gary109 commited on
Commit
1528249
1 Parent(s): a2dcb06

End of training

Browse files
all_results.json CHANGED
@@ -1,14 +1,14 @@
1
  {
2
  "epoch": 10.0,
3
- "eval_loss": 0.6743320226669312,
4
- "eval_runtime": 128.2267,
5
  "eval_samples": 532,
6
- "eval_samples_per_second": 4.149,
7
- "eval_steps_per_second": 0.523,
8
- "eval_wer": 0.4303528958025326,
9
- "train_loss": 0.6841171807712979,
10
- "train_runtime": 19401.0857,
11
  "train_samples": 4617,
12
- "train_samples_per_second": 2.38,
13
  "train_steps_per_second": 0.037
14
  }
 
1
  {
2
  "epoch": 10.0,
3
+ "eval_loss": 0.6776940822601318,
4
+ "eval_runtime": 127.7608,
5
  "eval_samples": 532,
6
+ "eval_samples_per_second": 4.164,
7
+ "eval_steps_per_second": 0.524,
8
+ "eval_wer": 0.4203895352574404,
9
+ "train_loss": 0.6085318591859605,
10
+ "train_runtime": 19393.8759,
11
  "train_samples": 4617,
12
+ "train_samples_per_second": 2.381,
13
  "train_steps_per_second": 0.037
14
  }
eval_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 10.0,
3
- "eval_loss": 0.6743320226669312,
4
- "eval_runtime": 128.2267,
5
  "eval_samples": 532,
6
- "eval_samples_per_second": 4.149,
7
- "eval_steps_per_second": 0.523,
8
- "eval_wer": 0.4303528958025326
9
  }
 
1
  {
2
  "epoch": 10.0,
3
+ "eval_loss": 0.6776940822601318,
4
+ "eval_runtime": 127.7608,
5
  "eval_samples": 532,
6
+ "eval_samples_per_second": 4.164,
7
+ "eval_steps_per_second": 0.524,
8
+ "eval_wer": 0.4203895352574404
9
  }
runs/Jul08_14-07-18_e43e87908b0c/events.out.tfevents.1657311726.e43e87908b0c.4172.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:457bb51610e36c4c8d553af5e35714bc681748f064ec0cd59d3f5fab4ed1af31
3
+ size 358
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "epoch": 10.0,
3
- "train_loss": 0.6841171807712979,
4
- "train_runtime": 19401.0857,
5
  "train_samples": 4617,
6
- "train_samples_per_second": 2.38,
7
  "train_steps_per_second": 0.037
8
  }
 
1
  {
2
  "epoch": 10.0,
3
+ "train_loss": 0.6085318591859605,
4
+ "train_runtime": 19393.8759,
5
  "train_samples": 4617,
6
+ "train_samples_per_second": 2.381,
7
  "train_steps_per_second": 0.037
8
  }
trainer_state.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "best_metric": 0.6743320226669312,
3
  "best_model_checkpoint": "ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1/checkpoint-720",
4
  "epoch": 9.996539792387543,
5
  "global_step": 720,
@@ -9,533 +9,533 @@
9
  "log_history": [
10
  {
11
  "epoch": 0.14,
12
- "learning_rate": 5e-06,
13
- "loss": 0.7293,
14
  "step": 10
15
  },
16
  {
17
  "epoch": 0.28,
18
- "learning_rate": 1e-05,
19
- "loss": 0.7451,
20
  "step": 20
21
  },
22
  {
23
  "epoch": 0.42,
24
- "learning_rate": 1.5e-05,
25
- "loss": 0.7259,
26
  "step": 30
27
  },
28
  {
29
  "epoch": 0.55,
30
- "learning_rate": 2e-05,
31
- "loss": 0.7445,
32
  "step": 40
33
  },
34
  {
35
  "epoch": 0.69,
36
- "learning_rate": 2.5e-05,
37
- "loss": 0.7222,
38
  "step": 50
39
  },
40
  {
41
  "epoch": 0.83,
42
- "learning_rate": 2.4626865671641793e-05,
43
- "loss": 0.7336,
44
  "step": 60
45
  },
46
  {
47
  "epoch": 0.97,
48
- "learning_rate": 2.4253731343283584e-05,
49
- "loss": 0.7295,
50
  "step": 70
51
  },
52
  {
53
  "epoch": 1.0,
54
- "eval_loss": 0.7118099927902222,
55
- "eval_runtime": 130.35,
56
- "eval_samples_per_second": 4.081,
57
- "eval_steps_per_second": 0.514,
58
- "eval_wer": 0.47457736067365175,
59
  "step": 72
60
  },
61
  {
62
  "epoch": 1.11,
63
- "learning_rate": 2.3880597014925373e-05,
64
- "loss": 0.767,
65
  "step": 80
66
  },
67
  {
68
  "epoch": 1.25,
69
- "learning_rate": 2.3507462686567168e-05,
70
- "loss": 0.729,
71
  "step": 90
72
  },
73
  {
74
  "epoch": 1.39,
75
- "learning_rate": 2.3134328358208956e-05,
76
- "loss": 0.7306,
77
  "step": 100
78
  },
79
  {
80
  "epoch": 1.53,
81
- "learning_rate": 2.2761194029850747e-05,
82
- "loss": 0.7362,
83
  "step": 110
84
  },
85
  {
86
  "epoch": 1.66,
87
- "learning_rate": 2.238805970149254e-05,
88
- "loss": 0.7128,
89
  "step": 120
90
  },
91
  {
92
  "epoch": 1.8,
93
- "learning_rate": 2.201492537313433e-05,
94
- "loss": 0.7593,
95
  "step": 130
96
  },
97
  {
98
  "epoch": 1.94,
99
- "learning_rate": 2.164179104477612e-05,
100
- "loss": 0.7062,
101
  "step": 140
102
  },
103
  {
104
  "epoch": 2.0,
105
- "eval_loss": 0.7108114361763,
106
- "eval_runtime": 129.5999,
107
- "eval_samples_per_second": 4.105,
108
  "eval_steps_per_second": 0.517,
109
- "eval_wer": 0.4676994279102655,
110
  "step": 144
111
  },
112
  {
113
  "epoch": 2.08,
114
- "learning_rate": 2.126865671641791e-05,
115
- "loss": 0.7321,
116
  "step": 150
117
  },
118
  {
119
  "epoch": 2.22,
120
- "learning_rate": 2.0895522388059702e-05,
121
- "loss": 0.7123,
122
  "step": 160
123
  },
124
  {
125
  "epoch": 2.36,
126
- "learning_rate": 2.0522388059701493e-05,
127
- "loss": 0.7404,
128
  "step": 170
129
  },
130
  {
131
  "epoch": 2.5,
132
- "learning_rate": 2.0149253731343285e-05,
133
- "loss": 0.7048,
134
  "step": 180
135
  },
136
  {
137
  "epoch": 2.64,
138
- "learning_rate": 1.9776119402985073e-05,
139
- "loss": 0.7002,
140
  "step": 190
141
  },
142
  {
143
  "epoch": 2.78,
144
- "learning_rate": 1.9402985074626868e-05,
145
- "loss": 0.6914,
146
  "step": 200
147
  },
148
  {
149
  "epoch": 2.91,
150
- "learning_rate": 1.9029850746268656e-05,
151
- "loss": 0.6932,
152
  "step": 210
153
  },
154
  {
155
  "epoch": 3.0,
156
- "eval_loss": 0.7351313829421997,
157
- "eval_runtime": 127.3643,
158
- "eval_samples_per_second": 4.177,
159
- "eval_steps_per_second": 0.526,
160
- "eval_wer": 0.48511923892781383,
161
  "step": 216
162
  },
163
  {
164
  "epoch": 3.06,
165
- "learning_rate": 1.865671641791045e-05,
166
- "loss": 0.7105,
167
  "step": 220
168
  },
169
  {
170
  "epoch": 3.19,
171
- "learning_rate": 1.828358208955224e-05,
172
- "loss": 0.702,
173
  "step": 230
174
  },
175
  {
176
  "epoch": 3.33,
177
- "learning_rate": 1.791044776119403e-05,
178
- "loss": 0.7372,
179
  "step": 240
180
  },
181
  {
182
  "epoch": 3.47,
183
- "learning_rate": 1.7537313432835823e-05,
184
- "loss": 0.6776,
185
  "step": 250
186
  },
187
  {
188
  "epoch": 3.61,
189
- "learning_rate": 1.716417910447761e-05,
190
- "loss": 0.6749,
191
  "step": 260
192
  },
193
  {
194
  "epoch": 3.75,
195
- "learning_rate": 1.6791044776119406e-05,
196
- "loss": 0.6834,
197
  "step": 270
198
  },
199
  {
200
  "epoch": 3.89,
201
- "learning_rate": 1.6417910447761194e-05,
202
- "loss": 0.6674,
203
  "step": 280
204
  },
205
  {
206
  "epoch": 4.0,
207
- "eval_loss": 0.7145407795906067,
208
- "eval_runtime": 127.1129,
209
- "eval_samples_per_second": 4.185,
210
- "eval_steps_per_second": 0.527,
211
- "eval_wer": 0.4479012663109854,
212
  "step": 288
213
  },
214
  {
215
  "epoch": 4.03,
216
- "learning_rate": 1.6044776119402986e-05,
217
- "loss": 0.702,
218
  "step": 290
219
  },
220
  {
221
  "epoch": 4.17,
222
- "learning_rate": 1.5671641791044777e-05,
223
- "loss": 0.6806,
224
  "step": 300
225
  },
226
  {
227
  "epoch": 4.3,
228
- "learning_rate": 1.529850746268657e-05,
229
- "loss": 0.683,
230
  "step": 310
231
  },
232
  {
233
  "epoch": 4.44,
234
- "learning_rate": 1.4925373134328357e-05,
235
- "loss": 0.6647,
236
  "step": 320
237
  },
238
  {
239
  "epoch": 4.58,
240
- "learning_rate": 1.455223880597015e-05,
241
- "loss": 0.7013,
242
  "step": 330
243
  },
244
  {
245
  "epoch": 4.72,
246
- "learning_rate": 1.417910447761194e-05,
247
- "loss": 0.6824,
248
  "step": 340
249
  },
250
  {
251
  "epoch": 4.86,
252
- "learning_rate": 1.3805970149253733e-05,
253
- "loss": 0.6746,
254
  "step": 350
255
  },
256
  {
257
  "epoch": 5.0,
258
- "learning_rate": 1.3432835820895523e-05,
259
- "loss": 0.6784,
260
  "step": 360
261
  },
262
  {
263
  "epoch": 5.0,
264
- "eval_loss": 0.7020928859710693,
265
- "eval_runtime": 128.5118,
266
- "eval_samples_per_second": 4.14,
267
- "eval_steps_per_second": 0.521,
268
- "eval_wer": 0.4596001799832873,
269
  "step": 360
270
  },
271
  {
272
  "epoch": 5.14,
273
- "learning_rate": 1.3059701492537313e-05,
274
- "loss": 0.6613,
275
  "step": 370
276
  },
277
  {
278
  "epoch": 5.28,
279
- "learning_rate": 1.2686567164179105e-05,
280
- "loss": 0.6596,
281
  "step": 380
282
  },
283
  {
284
  "epoch": 5.42,
285
- "learning_rate": 1.2313432835820896e-05,
286
- "loss": 0.6584,
287
  "step": 390
288
  },
289
  {
290
  "epoch": 5.55,
291
- "learning_rate": 1.1940298507462686e-05,
292
- "loss": 0.6721,
293
  "step": 400
294
  },
295
  {
296
  "epoch": 5.69,
297
- "learning_rate": 1.1567164179104478e-05,
298
- "loss": 0.679,
299
  "step": 410
300
  },
301
  {
302
  "epoch": 5.83,
303
- "learning_rate": 1.119402985074627e-05,
304
- "loss": 0.7022,
305
  "step": 420
306
  },
307
  {
308
  "epoch": 5.97,
309
- "learning_rate": 1.082089552238806e-05,
310
- "loss": 0.6493,
311
  "step": 430
312
  },
313
  {
314
  "epoch": 6.0,
315
- "eval_loss": 0.7246174216270447,
316
- "eval_runtime": 129.6572,
317
- "eval_samples_per_second": 4.103,
318
- "eval_steps_per_second": 0.517,
319
- "eval_wer": 0.4651282380921772,
320
  "step": 432
321
  },
322
  {
323
  "epoch": 6.11,
324
- "learning_rate": 1.0447761194029851e-05,
325
- "loss": 0.7152,
326
  "step": 440
327
  },
328
  {
329
  "epoch": 6.25,
330
- "learning_rate": 1.0074626865671643e-05,
331
- "loss": 0.6432,
332
  "step": 450
333
  },
334
  {
335
  "epoch": 6.39,
336
- "learning_rate": 9.701492537313434e-06,
337
- "loss": 0.6463,
338
  "step": 460
339
  },
340
  {
341
  "epoch": 6.53,
342
- "learning_rate": 9.328358208955226e-06,
343
- "loss": 0.6777,
344
  "step": 470
345
  },
346
  {
347
  "epoch": 6.66,
348
- "learning_rate": 8.955223880597016e-06,
349
- "loss": 0.676,
350
  "step": 480
351
  },
352
  {
353
  "epoch": 6.8,
354
- "learning_rate": 8.582089552238805e-06,
355
- "loss": 0.6484,
356
  "step": 490
357
  },
358
  {
359
  "epoch": 6.94,
360
- "learning_rate": 8.208955223880597e-06,
361
- "loss": 0.6338,
362
  "step": 500
363
  },
364
  {
365
  "epoch": 7.0,
366
- "eval_loss": 0.7020286917686462,
367
- "eval_runtime": 128.8637,
368
- "eval_samples_per_second": 4.128,
369
- "eval_steps_per_second": 0.52,
370
- "eval_wer": 0.44552291572925373,
371
  "step": 504
372
  },
373
  {
374
  "epoch": 7.08,
375
- "learning_rate": 7.835820895522389e-06,
376
- "loss": 0.6875,
377
  "step": 510
378
  },
379
  {
380
  "epoch": 7.22,
381
- "learning_rate": 7.4626865671641785e-06,
382
- "loss": 0.6397,
383
  "step": 520
384
  },
385
  {
386
  "epoch": 7.36,
387
- "learning_rate": 7.08955223880597e-06,
388
- "loss": 0.6695,
389
  "step": 530
390
  },
391
  {
392
  "epoch": 7.5,
393
- "learning_rate": 6.716417910447762e-06,
394
- "loss": 0.6582,
395
  "step": 540
396
  },
397
  {
398
  "epoch": 7.64,
399
- "learning_rate": 6.343283582089552e-06,
400
- "loss": 0.6335,
401
  "step": 550
402
  },
403
  {
404
  "epoch": 7.78,
405
- "learning_rate": 5.970149253731343e-06,
406
- "loss": 0.6542,
407
  "step": 560
408
  },
409
  {
410
  "epoch": 7.91,
411
- "learning_rate": 5.597014925373135e-06,
412
- "loss": 0.6464,
413
  "step": 570
414
  },
415
  {
416
  "epoch": 8.0,
417
- "eval_loss": 0.6980531811714172,
418
- "eval_runtime": 129.6171,
419
- "eval_samples_per_second": 4.104,
420
- "eval_steps_per_second": 0.517,
421
- "eval_wer": 0.439737738638555,
422
  "step": 576
423
  },
424
  {
425
  "epoch": 8.06,
426
- "learning_rate": 5.2238805970149255e-06,
427
- "loss": 0.6731,
428
  "step": 580
429
  },
430
  {
431
  "epoch": 8.19,
432
- "learning_rate": 4.850746268656717e-06,
433
- "loss": 0.632,
434
  "step": 590
435
  },
436
  {
437
  "epoch": 8.33,
438
- "learning_rate": 4.477611940298508e-06,
439
- "loss": 0.644,
440
  "step": 600
441
  },
442
  {
443
  "epoch": 8.47,
444
- "learning_rate": 4.1044776119402985e-06,
445
- "loss": 0.6391,
446
  "step": 610
447
  },
448
  {
449
  "epoch": 8.61,
450
- "learning_rate": 3.7313432835820893e-06,
451
- "loss": 0.6551,
452
  "step": 620
453
  },
454
  {
455
  "epoch": 8.75,
456
- "learning_rate": 3.358208955223881e-06,
457
- "loss": 0.6572,
458
  "step": 630
459
  },
460
  {
461
  "epoch": 8.89,
462
- "learning_rate": 2.9850746268656716e-06,
463
- "loss": 0.6227,
464
  "step": 640
465
  },
466
  {
467
  "epoch": 9.0,
468
- "eval_loss": 0.6833487153053284,
469
- "eval_runtime": 129.1282,
470
- "eval_samples_per_second": 4.12,
471
- "eval_steps_per_second": 0.519,
472
- "eval_wer": 0.43639519187504017,
473
  "step": 648
474
  },
475
  {
476
  "epoch": 9.03,
477
- "learning_rate": 2.6119402985074627e-06,
478
- "loss": 0.6722,
479
  "step": 650
480
  },
481
  {
482
  "epoch": 9.17,
483
- "learning_rate": 2.238805970149254e-06,
484
- "loss": 0.6375,
485
  "step": 660
486
  },
487
  {
488
  "epoch": 9.3,
489
- "learning_rate": 1.8656716417910446e-06,
490
- "loss": 0.6752,
491
  "step": 670
492
  },
493
  {
494
  "epoch": 9.44,
495
- "learning_rate": 1.4925373134328358e-06,
496
- "loss": 0.6428,
497
  "step": 680
498
  },
499
  {
500
  "epoch": 9.58,
501
- "learning_rate": 1.119402985074627e-06,
502
- "loss": 0.6649,
503
  "step": 690
504
  },
505
  {
506
  "epoch": 9.72,
507
- "learning_rate": 7.462686567164179e-07,
508
- "loss": 0.6376,
509
  "step": 700
510
  },
511
  {
512
  "epoch": 9.86,
513
- "learning_rate": 3.7313432835820895e-07,
514
- "loss": 0.6634,
515
  "step": 710
516
  },
517
  {
518
  "epoch": 10.0,
519
- "learning_rate": 0.0,
520
- "loss": 0.6627,
521
  "step": 720
522
  },
523
  {
524
  "epoch": 10.0,
525
- "eval_loss": 0.6743320226669312,
526
- "eval_runtime": 130.183,
527
- "eval_samples_per_second": 4.087,
528
- "eval_steps_per_second": 0.515,
529
- "eval_wer": 0.4303528958025326,
530
  "step": 720
531
  },
532
  {
533
  "epoch": 10.0,
534
  "step": 720,
535
  "total_flos": 2.7197958559499194e+19,
536
- "train_loss": 0.6841171807712979,
537
- "train_runtime": 19401.0857,
538
- "train_samples_per_second": 2.38,
539
  "train_steps_per_second": 0.037
540
  }
541
  ],
 
1
  {
2
+ "best_metric": 0.6776940822601318,
3
  "best_model_checkpoint": "ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1/checkpoint-720",
4
  "epoch": 9.996539792387543,
5
  "global_step": 720,
 
9
  "log_history": [
10
  {
11
  "epoch": 0.14,
12
+ "learning_rate": 2.0000000000000003e-06,
13
+ "loss": 0.634,
14
  "step": 10
15
  },
16
  {
17
  "epoch": 0.28,
18
+ "learning_rate": 4.000000000000001e-06,
19
+ "loss": 0.6419,
20
  "step": 20
21
  },
22
  {
23
  "epoch": 0.42,
24
+ "learning_rate": 6e-06,
25
+ "loss": 0.6223,
26
  "step": 30
27
  },
28
  {
29
  "epoch": 0.55,
30
+ "learning_rate": 8.000000000000001e-06,
31
+ "loss": 0.6284,
32
  "step": 40
33
  },
34
  {
35
  "epoch": 0.69,
36
+ "learning_rate": 1e-05,
37
+ "loss": 0.6029,
38
  "step": 50
39
  },
40
  {
41
  "epoch": 0.83,
42
+ "learning_rate": 9.850746268656717e-06,
43
+ "loss": 0.6024,
44
  "step": 60
45
  },
46
  {
47
  "epoch": 0.97,
48
+ "learning_rate": 9.701492537313434e-06,
49
+ "loss": 0.6227,
50
  "step": 70
51
  },
52
  {
53
  "epoch": 1.0,
54
+ "eval_loss": 0.7283244729042053,
55
+ "eval_runtime": 128.3245,
56
+ "eval_samples_per_second": 4.146,
57
+ "eval_steps_per_second": 0.522,
58
+ "eval_wer": 0.46589959503760364,
59
  "step": 72
60
  },
61
  {
62
  "epoch": 1.11,
63
+ "learning_rate": 9.552238805970149e-06,
64
+ "loss": 0.6452,
65
  "step": 80
66
  },
67
  {
68
  "epoch": 1.25,
69
+ "learning_rate": 9.402985074626867e-06,
70
+ "loss": 0.6194,
71
  "step": 90
72
  },
73
  {
74
  "epoch": 1.39,
75
+ "learning_rate": 9.253731343283582e-06,
76
+ "loss": 0.6242,
77
  "step": 100
78
  },
79
  {
80
  "epoch": 1.53,
81
+ "learning_rate": 9.104477611940299e-06,
82
+ "loss": 0.6169,
83
  "step": 110
84
  },
85
  {
86
  "epoch": 1.66,
87
+ "learning_rate": 8.955223880597016e-06,
88
+ "loss": 0.6081,
89
  "step": 120
90
  },
91
  {
92
  "epoch": 1.8,
93
+ "learning_rate": 8.805970149253732e-06,
94
+ "loss": 0.6511,
95
  "step": 130
96
  },
97
  {
98
  "epoch": 1.94,
99
+ "learning_rate": 8.656716417910447e-06,
100
+ "loss": 0.6027,
101
  "step": 140
102
  },
103
  {
104
  "epoch": 2.0,
105
+ "eval_loss": 0.7168184518814087,
106
+ "eval_runtime": 129.6874,
107
+ "eval_samples_per_second": 4.102,
108
  "eval_steps_per_second": 0.517,
109
+ "eval_wer": 0.4427588866748088,
110
  "step": 144
111
  },
112
  {
113
  "epoch": 2.08,
114
+ "learning_rate": 8.507462686567165e-06,
115
+ "loss": 0.6228,
116
  "step": 150
117
  },
118
  {
119
  "epoch": 2.22,
120
+ "learning_rate": 8.35820895522388e-06,
121
+ "loss": 0.6092,
122
  "step": 160
123
  },
124
  {
125
  "epoch": 2.36,
126
+ "learning_rate": 8.208955223880599e-06,
127
+ "loss": 0.625,
128
  "step": 170
129
  },
130
  {
131
  "epoch": 2.5,
132
+ "learning_rate": 8.059701492537314e-06,
133
+ "loss": 0.5997,
134
  "step": 180
135
  },
136
  {
137
  "epoch": 2.64,
138
+ "learning_rate": 7.91044776119403e-06,
139
+ "loss": 0.5925,
140
  "step": 190
141
  },
142
  {
143
  "epoch": 2.78,
144
+ "learning_rate": 7.761194029850747e-06,
145
+ "loss": 0.5989,
146
  "step": 200
147
  },
148
  {
149
  "epoch": 2.91,
150
+ "learning_rate": 7.611940298507463e-06,
151
+ "loss": 0.5973,
152
  "step": 210
153
  },
154
  {
155
  "epoch": 3.0,
156
+ "eval_loss": 0.706081211566925,
157
+ "eval_runtime": 130.3715,
158
+ "eval_samples_per_second": 4.081,
159
+ "eval_steps_per_second": 0.514,
160
+ "eval_wer": 0.4312528122388635,
161
  "step": 216
162
  },
163
  {
164
  "epoch": 3.06,
165
+ "learning_rate": 7.46268656716418e-06,
166
+ "loss": 0.6123,
167
  "step": 220
168
  },
169
  {
170
  "epoch": 3.19,
171
+ "learning_rate": 7.313432835820896e-06,
172
+ "loss": 0.6008,
173
  "step": 230
174
  },
175
  {
176
  "epoch": 3.33,
177
+ "learning_rate": 7.164179104477612e-06,
178
+ "loss": 0.6303,
179
  "step": 240
180
  },
181
  {
182
  "epoch": 3.47,
183
+ "learning_rate": 7.014925373134329e-06,
184
+ "loss": 0.5887,
185
  "step": 250
186
  },
187
  {
188
  "epoch": 3.61,
189
+ "learning_rate": 6.865671641791045e-06,
190
+ "loss": 0.5873,
191
  "step": 260
192
  },
193
  {
194
  "epoch": 3.75,
195
+ "learning_rate": 6.7164179104477625e-06,
196
+ "loss": 0.589,
197
  "step": 270
198
  },
199
  {
200
  "epoch": 3.89,
201
+ "learning_rate": 6.567164179104478e-06,
202
+ "loss": 0.5773,
203
  "step": 280
204
  },
205
  {
206
  "epoch": 4.0,
207
+ "eval_loss": 0.7067840695381165,
208
+ "eval_runtime": 128.8032,
209
+ "eval_samples_per_second": 4.13,
210
+ "eval_steps_per_second": 0.52,
211
+ "eval_wer": 0.4336311628205952,
212
  "step": 288
213
  },
214
  {
215
  "epoch": 4.03,
216
+ "learning_rate": 6.417910447761194e-06,
217
+ "loss": 0.6105,
218
  "step": 290
219
  },
220
  {
221
  "epoch": 4.17,
222
+ "learning_rate": 6.2686567164179116e-06,
223
+ "loss": 0.5964,
224
  "step": 300
225
  },
226
  {
227
  "epoch": 4.3,
228
+ "learning_rate": 6.119402985074627e-06,
229
+ "loss": 0.5959,
230
  "step": 310
231
  },
232
  {
233
  "epoch": 4.44,
234
+ "learning_rate": 5.970149253731343e-06,
235
+ "loss": 0.5843,
236
  "step": 320
237
  },
238
  {
239
  "epoch": 4.58,
240
+ "learning_rate": 5.820895522388061e-06,
241
+ "loss": 0.6068,
242
  "step": 330
243
  },
244
  {
245
  "epoch": 4.72,
246
+ "learning_rate": 5.671641791044776e-06,
247
+ "loss": 0.6045,
248
  "step": 340
249
  },
250
  {
251
  "epoch": 4.86,
252
+ "learning_rate": 5.522388059701493e-06,
253
+ "loss": 0.5906,
254
  "step": 350
255
  },
256
  {
257
  "epoch": 5.0,
258
+ "learning_rate": 5.37313432835821e-06,
259
+ "loss": 0.5965,
260
  "step": 360
261
  },
262
  {
263
  "epoch": 5.0,
264
+ "eval_loss": 0.7037749886512756,
265
+ "eval_runtime": 129.081,
266
+ "eval_samples_per_second": 4.121,
267
+ "eval_steps_per_second": 0.519,
268
+ "eval_wer": 0.4339525615478563,
269
  "step": 360
270
  },
271
  {
272
  "epoch": 5.14,
273
+ "learning_rate": 5.2238805970149255e-06,
274
+ "loss": 0.5885,
275
  "step": 370
276
  },
277
  {
278
  "epoch": 5.28,
279
+ "learning_rate": 5.074626865671642e-06,
280
+ "loss": 0.5907,
281
  "step": 380
282
  },
283
  {
284
  "epoch": 5.42,
285
+ "learning_rate": 4.925373134328359e-06,
286
+ "loss": 0.5862,
287
  "step": 390
288
  },
289
  {
290
  "epoch": 5.55,
291
+ "learning_rate": 4.7761194029850745e-06,
292
+ "loss": 0.6002,
293
  "step": 400
294
  },
295
  {
296
  "epoch": 5.69,
297
+ "learning_rate": 4.626865671641791e-06,
298
+ "loss": 0.6035,
299
  "step": 410
300
  },
301
  {
302
  "epoch": 5.83,
303
+ "learning_rate": 4.477611940298508e-06,
304
+ "loss": 0.6163,
305
  "step": 420
306
  },
307
  {
308
  "epoch": 5.97,
309
+ "learning_rate": 4.3283582089552236e-06,
310
+ "loss": 0.5854,
311
  "step": 430
312
  },
313
  {
314
  "epoch": 6.0,
315
+ "eval_loss": 0.705437958240509,
316
+ "eval_runtime": 128.9432,
317
+ "eval_samples_per_second": 4.126,
318
+ "eval_steps_per_second": 0.52,
319
+ "eval_wer": 0.43086713376615027,
320
  "step": 432
321
  },
322
  {
323
  "epoch": 6.11,
324
+ "learning_rate": 4.17910447761194e-06,
325
+ "loss": 0.6481,
326
  "step": 440
327
  },
328
  {
329
  "epoch": 6.25,
330
+ "learning_rate": 4.029850746268657e-06,
331
+ "loss": 0.5806,
332
  "step": 450
333
  },
334
  {
335
  "epoch": 6.39,
336
+ "learning_rate": 3.8805970149253735e-06,
337
+ "loss": 0.5838,
338
  "step": 460
339
  },
340
  {
341
  "epoch": 6.53,
342
+ "learning_rate": 3.73134328358209e-06,
343
+ "loss": 0.6078,
344
  "step": 470
345
  },
346
  {
347
  "epoch": 6.66,
348
+ "learning_rate": 3.582089552238806e-06,
349
+ "loss": 0.6097,
350
  "step": 480
351
  },
352
  {
353
  "epoch": 6.8,
354
+ "learning_rate": 3.447761194029851e-06,
355
+ "loss": 0.5909,
356
  "step": 490
357
  },
358
  {
359
  "epoch": 6.94,
360
+ "learning_rate": 3.2985074626865677e-06,
361
+ "loss": 0.5832,
362
  "step": 500
363
  },
364
  {
365
  "epoch": 7.0,
366
+ "eval_loss": 0.699042022228241,
367
+ "eval_runtime": 127.7248,
368
+ "eval_samples_per_second": 4.165,
369
+ "eval_steps_per_second": 0.525,
370
+ "eval_wer": 0.42527479591180817,
371
  "step": 504
372
  },
373
  {
374
  "epoch": 7.08,
375
+ "learning_rate": 3.1492537313432834e-06,
376
+ "loss": 0.6283,
377
  "step": 510
378
  },
379
  {
380
  "epoch": 7.22,
381
+ "learning_rate": 3e-06,
382
+ "loss": 0.5905,
383
  "step": 520
384
  },
385
  {
386
  "epoch": 7.36,
387
+ "learning_rate": 2.8507462686567167e-06,
388
+ "loss": 0.6143,
389
  "step": 530
390
  },
391
  {
392
  "epoch": 7.5,
393
+ "learning_rate": 2.7014925373134333e-06,
394
+ "loss": 0.6072,
395
  "step": 540
396
  },
397
  {
398
  "epoch": 7.64,
399
+ "learning_rate": 2.5522388059701496e-06,
400
+ "loss": 0.5894,
401
  "step": 550
402
  },
403
  {
404
  "epoch": 7.78,
405
+ "learning_rate": 2.402985074626866e-06,
406
+ "loss": 0.604,
407
  "step": 560
408
  },
409
  {
410
  "epoch": 7.91,
411
+ "learning_rate": 2.2537313432835824e-06,
412
+ "loss": 0.5968,
413
  "step": 570
414
  },
415
  {
416
  "epoch": 8.0,
417
+ "eval_loss": 0.698808491230011,
418
+ "eval_runtime": 129.2631,
419
+ "eval_samples_per_second": 4.116,
420
+ "eval_steps_per_second": 0.518,
421
+ "eval_wer": 0.42315356431188533,
422
  "step": 576
423
  },
424
  {
425
  "epoch": 8.06,
426
+ "learning_rate": 2.1044776119402986e-06,
427
+ "loss": 0.6225,
428
  "step": 580
429
  },
430
  {
431
  "epoch": 8.19,
432
+ "learning_rate": 1.955223880597015e-06,
433
+ "loss": 0.5931,
434
  "step": 590
435
  },
436
  {
437
  "epoch": 8.33,
438
+ "learning_rate": 1.8059701492537314e-06,
439
+ "loss": 0.6011,
440
  "step": 600
441
  },
442
  {
443
  "epoch": 8.47,
444
+ "learning_rate": 1.6567164179104479e-06,
445
+ "loss": 0.5992,
446
  "step": 610
447
  },
448
  {
449
  "epoch": 8.61,
450
+ "learning_rate": 1.5074626865671643e-06,
451
+ "loss": 0.6182,
452
  "step": 620
453
  },
454
  {
455
  "epoch": 8.75,
456
+ "learning_rate": 1.3582089552238805e-06,
457
+ "loss": 0.6231,
458
  "step": 630
459
  },
460
  {
461
  "epoch": 8.89,
462
+ "learning_rate": 1.2089552238805971e-06,
463
+ "loss": 0.5895,
464
  "step": 640
465
  },
466
  {
467
  "epoch": 9.0,
468
+ "eval_loss": 0.6866586208343506,
469
+ "eval_runtime": 126.5198,
470
+ "eval_samples_per_second": 4.205,
471
+ "eval_steps_per_second": 0.53,
472
+ "eval_wer": 0.4221893681301022,
473
  "step": 648
474
  },
475
  {
476
  "epoch": 9.03,
477
+ "learning_rate": 1.0597014925373136e-06,
478
+ "loss": 0.6387,
479
  "step": 650
480
  },
481
  {
482
  "epoch": 9.17,
483
+ "learning_rate": 9.104477611940299e-07,
484
+ "loss": 0.6074,
485
  "step": 660
486
  },
487
  {
488
  "epoch": 9.3,
489
+ "learning_rate": 7.611940298507463e-07,
490
+ "loss": 0.6425,
491
  "step": 670
492
  },
493
  {
494
  "epoch": 9.44,
495
+ "learning_rate": 6.119402985074627e-07,
496
+ "loss": 0.6131,
497
  "step": 680
498
  },
499
  {
500
  "epoch": 9.58,
501
+ "learning_rate": 4.6268656716417914e-07,
502
+ "loss": 0.6328,
503
  "step": 690
504
  },
505
  {
506
  "epoch": 9.72,
507
+ "learning_rate": 3.134328358208955e-07,
508
+ "loss": 0.61,
509
  "step": 700
510
  },
511
  {
512
  "epoch": 9.86,
513
+ "learning_rate": 1.6417910447761195e-07,
514
+ "loss": 0.6417,
515
  "step": 710
516
  },
517
  {
518
  "epoch": 10.0,
519
+ "learning_rate": 1.492537313432836e-08,
520
+ "loss": 0.6378,
521
  "step": 720
522
  },
523
  {
524
  "epoch": 10.0,
525
+ "eval_loss": 0.6776940822601318,
526
+ "eval_runtime": 127.2273,
527
+ "eval_samples_per_second": 4.181,
528
+ "eval_steps_per_second": 0.527,
529
+ "eval_wer": 0.4203895352574404,
530
  "step": 720
531
  },
532
  {
533
  "epoch": 10.0,
534
  "step": 720,
535
  "total_flos": 2.7197958559499194e+19,
536
+ "train_loss": 0.6085318591859605,
537
+ "train_runtime": 19393.8759,
538
+ "train_samples_per_second": 2.381,
539
  "train_steps_per_second": 0.037
540
  }
541
  ],