hts98 commited on
Commit
38b2ce5
1 Parent(s): 406cd6c

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_loss": 0.8695769309997559,
4
+ "eval_runtime": 54.9678,
5
+ "eval_samples": 842,
6
+ "eval_samples_per_second": 15.318,
7
+ "eval_steps_per_second": 1.928,
8
+ "eval_wer": 0.457239344874995,
9
+ "train_loss": 0.8170821473989913,
10
+ "train_runtime": 40557.4985,
11
+ "train_samples": 3350,
12
+ "train_samples_per_second": 4.13,
13
+ "train_steps_per_second": 0.413
14
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_loss": 0.8695769309997559,
4
+ "eval_runtime": 54.9678,
5
+ "eval_samples": 842,
6
+ "eval_samples_per_second": 15.318,
7
+ "eval_steps_per_second": 1.928,
8
+ "eval_wer": 0.457239344874995
9
+ }
runs/Jun30_09-39-00_cbe1b177ca84/events.out.tfevents.1688159403.cbe1b177ca84.1443.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:797e7777b87ade08875b0d93dcc824d2c0de3810fd992048b6f36280de66dd30
3
+ size 412
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "train_loss": 0.8170821473989913,
4
+ "train_runtime": 40557.4985,
5
+ "train_samples": 3350,
6
+ "train_samples_per_second": 4.13,
7
+ "train_steps_per_second": 0.413
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,673 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.8695769309997559,
3
+ "best_model_checkpoint": "./wav2vec2-large-robust-paper/checkpoint-4355",
4
+ "epoch": 50.0,
5
+ "global_step": 16750,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 1.0,
12
+ "eval_loss": 3.9162847995758057,
13
+ "eval_runtime": 53.7666,
14
+ "eval_samples_per_second": 15.66,
15
+ "eval_steps_per_second": 1.971,
16
+ "eval_wer": 1.0,
17
+ "step": 335
18
+ },
19
+ {
20
+ "epoch": 1.49,
21
+ "learning_rate": 4.976729944886712e-05,
22
+ "loss": 7.1369,
23
+ "step": 500
24
+ },
25
+ {
26
+ "epoch": 2.0,
27
+ "eval_loss": 3.342172145843506,
28
+ "eval_runtime": 53.6439,
29
+ "eval_samples_per_second": 15.696,
30
+ "eval_steps_per_second": 1.976,
31
+ "eval_wer": 1.0,
32
+ "step": 670
33
+ },
34
+ {
35
+ "epoch": 2.99,
36
+ "learning_rate": 4.82363747703613e-05,
37
+ "loss": 3.3448,
38
+ "step": 1000
39
+ },
40
+ {
41
+ "epoch": 3.0,
42
+ "eval_loss": 3.335538864135742,
43
+ "eval_runtime": 53.7095,
44
+ "eval_samples_per_second": 15.677,
45
+ "eval_steps_per_second": 1.974,
46
+ "eval_wer": 1.0,
47
+ "step": 1005
48
+ },
49
+ {
50
+ "epoch": 4.0,
51
+ "eval_loss": 3.3262581825256348,
52
+ "eval_runtime": 53.7592,
53
+ "eval_samples_per_second": 15.662,
54
+ "eval_steps_per_second": 1.972,
55
+ "eval_wer": 1.0,
56
+ "step": 1340
57
+ },
58
+ {
59
+ "epoch": 4.48,
60
+ "learning_rate": 4.6705450091855484e-05,
61
+ "loss": 3.3277,
62
+ "step": 1500
63
+ },
64
+ {
65
+ "epoch": 5.0,
66
+ "eval_loss": 2.8927650451660156,
67
+ "eval_runtime": 54.3633,
68
+ "eval_samples_per_second": 15.488,
69
+ "eval_steps_per_second": 1.95,
70
+ "eval_wer": 1.0079421226156946,
71
+ "step": 1675
72
+ },
73
+ {
74
+ "epoch": 5.97,
75
+ "learning_rate": 4.517452541334967e-05,
76
+ "loss": 2.6655,
77
+ "step": 2000
78
+ },
79
+ {
80
+ "epoch": 6.0,
81
+ "eval_loss": 1.7821804285049438,
82
+ "eval_runtime": 54.4942,
83
+ "eval_samples_per_second": 15.451,
84
+ "eval_steps_per_second": 1.945,
85
+ "eval_wer": 0.8787591601372186,
86
+ "step": 2010
87
+ },
88
+ {
89
+ "epoch": 7.0,
90
+ "eval_loss": 1.3193254470825195,
91
+ "eval_runtime": 54.3211,
92
+ "eval_samples_per_second": 15.5,
93
+ "eval_steps_per_second": 1.951,
94
+ "eval_wer": 0.7055407984836553,
95
+ "step": 2345
96
+ },
97
+ {
98
+ "epoch": 7.46,
99
+ "learning_rate": 4.364360073484385e-05,
100
+ "loss": 1.4617,
101
+ "step": 2500
102
+ },
103
+ {
104
+ "epoch": 8.0,
105
+ "eval_loss": 1.1407958269119263,
106
+ "eval_runtime": 54.3418,
107
+ "eval_samples_per_second": 15.495,
108
+ "eval_steps_per_second": 1.951,
109
+ "eval_wer": 0.607031781838568,
110
+ "step": 2680
111
+ },
112
+ {
113
+ "epoch": 8.96,
114
+ "learning_rate": 4.211267605633803e-05,
115
+ "loss": 1.0805,
116
+ "step": 3000
117
+ },
118
+ {
119
+ "epoch": 9.0,
120
+ "eval_loss": 1.0108448266983032,
121
+ "eval_runtime": 54.8305,
122
+ "eval_samples_per_second": 15.356,
123
+ "eval_steps_per_second": 1.933,
124
+ "eval_wer": 0.542240079020783,
125
+ "step": 3015
126
+ },
127
+ {
128
+ "epoch": 10.0,
129
+ "eval_loss": 0.9516933560371399,
130
+ "eval_runtime": 54.4768,
131
+ "eval_samples_per_second": 15.456,
132
+ "eval_steps_per_second": 1.946,
133
+ "eval_wer": 0.515383691285022,
134
+ "step": 3350
135
+ },
136
+ {
137
+ "epoch": 10.45,
138
+ "learning_rate": 4.058175137783221e-05,
139
+ "loss": 0.8759,
140
+ "step": 3500
141
+ },
142
+ {
143
+ "epoch": 11.0,
144
+ "eval_loss": 0.9081984758377075,
145
+ "eval_runtime": 54.3156,
146
+ "eval_samples_per_second": 15.502,
147
+ "eval_steps_per_second": 1.952,
148
+ "eval_wer": 0.49020916480905535,
149
+ "step": 3685
150
+ },
151
+ {
152
+ "epoch": 11.94,
153
+ "learning_rate": 3.905082669932639e-05,
154
+ "loss": 0.7462,
155
+ "step": 4000
156
+ },
157
+ {
158
+ "epoch": 12.0,
159
+ "eval_loss": 0.875782310962677,
160
+ "eval_runtime": 54.5818,
161
+ "eval_samples_per_second": 15.426,
162
+ "eval_steps_per_second": 1.942,
163
+ "eval_wer": 0.47056075390098373,
164
+ "step": 4020
165
+ },
166
+ {
167
+ "epoch": 13.0,
168
+ "eval_loss": 0.8695769309997559,
169
+ "eval_runtime": 54.4851,
170
+ "eval_samples_per_second": 15.454,
171
+ "eval_steps_per_second": 1.945,
172
+ "eval_wer": 0.457239344874995,
173
+ "step": 4355
174
+ },
175
+ {
176
+ "epoch": 13.43,
177
+ "learning_rate": 3.7519902020820577e-05,
178
+ "loss": 0.6429,
179
+ "step": 4500
180
+ },
181
+ {
182
+ "epoch": 14.0,
183
+ "eval_loss": 0.8731432557106018,
184
+ "eval_runtime": 55.2734,
185
+ "eval_samples_per_second": 15.233,
186
+ "eval_steps_per_second": 1.918,
187
+ "eval_wer": 0.45348852730354927,
188
+ "step": 4690
189
+ },
190
+ {
191
+ "epoch": 14.93,
192
+ "learning_rate": 3.598897734231476e-05,
193
+ "loss": 0.5672,
194
+ "step": 5000
195
+ },
196
+ {
197
+ "epoch": 15.0,
198
+ "eval_loss": 0.8749489784240723,
199
+ "eval_runtime": 54.2876,
200
+ "eval_samples_per_second": 15.51,
201
+ "eval_steps_per_second": 1.953,
202
+ "eval_wer": 0.450792210045784,
203
+ "step": 5025
204
+ },
205
+ {
206
+ "epoch": 16.0,
207
+ "eval_loss": 0.8752743005752563,
208
+ "eval_runtime": 54.5671,
209
+ "eval_samples_per_second": 15.431,
210
+ "eval_steps_per_second": 1.943,
211
+ "eval_wer": 0.4512460456238237,
212
+ "step": 5360
213
+ },
214
+ {
215
+ "epoch": 16.42,
216
+ "learning_rate": 3.445805266380894e-05,
217
+ "loss": 0.4959,
218
+ "step": 5500
219
+ },
220
+ {
221
+ "epoch": 17.0,
222
+ "eval_loss": 0.9039203524589539,
223
+ "eval_runtime": 54.5236,
224
+ "eval_samples_per_second": 15.443,
225
+ "eval_steps_per_second": 1.944,
226
+ "eval_wer": 0.4486965575236595,
227
+ "step": 5695
228
+ },
229
+ {
230
+ "epoch": 17.91,
231
+ "learning_rate": 3.2927127985303126e-05,
232
+ "loss": 0.4456,
233
+ "step": 6000
234
+ },
235
+ {
236
+ "epoch": 18.0,
237
+ "eval_loss": 0.9161030650138855,
238
+ "eval_runtime": 54.3598,
239
+ "eval_samples_per_second": 15.489,
240
+ "eval_steps_per_second": 1.95,
241
+ "eval_wer": 0.443303923008129,
242
+ "step": 6030
243
+ },
244
+ {
245
+ "epoch": 19.0,
246
+ "eval_loss": 0.9505767226219177,
247
+ "eval_runtime": 54.2942,
248
+ "eval_samples_per_second": 15.508,
249
+ "eval_steps_per_second": 1.952,
250
+ "eval_wer": 0.4429835684824539,
251
+ "step": 6365
252
+ },
253
+ {
254
+ "epoch": 19.4,
255
+ "learning_rate": 3.13962033067973e-05,
256
+ "loss": 0.392,
257
+ "step": 6500
258
+ },
259
+ {
260
+ "epoch": 20.0,
261
+ "eval_loss": 0.9411688446998596,
262
+ "eval_runtime": 54.3336,
263
+ "eval_samples_per_second": 15.497,
264
+ "eval_steps_per_second": 1.951,
265
+ "eval_wer": 0.4438511953228239,
266
+ "step": 6700
267
+ },
268
+ {
269
+ "epoch": 20.9,
270
+ "learning_rate": 2.9865278628291486e-05,
271
+ "loss": 0.3594,
272
+ "step": 7000
273
+ },
274
+ {
275
+ "epoch": 21.0,
276
+ "eval_loss": 0.9884207844734192,
277
+ "eval_runtime": 54.3631,
278
+ "eval_samples_per_second": 15.488,
279
+ "eval_steps_per_second": 1.95,
280
+ "eval_wer": 0.44162206174833485,
281
+ "step": 7035
282
+ },
283
+ {
284
+ "epoch": 22.0,
285
+ "eval_loss": 1.022162914276123,
286
+ "eval_runtime": 54.656,
287
+ "eval_samples_per_second": 15.405,
288
+ "eval_steps_per_second": 1.939,
289
+ "eval_wer": 0.45101912783480386,
290
+ "step": 7370
291
+ },
292
+ {
293
+ "epoch": 22.39,
294
+ "learning_rate": 2.833435394978567e-05,
295
+ "loss": 0.3175,
296
+ "step": 7500
297
+ },
298
+ {
299
+ "epoch": 23.0,
300
+ "eval_loss": 1.0344542264938354,
301
+ "eval_runtime": 54.3628,
302
+ "eval_samples_per_second": 15.489,
303
+ "eval_steps_per_second": 1.95,
304
+ "eval_wer": 0.4439312839542427,
305
+ "step": 7705
306
+ },
307
+ {
308
+ "epoch": 23.88,
309
+ "learning_rate": 2.6803429271279856e-05,
310
+ "loss": 0.2947,
311
+ "step": 8000
312
+ },
313
+ {
314
+ "epoch": 24.0,
315
+ "eval_loss": 1.084883213043213,
316
+ "eval_runtime": 54.6337,
317
+ "eval_samples_per_second": 15.412,
318
+ "eval_steps_per_second": 1.94,
319
+ "eval_wer": 0.4464807720544069,
320
+ "step": 8040
321
+ },
322
+ {
323
+ "epoch": 25.0,
324
+ "eval_loss": 1.0879119634628296,
325
+ "eval_runtime": 54.4682,
326
+ "eval_samples_per_second": 15.459,
327
+ "eval_steps_per_second": 1.946,
328
+ "eval_wer": 0.4472416140528852,
329
+ "step": 8375
330
+ },
331
+ {
332
+ "epoch": 25.37,
333
+ "learning_rate": 2.527556644213105e-05,
334
+ "loss": 0.2674,
335
+ "step": 8500
336
+ },
337
+ {
338
+ "epoch": 26.0,
339
+ "eval_loss": 1.1071442365646362,
340
+ "eval_runtime": 54.3908,
341
+ "eval_samples_per_second": 15.481,
342
+ "eval_steps_per_second": 1.949,
343
+ "eval_wer": 0.4512460456238237,
344
+ "step": 8710
345
+ },
346
+ {
347
+ "epoch": 26.87,
348
+ "learning_rate": 2.3744641763625232e-05,
349
+ "loss": 0.2521,
350
+ "step": 9000
351
+ },
352
+ {
353
+ "epoch": 27.0,
354
+ "eval_loss": 1.1146758794784546,
355
+ "eval_runtime": 54.502,
356
+ "eval_samples_per_second": 15.449,
357
+ "eval_steps_per_second": 1.945,
358
+ "eval_wer": 0.4494307033116649,
359
+ "step": 9045
360
+ },
361
+ {
362
+ "epoch": 28.0,
363
+ "eval_loss": 1.1426175832748413,
364
+ "eval_runtime": 54.4541,
365
+ "eval_samples_per_second": 15.463,
366
+ "eval_steps_per_second": 1.947,
367
+ "eval_wer": 0.4525408118317605,
368
+ "step": 9380
369
+ },
370
+ {
371
+ "epoch": 28.36,
372
+ "learning_rate": 2.2216778934476424e-05,
373
+ "loss": 0.2321,
374
+ "step": 9500
375
+ },
376
+ {
377
+ "epoch": 29.0,
378
+ "eval_loss": 1.1591500043869019,
379
+ "eval_runtime": 54.6264,
380
+ "eval_samples_per_second": 15.414,
381
+ "eval_steps_per_second": 1.94,
382
+ "eval_wer": 0.4440380687961344,
383
+ "step": 9715
384
+ },
385
+ {
386
+ "epoch": 29.85,
387
+ "learning_rate": 2.0685854255970608e-05,
388
+ "loss": 0.2235,
389
+ "step": 10000
390
+ },
391
+ {
392
+ "epoch": 30.0,
393
+ "eval_loss": 1.178154706954956,
394
+ "eval_runtime": 54.3939,
395
+ "eval_samples_per_second": 15.48,
396
+ "eval_steps_per_second": 1.949,
397
+ "eval_wer": 0.44495908805745027,
398
+ "step": 10050
399
+ },
400
+ {
401
+ "epoch": 31.0,
402
+ "eval_loss": 1.2050156593322754,
403
+ "eval_runtime": 54.4103,
404
+ "eval_samples_per_second": 15.475,
405
+ "eval_steps_per_second": 1.948,
406
+ "eval_wer": 0.4436643218495135,
407
+ "step": 10385
408
+ },
409
+ {
410
+ "epoch": 31.34,
411
+ "learning_rate": 1.9154929577464788e-05,
412
+ "loss": 0.2071,
413
+ "step": 10500
414
+ },
415
+ {
416
+ "epoch": 32.0,
417
+ "eval_loss": 1.2223504781723022,
418
+ "eval_runtime": 54.4403,
419
+ "eval_samples_per_second": 15.466,
420
+ "eval_steps_per_second": 1.947,
421
+ "eval_wer": 0.4399535485937771,
422
+ "step": 10720
423
+ },
424
+ {
425
+ "epoch": 32.84,
426
+ "learning_rate": 1.7624004898958974e-05,
427
+ "loss": 0.1951,
428
+ "step": 11000
429
+ },
430
+ {
431
+ "epoch": 33.0,
432
+ "eval_loss": 1.2269729375839233,
433
+ "eval_runtime": 54.7494,
434
+ "eval_samples_per_second": 15.379,
435
+ "eval_steps_per_second": 1.936,
436
+ "eval_wer": 0.4470680886848112,
437
+ "step": 11055
438
+ },
439
+ {
440
+ "epoch": 34.0,
441
+ "eval_loss": 1.246551275253296,
442
+ "eval_runtime": 54.3927,
443
+ "eval_samples_per_second": 15.48,
444
+ "eval_steps_per_second": 1.949,
445
+ "eval_wer": 0.44833615868227505,
446
+ "step": 11390
447
+ },
448
+ {
449
+ "epoch": 34.33,
450
+ "learning_rate": 1.6093080220453154e-05,
451
+ "loss": 0.1892,
452
+ "step": 11500
453
+ },
454
+ {
455
+ "epoch": 35.0,
456
+ "eval_loss": 1.232478380203247,
457
+ "eval_runtime": 54.41,
458
+ "eval_samples_per_second": 15.475,
459
+ "eval_steps_per_second": 1.948,
460
+ "eval_wer": 0.44286343553532576,
461
+ "step": 11725
462
+ },
463
+ {
464
+ "epoch": 35.82,
465
+ "learning_rate": 1.4562155541947337e-05,
466
+ "loss": 0.1809,
467
+ "step": 12000
468
+ },
469
+ {
470
+ "epoch": 36.0,
471
+ "eval_loss": 1.2754629850387573,
472
+ "eval_runtime": 54.5117,
473
+ "eval_samples_per_second": 15.446,
474
+ "eval_steps_per_second": 1.945,
475
+ "eval_wer": 0.44272995448296115,
476
+ "step": 12060
477
+ },
478
+ {
479
+ "epoch": 37.0,
480
+ "eval_loss": 1.2675007581710815,
481
+ "eval_runtime": 54.594,
482
+ "eval_samples_per_second": 15.423,
483
+ "eval_steps_per_second": 1.942,
484
+ "eval_wer": 0.4421693340630298,
485
+ "step": 12395
486
+ },
487
+ {
488
+ "epoch": 37.31,
489
+ "learning_rate": 1.3031230863441517e-05,
490
+ "loss": 0.1746,
491
+ "step": 12500
492
+ },
493
+ {
494
+ "epoch": 38.0,
495
+ "eval_loss": 1.3022228479385376,
496
+ "eval_runtime": 54.757,
497
+ "eval_samples_per_second": 15.377,
498
+ "eval_steps_per_second": 1.936,
499
+ "eval_wer": 0.4417688909059359,
500
+ "step": 12730
501
+ },
502
+ {
503
+ "epoch": 38.81,
504
+ "learning_rate": 1.15003061849357e-05,
505
+ "loss": 0.1656,
506
+ "step": 13000
507
+ },
508
+ {
509
+ "epoch": 39.0,
510
+ "eval_loss": 1.3178575038909912,
511
+ "eval_runtime": 54.82,
512
+ "eval_samples_per_second": 15.359,
513
+ "eval_steps_per_second": 1.934,
514
+ "eval_wer": 0.4407544349079648,
515
+ "step": 13065
516
+ },
517
+ {
518
+ "epoch": 40.0,
519
+ "eval_loss": 1.293350338935852,
520
+ "eval_runtime": 54.6833,
521
+ "eval_samples_per_second": 15.398,
522
+ "eval_steps_per_second": 1.938,
523
+ "eval_wer": 0.4425430810096507,
524
+ "step": 13400
525
+ },
526
+ {
527
+ "epoch": 40.3,
528
+ "learning_rate": 9.972443355786895e-06,
529
+ "loss": 0.1614,
530
+ "step": 13500
531
+ },
532
+ {
533
+ "epoch": 41.0,
534
+ "eval_loss": 1.3303664922714233,
535
+ "eval_runtime": 54.5126,
536
+ "eval_samples_per_second": 15.446,
537
+ "eval_steps_per_second": 1.945,
538
+ "eval_wer": 0.44259647343059655,
539
+ "step": 13735
540
+ },
541
+ {
542
+ "epoch": 41.79,
543
+ "learning_rate": 8.441518677281078e-06,
544
+ "loss": 0.1564,
545
+ "step": 14000
546
+ },
547
+ {
548
+ "epoch": 42.0,
549
+ "eval_loss": 1.3147845268249512,
550
+ "eval_runtime": 54.5529,
551
+ "eval_samples_per_second": 15.435,
552
+ "eval_steps_per_second": 1.943,
553
+ "eval_wer": 0.44196911248448284,
554
+ "step": 14070
555
+ },
556
+ {
557
+ "epoch": 43.0,
558
+ "eval_loss": 1.3266962766647339,
559
+ "eval_runtime": 54.7201,
560
+ "eval_samples_per_second": 15.387,
561
+ "eval_steps_per_second": 1.937,
562
+ "eval_wer": 0.44325053058718317,
563
+ "step": 14405
564
+ },
565
+ {
566
+ "epoch": 43.28,
567
+ "learning_rate": 6.910593998775261e-06,
568
+ "loss": 0.1546,
569
+ "step": 14500
570
+ },
571
+ {
572
+ "epoch": 44.0,
573
+ "eval_loss": 1.3331360816955566,
574
+ "eval_runtime": 54.6617,
575
+ "eval_samples_per_second": 15.404,
576
+ "eval_steps_per_second": 1.939,
577
+ "eval_wer": 0.4412750110121868,
578
+ "step": 14740
579
+ },
580
+ {
581
+ "epoch": 44.78,
582
+ "learning_rate": 5.379669320269443e-06,
583
+ "loss": 0.1515,
584
+ "step": 15000
585
+ },
586
+ {
587
+ "epoch": 45.0,
588
+ "eval_loss": 1.344504475593567,
589
+ "eval_runtime": 54.543,
590
+ "eval_samples_per_second": 15.437,
591
+ "eval_steps_per_second": 1.943,
592
+ "eval_wer": 0.43875221912249557,
593
+ "step": 15075
594
+ },
595
+ {
596
+ "epoch": 46.0,
597
+ "eval_loss": 1.352994441986084,
598
+ "eval_runtime": 54.5619,
599
+ "eval_samples_per_second": 15.432,
600
+ "eval_steps_per_second": 1.943,
601
+ "eval_wer": 0.4372438832307754,
602
+ "step": 15410
603
+ },
604
+ {
605
+ "epoch": 46.27,
606
+ "learning_rate": 3.851806491120637e-06,
607
+ "loss": 0.147,
608
+ "step": 15500
609
+ },
610
+ {
611
+ "epoch": 47.0,
612
+ "eval_loss": 1.34428870677948,
613
+ "eval_runtime": 54.5828,
614
+ "eval_samples_per_second": 15.426,
615
+ "eval_steps_per_second": 1.942,
616
+ "eval_wer": 0.4384585608072934,
617
+ "step": 15745
618
+ },
619
+ {
620
+ "epoch": 47.76,
621
+ "learning_rate": 2.3208818126148195e-06,
622
+ "loss": 0.1447,
623
+ "step": 16000
624
+ },
625
+ {
626
+ "epoch": 48.0,
627
+ "eval_loss": 1.3502939939498901,
628
+ "eval_runtime": 54.5282,
629
+ "eval_samples_per_second": 15.442,
630
+ "eval_steps_per_second": 1.944,
631
+ "eval_wer": 0.436856788178918,
632
+ "step": 16080
633
+ },
634
+ {
635
+ "epoch": 49.0,
636
+ "eval_loss": 1.358969807624817,
637
+ "eval_runtime": 55.1854,
638
+ "eval_samples_per_second": 15.258,
639
+ "eval_steps_per_second": 1.921,
640
+ "eval_wer": 0.4393261876476634,
641
+ "step": 16415
642
+ },
643
+ {
644
+ "epoch": 49.25,
645
+ "learning_rate": 7.930189834660135e-07,
646
+ "loss": 0.1437,
647
+ "step": 16500
648
+ },
649
+ {
650
+ "epoch": 50.0,
651
+ "eval_loss": 1.3668413162231445,
652
+ "eval_runtime": 54.6805,
653
+ "eval_samples_per_second": 15.399,
654
+ "eval_steps_per_second": 1.939,
655
+ "eval_wer": 0.43717714270459307,
656
+ "step": 16750
657
+ },
658
+ {
659
+ "epoch": 50.0,
660
+ "step": 16750,
661
+ "total_flos": 1.4087701795050537e+20,
662
+ "train_loss": 0.8170821473989913,
663
+ "train_runtime": 40557.4985,
664
+ "train_samples_per_second": 4.13,
665
+ "train_steps_per_second": 0.413
666
+ }
667
+ ],
668
+ "max_steps": 16750,
669
+ "num_train_epochs": 50,
670
+ "total_flos": 1.4087701795050537e+20,
671
+ "trial_name": null,
672
+ "trial_params": null
673
+ }