Plim commited on
Commit
f313a54
1 Parent(s): a629eee

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_loss": 0.2463931441307068,
4
+ "eval_runtime": 1008.501,
5
+ "eval_samples": 15941,
6
+ "eval_samples_per_second": 15.807,
7
+ "eval_steps_per_second": 0.989,
8
+ "eval_wer": 0.22199184819353135,
9
+ "train_loss": 1.7380384819737829,
10
+ "train_runtime": 214447.5226,
11
+ "train_samples": 395042,
12
+ "train_samples_per_second": 9.211,
13
+ "train_steps_per_second": 0.072
14
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_loss": 0.2463931441307068,
4
+ "eval_runtime": 1008.501,
5
+ "eval_samples": 15941,
6
+ "eval_samples_per_second": 15.807,
7
+ "eval_steps_per_second": 0.989,
8
+ "eval_wer": 0.22199184819353135
9
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "train_loss": 1.7380384819737829,
4
+ "train_runtime": 214447.5226,
5
+ "train_samples": 395042,
6
+ "train_samples_per_second": 9.211,
7
+ "train_steps_per_second": 0.072
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1084 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.2463931441307068,
3
+ "best_model_checkpoint": "./checkpoint-6000",
4
+ "epoch": 4.999878498238225,
5
+ "global_step": 15430,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 3.7125e-06,
13
+ "loss": 7.2804,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.06,
18
+ "learning_rate": 7.4625e-06,
19
+ "loss": 2.9939,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.1,
24
+ "learning_rate": 1.1212499999999998e-05,
25
+ "loss": 2.8161,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.13,
30
+ "learning_rate": 1.49625e-05,
31
+ "loss": 1.8175,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.16,
36
+ "learning_rate": 1.8675e-05,
37
+ "loss": 1.3167,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.19,
42
+ "learning_rate": 2.2424999999999996e-05,
43
+ "loss": 1.1703,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.23,
48
+ "learning_rate": 2.6174999999999996e-05,
49
+ "loss": 1.1128,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.26,
54
+ "learning_rate": 2.9925e-05,
55
+ "loss": 1.0763,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.29,
60
+ "learning_rate": 3.3675e-05,
61
+ "loss": 1.0522,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.32,
66
+ "learning_rate": 3.7424999999999995e-05,
67
+ "loss": 1.0326,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.32,
72
+ "eval_loss": 0.30916452407836914,
73
+ "eval_runtime": 1003.5092,
74
+ "eval_samples_per_second": 15.885,
75
+ "eval_steps_per_second": 0.994,
76
+ "eval_wer": 0.2718170286472251,
77
+ "step": 1000
78
+ },
79
+ {
80
+ "epoch": 0.36,
81
+ "learning_rate": 4.1175e-05,
82
+ "loss": 1.0398,
83
+ "step": 1100
84
+ },
85
+ {
86
+ "epoch": 0.39,
87
+ "learning_rate": 4.4924999999999994e-05,
88
+ "loss": 1.0289,
89
+ "step": 1200
90
+ },
91
+ {
92
+ "epoch": 0.42,
93
+ "learning_rate": 4.8675e-05,
94
+ "loss": 1.0403,
95
+ "step": 1300
96
+ },
97
+ {
98
+ "epoch": 0.45,
99
+ "learning_rate": 5.2424999999999994e-05,
100
+ "loss": 1.0439,
101
+ "step": 1400
102
+ },
103
+ {
104
+ "epoch": 0.49,
105
+ "learning_rate": 5.6175e-05,
106
+ "loss": 1.0455,
107
+ "step": 1500
108
+ },
109
+ {
110
+ "epoch": 0.52,
111
+ "learning_rate": 5.9925e-05,
112
+ "loss": 1.0567,
113
+ "step": 1600
114
+ },
115
+ {
116
+ "epoch": 0.55,
117
+ "learning_rate": 6.36375e-05,
118
+ "loss": 1.0687,
119
+ "step": 1700
120
+ },
121
+ {
122
+ "epoch": 0.58,
123
+ "learning_rate": 6.738749999999999e-05,
124
+ "loss": 1.0601,
125
+ "step": 1800
126
+ },
127
+ {
128
+ "epoch": 0.62,
129
+ "learning_rate": 7.11375e-05,
130
+ "loss": 1.0657,
131
+ "step": 1900
132
+ },
133
+ {
134
+ "epoch": 0.65,
135
+ "learning_rate": 7.48875e-05,
136
+ "loss": 1.0828,
137
+ "step": 2000
138
+ },
139
+ {
140
+ "epoch": 0.65,
141
+ "eval_loss": 0.28432372212409973,
142
+ "eval_runtime": 1004.4407,
143
+ "eval_samples_per_second": 15.871,
144
+ "eval_steps_per_second": 0.993,
145
+ "eval_wer": 0.26055517279125323,
146
+ "step": 2000
147
+ },
148
+ {
149
+ "epoch": 0.68,
150
+ "learning_rate": 7.4463886820551e-05,
151
+ "loss": 1.09,
152
+ "step": 2100
153
+ },
154
+ {
155
+ "epoch": 0.71,
156
+ "learning_rate": 7.390543559195829e-05,
157
+ "loss": 1.0696,
158
+ "step": 2200
159
+ },
160
+ {
161
+ "epoch": 0.75,
162
+ "learning_rate": 7.33469843633656e-05,
163
+ "loss": 1.0794,
164
+ "step": 2300
165
+ },
166
+ {
167
+ "epoch": 0.78,
168
+ "learning_rate": 7.279411764705881e-05,
169
+ "loss": 1.0906,
170
+ "step": 2400
171
+ },
172
+ {
173
+ "epoch": 0.81,
174
+ "learning_rate": 7.223566641846612e-05,
175
+ "loss": 1.0739,
176
+ "step": 2500
177
+ },
178
+ {
179
+ "epoch": 0.84,
180
+ "learning_rate": 7.16772151898734e-05,
181
+ "loss": 1.0753,
182
+ "step": 2600
183
+ },
184
+ {
185
+ "epoch": 0.87,
186
+ "learning_rate": 7.111876396128072e-05,
187
+ "loss": 1.0694,
188
+ "step": 2700
189
+ },
190
+ {
191
+ "epoch": 0.91,
192
+ "learning_rate": 7.0560312732688e-05,
193
+ "loss": 1.0652,
194
+ "step": 2800
195
+ },
196
+ {
197
+ "epoch": 0.94,
198
+ "learning_rate": 7.00018615040953e-05,
199
+ "loss": 1.0627,
200
+ "step": 2900
201
+ },
202
+ {
203
+ "epoch": 0.97,
204
+ "learning_rate": 6.94434102755026e-05,
205
+ "loss": 1.0771,
206
+ "step": 3000
207
+ },
208
+ {
209
+ "epoch": 0.97,
210
+ "eval_loss": 0.27740851044654846,
211
+ "eval_runtime": 1003.9914,
212
+ "eval_samples_per_second": 15.878,
213
+ "eval_steps_per_second": 0.993,
214
+ "eval_wer": 0.24881682903743005,
215
+ "step": 3000
216
+ },
217
+ {
218
+ "epoch": 1.0,
219
+ "learning_rate": 6.88849590469099e-05,
220
+ "loss": 1.0548,
221
+ "step": 3100
222
+ },
223
+ {
224
+ "epoch": 1.04,
225
+ "learning_rate": 6.832650781831719e-05,
226
+ "loss": 1.0594,
227
+ "step": 3200
228
+ },
229
+ {
230
+ "epoch": 1.07,
231
+ "learning_rate": 6.776805658972449e-05,
232
+ "loss": 1.0605,
233
+ "step": 3300
234
+ },
235
+ {
236
+ "epoch": 1.1,
237
+ "learning_rate": 6.720960536113179e-05,
238
+ "loss": 1.0501,
239
+ "step": 3400
240
+ },
241
+ {
242
+ "epoch": 1.13,
243
+ "learning_rate": 6.665673864482501e-05,
244
+ "loss": 1.06,
245
+ "step": 3500
246
+ },
247
+ {
248
+ "epoch": 1.17,
249
+ "learning_rate": 6.609828741623231e-05,
250
+ "loss": 1.0539,
251
+ "step": 3600
252
+ },
253
+ {
254
+ "epoch": 1.2,
255
+ "learning_rate": 6.553983618763961e-05,
256
+ "loss": 1.0483,
257
+ "step": 3700
258
+ },
259
+ {
260
+ "epoch": 1.23,
261
+ "learning_rate": 6.49813849590469e-05,
262
+ "loss": 1.0507,
263
+ "step": 3800
264
+ },
265
+ {
266
+ "epoch": 1.26,
267
+ "learning_rate": 6.44229337304542e-05,
268
+ "loss": 1.0416,
269
+ "step": 3900
270
+ },
271
+ {
272
+ "epoch": 1.3,
273
+ "learning_rate": 6.38644825018615e-05,
274
+ "loss": 1.0306,
275
+ "step": 4000
276
+ },
277
+ {
278
+ "epoch": 1.3,
279
+ "eval_loss": 0.25875675678253174,
280
+ "eval_runtime": 1004.8682,
281
+ "eval_samples_per_second": 15.864,
282
+ "eval_steps_per_second": 0.992,
283
+ "eval_wer": 0.23513389953832187,
284
+ "step": 4000
285
+ },
286
+ {
287
+ "epoch": 1.33,
288
+ "learning_rate": 6.33060312732688e-05,
289
+ "loss": 1.0413,
290
+ "step": 4100
291
+ },
292
+ {
293
+ "epoch": 1.36,
294
+ "learning_rate": 6.275316455696202e-05,
295
+ "loss": 1.0475,
296
+ "step": 4200
297
+ },
298
+ {
299
+ "epoch": 1.39,
300
+ "learning_rate": 6.219471332836931e-05,
301
+ "loss": 1.0328,
302
+ "step": 4300
303
+ },
304
+ {
305
+ "epoch": 1.43,
306
+ "learning_rate": 6.163626209977662e-05,
307
+ "loss": 1.036,
308
+ "step": 4400
309
+ },
310
+ {
311
+ "epoch": 1.46,
312
+ "learning_rate": 6.10778108711839e-05,
313
+ "loss": 1.0248,
314
+ "step": 4500
315
+ },
316
+ {
317
+ "epoch": 1.49,
318
+ "learning_rate": 6.051935964259121e-05,
319
+ "loss": 1.0184,
320
+ "step": 4600
321
+ },
322
+ {
323
+ "epoch": 1.52,
324
+ "learning_rate": 5.996090841399851e-05,
325
+ "loss": 1.0268,
326
+ "step": 4700
327
+ },
328
+ {
329
+ "epoch": 1.56,
330
+ "learning_rate": 5.94024571854058e-05,
331
+ "loss": 1.0177,
332
+ "step": 4800
333
+ },
334
+ {
335
+ "epoch": 1.59,
336
+ "learning_rate": 5.88440059568131e-05,
337
+ "loss": 1.0155,
338
+ "step": 4900
339
+ },
340
+ {
341
+ "epoch": 1.62,
342
+ "learning_rate": 5.82855547282204e-05,
343
+ "loss": 1.0052,
344
+ "step": 5000
345
+ },
346
+ {
347
+ "epoch": 1.62,
348
+ "eval_loss": 0.24826321005821228,
349
+ "eval_runtime": 1000.3482,
350
+ "eval_samples_per_second": 15.935,
351
+ "eval_steps_per_second": 0.997,
352
+ "eval_wer": 0.22835360553241082,
353
+ "step": 5000
354
+ },
355
+ {
356
+ "epoch": 1.65,
357
+ "learning_rate": 5.7727103499627696e-05,
358
+ "loss": 1.0076,
359
+ "step": 5100
360
+ },
361
+ {
362
+ "epoch": 1.68,
363
+ "learning_rate": 5.717423678332092e-05,
364
+ "loss": 1.0107,
365
+ "step": 5200
366
+ },
367
+ {
368
+ "epoch": 1.72,
369
+ "learning_rate": 5.661578555472821e-05,
370
+ "loss": 1.0135,
371
+ "step": 5300
372
+ },
373
+ {
374
+ "epoch": 1.75,
375
+ "learning_rate": 5.605733432613552e-05,
376
+ "loss": 1.012,
377
+ "step": 5400
378
+ },
379
+ {
380
+ "epoch": 1.78,
381
+ "learning_rate": 5.549888309754281e-05,
382
+ "loss": 0.9943,
383
+ "step": 5500
384
+ },
385
+ {
386
+ "epoch": 1.81,
387
+ "learning_rate": 5.4940431868950106e-05,
388
+ "loss": 1.0049,
389
+ "step": 5600
390
+ },
391
+ {
392
+ "epoch": 1.85,
393
+ "learning_rate": 5.4381980640357404e-05,
394
+ "loss": 1.0082,
395
+ "step": 5700
396
+ },
397
+ {
398
+ "epoch": 1.88,
399
+ "learning_rate": 5.38235294117647e-05,
400
+ "loss": 0.9948,
401
+ "step": 5800
402
+ },
403
+ {
404
+ "epoch": 1.91,
405
+ "learning_rate": 5.326507818317199e-05,
406
+ "loss": 0.9925,
407
+ "step": 5900
408
+ },
409
+ {
410
+ "epoch": 1.94,
411
+ "learning_rate": 5.2706626954579296e-05,
412
+ "loss": 0.9865,
413
+ "step": 6000
414
+ },
415
+ {
416
+ "epoch": 1.94,
417
+ "eval_loss": 0.2463931441307068,
418
+ "eval_runtime": 999.3892,
419
+ "eval_samples_per_second": 15.951,
420
+ "eval_steps_per_second": 0.998,
421
+ "eval_wer": 0.22199184819353135,
422
+ "step": 6000
423
+ },
424
+ {
425
+ "epoch": 1.98,
426
+ "learning_rate": 5.2148175725986594e-05,
427
+ "loss": 0.987,
428
+ "step": 6100
429
+ },
430
+ {
431
+ "epoch": 2.01,
432
+ "learning_rate": 5.158972449739389e-05,
433
+ "loss": 0.9789,
434
+ "step": 6200
435
+ },
436
+ {
437
+ "epoch": 2.04,
438
+ "learning_rate": 5.103127326880119e-05,
439
+ "loss": 0.9722,
440
+ "step": 6300
441
+ },
442
+ {
443
+ "epoch": 2.07,
444
+ "learning_rate": 5.047282204020848e-05,
445
+ "loss": 0.9747,
446
+ "step": 6400
447
+ },
448
+ {
449
+ "epoch": 2.11,
450
+ "learning_rate": 4.9914370811615785e-05,
451
+ "loss": 0.9771,
452
+ "step": 6500
453
+ },
454
+ {
455
+ "epoch": 2.14,
456
+ "learning_rate": 4.9355919583023076e-05,
457
+ "loss": 0.9777,
458
+ "step": 6600
459
+ },
460
+ {
461
+ "epoch": 2.17,
462
+ "learning_rate": 4.88030528667163e-05,
463
+ "loss": 0.9804,
464
+ "step": 6700
465
+ },
466
+ {
467
+ "epoch": 2.2,
468
+ "learning_rate": 4.82446016381236e-05,
469
+ "loss": 0.9575,
470
+ "step": 6800
471
+ },
472
+ {
473
+ "epoch": 2.24,
474
+ "learning_rate": 4.76861504095309e-05,
475
+ "loss": 0.9613,
476
+ "step": 6900
477
+ },
478
+ {
479
+ "epoch": 2.27,
480
+ "learning_rate": 4.713328369322412e-05,
481
+ "loss": 0.978,
482
+ "step": 7000
483
+ },
484
+ {
485
+ "epoch": 2.27,
486
+ "eval_loss": 0.2513602077960968,
487
+ "eval_runtime": 1010.6008,
488
+ "eval_samples_per_second": 15.774,
489
+ "eval_steps_per_second": 0.987,
490
+ "eval_wer": 0.21723340824066503,
491
+ "step": 7000
492
+ },
493
+ {
494
+ "epoch": 2.3,
495
+ "learning_rate": 4.658041697691735e-05,
496
+ "loss": 0.9837,
497
+ "step": 7100
498
+ },
499
+ {
500
+ "epoch": 2.33,
501
+ "learning_rate": 4.602196574832464e-05,
502
+ "loss": 1.0137,
503
+ "step": 7200
504
+ },
505
+ {
506
+ "epoch": 2.37,
507
+ "learning_rate": 4.546351451973194e-05,
508
+ "loss": 1.0971,
509
+ "step": 7300
510
+ },
511
+ {
512
+ "epoch": 2.4,
513
+ "learning_rate": 4.4905063291139234e-05,
514
+ "loss": 1.1683,
515
+ "step": 7400
516
+ },
517
+ {
518
+ "epoch": 2.43,
519
+ "learning_rate": 4.434661206254654e-05,
520
+ "loss": 1.2993,
521
+ "step": 7500
522
+ },
523
+ {
524
+ "epoch": 2.46,
525
+ "learning_rate": 4.378816083395383e-05,
526
+ "loss": 1.3806,
527
+ "step": 7600
528
+ },
529
+ {
530
+ "epoch": 2.5,
531
+ "learning_rate": 4.322970960536113e-05,
532
+ "loss": 1.4648,
533
+ "step": 7700
534
+ },
535
+ {
536
+ "epoch": 2.53,
537
+ "learning_rate": 4.2671258376768425e-05,
538
+ "loss": 1.5643,
539
+ "step": 7800
540
+ },
541
+ {
542
+ "epoch": 2.56,
543
+ "learning_rate": 4.211280714817572e-05,
544
+ "loss": 1.5596,
545
+ "step": 7900
546
+ },
547
+ {
548
+ "epoch": 2.59,
549
+ "learning_rate": 4.155435591958301e-05,
550
+ "loss": 1.7438,
551
+ "step": 8000
552
+ },
553
+ {
554
+ "epoch": 2.59,
555
+ "eval_loss": 0.7983415722846985,
556
+ "eval_runtime": 1011.5068,
557
+ "eval_samples_per_second": 15.76,
558
+ "eval_steps_per_second": 0.986,
559
+ "eval_wer": 0.5072471233652923,
560
+ "step": 8000
561
+ },
562
+ {
563
+ "epoch": 2.62,
564
+ "learning_rate": 4.099590469099032e-05,
565
+ "loss": 1.7779,
566
+ "step": 8100
567
+ },
568
+ {
569
+ "epoch": 2.66,
570
+ "learning_rate": 4.0437453462397615e-05,
571
+ "loss": 1.9193,
572
+ "step": 8200
573
+ },
574
+ {
575
+ "epoch": 2.69,
576
+ "learning_rate": 3.987900223380491e-05,
577
+ "loss": 1.8549,
578
+ "step": 8300
579
+ },
580
+ {
581
+ "epoch": 2.72,
582
+ "learning_rate": 3.932055100521221e-05,
583
+ "loss": 1.9323,
584
+ "step": 8400
585
+ },
586
+ {
587
+ "epoch": 2.75,
588
+ "learning_rate": 3.87620997766195e-05,
589
+ "loss": 2.0832,
590
+ "step": 8500
591
+ },
592
+ {
593
+ "epoch": 2.79,
594
+ "learning_rate": 3.8203648548026806e-05,
595
+ "loss": 2.268,
596
+ "step": 8600
597
+ },
598
+ {
599
+ "epoch": 2.82,
600
+ "learning_rate": 3.76451973194341e-05,
601
+ "loss": 1.9962,
602
+ "step": 8700
603
+ },
604
+ {
605
+ "epoch": 2.85,
606
+ "learning_rate": 3.7086746090841395e-05,
607
+ "loss": 2.2105,
608
+ "step": 8800
609
+ },
610
+ {
611
+ "epoch": 2.88,
612
+ "learning_rate": 3.65282948622487e-05,
613
+ "loss": 2.3688,
614
+ "step": 8900
615
+ },
616
+ {
617
+ "epoch": 2.92,
618
+ "learning_rate": 3.596984363365599e-05,
619
+ "loss": 2.3309,
620
+ "step": 9000
621
+ },
622
+ {
623
+ "epoch": 2.92,
624
+ "eval_loss": 1.8916891813278198,
625
+ "eval_runtime": 1006.0568,
626
+ "eval_samples_per_second": 15.845,
627
+ "eval_steps_per_second": 0.991,
628
+ "eval_wer": 0.9416173544619229,
629
+ "step": 9000
630
+ },
631
+ {
632
+ "epoch": 2.95,
633
+ "learning_rate": 3.541139240506329e-05,
634
+ "loss": 2.2688,
635
+ "step": 9100
636
+ },
637
+ {
638
+ "epoch": 2.98,
639
+ "learning_rate": 3.4852941176470585e-05,
640
+ "loss": 2.0991,
641
+ "step": 9200
642
+ },
643
+ {
644
+ "epoch": 3.01,
645
+ "learning_rate": 3.430007446016381e-05,
646
+ "loss": 2.0663,
647
+ "step": 9300
648
+ },
649
+ {
650
+ "epoch": 3.05,
651
+ "learning_rate": 3.374162323157111e-05,
652
+ "loss": 2.0995,
653
+ "step": 9400
654
+ },
655
+ {
656
+ "epoch": 3.08,
657
+ "learning_rate": 3.3183172002978406e-05,
658
+ "loss": 2.2691,
659
+ "step": 9500
660
+ },
661
+ {
662
+ "epoch": 3.11,
663
+ "learning_rate": 3.26247207743857e-05,
664
+ "loss": 2.1691,
665
+ "step": 9600
666
+ },
667
+ {
668
+ "epoch": 3.14,
669
+ "learning_rate": 3.2066269545792995e-05,
670
+ "loss": 2.2706,
671
+ "step": 9700
672
+ },
673
+ {
674
+ "epoch": 3.18,
675
+ "learning_rate": 3.150781831720029e-05,
676
+ "loss": 2.0075,
677
+ "step": 9800
678
+ },
679
+ {
680
+ "epoch": 3.21,
681
+ "learning_rate": 3.094936708860759e-05,
682
+ "loss": 2.1615,
683
+ "step": 9900
684
+ },
685
+ {
686
+ "epoch": 3.24,
687
+ "learning_rate": 3.0390915860014888e-05,
688
+ "loss": 2.1834,
689
+ "step": 10000
690
+ },
691
+ {
692
+ "epoch": 3.24,
693
+ "eval_loss": 1.7495689392089844,
694
+ "eval_runtime": 1013.0296,
695
+ "eval_samples_per_second": 15.736,
696
+ "eval_steps_per_second": 0.984,
697
+ "eval_wer": 0.9029832005820879,
698
+ "step": 10000
699
+ },
700
+ {
701
+ "epoch": 3.27,
702
+ "learning_rate": 2.983246463142219e-05,
703
+ "loss": 2.2615,
704
+ "step": 10100
705
+ },
706
+ {
707
+ "epoch": 3.31,
708
+ "learning_rate": 2.9274013402829486e-05,
709
+ "loss": 2.387,
710
+ "step": 10200
711
+ },
712
+ {
713
+ "epoch": 3.34,
714
+ "learning_rate": 2.871556217423678e-05,
715
+ "loss": 2.3751,
716
+ "step": 10300
717
+ },
718
+ {
719
+ "epoch": 3.37,
720
+ "learning_rate": 2.8157110945644078e-05,
721
+ "loss": 2.3223,
722
+ "step": 10400
723
+ },
724
+ {
725
+ "epoch": 3.4,
726
+ "learning_rate": 2.7604244229337304e-05,
727
+ "loss": 2.2708,
728
+ "step": 10500
729
+ },
730
+ {
731
+ "epoch": 3.43,
732
+ "learning_rate": 2.7045793000744598e-05,
733
+ "loss": 2.2369,
734
+ "step": 10600
735
+ },
736
+ {
737
+ "epoch": 3.47,
738
+ "learning_rate": 2.6487341772151896e-05,
739
+ "loss": 2.3768,
740
+ "step": 10700
741
+ },
742
+ {
743
+ "epoch": 3.5,
744
+ "learning_rate": 2.5928890543559194e-05,
745
+ "loss": 2.2929,
746
+ "step": 10800
747
+ },
748
+ {
749
+ "epoch": 3.53,
750
+ "learning_rate": 2.5376023827252416e-05,
751
+ "loss": 2.3779,
752
+ "step": 10900
753
+ },
754
+ {
755
+ "epoch": 3.56,
756
+ "learning_rate": 2.4817572598659713e-05,
757
+ "loss": 2.3047,
758
+ "step": 11000
759
+ },
760
+ {
761
+ "epoch": 3.56,
762
+ "eval_loss": 1.5376594066619873,
763
+ "eval_runtime": 1009.9366,
764
+ "eval_samples_per_second": 15.784,
765
+ "eval_steps_per_second": 0.987,
766
+ "eval_wer": 0.874715877993342,
767
+ "step": 11000
768
+ },
769
+ {
770
+ "epoch": 3.6,
771
+ "learning_rate": 2.4259121370067014e-05,
772
+ "loss": 2.0797,
773
+ "step": 11100
774
+ },
775
+ {
776
+ "epoch": 3.63,
777
+ "learning_rate": 2.3700670141474312e-05,
778
+ "loss": 2.1414,
779
+ "step": 11200
780
+ },
781
+ {
782
+ "epoch": 3.66,
783
+ "learning_rate": 2.3142218912881606e-05,
784
+ "loss": 2.1064,
785
+ "step": 11300
786
+ },
787
+ {
788
+ "epoch": 3.69,
789
+ "learning_rate": 2.2583767684288904e-05,
790
+ "loss": 2.0681,
791
+ "step": 11400
792
+ },
793
+ {
794
+ "epoch": 3.73,
795
+ "learning_rate": 2.2025316455696202e-05,
796
+ "loss": 2.0853,
797
+ "step": 11500
798
+ },
799
+ {
800
+ "epoch": 3.76,
801
+ "learning_rate": 2.14668652271035e-05,
802
+ "loss": 2.1272,
803
+ "step": 11600
804
+ },
805
+ {
806
+ "epoch": 3.79,
807
+ "learning_rate": 2.0908413998510794e-05,
808
+ "loss": 2.1257,
809
+ "step": 11700
810
+ },
811
+ {
812
+ "epoch": 3.82,
813
+ "learning_rate": 2.034996276991809e-05,
814
+ "loss": 2.0966,
815
+ "step": 11800
816
+ },
817
+ {
818
+ "epoch": 3.86,
819
+ "learning_rate": 1.979151154132539e-05,
820
+ "loss": 2.2297,
821
+ "step": 11900
822
+ },
823
+ {
824
+ "epoch": 3.89,
825
+ "learning_rate": 1.9233060312732687e-05,
826
+ "loss": 2.1378,
827
+ "step": 12000
828
+ },
829
+ {
830
+ "epoch": 3.89,
831
+ "eval_loss": 1.3500523567199707,
832
+ "eval_runtime": 1013.8007,
833
+ "eval_samples_per_second": 15.724,
834
+ "eval_steps_per_second": 0.983,
835
+ "eval_wer": 0.7923285448445941,
836
+ "step": 12000
837
+ },
838
+ {
839
+ "epoch": 3.92,
840
+ "learning_rate": 1.8680193596425912e-05,
841
+ "loss": 2.2683,
842
+ "step": 12100
843
+ },
844
+ {
845
+ "epoch": 3.95,
846
+ "learning_rate": 1.8127326880119135e-05,
847
+ "loss": 2.311,
848
+ "step": 12200
849
+ },
850
+ {
851
+ "epoch": 3.99,
852
+ "learning_rate": 1.7568875651526432e-05,
853
+ "loss": 2.0594,
854
+ "step": 12300
855
+ },
856
+ {
857
+ "epoch": 4.02,
858
+ "learning_rate": 1.701042442293373e-05,
859
+ "loss": 1.9798,
860
+ "step": 12400
861
+ },
862
+ {
863
+ "epoch": 4.05,
864
+ "learning_rate": 1.6457557706626952e-05,
865
+ "loss": 1.8521,
866
+ "step": 12500
867
+ },
868
+ {
869
+ "epoch": 4.08,
870
+ "learning_rate": 1.589910647803425e-05,
871
+ "loss": 1.8466,
872
+ "step": 12600
873
+ },
874
+ {
875
+ "epoch": 4.12,
876
+ "learning_rate": 1.5340655249441547e-05,
877
+ "loss": 1.9372,
878
+ "step": 12700
879
+ },
880
+ {
881
+ "epoch": 4.15,
882
+ "learning_rate": 1.4782204020848843e-05,
883
+ "loss": 1.9591,
884
+ "step": 12800
885
+ },
886
+ {
887
+ "epoch": 4.18,
888
+ "learning_rate": 1.4223752792256143e-05,
889
+ "loss": 1.9641,
890
+ "step": 12900
891
+ },
892
+ {
893
+ "epoch": 4.21,
894
+ "learning_rate": 1.3670886075949367e-05,
895
+ "loss": 1.9812,
896
+ "step": 13000
897
+ },
898
+ {
899
+ "epoch": 4.21,
900
+ "eval_loss": 1.266192078590393,
901
+ "eval_runtime": 1008.4278,
902
+ "eval_samples_per_second": 15.808,
903
+ "eval_steps_per_second": 0.989,
904
+ "eval_wer": 0.7696760526197176,
905
+ "step": 13000
906
+ },
907
+ {
908
+ "epoch": 4.24,
909
+ "learning_rate": 1.3112434847356663e-05,
910
+ "loss": 2.1185,
911
+ "step": 13100
912
+ },
913
+ {
914
+ "epoch": 4.28,
915
+ "learning_rate": 1.255398361876396e-05,
916
+ "loss": 2.2783,
917
+ "step": 13200
918
+ },
919
+ {
920
+ "epoch": 4.31,
921
+ "learning_rate": 1.2001116902457184e-05,
922
+ "loss": 2.3792,
923
+ "step": 13300
924
+ },
925
+ {
926
+ "epoch": 4.34,
927
+ "learning_rate": 1.1442665673864482e-05,
928
+ "loss": 2.4678,
929
+ "step": 13400
930
+ },
931
+ {
932
+ "epoch": 4.37,
933
+ "learning_rate": 1.0884214445271778e-05,
934
+ "loss": 2.5792,
935
+ "step": 13500
936
+ },
937
+ {
938
+ "epoch": 4.41,
939
+ "learning_rate": 1.0325763216679075e-05,
940
+ "loss": 2.6183,
941
+ "step": 13600
942
+ },
943
+ {
944
+ "epoch": 4.44,
945
+ "learning_rate": 9.767311988086373e-06,
946
+ "loss": 2.6677,
947
+ "step": 13700
948
+ },
949
+ {
950
+ "epoch": 4.47,
951
+ "learning_rate": 9.20886075949367e-06,
952
+ "loss": 2.7043,
953
+ "step": 13800
954
+ },
955
+ {
956
+ "epoch": 4.5,
957
+ "learning_rate": 8.650409530900967e-06,
958
+ "loss": 2.6737,
959
+ "step": 13900
960
+ },
961
+ {
962
+ "epoch": 4.54,
963
+ "learning_rate": 8.091958302308264e-06,
964
+ "loss": 2.6855,
965
+ "step": 14000
966
+ },
967
+ {
968
+ "epoch": 4.54,
969
+ "eval_loss": 2.4120402336120605,
970
+ "eval_runtime": 1002.3228,
971
+ "eval_samples_per_second": 15.904,
972
+ "eval_steps_per_second": 0.995,
973
+ "eval_wer": 0.9901740468632286,
974
+ "step": 14000
975
+ },
976
+ {
977
+ "epoch": 4.57,
978
+ "learning_rate": 7.533507073715562e-06,
979
+ "loss": 2.7223,
980
+ "step": 14100
981
+ },
982
+ {
983
+ "epoch": 4.6,
984
+ "learning_rate": 6.975055845122859e-06,
985
+ "loss": 2.7861,
986
+ "step": 14200
987
+ },
988
+ {
989
+ "epoch": 4.63,
990
+ "learning_rate": 6.416604616530156e-06,
991
+ "loss": 2.7397,
992
+ "step": 14300
993
+ },
994
+ {
995
+ "epoch": 4.67,
996
+ "learning_rate": 5.8581533879374525e-06,
997
+ "loss": 2.7574,
998
+ "step": 14400
999
+ },
1000
+ {
1001
+ "epoch": 4.7,
1002
+ "learning_rate": 5.3052866716306765e-06,
1003
+ "loss": 2.8119,
1004
+ "step": 14500
1005
+ },
1006
+ {
1007
+ "epoch": 4.73,
1008
+ "learning_rate": 4.746835443037975e-06,
1009
+ "loss": 2.725,
1010
+ "step": 14600
1011
+ },
1012
+ {
1013
+ "epoch": 4.76,
1014
+ "learning_rate": 4.188384214445272e-06,
1015
+ "loss": 2.7319,
1016
+ "step": 14700
1017
+ },
1018
+ {
1019
+ "epoch": 4.8,
1020
+ "learning_rate": 3.629932985852568e-06,
1021
+ "loss": 2.7426,
1022
+ "step": 14800
1023
+ },
1024
+ {
1025
+ "epoch": 4.83,
1026
+ "learning_rate": 3.071481757259866e-06,
1027
+ "loss": 2.7441,
1028
+ "step": 14900
1029
+ },
1030
+ {
1031
+ "epoch": 4.86,
1032
+ "learning_rate": 2.513030528667163e-06,
1033
+ "loss": 2.7482,
1034
+ "step": 15000
1035
+ },
1036
+ {
1037
+ "epoch": 4.86,
1038
+ "eval_loss": 2.5340933799743652,
1039
+ "eval_runtime": 1000.6978,
1040
+ "eval_samples_per_second": 15.93,
1041
+ "eval_steps_per_second": 0.996,
1042
+ "eval_wer": 0.9874181438864671,
1043
+ "step": 15000
1044
+ },
1045
+ {
1046
+ "epoch": 4.89,
1047
+ "learning_rate": 1.95457930007446e-06,
1048
+ "loss": 2.752,
1049
+ "step": 15100
1050
+ },
1051
+ {
1052
+ "epoch": 4.93,
1053
+ "learning_rate": 1.3961280714817571e-06,
1054
+ "loss": 2.7476,
1055
+ "step": 15200
1056
+ },
1057
+ {
1058
+ "epoch": 4.96,
1059
+ "learning_rate": 8.376768428890544e-07,
1060
+ "loss": 2.7484,
1061
+ "step": 15300
1062
+ },
1063
+ {
1064
+ "epoch": 4.99,
1065
+ "learning_rate": 2.7922561429635143e-07,
1066
+ "loss": 2.7494,
1067
+ "step": 15400
1068
+ },
1069
+ {
1070
+ "epoch": 5.0,
1071
+ "step": 15430,
1072
+ "total_flos": 8.828199178194905e+20,
1073
+ "train_loss": 1.7380384819737829,
1074
+ "train_runtime": 214447.5226,
1075
+ "train_samples_per_second": 9.211,
1076
+ "train_steps_per_second": 0.072
1077
+ }
1078
+ ],
1079
+ "max_steps": 15430,
1080
+ "num_train_epochs": 5,
1081
+ "total_flos": 8.828199178194905e+20,
1082
+ "trial_name": null,
1083
+ "trial_params": null
1084
+ }
wandb/run-20220201_234440-21bzabyi/files/output.log CHANGED
The diff for this file is too large to render. See raw diff
 
wandb/run-20220201_234440-21bzabyi/files/wandb-summary.json CHANGED
The diff for this file is too large to render. See raw diff
 
wandb/run-20220201_234440-21bzabyi/logs/debug-internal.log CHANGED
The diff for this file is too large to render. See raw diff
 
wandb/run-20220201_234440-21bzabyi/run-21bzabyi.wandb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7b26cb8439f0905381b10881efcc9807ea476b5f62302e04c02c9673b406db0d
3
- size 220974870
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c339629633fb35955298de60c5906bab613d768759a30bc3e4d592d8eead8e2
3
+ size 221118605