jonatasgrosman commited on
Commit
06a5e79
1 Parent(s): 2d65bb5

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 11.97,
3
+ "eval_cer": 5.74468085106383,
4
+ "eval_loss": 0.3191336393356323,
5
+ "eval_runtime": 193.2304,
6
+ "eval_samples": 1000,
7
+ "eval_samples_per_second": 5.175,
8
+ "eval_steps_per_second": 0.326,
9
+ "eval_wer": 14.884437596302003,
10
+ "train_loss": 1.8605577674278846,
11
+ "train_runtime": 13282.9377,
12
+ "train_samples_per_second": 24.091,
13
+ "train_steps_per_second": 0.753
14
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 11.97,
3
+ "eval_cer": 5.74468085106383,
4
+ "eval_loss": 0.3191336393356323,
5
+ "eval_runtime": 193.2304,
6
+ "eval_samples": 1000,
7
+ "eval_samples_per_second": 5.175,
8
+ "eval_steps_per_second": 0.326,
9
+ "eval_wer": 14.884437596302003
10
+ }
runs/Dec12_12-25-12_jonatas-ubuntu/events.out.tfevents.1670872635.jonatas-ubuntu.1334127.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2606f0cce117ae908b4a1d92ae1e431374d7a522942f134eee47dcbed45ed0e
3
+ size 405
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 11.97,
3
+ "train_loss": 1.8605577674278846,
4
+ "train_runtime": 13282.9377,
5
+ "train_samples_per_second": 24.091,
6
+ "train_steps_per_second": 0.753
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,935 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 14.884437596302003,
3
+ "best_model_checkpoint": "whisper-small-pt-cv11-v4/checkpoint-5000",
4
+ "epoch": 11.970534069981584,
5
+ "global_step": 6500,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.09,
12
+ "learning_rate": 2.2e-07,
13
+ "loss": 6.7515,
14
+ "step": 50
15
+ },
16
+ {
17
+ "epoch": 0.18,
18
+ "learning_rate": 4.7000000000000005e-07,
19
+ "loss": 5.8564,
20
+ "step": 100
21
+ },
22
+ {
23
+ "epoch": 0.28,
24
+ "learning_rate": 7.2e-07,
25
+ "loss": 5.5525,
26
+ "step": 150
27
+ },
28
+ {
29
+ "epoch": 0.37,
30
+ "learning_rate": 9.7e-07,
31
+ "loss": 5.4545,
32
+ "step": 200
33
+ },
34
+ {
35
+ "epoch": 0.46,
36
+ "learning_rate": 1.2200000000000002e-06,
37
+ "loss": 4.2057,
38
+ "step": 250
39
+ },
40
+ {
41
+ "epoch": 0.55,
42
+ "learning_rate": 1.4700000000000001e-06,
43
+ "loss": 3.82,
44
+ "step": 300
45
+ },
46
+ {
47
+ "epoch": 0.64,
48
+ "learning_rate": 1.72e-06,
49
+ "loss": 3.5564,
50
+ "step": 350
51
+ },
52
+ {
53
+ "epoch": 0.74,
54
+ "learning_rate": 1.97e-06,
55
+ "loss": 3.2007,
56
+ "step": 400
57
+ },
58
+ {
59
+ "epoch": 0.83,
60
+ "learning_rate": 2.2200000000000003e-06,
61
+ "loss": 3.0785,
62
+ "step": 450
63
+ },
64
+ {
65
+ "epoch": 0.92,
66
+ "learning_rate": 2.47e-06,
67
+ "loss": 2.9379,
68
+ "step": 500
69
+ },
70
+ {
71
+ "epoch": 0.92,
72
+ "eval_cer": 7.057198121027907,
73
+ "eval_loss": 0.47825536131858826,
74
+ "eval_runtime": 193.582,
75
+ "eval_samples_per_second": 5.166,
76
+ "eval_steps_per_second": 0.325,
77
+ "eval_wer": 17.380585516178737,
78
+ "step": 500
79
+ },
80
+ {
81
+ "epoch": 1.01,
82
+ "learning_rate": 2.7200000000000002e-06,
83
+ "loss": 2.83,
84
+ "step": 550
85
+ },
86
+ {
87
+ "epoch": 1.1,
88
+ "learning_rate": 2.97e-06,
89
+ "loss": 2.8714,
90
+ "step": 600
91
+ },
92
+ {
93
+ "epoch": 1.2,
94
+ "learning_rate": 3.2200000000000005e-06,
95
+ "loss": 2.5881,
96
+ "step": 650
97
+ },
98
+ {
99
+ "epoch": 1.29,
100
+ "learning_rate": 3.4700000000000002e-06,
101
+ "loss": 2.2805,
102
+ "step": 700
103
+ },
104
+ {
105
+ "epoch": 1.38,
106
+ "learning_rate": 3.7200000000000004e-06,
107
+ "loss": 2.5818,
108
+ "step": 750
109
+ },
110
+ {
111
+ "epoch": 1.47,
112
+ "learning_rate": 3.97e-06,
113
+ "loss": 2.9589,
114
+ "step": 800
115
+ },
116
+ {
117
+ "epoch": 1.57,
118
+ "learning_rate": 4.22e-06,
119
+ "loss": 2.5782,
120
+ "step": 850
121
+ },
122
+ {
123
+ "epoch": 1.66,
124
+ "learning_rate": 4.47e-06,
125
+ "loss": 2.3677,
126
+ "step": 900
127
+ },
128
+ {
129
+ "epoch": 1.75,
130
+ "learning_rate": 4.7200000000000005e-06,
131
+ "loss": 2.5862,
132
+ "step": 950
133
+ },
134
+ {
135
+ "epoch": 1.84,
136
+ "learning_rate": 4.970000000000001e-06,
137
+ "loss": 2.1727,
138
+ "step": 1000
139
+ },
140
+ {
141
+ "epoch": 1.84,
142
+ "eval_cer": 6.797457861287648,
143
+ "eval_loss": 0.3721262812614441,
144
+ "eval_runtime": 195.8834,
145
+ "eval_samples_per_second": 5.105,
146
+ "eval_steps_per_second": 0.322,
147
+ "eval_wer": 17.272727272727273,
148
+ "step": 1000
149
+ },
150
+ {
151
+ "epoch": 1.93,
152
+ "learning_rate": 4.9755555555555554e-06,
153
+ "loss": 2.0283,
154
+ "step": 1050
155
+ },
156
+ {
157
+ "epoch": 2.03,
158
+ "learning_rate": 4.947777777777778e-06,
159
+ "loss": 2.1114,
160
+ "step": 1100
161
+ },
162
+ {
163
+ "epoch": 2.12,
164
+ "learning_rate": 4.92e-06,
165
+ "loss": 2.0833,
166
+ "step": 1150
167
+ },
168
+ {
169
+ "epoch": 2.21,
170
+ "learning_rate": 4.892222222222222e-06,
171
+ "loss": 2.0559,
172
+ "step": 1200
173
+ },
174
+ {
175
+ "epoch": 2.3,
176
+ "learning_rate": 4.864444444444445e-06,
177
+ "loss": 1.9963,
178
+ "step": 1250
179
+ },
180
+ {
181
+ "epoch": 2.39,
182
+ "learning_rate": 4.836666666666667e-06,
183
+ "loss": 2.1794,
184
+ "step": 1300
185
+ },
186
+ {
187
+ "epoch": 2.49,
188
+ "learning_rate": 4.808888888888889e-06,
189
+ "loss": 2.1072,
190
+ "step": 1350
191
+ },
192
+ {
193
+ "epoch": 2.58,
194
+ "learning_rate": 4.781111111111111e-06,
195
+ "loss": 1.8907,
196
+ "step": 1400
197
+ },
198
+ {
199
+ "epoch": 2.67,
200
+ "learning_rate": 4.753333333333333e-06,
201
+ "loss": 1.8439,
202
+ "step": 1450
203
+ },
204
+ {
205
+ "epoch": 2.76,
206
+ "learning_rate": 4.725555555555556e-06,
207
+ "loss": 1.7856,
208
+ "step": 1500
209
+ },
210
+ {
211
+ "epoch": 2.76,
212
+ "eval_cer": 6.402321083172147,
213
+ "eval_loss": 0.3466031551361084,
214
+ "eval_runtime": 198.2589,
215
+ "eval_samples_per_second": 5.044,
216
+ "eval_steps_per_second": 0.318,
217
+ "eval_wer": 16.37904468412943,
218
+ "step": 1500
219
+ },
220
+ {
221
+ "epoch": 2.85,
222
+ "learning_rate": 4.697777777777778e-06,
223
+ "loss": 1.9889,
224
+ "step": 1550
225
+ },
226
+ {
227
+ "epoch": 2.95,
228
+ "learning_rate": 4.670000000000001e-06,
229
+ "loss": 1.9626,
230
+ "step": 1600
231
+ },
232
+ {
233
+ "epoch": 3.04,
234
+ "learning_rate": 4.642222222222222e-06,
235
+ "loss": 2.2689,
236
+ "step": 1650
237
+ },
238
+ {
239
+ "epoch": 3.13,
240
+ "learning_rate": 4.614444444444445e-06,
241
+ "loss": 1.5772,
242
+ "step": 1700
243
+ },
244
+ {
245
+ "epoch": 3.22,
246
+ "learning_rate": 4.586666666666667e-06,
247
+ "loss": 1.8441,
248
+ "step": 1750
249
+ },
250
+ {
251
+ "epoch": 3.31,
252
+ "learning_rate": 4.558888888888889e-06,
253
+ "loss": 1.6514,
254
+ "step": 1800
255
+ },
256
+ {
257
+ "epoch": 3.41,
258
+ "learning_rate": 4.531111111111112e-06,
259
+ "loss": 1.9828,
260
+ "step": 1850
261
+ },
262
+ {
263
+ "epoch": 3.5,
264
+ "learning_rate": 4.503333333333333e-06,
265
+ "loss": 1.8731,
266
+ "step": 1900
267
+ },
268
+ {
269
+ "epoch": 3.59,
270
+ "learning_rate": 4.475555555555556e-06,
271
+ "loss": 1.6879,
272
+ "step": 1950
273
+ },
274
+ {
275
+ "epoch": 3.68,
276
+ "learning_rate": 4.447777777777778e-06,
277
+ "loss": 1.7803,
278
+ "step": 2000
279
+ },
280
+ {
281
+ "epoch": 3.68,
282
+ "eval_cer": 6.20889748549323,
283
+ "eval_loss": 0.33716297149658203,
284
+ "eval_runtime": 196.0701,
285
+ "eval_samples_per_second": 5.1,
286
+ "eval_steps_per_second": 0.321,
287
+ "eval_wer": 15.901386748844375,
288
+ "step": 2000
289
+ },
290
+ {
291
+ "epoch": 3.78,
292
+ "learning_rate": 4.42e-06,
293
+ "loss": 1.8693,
294
+ "step": 2050
295
+ },
296
+ {
297
+ "epoch": 3.87,
298
+ "learning_rate": 4.392222222222223e-06,
299
+ "loss": 1.7551,
300
+ "step": 2100
301
+ },
302
+ {
303
+ "epoch": 3.96,
304
+ "learning_rate": 4.364444444444445e-06,
305
+ "loss": 2.2832,
306
+ "step": 2150
307
+ },
308
+ {
309
+ "epoch": 4.05,
310
+ "learning_rate": 4.336666666666667e-06,
311
+ "loss": 1.9009,
312
+ "step": 2200
313
+ },
314
+ {
315
+ "epoch": 4.14,
316
+ "learning_rate": 4.308888888888889e-06,
317
+ "loss": 1.5534,
318
+ "step": 2250
319
+ },
320
+ {
321
+ "epoch": 4.24,
322
+ "learning_rate": 4.281111111111112e-06,
323
+ "loss": 1.5566,
324
+ "step": 2300
325
+ },
326
+ {
327
+ "epoch": 4.33,
328
+ "learning_rate": 4.253333333333334e-06,
329
+ "loss": 1.5025,
330
+ "step": 2350
331
+ },
332
+ {
333
+ "epoch": 4.42,
334
+ "learning_rate": 4.225555555555556e-06,
335
+ "loss": 1.4498,
336
+ "step": 2400
337
+ },
338
+ {
339
+ "epoch": 4.51,
340
+ "learning_rate": 4.197777777777779e-06,
341
+ "loss": 1.4484,
342
+ "step": 2450
343
+ },
344
+ {
345
+ "epoch": 4.6,
346
+ "learning_rate": 4.17e-06,
347
+ "loss": 1.8312,
348
+ "step": 2500
349
+ },
350
+ {
351
+ "epoch": 4.6,
352
+ "eval_cer": 6.090080132633324,
353
+ "eval_loss": 0.3303045630455017,
354
+ "eval_runtime": 197.9104,
355
+ "eval_samples_per_second": 5.053,
356
+ "eval_steps_per_second": 0.318,
357
+ "eval_wer": 15.747303543913713,
358
+ "step": 2500
359
+ },
360
+ {
361
+ "epoch": 4.7,
362
+ "learning_rate": 4.142222222222223e-06,
363
+ "loss": 1.5531,
364
+ "step": 2550
365
+ },
366
+ {
367
+ "epoch": 4.79,
368
+ "learning_rate": 4.114444444444445e-06,
369
+ "loss": 1.6399,
370
+ "step": 2600
371
+ },
372
+ {
373
+ "epoch": 4.88,
374
+ "learning_rate": 4.086666666666667e-06,
375
+ "loss": 1.68,
376
+ "step": 2650
377
+ },
378
+ {
379
+ "epoch": 4.97,
380
+ "learning_rate": 4.05888888888889e-06,
381
+ "loss": 1.5905,
382
+ "step": 2700
383
+ },
384
+ {
385
+ "epoch": 5.06,
386
+ "learning_rate": 4.031111111111111e-06,
387
+ "loss": 1.4962,
388
+ "step": 2750
389
+ },
390
+ {
391
+ "epoch": 5.16,
392
+ "learning_rate": 4.003333333333334e-06,
393
+ "loss": 1.283,
394
+ "step": 2800
395
+ },
396
+ {
397
+ "epoch": 5.25,
398
+ "learning_rate": 3.975555555555556e-06,
399
+ "loss": 1.5967,
400
+ "step": 2850
401
+ },
402
+ {
403
+ "epoch": 5.34,
404
+ "learning_rate": 3.9477777777777786e-06,
405
+ "loss": 1.627,
406
+ "step": 2900
407
+ },
408
+ {
409
+ "epoch": 5.43,
410
+ "learning_rate": 3.920000000000001e-06,
411
+ "loss": 1.6596,
412
+ "step": 2950
413
+ },
414
+ {
415
+ "epoch": 5.52,
416
+ "learning_rate": 3.892222222222223e-06,
417
+ "loss": 1.6403,
418
+ "step": 3000
419
+ },
420
+ {
421
+ "epoch": 5.52,
422
+ "eval_cer": 6.189555125725339,
423
+ "eval_loss": 0.3256286084651947,
424
+ "eval_runtime": 196.2193,
425
+ "eval_samples_per_second": 5.096,
426
+ "eval_steps_per_second": 0.321,
427
+ "eval_wer": 15.947611710323576,
428
+ "step": 3000
429
+ },
430
+ {
431
+ "epoch": 5.62,
432
+ "learning_rate": 3.864444444444445e-06,
433
+ "loss": 1.6811,
434
+ "step": 3050
435
+ },
436
+ {
437
+ "epoch": 5.71,
438
+ "learning_rate": 3.836666666666667e-06,
439
+ "loss": 1.3881,
440
+ "step": 3100
441
+ },
442
+ {
443
+ "epoch": 5.8,
444
+ "learning_rate": 3.808888888888889e-06,
445
+ "loss": 1.9165,
446
+ "step": 3150
447
+ },
448
+ {
449
+ "epoch": 5.89,
450
+ "learning_rate": 3.7811111111111115e-06,
451
+ "loss": 1.7782,
452
+ "step": 3200
453
+ },
454
+ {
455
+ "epoch": 5.99,
456
+ "learning_rate": 3.753333333333334e-06,
457
+ "loss": 1.6915,
458
+ "step": 3250
459
+ },
460
+ {
461
+ "epoch": 6.08,
462
+ "learning_rate": 3.7255555555555556e-06,
463
+ "loss": 1.6901,
464
+ "step": 3300
465
+ },
466
+ {
467
+ "epoch": 6.17,
468
+ "learning_rate": 3.697777777777778e-06,
469
+ "loss": 1.8462,
470
+ "step": 3350
471
+ },
472
+ {
473
+ "epoch": 6.26,
474
+ "learning_rate": 3.6700000000000004e-06,
475
+ "loss": 1.3781,
476
+ "step": 3400
477
+ },
478
+ {
479
+ "epoch": 6.35,
480
+ "learning_rate": 3.6422222222222225e-06,
481
+ "loss": 1.4356,
482
+ "step": 3450
483
+ },
484
+ {
485
+ "epoch": 6.45,
486
+ "learning_rate": 3.614444444444445e-06,
487
+ "loss": 1.536,
488
+ "step": 3500
489
+ },
490
+ {
491
+ "epoch": 6.45,
492
+ "eval_cer": 6.09284332688588,
493
+ "eval_loss": 0.32352736592292786,
494
+ "eval_runtime": 196.5479,
495
+ "eval_samples_per_second": 5.088,
496
+ "eval_steps_per_second": 0.321,
497
+ "eval_wer": 15.500770416024654,
498
+ "step": 3500
499
+ },
500
+ {
501
+ "epoch": 6.54,
502
+ "learning_rate": 3.5866666666666673e-06,
503
+ "loss": 1.4201,
504
+ "step": 3550
505
+ },
506
+ {
507
+ "epoch": 6.63,
508
+ "learning_rate": 3.558888888888889e-06,
509
+ "loss": 1.4551,
510
+ "step": 3600
511
+ },
512
+ {
513
+ "epoch": 6.72,
514
+ "learning_rate": 3.5311111111111114e-06,
515
+ "loss": 1.5927,
516
+ "step": 3650
517
+ },
518
+ {
519
+ "epoch": 6.81,
520
+ "learning_rate": 3.5033333333333334e-06,
521
+ "loss": 1.357,
522
+ "step": 3700
523
+ },
524
+ {
525
+ "epoch": 6.91,
526
+ "learning_rate": 3.475555555555556e-06,
527
+ "loss": 1.5396,
528
+ "step": 3750
529
+ },
530
+ {
531
+ "epoch": 7.0,
532
+ "learning_rate": 3.4477777777777783e-06,
533
+ "loss": 1.5673,
534
+ "step": 3800
535
+ },
536
+ {
537
+ "epoch": 7.09,
538
+ "learning_rate": 3.4200000000000007e-06,
539
+ "loss": 1.3435,
540
+ "step": 3850
541
+ },
542
+ {
543
+ "epoch": 7.18,
544
+ "learning_rate": 3.3922222222222223e-06,
545
+ "loss": 1.4499,
546
+ "step": 3900
547
+ },
548
+ {
549
+ "epoch": 7.27,
550
+ "learning_rate": 3.3644444444444447e-06,
551
+ "loss": 1.4198,
552
+ "step": 3950
553
+ },
554
+ {
555
+ "epoch": 7.37,
556
+ "learning_rate": 3.3366666666666668e-06,
557
+ "loss": 1.4223,
558
+ "step": 4000
559
+ },
560
+ {
561
+ "epoch": 7.37,
562
+ "eval_cer": 6.0735009671179885,
563
+ "eval_loss": 0.32085078954696655,
564
+ "eval_runtime": 197.2684,
565
+ "eval_samples_per_second": 5.069,
566
+ "eval_steps_per_second": 0.319,
567
+ "eval_wer": 15.362095531587059,
568
+ "step": 4000
569
+ },
570
+ {
571
+ "epoch": 7.46,
572
+ "learning_rate": 3.308888888888889e-06,
573
+ "loss": 1.5456,
574
+ "step": 4050
575
+ },
576
+ {
577
+ "epoch": 7.55,
578
+ "learning_rate": 3.2811111111111116e-06,
579
+ "loss": 1.2197,
580
+ "step": 4100
581
+ },
582
+ {
583
+ "epoch": 7.64,
584
+ "learning_rate": 3.2533333333333332e-06,
585
+ "loss": 1.4893,
586
+ "step": 4150
587
+ },
588
+ {
589
+ "epoch": 7.73,
590
+ "learning_rate": 3.2255555555555557e-06,
591
+ "loss": 1.4071,
592
+ "step": 4200
593
+ },
594
+ {
595
+ "epoch": 7.83,
596
+ "learning_rate": 3.197777777777778e-06,
597
+ "loss": 1.5669,
598
+ "step": 4250
599
+ },
600
+ {
601
+ "epoch": 7.92,
602
+ "learning_rate": 3.17e-06,
603
+ "loss": 1.3333,
604
+ "step": 4300
605
+ },
606
+ {
607
+ "epoch": 8.01,
608
+ "learning_rate": 3.1422222222222226e-06,
609
+ "loss": 1.4102,
610
+ "step": 4350
611
+ },
612
+ {
613
+ "epoch": 8.1,
614
+ "learning_rate": 3.114444444444445e-06,
615
+ "loss": 1.525,
616
+ "step": 4400
617
+ },
618
+ {
619
+ "epoch": 8.2,
620
+ "learning_rate": 3.0866666666666666e-06,
621
+ "loss": 1.6757,
622
+ "step": 4450
623
+ },
624
+ {
625
+ "epoch": 8.29,
626
+ "learning_rate": 3.058888888888889e-06,
627
+ "loss": 1.4652,
628
+ "step": 4500
629
+ },
630
+ {
631
+ "epoch": 8.29,
632
+ "eval_cer": 5.932578060237635,
633
+ "eval_loss": 0.3209191560745239,
634
+ "eval_runtime": 200.0634,
635
+ "eval_samples_per_second": 4.998,
636
+ "eval_steps_per_second": 0.315,
637
+ "eval_wer": 15.26964560862866,
638
+ "step": 4500
639
+ },
640
+ {
641
+ "epoch": 8.38,
642
+ "learning_rate": 3.0311111111111115e-06,
643
+ "loss": 1.6026,
644
+ "step": 4550
645
+ },
646
+ {
647
+ "epoch": 8.47,
648
+ "learning_rate": 3.0033333333333335e-06,
649
+ "loss": 1.425,
650
+ "step": 4600
651
+ },
652
+ {
653
+ "epoch": 8.56,
654
+ "learning_rate": 2.975555555555556e-06,
655
+ "loss": 1.548,
656
+ "step": 4650
657
+ },
658
+ {
659
+ "epoch": 8.66,
660
+ "learning_rate": 2.9477777777777784e-06,
661
+ "loss": 1.1192,
662
+ "step": 4700
663
+ },
664
+ {
665
+ "epoch": 8.75,
666
+ "learning_rate": 2.92e-06,
667
+ "loss": 1.5373,
668
+ "step": 4750
669
+ },
670
+ {
671
+ "epoch": 8.84,
672
+ "learning_rate": 2.8922222222222224e-06,
673
+ "loss": 1.4464,
674
+ "step": 4800
675
+ },
676
+ {
677
+ "epoch": 8.93,
678
+ "learning_rate": 2.864444444444445e-06,
679
+ "loss": 1.0499,
680
+ "step": 4850
681
+ },
682
+ {
683
+ "epoch": 9.02,
684
+ "learning_rate": 2.836666666666667e-06,
685
+ "loss": 1.2791,
686
+ "step": 4900
687
+ },
688
+ {
689
+ "epoch": 9.12,
690
+ "learning_rate": 2.8088888888888893e-06,
691
+ "loss": 1.2165,
692
+ "step": 4950
693
+ },
694
+ {
695
+ "epoch": 9.21,
696
+ "learning_rate": 2.781111111111111e-06,
697
+ "loss": 1.2572,
698
+ "step": 5000
699
+ },
700
+ {
701
+ "epoch": 9.21,
702
+ "eval_cer": 5.74468085106383,
703
+ "eval_loss": 0.3191336393356323,
704
+ "eval_runtime": 195.9428,
705
+ "eval_samples_per_second": 5.104,
706
+ "eval_steps_per_second": 0.322,
707
+ "eval_wer": 14.884437596302003,
708
+ "step": 5000
709
+ },
710
+ {
711
+ "epoch": 9.3,
712
+ "learning_rate": 2.7533333333333334e-06,
713
+ "loss": 1.4301,
714
+ "step": 5050
715
+ },
716
+ {
717
+ "epoch": 9.39,
718
+ "learning_rate": 2.725555555555556e-06,
719
+ "loss": 1.4087,
720
+ "step": 5100
721
+ },
722
+ {
723
+ "epoch": 9.48,
724
+ "learning_rate": 2.6977777777777783e-06,
725
+ "loss": 1.6344,
726
+ "step": 5150
727
+ },
728
+ {
729
+ "epoch": 9.58,
730
+ "learning_rate": 2.6700000000000003e-06,
731
+ "loss": 1.4237,
732
+ "step": 5200
733
+ },
734
+ {
735
+ "epoch": 9.67,
736
+ "learning_rate": 2.6422222222222227e-06,
737
+ "loss": 1.5428,
738
+ "step": 5250
739
+ },
740
+ {
741
+ "epoch": 9.76,
742
+ "learning_rate": 2.6144444444444443e-06,
743
+ "loss": 1.2484,
744
+ "step": 5300
745
+ },
746
+ {
747
+ "epoch": 9.85,
748
+ "learning_rate": 2.5866666666666667e-06,
749
+ "loss": 1.424,
750
+ "step": 5350
751
+ },
752
+ {
753
+ "epoch": 9.94,
754
+ "learning_rate": 2.558888888888889e-06,
755
+ "loss": 1.4562,
756
+ "step": 5400
757
+ },
758
+ {
759
+ "epoch": 10.04,
760
+ "learning_rate": 2.531111111111111e-06,
761
+ "loss": 1.2812,
762
+ "step": 5450
763
+ },
764
+ {
765
+ "epoch": 10.13,
766
+ "learning_rate": 2.5033333333333336e-06,
767
+ "loss": 1.7142,
768
+ "step": 5500
769
+ },
770
+ {
771
+ "epoch": 10.13,
772
+ "eval_cer": 5.8469190384084,
773
+ "eval_loss": 0.3182465732097626,
774
+ "eval_runtime": 195.9115,
775
+ "eval_samples_per_second": 5.104,
776
+ "eval_steps_per_second": 0.322,
777
+ "eval_wer": 15.007704160246533,
778
+ "step": 5500
779
+ },
780
+ {
781
+ "epoch": 10.22,
782
+ "learning_rate": 2.4755555555555557e-06,
783
+ "loss": 1.238,
784
+ "step": 5550
785
+ },
786
+ {
787
+ "epoch": 10.31,
788
+ "learning_rate": 2.447777777777778e-06,
789
+ "loss": 1.178,
790
+ "step": 5600
791
+ },
792
+ {
793
+ "epoch": 10.41,
794
+ "learning_rate": 2.42e-06,
795
+ "loss": 1.3038,
796
+ "step": 5650
797
+ },
798
+ {
799
+ "epoch": 10.5,
800
+ "learning_rate": 2.3922222222222226e-06,
801
+ "loss": 1.3901,
802
+ "step": 5700
803
+ },
804
+ {
805
+ "epoch": 10.59,
806
+ "learning_rate": 2.3644444444444446e-06,
807
+ "loss": 1.4346,
808
+ "step": 5750
809
+ },
810
+ {
811
+ "epoch": 10.68,
812
+ "learning_rate": 2.3366666666666666e-06,
813
+ "loss": 1.3559,
814
+ "step": 5800
815
+ },
816
+ {
817
+ "epoch": 10.77,
818
+ "learning_rate": 2.308888888888889e-06,
819
+ "loss": 1.2192,
820
+ "step": 5850
821
+ },
822
+ {
823
+ "epoch": 10.87,
824
+ "learning_rate": 2.281111111111111e-06,
825
+ "loss": 1.0437,
826
+ "step": 5900
827
+ },
828
+ {
829
+ "epoch": 10.96,
830
+ "learning_rate": 2.2533333333333335e-06,
831
+ "loss": 1.8052,
832
+ "step": 5950
833
+ },
834
+ {
835
+ "epoch": 11.05,
836
+ "learning_rate": 2.225555555555556e-06,
837
+ "loss": 1.4195,
838
+ "step": 6000
839
+ },
840
+ {
841
+ "epoch": 11.05,
842
+ "eval_cer": 5.885603757944184,
843
+ "eval_loss": 0.3170944154262543,
844
+ "eval_runtime": 198.0232,
845
+ "eval_samples_per_second": 5.05,
846
+ "eval_steps_per_second": 0.318,
847
+ "eval_wer": 15.069337442218798,
848
+ "step": 6000
849
+ },
850
+ {
851
+ "epoch": 11.14,
852
+ "learning_rate": 2.197777777777778e-06,
853
+ "loss": 1.1544,
854
+ "step": 6050
855
+ },
856
+ {
857
+ "epoch": 11.23,
858
+ "learning_rate": 2.17e-06,
859
+ "loss": 1.3013,
860
+ "step": 6100
861
+ },
862
+ {
863
+ "epoch": 11.33,
864
+ "learning_rate": 2.1422222222222224e-06,
865
+ "loss": 1.4788,
866
+ "step": 6150
867
+ },
868
+ {
869
+ "epoch": 11.42,
870
+ "learning_rate": 2.1144444444444444e-06,
871
+ "loss": 1.345,
872
+ "step": 6200
873
+ },
874
+ {
875
+ "epoch": 11.51,
876
+ "learning_rate": 2.086666666666667e-06,
877
+ "loss": 1.4488,
878
+ "step": 6250
879
+ },
880
+ {
881
+ "epoch": 11.6,
882
+ "learning_rate": 2.0588888888888893e-06,
883
+ "loss": 1.0227,
884
+ "step": 6300
885
+ },
886
+ {
887
+ "epoch": 11.69,
888
+ "learning_rate": 2.0311111111111113e-06,
889
+ "loss": 1.4288,
890
+ "step": 6350
891
+ },
892
+ {
893
+ "epoch": 11.79,
894
+ "learning_rate": 2.0033333333333334e-06,
895
+ "loss": 1.0756,
896
+ "step": 6400
897
+ },
898
+ {
899
+ "epoch": 11.88,
900
+ "learning_rate": 1.975555555555556e-06,
901
+ "loss": 1.3951,
902
+ "step": 6450
903
+ },
904
+ {
905
+ "epoch": 11.97,
906
+ "learning_rate": 1.947777777777778e-06,
907
+ "loss": 1.3965,
908
+ "step": 6500
909
+ },
910
+ {
911
+ "epoch": 11.97,
912
+ "eval_cer": 5.857971815418623,
913
+ "eval_loss": 0.3166500926017761,
914
+ "eval_runtime": 195.9821,
915
+ "eval_samples_per_second": 5.103,
916
+ "eval_steps_per_second": 0.321,
917
+ "eval_wer": 15.053929121725732,
918
+ "step": 6500
919
+ },
920
+ {
921
+ "epoch": 11.97,
922
+ "step": 6500,
923
+ "total_flos": 6.002576326656e+19,
924
+ "train_loss": 1.8605577674278846,
925
+ "train_runtime": 13282.9377,
926
+ "train_samples_per_second": 24.091,
927
+ "train_steps_per_second": 0.753
928
+ }
929
+ ],
930
+ "max_steps": 10000,
931
+ "num_train_epochs": 19,
932
+ "total_flos": 6.002576326656e+19,
933
+ "trial_name": null,
934
+ "trial_params": null
935
+ }