fsicoli commited on
Commit
785607c
1 Parent(s): 6de378e

Upload folder using huggingface_hub (#3)

Browse files

- a9e3e1fe0bdb73baa1de20ba85c1e2663b473b7c840f5592a50e1b10723335da (a7bb8eb8108d58e25a9a04600798312af7be11cd)
- ff636a68aab72060c1f1b0de4c3388fc6cf5bca61148c649a2301a33d9647b98 (eb5651f96472474180a05ed349cce05c03d53797)
- 506bcd40e40f157b7ac9029fc6781fa47bfd058769b78122fbe246fb5cb0d65b (1abd3f00b6a7979b2d08c918bd632e14b683a06b)
- 75d0a6eadc47e39e045bed3a92f4b167230b04ff777c90488e5a622894df6c15 (7730173a690a8ccd9cd39b0476164b00b41571aa)
- 727abfb80cd7472f1e3e6ee06a5502a0cd6ffc1609cb50638a304dbdc8d86d26 (da4c9a52c733905457bd680cc0e341c0a88c6f44)
- 05f67ae9e22537d6f13f10989169fbee1be451c01ffd18b411a0db4c023b5fd4 (5f0a67b96867b67db538387c9806c8b00bbdbfce)
- 00db7ccc238197fb309448673d67751989b679f3087fee1e9412ba8ab55f8648 (438c64ea3c50300f42392f854965b0d50cffb2d4)
- 04e38f9271ac6629ea37d2435dbb6cda94d5b10ad957b2e880cf1fb35fbf16be (7a23cd5df1934121cfc54f21977c2d3c2757bda0)
- 3774e99c218029955d066182062a139e2ea1af85dfb577a5f2d456579aaeeead (9af935a62fb67453e6a8da6152ad32bed1f012ac)
- 97a201a3a482857fa04573bc315443ac546c9885295531df98b3d46a0cfbfcb7 (2c8f015b0c781af64f9c99ad4b4351a37079a37b)
- 19fe2ba494d55ebc806356772d026842497997698565efe2c6e65a5b40a2d572 (98d6b9a09d4eb9d5b76e36a25078fd258e496a35)
- d43a5e69cd8114d6b646266452b3d0226ca850c9806a2018c94fabbe7070e81d (1f1d6d9cd5eeb7e0c67e8abfefefb76396a40c2a)
- f7c7b967f2bccbc5b3ac50e22c6321a1541f3aa5257597a541ecc235d0ec2b01 (b50e049487975d6e7f33224b0f6668582d05e52f)
- fcdeee75912c17ba55bb9117758839bf9103be58a487da5dfce504b1c8bfb30f (19be26ff99b7daa2867b301ec18311343dd27abd)
- 38cdd076421efe79688bbb3eec2f985d36427ff44f0013b8f5e684687513c5af (2afa10337d30e10bde60c71f132a53b942a48ac2)
- 437916f31f88890597dfd8c3b75f1625274c8ecf732ff04740052eba3919ec22 (00ffbb3a9f6930ac6c4c5a059cb1505edd05757c)
- 33d4cdefdfe5a65f442f0e42bc45d33d6627d35ca6548d7254f191626170f210 (cc173dcc72866ca90af943ec5e9dd1f3a485d515)

This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. README.md +80 -0
  2. all_results.json +14 -0
  3. checkpoint-1000/config.json +51 -0
  4. checkpoint-1000/generation_config.json +264 -0
  5. checkpoint-1000/model-00001-of-00002.safetensors +3 -0
  6. checkpoint-1000/model-00002-of-00002.safetensors +3 -0
  7. checkpoint-1000/model.safetensors.index.json +0 -0
  8. checkpoint-1000/optimizer.pt +3 -0
  9. checkpoint-1000/preprocessor_config.json +14 -0
  10. checkpoint-1000/rng_state.pth +0 -0
  11. checkpoint-1000/scheduler.pt +0 -0
  12. checkpoint-1000/trainer_state.json +270 -0
  13. checkpoint-1000/training_args.bin +0 -0
  14. checkpoint-2000/config.json +51 -0
  15. checkpoint-2000/generation_config.json +264 -0
  16. checkpoint-2000/model-00001-of-00002.safetensors +3 -0
  17. checkpoint-2000/model-00002-of-00002.safetensors +3 -0
  18. checkpoint-2000/model.safetensors.index.json +0 -0
  19. checkpoint-2000/optimizer.pt +3 -0
  20. checkpoint-2000/preprocessor_config.json +14 -0
  21. checkpoint-2000/rng_state.pth +0 -0
  22. checkpoint-2000/scheduler.pt +0 -0
  23. checkpoint-2000/trainer_state.json +519 -0
  24. checkpoint-2000/training_args.bin +0 -0
  25. checkpoint-3000/config.json +51 -0
  26. checkpoint-3000/generation_config.json +264 -0
  27. checkpoint-3000/model-00001-of-00002.safetensors +3 -0
  28. checkpoint-3000/model-00002-of-00002.safetensors +3 -0
  29. checkpoint-3000/model.safetensors.index.json +0 -0
  30. checkpoint-3000/optimizer.pt +3 -0
  31. checkpoint-3000/preprocessor_config.json +14 -0
  32. checkpoint-3000/rng_state.pth +0 -0
  33. checkpoint-3000/scheduler.pt +0 -0
  34. checkpoint-3000/trainer_state.json +768 -0
  35. checkpoint-3000/training_args.bin +0 -0
  36. checkpoint-4000/config.json +51 -0
  37. checkpoint-4000/generation_config.json +265 -0
  38. checkpoint-4000/model-00001-of-00002.safetensors +3 -0
  39. checkpoint-4000/model-00002-of-00002.safetensors +3 -0
  40. checkpoint-4000/model.safetensors.index.json +0 -0
  41. checkpoint-4000/optimizer.pt +3 -0
  42. checkpoint-4000/preprocessor_config.json +14 -0
  43. checkpoint-4000/rng_state.pth +0 -0
  44. checkpoint-4000/scheduler.pt +0 -0
  45. checkpoint-4000/trainer_state.json +1017 -0
  46. checkpoint-4000/training_args.bin +0 -0
  47. checkpoint-5000/config.json +51 -0
  48. checkpoint-5000/generation_config.json +265 -0
  49. checkpoint-5000/model-00001-of-00002.safetensors +3 -0
  50. checkpoint-5000/model-00002-of-00002.safetensors +3 -0
README.md ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: openai/whisper-large-v3
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - mozilla-foundation/common_voice_16_0
8
+ metrics:
9
+ - wer
10
+ model-index:
11
+ - name: whisper-large-v3-pt-cv16-cuda
12
+ results:
13
+ - task:
14
+ name: Automatic Speech Recognition
15
+ type: automatic-speech-recognition
16
+ dataset:
17
+ name: mozilla-foundation/common_voice_16_0 pt
18
+ type: mozilla-foundation/common_voice_16_0
19
+ split: None
20
+ args: pt
21
+ metrics:
22
+ - name: Wer
23
+ type: wer
24
+ value: 0.9998545572074984
25
+ ---
26
+
27
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
28
+ should probably proofread and complete it, then remove this comment. -->
29
+
30
+ # whisper-large-v3-pt-cv16-cuda
31
+
32
+ This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the mozilla-foundation/common_voice_16_0 pt dataset.
33
+ It achieves the following results on the evaluation set:
34
+ - Loss: 0.1325
35
+ - Wer: 0.9999
36
+
37
+ ## Model description
38
+
39
+ More information needed
40
+
41
+ ## Intended uses & limitations
42
+
43
+ More information needed
44
+
45
+ ## Training and evaluation data
46
+
47
+ More information needed
48
+
49
+ ## Training procedure
50
+
51
+ ### Training hyperparameters
52
+
53
+ The following hyperparameters were used during training:
54
+ - learning_rate: 1e-06
55
+ - train_batch_size: 8
56
+ - eval_batch_size: 8
57
+ - seed: 42
58
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
+ - lr_scheduler_type: linear
60
+ - lr_scheduler_warmup_steps: 2000
61
+ - training_steps: 5000
62
+ - mixed_precision_training: Native AMP
63
+
64
+ ### Training results
65
+
66
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
67
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
68
+ | 0.199 | 0.26 | 1000 | 0.1563 | 0.1124 |
69
+ | 0.1654 | 0.52 | 2000 | 0.1500 | 0.1052 |
70
+ | 0.1794 | 0.77 | 3000 | 0.1379 | 0.0997 |
71
+ | 0.0821 | 1.03 | 4000 | 0.1321 | 1.0007 |
72
+ | 0.1292 | 1.29 | 5000 | 0.1325 | 0.9999 |
73
+
74
+
75
+ ### Framework versions
76
+
77
+ - Transformers 4.37.0.dev0
78
+ - Pytorch 2.2.0.dev20231212
79
+ - Datasets 2.15.1.dev0
80
+ - Tokenizers 0.15.0
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.29,
3
+ "eval_loss": 0.132488414645195,
4
+ "eval_runtime": 171706.1904,
5
+ "eval_samples": 9399,
6
+ "eval_samples_per_second": 0.055,
7
+ "eval_steps_per_second": 0.007,
8
+ "eval_wer": 0.9998545572074984,
9
+ "train_loss": 0.05041759390830994,
10
+ "train_runtime": 1177508.0135,
11
+ "train_samples": 30998,
12
+ "train_samples_per_second": 0.034,
13
+ "train_steps_per_second": 0.004
14
+ }
checkpoint-1000/config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-large-v3",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": true,
6
+ "architectures": [
7
+ "WhisperForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": [
11
+ 220,
12
+ 50257
13
+ ],
14
+ "bos_token_id": 50257,
15
+ "classifier_proj_size": 256,
16
+ "d_model": 1280,
17
+ "decoder_attention_heads": 20,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 32,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 20,
24
+ "encoder_ffn_dim": 5120,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 32,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 32,
43
+ "num_mel_bins": 128,
44
+ "pad_token_id": 50256,
45
+ "scale_embedding": false,
46
+ "torch_dtype": "float32",
47
+ "transformers_version": "4.37.0.dev0",
48
+ "use_cache": true,
49
+ "use_weighted_layer_sum": false,
50
+ "vocab_size": 51866
51
+ }
checkpoint-1000/generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 7,
5
+ 0
6
+ ],
7
+ [
8
+ 10,
9
+ 17
10
+ ],
11
+ [
12
+ 12,
13
+ 18
14
+ ],
15
+ [
16
+ 13,
17
+ 12
18
+ ],
19
+ [
20
+ 16,
21
+ 1
22
+ ],
23
+ [
24
+ 17,
25
+ 14
26
+ ],
27
+ [
28
+ 19,
29
+ 11
30
+ ],
31
+ [
32
+ 21,
33
+ 4
34
+ ],
35
+ [
36
+ 24,
37
+ 1
38
+ ],
39
+ [
40
+ 25,
41
+ 6
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50360
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|yue|>": 50358,
162
+ "<|zh|>": 50260
163
+ },
164
+ "max_initial_timestamp_index": 1,
165
+ "max_length": 448,
166
+ "no_timestamps_token_id": 50364,
167
+ "pad_token_id": 50257,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50359,
254
+ 50360,
255
+ 50361,
256
+ 50362,
257
+ 50363
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50360,
261
+ "translate": 50359
262
+ },
263
+ "transformers_version": "4.37.0.dev0"
264
+ }
checkpoint-1000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14da2025731d949006200c34abfb6aa1be4cc2350ecc27dc6af559603a86225d
3
+ size 4993448880
checkpoint-1000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48b38b2d250388274dbab6a134b7ff4184cb3525599666efa7cdefed5c99d82c
3
+ size 1180663192
checkpoint-1000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8b7a6a42c8f342e4252296a6744e5aeb820e8d8de08a16c9dd10adc396c1293
3
+ size 12333660476
checkpoint-1000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 128,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-1000/rng_state.pth ADDED
Binary file (14.3 kB). View file
 
checkpoint-1000/scheduler.pt ADDED
Binary file (1.06 kB). View file
 
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,270 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.25806451612903225,
5
+ "eval_steps": 1000,
6
+ "global_step": 1000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 1.1499999999999999e-08,
14
+ "loss": 0.7822,
15
+ "step": 25
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 2.4e-08,
20
+ "loss": 1.3149,
21
+ "step": 50
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 3.6499999999999996e-08,
26
+ "loss": 0.7809,
27
+ "step": 75
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 4.9e-08,
32
+ "loss": 1.2916,
33
+ "step": 100
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 6.15e-08,
38
+ "loss": 0.7645,
39
+ "step": 125
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 7.399999999999999e-08,
44
+ "loss": 1.3379,
45
+ "step": 150
46
+ },
47
+ {
48
+ "epoch": 0.05,
49
+ "learning_rate": 8.649999999999999e-08,
50
+ "loss": 0.7722,
51
+ "step": 175
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 9.9e-08,
56
+ "loss": 1.281,
57
+ "step": 200
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 1.115e-07,
62
+ "loss": 0.6998,
63
+ "step": 225
64
+ },
65
+ {
66
+ "epoch": 0.06,
67
+ "learning_rate": 1.24e-07,
68
+ "loss": 1.1532,
69
+ "step": 250
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "learning_rate": 1.365e-07,
74
+ "loss": 0.583,
75
+ "step": 275
76
+ },
77
+ {
78
+ "epoch": 0.08,
79
+ "learning_rate": 1.49e-07,
80
+ "loss": 0.6693,
81
+ "step": 300
82
+ },
83
+ {
84
+ "epoch": 0.08,
85
+ "learning_rate": 1.615e-07,
86
+ "loss": 0.3395,
87
+ "step": 325
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "learning_rate": 1.7399999999999997e-07,
92
+ "loss": 0.3597,
93
+ "step": 350
94
+ },
95
+ {
96
+ "epoch": 0.1,
97
+ "learning_rate": 1.8649999999999998e-07,
98
+ "loss": 0.22,
99
+ "step": 375
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "learning_rate": 1.99e-07,
104
+ "loss": 0.3449,
105
+ "step": 400
106
+ },
107
+ {
108
+ "epoch": 0.11,
109
+ "learning_rate": 2.1149999999999998e-07,
110
+ "loss": 0.2123,
111
+ "step": 425
112
+ },
113
+ {
114
+ "epoch": 0.12,
115
+ "learning_rate": 2.24e-07,
116
+ "loss": 0.3362,
117
+ "step": 450
118
+ },
119
+ {
120
+ "epoch": 0.12,
121
+ "learning_rate": 2.3649999999999998e-07,
122
+ "loss": 0.175,
123
+ "step": 475
124
+ },
125
+ {
126
+ "epoch": 0.13,
127
+ "learning_rate": 2.4899999999999997e-07,
128
+ "loss": 0.3051,
129
+ "step": 500
130
+ },
131
+ {
132
+ "epoch": 0.14,
133
+ "learning_rate": 2.615e-07,
134
+ "loss": 0.176,
135
+ "step": 525
136
+ },
137
+ {
138
+ "epoch": 0.14,
139
+ "learning_rate": 2.74e-07,
140
+ "loss": 0.2517,
141
+ "step": 550
142
+ },
143
+ {
144
+ "epoch": 0.15,
145
+ "learning_rate": 2.865e-07,
146
+ "loss": 0.1863,
147
+ "step": 575
148
+ },
149
+ {
150
+ "epoch": 0.15,
151
+ "learning_rate": 2.9899999999999996e-07,
152
+ "loss": 0.2614,
153
+ "step": 600
154
+ },
155
+ {
156
+ "epoch": 0.16,
157
+ "learning_rate": 3.115e-07,
158
+ "loss": 0.1942,
159
+ "step": 625
160
+ },
161
+ {
162
+ "epoch": 0.17,
163
+ "learning_rate": 3.24e-07,
164
+ "loss": 0.2901,
165
+ "step": 650
166
+ },
167
+ {
168
+ "epoch": 0.17,
169
+ "learning_rate": 3.3650000000000003e-07,
170
+ "loss": 0.1587,
171
+ "step": 675
172
+ },
173
+ {
174
+ "epoch": 0.18,
175
+ "learning_rate": 3.4899999999999996e-07,
176
+ "loss": 0.2128,
177
+ "step": 700
178
+ },
179
+ {
180
+ "epoch": 0.19,
181
+ "learning_rate": 3.6149999999999995e-07,
182
+ "loss": 0.1647,
183
+ "step": 725
184
+ },
185
+ {
186
+ "epoch": 0.19,
187
+ "learning_rate": 3.74e-07,
188
+ "loss": 0.2473,
189
+ "step": 750
190
+ },
191
+ {
192
+ "epoch": 0.2,
193
+ "learning_rate": 3.8649999999999997e-07,
194
+ "loss": 0.1415,
195
+ "step": 775
196
+ },
197
+ {
198
+ "epoch": 0.21,
199
+ "learning_rate": 3.99e-07,
200
+ "loss": 0.2367,
201
+ "step": 800
202
+ },
203
+ {
204
+ "epoch": 0.21,
205
+ "learning_rate": 4.1149999999999995e-07,
206
+ "loss": 0.1498,
207
+ "step": 825
208
+ },
209
+ {
210
+ "epoch": 0.22,
211
+ "learning_rate": 4.24e-07,
212
+ "loss": 0.2161,
213
+ "step": 850
214
+ },
215
+ {
216
+ "epoch": 0.23,
217
+ "learning_rate": 4.3649999999999997e-07,
218
+ "loss": 0.144,
219
+ "step": 875
220
+ },
221
+ {
222
+ "epoch": 0.23,
223
+ "learning_rate": 4.49e-07,
224
+ "loss": 0.2316,
225
+ "step": 900
226
+ },
227
+ {
228
+ "epoch": 0.24,
229
+ "learning_rate": 4.615e-07,
230
+ "loss": 0.1449,
231
+ "step": 925
232
+ },
233
+ {
234
+ "epoch": 0.25,
235
+ "learning_rate": 4.7399999999999993e-07,
236
+ "loss": 0.1513,
237
+ "step": 950
238
+ },
239
+ {
240
+ "epoch": 0.25,
241
+ "learning_rate": 4.864999999999999e-07,
242
+ "loss": 0.1855,
243
+ "step": 975
244
+ },
245
+ {
246
+ "epoch": 0.26,
247
+ "learning_rate": 4.99e-07,
248
+ "loss": 0.199,
249
+ "step": 1000
250
+ },
251
+ {
252
+ "epoch": 0.26,
253
+ "eval_loss": 0.15628743171691895,
254
+ "eval_runtime": 184706.1006,
255
+ "eval_samples_per_second": 0.051,
256
+ "eval_steps_per_second": 0.006,
257
+ "eval_wer": 0.1124272786037492,
258
+ "step": 1000
259
+ }
260
+ ],
261
+ "logging_steps": 25,
262
+ "max_steps": 5000,
263
+ "num_input_tokens_seen": 0,
264
+ "num_train_epochs": 2,
265
+ "save_steps": 1000,
266
+ "total_flos": 2.717998645248e+19,
267
+ "train_batch_size": 8,
268
+ "trial_name": null,
269
+ "trial_params": null
270
+ }
checkpoint-1000/training_args.bin ADDED
Binary file (4.86 kB). View file
 
checkpoint-2000/config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-large-v3",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": true,
6
+ "architectures": [
7
+ "WhisperForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": [
11
+ 220,
12
+ 50257
13
+ ],
14
+ "bos_token_id": 50257,
15
+ "classifier_proj_size": 256,
16
+ "d_model": 1280,
17
+ "decoder_attention_heads": 20,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 32,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 20,
24
+ "encoder_ffn_dim": 5120,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 32,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 32,
43
+ "num_mel_bins": 128,
44
+ "pad_token_id": 50256,
45
+ "scale_embedding": false,
46
+ "torch_dtype": "float32",
47
+ "transformers_version": "4.37.0.dev0",
48
+ "use_cache": true,
49
+ "use_weighted_layer_sum": false,
50
+ "vocab_size": 51866
51
+ }
checkpoint-2000/generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 7,
5
+ 0
6
+ ],
7
+ [
8
+ 10,
9
+ 17
10
+ ],
11
+ [
12
+ 12,
13
+ 18
14
+ ],
15
+ [
16
+ 13,
17
+ 12
18
+ ],
19
+ [
20
+ 16,
21
+ 1
22
+ ],
23
+ [
24
+ 17,
25
+ 14
26
+ ],
27
+ [
28
+ 19,
29
+ 11
30
+ ],
31
+ [
32
+ 21,
33
+ 4
34
+ ],
35
+ [
36
+ 24,
37
+ 1
38
+ ],
39
+ [
40
+ 25,
41
+ 6
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50360
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|yue|>": 50358,
162
+ "<|zh|>": 50260
163
+ },
164
+ "max_initial_timestamp_index": 1,
165
+ "max_length": 448,
166
+ "no_timestamps_token_id": 50364,
167
+ "pad_token_id": 50257,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50359,
254
+ 50360,
255
+ 50361,
256
+ 50362,
257
+ 50363
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50360,
261
+ "translate": 50359
262
+ },
263
+ "transformers_version": "4.37.0.dev0"
264
+ }
checkpoint-2000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:156e0fd71ba8d6d1a308460a813997db7d653c1f820bd2d5995abd49692f14a1
3
+ size 4993448880
checkpoint-2000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c05a1c27b9a9a440db979127057f5f66ec41e211bf68bb2ea3a227c4403c8e60
3
+ size 1180663192
checkpoint-2000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4da10ec2e0bf47d4b44db41bb1bfc93e007a4873e0e2ff5dbc5e667cb0168b4f
3
+ size 12333660476
checkpoint-2000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 128,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-2000/rng_state.pth ADDED
Binary file (14.2 kB). View file
 
checkpoint-2000/scheduler.pt ADDED
Binary file (1.06 kB). View file
 
checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,519 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.5161290322580645,
5
+ "eval_steps": 1000,
6
+ "global_step": 2000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 1.1499999999999999e-08,
14
+ "loss": 0.7822,
15
+ "step": 25
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 2.4e-08,
20
+ "loss": 1.3149,
21
+ "step": 50
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 3.6499999999999996e-08,
26
+ "loss": 0.7809,
27
+ "step": 75
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 4.9e-08,
32
+ "loss": 1.2916,
33
+ "step": 100
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 6.15e-08,
38
+ "loss": 0.7645,
39
+ "step": 125
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 7.399999999999999e-08,
44
+ "loss": 1.3379,
45
+ "step": 150
46
+ },
47
+ {
48
+ "epoch": 0.05,
49
+ "learning_rate": 8.649999999999999e-08,
50
+ "loss": 0.7722,
51
+ "step": 175
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 9.9e-08,
56
+ "loss": 1.281,
57
+ "step": 200
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 1.115e-07,
62
+ "loss": 0.6998,
63
+ "step": 225
64
+ },
65
+ {
66
+ "epoch": 0.06,
67
+ "learning_rate": 1.24e-07,
68
+ "loss": 1.1532,
69
+ "step": 250
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "learning_rate": 1.365e-07,
74
+ "loss": 0.583,
75
+ "step": 275
76
+ },
77
+ {
78
+ "epoch": 0.08,
79
+ "learning_rate": 1.49e-07,
80
+ "loss": 0.6693,
81
+ "step": 300
82
+ },
83
+ {
84
+ "epoch": 0.08,
85
+ "learning_rate": 1.615e-07,
86
+ "loss": 0.3395,
87
+ "step": 325
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "learning_rate": 1.7399999999999997e-07,
92
+ "loss": 0.3597,
93
+ "step": 350
94
+ },
95
+ {
96
+ "epoch": 0.1,
97
+ "learning_rate": 1.8649999999999998e-07,
98
+ "loss": 0.22,
99
+ "step": 375
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "learning_rate": 1.99e-07,
104
+ "loss": 0.3449,
105
+ "step": 400
106
+ },
107
+ {
108
+ "epoch": 0.11,
109
+ "learning_rate": 2.1149999999999998e-07,
110
+ "loss": 0.2123,
111
+ "step": 425
112
+ },
113
+ {
114
+ "epoch": 0.12,
115
+ "learning_rate": 2.24e-07,
116
+ "loss": 0.3362,
117
+ "step": 450
118
+ },
119
+ {
120
+ "epoch": 0.12,
121
+ "learning_rate": 2.3649999999999998e-07,
122
+ "loss": 0.175,
123
+ "step": 475
124
+ },
125
+ {
126
+ "epoch": 0.13,
127
+ "learning_rate": 2.4899999999999997e-07,
128
+ "loss": 0.3051,
129
+ "step": 500
130
+ },
131
+ {
132
+ "epoch": 0.14,
133
+ "learning_rate": 2.615e-07,
134
+ "loss": 0.176,
135
+ "step": 525
136
+ },
137
+ {
138
+ "epoch": 0.14,
139
+ "learning_rate": 2.74e-07,
140
+ "loss": 0.2517,
141
+ "step": 550
142
+ },
143
+ {
144
+ "epoch": 0.15,
145
+ "learning_rate": 2.865e-07,
146
+ "loss": 0.1863,
147
+ "step": 575
148
+ },
149
+ {
150
+ "epoch": 0.15,
151
+ "learning_rate": 2.9899999999999996e-07,
152
+ "loss": 0.2614,
153
+ "step": 600
154
+ },
155
+ {
156
+ "epoch": 0.16,
157
+ "learning_rate": 3.115e-07,
158
+ "loss": 0.1942,
159
+ "step": 625
160
+ },
161
+ {
162
+ "epoch": 0.17,
163
+ "learning_rate": 3.24e-07,
164
+ "loss": 0.2901,
165
+ "step": 650
166
+ },
167
+ {
168
+ "epoch": 0.17,
169
+ "learning_rate": 3.3650000000000003e-07,
170
+ "loss": 0.1587,
171
+ "step": 675
172
+ },
173
+ {
174
+ "epoch": 0.18,
175
+ "learning_rate": 3.4899999999999996e-07,
176
+ "loss": 0.2128,
177
+ "step": 700
178
+ },
179
+ {
180
+ "epoch": 0.19,
181
+ "learning_rate": 3.6149999999999995e-07,
182
+ "loss": 0.1647,
183
+ "step": 725
184
+ },
185
+ {
186
+ "epoch": 0.19,
187
+ "learning_rate": 3.74e-07,
188
+ "loss": 0.2473,
189
+ "step": 750
190
+ },
191
+ {
192
+ "epoch": 0.2,
193
+ "learning_rate": 3.8649999999999997e-07,
194
+ "loss": 0.1415,
195
+ "step": 775
196
+ },
197
+ {
198
+ "epoch": 0.21,
199
+ "learning_rate": 3.99e-07,
200
+ "loss": 0.2367,
201
+ "step": 800
202
+ },
203
+ {
204
+ "epoch": 0.21,
205
+ "learning_rate": 4.1149999999999995e-07,
206
+ "loss": 0.1498,
207
+ "step": 825
208
+ },
209
+ {
210
+ "epoch": 0.22,
211
+ "learning_rate": 4.24e-07,
212
+ "loss": 0.2161,
213
+ "step": 850
214
+ },
215
+ {
216
+ "epoch": 0.23,
217
+ "learning_rate": 4.3649999999999997e-07,
218
+ "loss": 0.144,
219
+ "step": 875
220
+ },
221
+ {
222
+ "epoch": 0.23,
223
+ "learning_rate": 4.49e-07,
224
+ "loss": 0.2316,
225
+ "step": 900
226
+ },
227
+ {
228
+ "epoch": 0.24,
229
+ "learning_rate": 4.615e-07,
230
+ "loss": 0.1449,
231
+ "step": 925
232
+ },
233
+ {
234
+ "epoch": 0.25,
235
+ "learning_rate": 4.7399999999999993e-07,
236
+ "loss": 0.1513,
237
+ "step": 950
238
+ },
239
+ {
240
+ "epoch": 0.25,
241
+ "learning_rate": 4.864999999999999e-07,
242
+ "loss": 0.1855,
243
+ "step": 975
244
+ },
245
+ {
246
+ "epoch": 0.26,
247
+ "learning_rate": 4.99e-07,
248
+ "loss": 0.199,
249
+ "step": 1000
250
+ },
251
+ {
252
+ "epoch": 0.26,
253
+ "eval_loss": 0.15628743171691895,
254
+ "eval_runtime": 184706.1006,
255
+ "eval_samples_per_second": 0.051,
256
+ "eval_steps_per_second": 0.006,
257
+ "eval_wer": 0.1124272786037492,
258
+ "step": 1000
259
+ },
260
+ {
261
+ "epoch": 0.26,
262
+ "learning_rate": 5.114999999999999e-07,
263
+ "loss": 0.1301,
264
+ "step": 1025
265
+ },
266
+ {
267
+ "epoch": 0.27,
268
+ "learning_rate": 5.24e-07,
269
+ "loss": 0.2193,
270
+ "step": 1050
271
+ },
272
+ {
273
+ "epoch": 0.28,
274
+ "learning_rate": 5.365e-07,
275
+ "loss": 0.1552,
276
+ "step": 1075
277
+ },
278
+ {
279
+ "epoch": 0.28,
280
+ "learning_rate": 5.490000000000001e-07,
281
+ "loss": 0.1865,
282
+ "step": 1100
283
+ },
284
+ {
285
+ "epoch": 0.29,
286
+ "learning_rate": 5.614999999999999e-07,
287
+ "loss": 0.1618,
288
+ "step": 1125
289
+ },
290
+ {
291
+ "epoch": 0.3,
292
+ "learning_rate": 5.739999999999999e-07,
293
+ "loss": 0.2259,
294
+ "step": 1150
295
+ },
296
+ {
297
+ "epoch": 0.3,
298
+ "learning_rate": 5.865e-07,
299
+ "loss": 0.1418,
300
+ "step": 1175
301
+ },
302
+ {
303
+ "epoch": 0.31,
304
+ "learning_rate": 5.989999999999999e-07,
305
+ "loss": 0.1916,
306
+ "step": 1200
307
+ },
308
+ {
309
+ "epoch": 0.32,
310
+ "learning_rate": 6.115e-07,
311
+ "loss": 0.1295,
312
+ "step": 1225
313
+ },
314
+ {
315
+ "epoch": 0.32,
316
+ "learning_rate": 6.24e-07,
317
+ "loss": 0.2112,
318
+ "step": 1250
319
+ },
320
+ {
321
+ "epoch": 0.33,
322
+ "learning_rate": 6.365e-07,
323
+ "loss": 0.1231,
324
+ "step": 1275
325
+ },
326
+ {
327
+ "epoch": 0.34,
328
+ "learning_rate": 6.49e-07,
329
+ "loss": 0.1914,
330
+ "step": 1300
331
+ },
332
+ {
333
+ "epoch": 0.34,
334
+ "learning_rate": 6.614999999999999e-07,
335
+ "loss": 0.1485,
336
+ "step": 1325
337
+ },
338
+ {
339
+ "epoch": 0.35,
340
+ "learning_rate": 6.74e-07,
341
+ "loss": 0.1958,
342
+ "step": 1350
343
+ },
344
+ {
345
+ "epoch": 0.35,
346
+ "learning_rate": 6.865e-07,
347
+ "loss": 0.1452,
348
+ "step": 1375
349
+ },
350
+ {
351
+ "epoch": 0.36,
352
+ "learning_rate": 6.989999999999999e-07,
353
+ "loss": 0.1624,
354
+ "step": 1400
355
+ },
356
+ {
357
+ "epoch": 0.37,
358
+ "learning_rate": 7.115e-07,
359
+ "loss": 0.1518,
360
+ "step": 1425
361
+ },
362
+ {
363
+ "epoch": 0.37,
364
+ "learning_rate": 7.24e-07,
365
+ "loss": 0.1935,
366
+ "step": 1450
367
+ },
368
+ {
369
+ "epoch": 0.38,
370
+ "learning_rate": 7.365e-07,
371
+ "loss": 0.1138,
372
+ "step": 1475
373
+ },
374
+ {
375
+ "epoch": 0.39,
376
+ "learning_rate": 7.489999999999999e-07,
377
+ "loss": 0.16,
378
+ "step": 1500
379
+ },
380
+ {
381
+ "epoch": 0.39,
382
+ "learning_rate": 7.614999999999999e-07,
383
+ "loss": 0.1279,
384
+ "step": 1525
385
+ },
386
+ {
387
+ "epoch": 0.4,
388
+ "learning_rate": 7.74e-07,
389
+ "loss": 0.1862,
390
+ "step": 1550
391
+ },
392
+ {
393
+ "epoch": 0.41,
394
+ "learning_rate": 7.864999999999999e-07,
395
+ "loss": 0.1537,
396
+ "step": 1575
397
+ },
398
+ {
399
+ "epoch": 0.41,
400
+ "learning_rate": 7.99e-07,
401
+ "loss": 0.1915,
402
+ "step": 1600
403
+ },
404
+ {
405
+ "epoch": 0.42,
406
+ "learning_rate": 8.115e-07,
407
+ "loss": 0.1303,
408
+ "step": 1625
409
+ },
410
+ {
411
+ "epoch": 0.43,
412
+ "learning_rate": 8.24e-07,
413
+ "loss": 0.1967,
414
+ "step": 1650
415
+ },
416
+ {
417
+ "epoch": 0.43,
418
+ "learning_rate": 8.365e-07,
419
+ "loss": 0.1418,
420
+ "step": 1675
421
+ },
422
+ {
423
+ "epoch": 0.44,
424
+ "learning_rate": 8.489999999999999e-07,
425
+ "loss": 0.194,
426
+ "step": 1700
427
+ },
428
+ {
429
+ "epoch": 0.45,
430
+ "learning_rate": 8.615e-07,
431
+ "loss": 0.1377,
432
+ "step": 1725
433
+ },
434
+ {
435
+ "epoch": 0.45,
436
+ "learning_rate": 8.739999999999999e-07,
437
+ "loss": 0.1534,
438
+ "step": 1750
439
+ },
440
+ {
441
+ "epoch": 0.46,
442
+ "learning_rate": 8.864999999999999e-07,
443
+ "loss": 0.1416,
444
+ "step": 1775
445
+ },
446
+ {
447
+ "epoch": 0.46,
448
+ "learning_rate": 8.99e-07,
449
+ "loss": 0.1869,
450
+ "step": 1800
451
+ },
452
+ {
453
+ "epoch": 0.47,
454
+ "learning_rate": 9.115e-07,
455
+ "loss": 0.1507,
456
+ "step": 1825
457
+ },
458
+ {
459
+ "epoch": 0.48,
460
+ "learning_rate": 9.24e-07,
461
+ "loss": 0.1571,
462
+ "step": 1850
463
+ },
464
+ {
465
+ "epoch": 0.48,
466
+ "learning_rate": 9.365e-07,
467
+ "loss": 0.1348,
468
+ "step": 1875
469
+ },
470
+ {
471
+ "epoch": 0.49,
472
+ "learning_rate": 9.489999999999999e-07,
473
+ "loss": 0.1981,
474
+ "step": 1900
475
+ },
476
+ {
477
+ "epoch": 0.5,
478
+ "learning_rate": 9.615e-07,
479
+ "loss": 0.1324,
480
+ "step": 1925
481
+ },
482
+ {
483
+ "epoch": 0.5,
484
+ "learning_rate": 9.74e-07,
485
+ "loss": 0.1712,
486
+ "step": 1950
487
+ },
488
+ {
489
+ "epoch": 0.51,
490
+ "learning_rate": 9.865e-07,
491
+ "loss": 0.1169,
492
+ "step": 1975
493
+ },
494
+ {
495
+ "epoch": 0.52,
496
+ "learning_rate": 9.989999999999999e-07,
497
+ "loss": 0.1654,
498
+ "step": 2000
499
+ },
500
+ {
501
+ "epoch": 0.52,
502
+ "eval_loss": 0.1500108540058136,
503
+ "eval_runtime": 188234.7285,
504
+ "eval_samples_per_second": 0.05,
505
+ "eval_steps_per_second": 0.006,
506
+ "eval_wer": 0.10521978021978022,
507
+ "step": 2000
508
+ }
509
+ ],
510
+ "logging_steps": 25,
511
+ "max_steps": 5000,
512
+ "num_input_tokens_seen": 0,
513
+ "num_train_epochs": 2,
514
+ "save_steps": 1000,
515
+ "total_flos": 5.435997290496e+19,
516
+ "train_batch_size": 8,
517
+ "trial_name": null,
518
+ "trial_params": null
519
+ }
checkpoint-2000/training_args.bin ADDED
Binary file (4.86 kB). View file
 
checkpoint-3000/config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-large-v3",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": true,
6
+ "architectures": [
7
+ "WhisperForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": [
11
+ 220,
12
+ 50257
13
+ ],
14
+ "bos_token_id": 50257,
15
+ "classifier_proj_size": 256,
16
+ "d_model": 1280,
17
+ "decoder_attention_heads": 20,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 32,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 20,
24
+ "encoder_ffn_dim": 5120,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 32,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 32,
43
+ "num_mel_bins": 128,
44
+ "pad_token_id": 50256,
45
+ "scale_embedding": false,
46
+ "torch_dtype": "float32",
47
+ "transformers_version": "4.37.0.dev0",
48
+ "use_cache": true,
49
+ "use_weighted_layer_sum": false,
50
+ "vocab_size": 51866
51
+ }
checkpoint-3000/generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 7,
5
+ 0
6
+ ],
7
+ [
8
+ 10,
9
+ 17
10
+ ],
11
+ [
12
+ 12,
13
+ 18
14
+ ],
15
+ [
16
+ 13,
17
+ 12
18
+ ],
19
+ [
20
+ 16,
21
+ 1
22
+ ],
23
+ [
24
+ 17,
25
+ 14
26
+ ],
27
+ [
28
+ 19,
29
+ 11
30
+ ],
31
+ [
32
+ 21,
33
+ 4
34
+ ],
35
+ [
36
+ 24,
37
+ 1
38
+ ],
39
+ [
40
+ 25,
41
+ 6
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50360
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|yue|>": 50358,
162
+ "<|zh|>": 50260
163
+ },
164
+ "max_initial_timestamp_index": 1,
165
+ "max_length": 448,
166
+ "no_timestamps_token_id": 50364,
167
+ "pad_token_id": 50257,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50359,
254
+ 50360,
255
+ 50361,
256
+ 50362,
257
+ 50363
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50360,
261
+ "translate": 50359
262
+ },
263
+ "transformers_version": "4.37.0.dev0"
264
+ }
checkpoint-3000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f878789c48bcae8bdb738a21db184b61ea25d207190b7a28b1886fd661820964
3
+ size 4993448880
checkpoint-3000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cebf488a1c5179e1630e18b09c98e0fc49f633401cd2860996acc46e8ce31123
3
+ size 1180663192
checkpoint-3000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b881cb7e883293384811c41e69a3af5ab73194ee3fd9c7fc959db40efbb8dce
3
+ size 12333660476
checkpoint-3000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 128,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-3000/rng_state.pth ADDED
Binary file (14.2 kB). View file
 
checkpoint-3000/scheduler.pt ADDED
Binary file (1.06 kB). View file
 
checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,768 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.7741935483870968,
5
+ "eval_steps": 1000,
6
+ "global_step": 3000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 1.1499999999999999e-08,
14
+ "loss": 0.7822,
15
+ "step": 25
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 2.4e-08,
20
+ "loss": 1.3149,
21
+ "step": 50
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 3.6499999999999996e-08,
26
+ "loss": 0.7809,
27
+ "step": 75
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 4.9e-08,
32
+ "loss": 1.2916,
33
+ "step": 100
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 6.15e-08,
38
+ "loss": 0.7645,
39
+ "step": 125
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 7.399999999999999e-08,
44
+ "loss": 1.3379,
45
+ "step": 150
46
+ },
47
+ {
48
+ "epoch": 0.05,
49
+ "learning_rate": 8.649999999999999e-08,
50
+ "loss": 0.7722,
51
+ "step": 175
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 9.9e-08,
56
+ "loss": 1.281,
57
+ "step": 200
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 1.115e-07,
62
+ "loss": 0.6998,
63
+ "step": 225
64
+ },
65
+ {
66
+ "epoch": 0.06,
67
+ "learning_rate": 1.24e-07,
68
+ "loss": 1.1532,
69
+ "step": 250
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "learning_rate": 1.365e-07,
74
+ "loss": 0.583,
75
+ "step": 275
76
+ },
77
+ {
78
+ "epoch": 0.08,
79
+ "learning_rate": 1.49e-07,
80
+ "loss": 0.6693,
81
+ "step": 300
82
+ },
83
+ {
84
+ "epoch": 0.08,
85
+ "learning_rate": 1.615e-07,
86
+ "loss": 0.3395,
87
+ "step": 325
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "learning_rate": 1.7399999999999997e-07,
92
+ "loss": 0.3597,
93
+ "step": 350
94
+ },
95
+ {
96
+ "epoch": 0.1,
97
+ "learning_rate": 1.8649999999999998e-07,
98
+ "loss": 0.22,
99
+ "step": 375
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "learning_rate": 1.99e-07,
104
+ "loss": 0.3449,
105
+ "step": 400
106
+ },
107
+ {
108
+ "epoch": 0.11,
109
+ "learning_rate": 2.1149999999999998e-07,
110
+ "loss": 0.2123,
111
+ "step": 425
112
+ },
113
+ {
114
+ "epoch": 0.12,
115
+ "learning_rate": 2.24e-07,
116
+ "loss": 0.3362,
117
+ "step": 450
118
+ },
119
+ {
120
+ "epoch": 0.12,
121
+ "learning_rate": 2.3649999999999998e-07,
122
+ "loss": 0.175,
123
+ "step": 475
124
+ },
125
+ {
126
+ "epoch": 0.13,
127
+ "learning_rate": 2.4899999999999997e-07,
128
+ "loss": 0.3051,
129
+ "step": 500
130
+ },
131
+ {
132
+ "epoch": 0.14,
133
+ "learning_rate": 2.615e-07,
134
+ "loss": 0.176,
135
+ "step": 525
136
+ },
137
+ {
138
+ "epoch": 0.14,
139
+ "learning_rate": 2.74e-07,
140
+ "loss": 0.2517,
141
+ "step": 550
142
+ },
143
+ {
144
+ "epoch": 0.15,
145
+ "learning_rate": 2.865e-07,
146
+ "loss": 0.1863,
147
+ "step": 575
148
+ },
149
+ {
150
+ "epoch": 0.15,
151
+ "learning_rate": 2.9899999999999996e-07,
152
+ "loss": 0.2614,
153
+ "step": 600
154
+ },
155
+ {
156
+ "epoch": 0.16,
157
+ "learning_rate": 3.115e-07,
158
+ "loss": 0.1942,
159
+ "step": 625
160
+ },
161
+ {
162
+ "epoch": 0.17,
163
+ "learning_rate": 3.24e-07,
164
+ "loss": 0.2901,
165
+ "step": 650
166
+ },
167
+ {
168
+ "epoch": 0.17,
169
+ "learning_rate": 3.3650000000000003e-07,
170
+ "loss": 0.1587,
171
+ "step": 675
172
+ },
173
+ {
174
+ "epoch": 0.18,
175
+ "learning_rate": 3.4899999999999996e-07,
176
+ "loss": 0.2128,
177
+ "step": 700
178
+ },
179
+ {
180
+ "epoch": 0.19,
181
+ "learning_rate": 3.6149999999999995e-07,
182
+ "loss": 0.1647,
183
+ "step": 725
184
+ },
185
+ {
186
+ "epoch": 0.19,
187
+ "learning_rate": 3.74e-07,
188
+ "loss": 0.2473,
189
+ "step": 750
190
+ },
191
+ {
192
+ "epoch": 0.2,
193
+ "learning_rate": 3.8649999999999997e-07,
194
+ "loss": 0.1415,
195
+ "step": 775
196
+ },
197
+ {
198
+ "epoch": 0.21,
199
+ "learning_rate": 3.99e-07,
200
+ "loss": 0.2367,
201
+ "step": 800
202
+ },
203
+ {
204
+ "epoch": 0.21,
205
+ "learning_rate": 4.1149999999999995e-07,
206
+ "loss": 0.1498,
207
+ "step": 825
208
+ },
209
+ {
210
+ "epoch": 0.22,
211
+ "learning_rate": 4.24e-07,
212
+ "loss": 0.2161,
213
+ "step": 850
214
+ },
215
+ {
216
+ "epoch": 0.23,
217
+ "learning_rate": 4.3649999999999997e-07,
218
+ "loss": 0.144,
219
+ "step": 875
220
+ },
221
+ {
222
+ "epoch": 0.23,
223
+ "learning_rate": 4.49e-07,
224
+ "loss": 0.2316,
225
+ "step": 900
226
+ },
227
+ {
228
+ "epoch": 0.24,
229
+ "learning_rate": 4.615e-07,
230
+ "loss": 0.1449,
231
+ "step": 925
232
+ },
233
+ {
234
+ "epoch": 0.25,
235
+ "learning_rate": 4.7399999999999993e-07,
236
+ "loss": 0.1513,
237
+ "step": 950
238
+ },
239
+ {
240
+ "epoch": 0.25,
241
+ "learning_rate": 4.864999999999999e-07,
242
+ "loss": 0.1855,
243
+ "step": 975
244
+ },
245
+ {
246
+ "epoch": 0.26,
247
+ "learning_rate": 4.99e-07,
248
+ "loss": 0.199,
249
+ "step": 1000
250
+ },
251
+ {
252
+ "epoch": 0.26,
253
+ "eval_loss": 0.15628743171691895,
254
+ "eval_runtime": 184706.1006,
255
+ "eval_samples_per_second": 0.051,
256
+ "eval_steps_per_second": 0.006,
257
+ "eval_wer": 0.1124272786037492,
258
+ "step": 1000
259
+ },
260
+ {
261
+ "epoch": 0.26,
262
+ "learning_rate": 5.114999999999999e-07,
263
+ "loss": 0.1301,
264
+ "step": 1025
265
+ },
266
+ {
267
+ "epoch": 0.27,
268
+ "learning_rate": 5.24e-07,
269
+ "loss": 0.2193,
270
+ "step": 1050
271
+ },
272
+ {
273
+ "epoch": 0.28,
274
+ "learning_rate": 5.365e-07,
275
+ "loss": 0.1552,
276
+ "step": 1075
277
+ },
278
+ {
279
+ "epoch": 0.28,
280
+ "learning_rate": 5.490000000000001e-07,
281
+ "loss": 0.1865,
282
+ "step": 1100
283
+ },
284
+ {
285
+ "epoch": 0.29,
286
+ "learning_rate": 5.614999999999999e-07,
287
+ "loss": 0.1618,
288
+ "step": 1125
289
+ },
290
+ {
291
+ "epoch": 0.3,
292
+ "learning_rate": 5.739999999999999e-07,
293
+ "loss": 0.2259,
294
+ "step": 1150
295
+ },
296
+ {
297
+ "epoch": 0.3,
298
+ "learning_rate": 5.865e-07,
299
+ "loss": 0.1418,
300
+ "step": 1175
301
+ },
302
+ {
303
+ "epoch": 0.31,
304
+ "learning_rate": 5.989999999999999e-07,
305
+ "loss": 0.1916,
306
+ "step": 1200
307
+ },
308
+ {
309
+ "epoch": 0.32,
310
+ "learning_rate": 6.115e-07,
311
+ "loss": 0.1295,
312
+ "step": 1225
313
+ },
314
+ {
315
+ "epoch": 0.32,
316
+ "learning_rate": 6.24e-07,
317
+ "loss": 0.2112,
318
+ "step": 1250
319
+ },
320
+ {
321
+ "epoch": 0.33,
322
+ "learning_rate": 6.365e-07,
323
+ "loss": 0.1231,
324
+ "step": 1275
325
+ },
326
+ {
327
+ "epoch": 0.34,
328
+ "learning_rate": 6.49e-07,
329
+ "loss": 0.1914,
330
+ "step": 1300
331
+ },
332
+ {
333
+ "epoch": 0.34,
334
+ "learning_rate": 6.614999999999999e-07,
335
+ "loss": 0.1485,
336
+ "step": 1325
337
+ },
338
+ {
339
+ "epoch": 0.35,
340
+ "learning_rate": 6.74e-07,
341
+ "loss": 0.1958,
342
+ "step": 1350
343
+ },
344
+ {
345
+ "epoch": 0.35,
346
+ "learning_rate": 6.865e-07,
347
+ "loss": 0.1452,
348
+ "step": 1375
349
+ },
350
+ {
351
+ "epoch": 0.36,
352
+ "learning_rate": 6.989999999999999e-07,
353
+ "loss": 0.1624,
354
+ "step": 1400
355
+ },
356
+ {
357
+ "epoch": 0.37,
358
+ "learning_rate": 7.115e-07,
359
+ "loss": 0.1518,
360
+ "step": 1425
361
+ },
362
+ {
363
+ "epoch": 0.37,
364
+ "learning_rate": 7.24e-07,
365
+ "loss": 0.1935,
366
+ "step": 1450
367
+ },
368
+ {
369
+ "epoch": 0.38,
370
+ "learning_rate": 7.365e-07,
371
+ "loss": 0.1138,
372
+ "step": 1475
373
+ },
374
+ {
375
+ "epoch": 0.39,
376
+ "learning_rate": 7.489999999999999e-07,
377
+ "loss": 0.16,
378
+ "step": 1500
379
+ },
380
+ {
381
+ "epoch": 0.39,
382
+ "learning_rate": 7.614999999999999e-07,
383
+ "loss": 0.1279,
384
+ "step": 1525
385
+ },
386
+ {
387
+ "epoch": 0.4,
388
+ "learning_rate": 7.74e-07,
389
+ "loss": 0.1862,
390
+ "step": 1550
391
+ },
392
+ {
393
+ "epoch": 0.41,
394
+ "learning_rate": 7.864999999999999e-07,
395
+ "loss": 0.1537,
396
+ "step": 1575
397
+ },
398
+ {
399
+ "epoch": 0.41,
400
+ "learning_rate": 7.99e-07,
401
+ "loss": 0.1915,
402
+ "step": 1600
403
+ },
404
+ {
405
+ "epoch": 0.42,
406
+ "learning_rate": 8.115e-07,
407
+ "loss": 0.1303,
408
+ "step": 1625
409
+ },
410
+ {
411
+ "epoch": 0.43,
412
+ "learning_rate": 8.24e-07,
413
+ "loss": 0.1967,
414
+ "step": 1650
415
+ },
416
+ {
417
+ "epoch": 0.43,
418
+ "learning_rate": 8.365e-07,
419
+ "loss": 0.1418,
420
+ "step": 1675
421
+ },
422
+ {
423
+ "epoch": 0.44,
424
+ "learning_rate": 8.489999999999999e-07,
425
+ "loss": 0.194,
426
+ "step": 1700
427
+ },
428
+ {
429
+ "epoch": 0.45,
430
+ "learning_rate": 8.615e-07,
431
+ "loss": 0.1377,
432
+ "step": 1725
433
+ },
434
+ {
435
+ "epoch": 0.45,
436
+ "learning_rate": 8.739999999999999e-07,
437
+ "loss": 0.1534,
438
+ "step": 1750
439
+ },
440
+ {
441
+ "epoch": 0.46,
442
+ "learning_rate": 8.864999999999999e-07,
443
+ "loss": 0.1416,
444
+ "step": 1775
445
+ },
446
+ {
447
+ "epoch": 0.46,
448
+ "learning_rate": 8.99e-07,
449
+ "loss": 0.1869,
450
+ "step": 1800
451
+ },
452
+ {
453
+ "epoch": 0.47,
454
+ "learning_rate": 9.115e-07,
455
+ "loss": 0.1507,
456
+ "step": 1825
457
+ },
458
+ {
459
+ "epoch": 0.48,
460
+ "learning_rate": 9.24e-07,
461
+ "loss": 0.1571,
462
+ "step": 1850
463
+ },
464
+ {
465
+ "epoch": 0.48,
466
+ "learning_rate": 9.365e-07,
467
+ "loss": 0.1348,
468
+ "step": 1875
469
+ },
470
+ {
471
+ "epoch": 0.49,
472
+ "learning_rate": 9.489999999999999e-07,
473
+ "loss": 0.1981,
474
+ "step": 1900
475
+ },
476
+ {
477
+ "epoch": 0.5,
478
+ "learning_rate": 9.615e-07,
479
+ "loss": 0.1324,
480
+ "step": 1925
481
+ },
482
+ {
483
+ "epoch": 0.5,
484
+ "learning_rate": 9.74e-07,
485
+ "loss": 0.1712,
486
+ "step": 1950
487
+ },
488
+ {
489
+ "epoch": 0.51,
490
+ "learning_rate": 9.865e-07,
491
+ "loss": 0.1169,
492
+ "step": 1975
493
+ },
494
+ {
495
+ "epoch": 0.52,
496
+ "learning_rate": 9.989999999999999e-07,
497
+ "loss": 0.1654,
498
+ "step": 2000
499
+ },
500
+ {
501
+ "epoch": 0.52,
502
+ "eval_loss": 0.1500108540058136,
503
+ "eval_runtime": 188234.7285,
504
+ "eval_samples_per_second": 0.05,
505
+ "eval_steps_per_second": 0.006,
506
+ "eval_wer": 0.10521978021978022,
507
+ "step": 2000
508
+ },
509
+ {
510
+ "epoch": 0.52,
511
+ "learning_rate": 9.923333333333332e-07,
512
+ "loss": 0.1202,
513
+ "step": 2025
514
+ },
515
+ {
516
+ "epoch": 0.53,
517
+ "learning_rate": 9.84e-07,
518
+ "loss": 0.186,
519
+ "step": 2050
520
+ },
521
+ {
522
+ "epoch": 0.54,
523
+ "learning_rate": 9.756666666666666e-07,
524
+ "loss": 0.1106,
525
+ "step": 2075
526
+ },
527
+ {
528
+ "epoch": 0.54,
529
+ "learning_rate": 9.673333333333332e-07,
530
+ "loss": 0.1926,
531
+ "step": 2100
532
+ },
533
+ {
534
+ "epoch": 0.55,
535
+ "learning_rate": 9.589999999999998e-07,
536
+ "loss": 0.1178,
537
+ "step": 2125
538
+ },
539
+ {
540
+ "epoch": 0.55,
541
+ "learning_rate": 9.506666666666667e-07,
542
+ "loss": 0.1744,
543
+ "step": 2150
544
+ },
545
+ {
546
+ "epoch": 0.56,
547
+ "learning_rate": 9.423333333333333e-07,
548
+ "loss": 0.1263,
549
+ "step": 2175
550
+ },
551
+ {
552
+ "epoch": 0.57,
553
+ "learning_rate": 9.34e-07,
554
+ "loss": 0.1818,
555
+ "step": 2200
556
+ },
557
+ {
558
+ "epoch": 0.57,
559
+ "learning_rate": 9.256666666666666e-07,
560
+ "loss": 0.1466,
561
+ "step": 2225
562
+ },
563
+ {
564
+ "epoch": 0.58,
565
+ "learning_rate": 9.173333333333333e-07,
566
+ "loss": 0.1357,
567
+ "step": 2250
568
+ },
569
+ {
570
+ "epoch": 0.59,
571
+ "learning_rate": 9.09e-07,
572
+ "loss": 0.148,
573
+ "step": 2275
574
+ },
575
+ {
576
+ "epoch": 0.59,
577
+ "learning_rate": 9.006666666666666e-07,
578
+ "loss": 0.1684,
579
+ "step": 2300
580
+ },
581
+ {
582
+ "epoch": 0.6,
583
+ "learning_rate": 8.923333333333333e-07,
584
+ "loss": 0.1368,
585
+ "step": 2325
586
+ },
587
+ {
588
+ "epoch": 0.61,
589
+ "learning_rate": 8.839999999999999e-07,
590
+ "loss": 0.188,
591
+ "step": 2350
592
+ },
593
+ {
594
+ "epoch": 0.61,
595
+ "learning_rate": 8.756666666666666e-07,
596
+ "loss": 0.1299,
597
+ "step": 2375
598
+ },
599
+ {
600
+ "epoch": 0.62,
601
+ "learning_rate": 8.673333333333332e-07,
602
+ "loss": 0.1461,
603
+ "step": 2400
604
+ },
605
+ {
606
+ "epoch": 0.63,
607
+ "learning_rate": 8.59e-07,
608
+ "loss": 0.1569,
609
+ "step": 2425
610
+ },
611
+ {
612
+ "epoch": 0.63,
613
+ "learning_rate": 8.506666666666667e-07,
614
+ "loss": 0.1527,
615
+ "step": 2450
616
+ },
617
+ {
618
+ "epoch": 0.64,
619
+ "learning_rate": 8.423333333333334e-07,
620
+ "loss": 0.1041,
621
+ "step": 2475
622
+ },
623
+ {
624
+ "epoch": 0.65,
625
+ "learning_rate": 8.34e-07,
626
+ "loss": 0.157,
627
+ "step": 2500
628
+ },
629
+ {
630
+ "epoch": 0.65,
631
+ "learning_rate": 8.256666666666666e-07,
632
+ "loss": 0.1364,
633
+ "step": 2525
634
+ },
635
+ {
636
+ "epoch": 0.66,
637
+ "learning_rate": 8.173333333333333e-07,
638
+ "loss": 0.1685,
639
+ "step": 2550
640
+ },
641
+ {
642
+ "epoch": 0.66,
643
+ "learning_rate": 8.09e-07,
644
+ "loss": 0.1344,
645
+ "step": 2575
646
+ },
647
+ {
648
+ "epoch": 0.67,
649
+ "learning_rate": 8.006666666666666e-07,
650
+ "loss": 0.1589,
651
+ "step": 2600
652
+ },
653
+ {
654
+ "epoch": 0.68,
655
+ "learning_rate": 7.923333333333333e-07,
656
+ "loss": 0.1334,
657
+ "step": 2625
658
+ },
659
+ {
660
+ "epoch": 0.68,
661
+ "learning_rate": 7.84e-07,
662
+ "loss": 0.1297,
663
+ "step": 2650
664
+ },
665
+ {
666
+ "epoch": 0.69,
667
+ "learning_rate": 7.756666666666665e-07,
668
+ "loss": 0.1326,
669
+ "step": 2675
670
+ },
671
+ {
672
+ "epoch": 0.7,
673
+ "learning_rate": 7.673333333333332e-07,
674
+ "loss": 0.179,
675
+ "step": 2700
676
+ },
677
+ {
678
+ "epoch": 0.7,
679
+ "learning_rate": 7.59e-07,
680
+ "loss": 0.1242,
681
+ "step": 2725
682
+ },
683
+ {
684
+ "epoch": 0.71,
685
+ "learning_rate": 7.506666666666667e-07,
686
+ "loss": 0.1714,
687
+ "step": 2750
688
+ },
689
+ {
690
+ "epoch": 0.72,
691
+ "learning_rate": 7.423333333333333e-07,
692
+ "loss": 0.1503,
693
+ "step": 2775
694
+ },
695
+ {
696
+ "epoch": 0.72,
697
+ "learning_rate": 7.34e-07,
698
+ "loss": 0.131,
699
+ "step": 2800
700
+ },
701
+ {
702
+ "epoch": 0.73,
703
+ "learning_rate": 7.256666666666667e-07,
704
+ "loss": 0.118,
705
+ "step": 2825
706
+ },
707
+ {
708
+ "epoch": 0.74,
709
+ "learning_rate": 7.173333333333333e-07,
710
+ "loss": 0.1729,
711
+ "step": 2850
712
+ },
713
+ {
714
+ "epoch": 0.74,
715
+ "learning_rate": 7.089999999999999e-07,
716
+ "loss": 0.1193,
717
+ "step": 2875
718
+ },
719
+ {
720
+ "epoch": 0.75,
721
+ "learning_rate": 7.006666666666666e-07,
722
+ "loss": 0.1723,
723
+ "step": 2900
724
+ },
725
+ {
726
+ "epoch": 0.75,
727
+ "learning_rate": 6.923333333333333e-07,
728
+ "loss": 0.1393,
729
+ "step": 2925
730
+ },
731
+ {
732
+ "epoch": 0.76,
733
+ "learning_rate": 6.84e-07,
734
+ "loss": 0.1259,
735
+ "step": 2950
736
+ },
737
+ {
738
+ "epoch": 0.77,
739
+ "learning_rate": 6.756666666666666e-07,
740
+ "loss": 0.1228,
741
+ "step": 2975
742
+ },
743
+ {
744
+ "epoch": 0.77,
745
+ "learning_rate": 6.673333333333334e-07,
746
+ "loss": 0.1794,
747
+ "step": 3000
748
+ },
749
+ {
750
+ "epoch": 0.77,
751
+ "eval_loss": 0.13788852095603943,
752
+ "eval_runtime": 198193.9708,
753
+ "eval_samples_per_second": 0.047,
754
+ "eval_steps_per_second": 0.006,
755
+ "eval_wer": 0.09970911441499677,
756
+ "step": 3000
757
+ }
758
+ ],
759
+ "logging_steps": 25,
760
+ "max_steps": 5000,
761
+ "num_input_tokens_seen": 0,
762
+ "num_train_epochs": 2,
763
+ "save_steps": 1000,
764
+ "total_flos": 8.153995935744e+19,
765
+ "train_batch_size": 8,
766
+ "trial_name": null,
767
+ "trial_params": null
768
+ }
checkpoint-3000/training_args.bin ADDED
Binary file (4.86 kB). View file
 
checkpoint-4000/config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-large-v3",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": true,
6
+ "architectures": [
7
+ "WhisperForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": [
11
+ 220,
12
+ 50257
13
+ ],
14
+ "bos_token_id": 50257,
15
+ "classifier_proj_size": 256,
16
+ "d_model": 1280,
17
+ "decoder_attention_heads": 20,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 32,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 20,
24
+ "encoder_ffn_dim": 5120,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 32,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 32,
43
+ "num_mel_bins": 128,
44
+ "pad_token_id": 50256,
45
+ "scale_embedding": false,
46
+ "torch_dtype": "float32",
47
+ "transformers_version": "4.37.0.dev0",
48
+ "use_cache": true,
49
+ "use_weighted_layer_sum": false,
50
+ "vocab_size": 51866
51
+ }
checkpoint-4000/generation_config.json ADDED
@@ -0,0 +1,265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 7,
5
+ 0
6
+ ],
7
+ [
8
+ 10,
9
+ 17
10
+ ],
11
+ [
12
+ 12,
13
+ 18
14
+ ],
15
+ [
16
+ 13,
17
+ 12
18
+ ],
19
+ [
20
+ 16,
21
+ 1
22
+ ],
23
+ [
24
+ 17,
25
+ 14
26
+ ],
27
+ [
28
+ 19,
29
+ 11
30
+ ],
31
+ [
32
+ 21,
33
+ 4
34
+ ],
35
+ [
36
+ 24,
37
+ 1
38
+ ],
39
+ [
40
+ 25,
41
+ 6
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50359
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|yue|>": 50358,
162
+ "<|zh|>": 50260
163
+ },
164
+ "max_initial_timestamp_index": 50,
165
+ "max_length": 448,
166
+ "no_timestamps_token_id": 50364,
167
+ "pad_token_id": 50257,
168
+ "prev_sot_token_id": 50362,
169
+ "return_timestamps": false,
170
+ "suppress_tokens": [
171
+ 1,
172
+ 2,
173
+ 7,
174
+ 8,
175
+ 9,
176
+ 10,
177
+ 14,
178
+ 25,
179
+ 26,
180
+ 27,
181
+ 28,
182
+ 29,
183
+ 31,
184
+ 58,
185
+ 59,
186
+ 60,
187
+ 61,
188
+ 62,
189
+ 63,
190
+ 90,
191
+ 91,
192
+ 92,
193
+ 93,
194
+ 359,
195
+ 503,
196
+ 522,
197
+ 542,
198
+ 873,
199
+ 893,
200
+ 902,
201
+ 918,
202
+ 922,
203
+ 931,
204
+ 1350,
205
+ 1853,
206
+ 1982,
207
+ 2460,
208
+ 2627,
209
+ 3246,
210
+ 3253,
211
+ 3268,
212
+ 3536,
213
+ 3846,
214
+ 3961,
215
+ 4183,
216
+ 4667,
217
+ 6585,
218
+ 6647,
219
+ 7273,
220
+ 9061,
221
+ 9383,
222
+ 10428,
223
+ 10929,
224
+ 11938,
225
+ 12033,
226
+ 12331,
227
+ 12562,
228
+ 13793,
229
+ 14157,
230
+ 14635,
231
+ 15265,
232
+ 15618,
233
+ 16553,
234
+ 16604,
235
+ 18362,
236
+ 18956,
237
+ 20075,
238
+ 21675,
239
+ 22520,
240
+ 26130,
241
+ 26161,
242
+ 26435,
243
+ 28279,
244
+ 29464,
245
+ 31650,
246
+ 32302,
247
+ 32470,
248
+ 36865,
249
+ 42863,
250
+ 47425,
251
+ 49870,
252
+ 50254,
253
+ 50258,
254
+ 50359,
255
+ 50360,
256
+ 50361,
257
+ 50362,
258
+ 50363
259
+ ],
260
+ "task_to_id": {
261
+ "transcribe": 50360,
262
+ "translate": 50359
263
+ },
264
+ "transformers_version": "4.37.0.dev0"
265
+ }
checkpoint-4000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fedb9cc896a6388bb4f5cddb373b7392782ec965512769af969c60e1af1a4e14
3
+ size 4993448880
checkpoint-4000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c6b0edf383dba5a20a86c58366da587af2b11f57f3238809e22174428275ba2
3
+ size 1180663192
checkpoint-4000/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80e4f108eb5557c44d32434b4917e37c96c5a7f16fb94640266fc8260e5fd15e
3
+ size 12333660476
checkpoint-4000/preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 128,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
checkpoint-4000/rng_state.pth ADDED
Binary file (14.3 kB). View file
 
checkpoint-4000/scheduler.pt ADDED
Binary file (1.06 kB). View file
 
checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,1017 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.032258064516129,
5
+ "eval_steps": 1000,
6
+ "global_step": 4000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 1.1499999999999999e-08,
14
+ "loss": 0.7822,
15
+ "step": 25
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 2.4e-08,
20
+ "loss": 1.3149,
21
+ "step": 50
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 3.6499999999999996e-08,
26
+ "loss": 0.7809,
27
+ "step": 75
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 4.9e-08,
32
+ "loss": 1.2916,
33
+ "step": 100
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 6.15e-08,
38
+ "loss": 0.7645,
39
+ "step": 125
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 7.399999999999999e-08,
44
+ "loss": 1.3379,
45
+ "step": 150
46
+ },
47
+ {
48
+ "epoch": 0.05,
49
+ "learning_rate": 8.649999999999999e-08,
50
+ "loss": 0.7722,
51
+ "step": 175
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 9.9e-08,
56
+ "loss": 1.281,
57
+ "step": 200
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 1.115e-07,
62
+ "loss": 0.6998,
63
+ "step": 225
64
+ },
65
+ {
66
+ "epoch": 0.06,
67
+ "learning_rate": 1.24e-07,
68
+ "loss": 1.1532,
69
+ "step": 250
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "learning_rate": 1.365e-07,
74
+ "loss": 0.583,
75
+ "step": 275
76
+ },
77
+ {
78
+ "epoch": 0.08,
79
+ "learning_rate": 1.49e-07,
80
+ "loss": 0.6693,
81
+ "step": 300
82
+ },
83
+ {
84
+ "epoch": 0.08,
85
+ "learning_rate": 1.615e-07,
86
+ "loss": 0.3395,
87
+ "step": 325
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "learning_rate": 1.7399999999999997e-07,
92
+ "loss": 0.3597,
93
+ "step": 350
94
+ },
95
+ {
96
+ "epoch": 0.1,
97
+ "learning_rate": 1.8649999999999998e-07,
98
+ "loss": 0.22,
99
+ "step": 375
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "learning_rate": 1.99e-07,
104
+ "loss": 0.3449,
105
+ "step": 400
106
+ },
107
+ {
108
+ "epoch": 0.11,
109
+ "learning_rate": 2.1149999999999998e-07,
110
+ "loss": 0.2123,
111
+ "step": 425
112
+ },
113
+ {
114
+ "epoch": 0.12,
115
+ "learning_rate": 2.24e-07,
116
+ "loss": 0.3362,
117
+ "step": 450
118
+ },
119
+ {
120
+ "epoch": 0.12,
121
+ "learning_rate": 2.3649999999999998e-07,
122
+ "loss": 0.175,
123
+ "step": 475
124
+ },
125
+ {
126
+ "epoch": 0.13,
127
+ "learning_rate": 2.4899999999999997e-07,
128
+ "loss": 0.3051,
129
+ "step": 500
130
+ },
131
+ {
132
+ "epoch": 0.14,
133
+ "learning_rate": 2.615e-07,
134
+ "loss": 0.176,
135
+ "step": 525
136
+ },
137
+ {
138
+ "epoch": 0.14,
139
+ "learning_rate": 2.74e-07,
140
+ "loss": 0.2517,
141
+ "step": 550
142
+ },
143
+ {
144
+ "epoch": 0.15,
145
+ "learning_rate": 2.865e-07,
146
+ "loss": 0.1863,
147
+ "step": 575
148
+ },
149
+ {
150
+ "epoch": 0.15,
151
+ "learning_rate": 2.9899999999999996e-07,
152
+ "loss": 0.2614,
153
+ "step": 600
154
+ },
155
+ {
156
+ "epoch": 0.16,
157
+ "learning_rate": 3.115e-07,
158
+ "loss": 0.1942,
159
+ "step": 625
160
+ },
161
+ {
162
+ "epoch": 0.17,
163
+ "learning_rate": 3.24e-07,
164
+ "loss": 0.2901,
165
+ "step": 650
166
+ },
167
+ {
168
+ "epoch": 0.17,
169
+ "learning_rate": 3.3650000000000003e-07,
170
+ "loss": 0.1587,
171
+ "step": 675
172
+ },
173
+ {
174
+ "epoch": 0.18,
175
+ "learning_rate": 3.4899999999999996e-07,
176
+ "loss": 0.2128,
177
+ "step": 700
178
+ },
179
+ {
180
+ "epoch": 0.19,
181
+ "learning_rate": 3.6149999999999995e-07,
182
+ "loss": 0.1647,
183
+ "step": 725
184
+ },
185
+ {
186
+ "epoch": 0.19,
187
+ "learning_rate": 3.74e-07,
188
+ "loss": 0.2473,
189
+ "step": 750
190
+ },
191
+ {
192
+ "epoch": 0.2,
193
+ "learning_rate": 3.8649999999999997e-07,
194
+ "loss": 0.1415,
195
+ "step": 775
196
+ },
197
+ {
198
+ "epoch": 0.21,
199
+ "learning_rate": 3.99e-07,
200
+ "loss": 0.2367,
201
+ "step": 800
202
+ },
203
+ {
204
+ "epoch": 0.21,
205
+ "learning_rate": 4.1149999999999995e-07,
206
+ "loss": 0.1498,
207
+ "step": 825
208
+ },
209
+ {
210
+ "epoch": 0.22,
211
+ "learning_rate": 4.24e-07,
212
+ "loss": 0.2161,
213
+ "step": 850
214
+ },
215
+ {
216
+ "epoch": 0.23,
217
+ "learning_rate": 4.3649999999999997e-07,
218
+ "loss": 0.144,
219
+ "step": 875
220
+ },
221
+ {
222
+ "epoch": 0.23,
223
+ "learning_rate": 4.49e-07,
224
+ "loss": 0.2316,
225
+ "step": 900
226
+ },
227
+ {
228
+ "epoch": 0.24,
229
+ "learning_rate": 4.615e-07,
230
+ "loss": 0.1449,
231
+ "step": 925
232
+ },
233
+ {
234
+ "epoch": 0.25,
235
+ "learning_rate": 4.7399999999999993e-07,
236
+ "loss": 0.1513,
237
+ "step": 950
238
+ },
239
+ {
240
+ "epoch": 0.25,
241
+ "learning_rate": 4.864999999999999e-07,
242
+ "loss": 0.1855,
243
+ "step": 975
244
+ },
245
+ {
246
+ "epoch": 0.26,
247
+ "learning_rate": 4.99e-07,
248
+ "loss": 0.199,
249
+ "step": 1000
250
+ },
251
+ {
252
+ "epoch": 0.26,
253
+ "eval_loss": 0.15628743171691895,
254
+ "eval_runtime": 184706.1006,
255
+ "eval_samples_per_second": 0.051,
256
+ "eval_steps_per_second": 0.006,
257
+ "eval_wer": 0.1124272786037492,
258
+ "step": 1000
259
+ },
260
+ {
261
+ "epoch": 0.26,
262
+ "learning_rate": 5.114999999999999e-07,
263
+ "loss": 0.1301,
264
+ "step": 1025
265
+ },
266
+ {
267
+ "epoch": 0.27,
268
+ "learning_rate": 5.24e-07,
269
+ "loss": 0.2193,
270
+ "step": 1050
271
+ },
272
+ {
273
+ "epoch": 0.28,
274
+ "learning_rate": 5.365e-07,
275
+ "loss": 0.1552,
276
+ "step": 1075
277
+ },
278
+ {
279
+ "epoch": 0.28,
280
+ "learning_rate": 5.490000000000001e-07,
281
+ "loss": 0.1865,
282
+ "step": 1100
283
+ },
284
+ {
285
+ "epoch": 0.29,
286
+ "learning_rate": 5.614999999999999e-07,
287
+ "loss": 0.1618,
288
+ "step": 1125
289
+ },
290
+ {
291
+ "epoch": 0.3,
292
+ "learning_rate": 5.739999999999999e-07,
293
+ "loss": 0.2259,
294
+ "step": 1150
295
+ },
296
+ {
297
+ "epoch": 0.3,
298
+ "learning_rate": 5.865e-07,
299
+ "loss": 0.1418,
300
+ "step": 1175
301
+ },
302
+ {
303
+ "epoch": 0.31,
304
+ "learning_rate": 5.989999999999999e-07,
305
+ "loss": 0.1916,
306
+ "step": 1200
307
+ },
308
+ {
309
+ "epoch": 0.32,
310
+ "learning_rate": 6.115e-07,
311
+ "loss": 0.1295,
312
+ "step": 1225
313
+ },
314
+ {
315
+ "epoch": 0.32,
316
+ "learning_rate": 6.24e-07,
317
+ "loss": 0.2112,
318
+ "step": 1250
319
+ },
320
+ {
321
+ "epoch": 0.33,
322
+ "learning_rate": 6.365e-07,
323
+ "loss": 0.1231,
324
+ "step": 1275
325
+ },
326
+ {
327
+ "epoch": 0.34,
328
+ "learning_rate": 6.49e-07,
329
+ "loss": 0.1914,
330
+ "step": 1300
331
+ },
332
+ {
333
+ "epoch": 0.34,
334
+ "learning_rate": 6.614999999999999e-07,
335
+ "loss": 0.1485,
336
+ "step": 1325
337
+ },
338
+ {
339
+ "epoch": 0.35,
340
+ "learning_rate": 6.74e-07,
341
+ "loss": 0.1958,
342
+ "step": 1350
343
+ },
344
+ {
345
+ "epoch": 0.35,
346
+ "learning_rate": 6.865e-07,
347
+ "loss": 0.1452,
348
+ "step": 1375
349
+ },
350
+ {
351
+ "epoch": 0.36,
352
+ "learning_rate": 6.989999999999999e-07,
353
+ "loss": 0.1624,
354
+ "step": 1400
355
+ },
356
+ {
357
+ "epoch": 0.37,
358
+ "learning_rate": 7.115e-07,
359
+ "loss": 0.1518,
360
+ "step": 1425
361
+ },
362
+ {
363
+ "epoch": 0.37,
364
+ "learning_rate": 7.24e-07,
365
+ "loss": 0.1935,
366
+ "step": 1450
367
+ },
368
+ {
369
+ "epoch": 0.38,
370
+ "learning_rate": 7.365e-07,
371
+ "loss": 0.1138,
372
+ "step": 1475
373
+ },
374
+ {
375
+ "epoch": 0.39,
376
+ "learning_rate": 7.489999999999999e-07,
377
+ "loss": 0.16,
378
+ "step": 1500
379
+ },
380
+ {
381
+ "epoch": 0.39,
382
+ "learning_rate": 7.614999999999999e-07,
383
+ "loss": 0.1279,
384
+ "step": 1525
385
+ },
386
+ {
387
+ "epoch": 0.4,
388
+ "learning_rate": 7.74e-07,
389
+ "loss": 0.1862,
390
+ "step": 1550
391
+ },
392
+ {
393
+ "epoch": 0.41,
394
+ "learning_rate": 7.864999999999999e-07,
395
+ "loss": 0.1537,
396
+ "step": 1575
397
+ },
398
+ {
399
+ "epoch": 0.41,
400
+ "learning_rate": 7.99e-07,
401
+ "loss": 0.1915,
402
+ "step": 1600
403
+ },
404
+ {
405
+ "epoch": 0.42,
406
+ "learning_rate": 8.115e-07,
407
+ "loss": 0.1303,
408
+ "step": 1625
409
+ },
410
+ {
411
+ "epoch": 0.43,
412
+ "learning_rate": 8.24e-07,
413
+ "loss": 0.1967,
414
+ "step": 1650
415
+ },
416
+ {
417
+ "epoch": 0.43,
418
+ "learning_rate": 8.365e-07,
419
+ "loss": 0.1418,
420
+ "step": 1675
421
+ },
422
+ {
423
+ "epoch": 0.44,
424
+ "learning_rate": 8.489999999999999e-07,
425
+ "loss": 0.194,
426
+ "step": 1700
427
+ },
428
+ {
429
+ "epoch": 0.45,
430
+ "learning_rate": 8.615e-07,
431
+ "loss": 0.1377,
432
+ "step": 1725
433
+ },
434
+ {
435
+ "epoch": 0.45,
436
+ "learning_rate": 8.739999999999999e-07,
437
+ "loss": 0.1534,
438
+ "step": 1750
439
+ },
440
+ {
441
+ "epoch": 0.46,
442
+ "learning_rate": 8.864999999999999e-07,
443
+ "loss": 0.1416,
444
+ "step": 1775
445
+ },
446
+ {
447
+ "epoch": 0.46,
448
+ "learning_rate": 8.99e-07,
449
+ "loss": 0.1869,
450
+ "step": 1800
451
+ },
452
+ {
453
+ "epoch": 0.47,
454
+ "learning_rate": 9.115e-07,
455
+ "loss": 0.1507,
456
+ "step": 1825
457
+ },
458
+ {
459
+ "epoch": 0.48,
460
+ "learning_rate": 9.24e-07,
461
+ "loss": 0.1571,
462
+ "step": 1850
463
+ },
464
+ {
465
+ "epoch": 0.48,
466
+ "learning_rate": 9.365e-07,
467
+ "loss": 0.1348,
468
+ "step": 1875
469
+ },
470
+ {
471
+ "epoch": 0.49,
472
+ "learning_rate": 9.489999999999999e-07,
473
+ "loss": 0.1981,
474
+ "step": 1900
475
+ },
476
+ {
477
+ "epoch": 0.5,
478
+ "learning_rate": 9.615e-07,
479
+ "loss": 0.1324,
480
+ "step": 1925
481
+ },
482
+ {
483
+ "epoch": 0.5,
484
+ "learning_rate": 9.74e-07,
485
+ "loss": 0.1712,
486
+ "step": 1950
487
+ },
488
+ {
489
+ "epoch": 0.51,
490
+ "learning_rate": 9.865e-07,
491
+ "loss": 0.1169,
492
+ "step": 1975
493
+ },
494
+ {
495
+ "epoch": 0.52,
496
+ "learning_rate": 9.989999999999999e-07,
497
+ "loss": 0.1654,
498
+ "step": 2000
499
+ },
500
+ {
501
+ "epoch": 0.52,
502
+ "eval_loss": 0.1500108540058136,
503
+ "eval_runtime": 188234.7285,
504
+ "eval_samples_per_second": 0.05,
505
+ "eval_steps_per_second": 0.006,
506
+ "eval_wer": 0.10521978021978022,
507
+ "step": 2000
508
+ },
509
+ {
510
+ "epoch": 0.52,
511
+ "learning_rate": 9.923333333333332e-07,
512
+ "loss": 0.1202,
513
+ "step": 2025
514
+ },
515
+ {
516
+ "epoch": 0.53,
517
+ "learning_rate": 9.84e-07,
518
+ "loss": 0.186,
519
+ "step": 2050
520
+ },
521
+ {
522
+ "epoch": 0.54,
523
+ "learning_rate": 9.756666666666666e-07,
524
+ "loss": 0.1106,
525
+ "step": 2075
526
+ },
527
+ {
528
+ "epoch": 0.54,
529
+ "learning_rate": 9.673333333333332e-07,
530
+ "loss": 0.1926,
531
+ "step": 2100
532
+ },
533
+ {
534
+ "epoch": 0.55,
535
+ "learning_rate": 9.589999999999998e-07,
536
+ "loss": 0.1178,
537
+ "step": 2125
538
+ },
539
+ {
540
+ "epoch": 0.55,
541
+ "learning_rate": 9.506666666666667e-07,
542
+ "loss": 0.1744,
543
+ "step": 2150
544
+ },
545
+ {
546
+ "epoch": 0.56,
547
+ "learning_rate": 9.423333333333333e-07,
548
+ "loss": 0.1263,
549
+ "step": 2175
550
+ },
551
+ {
552
+ "epoch": 0.57,
553
+ "learning_rate": 9.34e-07,
554
+ "loss": 0.1818,
555
+ "step": 2200
556
+ },
557
+ {
558
+ "epoch": 0.57,
559
+ "learning_rate": 9.256666666666666e-07,
560
+ "loss": 0.1466,
561
+ "step": 2225
562
+ },
563
+ {
564
+ "epoch": 0.58,
565
+ "learning_rate": 9.173333333333333e-07,
566
+ "loss": 0.1357,
567
+ "step": 2250
568
+ },
569
+ {
570
+ "epoch": 0.59,
571
+ "learning_rate": 9.09e-07,
572
+ "loss": 0.148,
573
+ "step": 2275
574
+ },
575
+ {
576
+ "epoch": 0.59,
577
+ "learning_rate": 9.006666666666666e-07,
578
+ "loss": 0.1684,
579
+ "step": 2300
580
+ },
581
+ {
582
+ "epoch": 0.6,
583
+ "learning_rate": 8.923333333333333e-07,
584
+ "loss": 0.1368,
585
+ "step": 2325
586
+ },
587
+ {
588
+ "epoch": 0.61,
589
+ "learning_rate": 8.839999999999999e-07,
590
+ "loss": 0.188,
591
+ "step": 2350
592
+ },
593
+ {
594
+ "epoch": 0.61,
595
+ "learning_rate": 8.756666666666666e-07,
596
+ "loss": 0.1299,
597
+ "step": 2375
598
+ },
599
+ {
600
+ "epoch": 0.62,
601
+ "learning_rate": 8.673333333333332e-07,
602
+ "loss": 0.1461,
603
+ "step": 2400
604
+ },
605
+ {
606
+ "epoch": 0.63,
607
+ "learning_rate": 8.59e-07,
608
+ "loss": 0.1569,
609
+ "step": 2425
610
+ },
611
+ {
612
+ "epoch": 0.63,
613
+ "learning_rate": 8.506666666666667e-07,
614
+ "loss": 0.1527,
615
+ "step": 2450
616
+ },
617
+ {
618
+ "epoch": 0.64,
619
+ "learning_rate": 8.423333333333334e-07,
620
+ "loss": 0.1041,
621
+ "step": 2475
622
+ },
623
+ {
624
+ "epoch": 0.65,
625
+ "learning_rate": 8.34e-07,
626
+ "loss": 0.157,
627
+ "step": 2500
628
+ },
629
+ {
630
+ "epoch": 0.65,
631
+ "learning_rate": 8.256666666666666e-07,
632
+ "loss": 0.1364,
633
+ "step": 2525
634
+ },
635
+ {
636
+ "epoch": 0.66,
637
+ "learning_rate": 8.173333333333333e-07,
638
+ "loss": 0.1685,
639
+ "step": 2550
640
+ },
641
+ {
642
+ "epoch": 0.66,
643
+ "learning_rate": 8.09e-07,
644
+ "loss": 0.1344,
645
+ "step": 2575
646
+ },
647
+ {
648
+ "epoch": 0.67,
649
+ "learning_rate": 8.006666666666666e-07,
650
+ "loss": 0.1589,
651
+ "step": 2600
652
+ },
653
+ {
654
+ "epoch": 0.68,
655
+ "learning_rate": 7.923333333333333e-07,
656
+ "loss": 0.1334,
657
+ "step": 2625
658
+ },
659
+ {
660
+ "epoch": 0.68,
661
+ "learning_rate": 7.84e-07,
662
+ "loss": 0.1297,
663
+ "step": 2650
664
+ },
665
+ {
666
+ "epoch": 0.69,
667
+ "learning_rate": 7.756666666666665e-07,
668
+ "loss": 0.1326,
669
+ "step": 2675
670
+ },
671
+ {
672
+ "epoch": 0.7,
673
+ "learning_rate": 7.673333333333332e-07,
674
+ "loss": 0.179,
675
+ "step": 2700
676
+ },
677
+ {
678
+ "epoch": 0.7,
679
+ "learning_rate": 7.59e-07,
680
+ "loss": 0.1242,
681
+ "step": 2725
682
+ },
683
+ {
684
+ "epoch": 0.71,
685
+ "learning_rate": 7.506666666666667e-07,
686
+ "loss": 0.1714,
687
+ "step": 2750
688
+ },
689
+ {
690
+ "epoch": 0.72,
691
+ "learning_rate": 7.423333333333333e-07,
692
+ "loss": 0.1503,
693
+ "step": 2775
694
+ },
695
+ {
696
+ "epoch": 0.72,
697
+ "learning_rate": 7.34e-07,
698
+ "loss": 0.131,
699
+ "step": 2800
700
+ },
701
+ {
702
+ "epoch": 0.73,
703
+ "learning_rate": 7.256666666666667e-07,
704
+ "loss": 0.118,
705
+ "step": 2825
706
+ },
707
+ {
708
+ "epoch": 0.74,
709
+ "learning_rate": 7.173333333333333e-07,
710
+ "loss": 0.1729,
711
+ "step": 2850
712
+ },
713
+ {
714
+ "epoch": 0.74,
715
+ "learning_rate": 7.089999999999999e-07,
716
+ "loss": 0.1193,
717
+ "step": 2875
718
+ },
719
+ {
720
+ "epoch": 0.75,
721
+ "learning_rate": 7.006666666666666e-07,
722
+ "loss": 0.1723,
723
+ "step": 2900
724
+ },
725
+ {
726
+ "epoch": 0.75,
727
+ "learning_rate": 6.923333333333333e-07,
728
+ "loss": 0.1393,
729
+ "step": 2925
730
+ },
731
+ {
732
+ "epoch": 0.76,
733
+ "learning_rate": 6.84e-07,
734
+ "loss": 0.1259,
735
+ "step": 2950
736
+ },
737
+ {
738
+ "epoch": 0.77,
739
+ "learning_rate": 6.756666666666666e-07,
740
+ "loss": 0.1228,
741
+ "step": 2975
742
+ },
743
+ {
744
+ "epoch": 0.77,
745
+ "learning_rate": 6.673333333333334e-07,
746
+ "loss": 0.1794,
747
+ "step": 3000
748
+ },
749
+ {
750
+ "epoch": 0.77,
751
+ "eval_loss": 0.13788852095603943,
752
+ "eval_runtime": 198193.9708,
753
+ "eval_samples_per_second": 0.047,
754
+ "eval_steps_per_second": 0.006,
755
+ "eval_wer": 0.09970911441499677,
756
+ "step": 3000
757
+ },
758
+ {
759
+ "epoch": 0.78,
760
+ "learning_rate": 6.59e-07,
761
+ "loss": 0.1334,
762
+ "step": 3025
763
+ },
764
+ {
765
+ "epoch": 0.79,
766
+ "learning_rate": 6.506666666666666e-07,
767
+ "loss": 0.1584,
768
+ "step": 3050
769
+ },
770
+ {
771
+ "epoch": 0.79,
772
+ "learning_rate": 6.423333333333333e-07,
773
+ "loss": 0.1331,
774
+ "step": 3075
775
+ },
776
+ {
777
+ "epoch": 0.8,
778
+ "learning_rate": 6.346666666666666e-07,
779
+ "loss": 0.1226,
780
+ "step": 3100
781
+ },
782
+ {
783
+ "epoch": 0.81,
784
+ "learning_rate": 6.263333333333332e-07,
785
+ "loss": 0.135,
786
+ "step": 3125
787
+ },
788
+ {
789
+ "epoch": 0.81,
790
+ "learning_rate": 6.18e-07,
791
+ "loss": 0.1538,
792
+ "step": 3150
793
+ },
794
+ {
795
+ "epoch": 0.82,
796
+ "learning_rate": 6.096666666666667e-07,
797
+ "loss": 0.1405,
798
+ "step": 3175
799
+ },
800
+ {
801
+ "epoch": 0.83,
802
+ "learning_rate": 6.013333333333334e-07,
803
+ "loss": 0.1534,
804
+ "step": 3200
805
+ },
806
+ {
807
+ "epoch": 0.83,
808
+ "learning_rate": 5.93e-07,
809
+ "loss": 0.1628,
810
+ "step": 3225
811
+ },
812
+ {
813
+ "epoch": 0.84,
814
+ "learning_rate": 5.846666666666667e-07,
815
+ "loss": 0.2002,
816
+ "step": 3250
817
+ },
818
+ {
819
+ "epoch": 0.85,
820
+ "learning_rate": 5.763333333333333e-07,
821
+ "loss": 0.1155,
822
+ "step": 3275
823
+ },
824
+ {
825
+ "epoch": 0.85,
826
+ "learning_rate": 5.679999999999999e-07,
827
+ "loss": 0.179,
828
+ "step": 3300
829
+ },
830
+ {
831
+ "epoch": 0.86,
832
+ "learning_rate": 5.596666666666666e-07,
833
+ "loss": 0.1471,
834
+ "step": 3325
835
+ },
836
+ {
837
+ "epoch": 0.86,
838
+ "learning_rate": 5.513333333333333e-07,
839
+ "loss": 0.1386,
840
+ "step": 3350
841
+ },
842
+ {
843
+ "epoch": 0.87,
844
+ "learning_rate": 5.43e-07,
845
+ "loss": 0.1185,
846
+ "step": 3375
847
+ },
848
+ {
849
+ "epoch": 0.88,
850
+ "learning_rate": 5.346666666666666e-07,
851
+ "loss": 0.1418,
852
+ "step": 3400
853
+ },
854
+ {
855
+ "epoch": 0.88,
856
+ "learning_rate": 5.263333333333333e-07,
857
+ "loss": 0.1015,
858
+ "step": 3425
859
+ },
860
+ {
861
+ "epoch": 0.89,
862
+ "learning_rate": 5.18e-07,
863
+ "loss": 0.1525,
864
+ "step": 3450
865
+ },
866
+ {
867
+ "epoch": 0.9,
868
+ "learning_rate": 5.096666666666667e-07,
869
+ "loss": 0.1212,
870
+ "step": 3475
871
+ },
872
+ {
873
+ "epoch": 0.9,
874
+ "learning_rate": 5.013333333333333e-07,
875
+ "loss": 0.1623,
876
+ "step": 3500
877
+ },
878
+ {
879
+ "epoch": 0.91,
880
+ "learning_rate": 4.93e-07,
881
+ "loss": 0.1178,
882
+ "step": 3525
883
+ },
884
+ {
885
+ "epoch": 0.92,
886
+ "learning_rate": 4.846666666666667e-07,
887
+ "loss": 0.1618,
888
+ "step": 3550
889
+ },
890
+ {
891
+ "epoch": 0.92,
892
+ "learning_rate": 4.763333333333333e-07,
893
+ "loss": 0.1187,
894
+ "step": 3575
895
+ },
896
+ {
897
+ "epoch": 0.93,
898
+ "learning_rate": 4.68e-07,
899
+ "loss": 0.1382,
900
+ "step": 3600
901
+ },
902
+ {
903
+ "epoch": 0.94,
904
+ "learning_rate": 4.5966666666666667e-07,
905
+ "loss": 0.1211,
906
+ "step": 3625
907
+ },
908
+ {
909
+ "epoch": 0.94,
910
+ "learning_rate": 4.5133333333333327e-07,
911
+ "loss": 0.1582,
912
+ "step": 3650
913
+ },
914
+ {
915
+ "epoch": 0.95,
916
+ "learning_rate": 4.43e-07,
917
+ "loss": 0.1201,
918
+ "step": 3675
919
+ },
920
+ {
921
+ "epoch": 0.95,
922
+ "learning_rate": 4.3466666666666664e-07,
923
+ "loss": 0.1571,
924
+ "step": 3700
925
+ },
926
+ {
927
+ "epoch": 0.96,
928
+ "learning_rate": 4.263333333333333e-07,
929
+ "loss": 0.1247,
930
+ "step": 3725
931
+ },
932
+ {
933
+ "epoch": 0.97,
934
+ "learning_rate": 4.1799999999999996e-07,
935
+ "loss": 0.1648,
936
+ "step": 3750
937
+ },
938
+ {
939
+ "epoch": 0.97,
940
+ "learning_rate": 4.0966666666666667e-07,
941
+ "loss": 0.1313,
942
+ "step": 3775
943
+ },
944
+ {
945
+ "epoch": 0.98,
946
+ "learning_rate": 4.0133333333333333e-07,
947
+ "loss": 0.1528,
948
+ "step": 3800
949
+ },
950
+ {
951
+ "epoch": 0.99,
952
+ "learning_rate": 3.93e-07,
953
+ "loss": 0.1252,
954
+ "step": 3825
955
+ },
956
+ {
957
+ "epoch": 0.99,
958
+ "learning_rate": 3.8466666666666664e-07,
959
+ "loss": 0.1674,
960
+ "step": 3850
961
+ },
962
+ {
963
+ "epoch": 1.0,
964
+ "learning_rate": 3.7633333333333335e-07,
965
+ "loss": 0.1192,
966
+ "step": 3875
967
+ },
968
+ {
969
+ "epoch": 1.01,
970
+ "learning_rate": 3.6799999999999996e-07,
971
+ "loss": 0.1054,
972
+ "step": 3900
973
+ },
974
+ {
975
+ "epoch": 1.01,
976
+ "learning_rate": 3.5966666666666667e-07,
977
+ "loss": 0.1353,
978
+ "step": 3925
979
+ },
980
+ {
981
+ "epoch": 1.02,
982
+ "learning_rate": 3.5133333333333333e-07,
983
+ "loss": 0.1004,
984
+ "step": 3950
985
+ },
986
+ {
987
+ "epoch": 1.03,
988
+ "learning_rate": 3.43e-07,
989
+ "loss": 0.1382,
990
+ "step": 3975
991
+ },
992
+ {
993
+ "epoch": 1.03,
994
+ "learning_rate": 3.3466666666666665e-07,
995
+ "loss": 0.0821,
996
+ "step": 4000
997
+ },
998
+ {
999
+ "epoch": 1.03,
1000
+ "eval_loss": 0.13208560645580292,
1001
+ "eval_runtime": 194476.4747,
1002
+ "eval_samples_per_second": 0.048,
1003
+ "eval_steps_per_second": 0.006,
1004
+ "eval_wer": 1.000743374272786,
1005
+ "step": 4000
1006
+ }
1007
+ ],
1008
+ "logging_steps": 25,
1009
+ "max_steps": 5000,
1010
+ "num_input_tokens_seen": 0,
1011
+ "num_train_epochs": 2,
1012
+ "save_steps": 1000,
1013
+ "total_flos": 1.0871315081330688e+20,
1014
+ "train_batch_size": 8,
1015
+ "trial_name": null,
1016
+ "trial_params": null
1017
+ }
checkpoint-4000/training_args.bin ADDED
Binary file (4.86 kB). View file
 
checkpoint-5000/config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-large-v3",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": true,
6
+ "architectures": [
7
+ "WhisperForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": [
11
+ 220,
12
+ 50257
13
+ ],
14
+ "bos_token_id": 50257,
15
+ "classifier_proj_size": 256,
16
+ "d_model": 1280,
17
+ "decoder_attention_heads": 20,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 32,
21
+ "decoder_start_token_id": 50258,
22
+ "dropout": 0.0,
23
+ "encoder_attention_heads": 20,
24
+ "encoder_ffn_dim": 5120,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 32,
27
+ "eos_token_id": 50257,
28
+ "forced_decoder_ids": null,
29
+ "init_std": 0.02,
30
+ "is_encoder_decoder": true,
31
+ "mask_feature_length": 10,
32
+ "mask_feature_min_masks": 0,
33
+ "mask_feature_prob": 0.0,
34
+ "mask_time_length": 10,
35
+ "mask_time_min_masks": 2,
36
+ "mask_time_prob": 0.05,
37
+ "max_length": 448,
38
+ "max_source_positions": 1500,
39
+ "max_target_positions": 448,
40
+ "median_filter_width": 7,
41
+ "model_type": "whisper",
42
+ "num_hidden_layers": 32,
43
+ "num_mel_bins": 128,
44
+ "pad_token_id": 50256,
45
+ "scale_embedding": false,
46
+ "torch_dtype": "float32",
47
+ "transformers_version": "4.37.0.dev0",
48
+ "use_cache": true,
49
+ "use_weighted_layer_sum": false,
50
+ "vocab_size": 51866
51
+ }
checkpoint-5000/generation_config.json ADDED
@@ -0,0 +1,265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 7,
5
+ 0
6
+ ],
7
+ [
8
+ 10,
9
+ 17
10
+ ],
11
+ [
12
+ 12,
13
+ 18
14
+ ],
15
+ [
16
+ 13,
17
+ 12
18
+ ],
19
+ [
20
+ 16,
21
+ 1
22
+ ],
23
+ [
24
+ 17,
25
+ 14
26
+ ],
27
+ [
28
+ 19,
29
+ 11
30
+ ],
31
+ [
32
+ 21,
33
+ 4
34
+ ],
35
+ [
36
+ 24,
37
+ 1
38
+ ],
39
+ [
40
+ 25,
41
+ 6
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50359
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|yue|>": 50358,
162
+ "<|zh|>": 50260
163
+ },
164
+ "max_initial_timestamp_index": 50,
165
+ "max_length": 448,
166
+ "no_timestamps_token_id": 50364,
167
+ "pad_token_id": 50257,
168
+ "prev_sot_token_id": 50362,
169
+ "return_timestamps": false,
170
+ "suppress_tokens": [
171
+ 1,
172
+ 2,
173
+ 7,
174
+ 8,
175
+ 9,
176
+ 10,
177
+ 14,
178
+ 25,
179
+ 26,
180
+ 27,
181
+ 28,
182
+ 29,
183
+ 31,
184
+ 58,
185
+ 59,
186
+ 60,
187
+ 61,
188
+ 62,
189
+ 63,
190
+ 90,
191
+ 91,
192
+ 92,
193
+ 93,
194
+ 359,
195
+ 503,
196
+ 522,
197
+ 542,
198
+ 873,
199
+ 893,
200
+ 902,
201
+ 918,
202
+ 922,
203
+ 931,
204
+ 1350,
205
+ 1853,
206
+ 1982,
207
+ 2460,
208
+ 2627,
209
+ 3246,
210
+ 3253,
211
+ 3268,
212
+ 3536,
213
+ 3846,
214
+ 3961,
215
+ 4183,
216
+ 4667,
217
+ 6585,
218
+ 6647,
219
+ 7273,
220
+ 9061,
221
+ 9383,
222
+ 10428,
223
+ 10929,
224
+ 11938,
225
+ 12033,
226
+ 12331,
227
+ 12562,
228
+ 13793,
229
+ 14157,
230
+ 14635,
231
+ 15265,
232
+ 15618,
233
+ 16553,
234
+ 16604,
235
+ 18362,
236
+ 18956,
237
+ 20075,
238
+ 21675,
239
+ 22520,
240
+ 26130,
241
+ 26161,
242
+ 26435,
243
+ 28279,
244
+ 29464,
245
+ 31650,
246
+ 32302,
247
+ 32470,
248
+ 36865,
249
+ 42863,
250
+ 47425,
251
+ 49870,
252
+ 50254,
253
+ 50258,
254
+ 50359,
255
+ 50360,
256
+ 50361,
257
+ 50362,
258
+ 50363
259
+ ],
260
+ "task_to_id": {
261
+ "transcribe": 50360,
262
+ "translate": 50359
263
+ },
264
+ "transformers_version": "4.37.0.dev0"
265
+ }
checkpoint-5000/model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:162e92f863126b6badbad6104326cd65b2875f74bb0340e9b4d53e3d7cc4e83b
3
+ size 4993448880
checkpoint-5000/model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a3553cfea4d55a22ddc4554e797d5f41b8d4ec64eff37633af911cf87725f80
3
+ size 1180663192