DrishtiSharma commited on
Commit
a796d67
1 Parent(s): edcdf01

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 +736 -0
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "eval_loss": 0.19872024655342102,
4
+ "eval_runtime": 92.1239,
5
+ "eval_samples": 1625,
6
+ "eval_samples_per_second": 17.639,
7
+ "eval_steps_per_second": 17.639,
8
+ "eval_wer": 0.1919862128392934,
9
+ "train_loss": 1.2004031839456644,
10
+ "train_runtime": 18933.7469,
11
+ "train_samples": 3548,
12
+ "train_samples_per_second": 18.739,
13
+ "train_steps_per_second": 0.586
14
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "eval_loss": 0.19872024655342102,
4
+ "eval_runtime": 92.1239,
5
+ "eval_samples": 1625,
6
+ "eval_samples_per_second": 17.639,
7
+ "eval_steps_per_second": 17.639,
8
+ "eval_wer": 0.1919862128392934
9
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "train_loss": 1.2004031839456644,
4
+ "train_runtime": 18933.7469,
5
+ "train_samples": 3548,
6
+ "train_samples_per_second": 18.739,
7
+ "train_steps_per_second": 0.586
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,736 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 100.0,
5
+ "global_step": 11100,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.9,
12
+ "learning_rate": 3.3949999999999997e-06,
13
+ "loss": 14.9391,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 1.8,
18
+ "learning_rate": 6.895e-06,
19
+ "loss": 7.9441,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 2.7,
24
+ "learning_rate": 1.0394999999999998e-05,
25
+ "loss": 4.8005,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 3.6,
30
+ "learning_rate": 1.3895e-05,
31
+ "loss": 4.0549,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 4.5,
36
+ "learning_rate": 1.7395e-05,
37
+ "loss": 3.6156,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 5.41,
42
+ "learning_rate": 2.0894999999999996e-05,
43
+ "loss": 3.3076,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 6.31,
48
+ "learning_rate": 2.4394999999999996e-05,
49
+ "loss": 3.2087,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 7.21,
54
+ "learning_rate": 2.7895e-05,
55
+ "loss": 3.1678,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 8.11,
60
+ "learning_rate": 3.1395e-05,
61
+ "loss": 3.1295,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 9.01,
66
+ "learning_rate": 3.4895e-05,
67
+ "loss": 3.0722,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 9.91,
72
+ "learning_rate": 3.8394999999999994e-05,
73
+ "loss": 3.0354,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 10.81,
78
+ "learning_rate": 4.1895e-05,
79
+ "loss": 3.0079,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 11.71,
84
+ "learning_rate": 4.5394999999999995e-05,
85
+ "loss": 2.8589,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 12.61,
90
+ "learning_rate": 4.8895e-05,
91
+ "loss": 2.317,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 13.51,
96
+ "learning_rate": 5.2395e-05,
97
+ "loss": 1.919,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 14.41,
102
+ "learning_rate": 5.589499999999999e-05,
103
+ "loss": 1.6821,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 15.32,
108
+ "learning_rate": 5.9394999999999996e-05,
109
+ "loss": 1.4925,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 16.22,
114
+ "learning_rate": 6.289499999999999e-05,
115
+ "loss": 1.3371,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 17.12,
120
+ "learning_rate": 6.639499999999999e-05,
121
+ "loss": 1.2374,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 18.02,
126
+ "learning_rate": 6.9895e-05,
127
+ "loss": 1.1721,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 18.02,
132
+ "eval_loss": 0.38306838274002075,
133
+ "eval_runtime": 92.7322,
134
+ "eval_samples_per_second": 17.524,
135
+ "eval_steps_per_second": 17.524,
136
+ "eval_wer": 0.4066350710900474,
137
+ "step": 2000
138
+ },
139
+ {
140
+ "epoch": 18.92,
141
+ "learning_rate": 6.925384615384615e-05,
142
+ "loss": 1.117,
143
+ "step": 2100
144
+ },
145
+ {
146
+ "epoch": 19.82,
147
+ "learning_rate": 6.848461538461538e-05,
148
+ "loss": 1.0605,
149
+ "step": 2200
150
+ },
151
+ {
152
+ "epoch": 20.72,
153
+ "learning_rate": 6.771538461538461e-05,
154
+ "loss": 1.0431,
155
+ "step": 2300
156
+ },
157
+ {
158
+ "epoch": 21.62,
159
+ "learning_rate": 6.694615384615384e-05,
160
+ "loss": 1.0137,
161
+ "step": 2400
162
+ },
163
+ {
164
+ "epoch": 22.52,
165
+ "learning_rate": 6.617692307692307e-05,
166
+ "loss": 0.9686,
167
+ "step": 2500
168
+ },
169
+ {
170
+ "epoch": 23.42,
171
+ "learning_rate": 6.54076923076923e-05,
172
+ "loss": 0.9676,
173
+ "step": 2600
174
+ },
175
+ {
176
+ "epoch": 24.32,
177
+ "learning_rate": 6.463846153846153e-05,
178
+ "loss": 0.9355,
179
+ "step": 2700
180
+ },
181
+ {
182
+ "epoch": 25.23,
183
+ "learning_rate": 6.386923076923076e-05,
184
+ "loss": 0.9065,
185
+ "step": 2800
186
+ },
187
+ {
188
+ "epoch": 26.13,
189
+ "learning_rate": 6.31e-05,
190
+ "loss": 0.8981,
191
+ "step": 2900
192
+ },
193
+ {
194
+ "epoch": 27.03,
195
+ "learning_rate": 6.233076923076922e-05,
196
+ "loss": 0.9019,
197
+ "step": 3000
198
+ },
199
+ {
200
+ "epoch": 27.93,
201
+ "learning_rate": 6.156153846153846e-05,
202
+ "loss": 0.879,
203
+ "step": 3100
204
+ },
205
+ {
206
+ "epoch": 28.83,
207
+ "learning_rate": 6.079230769230769e-05,
208
+ "loss": 0.8523,
209
+ "step": 3200
210
+ },
211
+ {
212
+ "epoch": 29.73,
213
+ "learning_rate": 6.0023076923076915e-05,
214
+ "loss": 0.8511,
215
+ "step": 3300
216
+ },
217
+ {
218
+ "epoch": 30.63,
219
+ "learning_rate": 5.925384615384615e-05,
220
+ "loss": 0.8445,
221
+ "step": 3400
222
+ },
223
+ {
224
+ "epoch": 31.53,
225
+ "learning_rate": 5.848461538461538e-05,
226
+ "loss": 0.8298,
227
+ "step": 3500
228
+ },
229
+ {
230
+ "epoch": 32.43,
231
+ "learning_rate": 5.771538461538461e-05,
232
+ "loss": 0.8099,
233
+ "step": 3600
234
+ },
235
+ {
236
+ "epoch": 33.33,
237
+ "learning_rate": 5.694615384615384e-05,
238
+ "loss": 0.8134,
239
+ "step": 3700
240
+ },
241
+ {
242
+ "epoch": 34.23,
243
+ "learning_rate": 5.617692307692307e-05,
244
+ "loss": 0.8022,
245
+ "step": 3800
246
+ },
247
+ {
248
+ "epoch": 35.14,
249
+ "learning_rate": 5.54076923076923e-05,
250
+ "loss": 0.7896,
251
+ "step": 3900
252
+ },
253
+ {
254
+ "epoch": 36.04,
255
+ "learning_rate": 5.463846153846154e-05,
256
+ "loss": 0.7849,
257
+ "step": 4000
258
+ },
259
+ {
260
+ "epoch": 36.04,
261
+ "eval_loss": 0.2190932035446167,
262
+ "eval_runtime": 92.9249,
263
+ "eval_samples_per_second": 17.487,
264
+ "eval_steps_per_second": 17.487,
265
+ "eval_wer": 0.24170616113744076,
266
+ "step": 4000
267
+ },
268
+ {
269
+ "epoch": 36.94,
270
+ "learning_rate": 5.386923076923076e-05,
271
+ "loss": 0.7758,
272
+ "step": 4100
273
+ },
274
+ {
275
+ "epoch": 37.84,
276
+ "learning_rate": 5.3099999999999996e-05,
277
+ "loss": 0.7633,
278
+ "step": 4200
279
+ },
280
+ {
281
+ "epoch": 38.74,
282
+ "learning_rate": 5.2330769230769226e-05,
283
+ "loss": 0.77,
284
+ "step": 4300
285
+ },
286
+ {
287
+ "epoch": 39.64,
288
+ "learning_rate": 5.1561538461538456e-05,
289
+ "loss": 0.751,
290
+ "step": 4400
291
+ },
292
+ {
293
+ "epoch": 40.54,
294
+ "learning_rate": 5.0792307692307686e-05,
295
+ "loss": 0.7549,
296
+ "step": 4500
297
+ },
298
+ {
299
+ "epoch": 41.44,
300
+ "learning_rate": 5.0030769230769225e-05,
301
+ "loss": 0.7342,
302
+ "step": 4600
303
+ },
304
+ {
305
+ "epoch": 42.34,
306
+ "learning_rate": 4.926153846153846e-05,
307
+ "loss": 0.741,
308
+ "step": 4700
309
+ },
310
+ {
311
+ "epoch": 43.24,
312
+ "learning_rate": 4.849230769230769e-05,
313
+ "loss": 0.7223,
314
+ "step": 4800
315
+ },
316
+ {
317
+ "epoch": 44.14,
318
+ "learning_rate": 4.7723076923076914e-05,
319
+ "loss": 0.7296,
320
+ "step": 4900
321
+ },
322
+ {
323
+ "epoch": 45.05,
324
+ "learning_rate": 4.695384615384615e-05,
325
+ "loss": 0.7195,
326
+ "step": 5000
327
+ },
328
+ {
329
+ "epoch": 45.95,
330
+ "learning_rate": 4.618461538461539e-05,
331
+ "loss": 0.7229,
332
+ "step": 5100
333
+ },
334
+ {
335
+ "epoch": 46.85,
336
+ "learning_rate": 4.541538461538461e-05,
337
+ "loss": 0.7113,
338
+ "step": 5200
339
+ },
340
+ {
341
+ "epoch": 47.75,
342
+ "learning_rate": 4.464615384615385e-05,
343
+ "loss": 0.7008,
344
+ "step": 5300
345
+ },
346
+ {
347
+ "epoch": 48.65,
348
+ "learning_rate": 4.387692307692307e-05,
349
+ "loss": 0.6895,
350
+ "step": 5400
351
+ },
352
+ {
353
+ "epoch": 49.55,
354
+ "learning_rate": 4.3107692307692306e-05,
355
+ "loss": 0.6903,
356
+ "step": 5500
357
+ },
358
+ {
359
+ "epoch": 50.45,
360
+ "learning_rate": 4.2338461538461536e-05,
361
+ "loss": 0.6828,
362
+ "step": 5600
363
+ },
364
+ {
365
+ "epoch": 51.35,
366
+ "learning_rate": 4.1569230769230766e-05,
367
+ "loss": 0.6756,
368
+ "step": 5700
369
+ },
370
+ {
371
+ "epoch": 52.25,
372
+ "learning_rate": 4.0799999999999996e-05,
373
+ "loss": 0.6854,
374
+ "step": 5800
375
+ },
376
+ {
377
+ "epoch": 53.15,
378
+ "learning_rate": 4.003076923076923e-05,
379
+ "loss": 0.6598,
380
+ "step": 5900
381
+ },
382
+ {
383
+ "epoch": 54.05,
384
+ "learning_rate": 3.9261538461538455e-05,
385
+ "loss": 0.6723,
386
+ "step": 6000
387
+ },
388
+ {
389
+ "epoch": 54.05,
390
+ "eval_loss": 0.20555555820465088,
391
+ "eval_runtime": 91.8886,
392
+ "eval_samples_per_second": 17.684,
393
+ "eval_steps_per_second": 17.684,
394
+ "eval_wer": 0.213356311934511,
395
+ "step": 6000
396
+ },
397
+ {
398
+ "epoch": 54.95,
399
+ "learning_rate": 3.849230769230769e-05,
400
+ "loss": 0.6571,
401
+ "step": 6100
402
+ },
403
+ {
404
+ "epoch": 55.86,
405
+ "learning_rate": 3.772307692307692e-05,
406
+ "loss": 0.6647,
407
+ "step": 6200
408
+ },
409
+ {
410
+ "epoch": 56.76,
411
+ "learning_rate": 3.695384615384615e-05,
412
+ "loss": 0.6464,
413
+ "step": 6300
414
+ },
415
+ {
416
+ "epoch": 57.66,
417
+ "learning_rate": 3.618461538461538e-05,
418
+ "loss": 0.6381,
419
+ "step": 6400
420
+ },
421
+ {
422
+ "epoch": 58.56,
423
+ "learning_rate": 3.541538461538462e-05,
424
+ "loss": 0.6454,
425
+ "step": 6500
426
+ },
427
+ {
428
+ "epoch": 59.46,
429
+ "learning_rate": 3.464615384615384e-05,
430
+ "loss": 0.6489,
431
+ "step": 6600
432
+ },
433
+ {
434
+ "epoch": 60.36,
435
+ "learning_rate": 3.387692307692308e-05,
436
+ "loss": 0.6424,
437
+ "step": 6700
438
+ },
439
+ {
440
+ "epoch": 61.26,
441
+ "learning_rate": 3.310769230769231e-05,
442
+ "loss": 0.6276,
443
+ "step": 6800
444
+ },
445
+ {
446
+ "epoch": 62.16,
447
+ "learning_rate": 3.233846153846154e-05,
448
+ "loss": 0.6283,
449
+ "step": 6900
450
+ },
451
+ {
452
+ "epoch": 63.06,
453
+ "learning_rate": 3.1576923076923076e-05,
454
+ "loss": 0.6247,
455
+ "step": 7000
456
+ },
457
+ {
458
+ "epoch": 63.96,
459
+ "learning_rate": 3.0807692307692305e-05,
460
+ "loss": 0.6172,
461
+ "step": 7100
462
+ },
463
+ {
464
+ "epoch": 64.86,
465
+ "learning_rate": 3.0038461538461535e-05,
466
+ "loss": 0.6214,
467
+ "step": 7200
468
+ },
469
+ {
470
+ "epoch": 65.77,
471
+ "learning_rate": 2.926923076923077e-05,
472
+ "loss": 0.605,
473
+ "step": 7300
474
+ },
475
+ {
476
+ "epoch": 66.67,
477
+ "learning_rate": 2.8499999999999998e-05,
478
+ "loss": 0.6067,
479
+ "step": 7400
480
+ },
481
+ {
482
+ "epoch": 67.57,
483
+ "learning_rate": 2.7730769230769228e-05,
484
+ "loss": 0.6092,
485
+ "step": 7500
486
+ },
487
+ {
488
+ "epoch": 68.47,
489
+ "learning_rate": 2.696153846153846e-05,
490
+ "loss": 0.6128,
491
+ "step": 7600
492
+ },
493
+ {
494
+ "epoch": 69.37,
495
+ "learning_rate": 2.619230769230769e-05,
496
+ "loss": 0.5973,
497
+ "step": 7700
498
+ },
499
+ {
500
+ "epoch": 70.27,
501
+ "learning_rate": 2.542307692307692e-05,
502
+ "loss": 0.6057,
503
+ "step": 7800
504
+ },
505
+ {
506
+ "epoch": 71.17,
507
+ "learning_rate": 2.4653846153846154e-05,
508
+ "loss": 0.588,
509
+ "step": 7900
510
+ },
511
+ {
512
+ "epoch": 72.07,
513
+ "learning_rate": 2.3884615384615384e-05,
514
+ "loss": 0.6015,
515
+ "step": 8000
516
+ },
517
+ {
518
+ "epoch": 72.07,
519
+ "eval_loss": 0.20075976848602295,
520
+ "eval_runtime": 91.8498,
521
+ "eval_samples_per_second": 17.692,
522
+ "eval_steps_per_second": 17.692,
523
+ "eval_wer": 0.2031021111589832,
524
+ "step": 8000
525
+ },
526
+ {
527
+ "epoch": 72.97,
528
+ "learning_rate": 2.3115384615384614e-05,
529
+ "loss": 0.5913,
530
+ "step": 8100
531
+ },
532
+ {
533
+ "epoch": 73.87,
534
+ "learning_rate": 2.2346153846153843e-05,
535
+ "loss": 0.5679,
536
+ "step": 8200
537
+ },
538
+ {
539
+ "epoch": 74.77,
540
+ "learning_rate": 2.1576923076923076e-05,
541
+ "loss": 0.5862,
542
+ "step": 8300
543
+ },
544
+ {
545
+ "epoch": 75.68,
546
+ "learning_rate": 2.0807692307692306e-05,
547
+ "loss": 0.579,
548
+ "step": 8400
549
+ },
550
+ {
551
+ "epoch": 76.58,
552
+ "learning_rate": 2.0038461538461536e-05,
553
+ "loss": 0.5774,
554
+ "step": 8500
555
+ },
556
+ {
557
+ "epoch": 77.48,
558
+ "learning_rate": 1.926923076923077e-05,
559
+ "loss": 0.5696,
560
+ "step": 8600
561
+ },
562
+ {
563
+ "epoch": 78.38,
564
+ "learning_rate": 1.85e-05,
565
+ "loss": 0.5745,
566
+ "step": 8700
567
+ },
568
+ {
569
+ "epoch": 79.28,
570
+ "learning_rate": 1.773076923076923e-05,
571
+ "loss": 0.5629,
572
+ "step": 8800
573
+ },
574
+ {
575
+ "epoch": 80.18,
576
+ "learning_rate": 1.696153846153846e-05,
577
+ "loss": 0.5698,
578
+ "step": 8900
579
+ },
580
+ {
581
+ "epoch": 81.08,
582
+ "learning_rate": 1.6192307692307692e-05,
583
+ "loss": 0.558,
584
+ "step": 9000
585
+ },
586
+ {
587
+ "epoch": 81.98,
588
+ "learning_rate": 1.543076923076923e-05,
589
+ "loss": 0.5513,
590
+ "step": 9100
591
+ },
592
+ {
593
+ "epoch": 82.88,
594
+ "learning_rate": 1.466153846153846e-05,
595
+ "loss": 0.5663,
596
+ "step": 9200
597
+ },
598
+ {
599
+ "epoch": 83.78,
600
+ "learning_rate": 1.3892307692307692e-05,
601
+ "loss": 0.557,
602
+ "step": 9300
603
+ },
604
+ {
605
+ "epoch": 84.68,
606
+ "learning_rate": 1.3123076923076922e-05,
607
+ "loss": 0.5463,
608
+ "step": 9400
609
+ },
610
+ {
611
+ "epoch": 85.59,
612
+ "learning_rate": 1.2353846153846153e-05,
613
+ "loss": 0.5456,
614
+ "step": 9500
615
+ },
616
+ {
617
+ "epoch": 86.49,
618
+ "learning_rate": 1.1584615384615385e-05,
619
+ "loss": 0.5491,
620
+ "step": 9600
621
+ },
622
+ {
623
+ "epoch": 87.39,
624
+ "learning_rate": 1.0815384615384614e-05,
625
+ "loss": 0.5525,
626
+ "step": 9700
627
+ },
628
+ {
629
+ "epoch": 88.29,
630
+ "learning_rate": 1.0046153846153846e-05,
631
+ "loss": 0.5458,
632
+ "step": 9800
633
+ },
634
+ {
635
+ "epoch": 89.19,
636
+ "learning_rate": 9.276923076923076e-06,
637
+ "loss": 0.544,
638
+ "step": 9900
639
+ },
640
+ {
641
+ "epoch": 90.09,
642
+ "learning_rate": 8.507692307692307e-06,
643
+ "loss": 0.5386,
644
+ "step": 10000
645
+ },
646
+ {
647
+ "epoch": 90.09,
648
+ "eval_loss": 0.19667646288871765,
649
+ "eval_runtime": 91.5699,
650
+ "eval_samples_per_second": 17.746,
651
+ "eval_steps_per_second": 17.746,
652
+ "eval_wer": 0.19534683326152522,
653
+ "step": 10000
654
+ },
655
+ {
656
+ "epoch": 90.99,
657
+ "learning_rate": 7.738461538461537e-06,
658
+ "loss": 0.5311,
659
+ "step": 10100
660
+ },
661
+ {
662
+ "epoch": 91.89,
663
+ "learning_rate": 6.9692307692307684e-06,
664
+ "loss": 0.5346,
665
+ "step": 10200
666
+ },
667
+ {
668
+ "epoch": 92.79,
669
+ "learning_rate": 6.199999999999999e-06,
670
+ "loss": 0.5336,
671
+ "step": 10300
672
+ },
673
+ {
674
+ "epoch": 93.69,
675
+ "learning_rate": 5.4307692307692306e-06,
676
+ "loss": 0.5298,
677
+ "step": 10400
678
+ },
679
+ {
680
+ "epoch": 94.59,
681
+ "learning_rate": 4.661538461538461e-06,
682
+ "loss": 0.5248,
683
+ "step": 10500
684
+ },
685
+ {
686
+ "epoch": 95.5,
687
+ "learning_rate": 3.892307692307692e-06,
688
+ "loss": 0.5355,
689
+ "step": 10600
690
+ },
691
+ {
692
+ "epoch": 96.4,
693
+ "learning_rate": 3.1307692307692306e-06,
694
+ "loss": 0.5227,
695
+ "step": 10700
696
+ },
697
+ {
698
+ "epoch": 97.3,
699
+ "learning_rate": 2.361538461538461e-06,
700
+ "loss": 0.5169,
701
+ "step": 10800
702
+ },
703
+ {
704
+ "epoch": 98.2,
705
+ "learning_rate": 1.5923076923076923e-06,
706
+ "loss": 0.5379,
707
+ "step": 10900
708
+ },
709
+ {
710
+ "epoch": 99.1,
711
+ "learning_rate": 8.23076923076923e-07,
712
+ "loss": 0.5211,
713
+ "step": 11000
714
+ },
715
+ {
716
+ "epoch": 100.0,
717
+ "learning_rate": 5.384615384615384e-08,
718
+ "loss": 0.5145,
719
+ "step": 11100
720
+ },
721
+ {
722
+ "epoch": 100.0,
723
+ "step": 11100,
724
+ "total_flos": 5.044434281896765e+19,
725
+ "train_loss": 1.2004031839456644,
726
+ "train_runtime": 18933.7469,
727
+ "train_samples_per_second": 18.739,
728
+ "train_steps_per_second": 0.586
729
+ }
730
+ ],
731
+ "max_steps": 11100,
732
+ "num_train_epochs": 100,
733
+ "total_flos": 5.044434281896765e+19,
734
+ "trial_name": null,
735
+ "trial_params": null
736
+ }