잤": 1231,
1236
+ "장": 1232,
1237
+ "잦": 1233,
1238
+ "재": 1234,
1239
+ "잭": 1235,
1240
+ "잰": 1236,
1241
+ "잴": 1237,
1242
+ "잼": 1238,
1243
+ "잽": 1239,
1244
+ "잿": 1240,
1245
+ "쟀": 1241,
1246
+ "쟁": 1242,
1247
+ "쟝": 1243,
1248
+ "쟤": 1244,
1249
+ "쟨": 1245,
1250
+ "저": 1246,
1251
+ "적": 1247,
1252
+ "전": 1248,
1253
+ "절": 1249,
1254
+ "젊": 1250,
1255
+ "점": 1251,
1256
+ "접": 1252,
1257
+ "젓": 1253,
1258
+ "정": 1254,
1259
+ "젖": 1255,
1260
+ "제": 1256,
1261
+ "젝": 1257,
1262
+ "젠": 1258,
1263
+ "젤": 1259,
1264
+ "젬": 1260,
1265
+ "젯": 1261,
1266
+ "져": 1262,
1267
+ "젼": 1263,
1268
+ "졌": 1264,
1269
+ "조": 1265,
1270
+ "족": 1266,
1271
+ "존": 1267,
1272
+ "졸": 1268,
1273
+ "좀": 1269,
1274
+ "좁": 1270,
1275
+ "좃": 1271,
1276
+ "종": 1272,
1277
+ "좆": 1273,
1278
+ "좋": 1274,
1279
+ "좌": 1275,
1280
+ "좍": 1276,
1281
+ "좔": 1277,
1282
+ "죄": 1278,
1283
+ "죠": 1279,
1284
+ "주": 1280,
1285
+ "죽": 1281,
1286
+ "준": 1282,
1287
+ "줄": 1283,
1288
+ "줌": 1284,
1289
+ "줍": 1285,
1290
+ "줏": 1286,
1291
+ "중": 1287,
1292
+ "줘": 1288,
1293
+ "줬": 1289,
1294
+ "줸": 1290,
1295
+ "쥐": 1291,
1296
+ "쥔": 1292,
1297
+ "쥘": 1293,
1298
+ "쥬": 1294,
1299
+ "즈": 1295,
1300
+ "즉": 1296,
1301
+ "즌": 1297,
1302
+ "즐": 1298,
1303
+ "즘": 1299,
1304
+ "즙": 1300,
1305
+ "증": 1301,
1306
+ "지": 1302,
1307
+ "직": 1303,
1308
+ "진": 1304,
1309
+ "질": 1305,
1310
+ "짊": 1306,
1311
+ "짐": 1307,
1312
+ "집": 1308,
1313
+ "짓": 1309,
1314
+ "징": 1310,
1315
+ "짖": 1311,
1316
+ "짚": 1312,
1317
+ "짜": 1313,
1318
+ "짝": 1314,
1319
+ "짠": 1315,
1320
+ "짤": 1316,
1321
+ "짥": 1317,
1322
+ "짧": 1318,
1323
+ "짬": 1319,
1324
+ "짭": 1320,
1325
+ "짯": 1321,
1326
+ "짰": 1322,
1327
+ "짱": 1323,
1328
+ "째": 1324,
1329
+ "짹": 1325,
1330
+ "짼": 1326,
1331
+ "쨋": 1327,
1332
+ "쨌": 1328,
1333
+ "쨍": 1329,
1334
+ "쨔": 1330,
1335
+ "쨰": 1331,
1336
+ "쩃": 1332,
1337
+ "쩄": 1333,
1338
+ "쩌": 1334,
1339
+ "쩍": 1335,
1340
+ "쩐": 1336,
1341
+ "쩔": 1337,
1342
+ "쩜": 1338,
1343
+ "쩝": 1339,
1344
+ "쩠": 1340,
1345
+ "쩡": 1341,
1346
+ "쩨": 1342,
1347
+ "쪄": 1343,
1348
+ "쪘": 1344,
1349
+ "쪼": 1345,
1350
+ "쪽": 1346,
1351
+ "쫀": 1347,
1352
+ "쫄": 1348,
1353
+ "쫌": 1349,
1354
+ "쫍": 1350,
1355
+ "쫑": 1351,
1356
+ "쫒": 1352,
1357
+ "쫓": 1353,
1358
+ "쫘": 1354,
1359
+ "쫙": 1355,
1360
+ "쬐": 1356,
1361
+ "쭈": 1357,
1362
+ "쭉": 1358,
1363
+ "쭌": 1359,
1364
+ "쭐": 1360,
1365
+ "쭘": 1361,
1366
+ "쭙": 1362,
1367
+ "쭝": 1363,
1368
+ "쭤": 1364,
1369
+ "쮸": 1365,
1370
+ "쯔": 1366,
1371
+ "쯕": 1367,
1372
+ "쯤": 1368,
1373
+ "쯧": 1369,
1374
+ "찌": 1370,
1375
+ "찍": 1371,
1376
+ "찐": 1372,
1377
+ "찔": 1373,
1378
+ "찜": 1374,
1379
+ "찝": 1375,
1380
+ "찡": 1376,
1381
+ "찢": 1377,
1382
+ "차": 1378,
1383
+ "착": 1379,
1384
+ "찬": 1380,
1385
+ "찮": 1381,
1386
+ "찰": 1382,
1387
+ "참": 1383,
1388
+ "찹": 1384,
1389
+ "찻": 1385,
1390
+ "찼": 1386,
1391
+ "창": 1387,
1392
+ "찾": 1388,
1393
+ "채": 1389,
1394
+ "책": 1390,
1395
+ "챌": 1391,
1396
+ "챔": 1392,
1397
+ "챕": 1393,
1398
+ "챗": 1394,
1399
+ "챘": 1395,
1400
+ "챙": 1396,
1401
+ "챠": 1397,
1402
+ "챴": 1398,
1403
+ "처": 1399,
1404
+ "척": 1400,
1405
+ "천": 1401,
1406
+ "철": 1402,
1407
+ "첨": 1403,
1408
+ "첩": 1404,
1409
+ "첫": 1405,
1410
+ "청": 1406,
1411
+ "체": 1407,
1412
+ "첵": 1408,
1413
+ "첸": 1409,
1414
+ "첼": 1410,
1415
+ "쳅": 1411,
1416
+ "쳇": 1412,
1417
+ "쳐": 1413,
1418
+ "쳤": 1414,
1419
+ "쳬": 1415,
1420
+ "초": 1416,
1421
+ "촉": 1417,
1422
+ "촌": 1418,
1423
+ "촐": 1419,
1424
+ "촘": 1420,
1425
+ "촛": 1421,
1426
+ "총": 1422,
1427
+ "촤": 1423,
1428
+ "촥": 1424,
1429
+ "촬": 1425,
1430
+ "최": 1426,
1431
+ "쵸": 1427,
1432
+ "추": 1428,
1433
+ "축": 1429,
1434
+ "춘": 1430,
1435
+ "출": 1431,
1436
+ "춤": 1432,
1437
+ "춥": 1433,
1438
+ "춧": 1434,
1439
+ "충": 1435,
1440
+ "춰": 1436,
1441
+ "춱": 1437,
1442
+ "췄": 1438,
1443
+ "췌": 1439,
1444
+ "취": 1440,
1445
+ "췬": 1441,
1446
+ "츄": 1442,
1447
+ "츈": 1443,
1448
+ "츠": 1444,
1449
+ "측": 1445,
1450
+ "츰": 1446,
1451
+ "층": 1447,
1452
+ "치": 1448,
1453
+ "칙": 1449,
1454
+ "친": 1450,
1455
+ "칠": 1451,
1456
+ "칡": 1452,
1457
+ "침": 1453,
1458
+ "칩": 1454,
1459
+ "칫": 1455,
1460
+ "칭": 1456,
1461
+ "카": 1457,
1462
+ "칵": 1458,
1463
+ "칸": 1459,
1464
+ "칼": 1460,
1465
+ "캄": 1461,
1466
+ "캅": 1462,
1467
+ "캇": 1463,
1468
+ "캉": 1464,
1469
+ "캐": 1465,
1470
+ "캔": 1466,
1471
+ "캘": 1467,
1472
+ "캠": 1468,
1473
+ "캡": 1469,
1474
+ "캣": 1470,
1475
+ "캤": 1471,
1476
+ "캥": 1472,
1477
+ "캬": 1473,
1478
+ "커": 1474,
1479
+ "컥": 1475,
1480
+ "컨": 1476,
1481
+ "컬": 1477,
1482
+ "컴": 1478,
1483
+ "컵": 1479,
1484
+ "컷": 1480,
1485
+ "컸": 1481,
1486
+ "케": 1482,
1487
+ "켁": 1483,
1488
+ "켄": 1484,
1489
+ "켈": 1485,
1490
+ "켓": 1486,
1491
+ "켜": 1487,
1492
+ "켠": 1488,
1493
+ "켤": 1489,
1494
+ "켰": 1490,
1495
+ "코": 1491,
1496
+ "콕": 1492,
1497
+ "콘": 1493,
1498
+ "콜": 1494,
1499
+ "콤": 1495,
1500
+ "콥": 1496,
1501
+ "콧": 1497,
1502
+ "콩": 1498,
1503
+ "콱": 1499,
1504
+ "콸": 1500,
1505
+ "쾅": 1501,
1506
+ "쾌": 1502,
1507
+ "쾨": 1503,
1508
+ "쿄": 1504,
1509
+ "쿠": 1505,
1510
+ "쿡": 1506,
1511
+ "쿤": 1507,
1512
+ "쿨": 1508,
1513
+ "쿰": 1509,
1514
+ "쿱": 1510,
1515
+ "쿵": 1511,
1516
+ "쿼": 1512,
1517
+ "퀀": 1513,
1518
+ "퀄": 1514,
1519
+ "퀘": 1515,
1520
+ "퀭": 1516,
1521
+ "퀴": 1517,
1522
+ "퀵": 1518,
1523
+ "퀸": 1519,
1524
+ "퀼": 1520,
1525
+ "큐": 1521,
1526
+ "큘": 1522,
1527
+ "크": 1523,
1528
+ "큰": 1524,
1529
+ "클": 1525,
1530
+ "큼": 1526,
1531
+ "큽": 1527,
1532
+ "킁": 1528,
1533
+ "키": 1529,
1534
+ "킥": 1530,
1535
+ "킨": 1531,
1536
+ "킬": 1532,
1537
+ "킴": 1533,
1538
+ "킵": 1534,
1539
+ "킷": 1535,
1540
+ "킹": 1536,
1541
+ "타": 1537,
1542
+ "탁": 1538,
1543
+ "탄": 1539,
1544
+ "탈": 1540,
1545
+ "탐": 1541,
1546
+ "탑": 1542,
1547
+ "탓": 1543,
1548
+ "탔": 1544,
1549
+ "탕": 1545,
1550
+ "태": 1546,
1551
+ "택": 1547,
1552
+ "탠": 1548,
1553
+ "탤": 1549,
1554
+ "탬": 1550,
1555
+ "탭": 1551,
1556
+ "탯": 1552,
1557
+ "탰": 1553,
1558
+ "탱": 1554,
1559
+ "터": 1555,
1560
+ "턱": 1556,
1561
+ "턴": 1557,
1562
+ "털": 1558,
1563
+ "텀": 1559,
1564
+ "텁": 1560,
1565
+ "텃": 1561,
1566
+ "텅": 1562,
1567
+ "테": 1563,
1568
+ "텍": 1564,
1569
+ "텐": 1565,
1570
+ "텔": 1566,
1571
+ "템": 1567,
1572
+ "텝": 1568,
1573
+ "텟": 1569,
1574
+ "텡": 1570,
1575
+ "텨": 1571,
1576
+ "텼": 1572,
1577
+ "토": 1573,
1578
+ "톡": 1574,
1579
+ "톤": 1575,
1580
+ "톨": 1576,
1581
+ "톰": 1577,
1582
+ "톱": 1578,
1583
+ "톳": 1579,
1584
+ "통": 1580,
1585
+ "퇴": 1581,
1586
+ "투": 1582,
1587
+ "툭": 1583,
1588
+ "툰": 1584,
1589
+ "툴": 1585,
1590
+ "툼": 1586,
1591
+ "퉁": 1587,
1592
+ "튀": 1588,
1593
+ "튄": 1589,
1594
+ "튕": 1590,
1595
+ "튜": 1591,
1596
+ "튠": 1592,
1597
+ "튤": 1593,
1598
+ "튱": 1594,
1599
+ "트": 1595,
1600
+ "특": 1596,
1601
+ "튼": 1597,
1602
+ "튿": 1598,
1603
+ "틀": 1599,
1604
+ "틈": 1600,
1605
+ "틋": 1601,
1606
+ "틍": 1602,
1607
+ "티": 1603,
1608
+ "틱": 1604,
1609
+ "틴": 1605,
1610
+ "틸": 1606,
1611
+ "팀": 1607,
1612
+ "팁": 1608,
1613
+ "팅": 1609,
1614
+ "파": 1610,
1615
+ "팍": 1611,
1616
+ "팎": 1612,
1617
+ "판": 1613,
1618
+ "팔": 1614,
1619
+ "팜": 1615,
1620
+ "팝": 1616,
1621
+ "팟": 1617,
1622
+ "팠": 1618,
1623
+ "팡": 1619,
1624
+ "팥": 1620,
1625
+ "패": 1621,
1626
+ "팩": 1622,
1627
+ "팬": 1623,
1628
+ "팰": 1624,
1629
+ "팸": 1625,
1630
+ "팻": 1626,
1631
+ "팼": 1627,
1632
+ "팽": 1628,
1633
+ "퍼": 1629,
1634
+ "퍽": 1630,
1635
+ "펀": 1631,
1636
+ "펄": 1632,
1637
+ "펌": 1633,
1638
+ "펍": 1634,
1639
+ "펐": 1635,
1640
+ "펑": 1636,
1641
+ "페": 1637,
1642
+ "펙": 1638,
1643
+ "펜": 1639,
1644
+ "펠": 1640,
1645
+ "펩": 1641,
1646
+ "펫": 1642,
1647
+ "펭": 1643,
1648
+ "펴": 1644,
1649
+ "편": 1645,
1650
+ "펼": 1646,
1651
+ "폄": 1647,
1652
+ "폈": 1648,
1653
+ "평": 1649,
1654
+ "폐": 1650,
1655
+ "포": 1651,
1656
+ "폭": 1652,
1657
+ "폰": 1653,
1658
+ "폴": 1654,
1659
+ "폼": 1655,
1660
+ "폿": 1656,
1661
+ "퐁": 1657,
1662
+ "표": 1658,
1663
+ "푯": 1659,
1664
+ "푸": 1660,
1665
+ "푹": 1661,
1666
+ "푼": 1662,
1667
+ "풀": 1663,
1668
+ "품": 1664,
1669
+ "풋": 1665,
1670
+ "풍": 1666,
1671
+ "퓨": 1667,
1672
+ "퓰": 1668,
1673
+ "프": 1669,
1674
+ "픈": 1670,
1675
+ "플": 1671,
1676
+ "픔": 1672,
1677
+ "픕": 1673,
1678
+ "픗": 1674,
1679
+ "피": 1675,
1680
+ "픽": 1676,
1681
+ "핀": 1677,
1682
+ "필": 1678,
1683
+ "핌": 1679,
1684
+ "핍": 1680,
1685
+ "핏": 1681,
1686
+ "핑": 1682,
1687
+ "하": 1683,
1688
+ "학": 1684,
1689
+ "한": 1685,
1690
+ "할": 1686,
1691
+ "핡": 1687,
1692
+ "핥": 1688,
1693
+ "핧": 1689,
1694
+ "함": 1690,
1695
+ "합": 1691,
1696
+ "핫": 1692,
1697
+ "핬": 1693,
1698
+ "항": 1694,
1699
+ "해": 1695,
1700
+ "핵": 1696,
1701
+ "핸": 1697,
1702
+ "핼": 1698,
1703
+ "햄": 1699,
1704
+ "햅": 1700,
1705
+ "햇": 1701,
1706
+ "했": 1702,
1707
+ "행": 1703,
1708
+ "햍": 1704,
1709
+ "햐": 1705,
1710
+ "햝": 1706,
1711
+ "향": 1707,
1712
+ "헀": 1708,
1713
+ "허": 1709,
1714
+ "헉": 1710,
1715
+ "헌": 1711,
1716
+ "헐": 1712,
1717
+ "험": 1713,
1718
+ "헛": 1714,
1719
+ "헝": 1715,
1720
+ "헤": 1716,
1721
+ "헥": 1717,
1722
+ "헨": 1718,
1723
+ "헬": 1719,
1724
+ "헷": 1720,
1725
+ "헹": 1721,
1726
+ "혀": 1722,
1727
+ "혁": 1723,
1728
+ "현": 1724,
1729
+ "혈": 1725,
1730
+ "혐": 1726,
1731
+ "협": 1727,
1732
+ "혓": 1728,
1733
+ "혔": 1729,
1734
+ "형": 1730,
1735
+ "혜": 1731,
1736
+ "호": 1732,
1737
+ "혹": 1733,
1738
+ "혼": 1734,
1739
+ "홀": 1735,
1740
+ "홈": 1736,
1741
+ "홉": 1737,
1742
+ "홋": 1738,
1743
+ "홍": 1739,
1744
+ "홑": 1740,
1745
+ "화": 1741,
1746
+ "확": 1742,
1747
+ "환": 1743,
1748
+ "활": 1744,
1749
+ "홧": 1745,
1750
+ "황": 1746,
1751
+ "홰": 1747,
1752
+ "회": 1748,
1753
+ "획": 1749,
1754
+ "횔": 1750,
1755
+ "횟": 1751,
1756
+ "횡": 1752,
1757
+ "효": 1753,
1758
+ "후": 1754,
1759
+ "훅": 1755,
1760
+ "훈": 1756,
1761
+ "훌": 1757,
1762
+ "훑": 1758,
1763
+ "훔": 1759,
1764
+ "훗": 1760,
1765
+ "훙": 1761,
1766
+ "훡": 1762,
1767
+ "훨": 1763,
1768
+ "훼": 1764,
1769
+ "휘": 1765,
1770
+ "휙": 1766,
1771
+ "휠": 1767,
1772
+ "휩": 1768,
1773
+ "휴": 1769,
1774
+ "흉": 1770,
1775
+ "흐": 1771,
1776
+ "흑": 1772,
1777
+ "흔": 1773,
1778
+ "흘": 1774,
1779
+ "흙": 1775,
1780
+ "흠": 1776,
1781
+ "흡": 1777,
1782
+ "흣": 1778,
1783
+ "흥": 1779,
1784
+ "흩": 1780,
1785
+ "희": 1781,
1786
+ "흰": 1782,
1787
+ "히": 1783,
1788
+ "힌": 1784,
1789
+ "힐": 1785,
1790
+ "힘": 1786,
1791
+ "힙": 1787
1792
+ }
checkpoint-62000/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "</s>": 1791,
3
+ "<s>": 1790
4
+ }
checkpoint-62000/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": 0.1,
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": false,
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": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.0,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_size": 1024,
60
+ "initializer_range": 0.02,
61
+ "intermediate_size": 4096,
62
+ "layer_norm_eps": 1e-05,
63
+ "layerdrop": 0.1,
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": 0.1,
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": 1789,
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": "4.34.1",
114
+ "use_weighted_layer_sum": false,
115
+ "vocab_size": 1792,
116
+ "xvector_output_dim": 512
117
+ }
checkpoint-62000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:661d98cff93e46bacca037c80db947346f7c838a8cec244b58b7f54cdb96e344
3
+ size 2504854198
checkpoint-62000/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-62000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:196351cd8c85539f3cee379731005320bd5ce61543332a6b8644c219041ffbc8
3
+ size 1269249578
checkpoint-62000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efea3d82b885b4b87e35137a5e0e6927ef58e31ec7256f3df17ec467c6e44b51
3
+ size 15024
checkpoint-62000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d103e5704b3f5feddc7fc5d9963ca950913df961ed659c72896bfdba4e8ec59f
3
+ size 15024
checkpoint-62000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:383dd9195375e58b75c9813b92b0def6380e1cd2e3f9d998b2f902db10f73e57
3
+ size 15088
checkpoint-62000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:614be9d84c9779744523ac7e4d90d81b655718f4b2867d71c6269f19eafb6121
3
+ size 15024
checkpoint-62000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b478c2bb2e511201f322eea69eecd0bf9e5e3c641c08bb31d6d642047072d572
3
+ size 1064
checkpoint-62000/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-62000/tokenizer_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "1788": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "1789": {
12
+ "content": "[PAD]",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "1790": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": true,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "1791": {
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-62000/trainer_state.json ADDED
@@ -0,0 +1,1073 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 29.238387172836596,
5
+ "eval_steps": 2000,
6
+ "global_step": 62000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.24,
13
+ "learning_rate": 0.00014849999999999998,
14
+ "loss": 15.6695,
15
+ "step": 500
16
+ },
17
+ {
18
+ "epoch": 0.47,
19
+ "learning_rate": 0.0002982,
20
+ "loss": 4.4987,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.71,
25
+ "learning_rate": 0.00029763258785942486,
26
+ "loss": 2.3147,
27
+ "step": 1500
28
+ },
29
+ {
30
+ "epoch": 0.94,
31
+ "learning_rate": 0.0002952364217252396,
32
+ "loss": 1.7721,
33
+ "step": 2000
34
+ },
35
+ {
36
+ "epoch": 0.94,
37
+ "eval_cer": 0.29026945911872865,
38
+ "eval_loss": 1.1368173360824585,
39
+ "eval_runtime": 2718.6813,
40
+ "eval_samples_per_second": 6.221,
41
+ "eval_steps_per_second": 0.389,
42
+ "eval_wer": 0.6589364856615053,
43
+ "step": 2000
44
+ },
45
+ {
46
+ "epoch": 1.18,
47
+ "learning_rate": 0.0002928402555910543,
48
+ "loss": 1.5735,
49
+ "step": 2500
50
+ },
51
+ {
52
+ "epoch": 1.41,
53
+ "learning_rate": 0.000290444089456869,
54
+ "loss": 1.4611,
55
+ "step": 3000
56
+ },
57
+ {
58
+ "epoch": 1.65,
59
+ "learning_rate": 0.00028805271565495203,
60
+ "loss": 1.4056,
61
+ "step": 3500
62
+ },
63
+ {
64
+ "epoch": 1.89,
65
+ "learning_rate": 0.00028565654952076674,
66
+ "loss": 1.3501,
67
+ "step": 4000
68
+ },
69
+ {
70
+ "epoch": 1.89,
71
+ "eval_cer": 0.2240432390931656,
72
+ "eval_loss": 0.8561204671859741,
73
+ "eval_runtime": 2698.4922,
74
+ "eval_samples_per_second": 6.268,
75
+ "eval_steps_per_second": 0.392,
76
+ "eval_wer": 0.5451323644839807,
77
+ "step": 4000
78
+ },
79
+ {
80
+ "epoch": 2.12,
81
+ "learning_rate": 0.00028326038338658146,
82
+ "loss": 1.2967,
83
+ "step": 4500
84
+ },
85
+ {
86
+ "epoch": 2.36,
87
+ "learning_rate": 0.00028086421725239617,
88
+ "loss": 1.244,
89
+ "step": 5000
90
+ },
91
+ {
92
+ "epoch": 2.59,
93
+ "learning_rate": 0.0002784728434504792,
94
+ "loss": 1.2361,
95
+ "step": 5500
96
+ },
97
+ {
98
+ "epoch": 2.83,
99
+ "learning_rate": 0.0002760766773162939,
100
+ "loss": 1.2133,
101
+ "step": 6000
102
+ },
103
+ {
104
+ "epoch": 2.83,
105
+ "eval_cer": 0.2002927101645895,
106
+ "eval_loss": 0.7505359053611755,
107
+ "eval_runtime": 2721.6856,
108
+ "eval_samples_per_second": 6.215,
109
+ "eval_steps_per_second": 0.389,
110
+ "eval_wer": 0.49743516558444134,
111
+ "step": 6000
112
+ },
113
+ {
114
+ "epoch": 3.07,
115
+ "learning_rate": 0.0002736805111821086,
116
+ "loss": 1.1755,
117
+ "step": 6500
118
+ },
119
+ {
120
+ "epoch": 3.3,
121
+ "learning_rate": 0.0002712843450479233,
122
+ "loss": 1.1403,
123
+ "step": 7000
124
+ },
125
+ {
126
+ "epoch": 3.54,
127
+ "learning_rate": 0.000268888178913738,
128
+ "loss": 1.1253,
129
+ "step": 7500
130
+ },
131
+ {
132
+ "epoch": 3.77,
133
+ "learning_rate": 0.00026649680511182103,
134
+ "loss": 1.0981,
135
+ "step": 8000
136
+ },
137
+ {
138
+ "epoch": 3.77,
139
+ "eval_cer": 0.18418492651709803,
140
+ "eval_loss": 0.6767656803131104,
141
+ "eval_runtime": 2740.3957,
142
+ "eval_samples_per_second": 6.172,
143
+ "eval_steps_per_second": 0.386,
144
+ "eval_wer": 0.46858375155594667,
145
+ "step": 8000
146
+ },
147
+ {
148
+ "epoch": 4.01,
149
+ "learning_rate": 0.00026410063897763575,
150
+ "loss": 1.0984,
151
+ "step": 8500
152
+ },
153
+ {
154
+ "epoch": 4.24,
155
+ "learning_rate": 0.00026170447284345046,
156
+ "loss": 1.0614,
157
+ "step": 9000
158
+ },
159
+ {
160
+ "epoch": 4.48,
161
+ "learning_rate": 0.00025930830670926517,
162
+ "loss": 1.0553,
163
+ "step": 9500
164
+ },
165
+ {
166
+ "epoch": 4.72,
167
+ "learning_rate": 0.0002569169329073482,
168
+ "loss": 1.0375,
169
+ "step": 10000
170
+ },
171
+ {
172
+ "epoch": 4.72,
173
+ "eval_cer": 0.17071197058066542,
174
+ "eval_loss": 0.64134681224823,
175
+ "eval_runtime": 2703.0312,
176
+ "eval_samples_per_second": 6.257,
177
+ "eval_steps_per_second": 0.391,
178
+ "eval_wer": 0.44041417913406117,
179
+ "step": 10000
180
+ },
181
+ {
182
+ "epoch": 4.95,
183
+ "learning_rate": 0.0002545207667731629,
184
+ "loss": 1.0351,
185
+ "step": 10500
186
+ },
187
+ {
188
+ "epoch": 5.19,
189
+ "learning_rate": 0.00025212460063897763,
190
+ "loss": 1.0087,
191
+ "step": 11000
192
+ },
193
+ {
194
+ "epoch": 5.42,
195
+ "learning_rate": 0.00024973322683706067,
196
+ "loss": 0.9934,
197
+ "step": 11500
198
+ },
199
+ {
200
+ "epoch": 5.66,
201
+ "learning_rate": 0.0002473370607028754,
202
+ "loss": 0.9927,
203
+ "step": 12000
204
+ },
205
+ {
206
+ "epoch": 5.66,
207
+ "eval_cer": 0.16338767301962598,
208
+ "eval_loss": 0.6106029152870178,
209
+ "eval_runtime": 2826.5991,
210
+ "eval_samples_per_second": 5.984,
211
+ "eval_steps_per_second": 0.374,
212
+ "eval_wer": 0.42461289632043386,
213
+ "step": 12000
214
+ },
215
+ {
216
+ "epoch": 5.89,
217
+ "learning_rate": 0.0002449408945686901,
218
+ "loss": 0.9821,
219
+ "step": 12500
220
+ },
221
+ {
222
+ "epoch": 6.13,
223
+ "learning_rate": 0.0002425447284345048,
224
+ "loss": 0.9637,
225
+ "step": 13000
226
+ },
227
+ {
228
+ "epoch": 6.37,
229
+ "learning_rate": 0.00024014856230031946,
230
+ "loss": 0.9497,
231
+ "step": 13500
232
+ },
233
+ {
234
+ "epoch": 6.6,
235
+ "learning_rate": 0.00023775239616613414,
236
+ "loss": 0.9439,
237
+ "step": 14000
238
+ },
239
+ {
240
+ "epoch": 6.6,
241
+ "eval_cer": 0.1613024947979602,
242
+ "eval_loss": 0.5999171733856201,
243
+ "eval_runtime": 2741.759,
244
+ "eval_samples_per_second": 6.169,
245
+ "eval_steps_per_second": 0.386,
246
+ "eval_wer": 0.41588373807768236,
247
+ "step": 14000
248
+ },
249
+ {
250
+ "epoch": 6.84,
251
+ "learning_rate": 0.00023536102236421723,
252
+ "loss": 0.948,
253
+ "step": 14500
254
+ },
255
+ {
256
+ "epoch": 7.07,
257
+ "learning_rate": 0.0002329696485623003,
258
+ "loss": 0.9367,
259
+ "step": 15000
260
+ },
261
+ {
262
+ "epoch": 7.31,
263
+ "learning_rate": 0.000230573482428115,
264
+ "loss": 0.9087,
265
+ "step": 15500
266
+ },
267
+ {
268
+ "epoch": 7.55,
269
+ "learning_rate": 0.0002281773162939297,
270
+ "loss": 0.9059,
271
+ "step": 16000
272
+ },
273
+ {
274
+ "epoch": 7.55,
275
+ "eval_cer": 0.15351797484699242,
276
+ "eval_loss": 0.5740103721618652,
277
+ "eval_runtime": 2702.3695,
278
+ "eval_samples_per_second": 6.259,
279
+ "eval_steps_per_second": 0.392,
280
+ "eval_wer": 0.39852056228147376,
281
+ "step": 16000
282
+ },
283
+ {
284
+ "epoch": 7.78,
285
+ "learning_rate": 0.00022578115015974438,
286
+ "loss": 0.9075,
287
+ "step": 16500
288
+ },
289
+ {
290
+ "epoch": 8.02,
291
+ "learning_rate": 0.0002233849840255591,
292
+ "loss": 0.8999,
293
+ "step": 17000
294
+ },
295
+ {
296
+ "epoch": 8.25,
297
+ "learning_rate": 0.00022098881789137377,
298
+ "loss": 0.8597,
299
+ "step": 17500
300
+ },
301
+ {
302
+ "epoch": 8.49,
303
+ "learning_rate": 0.00021859265175718849,
304
+ "loss": 0.8772,
305
+ "step": 18000
306
+ },
307
+ {
308
+ "epoch": 8.49,
309
+ "eval_cer": 0.14781427082015555,
310
+ "eval_loss": 0.5568912625312805,
311
+ "eval_runtime": 2727.4118,
312
+ "eval_samples_per_second": 6.201,
313
+ "eval_steps_per_second": 0.388,
314
+ "eval_wer": 0.3954364182701837,
315
+ "step": 18000
316
+ },
317
+ {
318
+ "epoch": 8.72,
319
+ "learning_rate": 0.00021619648562300317,
320
+ "loss": 0.8785,
321
+ "step": 18500
322
+ },
323
+ {
324
+ "epoch": 8.96,
325
+ "learning_rate": 0.00021380511182108623,
326
+ "loss": 0.8702,
327
+ "step": 19000
328
+ },
329
+ {
330
+ "epoch": 9.2,
331
+ "learning_rate": 0.00021140894568690095,
332
+ "loss": 0.8447,
333
+ "step": 19500
334
+ },
335
+ {
336
+ "epoch": 9.43,
337
+ "learning_rate": 0.00020901277955271563,
338
+ "loss": 0.8483,
339
+ "step": 20000
340
+ },
341
+ {
342
+ "epoch": 9.43,
343
+ "eval_cer": 0.14274528108464166,
344
+ "eval_loss": 0.5406663417816162,
345
+ "eval_runtime": 2824.1098,
346
+ "eval_samples_per_second": 5.989,
347
+ "eval_steps_per_second": 0.375,
348
+ "eval_wer": 0.3784141632772796,
349
+ "step": 20000
350
+ },
351
+ {
352
+ "epoch": 9.67,
353
+ "learning_rate": 0.00020661661341853031,
354
+ "loss": 0.8358,
355
+ "step": 20500
356
+ },
357
+ {
358
+ "epoch": 9.9,
359
+ "learning_rate": 0.00020422523961661338,
360
+ "loss": 0.8391,
361
+ "step": 21000
362
+ },
363
+ {
364
+ "epoch": 10.14,
365
+ "learning_rate": 0.00020182907348242812,
366
+ "loss": 0.8215,
367
+ "step": 21500
368
+ },
369
+ {
370
+ "epoch": 10.37,
371
+ "learning_rate": 0.00019943769968051118,
372
+ "loss": 0.81,
373
+ "step": 22000
374
+ },
375
+ {
376
+ "epoch": 10.37,
377
+ "eval_cer": 0.1415456492625536,
378
+ "eval_loss": 0.5282983779907227,
379
+ "eval_runtime": 2742.622,
380
+ "eval_samples_per_second": 6.167,
381
+ "eval_steps_per_second": 0.386,
382
+ "eval_wer": 0.37441032593614476,
383
+ "step": 22000
384
+ },
385
+ {
386
+ "epoch": 10.61,
387
+ "learning_rate": 0.00019704153354632587,
388
+ "loss": 0.8108,
389
+ "step": 22500
390
+ },
391
+ {
392
+ "epoch": 10.85,
393
+ "learning_rate": 0.00019464536741214058,
394
+ "loss": 0.8175,
395
+ "step": 23000
396
+ },
397
+ {
398
+ "epoch": 11.08,
399
+ "learning_rate": 0.00019224920127795526,
400
+ "loss": 0.8001,
401
+ "step": 23500
402
+ },
403
+ {
404
+ "epoch": 11.32,
405
+ "learning_rate": 0.00018985782747603833,
406
+ "loss": 0.793,
407
+ "step": 24000
408
+ },
409
+ {
410
+ "epoch": 11.32,
411
+ "eval_cer": 0.13662715879199255,
412
+ "eval_loss": 0.517921507358551,
413
+ "eval_runtime": 2729.8979,
414
+ "eval_samples_per_second": 6.196,
415
+ "eval_steps_per_second": 0.388,
416
+ "eval_wer": 0.36633129573690426,
417
+ "step": 24000
418
+ },
419
+ {
420
+ "epoch": 11.55,
421
+ "learning_rate": 0.000187461661341853,
422
+ "loss": 0.7827,
423
+ "step": 24500
424
+ },
425
+ {
426
+ "epoch": 11.79,
427
+ "learning_rate": 0.00018506549520766772,
428
+ "loss": 0.7899,
429
+ "step": 25000
430
+ },
431
+ {
432
+ "epoch": 12.03,
433
+ "learning_rate": 0.0001826693290734824,
434
+ "loss": 0.7806,
435
+ "step": 25500
436
+ },
437
+ {
438
+ "epoch": 12.26,
439
+ "learning_rate": 0.00018027316293929712,
440
+ "loss": 0.7577,
441
+ "step": 26000
442
+ },
443
+ {
444
+ "epoch": 12.26,
445
+ "eval_cer": 0.1359182854425769,
446
+ "eval_loss": 0.5058821439743042,
447
+ "eval_runtime": 2722.7634,
448
+ "eval_samples_per_second": 6.212,
449
+ "eval_steps_per_second": 0.389,
450
+ "eval_wer": 0.35946530932616605,
451
+ "step": 26000
452
+ },
453
+ {
454
+ "epoch": 12.5,
455
+ "learning_rate": 0.00017788178913738016,
456
+ "loss": 0.762,
457
+ "step": 26500
458
+ },
459
+ {
460
+ "epoch": 12.73,
461
+ "learning_rate": 0.00017548562300319487,
462
+ "loss": 0.7595,
463
+ "step": 27000
464
+ },
465
+ {
466
+ "epoch": 12.97,
467
+ "learning_rate": 0.00017308945686900955,
468
+ "loss": 0.7629,
469
+ "step": 27500
470
+ },
471
+ {
472
+ "epoch": 13.2,
473
+ "learning_rate": 0.00017069329073482426,
474
+ "loss": 0.7379,
475
+ "step": 28000
476
+ },
477
+ {
478
+ "epoch": 13.2,
479
+ "eval_cer": 0.13330526921919236,
480
+ "eval_loss": 0.4969228506088257,
481
+ "eval_runtime": 2824.8712,
482
+ "eval_samples_per_second": 5.988,
483
+ "eval_steps_per_second": 0.375,
484
+ "eval_wer": 0.35324945095893884,
485
+ "step": 28000
486
+ },
487
+ {
488
+ "epoch": 13.44,
489
+ "learning_rate": 0.00016830191693290736,
490
+ "loss": 0.737,
491
+ "step": 28500
492
+ },
493
+ {
494
+ "epoch": 13.68,
495
+ "learning_rate": 0.000165905750798722,
496
+ "loss": 0.7444,
497
+ "step": 29000
498
+ },
499
+ {
500
+ "epoch": 13.91,
501
+ "learning_rate": 0.00016350958466453675,
502
+ "loss": 0.7372,
503
+ "step": 29500
504
+ },
505
+ {
506
+ "epoch": 14.15,
507
+ "learning_rate": 0.00016111341853035144,
508
+ "loss": 0.7328,
509
+ "step": 30000
510
+ },
511
+ {
512
+ "epoch": 14.15,
513
+ "eval_cer": 0.13079476698787718,
514
+ "eval_loss": 0.4908413589000702,
515
+ "eval_runtime": 2825.8542,
516
+ "eval_samples_per_second": 5.985,
517
+ "eval_steps_per_second": 0.374,
518
+ "eval_wer": 0.3475251528197322,
519
+ "step": 30000
520
+ },
521
+ {
522
+ "epoch": 14.38,
523
+ "learning_rate": 0.0001587172523961661,
524
+ "loss": 0.7184,
525
+ "step": 30500
526
+ },
527
+ {
528
+ "epoch": 14.62,
529
+ "learning_rate": 0.00015632108626198083,
530
+ "loss": 0.7216,
531
+ "step": 31000
532
+ },
533
+ {
534
+ "epoch": 14.85,
535
+ "learning_rate": 0.0001539297124600639,
536
+ "loss": 0.7238,
537
+ "step": 31500
538
+ },
539
+ {
540
+ "epoch": 15.09,
541
+ "learning_rate": 0.00015153354632587858,
542
+ "loss": 0.7119,
543
+ "step": 32000
544
+ },
545
+ {
546
+ "epoch": 15.09,
547
+ "eval_cer": 0.12864851660072327,
548
+ "eval_loss": 0.4887321889400482,
549
+ "eval_runtime": 2731.0336,
550
+ "eval_samples_per_second": 6.193,
551
+ "eval_steps_per_second": 0.387,
552
+ "eval_wer": 0.34784228845071313,
553
+ "step": 32000
554
+ },
555
+ {
556
+ "epoch": 15.33,
557
+ "learning_rate": 0.00014914217252396165,
558
+ "loss": 0.7124,
559
+ "step": 32500
560
+ },
561
+ {
562
+ "epoch": 15.56,
563
+ "learning_rate": 0.00014674600638977636,
564
+ "loss": 0.7294,
565
+ "step": 33000
566
+ },
567
+ {
568
+ "epoch": 15.8,
569
+ "learning_rate": 0.00014434984025559104,
570
+ "loss": 0.7545,
571
+ "step": 33500
572
+ },
573
+ {
574
+ "epoch": 16.03,
575
+ "learning_rate": 0.00014195367412140575,
576
+ "loss": 0.7572,
577
+ "step": 34000
578
+ },
579
+ {
580
+ "epoch": 16.03,
581
+ "eval_cer": 0.13271854020075294,
582
+ "eval_loss": 0.5169993042945862,
583
+ "eval_runtime": 2729.0002,
584
+ "eval_samples_per_second": 6.198,
585
+ "eval_steps_per_second": 0.388,
586
+ "eval_wer": 0.3576893497926726,
587
+ "step": 34000
588
+ },
589
+ {
590
+ "epoch": 16.27,
591
+ "learning_rate": 0.00013956230031948882,
592
+ "loss": 0.7687,
593
+ "step": 34500
594
+ },
595
+ {
596
+ "epoch": 16.51,
597
+ "learning_rate": 0.0001371661341853035,
598
+ "loss": 0.7884,
599
+ "step": 35000
600
+ },
601
+ {
602
+ "epoch": 16.74,
603
+ "learning_rate": 0.00013476996805111819,
604
+ "loss": 0.8156,
605
+ "step": 35500
606
+ },
607
+ {
608
+ "epoch": 16.98,
609
+ "learning_rate": 0.0001323738019169329,
610
+ "loss": 0.8198,
611
+ "step": 36000
612
+ },
613
+ {
614
+ "epoch": 16.98,
615
+ "eval_cer": 0.1431662427967562,
616
+ "eval_loss": 0.5838645696640015,
617
+ "eval_runtime": 2730.4526,
618
+ "eval_samples_per_second": 6.195,
619
+ "eval_steps_per_second": 0.387,
620
+ "eval_wer": 0.38254485487080686,
621
+ "step": 36000
622
+ },
623
+ {
624
+ "epoch": 17.21,
625
+ "learning_rate": 0.00012997763578274758,
626
+ "loss": 0.819,
627
+ "step": 36500
628
+ },
629
+ {
630
+ "epoch": 17.45,
631
+ "learning_rate": 0.00012758626198083067,
632
+ "loss": 0.8411,
633
+ "step": 37000
634
+ },
635
+ {
636
+ "epoch": 17.68,
637
+ "learning_rate": 0.00012519009584664536,
638
+ "loss": 0.8366,
639
+ "step": 37500
640
+ },
641
+ {
642
+ "epoch": 17.92,
643
+ "learning_rate": 0.00012279392971246007,
644
+ "loss": 0.8008,
645
+ "step": 38000
646
+ },
647
+ {
648
+ "epoch": 17.92,
649
+ "eval_cer": 0.13762394377870937,
650
+ "eval_loss": 0.5447062253952026,
651
+ "eval_runtime": 2738.2931,
652
+ "eval_samples_per_second": 6.177,
653
+ "eval_steps_per_second": 0.386,
654
+ "eval_wer": 0.36609344401366856,
655
+ "step": 38000
656
+ },
657
+ {
658
+ "epoch": 18.16,
659
+ "learning_rate": 0.00012039776357827474,
660
+ "loss": 0.8032,
661
+ "step": 38500
662
+ },
663
+ {
664
+ "epoch": 18.39,
665
+ "learning_rate": 0.00011800159744408944,
666
+ "loss": 0.7753,
667
+ "step": 39000
668
+ },
669
+ {
670
+ "epoch": 18.63,
671
+ "learning_rate": 0.00011560543130990414,
672
+ "loss": 0.7608,
673
+ "step": 39500
674
+ },
675
+ {
676
+ "epoch": 18.86,
677
+ "learning_rate": 0.00011321405750798721,
678
+ "loss": 0.759,
679
+ "step": 40000
680
+ },
681
+ {
682
+ "epoch": 18.86,
683
+ "eval_cer": 0.1336804268071908,
684
+ "eval_loss": 0.49982598423957825,
685
+ "eval_runtime": 2725.5181,
686
+ "eval_samples_per_second": 6.206,
687
+ "eval_steps_per_second": 0.388,
688
+ "eval_wer": 0.3533921619928803,
689
+ "step": 40000
690
+ },
691
+ {
692
+ "epoch": 19.1,
693
+ "learning_rate": 0.00011081789137380191,
694
+ "loss": 0.7285,
695
+ "step": 40500
696
+ },
697
+ {
698
+ "epoch": 19.34,
699
+ "learning_rate": 0.00010842172523961661,
700
+ "loss": 0.7036,
701
+ "step": 41000
702
+ },
703
+ {
704
+ "epoch": 19.57,
705
+ "learning_rate": 0.00010602555910543131,
706
+ "loss": 0.6953,
707
+ "step": 41500
708
+ },
709
+ {
710
+ "epoch": 19.81,
711
+ "learning_rate": 0.00010363418530351436,
712
+ "loss": 0.6907,
713
+ "step": 42000
714
+ },
715
+ {
716
+ "epoch": 19.81,
717
+ "eval_cer": 0.12877502322923437,
718
+ "eval_loss": 0.47100237011909485,
719
+ "eval_runtime": 2667.1801,
720
+ "eval_samples_per_second": 6.342,
721
+ "eval_steps_per_second": 0.397,
722
+ "eval_wer": 0.34119829698166165,
723
+ "step": 42000
724
+ },
725
+ {
726
+ "epoch": 20.04,
727
+ "learning_rate": 0.00010123801916932906,
728
+ "loss": 0.6858,
729
+ "step": 42500
730
+ },
731
+ {
732
+ "epoch": 20.28,
733
+ "learning_rate": 9.884664536741213e-05,
734
+ "loss": 0.6603,
735
+ "step": 43000
736
+ },
737
+ {
738
+ "epoch": 20.51,
739
+ "learning_rate": 9.645047923322683e-05,
740
+ "loss": 0.6609,
741
+ "step": 43500
742
+ },
743
+ {
744
+ "epoch": 20.75,
745
+ "learning_rate": 9.405431309904153e-05,
746
+ "loss": 0.659,
747
+ "step": 44000
748
+ },
749
+ {
750
+ "epoch": 20.75,
751
+ "eval_cer": 0.12423387149543921,
752
+ "eval_loss": 0.4578304886817932,
753
+ "eval_runtime": 2665.8908,
754
+ "eval_samples_per_second": 6.345,
755
+ "eval_steps_per_second": 0.397,
756
+ "eval_wer": 0.3324532819573611,
757
+ "step": 44000
758
+ },
759
+ {
760
+ "epoch": 20.99,
761
+ "learning_rate": 9.165814696485623e-05,
762
+ "loss": 0.6567,
763
+ "step": 44500
764
+ },
765
+ {
766
+ "epoch": 21.22,
767
+ "learning_rate": 8.926198083067093e-05,
768
+ "loss": 0.6437,
769
+ "step": 45000
770
+ },
771
+ {
772
+ "epoch": 21.46,
773
+ "learning_rate": 8.686581469648561e-05,
774
+ "loss": 0.6371,
775
+ "step": 45500
776
+ },
777
+ {
778
+ "epoch": 21.69,
779
+ "learning_rate": 8.447444089456868e-05,
780
+ "loss": 0.6345,
781
+ "step": 46000
782
+ },
783
+ {
784
+ "epoch": 21.69,
785
+ "eval_cer": 0.12205708502554125,
786
+ "eval_loss": 0.45305466651916504,
787
+ "eval_runtime": 2667.1819,
788
+ "eval_samples_per_second": 6.342,
789
+ "eval_steps_per_second": 0.397,
790
+ "eval_wer": 0.3256982930174662,
791
+ "step": 46000
792
+ },
793
+ {
794
+ "epoch": 21.93,
795
+ "learning_rate": 8.207827476038337e-05,
796
+ "loss": 0.6418,
797
+ "step": 46500
798
+ },
799
+ {
800
+ "epoch": 22.16,
801
+ "learning_rate": 7.968210862619807e-05,
802
+ "loss": 0.6306,
803
+ "step": 47000
804
+ },
805
+ {
806
+ "epoch": 22.4,
807
+ "learning_rate": 7.728594249201278e-05,
808
+ "loss": 0.6213,
809
+ "step": 47500
810
+ },
811
+ {
812
+ "epoch": 22.64,
813
+ "learning_rate": 7.489456869009583e-05,
814
+ "loss": 0.6242,
815
+ "step": 48000
816
+ },
817
+ {
818
+ "epoch": 22.64,
819
+ "eval_cer": 0.12094251800538308,
820
+ "eval_loss": 0.4497627019882202,
821
+ "eval_runtime": 2727.8154,
822
+ "eval_samples_per_second": 6.201,
823
+ "eval_steps_per_second": 0.388,
824
+ "eval_wer": 0.32180545314717474,
825
+ "step": 48000
826
+ },
827
+ {
828
+ "epoch": 22.87,
829
+ "learning_rate": 7.249840255591053e-05,
830
+ "loss": 0.6294,
831
+ "step": 48500
832
+ },
833
+ {
834
+ "epoch": 23.11,
835
+ "learning_rate": 7.010223642172524e-05,
836
+ "loss": 0.6141,
837
+ "step": 49000
838
+ },
839
+ {
840
+ "epoch": 23.34,
841
+ "learning_rate": 6.770607028753993e-05,
842
+ "loss": 0.6155,
843
+ "step": 49500
844
+ },
845
+ {
846
+ "epoch": 23.58,
847
+ "learning_rate": 6.530990415335462e-05,
848
+ "loss": 0.6163,
849
+ "step": 50000
850
+ },
851
+ {
852
+ "epoch": 23.58,
853
+ "eval_cer": 0.11941571386818009,
854
+ "eval_loss": 0.45521289110183716,
855
+ "eval_runtime": 2664.8843,
856
+ "eval_samples_per_second": 6.347,
857
+ "eval_steps_per_second": 0.397,
858
+ "eval_wer": 0.3188402349975026,
859
+ "step": 50000
860
+ },
861
+ {
862
+ "epoch": 23.82,
863
+ "learning_rate": 6.291373801916932e-05,
864
+ "loss": 0.6167,
865
+ "step": 50500
866
+ },
867
+ {
868
+ "epoch": 24.05,
869
+ "learning_rate": 6.0522364217252394e-05,
870
+ "loss": 0.6179,
871
+ "step": 51000
872
+ },
873
+ {
874
+ "epoch": 24.29,
875
+ "learning_rate": 5.8126198083067085e-05,
876
+ "loss": 0.6154,
877
+ "step": 51500
878
+ },
879
+ {
880
+ "epoch": 24.52,
881
+ "learning_rate": 5.573482428115016e-05,
882
+ "loss": 0.6121,
883
+ "step": 52000
884
+ },
885
+ {
886
+ "epoch": 24.52,
887
+ "eval_cer": 0.1153500525656853,
888
+ "eval_loss": 0.46334853768348694,
889
+ "eval_runtime": 2666.3375,
890
+ "eval_samples_per_second": 6.344,
891
+ "eval_steps_per_second": 0.397,
892
+ "eval_wer": 0.3136947093848362,
893
+ "step": 52000
894
+ },
895
+ {
896
+ "epoch": 24.76,
897
+ "learning_rate": 5.3338658146964855e-05,
898
+ "loss": 0.6227,
899
+ "step": 52500
900
+ },
901
+ {
902
+ "epoch": 24.99,
903
+ "learning_rate": 5.0942492012779546e-05,
904
+ "loss": 0.6156,
905
+ "step": 53000
906
+ },
907
+ {
908
+ "epoch": 25.23,
909
+ "learning_rate": 4.854632587859424e-05,
910
+ "loss": 0.6159,
911
+ "step": 53500
912
+ },
913
+ {
914
+ "epoch": 25.47,
915
+ "learning_rate": 4.615015974440894e-05,
916
+ "loss": 0.6054,
917
+ "step": 54000
918
+ },
919
+ {
920
+ "epoch": 25.47,
921
+ "eval_cer": 0.11759009234983882,
922
+ "eval_loss": 0.46227386593818665,
923
+ "eval_runtime": 2666.9277,
924
+ "eval_samples_per_second": 6.342,
925
+ "eval_steps_per_second": 0.397,
926
+ "eval_wer": 0.3171356309809798,
927
+ "step": 54000
928
+ },
929
+ {
930
+ "epoch": 25.7,
931
+ "learning_rate": 4.375399361022364e-05,
932
+ "loss": 0.6051,
933
+ "step": 54500
934
+ },
935
+ {
936
+ "epoch": 25.94,
937
+ "learning_rate": 4.1362619808306704e-05,
938
+ "loss": 0.5986,
939
+ "step": 55000
940
+ },
941
+ {
942
+ "epoch": 26.17,
943
+ "learning_rate": 3.89664536741214e-05,
944
+ "loss": 0.5916,
945
+ "step": 55500
946
+ },
947
+ {
948
+ "epoch": 26.41,
949
+ "learning_rate": 3.65702875399361e-05,
950
+ "loss": 0.591,
951
+ "step": 56000
952
+ },
953
+ {
954
+ "epoch": 26.41,
955
+ "eval_cer": 0.11455829556310718,
956
+ "eval_loss": 0.4413212835788727,
957
+ "eval_runtime": 2669.4551,
958
+ "eval_samples_per_second": 6.336,
959
+ "eval_steps_per_second": 0.396,
960
+ "eval_wer": 0.31158575743881267,
961
+ "step": 56000
962
+ },
963
+ {
964
+ "epoch": 26.64,
965
+ "learning_rate": 3.41741214057508e-05,
966
+ "loss": 0.5904,
967
+ "step": 56500
968
+ },
969
+ {
970
+ "epoch": 26.88,
971
+ "learning_rate": 3.178274760383386e-05,
972
+ "loss": 0.5887,
973
+ "step": 57000
974
+ },
975
+ {
976
+ "epoch": 27.12,
977
+ "learning_rate": 2.9386581469648557e-05,
978
+ "loss": 0.5768,
979
+ "step": 57500
980
+ },
981
+ {
982
+ "epoch": 27.35,
983
+ "learning_rate": 2.6990415335463258e-05,
984
+ "loss": 0.5713,
985
+ "step": 58000
986
+ },
987
+ {
988
+ "epoch": 27.35,
989
+ "eval_cer": 0.11345245313801873,
990
+ "eval_loss": 0.4338010549545288,
991
+ "eval_runtime": 2668.5066,
992
+ "eval_samples_per_second": 6.338,
993
+ "eval_steps_per_second": 0.396,
994
+ "eval_wer": 0.3092706673326515,
995
+ "step": 58000
996
+ },
997
+ {
998
+ "epoch": 27.59,
999
+ "learning_rate": 2.4594249201277952e-05,
1000
+ "loss": 0.5653,
1001
+ "step": 58500
1002
+ },
1003
+ {
1004
+ "epoch": 27.82,
1005
+ "learning_rate": 2.219808306709265e-05,
1006
+ "loss": 0.569,
1007
+ "step": 59000
1008
+ },
1009
+ {
1010
+ "epoch": 28.06,
1011
+ "learning_rate": 1.980670926517572e-05,
1012
+ "loss": 0.5748,
1013
+ "step": 59500
1014
+ },
1015
+ {
1016
+ "epoch": 28.3,
1017
+ "learning_rate": 1.7410543130990413e-05,
1018
+ "loss": 0.5703,
1019
+ "step": 60000
1020
+ },
1021
+ {
1022
+ "epoch": 28.3,
1023
+ "eval_cer": 0.11209795975344294,
1024
+ "eval_loss": 0.42797738313674927,
1025
+ "eval_runtime": 2667.5384,
1026
+ "eval_samples_per_second": 6.341,
1027
+ "eval_steps_per_second": 0.397,
1028
+ "eval_wer": 0.30612309619516526,
1029
+ "step": 60000
1030
+ },
1031
+ {
1032
+ "epoch": 28.53,
1033
+ "learning_rate": 1.501437699680511e-05,
1034
+ "loss": 0.5606,
1035
+ "step": 60500
1036
+ },
1037
+ {
1038
+ "epoch": 28.77,
1039
+ "learning_rate": 1.2623003194888177e-05,
1040
+ "loss": 0.5647,
1041
+ "step": 61000
1042
+ },
1043
+ {
1044
+ "epoch": 29.0,
1045
+ "learning_rate": 1.0226837060702875e-05,
1046
+ "loss": 0.5567,
1047
+ "step": 61500
1048
+ },
1049
+ {
1050
+ "epoch": 29.24,
1051
+ "learning_rate": 7.830670926517571e-06,
1052
+ "loss": 0.5576,
1053
+ "step": 62000
1054
+ },
1055
+ {
1056
+ "epoch": 29.24,
1057
+ "eval_cer": 0.11193437359588548,
1058
+ "eval_loss": 0.42482054233551025,
1059
+ "eval_runtime": 2665.5981,
1060
+ "eval_samples_per_second": 6.345,
1061
+ "eval_steps_per_second": 0.397,
1062
+ "eval_wer": 0.30466427229265275,
1063
+ "step": 62000
1064
+ }
1065
+ ],
1066
+ "logging_steps": 500,
1067
+ "max_steps": 63600,
1068
+ "num_train_epochs": 30,
1069
+ "save_steps": 2000,
1070
+ "total_flos": 4.163828299513819e+20,
1071
+ "trial_name": null,
1072
+ "trial_params": null
1073
+ }
checkpoint-62000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dffbeeea609cd56167ef7580a93c1e52573b1cb0b44833565d5736f74bf636b
3
+ size 4536
checkpoint-62000/vocab.json ADDED
@@ -0,0 +1,1792 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "0": 1,
3
+ "1": 2,
4
+ "2": 3,
5
+ "3": 4,
6
+ "4": 5,
7
+ "5": 6,
8
+ "6": 7,
9
+ "7": 8,
10
+ "8": 9,
11
+ "9": 10,
12
+ "=": 11,
13
+ "[PAD]": 1789,
14
+ "[UNK]": 1788,
15
+ "_": 12,
16
+ "a": 13,
17
+ "b": 14,
18
+ "c": 15,
19
+ "d": 16,
20
+ "e": 17,
21
+ "f": 18,
22
+ "g": 19,
23
+ "h": 20,
24
+ "i": 21,
25
+ "j": 22,
26
+ "k": 23,
27
+ "l": 24,
28
+ "m": 25,
29
+ "n": 26,
30
+ "o": 27,
31
+ "p": 28,
32
+ "q": 29,
33
+ "r": 30,
34
+ "s": 31,
35
+ "t": 32,
36
+ "u": 33,
37
+ "v": 34,
38
+ "w": 35,
39
+ "x": 36,
40
+ "y": 37,
41
+ "z": 38,
42
+ "|": 0,
43
+ "ㄱ": 39,
44
+ "ㄴ": 40,
45
+ "ㄹ": 41,
46
+ "ㅁ": 42,
47
+ "ㅅ": 43,
48
+ "ㅇ": 44,
49
+ "ㅈ": 45,
50
+ "ㅋ": 46,
51
+ "ㅍ": 47,
52
+ "ㅠ": 48,
53
+ "ㅡ": 49,
54
+ "㎛": 50,
55
+ "가": 51,
56
+ "각": 52,
57
+ "간": 53,
58
+ "갇": 54,
59
+ "갈": 55,
60
+ "갉": 56,
61
+ "감": 57,
62
+ "갑": 58,
63
+ "값": 59,
64
+ "갓": 60,
65
+ "갔": 61,
66
+ "강": 62,
67
+ "갖": 63,
68
+ "같": 64,
69
+ "갚": 65,
70
+ "갛": 66,
71
+ "개": 67,
72
+ "객": 68,
73
+ "갠": 69,
74
+ "갤": 70,
75
+ "갬": 71,
76
+ "갭": 72,
77
+ "갯": 73,
78
+ "갰": 74,
79
+ "갱": 75,
80
+ "갸": 76,
81
+ "걀": 77,
82
+ "걍": 78,
83
+ "걔": 79,
84
+ "걘": 80,
85
+ "거": 81,
86
+ "걱": 82,
87
+ "건": 83,
88
+ "걷": 84,
89
+ "걸": 85,
90
+ "검": 86,
91
+ "겁": 87,
92
+ "것": 88,
93
+ "겄": 89,
94
+ "겅": 90,
95
+ "겉": 91,
96
+ "겊": 92,
97
+ "겋": 93,
98
+ "게": 94,
99
+ "겍": 95,
100
+ "겐": 96,
101
+ "겔": 97,
102
+ "겜": 98,
103
+ "겟": 99,
104
+ "겠": 100,
105
+ "겨": 101,
106
+ "격": 102,
107
+ "겪": 103,
108
+ "견": 104,
109
+ "결": 105,
110
+ "겸": 106,
111
+ "겹": 107,
112
+ "겼": 108,
113
+ "경": 109,
114
+ "곁": 110,
115
+ "계": 111,
116
+ "곗": 112,
117
+ "곘": 113,
118
+ "고": 114,
119
+ "곡": 115,
120
+ "곤": 116,
121
+ "곧": 117,
122
+ "골": 118,
123
+ "곪": 119,
124
+ "곯": 120,
125
+ "곰": 121,
126
+ "곱": 122,
127
+ "곳": 123,
128
+ "공": 124,
129
+ "과": 125,
130
+ "곽": 126,
131
+ "관": 127,
132
+ "괄": 128,
133
+ "괌": 129,
134
+ "광": 130,
135
+ "괘": 131,
136
+ "괜": 132,
137
+ "괴": 133,
138
+ "굉": 134,
139
+ "교": 135,
140
+ "굘": 136,
141
+ "굥": 137,
142
+ "구": 138,
143
+ "국": 139,
144
+ "군": 140,
145
+ "굳": 141,
146
+ "굴": 142,
147
+ "굵": 143,
148
+ "굶": 144,
149
+ "굷": 145,
150
+ "굼": 146,
151
+ "굽": 147,
152
+ "굿": 148,
153
+ "궁": 149,
154
+ "궈": 150,
155
+ "권": 151,
156
+ "궐": 152,
157
+ "궜": 153,
158
+ "궤": 154,
159
+ "귀": 155,
160
+ "귄": 156,
161
+ "귓": 157,
162
+ "규": 158,
163
+ "균": 159,
164
+ "귤": 160,
165
+ "그": 161,
166
+ "극": 162,
167
+ "근": 163,
168
+ "귿": 164,
169
+ "글": 165,
170
+ "긁": 166,
171
+ "금": 167,
172
+ "급": 168,
173
+ "긋": 169,
174
+ "긍": 170,
175
+ "긎": 171,
176
+ "긓": 172,
177
+ "기": 173,
178
+ "긱": 174,
179
+ "긴": 175,
180
+ "길": 176,
181
+ "김": 177,
182
+ "깁": 178,
183
+ "깃": 179,
184
+ "깄": 180,
185
+ "깅": 181,
186
+ "깊": 182,
187
+ "까": 183,
188
+ "깍": 184,
189
+ "깎": 185,
190
+ "깐": 186,
191
+ "깔": 187,
192
+ "깜": 188,
193
+ "깝": 189,
194
+ "깠": 190,
195
+ "깡": 191,
196
+ "깥": 192,
197
+ "깨": 193,
198
+ "깬": 194,
199
+ "깰": 195,
200
+ "깸": 196,
201
+ "깹": 197,
202
+ "깻": 198,
203
+ "깼": 199,
204
+ "깽": 200,
205
+ "꺄": 201,
206
+ "꺠": 202,
207
+ "꺵": 203,
208
+ "꺼": 204,
209
+ "꺽": 205,
210
+ "꺾": 206,
211
+ "껀": 207,
212
+ "껄": 208,
213
+ "껌": 209,
214
+ "껍": 210,
215
+ "껏": 211,
216
+ "껐": 212,
217
+ "껑": 213,
218
+ "께": 214,
219
+ "껬": 215,
220
+ "껴": 216,
221
+ "꼈": 217,
222
+ "꼐": 218,
223
+ "꼬": 219,
224
+ "꼭": 220,
225
+ "꼰": 221,
226
+ "꼴": 222,
227
+ "꼼": 223,
228
+ "꼽": 224,
229
+ "꼿": 225,
230
+ "꽁": 226,
231
+ "꽂": 227,
232
+ "꽃": 228,
233
+ "꽈": 229,
234
+ "꽉": 230,
235
+ "꽝": 231,
236
+ "꽤": 232,
237
+ "꽥": 233,
238
+ "꽹": 234,
239
+ "꾀": 235,
240
+ "꾸": 236,
241
+ "꾹": 237,
242
+ "꾼": 238,
243
+ "꿀": 239,
244
+ "꿇": 240,
245
+ "꿈": 241,
246
+ "꿉": 242,
247
+ "꿋": 243,
248
+ "꿍": 244,
249
+ "꿔": 245,
250
+ "꿨": 246,
251
+ "꿩": 247,
252
+ "꿰": 248,
253
+ "뀌": 249,
254
+ "뀐": 250,
255
+ "뀔": 251,
256
+ "뀝": 252,
257
+ "뀨": 253,
258
+ "끄": 254,
259
+ "끅": 255,
260
+ "끈": 256,
261
+ "끊": 257,
262
+ "끌": 258,
263
+ "끍": 259,
264
+ "끓": 260,
265
+ "끔": 261,
266
+ "끕": 262,
267
+ "끗": 263,
268
+ "끙": 264,
269
+ "끝": 265,
270
+ "끼": 266,
271
+ "끽": 267,
272
+ "낀": 268,
273
+ "낄": 269,
274
+ "낌": 270,
275
+ "낍": 271,
276
+ "낑": 272,
277
+ "나": 273,
278
+ "낙": 274,
279
+ "낚": 275,
280
+ "난": 276,
281
+ "날": 277,
282
+ "낡": 278,
283
+ "남": 279,
284
+ "납": 280,
285
+ "낫": 281,
286
+ "났": 282,
287
+ "낭": 283,
288
+ "낮": 284,
289
+ "낯": 285,
290
+ "낱": 286,
291
+ "낳": 287,
292
+ "내": 288,
293
+ "낸": 289,
294
+ "낼": 290,
295
+ "냄": 291,
296
+ "냅": 292,
297
+ "냈": 293,
298
+ "냉": 294,
299
+ "냐": 295,
300
+ "냠": 296,
301
+ "냥": 297,
302
+ "너": 298,
303
+ "넉": 299,
304
+ "넋": 300,
305
+ "넌": 301,
306
+ "널": 302,
307
+ "넒": 303,
308
+ "넓": 304,
309
+ "넘": 305,
310
+ "넙": 306,
311
+ "넛": 307,
312
+ "넜": 308,
313
+ "넝": 309,
314
+ "넣": 310,
315
+ "네": 311,
316
+ "넥": 312,
317
+ "넬": 313,
318
+ "넴": 314,
319
+ "넵": 315,
320
+ "넷": 316,
321
+ "녀": 317,
322
+ "녁": 318,
323
+ "년": 319,
324
+ "념": 320,
325
+ "녔": 321,
326
+ "녕": 322,
327
+ "녜": 323,
328
+ "노": 324,
329
+ "녹": 325,
330
+ "논": 326,
331
+ "놀": 327,
332
+ "놈": 328,
333
+ "놉": 329,
334
+ "농": 330,
335
+ "높": 331,
336
+ "놓": 332,
337
+ "놔": 333,
338
+ "놨": 334,
339
+ "뇌": 335,
340
+ "뇨": 336,
341
+ "뇰": 337,
342
+ "뇽": 338,
343
+ "누": 339,
344
+ "눅": 340,
345
+ "눈": 341,
346
+ "눌": 342,
347
+ "눔": 343,
348
+ "눕": 344,
349
+ "눗": 345,
350
+ "눙": 346,
351
+ "눠": 347,
352
+ "눴": 348,
353
+ "뉘": 349,
354
+ "뉜": 350,
355
+ "뉴": 351,
356
+ "늄": 352,
357
+ "느": 353,
358
+ "늑": 354,
359
+ "는": 355,
360
+ "늘": 356,
361
+ "늙": 357,
362
+ "늠": 358,
363
+ "능": 359,
364
+ "늦": 360,
365
+ "늬": 361,
366
+ "니": 362,
367
+ "닉": 363,
368
+ "닌": 364,
369
+ "닐": 365,
370
+ "님": 366,
371
+ "닙": 367,
372
+ "닛": 368,
373
+ "닝": 369,
374
+ "닢": 370,
375
+ "다": 371,
376
+ "닥": 372,
377
+ "닦": 373,
378
+ "단": 374,
379
+ "닫": 375,
380
+ "달": 376,
381
+ "닭": 377,
382
+ "닮": 378,
383
+ "닳": 379,
384
+ "담": 380,
385
+ "답": 381,
386
+ "닷": 382,
387
+ "당": 383,
388
+ "닻": 384,
389
+ "닿": 385,
390
+ "대": 386,
391
+ "댁": 387,
392
+ "댄": 388,
393
+ "댈": 389,
394
+ "댐": 390,
395
+ "댑": 391,
396
+ "댓": 392,
397
+ "댔": 393,
398
+ "댕": 394,
399
+ "댜": 395,
400
+ "더": 396,
401
+ "덕": 397,
402
+ "던": 398,
403
+ "덜": 399,
404
+ "덞": 400,
405
+ "덟": 401,
406
+ "덤": 402,
407
+ "덥": 403,
408
+ "덧": 404,
409
+ "덩": 405,
410
+ "덫": 406,
411
+ "덮": 407,
412
+ "데": 408,
413
+ "덱": 409,
414
+ "덴": 410,
415
+ "델": 411,
416
+ "뎀": 412,
417
+ "뎁": 413,
418
+ "뎃": 414,
419
+ "뎅": 415,
420
+ "뎌": 416,
421
+ "도": 417,
422
+ "독": 418,
423
+ "돈": 419,
424
+ "돋": 420,
425
+ "돌": 421,
426
+ "돔": 422,
427
+ "돕": 423,
428
+ "돗": 424,
429
+ "동": 425,
430
+ "돠": 426,
431
+ "돼": 427,
432
+ "됄": 428,
433
+ "됐": 429,
434
+ "되": 430,
435
+ "된": 431,
436
+ "될": 432,
437
+ "됨": 433,
438
+ "됩": 434,
439
+ "됬": 435,
440
+ "두": 436,
441
+ "둑": 437,
442
+ "둔": 438,
443
+ "둘": 439,
444
+ "둠": 440,
445
+ "둡": 441,
446
+ "둥": 442,
447
+ "둬": 443,
448
+ "뒀": 444,
449
+ "뒤": 445,
450
+ "뒨": 446,
451
+ "뒷": 447,
452
+ "뒹": 448,
453
+ "듀": 449,
454
+ "듄": 450,
455
+ "드": 451,
456
+ "득": 452,
457
+ "든": 453,
458
+ "듣": 454,
459
+ "들": 455,
460
+ "듦": 456,
461
+ "듬": 457,
462
+ "듭": 458,
463
+ "듯": 459,
464
+ "등": 460,
465
+ "디": 461,
466
+ "딕": 462,
467
+ "딘": 463,
468
+ "딛": 464,
469
+ "딜": 465,
470
+ "딤": 466,
471
+ "딧": 467,
472
+ "딨": 468,
473
+ "딩": 469,
474
+ "딪": 470,
475
+ "따": 471,
476
+ "딱": 472,
477
+ "딲": 473,
478
+ "딴": 474,
479
+ "딸": 475,
480
+ "땀": 476,
481
+ "땄": 477,
482
+ "땅": 478,
483
+ "때": 479,
484
+ "땐": 480,
485
+ "땜": 481,
486
+ "땠": 482,
487
+ "땡": 483,
488
+ "떄": 484,
489
+ "떙": 485,
490
+ "떠": 486,
491
+ "떡": 487,
492
+ "떤": 488,
493
+ "떨": 489,
494
+ "떫": 490,
495
+ "떱": 491,
496
+ "떳": 492,
497
+ "떴": 493,
498
+ "떻": 494,
499
+ "떼": 495,
500
+ "뗀": 496,
501
+ "뗏": 497,
502
+ "뗐": 498,
503
+ "또": 499,
504
+ "똑": 500,
505
+ "똘": 501,
506
+ "똠": 502,
507
+ "똥": 503,
508
+ "뙤": 504,
509
+ "뚜": 505,
510
+ "뚝": 506,
511
+ "뚤": 507,
512
+ "뚫": 508,
513
+ "뚱": 509,
514
+ "뛰": 510,
515
+ "뛴": 511,
516
+ "뛸": 512,
517
+ "뜀": 513,
518
+ "뜁": 514,
519
+ "뜨": 515,
520
+ "뜩": 516,
521
+ "뜬": 517,
522
+ "뜯": 518,
523
+ "뜰": 519,
524
+ "뜸": 520,
525
+ "뜹": 521,
526
+ "뜻": 522,
527
+ "뜽": 523,
528
+ "띄": 524,
529
+ "띈": 525,
530
+ "띌": 526,
531
+ "띕": 527,
532
+ "띠": 528,
533
+ "띡": 529,
534
+ "띤": 530,
535
+ "띵": 531,
536
+ "라": 532,
537
+ "락": 533,
538
+ "란": 534,
539
+ "랄": 535,
540
+ "람": 536,
541
+ "랍": 537,
542
+ "랏": 538,
543
+ "랐": 539,
544
+ "랑": 540,
545
+ "랖": 541,
546
+ "랗": 542,
547
+ "래": 543,
548
+ "랙": 544,
549
+ "랜": 545,
550
+ "랠": 546,
551
+ "램": 547,
552
+ "랩": 548,
553
+ "랫": 549,
554
+ "랬": 550,
555
+ "랭": 551,
556
+ "랴": 552,
557
+ "략": 553,
558
+ "럈": 554,
559
+ "량": 555,
560
+ "러": 556,
561
+ "럭": 557,
562
+ "런": 558,
563
+ "럴": 559,
564
+ "럼": 560,
565
+ "럽": 561,
566
+ "럿": 562,
567
+ "렀": 563,
568
+ "렁": 564,
569
+ "렇": 565,
570
+ "레": 566,
571
+ "렉": 567,
572
+ "렌": 568,
573
+ "렐": 569,
574
+ "렘": 570,
575
+ "렙": 571,
576
+ "렛": 572,
577
+ "렝": 573,
578
+ "려": 574,
579
+ "력": 575,
580
+ "련": 576,
581
+ "렬": 577,
582
+ "렴": 578,
583
+ "렵": 579,
584
+ "렷": 580,
585
+ "렸": 581,
586
+ "령": 582,
587
+ "렼": 583,
588
+ "례": 584,
589
+ "로": 585,
590
+ "록": 586,
591
+ "론": 587,
592
+ "롤": 588,
593
+ "롬": 589,
594
+ "롭": 590,
595
+ "롯": 591,
596
+ "롱": 592,
597
+ "롷": 593,
598
+ "롸": 594,
599
+ "뢰": 595,
600
+ "료": 596,
601
+ "룐": 597,
602
+ "룟": 598,
603
+ "룡": 599,
604
+ "루": 600,
605
+ "룩": 601,
606
+ "룬": 602,
607
+ "룰": 603,
608
+ "룸": 604,
609
+ "룹": 605,
610
+ "룻": 606,
611
+ "룽": 607,
612
+ "뤄": 608,
613
+ "뤘": 609,
614
+ "뤼": 610,
615
+ "류": 611,
616
+ "륙": 612,
617
+ "륜": 613,
618
+ "률": 614,
619
+ "륨": 615,
620
+ "륭": 616,
621
+ "르": 617,
622
+ "륵": 618,
623
+ "른": 619,
624
+ "를": 620,
625
+ "름": 621,
626
+ "릅": 622,
627
+ "릇": 623,
628
+ "릉": 624,
629
+ "릎": 625,
630
+ "릏": 626,
631
+ "리": 627,
632
+ "릭": 628,
633
+ "린": 629,
634
+ "릴": 630,
635
+ "림": 631,
636
+ "립": 632,
637
+ "릿": 633,
638
+ "링": 634,
639
+ "마": 635,
640
+ "막": 636,
641
+ "만": 637,
642
+ "많": 638,
643
+ "맏": 639,
644
+ "말": 640,
645
+ "맑": 641,
646
+ "맘": 642,
647
+ "맙": 643,
648
+ "맛": 644,
649
+ "망": 645,
650
+ "맞": 646,
651
+ "맡": 647,
652
+ "맣": 648,
653
+ "매": 649,
654
+ "맥": 650,
655
+ "맨": 651,
656
+ "맬": 652,
657
+ "맴": 653,
658
+ "맵": 654,
659
+ "맷": 655,
660
+ "맸": 656,
661
+ "맹": 657,
662
+ "맺": 658,
663
+ "먀": 659,
664
+ "머": 660,
665
+ "먹": 661,
666
+ "먼": 662,
667
+ "멀": 663,
668
+ "멈": 664,
669
+ "멉": 665,
670
+ "멋": 666,
671
+ "멍": 667,
672
+ "멎": 668,
673
+ "멓": 669,
674
+ "메": 670,
675
+ "멕": 671,
676
+ "멘": 672,
677
+ "멜": 673,
678
+ "멤": 674,
679
+ "멧": 675,
680
+ "멩": 676,
681
+ "며": 677,
682
+ "멱": 678,
683
+ "면": 679,
684
+ "멸": 680,
685
+ "명": 681,
686
+ "몇": 682,
687
+ "모": 683,
688
+ "목": 684,
689
+ "몫": 685,
690
+ "몬": 686,
691
+ "몰": 687,
692
+ "몸": 688,
693
+ "몹": 689,
694
+ "못": 690,
695
+ "���": 691,
696
+ "묘": 692,
697
+ "무": 693,
698
+ "묵": 694,
699
+ "묶": 695,
700
+ "문": 696,
701
+ "묻": 697,
702
+ "물": 698,
703
+ "묽": 699,
704
+ "뭅": 700,
705
+ "뭇": 701,
706
+ "뭉": 702,
707
+ "뭍": 703,
708
+ "뭐": 704,
709
+ "뭔": 705,
710
+ "뭘": 706,
711
+ "뭡": 707,
712
+ "뭣": 708,
713
+ "뮈": 709,
714
+ "뮤": 710,
715
+ "뮬": 711,
716
+ "뮹": 712,
717
+ "므": 713,
718
+ "믄": 714,
719
+ "믈": 715,
720
+ "미": 716,
721
+ "믹": 717,
722
+ "민": 718,
723
+ "믿": 719,
724
+ "밀": 720,
725
+ "밈": 721,
726
+ "밉": 722,
727
+ "밋": 723,
728
+ "밌": 724,
729
+ "밍": 725,
730
+ "및": 726,
731
+ "밑": 727,
732
+ "바": 728,
733
+ "박": 729,
734
+ "밖": 730,
735
+ "반": 731,
736
+ "받": 732,
737
+ "발": 733,
738
+ "밝": 734,
739
+ "밟": 735,
740
+ "밤": 736,
741
+ "밥": 737,
742
+ "밧": 738,
743
+ "방": 739,
744
+ "밭": 740,
745
+ "배": 741,
746
+ "백": 742,
747
+ "밴": 743,
748
+ "밸": 744,
749
+ "뱀": 745,
750
+ "뱁": 746,
751
+ "뱃": 747,
752
+ "뱅": 748,
753
+ "뱉": 749,
754
+ "버": 750,
755
+ "벅": 751,
756
+ "번": 752,
757
+ "벌": 753,
758
+ "범": 754,
759
+ "법": 755,
760
+ "벗": 756,
761
+ "벙": 757,
762
+ "벚": 758,
763
+ "베": 759,
764
+ "벡": 760,
765
+ "벤": 761,
766
+ "벨": 762,
767
+ "벳": 763,
768
+ "벵": 764,
769
+ "벼": 765,
770
+ "벽": 766,
771
+ "변": 767,
772
+ "별": 768,
773
+ "볍": 769,
774
+ "볐": 770,
775
+ "병": 771,
776
+ "볕": 772,
777
+ "보": 773,
778
+ "복": 774,
779
+ "볶": 775,
780
+ "본": 776,
781
+ "볼": 777,
782
+ "봄": 778,
783
+ "봅": 779,
784
+ "봇": 780,
785
+ "봈": 781,
786
+ "봉": 782,
787
+ "봐": 783,
788
+ "봤": 784,
789
+ "봬": 785,
790
+ "뵀": 786,
791
+ "뵈": 787,
792
+ "뵌": 788,
793
+ "뵐": 789,
794
+ "뵙": 790,
795
+ "뵜": 791,
796
+ "부": 792,
797
+ "북": 793,
798
+ "분": 794,
799
+ "불": 795,
800
+ "붉": 796,
801
+ "붐": 797,
802
+ "붑": 798,
803
+ "붓": 799,
804
+ "붕": 800,
805
+ "붙": 801,
806
+ "뷔": 802,
807
+ "뷰": 803,
808
+ "뷸": 804,
809
+ "브": 805,
810
+ "븐": 806,
811
+ "블": 807,
812
+ "비": 808,
813
+ "빅": 809,
814
+ "빈": 810,
815
+ "빌": 811,
816
+ "빔": 812,
817
+ "빕": 813,
818
+ "빗": 814,
819
+ "빙": 815,
820
+ "빚": 816,
821
+ "빛": 817,
822
+ "빠": 818,
823
+ "빡": 819,
824
+ "빤": 820,
825
+ "빨": 821,
826
+ "빰": 822,
827
+ "빳": 823,
828
+ "빴": 824,
829
+ "빵": 825,
830
+ "빼": 826,
831
+ "빽": 827,
832
+ "뺀": 828,
833
+ "뺄": 829,
834
+ "뺍": 830,
835
+ "뺏": 831,
836
+ "뺐": 832,
837
+ "뺑": 833,
838
+ "뺨": 834,
839
+ "뺭": 835,
840
+ "뺴": 836,
841
+ "뺼": 837,
842
+ "뻇": 838,
843
+ "뻐": 839,
844
+ "뻑": 840,
845
+ "뻔": 841,
846
+ "뻗": 842,
847
+ "뻘": 843,
848
+ "뻣": 844,
849
+ "뻤": 845,
850
+ "뻥": 846,
851
+ "뼈": 847,
852
+ "뼘": 848,
853
+ "뼝": 849,
854
+ "뽀": 850,
855
+ "뽁": 851,
856
+ "뽄": 852,
857
+ "뽑": 853,
858
+ "뽕": 854,
859
+ "뾰": 855,
860
+ "뾱": 856,
861
+ "뿅": 857,
862
+ "뿌": 858,
863
+ "뿍": 859,
864
+ "뿐": 860,
865
+ "뿔": 861,
866
+ "뿜": 862,
867
+ "뿡": 863,
868
+ "쁘": 864,
869
+ "쁜": 865,
870
+ "쁠": 866,
871
+ "쁨": 867,
872
+ "쁩": 868,
873
+ "삐": 869,
874
+ "삔": 870,
875
+ "삘": 871,
876
+ "삥": 872,
877
+ "사": 873,
878
+ "삭": 874,
879
+ "산": 875,
880
+ "살": 876,
881
+ "삶": 877,
882
+ "삼": 878,
883
+ "삽": 879,
884
+ "삿": 880,
885
+ "샀": 881,
886
+ "상": 882,
887
+ "새": 883,
888
+ "색": 884,
889
+ "샊": 885,
890
+ "샌": 886,
891
+ "샐": 887,
892
+ "샘": 888,
893
+ "샙": 889,
894
+ "샛": 890,
895
+ "샜": 891,
896
+ "생": 892,
897
+ "샤": 893,
898
+ "샥": 894,
899
+ "샨": 895,
900
+ "샬": 896,
901
+ "샴": 897,
902
+ "샵": 898,
903
+ "샷": 899,
904
+ "샹": 900,
905
+ "섀": 901,
906
+ "서": 902,
907
+ "석": 903,
908
+ "섞": 904,
909
+ "선": 905,
910
+ "섣": 906,
911
+ "설": 907,
912
+ "섬": 908,
913
+ "섭": 909,
914
+ "섯": 910,
915
+ "섰": 911,
916
+ "성": 912,
917
+ "세": 913,
918
+ "섹": 914,
919
+ "센": 915,
920
+ "셀": 916,
921
+ "셈": 917,
922
+ "셉": 918,
923
+ "셋": 919,
924
+ "셌": 920,
925
+ "셔": 921,
926
+ "션": 922,
927
+ "셜": 923,
928
+ "셥": 924,
929
+ "셧": 925,
930
+ "셨": 926,
931
+ "셰": 927,
932
+ "소": 928,
933
+ "속": 929,
934
+ "솎": 930,
935
+ "손": 931,
936
+ "솔": 932,
937
+ "솜": 933,
938
+ "솟": 934,
939
+ "송": 935,
940
+ "솥": 936,
941
+ "솨": 937,
942
+ "쇄": 938,
943
+ "쇠": 939,
944
+ "쇼": 940,
945
+ "쇽": 941,
946
+ "숀": 942,
947
+ "숌": 943,
948
+ "숍": 944,
949
+ "숏": 945,
950
+ "숑": 946,
951
+ "수": 947,
952
+ "숙": 948,
953
+ "순": 949,
954
+ "숟": 950,
955
+ "술": 951,
956
+ "숨": 952,
957
+ "숩": 953,
958
+ "숫": 954,
959
+ "숭": 955,
960
+ "숯": 956,
961
+ "숱": 957,
962
+ "숲": 958,
963
+ "숴": 959,
964
+ "쉐": 960,
965
+ "쉘": 961,
966
+ "쉣": 962,
967
+ "쉬": 963,
968
+ "쉭": 964,
969
+ "쉰": 965,
970
+ "쉴": 966,
971
+ "쉼": 967,
972
+ "쉽": 968,
973
+ "쉿": 969,
974
+ "슁": 970,
975
+ "슈": 971,
976
+ "슉": 972,
977
+ "슐": 973,
978
+ "슘": 974,
979
+ "슛": 975,
980
+ "슝": 976,
981
+ "스": 977,
982
+ "슥": 978,
983
+ "슨": 979,
984
+ "슬": 980,
985
+ "슴": 981,
986
+ "습": 982,
987
+ "슷": 983,
988
+ "승": 984,
989
+ "시": 985,
990
+ "식": 986,
991
+ "신": 987,
992
+ "싣": 988,
993
+ "실": 989,
994
+ "싫": 990,
995
+ "심": 991,
996
+ "십": 992,
997
+ "싯": 993,
998
+ "싱": 994,
999
+ "싴": 995,
1000
+ "싶": 996,
1001
+ "싸": 997,
1002
+ "싹": 998,
1003
+ "싼": 999,
1004
+ "쌀": 1000,
1005
+ "쌈": 1001,
1006
+ "쌉": 1002,
1007
+ "쌌": 1003,
1008
+ "쌍": 1004,
1009
+ "쌓": 1005,
1010
+ "쌔": 1006,
1011
+ "쌘": 1007,
1012
+ "쌤": 1008,
1013
+ "쌩": 1009,
1014
+ "쌰": 1010,
1015
+ "썜": 1011,
1016
+ "써": 1012,
1017
+ "썩": 1013,
1018
+ "썬": 1014,
1019
+ "썰": 1015,
1020
+ "썸": 1016,
1021
+ "썹": 1017,
1022
+ "썻": 1018,
1023
+ "썼": 1019,
1024
+ "썽": 1020,
1025
+ "쎄": 1021,
1026
+ "쎈": 1022,
1027
+ "쎌": 1023,
1028
+ "쎘": 1024,
1029
+ "쎼": 1025,
1030
+ "쏘": 1026,
1031
+ "쏙": 1027,
1032
+ "쏜": 1028,
1033
+ "쏟": 1029,
1034
+ "쏠": 1030,
1035
+ "쏩": 1031,
1036
+ "쏭": 1032,
1037
+ "쏴": 1033,
1038
+ "쏵": 1034,
1039
+ "쐈": 1035,
1040
+ "쐬": 1036,
1041
+ "쑈": 1037,
1042
+ "쑤": 1038,
1043
+ "쑥": 1039,
1044
+ "쒸": 1040,
1045
+ "쓰": 1041,
1046
+ "쓱": 1042,
1047
+ "쓴": 1043,
1048
+ "쓸": 1044,
1049
+ "씀": 1045,
1050
+ "씁": 1046,
1051
+ "씌": 1047,
1052
+ "씨": 1048,
1053
+ "씩": 1049,
1054
+ "씬": 1050,
1055
+ "씸": 1051,
1056
+ "씹": 1052,
1057
+ "씻": 1053,
1058
+ "씽": 1054,
1059
+ "아": 1055,
1060
+ "악": 1056,
1061
+ "안": 1057,
1062
+ "앉": 1058,
1063
+ "않": 1059,
1064
+ "알": 1060,
1065
+ "앓": 1061,
1066
+ "암": 1062,
1067
+ "압": 1063,
1068
+ "앗": 1064,
1069
+ "았": 1065,
1070
+ "앙": 1066,
1071
+ "앞": 1067,
1072
+ "애": 1068,
1073
+ "액": 1069,
1074
+ "앤": 1070,
1075
+ "앨": 1071,
1076
+ "앰": 1072,
1077
+ "앱": 1073,
1078
+ "앳": 1074,
1079
+ "앴": 1075,
1080
+ "앵": 1076,
1081
+ "야": 1077,
1082
+ "약": 1078,
1083
+ "얀": 1079,
1084
+ "얄": 1080,
1085
+ "얇": 1081,
1086
+ "얌": 1082,
1087
+ "얍": 1083,
1088
+ "양": 1084,
1089
+ "얕": 1085,
1090
+ "얗": 1086,
1091
+ "얘": 1087,
1092
+ "얜": 1088,
1093
+ "어": 1089,
1094
+ "억": 1090,
1095
+ "언": 1091,
1096
+ "얹": 1092,
1097
+ "얻": 1093,
1098
+ "얼": 1094,
1099
+ "얽": 1095,
1100
+ "엄": 1096,
1101
+ "업": 1097,
1102
+ "없": 1098,
1103
+ "엇": 1099,
1104
+ "었": 1100,
1105
+ "엉": 1101,
1106
+ "엊": 1102,
1107
+ "엌": 1103,
1108
+ "엎": 1104,
1109
+ "에": 1105,
1110
+ "엑": 1106,
1111
+ "엔": 1107,
1112
+ "엘": 1108,
1113
+ "엠": 1109,
1114
+ "엣": 1110,
1115
+ "엥": 1111,
1116
+ "여": 1112,
1117
+ "역": 1113,
1118
+ "엮": 1114,
1119
+ "연": 1115,
1120
+ "열": 1116,
1121
+ "염": 1117,
1122
+ "엽": 1118,
1123
+ "엿": 1119,
1124
+ "였": 1120,
1125
+ "영": 1121,
1126
+ "옅": 1122,
1127
+ "옆": 1123,
1128
+ "옇": 1124,
1129
+ "예": 1125,
1130
+ "옌": 1126,
1131
+ "옐": 1127,
1132
+ "옛": 1128,
1133
+ "오": 1129,
1134
+ "옥": 1130,
1135
+ "온": 1131,
1136
+ "올": 1132,
1137
+ "옮": 1133,
1138
+ "옳": 1134,
1139
+ "옴": 1135,
1140
+ "옵": 1136,
1141
+ "옷": 1137,
1142
+ "옹": 1138,
1143
+ "옻": 1139,
1144
+ "와": 1140,
1145
+ "왁": 1141,
1146
+ "완": 1142,
1147
+ "왈": 1143,
1148
+ "왓": 1144,
1149
+ "왔": 1145,
1150
+ "왕": 1146,
1151
+ "왜": 1147,
1152
+ "왠": 1148,
1153
+ "왤": 1149,
1154
+ "외": 1150,
1155
+ "왼": 1151,
1156
+ "욀": 1152,
1157
+ "요": 1153,
1158
+ "욕": 1154,
1159
+ "욘": 1155,
1160
+ "욜": 1156,
1161
+ "욤": 1157,
1162
+ "욥": 1158,
1163
+ "용": 1159,
1164
+ "우": 1160,
1165
+ "욱": 1161,
1166
+ "운": 1162,
1167
+ "울": 1163,
1168
+ "움": 1164,
1169
+ "웁": 1165,
1170
+ "웃": 1166,
1171
+ "웅": 1167,
1172
+ "워": 1168,
1173
+ "웍": 1169,
1174
+ "원": 1170,
1175
+ "월": 1171,
1176
+ "웜": 1172,
1177
+ "웠": 1173,
1178
+ "웨": 1174,
1179
+ "웩": 1175,
1180
+ "웬": 1176,
1181
+ "웰": 1177,
1182
+ "웹": 1178,
1183
+ "웻": 1179,
1184
+ "웽": 1180,
1185
+ "위": 1181,
1186
+ "윅": 1182,
1187
+ "윈": 1183,
1188
+ "윌": 1184,
1189
+ "윔": 1185,
1190
+ "윗": 1186,
1191
+ "윙": 1187,
1192
+ "유": 1188,
1193
+ "육": 1189,
1194
+ "윤": 1190,
1195
+ "율": 1191,
1196
+ "윱": 1192,
1197
+ "윳": 1193,
1198
+ "융": 1194,
1199
+ "윶": 1195,
1200
+ "윷": 1196,
1201
+ "으": 1197,
1202
+ "윽": 1198,
1203
+ "은": 1199,
1204
+ "읃": 1200,
1205
+ "을": 1201,
1206
+ "읊": 1202,
1207
+ "음": 1203,
1208
+ "읍": 1204,
1209
+ "읏": 1205,
1210
+ "응": 1206,
1211
+ "읒": 1207,
1212
+ "의": 1208,
1213
+ "이": 1209,
1214
+ "익": 1210,
1215
+ "인": 1211,
1216
+ "일": 1212,
1217
+ "읽": 1213,
1218
+ "잃": 1214,
1219
+ "임": 1215,
1220
+ "입": 1216,
1221
+ "잇": 1217,
1222
+ "있": 1218,
1223
+ "잉": 1219,
1224
+ "잊": 1220,
1225
+ "잌": 1221,
1226
+ "잎": 1222,
1227
+ "자": 1223,
1228
+ "작": 1224,
1229
+ "잔": 1225,
1230
+ "잖": 1226,
1231
+ "잘": 1227,
1232
+ "잠": 1228,
1233
+ "잡": 1229,
1234
+ "잣": 1230,
1235
+ "잤": 1231,
1236
+ "장": 1232,
1237
+ "잦": 1233,
1238
+ "재": 1234,
1239
+ "잭": 1235,
1240
+ "잰": 1236,
1241
+ "잴": 1237,
1242
+ "잼": 1238,
1243
+ "잽": 1239,
1244
+ "잿": 1240,
1245
+ "쟀": 1241,
1246
+ "쟁": 1242,
1247
+ "쟝": 1243,
1248
+ "쟤": 1244,
1249
+ "쟨": 1245,
1250
+ "저": 1246,
1251
+ "적": 1247,
1252
+ "전": 1248,
1253
+ "절": 1249,
1254
+ "젊": 1250,
1255
+ "점": 1251,
1256
+ "접": 1252,
1257
+ "젓": 1253,
1258
+ "정": 1254,
1259
+ "젖": 1255,
1260
+ "제": 1256,
1261
+ "젝": 1257,
1262
+ "젠": 1258,
1263
+ "젤": 1259,
1264
+ "젬": 1260,
1265
+ "젯": 1261,
1266
+ "져": 1262,
1267
+ "젼": 1263,
1268
+ "졌": 1264,
1269
+ "조": 1265,
1270
+ "족": 1266,
1271
+ "존": 1267,
1272
+ "졸": 1268,
1273
+ "좀": 1269,
1274
+ "좁": 1270,
1275
+ "좃": 1271,
1276
+ "종": 1272,
1277
+ "좆": 1273,
1278
+ "좋": 1274,
1279
+ "좌": 1275,
1280
+ "좍": 1276,
1281
+ "좔": 1277,
1282
+ "죄": 1278,
1283
+ "죠": 1279,
1284
+ "주": 1280,
1285
+ "죽": 1281,
1286
+ "준": 1282,
1287
+ "줄": 1283,
1288
+ "줌": 1284,
1289
+ "줍": 1285,
1290
+ "줏": 1286,
1291
+ "중": 1287,
1292
+ "줘": 1288,
1293
+ "줬": 1289,
1294
+ "줸": 1290,
1295
+ "쥐": 1291,
1296
+ "쥔": 1292,
1297
+ "쥘": 1293,
1298
+ "쥬": 1294,
1299
+ "즈": 1295,
1300
+ "즉": 1296,
1301
+ "즌": 1297,
1302
+ "즐": 1298,
1303
+ "즘": 1299,
1304
+ "즙": 1300,
1305
+ "증": 1301,
1306
+ "지": 1302,
1307
+ "직": 1303,
1308
+ "진": 1304,
1309
+ "질": 1305,
1310
+ "짊": 1306,
1311
+ "짐": 1307,
1312
+ "집": 1308,
1313
+ "짓": 1309,
1314
+ "징": 1310,
1315
+ "짖": 1311,
1316
+ "짚": 1312,
1317
+ "짜": 1313,
1318
+ "짝": 1314,
1319
+ "짠": 1315,
1320
+ "짤": 1316,
1321
+ "짥": 1317,
1322
+ "짧": 1318,
1323
+ "짬": 1319,
1324
+ "짭": 1320,
1325
+ "짯": 1321,
1326
+ "짰": 1322,
1327
+ "짱": 1323,
1328
+ "째": 1324,
1329
+ "짹": 1325,
1330
+ "짼": 1326,
1331
+ "쨋": 1327,
1332
+ "쨌": 1328,
1333
+ "쨍": 1329,
1334
+ "쨔": 1330,
1335
+ "쨰": 1331,
1336
+ "쩃": 1332,
1337
+ "쩄": 1333,
1338
+ "쩌": 1334,
1339
+ "쩍": 1335,
1340
+ "쩐": 1336,
1341
+ "쩔": 1337,
1342
+ "쩜": 1338,
1343
+ "쩝": 1339,
1344
+ "쩠": 1340,
1345
+ "쩡": 1341,
1346
+ "쩨": 1342,
1347
+ "쪄": 1343,
1348
+ "쪘": 1344,
1349
+ "쪼": 1345,
1350
+ "쪽": 1346,
1351
+ "쫀": 1347,
1352
+ "쫄": 1348,
1353
+ "쫌": 1349,
1354
+ "쫍": 1350,
1355
+ "쫑": 1351,
1356
+ "쫒": 1352,
1357
+ "쫓": 1353,
1358
+ "쫘": 1354,
1359
+ "쫙": 1355,
1360
+ "쬐": 1356,
1361
+ "쭈": 1357,
1362
+ "쭉": 1358,
1363
+ "쭌": 1359,
1364
+ "쭐": 1360,
1365
+ "쭘": 1361,
1366
+ "쭙": 1362,
1367
+ "쭝": 1363,
1368
+ "쭤": 1364,
1369
+ "쮸": 1365,
1370
+ "쯔": 1366,
1371
+ "쯕": 1367,
1372
+ "쯤": 1368,
1373
+ "쯧": 1369,
1374
+ "찌": 1370,
1375
+ "찍": 1371,
1376
+ "찐": 1372,
1377
+ "찔": 1373,
1378
+ "찜": 1374,
1379
+ "찝": 1375,
1380
+ "찡": 1376,
1381
+ "찢": 1377,
1382
+ "차": 1378,
1383
+ "착": 1379,
1384
+ "찬": 1380,
1385
+ "찮": 1381,
1386
+ "찰": 1382,
1387
+ "참": 1383,
1388
+ "찹": 1384,
1389
+ "찻": 1385,
1390
+ "찼": 1386,
1391
+ "창": 1387,
1392
+ "찾": 1388,
1393
+ "채": 1389,
1394
+ "책": 1390,
1395
+ "챌": 1391,
1396
+ "챔": 1392,
1397
+ "챕": 1393,
1398
+ "챗": 1394,
1399
+ "챘": 1395,
1400
+ "챙": 1396,
1401
+ "챠": 1397,
1402
+ "챴": 1398,
1403
+ "처": 1399,
1404
+ "척": 1400,
1405
+ "천": 1401,
1406
+ "철": 1402,
1407
+ "첨": 1403,
1408
+ "첩": 1404,
1409
+ "첫": 1405,
1410
+ "청": 1406,
1411
+ "체": 1407,
1412
+ "첵": 1408,
1413
+ "첸": 1409,
1414
+ "첼": 1410,
1415
+ "쳅": 1411,
1416
+ "쳇": 1412,
1417
+ "쳐": 1413,
1418
+ "쳤": 1414,
1419
+ "쳬": 1415,
1420
+ "초": 1416,
1421
+ "촉": 1417,
1422
+ "촌": 1418,
1423
+ "촐": 1419,
1424
+ "촘": 1420,
1425
+ "촛": 1421,
1426
+ "총": 1422,
1427
+ "촤": 1423,
1428
+ "촥": 1424,
1429
+ "촬": 1425,
1430
+ "최": 1426,
1431
+ "쵸": 1427,
1432
+ "추": 1428,
1433
+ "축": 1429,
1434
+ "춘": 1430,
1435
+ "출": 1431,
1436
+ "춤": 1432,
1437
+ "춥": 1433,
1438
+ "춧": 1434,
1439
+ "충": 1435,
1440
+ "춰": 1436,
1441
+ "춱": 1437,
1442
+ "췄": 1438,
1443
+ "췌": 1439,
1444
+ "취": 1440,
1445
+ "췬": 1441,
1446
+ "츄": 1442,
1447
+ "츈": 1443,
1448
+ "츠": 1444,
1449
+ "측": 1445,
1450
+ "츰": 1446,
1451
+ "층": 1447,
1452
+ "치": 1448,
1453
+ "칙": 1449,
1454
+ "친": 1450,
1455
+ "칠": 1451,
1456
+ "칡": 1452,
1457
+ "침": 1453,
1458
+ "칩": 1454,
1459
+ "칫": 1455,
1460
+ "칭": 1456,
1461
+ "카": 1457,
1462
+ "칵": 1458,
1463
+ "칸": 1459,
1464
+ "칼": 1460,
1465
+ "캄": 1461,
1466
+ "캅": 1462,
1467
+ "캇": 1463,
1468
+ "캉": 1464,
1469
+ "캐": 1465,
1470
+ "캔": 1466,
1471
+ "캘": 1467,
1472
+ "캠": 1468,
1473
+ "캡": 1469,
1474
+ "캣": 1470,
1475
+ "캤": 1471,
1476
+ "캥": 1472,
1477
+ "캬": 1473,
1478
+ "커": 1474,
1479
+ "컥": 1475,
1480
+ "컨": 1476,
1481
+ "컬": 1477,
1482
+ "컴": 1478,
1483
+ "컵": 1479,
1484
+ "컷": 1480,
1485
+ "컸": 1481,
1486
+ "케": 1482,
1487
+ "켁": 1483,
1488
+ "켄": 1484,
1489
+ "켈": 1485,
1490
+ "켓": 1486,
1491
+ "켜": 1487,
1492
+ "켠": 1488,
1493
+ "켤": 1489,
1494
+ "켰": 1490,
1495
+ "코": 1491,
1496
+ "콕": 1492,
1497
+ "콘": 1493,
1498
+ "콜": 1494,
1499
+ "콤": 1495,
1500
+ "콥": 1496,
1501
+ "콧": 1497,
1502
+ "콩": 1498,
1503
+ "콱": 1499,
1504
+ "콸": 1500,
1505
+ "쾅": 1501,
1506
+ "쾌": 1502,
1507
+ "쾨": 1503,
1508
+ "쿄": 1504,
1509
+ "쿠": 1505,
1510
+ "쿡": 1506,
1511
+ "쿤": 1507,
1512
+ "쿨": 1508,
1513
+ "쿰": 1509,
1514
+ "쿱": 1510,
1515
+ "쿵": 1511,
1516
+ "쿼": 1512,
1517
+ "퀀": 1513,
1518
+ "퀄": 1514,
1519
+ "퀘": 1515,
1520
+ "퀭": 1516,
1521
+ "퀴": 1517,
1522
+ "퀵": 1518,
1523
+ "퀸": 1519,
1524
+ "퀼": 1520,
1525
+ "큐": 1521,
1526
+ "큘": 1522,
1527
+ "크": 1523,
1528
+ "큰": 1524,
1529
+ "클": 1525,
1530
+ "큼": 1526,
1531
+ "큽": 1527,
1532
+ "킁": 1528,
1533
+ "키": 1529,
1534
+ "킥": 1530,
1535
+ "킨": 1531,
1536
+ "킬": 1532,
1537
+ "킴": 1533,
1538
+ "킵": 1534,
1539
+ "킷": 1535,
1540
+ "킹": 1536,
1541
+ "타": 1537,
1542
+ "탁": 1538,
1543
+ "탄": 1539,
1544
+ "탈": 1540,
1545
+ "탐": 1541,
1546
+ "탑": 1542,
1547
+ "탓": 1543,
1548
+ "탔": 1544,
1549
+ "탕": 1545,
1550
+ "태": 1546,
1551
+ "택": 1547,
1552
+ "탠": 1548,
1553
+ "탤": 1549,
1554
+ "탬": 1550,
1555
+ "탭": 1551,
1556
+ "탯": 1552,
1557
+ "탰": 1553,
1558
+ "탱": 1554,
1559
+ "터": 1555,
1560
+ "턱": 1556,
1561
+ "턴": 1557,
1562
+ "털": 1558,
1563
+ "텀": 1559,
1564
+ "텁": 1560,
1565
+ "텃": 1561,
1566
+ "텅": 1562,
1567
+ "테": 1563,
1568
+ "텍": 1564,
1569
+ "텐": 1565,
1570
+ "텔": 1566,
1571
+ "템": 1567,
1572
+ "텝": 1568,
1573
+ "텟": 1569,
1574
+ "텡": 1570,
1575
+ "텨": 1571,
1576
+ "텼": 1572,
1577
+ "토": 1573,
1578
+ "톡": 1574,
1579
+ "톤": 1575,
1580
+ "톨": 1576,
1581
+ "톰": 1577,
1582
+ "톱": 1578,
1583
+ "톳": 1579,
1584
+ "통": 1580,
1585
+ "퇴": 1581,
1586
+ "투": 1582,
1587
+ "툭": 1583,
1588
+ "툰": 1584,
1589
+ "툴": 1585,
1590
+ "툼": 1586,
1591
+ "퉁": 1587,
1592
+ "튀": 1588,
1593
+ "튄": 1589,
1594
+ "튕": 1590,
1595
+ "튜": 1591,
1596
+ "튠": 1592,
1597
+ "튤": 1593,
1598
+ "튱": 1594,
1599
+ "트": 1595,
1600
+ "특": 1596,
1601
+ "튼": 1597,
1602
+ "튿": 1598,
1603
+ "틀": 1599,
1604
+ "틈": 1600,
1605
+ "틋": 1601,
1606
+ "틍": 1602,
1607
+ "티": 1603,
1608
+ "틱": 1604,
1609
+ "틴": 1605,
1610
+ "틸": 1606,
1611
+ "팀": 1607,
1612
+ "팁": 1608,
1613
+ "팅": 1609,
1614
+ "파": 1610,
1615
+ "팍": 1611,
1616
+ "팎": 1612,
1617
+ "판": 1613,
1618
+ "팔": 1614,
1619
+ "팜": 1615,
1620
+ "팝": 1616,
1621
+ "팟": 1617,
1622
+ "팠": 1618,
1623
+ "팡": 1619,
1624
+ "팥": 1620,
1625
+ "패": 1621,
1626
+ "팩": 1622,
1627
+ "팬": 1623,
1628
+ "팰": 1624,
1629
+ "팸": 1625,
1630
+ "팻": 1626,
1631
+ "팼": 1627,
1632
+ "팽": 1628,
1633
+ "퍼": 1629,
1634
+ "퍽": 1630,
1635
+ "펀": 1631,
1636
+ "펄": 1632,
1637
+ "펌": 1633,
1638
+ "펍": 1634,
1639
+ "펐": 1635,
1640
+ "펑": 1636,
1641
+ "페": 1637,
1642
+ "펙": 1638,
1643
+ "펜": 1639,
1644
+ "펠": 1640,
1645
+ "펩": 1641,
1646
+ "펫": 1642,
1647
+ "펭": 1643,
1648
+ "펴": 1644,
1649
+ "편": 1645,
1650
+ "펼": 1646,
1651
+ "폄": 1647,
1652
+ "폈": 1648,
1653
+ "평": 1649,
1654
+ "폐": 1650,
1655
+ "포": 1651,
1656
+ "폭": 1652,
1657
+ "폰": 1653,
1658
+ "폴": 1654,
1659
+ "폼": 1655,
1660
+ "폿": 1656,
1661
+ "퐁": 1657,
1662
+ "표": 1658,
1663
+ "푯": 1659,
1664
+ "푸": 1660,
1665
+ "푹": 1661,
1666
+ "푼": 1662,
1667
+ "풀": 1663,
1668
+ "품": 1664,
1669
+ "풋": 1665,
1670
+ "풍": 1666,
1671
+ "퓨": 1667,
1672
+ "퓰": 1668,
1673
+ "프": 1669,
1674
+ "픈": 1670,
1675
+ "플": 1671,
1676
+ "픔": 1672,
1677
+ "픕": 1673,
1678
+ "픗": 1674,
1679
+ "피": 1675,
1680
+ "픽": 1676,
1681
+ "핀": 1677,
1682
+ "필": 1678,
1683
+ "핌": 1679,
1684
+ "핍": 1680,
1685
+ "핏": 1681,
1686
+ "핑": 1682,
1687
+ "하": 1683,
1688
+ "학": 1684,
1689
+ "한": 1685,
1690
+ "할": 1686,
1691
+ "핡": 1687,
1692
+ "핥": 1688,
1693
+ "핧": 1689,
1694
+ "함": 1690,
1695
+ "합": 1691,
1696
+ "핫": 1692,
1697
+ "핬": 1693,
1698
+ "항": 1694,
1699
+ "해": 1695,
1700
+ "핵": 1696,
1701
+ "핸": 1697,
1702
+ "핼": 1698,
1703
+ "햄": 1699,
1704
+ "햅": 1700,
1705
+ "햇": 1701,
1706
+ "했": 1702,
1707
+ "행": 1703,
1708
+ "햍": 1704,
1709
+ "햐": 1705,
1710
+ "햝": 1706,
1711
+ "향": 1707,
1712
+ "헀": 1708,
1713
+ "허": 1709,
1714
+ "헉": 1710,
1715
+ "헌": 1711,
1716
+ "헐": 1712,
1717
+ "험": 1713,
1718
+ "헛": 1714,
1719
+ "헝": 1715,
1720
+ "헤": 1716,
1721
+ "헥": 1717,
1722
+ "헨": 1718,
1723
+ "헬": 1719,
1724
+ "헷": 1720,
1725
+ "헹": 1721,
1726
+ "혀": 1722,
1727
+ "혁": 1723,
1728
+ "현": 1724,
1729
+ "혈": 1725,
1730
+ "혐": 1726,
1731
+ "협": 1727,
1732
+ "혓": 1728,
1733
+ "혔": 1729,
1734
+ "형": 1730,
1735
+ "혜": 1731,
1736
+ "호": 1732,
1737
+ "혹": 1733,
1738
+ "혼": 1734,
1739
+ "홀": 1735,
1740
+ "홈": 1736,
1741
+ "홉": 1737,
1742
+ "홋": 1738,
1743
+ "홍": 1739,
1744
+ "홑": 1740,
1745
+ "화": 1741,
1746
+ "확": 1742,
1747
+ "환": 1743,
1748
+ "활": 1744,
1749
+ "홧": 1745,
1750
+ "황": 1746,
1751
+ "홰": 1747,
1752
+ "회": 1748,
1753
+ "획": 1749,
1754
+ "횔": 1750,
1755
+ "횟": 1751,
1756
+ "횡": 1752,
1757
+ "효": 1753,
1758
+ "후": 1754,
1759
+ "훅": 1755,
1760
+ "훈": 1756,
1761
+ "훌": 1757,
1762
+ "훑": 1758,
1763
+ "훔": 1759,
1764
+ "훗": 1760,
1765
+ "훙": 1761,
1766
+ "훡": 1762,
1767
+ "훨": 1763,
1768
+ "훼": 1764,
1769
+ "휘": 1765,
1770
+ "휙": 1766,
1771
+ "휠": 1767,
1772
+ "휩": 1768,
1773
+ "휴": 1769,
1774
+ "흉": 1770,
1775
+ "흐": 1771,
1776
+ "흑": 1772,
1777
+ "흔": 1773,
1778
+ "흘": 1774,
1779
+ "흙": 1775,
1780
+ "흠": 1776,
1781
+ "흡": 1777,
1782
+ "흣": 1778,
1783
+ "흥": 1779,
1784
+ "흩": 1780,
1785
+ "희": 1781,
1786
+ "흰": 1782,
1787
+ "히": 1783,
1788
+ "힌": 1784,
1789
+ "힐": 1785,
1790
+ "힘": 1786,
1791
+ "힙": 1787
1792
+ }
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": 0.1,
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": false,
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": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.0,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_size": 1024,
60
+ "initializer_range": 0.02,
61
+ "intermediate_size": 4096,
62
+ "layer_norm_eps": 1e-05,
63
+ "layerdrop": 0.1,
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": 0.1,
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": 1789,
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": "4.34.1",
114
+ "use_weighted_layer_sum": false,
115
+ "vocab_size": 1792,
116
+ "xvector_output_dim": 512
117
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 29.99,
3
+ "eval_cer": 0.11131710849470199,
4
+ "eval_loss": 0.4239382743835449,
5
+ "eval_runtime": 2675.296,
6
+ "eval_samples": 16914,
7
+ "eval_samples_per_second": 6.322,
8
+ "eval_steps_per_second": 0.395,
9
+ "eval_wer": 0.30376043574435696
10
+ }
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
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f02144df03af73dedf4c1427aa768bfd0718e2a4d7f0a65c9d69477a6bd62815
3
+ size 1269249578
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
+ "1788": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "1789": {
12
+ "content": "[PAD]",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "1790": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": true,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "1791": {
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": 29.99,
3
+ "train_loss": 0.9622837933354408,
4
+ "train_runtime": 156348.4692,
5
+ "train_samples": 135710,
6
+ "train_samples_per_second": 26.04,
7
+ "train_steps_per_second": 0.407
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1100 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 29.992926196651734,
5
+ "eval_steps": 2000,
6
+ "global_step": 63600,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.24,
13
+ "learning_rate": 0.00014849999999999998,
14
+ "loss": 15.6695,
15
+ "step": 500
16
+ },
17
+ {
18
+ "epoch": 0.47,
19
+ "learning_rate": 0.0002982,
20
+ "loss": 4.4987,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.71,
25
+ "learning_rate": 0.00029763258785942486,
26
+ "loss": 2.3147,
27
+ "step": 1500
28
+ },
29
+ {
30
+ "epoch": 0.94,
31
+ "learning_rate": 0.0002952364217252396,
32
+ "loss": 1.7721,
33
+ "step": 2000
34
+ },
35
+ {
36
+ "epoch": 0.94,
37
+ "eval_cer": 0.29026945911872865,
38
+ "eval_loss": 1.1368173360824585,
39
+ "eval_runtime": 2718.6813,
40
+ "eval_samples_per_second": 6.221,
41
+ "eval_steps_per_second": 0.389,
42
+ "eval_wer": 0.6589364856615053,
43
+ "step": 2000
44
+ },
45
+ {
46
+ "epoch": 1.18,
47
+ "learning_rate": 0.0002928402555910543,
48
+ "loss": 1.5735,
49
+ "step": 2500
50
+ },
51
+ {
52
+ "epoch": 1.41,
53
+ "learning_rate": 0.000290444089456869,
54
+ "loss": 1.4611,
55
+ "step": 3000
56
+ },
57
+ {
58
+ "epoch": 1.65,
59
+ "learning_rate": 0.00028805271565495203,
60
+ "loss": 1.4056,
61
+ "step": 3500
62
+ },
63
+ {
64
+ "epoch": 1.89,
65
+ "learning_rate": 0.00028565654952076674,
66
+ "loss": 1.3501,
67
+ "step": 4000
68
+ },
69
+ {
70
+ "epoch": 1.89,
71
+ "eval_cer": 0.2240432390931656,
72
+ "eval_loss": 0.8561204671859741,
73
+ "eval_runtime": 2698.4922,
74
+ "eval_samples_per_second": 6.268,
75
+ "eval_steps_per_second": 0.392,
76
+ "eval_wer": 0.5451323644839807,
77
+ "step": 4000
78
+ },
79
+ {
80
+ "epoch": 2.12,
81
+ "learning_rate": 0.00028326038338658146,
82
+ "loss": 1.2967,
83
+ "step": 4500
84
+ },
85
+ {
86
+ "epoch": 2.36,
87
+ "learning_rate": 0.00028086421725239617,
88
+ "loss": 1.244,
89
+ "step": 5000
90
+ },
91
+ {
92
+ "epoch": 2.59,
93
+ "learning_rate": 0.0002784728434504792,
94
+ "loss": 1.2361,
95
+ "step": 5500
96
+ },
97
+ {
98
+ "epoch": 2.83,
99
+ "learning_rate": 0.0002760766773162939,
100
+ "loss": 1.2133,
101
+ "step": 6000
102
+ },
103
+ {
104
+ "epoch": 2.83,
105
+ "eval_cer": 0.2002927101645895,
106
+ "eval_loss": 0.7505359053611755,
107
+ "eval_runtime": 2721.6856,
108
+ "eval_samples_per_second": 6.215,
109
+ "eval_steps_per_second": 0.389,
110
+ "eval_wer": 0.49743516558444134,
111
+ "step": 6000
112
+ },
113
+ {
114
+ "epoch": 3.07,
115
+ "learning_rate": 0.0002736805111821086,
116
+ "loss": 1.1755,
117
+ "step": 6500
118
+ },
119
+ {
120
+ "epoch": 3.3,
121
+ "learning_rate": 0.0002712843450479233,
122
+ "loss": 1.1403,
123
+ "step": 7000
124
+ },
125
+ {
126
+ "epoch": 3.54,
127
+ "learning_rate": 0.000268888178913738,
128
+ "loss": 1.1253,
129
+ "step": 7500
130
+ },
131
+ {
132
+ "epoch": 3.77,
133
+ "learning_rate": 0.00026649680511182103,
134
+ "loss": 1.0981,
135
+ "step": 8000
136
+ },
137
+ {
138
+ "epoch": 3.77,
139
+ "eval_cer": 0.18418492651709803,
140
+ "eval_loss": 0.6767656803131104,
141
+ "eval_runtime": 2740.3957,
142
+ "eval_samples_per_second": 6.172,
143
+ "eval_steps_per_second": 0.386,
144
+ "eval_wer": 0.46858375155594667,
145
+ "step": 8000
146
+ },
147
+ {
148
+ "epoch": 4.01,
149
+ "learning_rate": 0.00026410063897763575,
150
+ "loss": 1.0984,
151
+ "step": 8500
152
+ },
153
+ {
154
+ "epoch": 4.24,
155
+ "learning_rate": 0.00026170447284345046,
156
+ "loss": 1.0614,
157
+ "step": 9000
158
+ },
159
+ {
160
+ "epoch": 4.48,
161
+ "learning_rate": 0.00025930830670926517,
162
+ "loss": 1.0553,
163
+ "step": 9500
164
+ },
165
+ {
166
+ "epoch": 4.72,
167
+ "learning_rate": 0.0002569169329073482,
168
+ "loss": 1.0375,
169
+ "step": 10000
170
+ },
171
+ {
172
+ "epoch": 4.72,
173
+ "eval_cer": 0.17071197058066542,
174
+ "eval_loss": 0.64134681224823,
175
+ "eval_runtime": 2703.0312,
176
+ "eval_samples_per_second": 6.257,
177
+ "eval_steps_per_second": 0.391,
178
+ "eval_wer": 0.44041417913406117,
179
+ "step": 10000
180
+ },
181
+ {
182
+ "epoch": 4.95,
183
+ "learning_rate": 0.0002545207667731629,
184
+ "loss": 1.0351,
185
+ "step": 10500
186
+ },
187
+ {
188
+ "epoch": 5.19,
189
+ "learning_rate": 0.00025212460063897763,
190
+ "loss": 1.0087,
191
+ "step": 11000
192
+ },
193
+ {
194
+ "epoch": 5.42,
195
+ "learning_rate": 0.00024973322683706067,
196
+ "loss": 0.9934,
197
+ "step": 11500
198
+ },
199
+ {
200
+ "epoch": 5.66,
201
+ "learning_rate": 0.0002473370607028754,
202
+ "loss": 0.9927,
203
+ "step": 12000
204
+ },
205
+ {
206
+ "epoch": 5.66,
207
+ "eval_cer": 0.16338767301962598,
208
+ "eval_loss": 0.6106029152870178,
209
+ "eval_runtime": 2826.5991,
210
+ "eval_samples_per_second": 5.984,
211
+ "eval_steps_per_second": 0.374,
212
+ "eval_wer": 0.42461289632043386,
213
+ "step": 12000
214
+ },
215
+ {
216
+ "epoch": 5.89,
217
+ "learning_rate": 0.0002449408945686901,
218
+ "loss": 0.9821,
219
+ "step": 12500
220
+ },
221
+ {
222
+ "epoch": 6.13,
223
+ "learning_rate": 0.0002425447284345048,
224
+ "loss": 0.9637,
225
+ "step": 13000
226
+ },
227
+ {
228
+ "epoch": 6.37,
229
+ "learning_rate": 0.00024014856230031946,
230
+ "loss": 0.9497,
231
+ "step": 13500
232
+ },
233
+ {
234
+ "epoch": 6.6,
235
+ "learning_rate": 0.00023775239616613414,
236
+ "loss": 0.9439,
237
+ "step": 14000
238
+ },
239
+ {
240
+ "epoch": 6.6,
241
+ "eval_cer": 0.1613024947979602,
242
+ "eval_loss": 0.5999171733856201,
243
+ "eval_runtime": 2741.759,
244
+ "eval_samples_per_second": 6.169,
245
+ "eval_steps_per_second": 0.386,
246
+ "eval_wer": 0.41588373807768236,
247
+ "step": 14000
248
+ },
249
+ {
250
+ "epoch": 6.84,
251
+ "learning_rate": 0.00023536102236421723,
252
+ "loss": 0.948,
253
+ "step": 14500
254
+ },
255
+ {
256
+ "epoch": 7.07,
257
+ "learning_rate": 0.0002329696485623003,
258
+ "loss": 0.9367,
259
+ "step": 15000
260
+ },
261
+ {
262
+ "epoch": 7.31,
263
+ "learning_rate": 0.000230573482428115,
264
+ "loss": 0.9087,
265
+ "step": 15500
266
+ },
267
+ {
268
+ "epoch": 7.55,
269
+ "learning_rate": 0.0002281773162939297,
270
+ "loss": 0.9059,
271
+ "step": 16000
272
+ },
273
+ {
274
+ "epoch": 7.55,
275
+ "eval_cer": 0.15351797484699242,
276
+ "eval_loss": 0.5740103721618652,
277
+ "eval_runtime": 2702.3695,
278
+ "eval_samples_per_second": 6.259,
279
+ "eval_steps_per_second": 0.392,
280
+ "eval_wer": 0.39852056228147376,
281
+ "step": 16000
282
+ },
283
+ {
284
+ "epoch": 7.78,
285
+ "learning_rate": 0.00022578115015974438,
286
+ "loss": 0.9075,
287
+ "step": 16500
288
+ },
289
+ {
290
+ "epoch": 8.02,
291
+ "learning_rate": 0.0002233849840255591,
292
+ "loss": 0.8999,
293
+ "step": 17000
294
+ },
295
+ {
296
+ "epoch": 8.25,
297
+ "learning_rate": 0.00022098881789137377,
298
+ "loss": 0.8597,
299
+ "step": 17500
300
+ },
301
+ {
302
+ "epoch": 8.49,
303
+ "learning_rate": 0.00021859265175718849,
304
+ "loss": 0.8772,
305
+ "step": 18000
306
+ },
307
+ {
308
+ "epoch": 8.49,
309
+ "eval_cer": 0.14781427082015555,
310
+ "eval_loss": 0.5568912625312805,
311
+ "eval_runtime": 2727.4118,
312
+ "eval_samples_per_second": 6.201,
313
+ "eval_steps_per_second": 0.388,
314
+ "eval_wer": 0.3954364182701837,
315
+ "step": 18000
316
+ },
317
+ {
318
+ "epoch": 8.72,
319
+ "learning_rate": 0.00021619648562300317,
320
+ "loss": 0.8785,
321
+ "step": 18500
322
+ },
323
+ {
324
+ "epoch": 8.96,
325
+ "learning_rate": 0.00021380511182108623,
326
+ "loss": 0.8702,
327
+ "step": 19000
328
+ },
329
+ {
330
+ "epoch": 9.2,
331
+ "learning_rate": 0.00021140894568690095,
332
+ "loss": 0.8447,
333
+ "step": 19500
334
+ },
335
+ {
336
+ "epoch": 9.43,
337
+ "learning_rate": 0.00020901277955271563,
338
+ "loss": 0.8483,
339
+ "step": 20000
340
+ },
341
+ {
342
+ "epoch": 9.43,
343
+ "eval_cer": 0.14274528108464166,
344
+ "eval_loss": 0.5406663417816162,
345
+ "eval_runtime": 2824.1098,
346
+ "eval_samples_per_second": 5.989,
347
+ "eval_steps_per_second": 0.375,
348
+ "eval_wer": 0.3784141632772796,
349
+ "step": 20000
350
+ },
351
+ {
352
+ "epoch": 9.67,
353
+ "learning_rate": 0.00020661661341853031,
354
+ "loss": 0.8358,
355
+ "step": 20500
356
+ },
357
+ {
358
+ "epoch": 9.9,
359
+ "learning_rate": 0.00020422523961661338,
360
+ "loss": 0.8391,
361
+ "step": 21000
362
+ },
363
+ {
364
+ "epoch": 10.14,
365
+ "learning_rate": 0.00020182907348242812,
366
+ "loss": 0.8215,
367
+ "step": 21500
368
+ },
369
+ {
370
+ "epoch": 10.37,
371
+ "learning_rate": 0.00019943769968051118,
372
+ "loss": 0.81,
373
+ "step": 22000
374
+ },
375
+ {
376
+ "epoch": 10.37,
377
+ "eval_cer": 0.1415456492625536,
378
+ "eval_loss": 0.5282983779907227,
379
+ "eval_runtime": 2742.622,
380
+ "eval_samples_per_second": 6.167,
381
+ "eval_steps_per_second": 0.386,
382
+ "eval_wer": 0.37441032593614476,
383
+ "step": 22000
384
+ },
385
+ {
386
+ "epoch": 10.61,
387
+ "learning_rate": 0.00019704153354632587,
388
+ "loss": 0.8108,
389
+ "step": 22500
390
+ },
391
+ {
392
+ "epoch": 10.85,
393
+ "learning_rate": 0.00019464536741214058,
394
+ "loss": 0.8175,
395
+ "step": 23000
396
+ },
397
+ {
398
+ "epoch": 11.08,
399
+ "learning_rate": 0.00019224920127795526,
400
+ "loss": 0.8001,
401
+ "step": 23500
402
+ },
403
+ {
404
+ "epoch": 11.32,
405
+ "learning_rate": 0.00018985782747603833,
406
+ "loss": 0.793,
407
+ "step": 24000
408
+ },
409
+ {
410
+ "epoch": 11.32,
411
+ "eval_cer": 0.13662715879199255,
412
+ "eval_loss": 0.517921507358551,
413
+ "eval_runtime": 2729.8979,
414
+ "eval_samples_per_second": 6.196,
415
+ "eval_steps_per_second": 0.388,
416
+ "eval_wer": 0.36633129573690426,
417
+ "step": 24000
418
+ },
419
+ {
420
+ "epoch": 11.55,
421
+ "learning_rate": 0.000187461661341853,
422
+ "loss": 0.7827,
423
+ "step": 24500
424
+ },
425
+ {
426
+ "epoch": 11.79,
427
+ "learning_rate": 0.00018506549520766772,
428
+ "loss": 0.7899,
429
+ "step": 25000
430
+ },
431
+ {
432
+ "epoch": 12.03,
433
+ "learning_rate": 0.0001826693290734824,
434
+ "loss": 0.7806,
435
+ "step": 25500
436
+ },
437
+ {
438
+ "epoch": 12.26,
439
+ "learning_rate": 0.00018027316293929712,
440
+ "loss": 0.7577,
441
+ "step": 26000
442
+ },
443
+ {
444
+ "epoch": 12.26,
445
+ "eval_cer": 0.1359182854425769,
446
+ "eval_loss": 0.5058821439743042,
447
+ "eval_runtime": 2722.7634,
448
+ "eval_samples_per_second": 6.212,
449
+ "eval_steps_per_second": 0.389,
450
+ "eval_wer": 0.35946530932616605,
451
+ "step": 26000
452
+ },
453
+ {
454
+ "epoch": 12.5,
455
+ "learning_rate": 0.00017788178913738016,
456
+ "loss": 0.762,
457
+ "step": 26500
458
+ },
459
+ {
460
+ "epoch": 12.73,
461
+ "learning_rate": 0.00017548562300319487,
462
+ "loss": 0.7595,
463
+ "step": 27000
464
+ },
465
+ {
466
+ "epoch": 12.97,
467
+ "learning_rate": 0.00017308945686900955,
468
+ "loss": 0.7629,
469
+ "step": 27500
470
+ },
471
+ {
472
+ "epoch": 13.2,
473
+ "learning_rate": 0.00017069329073482426,
474
+ "loss": 0.7379,
475
+ "step": 28000
476
+ },
477
+ {
478
+ "epoch": 13.2,
479
+ "eval_cer": 0.13330526921919236,
480
+ "eval_loss": 0.4969228506088257,
481
+ "eval_runtime": 2824.8712,
482
+ "eval_samples_per_second": 5.988,
483
+ "eval_steps_per_second": 0.375,
484
+ "eval_wer": 0.35324945095893884,
485
+ "step": 28000
486
+ },
487
+ {
488
+ "epoch": 13.44,
489
+ "learning_rate": 0.00016830191693290736,
490
+ "loss": 0.737,
491
+ "step": 28500
492
+ },
493
+ {
494
+ "epoch": 13.68,
495
+ "learning_rate": 0.000165905750798722,
496
+ "loss": 0.7444,
497
+ "step": 29000
498
+ },
499
+ {
500
+ "epoch": 13.91,
501
+ "learning_rate": 0.00016350958466453675,
502
+ "loss": 0.7372,
503
+ "step": 29500
504
+ },
505
+ {
506
+ "epoch": 14.15,
507
+ "learning_rate": 0.00016111341853035144,
508
+ "loss": 0.7328,
509
+ "step": 30000
510
+ },
511
+ {
512
+ "epoch": 14.15,
513
+ "eval_cer": 0.13079476698787718,
514
+ "eval_loss": 0.4908413589000702,
515
+ "eval_runtime": 2825.8542,
516
+ "eval_samples_per_second": 5.985,
517
+ "eval_steps_per_second": 0.374,
518
+ "eval_wer": 0.3475251528197322,
519
+ "step": 30000
520
+ },
521
+ {
522
+ "epoch": 14.38,
523
+ "learning_rate": 0.0001587172523961661,
524
+ "loss": 0.7184,
525
+ "step": 30500
526
+ },
527
+ {
528
+ "epoch": 14.62,
529
+ "learning_rate": 0.00015632108626198083,
530
+ "loss": 0.7216,
531
+ "step": 31000
532
+ },
533
+ {
534
+ "epoch": 14.85,
535
+ "learning_rate": 0.0001539297124600639,
536
+ "loss": 0.7238,
537
+ "step": 31500
538
+ },
539
+ {
540
+ "epoch": 15.09,
541
+ "learning_rate": 0.00015153354632587858,
542
+ "loss": 0.7119,
543
+ "step": 32000
544
+ },
545
+ {
546
+ "epoch": 15.09,
547
+ "eval_cer": 0.12864851660072327,
548
+ "eval_loss": 0.4887321889400482,
549
+ "eval_runtime": 2731.0336,
550
+ "eval_samples_per_second": 6.193,
551
+ "eval_steps_per_second": 0.387,
552
+ "eval_wer": 0.34784228845071313,
553
+ "step": 32000
554
+ },
555
+ {
556
+ "epoch": 15.33,
557
+ "learning_rate": 0.00014914217252396165,
558
+ "loss": 0.7124,
559
+ "step": 32500
560
+ },
561
+ {
562
+ "epoch": 15.56,
563
+ "learning_rate": 0.00014674600638977636,
564
+ "loss": 0.7294,
565
+ "step": 33000
566
+ },
567
+ {
568
+ "epoch": 15.8,
569
+ "learning_rate": 0.00014434984025559104,
570
+ "loss": 0.7545,
571
+ "step": 33500
572
+ },
573
+ {
574
+ "epoch": 16.03,
575
+ "learning_rate": 0.00014195367412140575,
576
+ "loss": 0.7572,
577
+ "step": 34000
578
+ },
579
+ {
580
+ "epoch": 16.03,
581
+ "eval_cer": 0.13271854020075294,
582
+ "eval_loss": 0.5169993042945862,
583
+ "eval_runtime": 2729.0002,
584
+ "eval_samples_per_second": 6.198,
585
+ "eval_steps_per_second": 0.388,
586
+ "eval_wer": 0.3576893497926726,
587
+ "step": 34000
588
+ },
589
+ {
590
+ "epoch": 16.27,
591
+ "learning_rate": 0.00013956230031948882,
592
+ "loss": 0.7687,
593
+ "step": 34500
594
+ },
595
+ {
596
+ "epoch": 16.51,
597
+ "learning_rate": 0.0001371661341853035,
598
+ "loss": 0.7884,
599
+ "step": 35000
600
+ },
601
+ {
602
+ "epoch": 16.74,
603
+ "learning_rate": 0.00013476996805111819,
604
+ "loss": 0.8156,
605
+ "step": 35500
606
+ },
607
+ {
608
+ "epoch": 16.98,
609
+ "learning_rate": 0.0001323738019169329,
610
+ "loss": 0.8198,
611
+ "step": 36000
612
+ },
613
+ {
614
+ "epoch": 16.98,
615
+ "eval_cer": 0.1431662427967562,
616
+ "eval_loss": 0.5838645696640015,
617
+ "eval_runtime": 2730.4526,
618
+ "eval_samples_per_second": 6.195,
619
+ "eval_steps_per_second": 0.387,
620
+ "eval_wer": 0.38254485487080686,
621
+ "step": 36000
622
+ },
623
+ {
624
+ "epoch": 17.21,
625
+ "learning_rate": 0.00012997763578274758,
626
+ "loss": 0.819,
627
+ "step": 36500
628
+ },
629
+ {
630
+ "epoch": 17.45,
631
+ "learning_rate": 0.00012758626198083067,
632
+ "loss": 0.8411,
633
+ "step": 37000
634
+ },
635
+ {
636
+ "epoch": 17.68,
637
+ "learning_rate": 0.00012519009584664536,
638
+ "loss": 0.8366,
639
+ "step": 37500
640
+ },
641
+ {
642
+ "epoch": 17.92,
643
+ "learning_rate": 0.00012279392971246007,
644
+ "loss": 0.8008,
645
+ "step": 38000
646
+ },
647
+ {
648
+ "epoch": 17.92,
649
+ "eval_cer": 0.13762394377870937,
650
+ "eval_loss": 0.5447062253952026,
651
+ "eval_runtime": 2738.2931,
652
+ "eval_samples_per_second": 6.177,
653
+ "eval_steps_per_second": 0.386,
654
+ "eval_wer": 0.36609344401366856,
655
+ "step": 38000
656
+ },
657
+ {
658
+ "epoch": 18.16,
659
+ "learning_rate": 0.00012039776357827474,
660
+ "loss": 0.8032,
661
+ "step": 38500
662
+ },
663
+ {
664
+ "epoch": 18.39,
665
+ "learning_rate": 0.00011800159744408944,
666
+ "loss": 0.7753,
667
+ "step": 39000
668
+ },
669
+ {
670
+ "epoch": 18.63,
671
+ "learning_rate": 0.00011560543130990414,
672
+ "loss": 0.7608,
673
+ "step": 39500
674
+ },
675
+ {
676
+ "epoch": 18.86,
677
+ "learning_rate": 0.00011321405750798721,
678
+ "loss": 0.759,
679
+ "step": 40000
680
+ },
681
+ {
682
+ "epoch": 18.86,
683
+ "eval_cer": 0.1336804268071908,
684
+ "eval_loss": 0.49982598423957825,
685
+ "eval_runtime": 2725.5181,
686
+ "eval_samples_per_second": 6.206,
687
+ "eval_steps_per_second": 0.388,
688
+ "eval_wer": 0.3533921619928803,
689
+ "step": 40000
690
+ },
691
+ {
692
+ "epoch": 19.1,
693
+ "learning_rate": 0.00011081789137380191,
694
+ "loss": 0.7285,
695
+ "step": 40500
696
+ },
697
+ {
698
+ "epoch": 19.34,
699
+ "learning_rate": 0.00010842172523961661,
700
+ "loss": 0.7036,
701
+ "step": 41000
702
+ },
703
+ {
704
+ "epoch": 19.57,
705
+ "learning_rate": 0.00010602555910543131,
706
+ "loss": 0.6953,
707
+ "step": 41500
708
+ },
709
+ {
710
+ "epoch": 19.81,
711
+ "learning_rate": 0.00010363418530351436,
712
+ "loss": 0.6907,
713
+ "step": 42000
714
+ },
715
+ {
716
+ "epoch": 19.81,
717
+ "eval_cer": 0.12877502322923437,
718
+ "eval_loss": 0.47100237011909485,
719
+ "eval_runtime": 2667.1801,
720
+ "eval_samples_per_second": 6.342,
721
+ "eval_steps_per_second": 0.397,
722
+ "eval_wer": 0.34119829698166165,
723
+ "step": 42000
724
+ },
725
+ {
726
+ "epoch": 20.04,
727
+ "learning_rate": 0.00010123801916932906,
728
+ "loss": 0.6858,
729
+ "step": 42500
730
+ },
731
+ {
732
+ "epoch": 20.28,
733
+ "learning_rate": 9.884664536741213e-05,
734
+ "loss": 0.6603,
735
+ "step": 43000
736
+ },
737
+ {
738
+ "epoch": 20.51,
739
+ "learning_rate": 9.645047923322683e-05,
740
+ "loss": 0.6609,
741
+ "step": 43500
742
+ },
743
+ {
744
+ "epoch": 20.75,
745
+ "learning_rate": 9.405431309904153e-05,
746
+ "loss": 0.659,
747
+ "step": 44000
748
+ },
749
+ {
750
+ "epoch": 20.75,
751
+ "eval_cer": 0.12423387149543921,
752
+ "eval_loss": 0.4578304886817932,
753
+ "eval_runtime": 2665.8908,
754
+ "eval_samples_per_second": 6.345,
755
+ "eval_steps_per_second": 0.397,
756
+ "eval_wer": 0.3324532819573611,
757
+ "step": 44000
758
+ },
759
+ {
760
+ "epoch": 20.99,
761
+ "learning_rate": 9.165814696485623e-05,
762
+ "loss": 0.6567,
763
+ "step": 44500
764
+ },
765
+ {
766
+ "epoch": 21.22,
767
+ "learning_rate": 8.926198083067093e-05,
768
+ "loss": 0.6437,
769
+ "step": 45000
770
+ },
771
+ {
772
+ "epoch": 21.46,
773
+ "learning_rate": 8.686581469648561e-05,
774
+ "loss": 0.6371,
775
+ "step": 45500
776
+ },
777
+ {
778
+ "epoch": 21.69,
779
+ "learning_rate": 8.447444089456868e-05,
780
+ "loss": 0.6345,
781
+ "step": 46000
782
+ },
783
+ {
784
+ "epoch": 21.69,
785
+ "eval_cer": 0.12205708502554125,
786
+ "eval_loss": 0.45305466651916504,
787
+ "eval_runtime": 2667.1819,
788
+ "eval_samples_per_second": 6.342,
789
+ "eval_steps_per_second": 0.397,
790
+ "eval_wer": 0.3256982930174662,
791
+ "step": 46000
792
+ },
793
+ {
794
+ "epoch": 21.93,
795
+ "learning_rate": 8.207827476038337e-05,
796
+ "loss": 0.6418,
797
+ "step": 46500
798
+ },
799
+ {
800
+ "epoch": 22.16,
801
+ "learning_rate": 7.968210862619807e-05,
802
+ "loss": 0.6306,
803
+ "step": 47000
804
+ },
805
+ {
806
+ "epoch": 22.4,
807
+ "learning_rate": 7.728594249201278e-05,
808
+ "loss": 0.6213,
809
+ "step": 47500
810
+ },
811
+ {
812
+ "epoch": 22.64,
813
+ "learning_rate": 7.489456869009583e-05,
814
+ "loss": 0.6242,
815
+ "step": 48000
816
+ },
817
+ {
818
+ "epoch": 22.64,
819
+ "eval_cer": 0.12094251800538308,
820
+ "eval_loss": 0.4497627019882202,
821
+ "eval_runtime": 2727.8154,
822
+ "eval_samples_per_second": 6.201,
823
+ "eval_steps_per_second": 0.388,
824
+ "eval_wer": 0.32180545314717474,
825
+ "step": 48000
826
+ },
827
+ {
828
+ "epoch": 22.87,
829
+ "learning_rate": 7.249840255591053e-05,
830
+ "loss": 0.6294,
831
+ "step": 48500
832
+ },
833
+ {
834
+ "epoch": 23.11,
835
+ "learning_rate": 7.010223642172524e-05,
836
+ "loss": 0.6141,
837
+ "step": 49000
838
+ },
839
+ {
840
+ "epoch": 23.34,
841
+ "learning_rate": 6.770607028753993e-05,
842
+ "loss": 0.6155,
843
+ "step": 49500
844
+ },
845
+ {
846
+ "epoch": 23.58,
847
+ "learning_rate": 6.530990415335462e-05,
848
+ "loss": 0.6163,
849
+ "step": 50000
850
+ },
851
+ {
852
+ "epoch": 23.58,
853
+ "eval_cer": 0.11941571386818009,
854
+ "eval_loss": 0.45521289110183716,
855
+ "eval_runtime": 2664.8843,
856
+ "eval_samples_per_second": 6.347,
857
+ "eval_steps_per_second": 0.397,
858
+ "eval_wer": 0.3188402349975026,
859
+ "step": 50000
860
+ },
861
+ {
862
+ "epoch": 23.82,
863
+ "learning_rate": 6.291373801916932e-05,
864
+ "loss": 0.6167,
865
+ "step": 50500
866
+ },
867
+ {
868
+ "epoch": 24.05,
869
+ "learning_rate": 6.0522364217252394e-05,
870
+ "loss": 0.6179,
871
+ "step": 51000
872
+ },
873
+ {
874
+ "epoch": 24.29,
875
+ "learning_rate": 5.8126198083067085e-05,
876
+ "loss": 0.6154,
877
+ "step": 51500
878
+ },
879
+ {
880
+ "epoch": 24.52,
881
+ "learning_rate": 5.573482428115016e-05,
882
+ "loss": 0.6121,
883
+ "step": 52000
884
+ },
885
+ {
886
+ "epoch": 24.52,
887
+ "eval_cer": 0.1153500525656853,
888
+ "eval_loss": 0.46334853768348694,
889
+ "eval_runtime": 2666.3375,
890
+ "eval_samples_per_second": 6.344,
891
+ "eval_steps_per_second": 0.397,
892
+ "eval_wer": 0.3136947093848362,
893
+ "step": 52000
894
+ },
895
+ {
896
+ "epoch": 24.76,
897
+ "learning_rate": 5.3338658146964855e-05,
898
+ "loss": 0.6227,
899
+ "step": 52500
900
+ },
901
+ {
902
+ "epoch": 24.99,
903
+ "learning_rate": 5.0942492012779546e-05,
904
+ "loss": 0.6156,
905
+ "step": 53000
906
+ },
907
+ {
908
+ "epoch": 25.23,
909
+ "learning_rate": 4.854632587859424e-05,
910
+ "loss": 0.6159,
911
+ "step": 53500
912
+ },
913
+ {
914
+ "epoch": 25.47,
915
+ "learning_rate": 4.615015974440894e-05,
916
+ "loss": 0.6054,
917
+ "step": 54000
918
+ },
919
+ {
920
+ "epoch": 25.47,
921
+ "eval_cer": 0.11759009234983882,
922
+ "eval_loss": 0.46227386593818665,
923
+ "eval_runtime": 2666.9277,
924
+ "eval_samples_per_second": 6.342,
925
+ "eval_steps_per_second": 0.397,
926
+ "eval_wer": 0.3171356309809798,
927
+ "step": 54000
928
+ },
929
+ {
930
+ "epoch": 25.7,
931
+ "learning_rate": 4.375399361022364e-05,
932
+ "loss": 0.6051,
933
+ "step": 54500
934
+ },
935
+ {
936
+ "epoch": 25.94,
937
+ "learning_rate": 4.1362619808306704e-05,
938
+ "loss": 0.5986,
939
+ "step": 55000
940
+ },
941
+ {
942
+ "epoch": 26.17,
943
+ "learning_rate": 3.89664536741214e-05,
944
+ "loss": 0.5916,
945
+ "step": 55500
946
+ },
947
+ {
948
+ "epoch": 26.41,
949
+ "learning_rate": 3.65702875399361e-05,
950
+ "loss": 0.591,
951
+ "step": 56000
952
+ },
953
+ {
954
+ "epoch": 26.41,
955
+ "eval_cer": 0.11455829556310718,
956
+ "eval_loss": 0.4413212835788727,
957
+ "eval_runtime": 2669.4551,
958
+ "eval_samples_per_second": 6.336,
959
+ "eval_steps_per_second": 0.396,
960
+ "eval_wer": 0.31158575743881267,
961
+ "step": 56000
962
+ },
963
+ {
964
+ "epoch": 26.64,
965
+ "learning_rate": 3.41741214057508e-05,
966
+ "loss": 0.5904,
967
+ "step": 56500
968
+ },
969
+ {
970
+ "epoch": 26.88,
971
+ "learning_rate": 3.178274760383386e-05,
972
+ "loss": 0.5887,
973
+ "step": 57000
974
+ },
975
+ {
976
+ "epoch": 27.12,
977
+ "learning_rate": 2.9386581469648557e-05,
978
+ "loss": 0.5768,
979
+ "step": 57500
980
+ },
981
+ {
982
+ "epoch": 27.35,
983
+ "learning_rate": 2.6990415335463258e-05,
984
+ "loss": 0.5713,
985
+ "step": 58000
986
+ },
987
+ {
988
+ "epoch": 27.35,
989
+ "eval_cer": 0.11345245313801873,
990
+ "eval_loss": 0.4338010549545288,
991
+ "eval_runtime": 2668.5066,
992
+ "eval_samples_per_second": 6.338,
993
+ "eval_steps_per_second": 0.396,
994
+ "eval_wer": 0.3092706673326515,
995
+ "step": 58000
996
+ },
997
+ {
998
+ "epoch": 27.59,
999
+ "learning_rate": 2.4594249201277952e-05,
1000
+ "loss": 0.5653,
1001
+ "step": 58500
1002
+ },
1003
+ {
1004
+ "epoch": 27.82,
1005
+ "learning_rate": 2.219808306709265e-05,
1006
+ "loss": 0.569,
1007
+ "step": 59000
1008
+ },
1009
+ {
1010
+ "epoch": 28.06,
1011
+ "learning_rate": 1.980670926517572e-05,
1012
+ "loss": 0.5748,
1013
+ "step": 59500
1014
+ },
1015
+ {
1016
+ "epoch": 28.3,
1017
+ "learning_rate": 1.7410543130990413e-05,
1018
+ "loss": 0.5703,
1019
+ "step": 60000
1020
+ },
1021
+ {
1022
+ "epoch": 28.3,
1023
+ "eval_cer": 0.11209795975344294,
1024
+ "eval_loss": 0.42797738313674927,
1025
+ "eval_runtime": 2667.5384,
1026
+ "eval_samples_per_second": 6.341,
1027
+ "eval_steps_per_second": 0.397,
1028
+ "eval_wer": 0.30612309619516526,
1029
+ "step": 60000
1030
+ },
1031
+ {
1032
+ "epoch": 28.53,
1033
+ "learning_rate": 1.501437699680511e-05,
1034
+ "loss": 0.5606,
1035
+ "step": 60500
1036
+ },
1037
+ {
1038
+ "epoch": 28.77,
1039
+ "learning_rate": 1.2623003194888177e-05,
1040
+ "loss": 0.5647,
1041
+ "step": 61000
1042
+ },
1043
+ {
1044
+ "epoch": 29.0,
1045
+ "learning_rate": 1.0226837060702875e-05,
1046
+ "loss": 0.5567,
1047
+ "step": 61500
1048
+ },
1049
+ {
1050
+ "epoch": 29.24,
1051
+ "learning_rate": 7.830670926517571e-06,
1052
+ "loss": 0.5576,
1053
+ "step": 62000
1054
+ },
1055
+ {
1056
+ "epoch": 29.24,
1057
+ "eval_cer": 0.11193437359588548,
1058
+ "eval_loss": 0.42482054233551025,
1059
+ "eval_runtime": 2665.5981,
1060
+ "eval_samples_per_second": 6.345,
1061
+ "eval_steps_per_second": 0.397,
1062
+ "eval_wer": 0.30466427229265275,
1063
+ "step": 62000
1064
+ },
1065
+ {
1066
+ "epoch": 29.47,
1067
+ "learning_rate": 5.434504792332268e-06,
1068
+ "loss": 0.5596,
1069
+ "step": 62500
1070
+ },
1071
+ {
1072
+ "epoch": 29.71,
1073
+ "learning_rate": 3.0431309904153355e-06,
1074
+ "loss": 0.5567,
1075
+ "step": 63000
1076
+ },
1077
+ {
1078
+ "epoch": 29.95,
1079
+ "learning_rate": 6.469648562300319e-07,
1080
+ "loss": 0.5581,
1081
+ "step": 63500
1082
+ },
1083
+ {
1084
+ "epoch": 29.99,
1085
+ "step": 63600,
1086
+ "total_flos": 4.271260053717039e+20,
1087
+ "train_loss": 0.9622837933354408,
1088
+ "train_runtime": 156348.4692,
1089
+ "train_samples_per_second": 26.04,
1090
+ "train_steps_per_second": 0.407
1091
+ }
1092
+ ],
1093
+ "logging_steps": 500,
1094
+ "max_steps": 63600,
1095
+ "num_train_epochs": 30,
1096
+ "save_steps": 2000,
1097
+ "total_flos": 4.271260053717039e+20,
1098
+ "trial_name": null,
1099
+ "trial_params": null
1100
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dffbeeea609cd56167ef7580a93c1e52573b1cb0b44833565d5736f74bf636b
3
+ size 4536
vocab.json ADDED
@@ -0,0 +1,1792 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "0": 1,
3
+ "1": 2,
4
+ "2": 3,
5
+ "3": 4,
6
+ "4": 5,
7
+ "5": 6,
8
+ "6": 7,
9
+ "7": 8,
10
+ "8": 9,
11
+ "9": 10,
12
+ "=": 11,
13
+ "[PAD]": 1789,
14
+ "[UNK]": 1788,
15
+ "_": 12,
16
+ "a": 13,
17
+ "b": 14,
18
+ "c": 15,
19
+ "d": 16,
20
+ "e": 17,
21
+ "f": 18,
22
+ "g": 19,
23
+ "h": 20,
24
+ "i": 21,
25
+ "j": 22,
26
+ "k": 23,
27
+ "l": 24,
28
+ "m": 25,
29
+ "n": 26,
30
+ "o": 27,
31
+ "p": 28,
32
+ "q": 29,
33
+ "r": 30,
34
+ "s": 31,
35
+ "t": 32,
36
+ "u": 33,
37
+ "v": 34,
38
+ "w": 35,
39
+ "x": 36,
40
+ "y": 37,
41
+ "z": 38,
42
+ "|": 0,
43
+ "ㄱ": 39,
44
+ "ㄴ": 40,
45
+ "ㄹ": 41,
46
+ "ㅁ": 42,
47
+ "ㅅ": 43,
48
+ "ㅇ": 44,
49
+ "ㅈ": 45,
50
+ "ㅋ": 46,
51
+ "ㅍ": 47,
52
+ "ㅠ": 48,
53
+ "ㅡ": 49,
54
+ "㎛": 50,
55
+ "가": 51,
56
+ "각": 52,
57
+ "간": 53,
58
+ "갇": 54,
59
+ "갈": 55,
60
+ "갉": 56,
61
+ "감": 57,
62
+ "갑": 58,
63
+ "값": 59,
64
+ "갓": 60,
65
+ "갔": 61,
66
+ "강": 62,
67
+ "갖": 63,
68
+ "같": 64,
69
+ "갚": 65,
70
+ "갛": 66,
71
+ "개": 67,
72
+ "객": 68,
73
+ "갠": 69,
74
+ "갤": 70,
75
+ "갬": 71,
76
+ "갭": 72,
77
+ "갯": 73,
78
+ "갰": 74,
79
+ "갱": 75,
80
+ "갸": 76,
81
+ "걀": 77,
82
+ "걍": 78,
83
+ "걔": 79,
84
+ "걘": 80,
85
+ "거": 81,
86
+ "걱": 82,
87
+ "건": 83,
88
+ "걷": 84,
89
+ "걸": 85,
90
+ "검": 86,
91
+ "겁": 87,
92
+ "것": 88,
93
+ "겄": 89,
94
+ "겅": 90,
95
+ "겉": 91,
96
+ "겊": 92,
97
+ "겋": 93,
98
+ "게": 94,
99
+ "겍": 95,
100
+ "겐": 96,
101
+ "겔": 97,
102
+ "겜": 98,
103
+ "겟": 99,
104
+ "겠": 100,
105
+ "겨": 101,
106
+ "격": 102,
107
+ "겪": 103,
108
+ "견": 104,
109
+ "결": 105,
110
+ "겸": 106,
111
+ "겹": 107,
112
+ "겼": 108,
113
+ "경": 109,
114
+ "곁": 110,
115
+ "계": 111,
116
+ "곗": 112,
117
+ "곘": 113,
118
+ "고": 114,
119
+ "곡": 115,
120
+ "곤": 116,
121
+ "곧": 117,
122
+ "골": 118,
123
+ "곪": 119,
124
+ "곯": 120,
125
+ "곰": 121,
126
+ "곱": 122,
127
+ "곳": 123,
128
+ "공": 124,
129
+ "과": 125,
130
+ "곽": 126,
131
+ "관": 127,
132
+ "괄": 128,
133
+ "괌": 129,
134
+ "광": 130,
135
+ "괘": 131,
136
+ "괜": 132,
137
+ "괴": 133,
138
+ "굉": 134,
139
+ "교": 135,
140
+ "굘": 136,
141
+ "굥": 137,
142
+ "구": 138,
143
+ "국": 139,
144
+ "군": 140,
145
+ "굳": 141,
146
+ "굴": 142,
147
+ "굵": 143,
148
+ "굶": 144,
149
+ "굷": 145,
150
+ "굼": 146,
151
+ "굽": 147,
152
+ "굿": 148,
153
+ "궁": 149,
154
+ "궈": 150,
155
+ "권": 151,
156
+ "궐": 152,
157
+ "궜": 153,
158
+ "궤": 154,
159
+ "귀": 155,
160
+ "귄": 156,
161
+ "귓": 157,
162
+ "규": 158,
163
+ "균": 159,
164
+ "귤": 160,
165
+ "그": 161,
166
+ "극": 162,
167
+ "근": 163,
168
+ "귿": 164,
169
+ "글": 165,
170
+ "긁": 166,
171
+ "금": 167,
172
+ "급": 168,
173
+ "긋": 169,
174
+ "긍": 170,
175
+ "긎": 171,
176
+ "긓": 172,
177
+ "기": 173,
178
+ "긱": 174,
179
+ "긴": 175,
180
+ "길": 176,
181
+ "김": 177,
182
+ "깁": 178,
183
+ "깃": 179,
184
+ "깄": 180,
185
+ "깅": 181,
186
+ "깊": 182,
187
+ "까": 183,
188
+ "깍": 184,
189
+ "깎": 185,
190
+ "깐": 186,
191
+ "깔": 187,
192
+ "깜": 188,
193
+ "깝": 189,
194
+ "깠": 190,
195
+ "깡": 191,
196
+ "깥": 192,
197
+ "깨": 193,
198
+ "깬": 194,
199
+ "깰": 195,
200
+ "깸": 196,
201
+ "깹": 197,
202
+ "깻": 198,
203
+ "깼": 199,
204
+ "깽": 200,
205
+ "꺄": 201,
206
+ "꺠": 202,
207
+ "꺵": 203,
208
+ "꺼": 204,
209
+ "꺽": 205,
210
+ "꺾": 206,
211
+ "껀": 207,
212
+ "껄": 208,
213
+ "껌": 209,
214
+ "껍": 210,
215
+ "껏": 211,
216
+ "껐": 212,
217
+ "껑": 213,
218
+ "께": 214,
219
+ "껬": 215,
220
+ "껴": 216,
221
+ "꼈": 217,
222
+ "꼐": 218,
223
+ "꼬": 219,
224
+ "꼭": 220,
225
+ "꼰": 221,
226
+ "꼴": 222,
227
+ "꼼": 223,
228
+ "꼽": 224,
229
+ "꼿": 225,
230
+ "꽁": 226,
231
+ "꽂": 227,
232
+ "꽃": 228,
233
+ "꽈": 229,
234
+ "꽉": 230,
235
+ "꽝": 231,
236
+ "꽤": 232,
237
+ "꽥": 233,
238
+ "꽹": 234,
239
+ "꾀": 235,
240
+ "꾸": 236,
241
+ "꾹": 237,
242
+ "꾼": 238,
243
+ "꿀": 239,
244
+ "꿇": 240,
245
+ "꿈": 241,
246
+ "꿉": 242,
247
+ "꿋": 243,
248
+ "꿍": 244,
249
+ "꿔": 245,
250
+ "꿨": 246,
251
+ "꿩": 247,
252
+ "꿰": 248,
253
+ "뀌": 249,
254
+ "뀐": 250,
255
+ "뀔": 251,
256
+ "뀝": 252,
257
+ "뀨": 253,
258
+ "끄": 254,
259
+ "끅": 255,
260
+ "끈": 256,
261
+ "끊": 257,
262
+ "끌": 258,
263
+ "끍": 259,
264
+ "끓": 260,
265
+ "끔": 261,
266
+ "끕": 262,
267
+ "끗": 263,
268
+ "끙": 264,
269
+ "끝": 265,
270
+ "끼": 266,
271
+ "끽": 267,
272
+ "낀": 268,
273
+ "낄": 269,
274
+ "낌": 270,
275
+ "낍": 271,
276
+ "낑": 272,
277
+ "나": 273,
278
+ "낙": 274,
279
+ "낚": 275,
280
+ "난": 276,
281
+ "날": 277,
282
+ "낡": 278,
283
+ "남": 279,
284
+ "납": 280,
285
+ "낫": 281,
286
+ "났": 282,
287
+ "낭": 283,
288
+ "낮": 284,
289
+ "낯": 285,
290
+ "낱": 286,
291
+ "낳": 287,
292
+ "내": 288,
293
+ "낸": 289,
294
+ "낼": 290,
295
+ "냄": 291,
296
+ "냅": 292,
297
+ "냈": 293,
298
+ "냉": 294,
299
+ "냐": 295,
300
+ "냠": 296,
301
+ "냥": 297,
302
+ "너": 298,
303
+ "넉": 299,
304
+ "넋": 300,
305
+ "넌": 301,
306
+ "널": 302,
307
+ "넒": 303,
308
+ "넓": 304,
309
+ "넘": 305,
310
+ "넙": 306,
311
+ "넛": 307,
312
+ "넜": 308,
313
+ "넝": 309,
314
+ "넣": 310,
315
+ "네": 311,
316
+ "넥": 312,
317
+ "넬": 313,
318
+ "넴": 314,
319
+ "넵": 315,
320
+ "넷": 316,
321
+ "녀": 317,
322
+ "녁": 318,
323
+ "년": 319,
324
+ "념": 320,
325
+ "녔": 321,
326
+ "녕": 322,
327
+ "녜": 323,
328
+ "노": 324,
329
+ "녹": 325,
330
+ "논": 326,
331
+ "놀": 327,
332
+ "놈": 328,
333
+ "놉": 329,
334
+ "농": 330,
335
+ "높": 331,
336
+ "놓": 332,
337
+ "놔": 333,
338
+ "놨": 334,
339
+ "뇌": 335,
340
+ "뇨": 336,
341
+ "뇰": 337,
342
+ "뇽": 338,
343
+ "누": 339,
344
+ "눅": 340,
345
+ "눈": 341,
346
+ "눌": 342,
347
+ "눔": 343,
348
+ "눕": 344,
349
+ "눗": 345,
350
+ "눙": 346,
351
+ "눠": 347,
352
+ "눴": 348,
353
+ "뉘": 349,
354
+ "뉜": 350,
355
+ "뉴": 351,
356
+ "늄": 352,
357
+ "느": 353,
358
+ "늑": 354,
359
+ "는": 355,
360
+ "늘": 356,
361
+ "늙": 357,
362
+ "늠": 358,
363
+ "능": 359,
364
+ "늦": 360,
365
+ "늬": 361,
366
+ "니": 362,
367
+ "닉": 363,
368
+ "닌": 364,
369
+ "닐": 365,
370
+ "님": 366,
371
+ "닙": 367,
372
+ "닛": 368,
373
+ "닝": 369,
374
+ "닢": 370,
375
+ "다": 371,
376
+ "닥": 372,
377
+ "닦": 373,
378
+ "단": 374,
379
+ "닫": 375,
380
+ "달": 376,
381
+ "닭": 377,
382
+ "닮": 378,
383
+ "닳": 379,
384
+ "담": 380,
385
+ "답": 381,
386
+ "닷": 382,
387
+ "당": 383,
388
+ "닻": 384,
389
+ "닿": 385,
390
+ "대": 386,
391
+ "댁": 387,
392
+ "댄": 388,
393
+ "댈": 389,
394
+ "댐": 390,
395
+ "댑": 391,
396
+ "댓": 392,
397
+ "댔": 393,
398
+ "댕": 394,
399
+ "댜": 395,
400
+ "더": 396,
401
+ "덕": 397,
402
+ "던": 398,
403
+ "덜": 399,
404
+ "덞": 400,
405
+ "덟": 401,
406
+ "덤": 402,
407
+ "덥": 403,
408
+ "덧": 404,
409
+ "덩": 405,
410
+ "덫": 406,
411
+ "덮": 407,
412
+ "데": 408,
413
+ "덱": 409,
414
+ "덴": 410,
415
+ "델": 411,
416
+ "뎀": 412,
417
+ "뎁": 413,
418
+ "뎃": 414,
419
+ "뎅": 415,
420
+ "뎌": 416,
421
+ "도": 417,
422
+ "독": 418,
423
+ "돈": 419,
424
+ "돋": 420,
425
+ "돌": 421,
426
+ "돔": 422,
427
+ "돕": 423,
428
+ "돗": 424,
429
+ "동": 425,
430
+ "돠": 426,
431
+ "돼": 427,
432
+ "됄": 428,
433
+ "됐": 429,
434
+ "되": 430,
435
+ "된": 431,
436
+ "될": 432,
437
+ "됨": 433,
438
+ "됩": 434,
439
+ "됬": 435,
440
+ "두": 436,
441
+ "둑": 437,
442
+ "둔": 438,
443
+ "둘": 439,
444
+ "둠": 440,
445
+ "둡": 441,
446
+ "둥": 442,
447
+ "둬": 443,
448
+ "뒀": 444,
449
+ "뒤": 445,
450
+ "뒨": 446,
451
+ "뒷": 447,
452
+ "뒹": 448,
453
+ "듀": 449,
454
+ "듄": 450,
455
+ "드": 451,
456
+ "득": 452,
457
+ "든": 453,
458
+ "듣": 454,
459
+ "들": 455,
460
+ "듦": 456,
461
+ "듬": 457,
462
+ "듭": 458,
463
+ "듯": 459,
464
+ "등": 460,
465
+ "디": 461,
466
+ "딕": 462,
467
+ "딘": 463,
468
+ "딛": 464,
469
+ "딜": 465,
470
+ "딤": 466,
471
+ "딧": 467,
472
+ "딨": 468,
473
+ "딩": 469,
474
+ "딪": 470,
475
+ "따": 471,
476
+ "딱": 472,
477
+ "딲": 473,
478
+ "딴": 474,
479
+ "딸": 475,
480
+ "땀": 476,
481
+ "땄": 477,
482
+ "땅": 478,
483
+ "때": 479,
484
+ "땐": 480,
485
+ "땜": 481,
486
+ "땠": 482,
487
+ "땡": 483,
488
+ "떄": 484,
489
+ "떙": 485,
490
+ "떠": 486,
491
+ "떡": 487,
492
+ "떤": 488,
493
+ "떨": 489,
494
+ "떫": 490,
495
+ "떱": 491,
496
+ "떳": 492,
497
+ "떴": 493,
498
+ "떻": 494,
499
+ "떼": 495,
500
+ "뗀": 496,
501
+ "뗏": 497,
502
+ "뗐": 498,
503
+ "또": 499,
504
+ "똑": 500,
505
+ "똘": 501,
506
+ "똠": 502,
507
+ "똥": 503,
508
+ "뙤": 504,
509
+ "뚜": 505,
510
+ "뚝": 506,
511
+ "뚤": 507,
512
+ "뚫": 508,
513
+ "뚱": 509,
514
+ "뛰": 510,
515
+ "뛴": 511,
516
+ "뛸": 512,
517
+ "뜀": 513,
518
+ "뜁": 514,
519
+ "뜨": 515,
520
+ "뜩": 516,
521
+ "뜬": 517,
522
+ "뜯": 518,
523
+ "뜰": 519,
524
+ "뜸": 520,
525
+ "뜹": 521,
526
+ "뜻": 522,
527
+ "뜽": 523,
528
+ "띄": 524,
529
+ "띈": 525,
530
+ "띌": 526,
531
+ "띕": 527,
532
+ "띠": 528,
533
+ "띡": 529,
534
+ "띤": 530,
535
+ "띵": 531,
536
+ "라": 532,
537
+ "락": 533,
538
+ "란": 534,
539
+ "랄": 535,
540
+ "람": 536,
541
+ "랍": 537,
542
+ "랏": 538,
543
+ "랐": 539,
544
+ "랑": 540,
545
+ "랖": 541,
546
+ "랗": 542,
547
+ "래": 543,
548
+ "랙": 544,
549
+ "랜": 545,
550
+ "랠": 546,
551
+ "램": 547,
552
+ "랩": 548,
553
+ "랫": 549,
554
+ "랬": 550,
555
+ "랭": 551,
556
+ "랴": 552,
557
+ "략": 553,
558
+ "럈": 554,
559
+ "량": 555,
560
+ "러": 556,
561
+ "럭": 557,
562
+ "런": 558,
563
+ "럴": 559,
564
+ "럼": 560,
565
+ "럽": 561,
566
+ "럿": 562,
567
+ "렀": 563,
568
+ "렁": 564,
569
+ "렇": 565,
570
+ "레": 566,
571
+ "렉": 567,
572
+ "렌": 568,
573
+ "렐": 569,
574
+ "렘": 570,
575
+ "렙": 571,
576
+ "렛": 572,
577
+ "렝": 573,
578
+ "려": 574,
579
+ "력": 575,
580
+ "련": 576,
581
+ "렬": 577,
582
+ "렴": 578,
583
+ "렵": 579,
584
+ "렷": 580,
585
+ "렸": 581,
586
+ "령": 582,
587
+ "렼": 583,
588
+ "례": 584,
589
+ "로": 585,
590
+ "록": 586,
591
+ "론": 587,
592
+ "롤": 588,
593
+ "롬": 589,
594
+ "롭": 590,
595
+ "롯": 591,
596
+ "롱": 592,
597
+ "롷": 593,
598
+ "롸": 594,
599
+ "뢰": 595,
600
+ "료": 596,
601
+ "룐": 597,
602
+ "룟": 598,
603
+ "룡": 599,
604
+ "루": 600,
605
+ "룩": 601,
606
+ "룬": 602,
607
+ "룰": 603,
608
+ "룸": 604,
609
+ "룹": 605,
610
+ "룻": 606,
611
+ "룽": 607,
612
+ "뤄": 608,
613
+ "뤘": 609,
614
+ "뤼": 610,
615
+ "류": 611,
616
+ "륙": 612,
617
+ "륜": 613,
618
+ "률": 614,
619
+ "륨": 615,
620
+ "륭": 616,
621
+ "르": 617,
622
+ "륵": 618,
623
+ "른": 619,
624
+ "를": 620,
625
+ "름": 621,
626
+ "릅": 622,
627
+ "릇": 623,
628
+ "릉": 624,
629
+ "릎": 625,
630
+ "릏": 626,
631
+ "리": 627,
632
+ "릭": 628,
633
+ "린": 629,
634
+ "릴": 630,
635
+ "림": 631,
636
+ "립": 632,
637
+ "릿": 633,
638
+ "링": 634,
639
+ "마": 635,
640
+ "막": 636,
641
+ "만": 637,
642
+ "많": 638,
643
+ "맏": 639,
644
+ "말": 640,
645
+ "맑": 641,
646
+ "맘": 642,
647
+ "맙": 643,
648
+ "맛": 644,
649
+ "망": 645,
650
+ "맞": 646,
651
+ "맡": 647,
652
+ "맣": 648,
653
+ "매": 649,
654
+ "맥": 650,
655
+ "맨": 651,
656
+ "맬": 652,
657
+ "맴": 653,
658
+ "맵": 654,
659
+ "맷": 655,
660
+ "맸": 656,
661
+ "맹": 657,
662
+ "맺": 658,
663
+ "먀": 659,
664
+ "머": 660,
665
+ "먹": 661,
666
+ "먼": 662,
667
+ "멀": 663,
668
+ "멈": 664,
669
+ "멉": 665,
670
+ "멋": 666,
671
+ "멍": 667,
672
+ "멎": 668,
673
+ "멓": 669,
674
+ "메": 670,
675
+ "멕": 671,
676
+ "멘": 672,
677
+ "멜": 673,
678
+ "멤": 674,
679
+ "멧": 675,
680
+ "멩": 676,
681
+ "며": 677,
682
+ "멱": 678,
683
+ "면": 679,
684
+ "멸": 680,
685
+ "명": 681,
686
+ "몇": 682,
687
+ "모": 683,
688
+ "목": 684,
689
+ "몫": 685,
690
+ "몬": 686,
691
+ "몰": 687,
692
+ "몸": 688,
693
+ "몹": 689,
694
+ "못": 690,
695
+ "���": 691,
696
+ "묘": 692,
697
+ "무": 693,
698
+ "묵": 694,
699
+ "묶": 695,
700
+ "문": 696,
701
+ "묻": 697,
702
+ "물": 698,
703
+ "묽": 699,
704
+ "뭅": 700,
705
+ "뭇": 701,
706
+ "뭉": 702,
707
+ "뭍": 703,
708
+ "뭐": 704,
709
+ "뭔": 705,
710
+ "뭘": 706,
711
+ "뭡": 707,
712
+ "뭣": 708,
713
+ "뮈": 709,
714
+ "뮤": 710,
715
+ "뮬": 711,
716
+ "뮹": 712,
717
+ "므": 713,
718
+ "믄": 714,
719
+ "믈": 715,
720
+ "미": 716,
721
+ "믹": 717,
722
+ "민": 718,
723
+ "믿": 719,
724
+ "밀": 720,
725
+ "밈": 721,
726
+ "밉": 722,
727
+ "밋": 723,
728
+ "밌": 724,
729
+ "밍": 725,
730
+ "및": 726,
731
+ "밑": 727,
732
+ "바": 728,
733
+ "박": 729,
734
+ "밖": 730,
735
+ "반": 731,
736
+ "받": 732,
737
+ "발": 733,
738
+ "밝": 734,
739
+ "밟": 735,
740
+ "밤": 736,
741
+ "밥": 737,
742
+ "밧": 738,
743
+ "방": 739,
744
+ "밭": 740,
745
+ "배": 741,
746
+ "백": 742,
747
+ "밴": 743,
748
+ "밸": 744,
749
+ "뱀": 745,
750
+ "뱁": 746,
751
+ "뱃": 747,
752
+ "뱅": 748,
753
+ "뱉": 749,
754
+ "버": 750,
755
+ "벅": 751,
756
+ "번": 752,
757
+ "벌": 753,
758
+ "범": 754,
759
+ "법": 755,
760
+ "벗": 756,
761
+ "벙": 757,
762
+ "벚": 758,
763
+ "베": 759,
764
+ "벡": 760,
765
+ "벤": 761,
766
+ "벨": 762,
767
+ "벳": 763,
768
+ "벵": 764,
769
+ "벼": 765,
770
+ "벽": 766,
771
+ "변": 767,
772
+ "별": 768,
773
+ "볍": 769,
774
+ "볐": 770,
775
+ "병": 771,
776
+ "볕": 772,
777
+ "보": 773,
778
+ "복": 774,
779
+ "볶": 775,
780
+ "본": 776,
781
+ "볼": 777,
782
+ "봄": 778,
783
+ "봅": 779,
784
+ "봇": 780,
785
+ "봈": 781,
786
+ "봉": 782,
787
+ "봐": 783,
788
+ "봤": 784,
789
+ "봬": 785,
790
+ "뵀": 786,
791
+ "뵈": 787,
792
+ "뵌": 788,
793
+ "뵐": 789,
794
+ "뵙": 790,
795
+ "뵜": 791,
796
+ "부": 792,
797
+ "북": 793,
798
+ "분": 794,
799
+ "불": 795,
800
+ "붉": 796,
801
+ "붐": 797,
802
+ "붑": 798,
803
+ "붓": 799,
804
+ "붕": 800,
805
+ "붙": 801,
806
+ "뷔": 802,
807
+ "뷰": 803,
808
+ "뷸": 804,
809
+ "브": 805,
810
+ "븐": 806,
811
+ "블": 807,
812
+ "비": 808,
813
+ "빅": 809,
814
+ "빈": 810,
815
+ "빌": 811,
816
+ "빔": 812,
817
+ "빕": 813,
818
+ "빗": 814,
819
+ "빙": 815,
820
+ "빚": 816,
821
+ "빛": 817,
822
+ "빠": 818,
823
+ "빡": 819,
824
+ "빤": 820,
825
+ "빨": 821,
826
+ "빰": 822,
827
+ "빳": 823,
828
+ "빴": 824,
829
+ "빵": 825,
830
+ "빼": 826,
831
+ "빽": 827,
832
+ "뺀": 828,
833
+ "뺄": 829,
834
+ "뺍": 830,
835
+ "뺏": 831,
836
+ "뺐": 832,
837
+ "뺑": 833,
838
+ "뺨": 834,
839
+ "뺭": 835,
840
+ "뺴": 836,
841
+ "뺼": 837,
842
+ "뻇": 838,
843
+ "뻐": 839,
844
+ "뻑": 840,
845
+ "뻔": 841,
846
+ "뻗": 842,
847
+ "뻘": 843,
848
+ "뻣": 844,
849
+ "뻤": 845,
850
+ "뻥": 846,
851
+ "뼈": 847,
852
+ "뼘": 848,
853
+ "뼝": 849,
854
+ "뽀": 850,
855
+ "뽁": 851,
856
+ "뽄": 852,
857
+ "뽑": 853,
858
+ "뽕": 854,
859
+ "뾰": 855,
860
+ "뾱": 856,
861
+ "뿅": 857,
862
+ "뿌": 858,
863
+ "뿍": 859,
864
+ "뿐": 860,
865
+ "뿔": 861,
866
+ "뿜": 862,
867
+ "뿡": 863,
868
+ "쁘": 864,
869
+ "쁜": 865,
870
+ "쁠": 866,
871
+ "쁨": 867,
872
+ "쁩": 868,
873
+ "삐": 869,
874
+ "삔": 870,
875
+ "삘": 871,
876
+ "삥": 872,
877
+ "사": 873,
878
+ "삭": 874,
879
+ "산": 875,
880
+ "살": 876,
881
+ "삶": 877,
882
+ "삼": 878,
883
+ "삽": 879,
884
+ "삿": 880,
885
+ "샀": 881,
886
+ "상": 882,
887
+ "새": 883,
888
+ "색": 884,
889
+ "샊": 885,
890
+ "샌": 886,
891
+ "샐": 887,
892
+ "샘": 888,
893
+ "샙": 889,
894
+ "샛": 890,
895
+ "샜": 891,
896
+ "생": 892,
897
+ "샤": 893,
898
+ "샥": 894,
899
+ "샨": 895,
900
+ "샬": 896,
901
+ "샴": 897,
902
+ "샵": 898,
903
+ "샷": 899,
904
+ "샹": 900,
905
+ "섀": 901,
906
+ "서": 902,
907
+ "석": 903,
908
+ "섞": 904,
909
+ "선": 905,
910
+ "섣": 906,
911
+ "설": 907,
912
+ "섬": 908,
913
+ "섭": 909,
914
+ "섯": 910,
915
+ "섰": 911,
916
+ "성": 912,
917
+ "세": 913,
918
+ "섹": 914,
919
+ "센": 915,
920
+ "셀": 916,
921
+ "셈": 917,
922
+ "셉": 918,
923
+ "셋": 919,
924
+ "셌": 920,
925
+ "셔": 921,
926
+ "션": 922,
927
+ "셜": 923,
928
+ "셥": 924,
929
+ "셧": 925,
930
+ "셨": 926,
931
+ "셰": 927,
932
+ "소": 928,
933
+ "속": 929,
934
+ "솎": 930,
935
+ "손": 931,
936
+ "솔": 932,
937
+ "솜": 933,
938
+ "솟": 934,
939
+ "송": 935,
940
+ "솥": 936,
941
+ "솨": 937,
942
+ "쇄": 938,
943
+ "쇠": 939,
944
+ "쇼": 940,
945
+ "쇽": 941,
946
+ "숀": 942,
947
+ "숌": 943,
948
+ "숍": 944,
949
+ "숏": 945,
950
+ "숑": 946,
951
+ "수": 947,
952
+ "숙": 948,
953
+ "순": 949,
954
+ "숟": 950,
955
+ "술": 951,
956
+ "숨": 952,
957
+ "숩": 953,
958
+ "숫": 954,
959
+ "숭": 955,
960
+ "숯": 956,
961
+ "숱": 957,
962
+ "숲": 958,
963
+ "숴": 959,
964
+ "쉐": 960,
965
+ "쉘": 961,
966
+ "쉣": 962,
967
+ "쉬": 963,
968
+ "쉭": 964,
969
+ "쉰": 965,
970
+ "쉴": 966,
971
+ "쉼": 967,
972
+ "쉽": 968,
973
+ "쉿": 969,
974
+ "슁": 970,
975
+ "슈": 971,
976
+ "슉": 972,
977
+ "슐": 973,
978
+ "슘": 974,
979
+ "슛": 975,
980
+ "슝": 976,
981
+ "스": 977,
982
+ "슥": 978,
983
+ "슨": 979,
984
+ "슬": 980,
985
+ "슴": 981,
986
+ "습": 982,
987
+ "슷": 983,
988
+ "승": 984,
989
+ "시": 985,
990
+ "식": 986,
991
+ "신": 987,
992
+ "싣": 988,
993
+ "실": 989,
994
+ "싫": 990,
995
+ "심": 991,
996
+ "십": 992,
997
+ "싯": 993,
998
+ "싱": 994,
999
+ "싴": 995,
1000
+ "싶": 996,
1001
+ "싸": 997,
1002
+ "싹": 998,
1003
+ "싼": 999,
1004
+ "쌀": 1000,
1005
+ "쌈": 1001,
1006
+ "쌉": 1002,
1007
+ "쌌": 1003,
1008
+ "쌍": 1004,
1009
+ "쌓": 1005,
1010
+ "쌔": 1006,
1011
+ "쌘": 1007,
1012
+ "쌤": 1008,
1013
+ "쌩": 1009,
1014
+ "쌰": 1010,
1015
+ "썜": 1011,
1016
+ "써": 1012,
1017
+ "썩": 1013,
1018
+ "썬": 1014,
1019
+ "썰": 1015,
1020
+ "썸": 1016,
1021
+ "썹": 1017,
1022
+ "썻": 1018,
1023
+ "썼": 1019,
1024
+ "썽": 1020,
1025
+ "쎄": 1021,
1026
+ "쎈": 1022,
1027
+ "쎌": 1023,
1028
+ "쎘": 1024,
1029
+ "쎼": 1025,
1030
+ "쏘": 1026,
1031
+ "쏙": 1027,
1032
+ "쏜": 1028,
1033
+ "쏟": 1029,
1034
+ "쏠": 1030,
1035
+ "쏩": 1031,
1036
+ "쏭": 1032,
1037
+ "쏴": 1033,
1038
+ "쏵": 1034,
1039
+ "쐈": 1035,
1040
+ "쐬": 1036,
1041
+ "쑈": 1037,
1042
+ "쑤": 1038,
1043
+ "쑥": 1039,
1044
+ "쒸": 1040,
1045
+ "쓰": 1041,
1046
+ "쓱": 1042,
1047
+ "쓴": 1043,
1048
+ "쓸": 1044,
1049
+ "씀": 1045,
1050
+ "씁": 1046,
1051
+ "씌": 1047,
1052
+ "씨": 1048,
1053
+ "씩": 1049,
1054
+ "씬": 1050,
1055
+ "씸": 1051,
1056
+ "씹": 1052,
1057
+ "씻": 1053,
1058
+ "씽": 1054,
1059
+ "아": 1055,
1060
+ "악": 1056,
1061
+ "안": 1057,
1062
+ "앉": 1058,
1063
+ "않": 1059,
1064
+ "알": 1060,
1065
+ "앓": 1061,
1066
+ "암": 1062,
1067
+ "압": 1063,
1068
+ "앗": 1064,
1069
+ "았": 1065,
1070
+ "앙": 1066,
1071
+ "앞": 1067,
1072
+ "애": 1068,
1073
+ "액": 1069,
1074
+ "앤": 1070,
1075
+ "앨": 1071,
1076
+ "앰": 1072,
1077
+ "앱": 1073,
1078
+ "앳": 1074,
1079
+ "앴": 1075,
1080
+ "앵": 1076,
1081
+ "야": 1077,
1082
+ "약": 1078,
1083
+ "얀": 1079,
1084
+ "얄": 1080,
1085
+ "얇": 1081,
1086
+ "얌": 1082,
1087
+ "얍": 1083,
1088
+ "양": 1084,
1089
+ "얕": 1085,
1090
+ "얗": 1086,
1091
+ "얘": 1087,
1092
+ "얜": 1088,
1093
+ "어": 1089,
1094
+ "억": 1090,
1095
+ "언": 1091,
1096
+ "얹": 1092,
1097
+ "얻": 1093,
1098
+ "얼": 1094,
1099
+ "얽": 1095,
1100
+ "엄": 1096,
1101
+ "업": 1097,
1102
+ "없": 1098,
1103
+ "엇": 1099,
1104
+ "었": 1100,
1105
+ "엉": 1101,
1106
+ "엊": 1102,
1107
+ "엌": 1103,
1108
+ "엎": 1104,
1109
+ "에": 1105,
1110
+ "엑": 1106,
1111
+ "엔": 1107,
1112
+ "엘": 1108,
1113
+ "엠": 1109,
1114
+ "엣": 1110,
1115
+ "엥": 1111,
1116
+ "여": 1112,
1117
+ "역": 1113,
1118
+ "엮": 1114,
1119
+ "연": 1115,
1120
+ "열": 1116,
1121
+ "염": 1117,
1122
+ "엽": 1118,
1123
+ "엿": 1119,
1124
+ "였": 1120,
1125
+ "영": 1121,
1126
+ "옅": 1122,
1127
+ "옆": 1123,
1128
+ "옇": 1124,
1129
+ "예": 1125,
1130
+ "옌": 1126,
1131
+ "옐": 1127,
1132
+ "옛": 1128,
1133
+ "오": 1129,
1134
+ "옥": 1130,
1135
+ "온": 1131,
1136
+ "올": 1132,
1137
+ "옮": 1133,
1138
+ "옳": 1134,
1139
+ "옴": 1135,
1140
+ "옵": 1136,
1141
+ "옷": 1137,
1142
+ "옹": 1138,
1143
+ "옻": 1139,
1144
+ "와": 1140,
1145
+ "왁": 1141,
1146
+ "완": 1142,
1147
+ "왈": 1143,
1148
+ "왓": 1144,
1149
+ "왔": 1145,
1150
+ "왕": 1146,
1151
+ "왜": 1147,
1152
+ "왠": 1148,
1153
+ "왤": 1149,
1154
+ "외": 1150,
1155
+ "왼": 1151,
1156
+ "욀": 1152,
1157
+ "요": 1153,
1158
+ "욕": 1154,
1159
+ "욘": 1155,
1160
+ "욜": 1156,
1161
+ "욤": 1157,
1162
+ "욥": 1158,
1163
+ "용": 1159,
1164
+ "우": 1160,
1165
+ "욱": 1161,
1166
+ "운": 1162,
1167
+ "울": 1163,
1168
+ "움": 1164,
1169
+ "웁": 1165,
1170
+ "웃": 1166,
1171
+ "웅": 1167,
1172
+ "워": 1168,
1173
+ "웍": 1169,
1174
+ "원": 1170,
1175
+ "월": 1171,
1176
+ "웜": 1172,
1177
+ "웠": 1173,
1178
+ "웨": 1174,
1179
+ "웩": 1175,
1180
+ "웬": 1176,
1181
+ "웰": 1177,
1182
+ "웹": 1178,
1183
+ "웻": 1179,
1184
+ "웽": 1180,
1185
+ "위": 1181,
1186
+ "윅": 1182,
1187
+ "윈": 1183,
1188
+ "윌": 1184,
1189
+ "윔": 1185,
1190
+ "윗": 1186,
1191
+ "윙": 1187,
1192
+ "유": 1188,
1193
+ "육": 1189,
1194
+ "윤": 1190,
1195
+ "율": 1191,
1196
+ "윱": 1192,
1197
+ "윳": 1193,
1198
+ "융": 1194,
1199
+ "윶": 1195,
1200
+ "윷": 1196,
1201
+ "으": 1197,
1202
+ "윽": 1198,
1203
+ "은": 1199,
1204
+ "읃": 1200,
1205
+ "을": 1201,
1206
+ "읊": 1202,
1207
+ "음": 1203,
1208
+ "읍": 1204,
1209
+ "읏": 1205,
1210
+ "응": 1206,
1211
+ "읒": 1207,
1212
+ "의": 1208,
1213
+ "이": 1209,
1214
+ "익": 1210,
1215
+ "인": 1211,
1216
+ "일": 1212,
1217
+ "읽": 1213,
1218
+ "잃": 1214,
1219
+ "임": 1215,
1220
+ "입": 1216,
1221
+ "잇": 1217,
1222
+ "있": 1218,
1223
+ "잉": 1219,
1224
+ "잊": 1220,
1225
+ "잌": 1221,
1226
+ "잎": 1222,
1227
+ "자": 1223,
1228
+ "작": 1224,
1229
+ "잔": 1225,
1230
+ "잖": 1226,
1231
+ "잘": 1227,
1232
+ "잠": 1228,
1233
+ "잡": 1229,
1234
+ "잣": 1230,
1235
+ "잤": 1231,
1236
+ "장": 1232,
1237
+ "잦": 1233,
1238
+ "재": 1234,
1239
+ "잭": 1235,
1240
+ "잰": 1236,
1241
+ "잴": 1237,
1242
+ "잼": 1238,
1243
+ "잽": 1239,
1244
+ "잿": 1240,
1245
+ "쟀": 1241,
1246
+ "쟁": 1242,
1247
+ "쟝": 1243,
1248
+ "쟤": 1244,
1249
+ "쟨": 1245,
1250
+ "저": 1246,
1251
+ "적": 1247,
1252
+ "전": 1248,
1253
+ "절": 1249,
1254
+ "젊": 1250,
1255
+ "점": 1251,
1256
+ "접": 1252,
1257
+ "젓": 1253,
1258
+ "정": 1254,
1259
+ "젖": 1255,
1260
+ "제": 1256,
1261
+ "젝": 1257,
1262
+ "젠": 1258,
1263
+ "젤": 1259,
1264
+ "젬": 1260,
1265
+ "젯": 1261,
1266
+ "져": 1262,
1267
+ "젼": 1263,
1268
+ "졌": 1264,
1269
+ "조": 1265,
1270
+ "족": 1266,
1271
+ "존": 1267,
1272
+ "졸": 1268,
1273
+ "좀": 1269,
1274
+ "좁": 1270,
1275
+ "좃": 1271,
1276
+ "종": 1272,
1277
+ "좆": 1273,
1278
+ "좋": 1274,
1279
+ "좌": 1275,
1280
+ "좍": 1276,
1281
+ "좔": 1277,
1282
+ "죄": 1278,
1283
+ "죠": 1279,
1284
+ "주": 1280,
1285
+ "죽": 1281,
1286
+ "준": 1282,
1287
+ "줄": 1283,
1288
+ "줌": 1284,
1289
+ "줍": 1285,
1290
+ "줏": 1286,
1291
+ "중": 1287,
1292
+ "줘": 1288,
1293
+ "줬": 1289,
1294
+ "줸": 1290,
1295
+ "쥐": 1291,
1296
+ "쥔": 1292,
1297
+ "쥘": 1293,
1298
+ "쥬": 1294,
1299
+ "즈": 1295,
1300
+ "즉": 1296,
1301
+ "즌": 1297,
1302
+ "즐": 1298,
1303
+ "즘": 1299,
1304
+ "즙": 1300,
1305
+ "증": 1301,
1306
+ "지": 1302,
1307
+ "직": 1303,
1308
+ "진": 1304,
1309
+ "질": 1305,
1310
+ "짊": 1306,
1311
+ "짐": 1307,
1312
+ "집": 1308,
1313
+ "짓": 1309,
1314
+ "징": 1310,
1315
+ "짖": 1311,
1316
+ "짚": 1312,
1317
+ "짜": 1313,
1318
+ "짝": 1314,
1319
+ "짠": 1315,
1320
+ "짤": 1316,
1321
+ "짥": 1317,
1322
+ "짧": 1318,
1323
+ "짬": 1319,
1324
+ "짭": 1320,
1325
+ "짯": 1321,
1326
+ "짰": 1322,
1327
+ "짱": 1323,
1328
+ "째": 1324,
1329
+ "짹": 1325,
1330
+ "짼": 1326,
1331
+ "쨋": 1327,
1332
+ "쨌": 1328,
1333
+ "쨍": 1329,
1334
+ "쨔": 1330,
1335
+ "쨰": 1331,
1336
+ "쩃": 1332,
1337
+ "쩄": 1333,
1338
+ "쩌": 1334,
1339
+ "쩍": 1335,
1340
+ "쩐": 1336,
1341
+ "쩔": 1337,
1342
+ "쩜": 1338,
1343
+ "쩝": 1339,
1344
+ "쩠": 1340,
1345
+ "쩡": 1341,
1346
+ "쩨": 1342,
1347
+ "쪄": 1343,
1348
+ "쪘": 1344,
1349
+ "쪼": 1345,
1350
+ "쪽": 1346,
1351
+ "쫀": 1347,
1352
+ "쫄": 1348,
1353
+ "쫌": 1349,
1354
+ "쫍": 1350,
1355
+ "쫑": 1351,
1356
+ "쫒": 1352,
1357
+ "쫓": 1353,
1358
+ "쫘": 1354,
1359
+ "쫙": 1355,
1360
+ "쬐": 1356,
1361
+ "쭈": 1357,
1362
+ "쭉": 1358,
1363
+ "쭌": 1359,
1364
+ "쭐": 1360,
1365
+ "쭘": 1361,
1366
+ "쭙": 1362,
1367
+ "쭝": 1363,
1368
+ "쭤": 1364,
1369
+ "쮸": 1365,
1370
+ "쯔": 1366,
1371
+ "쯕": 1367,
1372
+ "쯤": 1368,
1373
+ "쯧": 1369,
1374
+ "찌": 1370,
1375
+ "찍": 1371,
1376
+ "찐": 1372,
1377
+ "찔": 1373,
1378
+ "찜": 1374,
1379
+ "찝": 1375,
1380
+ "찡": 1376,
1381
+ "찢": 1377,
1382
+ "차": 1378,
1383
+ "착": 1379,
1384
+ "찬": 1380,
1385
+ "찮": 1381,
1386
+ "찰": 1382,
1387
+ "참": 1383,
1388
+ "찹": 1384,
1389
+ "찻": 1385,
1390
+ "찼": 1386,
1391
+ "창": 1387,
1392
+ "찾": 1388,
1393
+ "채": 1389,
1394
+ "책": 1390,
1395
+ "챌": 1391,
1396
+ "챔": 1392,
1397
+ "챕": 1393,
1398
+ "챗": 1394,
1399
+ "챘": 1395,
1400
+ "챙": 1396,
1401
+ "챠": 1397,
1402
+ "챴": 1398,
1403
+ "처": 1399,
1404
+ "척": 1400,
1405
+ "천": 1401,
1406
+ "철": 1402,
1407
+ "첨": 1403,
1408
+ "첩": 1404,
1409
+ "첫": 1405,
1410
+ "청": 1406,
1411
+ "체": 1407,
1412
+ "첵": 1408,
1413
+ "첸": 1409,
1414
+ "첼": 1410,
1415
+ "쳅": 1411,
1416
+ "쳇": 1412,
1417
+ "쳐": 1413,
1418
+ "쳤": 1414,
1419
+ "쳬": 1415,
1420
+ "초": 1416,
1421
+ "촉": 1417,
1422
+ "촌": 1418,
1423
+ "촐": 1419,
1424
+ "촘": 1420,
1425
+ "촛": 1421,
1426
+ "총": 1422,
1427
+ "촤": 1423,
1428
+ "촥": 1424,
1429
+ "촬": 1425,
1430
+ "최": 1426,
1431
+ "쵸": 1427,
1432
+ "추": 1428,
1433
+ "축": 1429,
1434
+ "춘": 1430,
1435
+ "출": 1431,
1436
+ "춤": 1432,
1437
+ "춥": 1433,
1438
+ "춧": 1434,
1439
+ "충": 1435,
1440
+ "춰": 1436,
1441
+ "춱": 1437,
1442
+ "췄": 1438,
1443
+ "췌": 1439,
1444
+ "취": 1440,
1445
+ "췬": 1441,
1446
+ "츄": 1442,
1447
+ "츈": 1443,
1448
+ "츠": 1444,
1449
+ "측": 1445,
1450
+ "츰": 1446,
1451
+ "층": 1447,
1452
+ "치": 1448,
1453
+ "칙": 1449,
1454
+ "친": 1450,
1455
+ "칠": 1451,
1456
+ "칡": 1452,
1457
+ "침": 1453,
1458
+ "칩": 1454,
1459
+ "칫": 1455,
1460
+ "칭": 1456,
1461
+ "카": 1457,
1462
+ "칵": 1458,
1463
+ "칸": 1459,
1464
+ "칼": 1460,
1465
+ "캄": 1461,
1466
+ "캅": 1462,
1467
+ "캇": 1463,
1468
+ "캉": 1464,
1469
+ "캐": 1465,
1470
+ "캔": 1466,
1471
+ "캘": 1467,
1472
+ "캠": 1468,
1473
+ "캡": 1469,
1474
+ "캣": 1470,
1475
+ "캤": 1471,
1476
+ "캥": 1472,
1477
+ "캬": 1473,
1478
+ "커": 1474,
1479
+ "컥": 1475,
1480
+ "컨": 1476,
1481
+ "컬": 1477,
1482
+ "컴": 1478,
1483
+ "컵": 1479,
1484
+ "컷": 1480,
1485
+ "컸": 1481,
1486
+ "케": 1482,
1487
+ "켁": 1483,
1488
+ "켄": 1484,
1489
+ "켈": 1485,
1490
+ "켓": 1486,
1491
+ "켜": 1487,
1492
+ "켠": 1488,
1493
+ "켤": 1489,
1494
+ "켰": 1490,
1495
+ "코": 1491,
1496
+ "콕": 1492,
1497
+ "콘": 1493,
1498
+ "콜": 1494,
1499
+ "콤": 1495,
1500
+ "콥": 1496,
1501
+ "콧": 1497,
1502
+ "콩": 1498,
1503
+ "콱": 1499,
1504
+ "콸": 1500,
1505
+ "쾅": 1501,
1506
+ "쾌": 1502,
1507
+ "쾨": 1503,
1508
+ "쿄": 1504,
1509
+ "쿠": 1505,
1510
+ "쿡": 1506,
1511
+ "쿤": 1507,
1512
+ "쿨": 1508,
1513
+ "쿰": 1509,
1514
+ "쿱": 1510,
1515
+ "쿵": 1511,
1516
+ "쿼": 1512,
1517
+ "퀀": 1513,
1518
+ "퀄": 1514,
1519
+ "퀘": 1515,
1520
+ "퀭": 1516,
1521
+ "퀴": 1517,
1522
+ "퀵": 1518,
1523
+ "퀸": 1519,
1524
+ "퀼": 1520,
1525
+ "큐": 1521,
1526
+ "큘": 1522,
1527
+ "크": 1523,
1528
+ "큰": 1524,
1529
+ "클": 1525,
1530
+ "큼": 1526,
1531
+ "큽": 1527,
1532
+ "킁": 1528,
1533
+ "키": 1529,
1534
+ "킥": 1530,
1535
+ "킨": 1531,
1536
+ "킬": 1532,
1537
+ "킴": 1533,
1538
+ "킵": 1534,
1539
+ "킷": 1535,
1540
+ "킹": 1536,
1541
+ "타": 1537,
1542
+ "탁": 1538,
1543
+ "탄": 1539,
1544
+ "탈": 1540,
1545
+ "탐": 1541,
1546
+ "탑": 1542,
1547
+ "탓": 1543,
1548
+ "탔": 1544,
1549
+ "탕": 1545,
1550
+ "태": 1546,
1551
+ "택": 1547,
1552
+ "탠": 1548,
1553
+ "탤": 1549,
1554
+ "탬": 1550,
1555
+ "탭": 1551,
1556
+ "탯": 1552,
1557
+ "탰": 1553,
1558
+ "탱": 1554,
1559
+ "터": 1555,
1560
+ "턱": 1556,
1561
+ "턴": 1557,
1562
+ "털": 1558,
1563
+ "텀": 1559,
1564
+ "텁": 1560,
1565
+ "텃": 1561,
1566
+ "텅": 1562,
1567
+ "테": 1563,
1568
+ "텍": 1564,
1569
+ "텐": 1565,
1570
+ "텔": 1566,
1571
+ "템": 1567,
1572
+ "텝": 1568,
1573
+ "텟": 1569,
1574
+ "텡": 1570,
1575
+ "텨": 1571,
1576
+ "텼": 1572,
1577
+ "토": 1573,
1578
+ "톡": 1574,
1579
+ "톤": 1575,
1580
+ "톨": 1576,
1581
+ "톰": 1577,
1582
+ "톱": 1578,
1583
+ "톳": 1579,
1584
+ "통": 1580,
1585
+ "퇴": 1581,
1586
+ "투": 1582,
1587
+ "툭": 1583,
1588
+ "툰": 1584,
1589
+ "툴": 1585,
1590
+ "툼": 1586,
1591
+ "퉁": 1587,
1592
+ "튀": 1588,
1593
+ "튄": 1589,
1594
+ "튕": 1590,
1595
+ "튜": 1591,
1596
+ "튠": 1592,
1597
+ "튤": 1593,
1598
+ "튱": 1594,
1599
+ "트": 1595,
1600
+ "특": 1596,
1601
+ "튼": 1597,
1602
+ "튿": 1598,
1603
+ "틀": 1599,
1604
+ "틈": 1600,
1605
+ "틋": 1601,
1606
+ "틍": 1602,
1607
+ "티": 1603,
1608
+ "틱": 1604,
1609
+ "틴": 1605,
1610
+ "틸": 1606,
1611
+ "팀": 1607,
1612
+ "팁": 1608,
1613
+ "팅": 1609,
1614
+ "파": 1610,
1615
+ "팍": 1611,
1616
+ "팎": 1612,
1617
+ "판": 1613,
1618
+ "팔": 1614,
1619
+ "팜": 1615,
1620
+ "팝": 1616,
1621
+ "팟": 1617,
1622
+ "팠": 1618,
1623
+ "팡": 1619,
1624
+ "팥": 1620,
1625
+ "패": 1621,
1626
+ "팩": 1622,
1627
+ "팬": 1623,
1628
+ "팰": 1624,
1629
+ "팸": 1625,
1630
+ "팻": 1626,
1631
+ "팼": 1627,
1632
+ "팽": 1628,
1633
+ "퍼": 1629,
1634
+ "퍽": 1630,
1635
+ "펀": 1631,
1636
+ "펄": 1632,
1637
+ "펌": 1633,
1638
+ "펍": 1634,
1639
+ "펐": 1635,
1640
+ "펑": 1636,
1641
+ "페": 1637,
1642
+ "펙": 1638,
1643
+ "펜": 1639,
1644
+ "펠": 1640,
1645
+ "펩": 1641,
1646
+ "펫": 1642,
1647
+ "펭": 1643,
1648
+ "펴": 1644,
1649
+ "편": 1645,
1650
+ "펼": 1646,
1651
+ "폄": 1647,
1652
+ "폈": 1648,
1653
+ "평": 1649,
1654
+ "폐": 1650,
1655
+ "포": 1651,
1656
+ "폭": 1652,
1657
+ "폰": 1653,
1658
+ "폴": 1654,
1659
+ "폼": 1655,
1660
+ "폿": 1656,
1661
+ "퐁": 1657,
1662
+ "표": 1658,
1663
+ "푯": 1659,
1664
+ "푸": 1660,
1665
+ "푹": 1661,
1666
+ "푼": 1662,
1667
+ "풀": 1663,
1668
+ "품": 1664,
1669
+ "풋": 1665,
1670
+ "풍": 1666,
1671
+ "퓨": 1667,
1672
+ "퓰": 1668,
1673
+ "프": 1669,
1674
+ "픈": 1670,
1675
+ "플": 1671,
1676
+ "픔": 1672,
1677
+ "픕": 1673,
1678
+ "픗": 1674,
1679
+ "피": 1675,
1680
+ "픽": 1676,
1681
+ "핀": 1677,
1682
+ "필": 1678,
1683
+ "핌": 1679,
1684
+ "핍": 1680,
1685
+ "핏": 1681,
1686
+ "핑": 1682,
1687
+ "하": 1683,
1688
+ "학": 1684,
1689
+ "한": 1685,
1690
+ "할": 1686,
1691
+ "핡": 1687,
1692
+ "핥": 1688,
1693
+ "핧": 1689,
1694
+ "함": 1690,
1695
+ "합": 1691,
1696
+ "핫": 1692,
1697
+ "핬": 1693,
1698
+ "항": 1694,
1699
+ "해": 1695,
1700
+ "핵": 1696,
1701
+ "핸": 1697,
1702
+ "핼": 1698,
1703
+ "햄": 1699,
1704
+ "햅": 1700,
1705
+ "햇": 1701,
1706
+ "했": 1702,
1707
+ "행": 1703,
1708
+ "햍": 1704,
1709
+ "햐": 1705,
1710
+ "햝": 1706,
1711
+ "향": 1707,
1712
+ "헀": 1708,
1713
+ "허": 1709,
1714
+ "헉": 1710,
1715
+ "헌": 1711,
1716
+ "헐": 1712,
1717
+ "험": 1713,
1718
+ "헛": 1714,
1719
+ "헝": 1715,
1720
+ "헤": 1716,
1721
+ "헥": 1717,
1722
+ "헨": 1718,
1723
+ "헬": 1719,
1724
+ "헷": 1720,
1725
+ "헹": 1721,
1726
+ "혀": 1722,
1727
+ "혁": 1723,
1728
+ "현": 1724,
1729
+ "혈": 1725,
1730
+ "혐": 1726,
1731
+ "협": 1727,
1732
+ "혓": 1728,
1733
+ "혔": 1729,
1734
+ "형": 1730,
1735
+ "혜": 1731,
1736
+ "호": 1732,
1737
+ "혹": 1733,
1738
+ "혼": 1734,
1739
+ "홀": 1735,
1740
+ "홈": 1736,
1741
+ "홉": 1737,
1742
+ "홋": 1738,
1743
+ "홍": 1739,
1744
+ "홑": 1740,
1745
+ "화": 1741,
1746
+ "확": 1742,
1747
+ "환": 1743,
1748
+ "활": 1744,
1749
+ "홧": 1745,
1750
+ "황": 1746,
1751
+ "홰": 1747,
1752
+ "회": 1748,
1753
+ "획": 1749,
1754
+ "횔": 1750,
1755
+ "횟": 1751,
1756
+ "횡": 1752,
1757
+ "효": 1753,
1758
+ "후": 1754,
1759
+ "훅": 1755,
1760
+ "훈": 1756,
1761
+ "훌": 1757,
1762
+ "훑": 1758,
1763
+ "훔": 1759,
1764
+ "훗": 1760,
1765
+ "훙": 1761,
1766
+ "훡": 1762,
1767
+ "훨": 1763,
1768
+ "훼": 1764,
1769
+ "휘": 1765,
1770
+ "휙": 1766,
1771
+ "휠": 1767,
1772
+ "휩": 1768,
1773
+ "휴": 1769,
1774
+ "흉": 1770,
1775
+ "흐": 1771,
1776
+ "흑": 1772,
1777
+ "흔": 1773,
1778
+ "흘": 1774,
1779
+ "흙": 1775,
1780
+ "흠": 1776,
1781
+ "흡": 1777,
1782
+ "흣": 1778,
1783
+ "흥": 1779,
1784
+ "흩": 1780,
1785
+ "희": 1781,
1786
+ "흰": 1782,
1787
+ "히": 1783,
1788
+ "힌": 1784,
1789
+ "힐": 1785,
1790
+ "힘": 1786,
1791
+ "힙": 1787
1792
+ }