oyqiz commited on
Commit
4b62033
1 Parent(s): 3951734

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +14 -0
  2. eval_results.json +9 -0
  3. train_results.json +8 -0
  4. trainer_state.json +589 -0
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 15.0,
3
+ "eval_loss": 0.5763075351715088,
4
+ "eval_runtime": 84.634,
5
+ "eval_samples": 5785,
6
+ "eval_samples_per_second": 68.353,
7
+ "eval_steps_per_second": 4.277,
8
+ "eval_wer": 0.4501877623149989,
9
+ "train_loss": 0.3679376628312752,
10
+ "train_runtime": 9694.3418,
11
+ "train_samples": 23541,
12
+ "train_samples_per_second": 36.425,
13
+ "train_steps_per_second": 4.554
14
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 15.0,
3
+ "eval_loss": 0.5763075351715088,
4
+ "eval_runtime": 84.634,
5
+ "eval_samples": 5785,
6
+ "eval_samples_per_second": 68.353,
7
+ "eval_steps_per_second": 4.277,
8
+ "eval_wer": 0.4501877623149989
9
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 15.0,
3
+ "train_loss": 0.3679376628312752,
4
+ "train_runtime": 9694.3418,
5
+ "train_samples": 23541,
6
+ "train_samples_per_second": 36.425,
7
+ "train_steps_per_second": 4.554
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,589 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 15.0,
5
+ "global_step": 44145,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.17,
12
+ "learning_rate": 0.00029759999999999997,
13
+ "loss": 2.9274,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.34,
18
+ "learning_rate": 0.00029660442204147095,
19
+ "loss": 0.9437,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.51,
24
+ "learning_rate": 0.0002931676022453889,
25
+ "loss": 0.8203,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.68,
30
+ "learning_rate": 0.0002897307824493069,
31
+ "loss": 0.7688,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.85,
36
+ "learning_rate": 0.0002862939626532249,
37
+ "loss": 0.7259,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 1.02,
42
+ "learning_rate": 0.0002828571428571428,
43
+ "loss": 0.6873,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 1.19,
48
+ "learning_rate": 0.0002794203230610608,
49
+ "loss": 0.6257,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 1.36,
54
+ "learning_rate": 0.0002759835032649788,
55
+ "loss": 0.6302,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 1.53,
60
+ "learning_rate": 0.00027255355710848893,
61
+ "loss": 0.6274,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 1.7,
66
+ "learning_rate": 0.0002691167373124069,
67
+ "loss": 0.6076,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 1.87,
72
+ "learning_rate": 0.00026567991751632485,
73
+ "loss": 0.5965,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 2.04,
78
+ "learning_rate": 0.00026224309772024287,
79
+ "loss": 0.5949,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 2.21,
84
+ "learning_rate": 0.00025880627792416083,
85
+ "loss": 0.543,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 2.38,
90
+ "learning_rate": 0.0002553694581280788,
91
+ "loss": 0.53,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 2.55,
96
+ "learning_rate": 0.00025193263833199676,
97
+ "loss": 0.5233,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 2.72,
102
+ "learning_rate": 0.0002484958185359147,
103
+ "loss": 0.5354,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 2.89,
108
+ "learning_rate": 0.00024505899873983274,
109
+ "loss": 0.5302,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 3.06,
114
+ "learning_rate": 0.00024162905258334286,
115
+ "loss": 0.5072,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 3.23,
120
+ "learning_rate": 0.00023819223278726085,
121
+ "loss": 0.4687,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 3.4,
126
+ "learning_rate": 0.00023475541299117878,
127
+ "loss": 0.4736,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 3.4,
132
+ "eval_loss": 0.624671459197998,
133
+ "eval_runtime": 84.9387,
134
+ "eval_samples_per_second": 68.108,
135
+ "eval_steps_per_second": 4.262,
136
+ "eval_wer": 0.6361829025844931,
137
+ "step": 10000
138
+ },
139
+ {
140
+ "epoch": 3.57,
141
+ "learning_rate": 0.00023131859319509677,
142
+ "loss": 0.4707,
143
+ "step": 10500
144
+ },
145
+ {
146
+ "epoch": 3.74,
147
+ "learning_rate": 0.00022788864703860692,
148
+ "loss": 0.4712,
149
+ "step": 11000
150
+ },
151
+ {
152
+ "epoch": 3.91,
153
+ "learning_rate": 0.00022445182724252488,
154
+ "loss": 0.4596,
155
+ "step": 11500
156
+ },
157
+ {
158
+ "epoch": 4.08,
159
+ "learning_rate": 0.00022101500744644287,
160
+ "loss": 0.4394,
161
+ "step": 12000
162
+ },
163
+ {
164
+ "epoch": 4.25,
165
+ "learning_rate": 0.00021757818765036084,
166
+ "loss": 0.4136,
167
+ "step": 12500
168
+ },
169
+ {
170
+ "epoch": 4.42,
171
+ "learning_rate": 0.00021414136785427883,
172
+ "loss": 0.4233,
173
+ "step": 13000
174
+ },
175
+ {
176
+ "epoch": 4.59,
177
+ "learning_rate": 0.00021071142169778895,
178
+ "loss": 0.4353,
179
+ "step": 13500
180
+ },
181
+ {
182
+ "epoch": 4.76,
183
+ "learning_rate": 0.00020727460190170694,
184
+ "loss": 0.4248,
185
+ "step": 14000
186
+ },
187
+ {
188
+ "epoch": 4.93,
189
+ "learning_rate": 0.00020383778210562493,
190
+ "loss": 0.418,
191
+ "step": 14500
192
+ },
193
+ {
194
+ "epoch": 5.1,
195
+ "learning_rate": 0.0002004009623095429,
196
+ "loss": 0.3875,
197
+ "step": 15000
198
+ },
199
+ {
200
+ "epoch": 5.27,
201
+ "learning_rate": 0.00019696414251346088,
202
+ "loss": 0.3812,
203
+ "step": 15500
204
+ },
205
+ {
206
+ "epoch": 5.44,
207
+ "learning_rate": 0.0001935273227173788,
208
+ "loss": 0.3817,
209
+ "step": 16000
210
+ },
211
+ {
212
+ "epoch": 5.61,
213
+ "learning_rate": 0.0001900905029212968,
214
+ "loss": 0.3821,
215
+ "step": 16500
216
+ },
217
+ {
218
+ "epoch": 5.78,
219
+ "learning_rate": 0.0001866536831252148,
220
+ "loss": 0.3684,
221
+ "step": 17000
222
+ },
223
+ {
224
+ "epoch": 5.95,
225
+ "learning_rate": 0.00018321686332913275,
226
+ "loss": 0.3791,
227
+ "step": 17500
228
+ },
229
+ {
230
+ "epoch": 6.12,
231
+ "learning_rate": 0.00017978004353305074,
232
+ "loss": 0.3536,
233
+ "step": 18000
234
+ },
235
+ {
236
+ "epoch": 6.29,
237
+ "learning_rate": 0.00017634322373696873,
238
+ "loss": 0.33,
239
+ "step": 18500
240
+ },
241
+ {
242
+ "epoch": 6.46,
243
+ "learning_rate": 0.00017290640394088667,
244
+ "loss": 0.3324,
245
+ "step": 19000
246
+ },
247
+ {
248
+ "epoch": 6.63,
249
+ "learning_rate": 0.00016946958414480466,
250
+ "loss": 0.3315,
251
+ "step": 19500
252
+ },
253
+ {
254
+ "epoch": 6.8,
255
+ "learning_rate": 0.00016603963798831478,
256
+ "loss": 0.3392,
257
+ "step": 20000
258
+ },
259
+ {
260
+ "epoch": 6.8,
261
+ "eval_loss": 0.7254078388214111,
262
+ "eval_runtime": 84.9657,
263
+ "eval_samples_per_second": 68.086,
264
+ "eval_steps_per_second": 4.261,
265
+ "eval_wer": 0.5604705102717031,
266
+ "step": 20000
267
+ },
268
+ {
269
+ "epoch": 6.97,
270
+ "learning_rate": 0.00016260281819223277,
271
+ "loss": 0.3343,
272
+ "step": 20500
273
+ },
274
+ {
275
+ "epoch": 7.14,
276
+ "learning_rate": 0.00015917974567533508,
277
+ "loss": 0.3206,
278
+ "step": 21000
279
+ },
280
+ {
281
+ "epoch": 7.31,
282
+ "learning_rate": 0.00015574292587925307,
283
+ "loss": 0.3032,
284
+ "step": 21500
285
+ },
286
+ {
287
+ "epoch": 7.48,
288
+ "learning_rate": 0.00015230610608317103,
289
+ "loss": 0.3005,
290
+ "step": 22000
291
+ },
292
+ {
293
+ "epoch": 7.65,
294
+ "learning_rate": 0.00014886928628708902,
295
+ "loss": 0.3025,
296
+ "step": 22500
297
+ },
298
+ {
299
+ "epoch": 7.82,
300
+ "learning_rate": 0.00014543246649100698,
301
+ "loss": 0.3011,
302
+ "step": 23000
303
+ },
304
+ {
305
+ "epoch": 7.99,
306
+ "learning_rate": 0.00014200252033451713,
307
+ "loss": 0.2985,
308
+ "step": 23500
309
+ },
310
+ {
311
+ "epoch": 8.15,
312
+ "learning_rate": 0.0001385657005384351,
313
+ "loss": 0.2646,
314
+ "step": 24000
315
+ },
316
+ {
317
+ "epoch": 8.32,
318
+ "learning_rate": 0.00013512888074235306,
319
+ "loss": 0.2726,
320
+ "step": 24500
321
+ },
322
+ {
323
+ "epoch": 8.49,
324
+ "learning_rate": 0.00013169206094627105,
325
+ "loss": 0.2647,
326
+ "step": 25000
327
+ },
328
+ {
329
+ "epoch": 8.66,
330
+ "learning_rate": 0.000128255241150189,
331
+ "loss": 0.2722,
332
+ "step": 25500
333
+ },
334
+ {
335
+ "epoch": 8.83,
336
+ "learning_rate": 0.000124818421354107,
337
+ "loss": 0.2637,
338
+ "step": 26000
339
+ },
340
+ {
341
+ "epoch": 9.0,
342
+ "learning_rate": 0.00012138160155802496,
343
+ "loss": 0.2593,
344
+ "step": 26500
345
+ },
346
+ {
347
+ "epoch": 9.17,
348
+ "learning_rate": 0.00011794478176194294,
349
+ "loss": 0.2332,
350
+ "step": 27000
351
+ },
352
+ {
353
+ "epoch": 9.34,
354
+ "learning_rate": 0.00011451483560545307,
355
+ "loss": 0.2425,
356
+ "step": 27500
357
+ },
358
+ {
359
+ "epoch": 9.51,
360
+ "learning_rate": 0.00011107801580937105,
361
+ "loss": 0.2392,
362
+ "step": 28000
363
+ },
364
+ {
365
+ "epoch": 9.68,
366
+ "learning_rate": 0.00010764119601328904,
367
+ "loss": 0.2327,
368
+ "step": 28500
369
+ },
370
+ {
371
+ "epoch": 9.85,
372
+ "learning_rate": 0.000104204376217207,
373
+ "loss": 0.2305,
374
+ "step": 29000
375
+ },
376
+ {
377
+ "epoch": 10.02,
378
+ "learning_rate": 0.00010077443006071715,
379
+ "loss": 0.2268,
380
+ "step": 29500
381
+ },
382
+ {
383
+ "epoch": 10.19,
384
+ "learning_rate": 9.733761026463512e-05,
385
+ "loss": 0.2085,
386
+ "step": 30000
387
+ },
388
+ {
389
+ "epoch": 10.19,
390
+ "eval_loss": 0.5465497970581055,
391
+ "eval_runtime": 84.7941,
392
+ "eval_samples_per_second": 68.224,
393
+ "eval_steps_per_second": 4.269,
394
+ "eval_wer": 0.5096918489065606,
395
+ "step": 30000
396
+ },
397
+ {
398
+ "epoch": 10.36,
399
+ "learning_rate": 9.390079046855309e-05,
400
+ "loss": 0.2124,
401
+ "step": 30500
402
+ },
403
+ {
404
+ "epoch": 10.53,
405
+ "learning_rate": 9.046397067247106e-05,
406
+ "loss": 0.2078,
407
+ "step": 31000
408
+ },
409
+ {
410
+ "epoch": 10.7,
411
+ "learning_rate": 8.702715087638905e-05,
412
+ "loss": 0.2046,
413
+ "step": 31500
414
+ },
415
+ {
416
+ "epoch": 10.87,
417
+ "learning_rate": 8.359033108030701e-05,
418
+ "loss": 0.2055,
419
+ "step": 32000
420
+ },
421
+ {
422
+ "epoch": 11.04,
423
+ "learning_rate": 8.015351128422499e-05,
424
+ "loss": 0.2009,
425
+ "step": 32500
426
+ },
427
+ {
428
+ "epoch": 11.21,
429
+ "learning_rate": 7.671669148814298e-05,
430
+ "loss": 0.179,
431
+ "step": 33000
432
+ },
433
+ {
434
+ "epoch": 11.38,
435
+ "learning_rate": 7.327987169206094e-05,
436
+ "loss": 0.1815,
437
+ "step": 33500
438
+ },
439
+ {
440
+ "epoch": 11.55,
441
+ "learning_rate": 6.984305189597892e-05,
442
+ "loss": 0.1835,
443
+ "step": 34000
444
+ },
445
+ {
446
+ "epoch": 11.72,
447
+ "learning_rate": 6.640623209989689e-05,
448
+ "loss": 0.1819,
449
+ "step": 34500
450
+ },
451
+ {
452
+ "epoch": 11.89,
453
+ "learning_rate": 6.296941230381487e-05,
454
+ "loss": 0.1771,
455
+ "step": 35000
456
+ },
457
+ {
458
+ "epoch": 12.06,
459
+ "learning_rate": 5.953259250773284e-05,
460
+ "loss": 0.1708,
461
+ "step": 35500
462
+ },
463
+ {
464
+ "epoch": 12.23,
465
+ "learning_rate": 5.609577271165081e-05,
466
+ "loss": 0.16,
467
+ "step": 36000
468
+ },
469
+ {
470
+ "epoch": 12.4,
471
+ "learning_rate": 5.265895291556879e-05,
472
+ "loss": 0.1643,
473
+ "step": 36500
474
+ },
475
+ {
476
+ "epoch": 12.57,
477
+ "learning_rate": 4.9222133119486765e-05,
478
+ "loss": 0.1578,
479
+ "step": 37000
480
+ },
481
+ {
482
+ "epoch": 12.74,
483
+ "learning_rate": 4.578531332340474e-05,
484
+ "loss": 0.1521,
485
+ "step": 37500
486
+ },
487
+ {
488
+ "epoch": 12.91,
489
+ "learning_rate": 4.2348493527322716e-05,
490
+ "loss": 0.1592,
491
+ "step": 38000
492
+ },
493
+ {
494
+ "epoch": 13.08,
495
+ "learning_rate": 3.8911673731240685e-05,
496
+ "loss": 0.1453,
497
+ "step": 38500
498
+ },
499
+ {
500
+ "epoch": 13.25,
501
+ "learning_rate": 3.548172757475083e-05,
502
+ "loss": 0.1437,
503
+ "step": 39000
504
+ },
505
+ {
506
+ "epoch": 13.42,
507
+ "learning_rate": 3.2044907778668804e-05,
508
+ "loss": 0.1353,
509
+ "step": 39500
510
+ },
511
+ {
512
+ "epoch": 13.59,
513
+ "learning_rate": 2.8614961622178944e-05,
514
+ "loss": 0.1387,
515
+ "step": 40000
516
+ },
517
+ {
518
+ "epoch": 13.59,
519
+ "eval_loss": 0.598371684551239,
520
+ "eval_runtime": 84.9323,
521
+ "eval_samples_per_second": 68.113,
522
+ "eval_steps_per_second": 4.262,
523
+ "eval_wer": 0.46322067594433397,
524
+ "step": 40000
525
+ },
526
+ {
527
+ "epoch": 13.76,
528
+ "learning_rate": 2.5178141826096916e-05,
529
+ "loss": 0.1388,
530
+ "step": 40500
531
+ },
532
+ {
533
+ "epoch": 13.93,
534
+ "learning_rate": 2.1741322030014892e-05,
535
+ "loss": 0.1376,
536
+ "step": 41000
537
+ },
538
+ {
539
+ "epoch": 14.1,
540
+ "learning_rate": 1.8304502233932868e-05,
541
+ "loss": 0.137,
542
+ "step": 41500
543
+ },
544
+ {
545
+ "epoch": 14.27,
546
+ "learning_rate": 1.486768243785084e-05,
547
+ "loss": 0.1257,
548
+ "step": 42000
549
+ },
550
+ {
551
+ "epoch": 14.44,
552
+ "learning_rate": 1.1430862641768814e-05,
553
+ "loss": 0.1243,
554
+ "step": 42500
555
+ },
556
+ {
557
+ "epoch": 14.61,
558
+ "learning_rate": 7.99404284568679e-06,
559
+ "loss": 0.1207,
560
+ "step": 43000
561
+ },
562
+ {
563
+ "epoch": 14.78,
564
+ "learning_rate": 4.557223049604766e-06,
565
+ "loss": 0.121,
566
+ "step": 43500
567
+ },
568
+ {
569
+ "epoch": 14.95,
570
+ "learning_rate": 1.1272768931149044e-06,
571
+ "loss": 0.1214,
572
+ "step": 44000
573
+ },
574
+ {
575
+ "epoch": 15.0,
576
+ "step": 44145,
577
+ "total_flos": 1.429924244829818e+19,
578
+ "train_loss": 0.3679376628312752,
579
+ "train_runtime": 9694.3418,
580
+ "train_samples_per_second": 36.425,
581
+ "train_steps_per_second": 4.554
582
+ }
583
+ ],
584
+ "max_steps": 44145,
585
+ "num_train_epochs": 15,
586
+ "total_flos": 1.429924244829818e+19,
587
+ "trial_name": null,
588
+ "trial_params": null
589
+ }