Hugging Face's logo

speech31
/
wavlm-large-english-phoneme

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Inference Endpoints
Model card Files Community
2
speech31 commited on
Commit
d448bf4
•
1 Parent(s): 276e64e

initial_commit

Browse files
Files changed (12) hide show
  1. config.json +99 -0
  2. optimizer.pt +3 -0
  3. preprocessor_config.json +9 -0
  4. pytorch_model.bin +3 -0
  5. rng_state.pth +3 -0
  6. scaler.pt +3 -0
  7. scheduler.pt +3 -0
  8. special_tokens_map.json +1 -0
  9. tokenizer_config.json +1 -0
  10. trainer_state.json +523 -0
  11. training_args.bin +3 -0
  12. vocab.json +1 -0
config.json ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "wavlm-english-phoneme",
3
+ "activation_dropout": 0.0,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForCTC"
10
+ ],
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 768,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": false,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "mean",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": true,
48
+ "eos_token_id": 2,
49
+ "feat_extract_activation": "gelu",
50
+ "feat_extract_dropout": 0.0,
51
+ "feat_extract_norm": "layer",
52
+ "feat_proj_dropout": 0.0,
53
+ "feat_quantizer_dropout": 0.0,
54
+ "final_dropout": 0.0,
55
+ "gradient_checkpointing": false,
56
+ "hidden_act": "gelu",
57
+ "hidden_dropout": 0.0,
58
+ "hidden_size": 1024,
59
+ "initializer_range": 0.02,
60
+ "intermediate_size": 4096,
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.0,
63
+ "mask_channel_length": 10,
64
+ "mask_channel_min_space": 1,
65
+ "mask_channel_other": 0.0,
66
+ "mask_channel_prob": 0.0,
67
+ "mask_channel_selection": "static",
68
+ "mask_feature_length": 10,
69
+ "mask_feature_min_masks": 0,
70
+ "mask_feature_prob": 0.0,
71
+ "mask_time_length": 10,
72
+ "mask_time_min_masks": 2,
73
+ "mask_time_min_space": 1,
74
+ "mask_time_other": 0.0,
75
+ "mask_time_prob": 0.05,
76
+ "mask_time_selection": "static",
77
+ "max_bucket_distance": 800,
78
+ "model_type": "wav2vec2",
79
+ "num_adapter_layers": 3,
80
+ "num_attention_heads": 16,
81
+ "num_buckets": 320,
82
+ "num_codevector_groups": 2,
83
+ "num_codevectors_per_group": 320,
84
+ "num_conv_pos_embedding_groups": 16,
85
+ "num_conv_pos_embeddings": 128,
86
+ "num_ctc_classes": 80,
87
+ "num_feat_extract_layers": 7,
88
+ "num_hidden_layers": 24,
89
+ "num_negatives": 100,
90
+ "output_hidden_size": 1024,
91
+ "pad_token_id": 42,
92
+ "proj_codevector_dim": 768,
93
+ "replace_prob": 0.5,
94
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
95
+ "torch_dtype": "float32",
96
+ "transformers_version": "4.11.3",
97
+ "use_weighted_layer_sum": false,
98
+ "vocab_size": 43
99
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:850c2cc5fff84b3019ffe486e962edac5b090d3888910f9d1dd49cd82d8214c1
3
+ size 2490411473
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a54c605fd6bc3bc3a8014b9052bd1a01476ded3d58db0a78b71d0ec72afb9af
3
+ size 1262083263
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02f4ee1516460814d377b0294bd1e309466827778a966acaf0e9faae984b68c4
3
+ size 14567
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e823e293f122f500d37aed0e4cbb93f2648f03767613829430ff15253fe550c6
3
+ size 559
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a974fa47f1eb085a288f566204a61aee1e72de1ef3e8ac82c9975238f32a04cb
3
+ size 623
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "tokenizer_class": "Wav2Vec2CTCTokenizer"}
trainer_state.json ADDED
@@ -0,0 +1,523 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.1746283918619156,
3
+ "best_model_checkpoint": "wavlm-english-phoneme/checkpoint-6500",
4
+ "epoch": 50.0,
5
+ "global_step": 6500,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.77,
12
+ "learning_rate": 2.97e-05,
13
+ "loss": 4.896,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 1.54,
18
+ "learning_rate": 5.97e-05,
19
+ "loss": 2.5812,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 2.31,
24
+ "learning_rate": 8.969999999999998e-05,
25
+ "loss": 2.4853,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 3.08,
30
+ "learning_rate": 0.0001197,
31
+ "loss": 2.2031,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 3.85,
36
+ "learning_rate": 0.00014969999999999998,
37
+ "loss": 1.8587,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 3.85,
42
+ "eval_cer": 0.7105220883534137,
43
+ "eval_loss": 1.3614039421081543,
44
+ "eval_runtime": 21.402,
45
+ "eval_samples_per_second": 21.587,
46
+ "eval_steps_per_second": 2.71,
47
+ "step": 500
48
+ },
49
+ {
50
+ "epoch": 4.62,
51
+ "learning_rate": 0.00017969999999999998,
52
+ "loss": 1.1164,
53
+ "step": 600
54
+ },
55
+ {
56
+ "epoch": 5.38,
57
+ "learning_rate": 0.00020969999999999997,
58
+ "loss": 0.7271,
59
+ "step": 700
60
+ },
61
+ {
62
+ "epoch": 6.15,
63
+ "learning_rate": 0.0002397,
64
+ "loss": 0.5784,
65
+ "step": 800
66
+ },
67
+ {
68
+ "epoch": 6.92,
69
+ "learning_rate": 0.0002697,
70
+ "loss": 0.4487,
71
+ "step": 900
72
+ },
73
+ {
74
+ "epoch": 7.69,
75
+ "learning_rate": 0.00029969999999999997,
76
+ "loss": 0.3517,
77
+ "step": 1000
78
+ },
79
+ {
80
+ "epoch": 7.69,
81
+ "eval_cer": 0.16610441767068274,
82
+ "eval_loss": 0.3352901339530945,
83
+ "eval_runtime": 21.5161,
84
+ "eval_samples_per_second": 21.472,
85
+ "eval_steps_per_second": 2.696,
86
+ "step": 1000
87
+ },
88
+ {
89
+ "epoch": 8.46,
90
+ "learning_rate": 0.00029459999999999995,
91
+ "loss": 0.3089,
92
+ "step": 1100
93
+ },
94
+ {
95
+ "epoch": 9.23,
96
+ "learning_rate": 0.00028914545454545454,
97
+ "loss": 0.2465,
98
+ "step": 1200
99
+ },
100
+ {
101
+ "epoch": 10.0,
102
+ "learning_rate": 0.00028369090909090907,
103
+ "loss": 0.2099,
104
+ "step": 1300
105
+ },
106
+ {
107
+ "epoch": 10.77,
108
+ "learning_rate": 0.0002782363636363636,
109
+ "loss": 0.1649,
110
+ "step": 1400
111
+ },
112
+ {
113
+ "epoch": 11.54,
114
+ "learning_rate": 0.00027278181818181813,
115
+ "loss": 0.1473,
116
+ "step": 1500
117
+ },
118
+ {
119
+ "epoch": 11.54,
120
+ "eval_cer": 0.11325301204819277,
121
+ "eval_loss": 0.23726579546928406,
122
+ "eval_runtime": 34.3746,
123
+ "eval_samples_per_second": 13.44,
124
+ "eval_steps_per_second": 1.687,
125
+ "step": 1500
126
+ },
127
+ {
128
+ "epoch": 12.31,
129
+ "learning_rate": 0.0002673272727272727,
130
+ "loss": 0.1335,
131
+ "step": 1600
132
+ },
133
+ {
134
+ "epoch": 13.08,
135
+ "learning_rate": 0.00026187272727272725,
136
+ "loss": 0.118,
137
+ "step": 1700
138
+ },
139
+ {
140
+ "epoch": 13.85,
141
+ "learning_rate": 0.00025641818181818183,
142
+ "loss": 0.1083,
143
+ "step": 1800
144
+ },
145
+ {
146
+ "epoch": 14.62,
147
+ "learning_rate": 0.00025096363636363636,
148
+ "loss": 0.106,
149
+ "step": 1900
150
+ },
151
+ {
152
+ "epoch": 15.38,
153
+ "learning_rate": 0.0002455090909090909,
154
+ "loss": 0.0799,
155
+ "step": 2000
156
+ },
157
+ {
158
+ "epoch": 15.38,
159
+ "eval_cer": 0.09397590361445783,
160
+ "eval_loss": 0.20781992375850677,
161
+ "eval_runtime": 21.3113,
162
+ "eval_samples_per_second": 21.679,
163
+ "eval_steps_per_second": 2.722,
164
+ "step": 2000
165
+ },
166
+ {
167
+ "epoch": 16.15,
168
+ "learning_rate": 0.00024005454545454545,
169
+ "loss": 0.0767,
170
+ "step": 2100
171
+ },
172
+ {
173
+ "epoch": 16.92,
174
+ "learning_rate": 0.00023459999999999998,
175
+ "loss": 0.0761,
176
+ "step": 2200
177
+ },
178
+ {
179
+ "epoch": 17.69,
180
+ "learning_rate": 0.00022914545454545454,
181
+ "loss": 0.0705,
182
+ "step": 2300
183
+ },
184
+ {
185
+ "epoch": 18.46,
186
+ "learning_rate": 0.00022369090909090907,
187
+ "loss": 0.0619,
188
+ "step": 2400
189
+ },
190
+ {
191
+ "epoch": 19.23,
192
+ "learning_rate": 0.00021823636363636363,
193
+ "loss": 0.0634,
194
+ "step": 2500
195
+ },
196
+ {
197
+ "epoch": 19.23,
198
+ "eval_cer": 0.08538152610441767,
199
+ "eval_loss": 0.20878814160823822,
200
+ "eval_runtime": 20.9398,
201
+ "eval_samples_per_second": 22.063,
202
+ "eval_steps_per_second": 2.77,
203
+ "step": 2500
204
+ },
205
+ {
206
+ "epoch": 20.0,
207
+ "learning_rate": 0.00021278181818181816,
208
+ "loss": 0.0556,
209
+ "step": 2600
210
+ },
211
+ {
212
+ "epoch": 20.77,
213
+ "learning_rate": 0.00020732727272727272,
214
+ "loss": 0.0562,
215
+ "step": 2700
216
+ },
217
+ {
218
+ "epoch": 21.54,
219
+ "learning_rate": 0.00020187272727272725,
220
+ "loss": 0.0556,
221
+ "step": 2800
222
+ },
223
+ {
224
+ "epoch": 22.31,
225
+ "learning_rate": 0.0001964181818181818,
226
+ "loss": 0.0483,
227
+ "step": 2900
228
+ },
229
+ {
230
+ "epoch": 23.08,
231
+ "learning_rate": 0.00019096363636363634,
232
+ "loss": 0.0472,
233
+ "step": 3000
234
+ },
235
+ {
236
+ "epoch": 23.08,
237
+ "eval_cer": 0.0767871485943775,
238
+ "eval_loss": 0.18513090908527374,
239
+ "eval_runtime": 21.0154,
240
+ "eval_samples_per_second": 21.984,
241
+ "eval_steps_per_second": 2.76,
242
+ "step": 3000
243
+ },
244
+ {
245
+ "epoch": 23.85,
246
+ "learning_rate": 0.00018550909090909087,
247
+ "loss": 0.044,
248
+ "step": 3100
249
+ },
250
+ {
251
+ "epoch": 24.62,
252
+ "learning_rate": 0.00018005454545454543,
253
+ "loss": 0.0409,
254
+ "step": 3200
255
+ },
256
+ {
257
+ "epoch": 25.38,
258
+ "learning_rate": 0.00017459999999999996,
259
+ "loss": 0.0386,
260
+ "step": 3300
261
+ },
262
+ {
263
+ "epoch": 26.15,
264
+ "learning_rate": 0.00016914545454545452,
265
+ "loss": 0.0397,
266
+ "step": 3400
267
+ },
268
+ {
269
+ "epoch": 26.92,
270
+ "learning_rate": 0.00016369090909090905,
271
+ "loss": 0.0337,
272
+ "step": 3500
273
+ },
274
+ {
275
+ "epoch": 26.92,
276
+ "eval_cer": 0.0757429718875502,
277
+ "eval_loss": 0.2069997787475586,
278
+ "eval_runtime": 21.3619,
279
+ "eval_samples_per_second": 21.627,
280
+ "eval_steps_per_second": 2.715,
281
+ "step": 3500
282
+ },
283
+ {
284
+ "epoch": 27.69,
285
+ "learning_rate": 0.00015823636363636364,
286
+ "loss": 0.0368,
287
+ "step": 3600
288
+ },
289
+ {
290
+ "epoch": 28.46,
291
+ "learning_rate": 0.0001527818181818182,
292
+ "loss": 0.0329,
293
+ "step": 3700
294
+ },
295
+ {
296
+ "epoch": 29.23,
297
+ "learning_rate": 0.0001473272727272727,
298
+ "loss": 0.0306,
299
+ "step": 3800
300
+ },
301
+ {
302
+ "epoch": 30.0,
303
+ "learning_rate": 0.00014187272727272726,
304
+ "loss": 0.0298,
305
+ "step": 3900
306
+ },
307
+ {
308
+ "epoch": 30.77,
309
+ "learning_rate": 0.0001364181818181818,
310
+ "loss": 0.0283,
311
+ "step": 4000
312
+ },
313
+ {
314
+ "epoch": 30.77,
315
+ "eval_cer": 0.07180722891566266,
316
+ "eval_loss": 0.19239871203899384,
317
+ "eval_runtime": 34.8631,
318
+ "eval_samples_per_second": 13.252,
319
+ "eval_steps_per_second": 1.664,
320
+ "step": 4000
321
+ },
322
+ {
323
+ "epoch": 31.54,
324
+ "learning_rate": 0.00013096363636363635,
325
+ "loss": 0.0256,
326
+ "step": 4100
327
+ },
328
+ {
329
+ "epoch": 32.31,
330
+ "learning_rate": 0.0001255090909090909,
331
+ "loss": 0.0257,
332
+ "step": 4200
333
+ },
334
+ {
335
+ "epoch": 33.08,
336
+ "learning_rate": 0.00012005454545454545,
337
+ "loss": 0.0238,
338
+ "step": 4300
339
+ },
340
+ {
341
+ "epoch": 33.85,
342
+ "learning_rate": 0.0001146,
343
+ "loss": 0.0228,
344
+ "step": 4400
345
+ },
346
+ {
347
+ "epoch": 34.62,
348
+ "learning_rate": 0.00010919999999999998,
349
+ "loss": 0.0242,
350
+ "step": 4500
351
+ },
352
+ {
353
+ "epoch": 34.62,
354
+ "eval_cer": 0.0668273092369478,
355
+ "eval_loss": 0.18180887401103973,
356
+ "eval_runtime": 20.8913,
357
+ "eval_samples_per_second": 22.115,
358
+ "eval_steps_per_second": 2.776,
359
+ "step": 4500
360
+ },
361
+ {
362
+ "epoch": 35.38,
363
+ "learning_rate": 0.00010374545454545453,
364
+ "loss": 0.0209,
365
+ "step": 4600
366
+ },
367
+ {
368
+ "epoch": 36.15,
369
+ "learning_rate": 9.829090909090907e-05,
370
+ "loss": 0.0192,
371
+ "step": 4700
372
+ },
373
+ {
374
+ "epoch": 36.92,
375
+ "learning_rate": 9.283636363636363e-05,
376
+ "loss": 0.0171,
377
+ "step": 4800
378
+ },
379
+ {
380
+ "epoch": 37.69,
381
+ "learning_rate": 8.738181818181818e-05,
382
+ "loss": 0.0205,
383
+ "step": 4900
384
+ },
385
+ {
386
+ "epoch": 38.46,
387
+ "learning_rate": 8.192727272727272e-05,
388
+ "loss": 0.0187,
389
+ "step": 5000
390
+ },
391
+ {
392
+ "epoch": 38.46,
393
+ "eval_cer": 0.06730923694779116,
394
+ "eval_loss": 0.19057846069335938,
395
+ "eval_runtime": 27.3046,
396
+ "eval_samples_per_second": 16.92,
397
+ "eval_steps_per_second": 2.124,
398
+ "step": 5000
399
+ },
400
+ {
401
+ "epoch": 39.23,
402
+ "learning_rate": 7.647272727272727e-05,
403
+ "loss": 0.0159,
404
+ "step": 5100
405
+ },
406
+ {
407
+ "epoch": 40.0,
408
+ "learning_rate": 7.101818181818181e-05,
409
+ "loss": 0.0154,
410
+ "step": 5200
411
+ },
412
+ {
413
+ "epoch": 40.77,
414
+ "learning_rate": 6.556363636363636e-05,
415
+ "loss": 0.0165,
416
+ "step": 5300
417
+ },
418
+ {
419
+ "epoch": 41.54,
420
+ "learning_rate": 6.01090909090909e-05,
421
+ "loss": 0.0135,
422
+ "step": 5400
423
+ },
424
+ {
425
+ "epoch": 42.31,
426
+ "learning_rate": 5.465454545454545e-05,
427
+ "loss": 0.0156,
428
+ "step": 5500
429
+ },
430
+ {
431
+ "epoch": 42.31,
432
+ "eval_cer": 0.06409638554216868,
433
+ "eval_loss": 0.17488864064216614,
434
+ "eval_runtime": 20.9298,
435
+ "eval_samples_per_second": 22.074,
436
+ "eval_steps_per_second": 2.771,
437
+ "step": 5500
438
+ },
439
+ {
440
+ "epoch": 43.08,
441
+ "learning_rate": 4.9199999999999997e-05,
442
+ "loss": 0.0137,
443
+ "step": 5600
444
+ },
445
+ {
446
+ "epoch": 43.85,
447
+ "learning_rate": 4.374545454545454e-05,
448
+ "loss": 0.0138,
449
+ "step": 5700
450
+ },
451
+ {
452
+ "epoch": 44.62,
453
+ "learning_rate": 3.8290909090909086e-05,
454
+ "loss": 0.0117,
455
+ "step": 5800
456
+ },
457
+ {
458
+ "epoch": 45.38,
459
+ "learning_rate": 3.283636363636364e-05,
460
+ "loss": 0.0132,
461
+ "step": 5900
462
+ },
463
+ {
464
+ "epoch": 46.15,
465
+ "learning_rate": 2.738181818181818e-05,
466
+ "loss": 0.0107,
467
+ "step": 6000
468
+ },
469
+ {
470
+ "epoch": 46.15,
471
+ "eval_cer": 0.060481927710843375,
472
+ "eval_loss": 0.17818936705589294,
473
+ "eval_runtime": 20.9013,
474
+ "eval_samples_per_second": 22.104,
475
+ "eval_steps_per_second": 2.775,
476
+ "step": 6000
477
+ },
478
+ {
479
+ "epoch": 46.92,
480
+ "learning_rate": 2.1927272727272727e-05,
481
+ "loss": 0.0124,
482
+ "step": 6100
483
+ },
484
+ {
485
+ "epoch": 47.69,
486
+ "learning_rate": 1.647272727272727e-05,
487
+ "loss": 0.0119,
488
+ "step": 6200
489
+ },
490
+ {
491
+ "epoch": 48.46,
492
+ "learning_rate": 1.1018181818181816e-05,
493
+ "loss": 0.0097,
494
+ "step": 6300
495
+ },
496
+ {
497
+ "epoch": 49.23,
498
+ "learning_rate": 5.563636363636363e-06,
499
+ "loss": 0.0105,
500
+ "step": 6400
501
+ },
502
+ {
503
+ "epoch": 50.0,
504
+ "learning_rate": 1.0909090909090908e-07,
505
+ "loss": 0.0099,
506
+ "step": 6500
507
+ },
508
+ {
509
+ "epoch": 50.0,
510
+ "eval_cer": 0.06088353413654619,
511
+ "eval_loss": 0.1746283918619156,
512
+ "eval_runtime": 27.3261,
513
+ "eval_samples_per_second": 16.907,
514
+ "eval_steps_per_second": 2.123,
515
+ "step": 6500
516
+ }
517
+ ],
518
+ "max_steps": 6500,
519
+ "num_train_epochs": 50,
520
+ "total_flos": 1.985767056261625e+19,
521
+ "trial_name": null,
522
+ "trial_params": null
523
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77f8a4e967a3b30b8f18e337b726e1abc24067af29f74d50b5398c3fef643022
3
+ size 2863
vocab.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"t͡ʃ": 1, "ɹ̩": 2, "æ": 3, "k": 4, "ej": 5, "m": 6, "ow": 7, "ʌ": 8, "z": 9, "u": 10, "ð": 11, "ə": 12, "f": 13, "ɛ": 14, "b": 15, "oj": 16, "d": 17, "θ": 18, "v": 19, "ɔ": 20, "ʊ": 21, "ɪ": 22, "aj": 23, "d͡ʒ": 24, "w": 25, "h": 26, "t": 27, "l": 28, "p": 29, "j": 30, "ɹ": 31, "ʃ": 32, "i": 33, "ʒ": 34, "ŋ": 35, "aw": 36, "n": 37, "ɑ": 38, "ɡ": 39, "s": 40, "|": 0, "[UNK]": 41, "[PAD]": 42}