MMars commited on
Commit
e79d273
1 Parent(s): b0bdbb3

Upload 7 files

Browse files
Files changed (7) hide show
  1. config.json +42 -0
  2. optimizer.pt +3 -0
  3. rng_state.pth +3 -0
  4. scaler.pt +3 -0
  5. scheduler.pt +3 -0
  6. trainer_state.json +1012 -0
  7. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "architectures": [
6
+ "WhisperForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "begin_suppress_tokens": [
10
+ 220,
11
+ 50257
12
+ ],
13
+ "bos_token_id": 50257,
14
+ "d_model": 768,
15
+ "decoder_attention_heads": 12,
16
+ "decoder_ffn_dim": 3072,
17
+ "decoder_layerdrop": 0.0,
18
+ "decoder_layers": 12,
19
+ "decoder_start_token_id": 50258,
20
+ "dropout": 0.0,
21
+ "encoder_attention_heads": 12,
22
+ "encoder_ffn_dim": 3072,
23
+ "encoder_layerdrop": 0.0,
24
+ "encoder_layers": 12,
25
+ "eos_token_id": 50257,
26
+ "forced_decoder_ids": null,
27
+ "init_std": 0.02,
28
+ "is_encoder_decoder": true,
29
+ "max_length": 448,
30
+ "max_source_positions": 1500,
31
+ "max_target_positions": 448,
32
+ "model_type": "whisper",
33
+ "num_hidden_layers": 12,
34
+ "num_mel_bins": 80,
35
+ "pad_token_id": 50257,
36
+ "scale_embedding": false,
37
+ "suppress_tokens": [],
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.26.0.dev0",
40
+ "use_cache": true,
41
+ "vocab_size": 51865
42
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:683d1e94d46b9bffdbd9a93f5dbf0ea4e9976c795713573b6b190587820c8f5d
3
+ size 1934161093
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de2f093149e1c490716f8f2972665fc399418947c53e76711433b3ac36b93b4d
3
+ size 14511
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44f7e2be3b4731e58cc18a29e9923afb5d0880c79f7d50afe3120769ef965929
3
+ size 557
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:277cda986867c60d4c536c9fb0285b0c0f822402d1380367bb87b63927a236fb
3
+ size 627
trainer_state.json ADDED
@@ -0,0 +1,1012 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 44.976586172068686,
3
+ "best_model_checkpoint": "./whisper-small-ar/whisper-small-ar/checkpoint-4000",
4
+ "epoch": 1.6625103906899419,
5
+ "global_step": 4000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 4.2000000000000006e-07,
13
+ "loss": 3.262,
14
+ "step": 25
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 9.200000000000001e-07,
19
+ "loss": 2.5217,
20
+ "step": 50
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 1.42e-06,
25
+ "loss": 1.7958,
26
+ "step": 75
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 1.9200000000000003e-06,
31
+ "loss": 1.2313,
32
+ "step": 100
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 2.42e-06,
37
+ "loss": 1.0969,
38
+ "step": 125
39
+ },
40
+ {
41
+ "epoch": 0.06,
42
+ "learning_rate": 2.92e-06,
43
+ "loss": 0.9611,
44
+ "step": 150
45
+ },
46
+ {
47
+ "epoch": 0.07,
48
+ "learning_rate": 3.4200000000000007e-06,
49
+ "loss": 0.9175,
50
+ "step": 175
51
+ },
52
+ {
53
+ "epoch": 0.08,
54
+ "learning_rate": 3.920000000000001e-06,
55
+ "loss": 0.819,
56
+ "step": 200
57
+ },
58
+ {
59
+ "epoch": 0.09,
60
+ "learning_rate": 4.42e-06,
61
+ "loss": 0.7742,
62
+ "step": 225
63
+ },
64
+ {
65
+ "epoch": 0.1,
66
+ "learning_rate": 4.92e-06,
67
+ "loss": 0.6782,
68
+ "step": 250
69
+ },
70
+ {
71
+ "epoch": 0.11,
72
+ "learning_rate": 5.420000000000001e-06,
73
+ "loss": 0.5778,
74
+ "step": 275
75
+ },
76
+ {
77
+ "epoch": 0.12,
78
+ "learning_rate": 5.92e-06,
79
+ "loss": 0.456,
80
+ "step": 300
81
+ },
82
+ {
83
+ "epoch": 0.14,
84
+ "learning_rate": 6.42e-06,
85
+ "loss": 0.4114,
86
+ "step": 325
87
+ },
88
+ {
89
+ "epoch": 0.15,
90
+ "learning_rate": 6.92e-06,
91
+ "loss": 0.4197,
92
+ "step": 350
93
+ },
94
+ {
95
+ "epoch": 0.16,
96
+ "learning_rate": 7.420000000000001e-06,
97
+ "loss": 0.3929,
98
+ "step": 375
99
+ },
100
+ {
101
+ "epoch": 0.17,
102
+ "learning_rate": 7.92e-06,
103
+ "loss": 0.4104,
104
+ "step": 400
105
+ },
106
+ {
107
+ "epoch": 0.18,
108
+ "learning_rate": 8.42e-06,
109
+ "loss": 0.4269,
110
+ "step": 425
111
+ },
112
+ {
113
+ "epoch": 0.19,
114
+ "learning_rate": 8.920000000000001e-06,
115
+ "loss": 0.4303,
116
+ "step": 450
117
+ },
118
+ {
119
+ "epoch": 0.2,
120
+ "learning_rate": 9.42e-06,
121
+ "loss": 0.3523,
122
+ "step": 475
123
+ },
124
+ {
125
+ "epoch": 0.21,
126
+ "learning_rate": 9.920000000000002e-06,
127
+ "loss": 0.3633,
128
+ "step": 500
129
+ },
130
+ {
131
+ "epoch": 0.22,
132
+ "learning_rate": 9.940000000000001e-06,
133
+ "loss": 0.361,
134
+ "step": 525
135
+ },
136
+ {
137
+ "epoch": 0.23,
138
+ "learning_rate": 9.86857142857143e-06,
139
+ "loss": 0.3368,
140
+ "step": 550
141
+ },
142
+ {
143
+ "epoch": 0.24,
144
+ "learning_rate": 9.797142857142858e-06,
145
+ "loss": 0.329,
146
+ "step": 575
147
+ },
148
+ {
149
+ "epoch": 0.25,
150
+ "learning_rate": 9.725714285714287e-06,
151
+ "loss": 0.3671,
152
+ "step": 600
153
+ },
154
+ {
155
+ "epoch": 0.26,
156
+ "learning_rate": 9.654285714285716e-06,
157
+ "loss": 0.3832,
158
+ "step": 625
159
+ },
160
+ {
161
+ "epoch": 0.27,
162
+ "learning_rate": 9.582857142857143e-06,
163
+ "loss": 0.3494,
164
+ "step": 650
165
+ },
166
+ {
167
+ "epoch": 0.28,
168
+ "learning_rate": 9.511428571428572e-06,
169
+ "loss": 0.3334,
170
+ "step": 675
171
+ },
172
+ {
173
+ "epoch": 0.29,
174
+ "learning_rate": 9.440000000000001e-06,
175
+ "loss": 0.3466,
176
+ "step": 700
177
+ },
178
+ {
179
+ "epoch": 0.3,
180
+ "learning_rate": 9.368571428571428e-06,
181
+ "loss": 0.3464,
182
+ "step": 725
183
+ },
184
+ {
185
+ "epoch": 0.31,
186
+ "learning_rate": 9.297142857142857e-06,
187
+ "loss": 0.3481,
188
+ "step": 750
189
+ },
190
+ {
191
+ "epoch": 0.32,
192
+ "learning_rate": 9.225714285714286e-06,
193
+ "loss": 0.3188,
194
+ "step": 775
195
+ },
196
+ {
197
+ "epoch": 0.33,
198
+ "learning_rate": 9.154285714285715e-06,
199
+ "loss": 0.3396,
200
+ "step": 800
201
+ },
202
+ {
203
+ "epoch": 0.34,
204
+ "learning_rate": 9.082857142857143e-06,
205
+ "loss": 0.3187,
206
+ "step": 825
207
+ },
208
+ {
209
+ "epoch": 0.35,
210
+ "learning_rate": 9.011428571428572e-06,
211
+ "loss": 0.3079,
212
+ "step": 850
213
+ },
214
+ {
215
+ "epoch": 0.36,
216
+ "learning_rate": 8.94e-06,
217
+ "loss": 0.311,
218
+ "step": 875
219
+ },
220
+ {
221
+ "epoch": 0.37,
222
+ "learning_rate": 8.86857142857143e-06,
223
+ "loss": 0.3162,
224
+ "step": 900
225
+ },
226
+ {
227
+ "epoch": 0.38,
228
+ "learning_rate": 8.797142857142857e-06,
229
+ "loss": 0.3474,
230
+ "step": 925
231
+ },
232
+ {
233
+ "epoch": 0.39,
234
+ "learning_rate": 8.725714285714286e-06,
235
+ "loss": 0.3154,
236
+ "step": 950
237
+ },
238
+ {
239
+ "epoch": 0.41,
240
+ "learning_rate": 8.654285714285715e-06,
241
+ "loss": 0.3315,
242
+ "step": 975
243
+ },
244
+ {
245
+ "epoch": 0.42,
246
+ "learning_rate": 8.582857142857144e-06,
247
+ "loss": 0.2811,
248
+ "step": 1000
249
+ },
250
+ {
251
+ "epoch": 0.42,
252
+ "eval_loss": 0.3930181860923767,
253
+ "eval_runtime": 2355.0366,
254
+ "eval_samples_per_second": 4.433,
255
+ "eval_steps_per_second": 0.554,
256
+ "eval_wer": 53.77834909558351,
257
+ "step": 1000
258
+ },
259
+ {
260
+ "epoch": 0.43,
261
+ "learning_rate": 8.511428571428571e-06,
262
+ "loss": 0.2884,
263
+ "step": 1025
264
+ },
265
+ {
266
+ "epoch": 0.44,
267
+ "learning_rate": 8.44e-06,
268
+ "loss": 0.3166,
269
+ "step": 1050
270
+ },
271
+ {
272
+ "epoch": 0.45,
273
+ "learning_rate": 8.36857142857143e-06,
274
+ "loss": 0.2723,
275
+ "step": 1075
276
+ },
277
+ {
278
+ "epoch": 0.46,
279
+ "learning_rate": 8.297142857142859e-06,
280
+ "loss": 0.2841,
281
+ "step": 1100
282
+ },
283
+ {
284
+ "epoch": 0.47,
285
+ "learning_rate": 8.225714285714288e-06,
286
+ "loss": 0.3115,
287
+ "step": 1125
288
+ },
289
+ {
290
+ "epoch": 0.48,
291
+ "learning_rate": 8.154285714285715e-06,
292
+ "loss": 0.3161,
293
+ "step": 1150
294
+ },
295
+ {
296
+ "epoch": 0.49,
297
+ "learning_rate": 8.082857142857144e-06,
298
+ "loss": 0.2771,
299
+ "step": 1175
300
+ },
301
+ {
302
+ "epoch": 0.5,
303
+ "learning_rate": 8.011428571428573e-06,
304
+ "loss": 0.2826,
305
+ "step": 1200
306
+ },
307
+ {
308
+ "epoch": 0.51,
309
+ "learning_rate": 7.94e-06,
310
+ "loss": 0.3057,
311
+ "step": 1225
312
+ },
313
+ {
314
+ "epoch": 0.52,
315
+ "learning_rate": 7.86857142857143e-06,
316
+ "loss": 0.2849,
317
+ "step": 1250
318
+ },
319
+ {
320
+ "epoch": 0.53,
321
+ "learning_rate": 7.797142857142858e-06,
322
+ "loss": 0.2959,
323
+ "step": 1275
324
+ },
325
+ {
326
+ "epoch": 0.54,
327
+ "learning_rate": 7.725714285714286e-06,
328
+ "loss": 0.3015,
329
+ "step": 1300
330
+ },
331
+ {
332
+ "epoch": 0.55,
333
+ "learning_rate": 7.654285714285715e-06,
334
+ "loss": 0.2686,
335
+ "step": 1325
336
+ },
337
+ {
338
+ "epoch": 0.56,
339
+ "learning_rate": 7.5828571428571444e-06,
340
+ "loss": 0.2554,
341
+ "step": 1350
342
+ },
343
+ {
344
+ "epoch": 0.57,
345
+ "learning_rate": 7.511428571428572e-06,
346
+ "loss": 0.3009,
347
+ "step": 1375
348
+ },
349
+ {
350
+ "epoch": 0.58,
351
+ "learning_rate": 7.440000000000001e-06,
352
+ "loss": 0.2896,
353
+ "step": 1400
354
+ },
355
+ {
356
+ "epoch": 0.59,
357
+ "learning_rate": 7.36857142857143e-06,
358
+ "loss": 0.271,
359
+ "step": 1425
360
+ },
361
+ {
362
+ "epoch": 0.6,
363
+ "learning_rate": 7.297142857142858e-06,
364
+ "loss": 0.2695,
365
+ "step": 1450
366
+ },
367
+ {
368
+ "epoch": 0.61,
369
+ "learning_rate": 7.225714285714286e-06,
370
+ "loss": 0.287,
371
+ "step": 1475
372
+ },
373
+ {
374
+ "epoch": 0.62,
375
+ "learning_rate": 7.154285714285715e-06,
376
+ "loss": 0.2473,
377
+ "step": 1500
378
+ },
379
+ {
380
+ "epoch": 0.63,
381
+ "learning_rate": 7.082857142857143e-06,
382
+ "loss": 0.2454,
383
+ "step": 1525
384
+ },
385
+ {
386
+ "epoch": 0.64,
387
+ "learning_rate": 7.011428571428572e-06,
388
+ "loss": 0.2952,
389
+ "step": 1550
390
+ },
391
+ {
392
+ "epoch": 0.65,
393
+ "learning_rate": 6.9400000000000005e-06,
394
+ "loss": 0.2765,
395
+ "step": 1575
396
+ },
397
+ {
398
+ "epoch": 0.67,
399
+ "learning_rate": 6.868571428571429e-06,
400
+ "loss": 0.2615,
401
+ "step": 1600
402
+ },
403
+ {
404
+ "epoch": 0.68,
405
+ "learning_rate": 6.797142857142858e-06,
406
+ "loss": 0.26,
407
+ "step": 1625
408
+ },
409
+ {
410
+ "epoch": 0.69,
411
+ "learning_rate": 6.725714285714287e-06,
412
+ "loss": 0.2687,
413
+ "step": 1650
414
+ },
415
+ {
416
+ "epoch": 0.7,
417
+ "learning_rate": 6.654285714285716e-06,
418
+ "loss": 0.2568,
419
+ "step": 1675
420
+ },
421
+ {
422
+ "epoch": 0.71,
423
+ "learning_rate": 6.582857142857143e-06,
424
+ "loss": 0.2634,
425
+ "step": 1700
426
+ },
427
+ {
428
+ "epoch": 0.72,
429
+ "learning_rate": 6.511428571428572e-06,
430
+ "loss": 0.2655,
431
+ "step": 1725
432
+ },
433
+ {
434
+ "epoch": 0.73,
435
+ "learning_rate": 6.440000000000001e-06,
436
+ "loss": 0.2457,
437
+ "step": 1750
438
+ },
439
+ {
440
+ "epoch": 0.74,
441
+ "learning_rate": 6.368571428571429e-06,
442
+ "loss": 0.2427,
443
+ "step": 1775
444
+ },
445
+ {
446
+ "epoch": 0.75,
447
+ "learning_rate": 6.297142857142857e-06,
448
+ "loss": 0.236,
449
+ "step": 1800
450
+ },
451
+ {
452
+ "epoch": 0.76,
453
+ "learning_rate": 6.225714285714286e-06,
454
+ "loss": 0.2553,
455
+ "step": 1825
456
+ },
457
+ {
458
+ "epoch": 0.77,
459
+ "learning_rate": 6.1542857142857145e-06,
460
+ "loss": 0.2521,
461
+ "step": 1850
462
+ },
463
+ {
464
+ "epoch": 0.78,
465
+ "learning_rate": 6.0828571428571435e-06,
466
+ "loss": 0.2536,
467
+ "step": 1875
468
+ },
469
+ {
470
+ "epoch": 0.79,
471
+ "learning_rate": 6.011428571428572e-06,
472
+ "loss": 0.2696,
473
+ "step": 1900
474
+ },
475
+ {
476
+ "epoch": 0.8,
477
+ "learning_rate": 5.94e-06,
478
+ "loss": 0.2436,
479
+ "step": 1925
480
+ },
481
+ {
482
+ "epoch": 0.81,
483
+ "learning_rate": 5.868571428571429e-06,
484
+ "loss": 0.2197,
485
+ "step": 1950
486
+ },
487
+ {
488
+ "epoch": 0.82,
489
+ "learning_rate": 5.797142857142858e-06,
490
+ "loss": 0.2526,
491
+ "step": 1975
492
+ },
493
+ {
494
+ "epoch": 0.83,
495
+ "learning_rate": 5.725714285714287e-06,
496
+ "loss": 0.2356,
497
+ "step": 2000
498
+ },
499
+ {
500
+ "epoch": 0.83,
501
+ "eval_loss": 0.3487941324710846,
502
+ "eval_runtime": 2297.3957,
503
+ "eval_samples_per_second": 4.544,
504
+ "eval_steps_per_second": 0.568,
505
+ "eval_wer": 47.79359103847213,
506
+ "step": 2000
507
+ },
508
+ {
509
+ "epoch": 0.84,
510
+ "learning_rate": 5.654285714285714e-06,
511
+ "loss": 0.2516,
512
+ "step": 2025
513
+ },
514
+ {
515
+ "epoch": 0.85,
516
+ "learning_rate": 5.582857142857143e-06,
517
+ "loss": 0.2467,
518
+ "step": 2050
519
+ },
520
+ {
521
+ "epoch": 0.86,
522
+ "learning_rate": 5.511428571428572e-06,
523
+ "loss": 0.23,
524
+ "step": 2075
525
+ },
526
+ {
527
+ "epoch": 0.87,
528
+ "learning_rate": 5.4400000000000004e-06,
529
+ "loss": 0.2571,
530
+ "step": 2100
531
+ },
532
+ {
533
+ "epoch": 0.88,
534
+ "learning_rate": 5.368571428571429e-06,
535
+ "loss": 0.2424,
536
+ "step": 2125
537
+ },
538
+ {
539
+ "epoch": 0.89,
540
+ "learning_rate": 5.297142857142858e-06,
541
+ "loss": 0.2402,
542
+ "step": 2150
543
+ },
544
+ {
545
+ "epoch": 0.9,
546
+ "learning_rate": 5.225714285714286e-06,
547
+ "loss": 0.2466,
548
+ "step": 2175
549
+ },
550
+ {
551
+ "epoch": 0.91,
552
+ "learning_rate": 5.154285714285715e-06,
553
+ "loss": 0.2415,
554
+ "step": 2200
555
+ },
556
+ {
557
+ "epoch": 0.92,
558
+ "learning_rate": 5.082857142857144e-06,
559
+ "loss": 0.2228,
560
+ "step": 2225
561
+ },
562
+ {
563
+ "epoch": 0.94,
564
+ "learning_rate": 5.011428571428571e-06,
565
+ "loss": 0.2359,
566
+ "step": 2250
567
+ },
568
+ {
569
+ "epoch": 0.95,
570
+ "learning_rate": 4.94e-06,
571
+ "loss": 0.237,
572
+ "step": 2275
573
+ },
574
+ {
575
+ "epoch": 0.96,
576
+ "learning_rate": 4.868571428571429e-06,
577
+ "loss": 0.232,
578
+ "step": 2300
579
+ },
580
+ {
581
+ "epoch": 0.97,
582
+ "learning_rate": 4.797142857142857e-06,
583
+ "loss": 0.246,
584
+ "step": 2325
585
+ },
586
+ {
587
+ "epoch": 0.98,
588
+ "learning_rate": 4.725714285714286e-06,
589
+ "loss": 0.2365,
590
+ "step": 2350
591
+ },
592
+ {
593
+ "epoch": 0.99,
594
+ "learning_rate": 4.6542857142857145e-06,
595
+ "loss": 0.2492,
596
+ "step": 2375
597
+ },
598
+ {
599
+ "epoch": 1.0,
600
+ "learning_rate": 4.5828571428571435e-06,
601
+ "loss": 0.2309,
602
+ "step": 2400
603
+ },
604
+ {
605
+ "epoch": 1.01,
606
+ "learning_rate": 4.514285714285714e-06,
607
+ "loss": 0.2007,
608
+ "step": 2425
609
+ },
610
+ {
611
+ "epoch": 1.02,
612
+ "learning_rate": 4.442857142857143e-06,
613
+ "loss": 0.1757,
614
+ "step": 2450
615
+ },
616
+ {
617
+ "epoch": 1.03,
618
+ "learning_rate": 4.371428571428572e-06,
619
+ "loss": 0.1718,
620
+ "step": 2475
621
+ },
622
+ {
623
+ "epoch": 1.04,
624
+ "learning_rate": 4.3e-06,
625
+ "loss": 0.1514,
626
+ "step": 2500
627
+ },
628
+ {
629
+ "epoch": 1.05,
630
+ "learning_rate": 4.228571428571429e-06,
631
+ "loss": 0.1563,
632
+ "step": 2525
633
+ },
634
+ {
635
+ "epoch": 1.06,
636
+ "learning_rate": 4.1571428571428575e-06,
637
+ "loss": 0.1723,
638
+ "step": 2550
639
+ },
640
+ {
641
+ "epoch": 1.07,
642
+ "learning_rate": 4.0857142857142865e-06,
643
+ "loss": 0.163,
644
+ "step": 2575
645
+ },
646
+ {
647
+ "epoch": 1.08,
648
+ "learning_rate": 4.014285714285715e-06,
649
+ "loss": 0.1618,
650
+ "step": 2600
651
+ },
652
+ {
653
+ "epoch": 1.09,
654
+ "learning_rate": 3.942857142857143e-06,
655
+ "loss": 0.1746,
656
+ "step": 2625
657
+ },
658
+ {
659
+ "epoch": 1.1,
660
+ "learning_rate": 3.871428571428572e-06,
661
+ "loss": 0.1623,
662
+ "step": 2650
663
+ },
664
+ {
665
+ "epoch": 1.11,
666
+ "learning_rate": 3.8000000000000005e-06,
667
+ "loss": 0.1673,
668
+ "step": 2675
669
+ },
670
+ {
671
+ "epoch": 1.12,
672
+ "learning_rate": 3.7285714285714286e-06,
673
+ "loss": 0.1554,
674
+ "step": 2700
675
+ },
676
+ {
677
+ "epoch": 1.13,
678
+ "learning_rate": 3.6571428571428576e-06,
679
+ "loss": 0.1571,
680
+ "step": 2725
681
+ },
682
+ {
683
+ "epoch": 1.14,
684
+ "learning_rate": 3.5857142857142862e-06,
685
+ "loss": 0.1643,
686
+ "step": 2750
687
+ },
688
+ {
689
+ "epoch": 1.15,
690
+ "learning_rate": 3.5142857142857144e-06,
691
+ "loss": 0.1504,
692
+ "step": 2775
693
+ },
694
+ {
695
+ "epoch": 1.16,
696
+ "learning_rate": 3.4428571428571434e-06,
697
+ "loss": 0.1582,
698
+ "step": 2800
699
+ },
700
+ {
701
+ "epoch": 1.17,
702
+ "learning_rate": 3.3714285714285716e-06,
703
+ "loss": 0.1427,
704
+ "step": 2825
705
+ },
706
+ {
707
+ "epoch": 1.18,
708
+ "learning_rate": 3.3000000000000006e-06,
709
+ "loss": 0.1559,
710
+ "step": 2850
711
+ },
712
+ {
713
+ "epoch": 1.19,
714
+ "learning_rate": 3.2285714285714288e-06,
715
+ "loss": 0.1617,
716
+ "step": 2875
717
+ },
718
+ {
719
+ "epoch": 1.21,
720
+ "learning_rate": 3.1571428571428573e-06,
721
+ "loss": 0.156,
722
+ "step": 2900
723
+ },
724
+ {
725
+ "epoch": 1.22,
726
+ "learning_rate": 3.085714285714286e-06,
727
+ "loss": 0.1486,
728
+ "step": 2925
729
+ },
730
+ {
731
+ "epoch": 1.23,
732
+ "learning_rate": 3.0142857142857145e-06,
733
+ "loss": 0.1604,
734
+ "step": 2950
735
+ },
736
+ {
737
+ "epoch": 1.24,
738
+ "learning_rate": 2.9428571428571427e-06,
739
+ "loss": 0.1587,
740
+ "step": 2975
741
+ },
742
+ {
743
+ "epoch": 1.25,
744
+ "learning_rate": 2.8714285714285717e-06,
745
+ "loss": 0.1705,
746
+ "step": 3000
747
+ },
748
+ {
749
+ "epoch": 1.25,
750
+ "eval_loss": 0.3322072923183441,
751
+ "eval_runtime": 2282.7387,
752
+ "eval_samples_per_second": 4.573,
753
+ "eval_steps_per_second": 0.572,
754
+ "eval_wer": 45.75888348177394,
755
+ "step": 3000
756
+ },
757
+ {
758
+ "epoch": 1.26,
759
+ "learning_rate": 2.8000000000000003e-06,
760
+ "loss": 0.1556,
761
+ "step": 3025
762
+ },
763
+ {
764
+ "epoch": 1.27,
765
+ "learning_rate": 2.728571428571429e-06,
766
+ "loss": 0.1552,
767
+ "step": 3050
768
+ },
769
+ {
770
+ "epoch": 1.28,
771
+ "learning_rate": 2.6571428571428575e-06,
772
+ "loss": 0.1631,
773
+ "step": 3075
774
+ },
775
+ {
776
+ "epoch": 1.29,
777
+ "learning_rate": 2.5857142857142856e-06,
778
+ "loss": 0.1511,
779
+ "step": 3100
780
+ },
781
+ {
782
+ "epoch": 1.3,
783
+ "learning_rate": 2.5142857142857147e-06,
784
+ "loss": 0.1587,
785
+ "step": 3125
786
+ },
787
+ {
788
+ "epoch": 1.31,
789
+ "learning_rate": 2.442857142857143e-06,
790
+ "loss": 0.1544,
791
+ "step": 3150
792
+ },
793
+ {
794
+ "epoch": 1.32,
795
+ "learning_rate": 2.371428571428572e-06,
796
+ "loss": 0.1549,
797
+ "step": 3175
798
+ },
799
+ {
800
+ "epoch": 1.33,
801
+ "learning_rate": 2.3000000000000004e-06,
802
+ "loss": 0.1451,
803
+ "step": 3200
804
+ },
805
+ {
806
+ "epoch": 1.34,
807
+ "learning_rate": 2.228571428571429e-06,
808
+ "loss": 0.1467,
809
+ "step": 3225
810
+ },
811
+ {
812
+ "epoch": 1.35,
813
+ "learning_rate": 2.157142857142857e-06,
814
+ "loss": 0.1543,
815
+ "step": 3250
816
+ },
817
+ {
818
+ "epoch": 1.36,
819
+ "learning_rate": 2.0857142857142858e-06,
820
+ "loss": 0.1642,
821
+ "step": 3275
822
+ },
823
+ {
824
+ "epoch": 1.37,
825
+ "learning_rate": 2.0142857142857144e-06,
826
+ "loss": 0.1514,
827
+ "step": 3300
828
+ },
829
+ {
830
+ "epoch": 1.38,
831
+ "learning_rate": 1.942857142857143e-06,
832
+ "loss": 0.1405,
833
+ "step": 3325
834
+ },
835
+ {
836
+ "epoch": 1.39,
837
+ "learning_rate": 1.8714285714285715e-06,
838
+ "loss": 0.1615,
839
+ "step": 3350
840
+ },
841
+ {
842
+ "epoch": 1.4,
843
+ "learning_rate": 1.8000000000000001e-06,
844
+ "loss": 0.1473,
845
+ "step": 3375
846
+ },
847
+ {
848
+ "epoch": 1.41,
849
+ "learning_rate": 1.7285714285714287e-06,
850
+ "loss": 0.1626,
851
+ "step": 3400
852
+ },
853
+ {
854
+ "epoch": 1.42,
855
+ "learning_rate": 1.657142857142857e-06,
856
+ "loss": 0.1577,
857
+ "step": 3425
858
+ },
859
+ {
860
+ "epoch": 1.43,
861
+ "learning_rate": 1.5857142857142857e-06,
862
+ "loss": 0.1629,
863
+ "step": 3450
864
+ },
865
+ {
866
+ "epoch": 1.44,
867
+ "learning_rate": 1.5142857142857145e-06,
868
+ "loss": 0.1587,
869
+ "step": 3475
870
+ },
871
+ {
872
+ "epoch": 1.45,
873
+ "learning_rate": 1.442857142857143e-06,
874
+ "loss": 0.1451,
875
+ "step": 3500
876
+ },
877
+ {
878
+ "epoch": 1.47,
879
+ "learning_rate": 1.3714285714285717e-06,
880
+ "loss": 0.1485,
881
+ "step": 3525
882
+ },
883
+ {
884
+ "epoch": 1.48,
885
+ "learning_rate": 1.3e-06,
886
+ "loss": 0.1477,
887
+ "step": 3550
888
+ },
889
+ {
890
+ "epoch": 1.49,
891
+ "learning_rate": 1.2285714285714286e-06,
892
+ "loss": 0.154,
893
+ "step": 3575
894
+ },
895
+ {
896
+ "epoch": 1.5,
897
+ "learning_rate": 1.1571428571428572e-06,
898
+ "loss": 0.149,
899
+ "step": 3600
900
+ },
901
+ {
902
+ "epoch": 1.51,
903
+ "learning_rate": 1.0857142857142858e-06,
904
+ "loss": 0.1551,
905
+ "step": 3625
906
+ },
907
+ {
908
+ "epoch": 1.52,
909
+ "learning_rate": 1.0142857142857144e-06,
910
+ "loss": 0.1613,
911
+ "step": 3650
912
+ },
913
+ {
914
+ "epoch": 1.53,
915
+ "learning_rate": 9.42857142857143e-07,
916
+ "loss": 0.1529,
917
+ "step": 3675
918
+ },
919
+ {
920
+ "epoch": 1.54,
921
+ "learning_rate": 8.714285714285716e-07,
922
+ "loss": 0.1574,
923
+ "step": 3700
924
+ },
925
+ {
926
+ "epoch": 1.55,
927
+ "learning_rate": 8.000000000000001e-07,
928
+ "loss": 0.1534,
929
+ "step": 3725
930
+ },
931
+ {
932
+ "epoch": 1.56,
933
+ "learning_rate": 7.285714285714287e-07,
934
+ "loss": 0.148,
935
+ "step": 3750
936
+ },
937
+ {
938
+ "epoch": 1.57,
939
+ "learning_rate": 6.571428571428571e-07,
940
+ "loss": 0.1441,
941
+ "step": 3775
942
+ },
943
+ {
944
+ "epoch": 1.58,
945
+ "learning_rate": 5.857142857142857e-07,
946
+ "loss": 0.1518,
947
+ "step": 3800
948
+ },
949
+ {
950
+ "epoch": 1.59,
951
+ "learning_rate": 5.142857142857143e-07,
952
+ "loss": 0.1447,
953
+ "step": 3825
954
+ },
955
+ {
956
+ "epoch": 1.6,
957
+ "learning_rate": 4.4285714285714286e-07,
958
+ "loss": 0.153,
959
+ "step": 3850
960
+ },
961
+ {
962
+ "epoch": 1.61,
963
+ "learning_rate": 3.7142857142857145e-07,
964
+ "loss": 0.1543,
965
+ "step": 3875
966
+ },
967
+ {
968
+ "epoch": 1.62,
969
+ "learning_rate": 3.0000000000000004e-07,
970
+ "loss": 0.1404,
971
+ "step": 3900
972
+ },
973
+ {
974
+ "epoch": 1.63,
975
+ "learning_rate": 2.285714285714286e-07,
976
+ "loss": 0.1513,
977
+ "step": 3925
978
+ },
979
+ {
980
+ "epoch": 1.64,
981
+ "learning_rate": 1.5714285714285717e-07,
982
+ "loss": 0.1519,
983
+ "step": 3950
984
+ },
985
+ {
986
+ "epoch": 1.65,
987
+ "learning_rate": 8.571428571428573e-08,
988
+ "loss": 0.1468,
989
+ "step": 3975
990
+ },
991
+ {
992
+ "epoch": 1.66,
993
+ "learning_rate": 1.4285714285714288e-08,
994
+ "loss": 0.1476,
995
+ "step": 4000
996
+ },
997
+ {
998
+ "epoch": 1.66,
999
+ "eval_loss": 0.3225502073764801,
1000
+ "eval_runtime": 2316.4817,
1001
+ "eval_samples_per_second": 4.507,
1002
+ "eval_steps_per_second": 0.563,
1003
+ "eval_wer": 44.976586172068686,
1004
+ "step": 4000
1005
+ }
1006
+ ],
1007
+ "max_steps": 4000,
1008
+ "num_train_epochs": 2,
1009
+ "total_flos": 1.84651368394752e+19,
1010
+ "trial_name": null,
1011
+ "trial_params": null
1012
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3151cbaa1e8c87f6519ec6e91dc1b97989a7671e4e34a50a72f498793100e474
3
+ size 3643