wannaphong commited on
Commit
458eef3
1 Parent(s): bd124fd

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1471 -0
trainer_state.json ADDED
@@ -0,0 +1,1471 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.0,
3
+ "best_model_checkpoint": "./models/checkpoint-1003",
4
+ "epoch": 97.0,
5
+ "global_step": 97291,
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
+ "learning_rate": 9.98e-05,
13
+ "loss": 6.8202,
14
+ "step": 1003
15
+ },
16
+ {
17
+ "epoch": 1.0,
18
+ "eval_loss": 3.5011327266693115,
19
+ "eval_runtime": 260.4478,
20
+ "eval_samples_per_second": 17.735,
21
+ "eval_steps_per_second": 0.372,
22
+ "eval_wer": 1.0,
23
+ "step": 1003
24
+ },
25
+ {
26
+ "epoch": 2.0,
27
+ "learning_rate": 9.899194360523666e-05,
28
+ "loss": 1.0758,
29
+ "step": 2006
30
+ },
31
+ {
32
+ "epoch": 2.0,
33
+ "eval_loss": 0.5050923228263855,
34
+ "eval_runtime": 330.4397,
35
+ "eval_samples_per_second": 13.978,
36
+ "eval_steps_per_second": 0.294,
37
+ "eval_wer": 0.4667033908679153,
38
+ "step": 2006
39
+ },
40
+ {
41
+ "epoch": 3.0,
42
+ "learning_rate": 9.798187311178248e-05,
43
+ "loss": 0.5297,
44
+ "step": 3009
45
+ },
46
+ {
47
+ "epoch": 3.0,
48
+ "eval_loss": 0.3798016309738159,
49
+ "eval_runtime": 325.2274,
50
+ "eval_samples_per_second": 14.202,
51
+ "eval_steps_per_second": 0.298,
52
+ "eval_wer": 0.39802913453299055,
53
+ "step": 3009
54
+ },
55
+ {
56
+ "epoch": 4.0,
57
+ "learning_rate": 9.69718026183283e-05,
58
+ "loss": 0.4266,
59
+ "step": 4012
60
+ },
61
+ {
62
+ "epoch": 4.0,
63
+ "eval_loss": 0.32800063490867615,
64
+ "eval_runtime": 361.1443,
65
+ "eval_samples_per_second": 12.79,
66
+ "eval_steps_per_second": 0.269,
67
+ "eval_wer": 0.36733382298934997,
68
+ "step": 4012
69
+ },
70
+ {
71
+ "epoch": 5.0,
72
+ "learning_rate": 9.596173212487412e-05,
73
+ "loss": 0.3668,
74
+ "step": 5015
75
+ },
76
+ {
77
+ "epoch": 5.0,
78
+ "eval_loss": 0.31621453166007996,
79
+ "eval_runtime": 367.1388,
80
+ "eval_samples_per_second": 12.581,
81
+ "eval_steps_per_second": 0.264,
82
+ "eval_wer": 0.3481760313379851,
83
+ "step": 5015
84
+ },
85
+ {
86
+ "epoch": 6.0,
87
+ "learning_rate": 9.495166163141995e-05,
88
+ "loss": 0.326,
89
+ "step": 6018
90
+ },
91
+ {
92
+ "epoch": 6.0,
93
+ "eval_loss": 0.29389986395835876,
94
+ "eval_runtime": 416.389,
95
+ "eval_samples_per_second": 11.093,
96
+ "eval_steps_per_second": 0.233,
97
+ "eval_wer": 0.33691394295507404,
98
+ "step": 6018
99
+ },
100
+ {
101
+ "epoch": 7.0,
102
+ "learning_rate": 9.394159113796577e-05,
103
+ "loss": 0.2961,
104
+ "step": 7021
105
+ },
106
+ {
107
+ "epoch": 7.0,
108
+ "eval_loss": 0.28012460470199585,
109
+ "eval_runtime": 270.2817,
110
+ "eval_samples_per_second": 17.09,
111
+ "eval_steps_per_second": 0.359,
112
+ "eval_wer": 0.3236320235034888,
113
+ "step": 7021
114
+ },
115
+ {
116
+ "epoch": 8.0,
117
+ "learning_rate": 9.293152064451159e-05,
118
+ "loss": 0.2751,
119
+ "step": 8024
120
+ },
121
+ {
122
+ "epoch": 8.0,
123
+ "eval_loss": 0.2852382957935333,
124
+ "eval_runtime": 304.5141,
125
+ "eval_samples_per_second": 15.168,
126
+ "eval_steps_per_second": 0.319,
127
+ "eval_wer": 0.31775615130370916,
128
+ "step": 8024
129
+ },
130
+ {
131
+ "epoch": 9.0,
132
+ "learning_rate": 9.192145015105741e-05,
133
+ "loss": 0.2561,
134
+ "step": 9027
135
+ },
136
+ {
137
+ "epoch": 9.0,
138
+ "eval_loss": 0.26253288984298706,
139
+ "eval_runtime": 284.9667,
140
+ "eval_samples_per_second": 16.209,
141
+ "eval_steps_per_second": 0.34,
142
+ "eval_wer": 0.30986044803525525,
143
+ "step": 9027
144
+ },
145
+ {
146
+ "epoch": 10.0,
147
+ "learning_rate": 9.091137965760322e-05,
148
+ "loss": 0.2432,
149
+ "step": 10030
150
+ },
151
+ {
152
+ "epoch": 10.0,
153
+ "eval_loss": 0.2666085958480835,
154
+ "eval_runtime": 406.9804,
155
+ "eval_samples_per_second": 11.349,
156
+ "eval_steps_per_second": 0.238,
157
+ "eval_wer": 0.30536173338229894,
158
+ "step": 10030
159
+ },
160
+ {
161
+ "epoch": 11.0,
162
+ "learning_rate": 8.990130916414905e-05,
163
+ "loss": 0.2242,
164
+ "step": 11033
165
+ },
166
+ {
167
+ "epoch": 11.0,
168
+ "eval_loss": 0.269761860370636,
169
+ "eval_runtime": 400.3745,
170
+ "eval_samples_per_second": 11.537,
171
+ "eval_steps_per_second": 0.242,
172
+ "eval_wer": 0.30089362223038313,
173
+ "step": 11033
174
+ },
175
+ {
176
+ "epoch": 12.0,
177
+ "learning_rate": 8.889123867069486e-05,
178
+ "loss": 0.2153,
179
+ "step": 12036
180
+ },
181
+ {
182
+ "epoch": 12.0,
183
+ "eval_loss": 0.2758134603500366,
184
+ "eval_runtime": 389.1981,
185
+ "eval_samples_per_second": 11.868,
186
+ "eval_steps_per_second": 0.249,
187
+ "eval_wer": 0.30067939772309954,
188
+ "step": 12036
189
+ },
190
+ {
191
+ "epoch": 13.0,
192
+ "learning_rate": 8.78821752265861e-05,
193
+ "loss": 0.2036,
194
+ "step": 13039
195
+ },
196
+ {
197
+ "epoch": 13.0,
198
+ "eval_loss": 0.2647706866264343,
199
+ "eval_runtime": 436.7457,
200
+ "eval_samples_per_second": 10.576,
201
+ "eval_steps_per_second": 0.222,
202
+ "eval_wer": 0.29743542661280453,
203
+ "step": 13039
204
+ },
205
+ {
206
+ "epoch": 14.0,
207
+ "learning_rate": 8.687210473313193e-05,
208
+ "loss": 0.1972,
209
+ "step": 14042
210
+ },
211
+ {
212
+ "epoch": 14.0,
213
+ "eval_loss": 0.2667146325111389,
214
+ "eval_runtime": 402.5926,
215
+ "eval_samples_per_second": 11.473,
216
+ "eval_steps_per_second": 0.241,
217
+ "eval_wer": 0.29238584894111885,
218
+ "step": 14042
219
+ },
220
+ {
221
+ "epoch": 15.0,
222
+ "learning_rate": 8.586203423967774e-05,
223
+ "loss": 0.186,
224
+ "step": 15045
225
+ },
226
+ {
227
+ "epoch": 15.0,
228
+ "eval_loss": 0.2569558024406433,
229
+ "eval_runtime": 397.1089,
230
+ "eval_samples_per_second": 11.632,
231
+ "eval_steps_per_second": 0.244,
232
+ "eval_wer": 0.29097808789325497,
233
+ "step": 15045
234
+ },
235
+ {
236
+ "epoch": 16.0,
237
+ "learning_rate": 8.485196374622358e-05,
238
+ "loss": 0.179,
239
+ "step": 16048
240
+ },
241
+ {
242
+ "epoch": 16.0,
243
+ "eval_loss": 0.270986944437027,
244
+ "eval_runtime": 443.1466,
245
+ "eval_samples_per_second": 10.423,
246
+ "eval_steps_per_second": 0.219,
247
+ "eval_wer": 0.289386705839148,
248
+ "step": 16048
249
+ },
250
+ {
251
+ "epoch": 17.0,
252
+ "learning_rate": 8.384390735146023e-05,
253
+ "loss": 0.1721,
254
+ "step": 17051
255
+ },
256
+ {
257
+ "epoch": 17.0,
258
+ "eval_loss": 0.2602141797542572,
259
+ "eval_runtime": 438.2891,
260
+ "eval_samples_per_second": 10.539,
261
+ "eval_steps_per_second": 0.221,
262
+ "eval_wer": 0.2873056677683927,
263
+ "step": 17051
264
+ },
265
+ {
266
+ "epoch": 18.0,
267
+ "learning_rate": 8.283383685800604e-05,
268
+ "loss": 0.1656,
269
+ "step": 18054
270
+ },
271
+ {
272
+ "epoch": 18.0,
273
+ "eval_loss": 0.26877614855766296,
274
+ "eval_runtime": 441.5132,
275
+ "eval_samples_per_second": 10.462,
276
+ "eval_steps_per_second": 0.22,
277
+ "eval_wer": 0.28403109315705716,
278
+ "step": 18054
279
+ },
280
+ {
281
+ "epoch": 19.0,
282
+ "learning_rate": 8.182376636455186e-05,
283
+ "loss": 0.1613,
284
+ "step": 19057
285
+ },
286
+ {
287
+ "epoch": 19.0,
288
+ "eval_loss": 0.28074949979782104,
289
+ "eval_runtime": 481.2314,
290
+ "eval_samples_per_second": 9.598,
291
+ "eval_steps_per_second": 0.202,
292
+ "eval_wer": 0.28369445464561144,
293
+ "step": 19057
294
+ },
295
+ {
296
+ "epoch": 20.0,
297
+ "learning_rate": 8.081369587109768e-05,
298
+ "loss": 0.1546,
299
+ "step": 20060
300
+ },
301
+ {
302
+ "epoch": 20.0,
303
+ "eval_loss": 0.26929938793182373,
304
+ "eval_runtime": 412.4546,
305
+ "eval_samples_per_second": 11.199,
306
+ "eval_steps_per_second": 0.235,
307
+ "eval_wer": 0.2782776349614396,
308
+ "step": 20060
309
+ },
310
+ {
311
+ "epoch": 21.0,
312
+ "learning_rate": 7.980463242698892e-05,
313
+ "loss": 0.1506,
314
+ "step": 21063
315
+ },
316
+ {
317
+ "epoch": 21.0,
318
+ "eval_loss": 0.2805509865283966,
319
+ "eval_runtime": 402.9001,
320
+ "eval_samples_per_second": 11.464,
321
+ "eval_steps_per_second": 0.241,
322
+ "eval_wer": 0.28161341657485617,
323
+ "step": 21063
324
+ },
325
+ {
326
+ "epoch": 22.0,
327
+ "learning_rate": 7.879456193353474e-05,
328
+ "loss": 0.1459,
329
+ "step": 22066
330
+ },
331
+ {
332
+ "epoch": 22.0,
333
+ "eval_loss": 0.28382739424705505,
334
+ "eval_runtime": 393.1186,
335
+ "eval_samples_per_second": 11.75,
336
+ "eval_steps_per_second": 0.247,
337
+ "eval_wer": 0.27885910148120946,
338
+ "step": 22066
339
+ },
340
+ {
341
+ "epoch": 23.0,
342
+ "learning_rate": 7.778449144008056e-05,
343
+ "loss": 0.14,
344
+ "step": 23069
345
+ },
346
+ {
347
+ "epoch": 23.0,
348
+ "eval_loss": 0.2746136486530304,
349
+ "eval_runtime": 425.1655,
350
+ "eval_samples_per_second": 10.864,
351
+ "eval_steps_per_second": 0.228,
352
+ "eval_wer": 0.2754315093646713,
353
+ "step": 23069
354
+ },
355
+ {
356
+ "epoch": 24.0,
357
+ "learning_rate": 7.677442094662639e-05,
358
+ "loss": 0.1367,
359
+ "step": 24072
360
+ },
361
+ {
362
+ "epoch": 24.0,
363
+ "eval_loss": 0.2868611514568329,
364
+ "eval_runtime": 427.1205,
365
+ "eval_samples_per_second": 10.814,
366
+ "eval_steps_per_second": 0.227,
367
+ "eval_wer": 0.27466642183865836,
368
+ "step": 24072
369
+ },
370
+ {
371
+ "epoch": 25.0,
372
+ "learning_rate": 7.576435045317221e-05,
373
+ "loss": 0.1303,
374
+ "step": 25075
375
+ },
376
+ {
377
+ "epoch": 25.0,
378
+ "eval_loss": 0.26962676644325256,
379
+ "eval_runtime": 414.5701,
380
+ "eval_samples_per_second": 11.142,
381
+ "eval_steps_per_second": 0.234,
382
+ "eval_wer": 0.27371771330640227,
383
+ "step": 25075
384
+ },
385
+ {
386
+ "epoch": 26.0,
387
+ "learning_rate": 7.475528700906345e-05,
388
+ "loss": 0.1299,
389
+ "step": 26078
390
+ },
391
+ {
392
+ "epoch": 26.0,
393
+ "eval_loss": 0.2933480143547058,
394
+ "eval_runtime": 423.3941,
395
+ "eval_samples_per_second": 10.909,
396
+ "eval_steps_per_second": 0.229,
397
+ "eval_wer": 0.27429917982617213,
398
+ "step": 26078
399
+ },
400
+ {
401
+ "epoch": 27.0,
402
+ "learning_rate": 7.374521651560927e-05,
403
+ "loss": 0.1241,
404
+ "step": 27081
405
+ },
406
+ {
407
+ "epoch": 27.0,
408
+ "eval_loss": 0.2790047824382782,
409
+ "eval_runtime": 413.5975,
410
+ "eval_samples_per_second": 11.168,
411
+ "eval_steps_per_second": 0.235,
412
+ "eval_wer": 0.27019831068674255,
413
+ "step": 27081
414
+ },
415
+ {
416
+ "epoch": 28.0,
417
+ "learning_rate": 7.273615307150052e-05,
418
+ "loss": 0.122,
419
+ "step": 28084
420
+ },
421
+ {
422
+ "epoch": 28.0,
423
+ "eval_loss": 0.29522669315338135,
424
+ "eval_runtime": 397.9723,
425
+ "eval_samples_per_second": 11.606,
426
+ "eval_steps_per_second": 0.244,
427
+ "eval_wer": 0.27243236626270045,
428
+ "step": 28084
429
+ },
430
+ {
431
+ "epoch": 29.0,
432
+ "learning_rate": 7.172708962739174e-05,
433
+ "loss": 0.1195,
434
+ "step": 29087
435
+ },
436
+ {
437
+ "epoch": 29.0,
438
+ "eval_loss": 0.2860444486141205,
439
+ "eval_runtime": 401.6884,
440
+ "eval_samples_per_second": 11.499,
441
+ "eval_steps_per_second": 0.241,
442
+ "eval_wer": 0.26967805116905375,
443
+ "step": 29087
444
+ },
445
+ {
446
+ "epoch": 30.0,
447
+ "learning_rate": 7.071701913393757e-05,
448
+ "loss": 0.1171,
449
+ "step": 30090
450
+ },
451
+ {
452
+ "epoch": 30.0,
453
+ "eval_loss": 0.30656594038009644,
454
+ "eval_runtime": 400.5872,
455
+ "eval_samples_per_second": 11.531,
456
+ "eval_steps_per_second": 0.242,
457
+ "eval_wer": 0.27148365773044436,
458
+ "step": 30090
459
+ },
460
+ {
461
+ "epoch": 31.0,
462
+ "learning_rate": 6.970694864048339e-05,
463
+ "loss": 0.1131,
464
+ "step": 31093
465
+ },
466
+ {
467
+ "epoch": 31.0,
468
+ "eval_loss": 0.3087522089481354,
469
+ "eval_runtime": 410.4211,
470
+ "eval_samples_per_second": 11.254,
471
+ "eval_steps_per_second": 0.236,
472
+ "eval_wer": 0.26646468355979924,
473
+ "step": 31093
474
+ },
475
+ {
476
+ "epoch": 32.0,
477
+ "learning_rate": 6.869687814702921e-05,
478
+ "loss": 0.1111,
479
+ "step": 32096
480
+ },
481
+ {
482
+ "epoch": 32.0,
483
+ "eval_loss": 0.305497407913208,
484
+ "eval_runtime": 397.2313,
485
+ "eval_samples_per_second": 11.628,
486
+ "eval_steps_per_second": 0.244,
487
+ "eval_wer": 0.26823968662014935,
488
+ "step": 32096
489
+ },
490
+ {
491
+ "epoch": 33.0,
492
+ "learning_rate": 6.768680765357503e-05,
493
+ "loss": 0.1081,
494
+ "step": 33099
495
+ },
496
+ {
497
+ "epoch": 33.0,
498
+ "eval_loss": 0.3094805181026459,
499
+ "eval_runtime": 398.4287,
500
+ "eval_samples_per_second": 11.593,
501
+ "eval_steps_per_second": 0.243,
502
+ "eval_wer": 0.2667401150691639,
503
+ "step": 33099
504
+ },
505
+ {
506
+ "epoch": 34.0,
507
+ "learning_rate": 6.667673716012085e-05,
508
+ "loss": 0.106,
509
+ "step": 34102
510
+ },
511
+ {
512
+ "epoch": 34.0,
513
+ "eval_loss": 0.31470388174057007,
514
+ "eval_runtime": 394.7251,
515
+ "eval_samples_per_second": 11.702,
516
+ "eval_steps_per_second": 0.246,
517
+ "eval_wer": 0.2654241645244216,
518
+ "step": 34102
519
+ },
520
+ {
521
+ "epoch": 35.0,
522
+ "learning_rate": 6.566666666666666e-05,
523
+ "loss": 0.1023,
524
+ "step": 35105
525
+ },
526
+ {
527
+ "epoch": 35.0,
528
+ "eval_loss": 0.32143208384513855,
529
+ "eval_runtime": 406.2767,
530
+ "eval_samples_per_second": 11.369,
531
+ "eval_steps_per_second": 0.239,
532
+ "eval_wer": 0.2643530419880034,
533
+ "step": 35105
534
+ },
535
+ {
536
+ "epoch": 36.0,
537
+ "learning_rate": 6.46565961732125e-05,
538
+ "loss": 0.1031,
539
+ "step": 36108
540
+ },
541
+ {
542
+ "epoch": 36.0,
543
+ "eval_loss": 0.3046073317527771,
544
+ "eval_runtime": 414.3275,
545
+ "eval_samples_per_second": 11.148,
546
+ "eval_steps_per_second": 0.234,
547
+ "eval_wer": 0.26306769494430166,
548
+ "step": 36108
549
+ },
550
+ {
551
+ "epoch": 37.0,
552
+ "learning_rate": 6.364853977844915e-05,
553
+ "loss": 0.1008,
554
+ "step": 37111
555
+ },
556
+ {
557
+ "epoch": 37.0,
558
+ "eval_loss": 0.3285815715789795,
559
+ "eval_runtime": 474.4359,
560
+ "eval_samples_per_second": 9.736,
561
+ "eval_steps_per_second": 0.204,
562
+ "eval_wer": 0.2653323540213,
563
+ "step": 37111
564
+ },
565
+ {
566
+ "epoch": 38.0,
567
+ "learning_rate": 6.263846928499497e-05,
568
+ "loss": 0.0977,
569
+ "step": 38114
570
+ },
571
+ {
572
+ "epoch": 38.0,
573
+ "eval_loss": 0.2976115047931671,
574
+ "eval_runtime": 433.6527,
575
+ "eval_samples_per_second": 10.651,
576
+ "eval_steps_per_second": 0.224,
577
+ "eval_wer": 0.2623638144203697,
578
+ "step": 38114
579
+ },
580
+ {
581
+ "epoch": 39.0,
582
+ "learning_rate": 6.162839879154079e-05,
583
+ "loss": 0.0951,
584
+ "step": 39117
585
+ },
586
+ {
587
+ "epoch": 39.0,
588
+ "eval_loss": 0.3340004086494446,
589
+ "eval_runtime": 479.8072,
590
+ "eval_samples_per_second": 9.627,
591
+ "eval_steps_per_second": 0.202,
592
+ "eval_wer": 0.26282286693597745,
593
+ "step": 39117
594
+ },
595
+ {
596
+ "epoch": 40.0,
597
+ "learning_rate": 6.0618328298086614e-05,
598
+ "loss": 0.0923,
599
+ "step": 40120
600
+ },
601
+ {
602
+ "epoch": 40.0,
603
+ "eval_loss": 0.3238993287086487,
604
+ "eval_runtime": 375.954,
605
+ "eval_samples_per_second": 12.286,
606
+ "eval_steps_per_second": 0.258,
607
+ "eval_wer": 0.26407761047863876,
608
+ "step": 40120
609
+ },
610
+ {
611
+ "epoch": 41.0,
612
+ "learning_rate": 5.960926485397785e-05,
613
+ "loss": 0.0931,
614
+ "step": 41123
615
+ },
616
+ {
617
+ "epoch": 41.0,
618
+ "eval_loss": 0.3135124444961548,
619
+ "eval_runtime": 457.0228,
620
+ "eval_samples_per_second": 10.107,
621
+ "eval_steps_per_second": 0.212,
622
+ "eval_wer": 0.2601909658464928,
623
+ "step": 41123
624
+ },
625
+ {
626
+ "epoch": 42.0,
627
+ "learning_rate": 5.860020140986908e-05,
628
+ "loss": 0.0902,
629
+ "step": 42126
630
+ },
631
+ {
632
+ "epoch": 42.0,
633
+ "eval_loss": 0.3057383894920349,
634
+ "eval_runtime": 459.7321,
635
+ "eval_samples_per_second": 10.047,
636
+ "eval_steps_per_second": 0.211,
637
+ "eval_wer": 0.25957889582568244,
638
+ "step": 42126
639
+ },
640
+ {
641
+ "epoch": 43.0,
642
+ "learning_rate": 5.759013091641491e-05,
643
+ "loss": 0.0899,
644
+ "step": 43129
645
+ },
646
+ {
647
+ "epoch": 43.0,
648
+ "eval_loss": 0.3292476534843445,
649
+ "eval_runtime": 439.0502,
650
+ "eval_samples_per_second": 10.52,
651
+ "eval_steps_per_second": 0.221,
652
+ "eval_wer": 0.26068062186314117,
653
+ "step": 43129
654
+ },
655
+ {
656
+ "epoch": 44.0,
657
+ "learning_rate": 5.658106747230615e-05,
658
+ "loss": 0.0873,
659
+ "step": 44132
660
+ },
661
+ {
662
+ "epoch": 44.0,
663
+ "eval_loss": 0.3199455142021179,
664
+ "eval_runtime": 360.0807,
665
+ "eval_samples_per_second": 12.828,
666
+ "eval_steps_per_second": 0.269,
667
+ "eval_wer": 0.2638633859713551,
668
+ "step": 44132
669
+ },
670
+ {
671
+ "epoch": 45.0,
672
+ "learning_rate": 5.5570996978851965e-05,
673
+ "loss": 0.0856,
674
+ "step": 45135
675
+ },
676
+ {
677
+ "epoch": 45.0,
678
+ "eval_loss": 0.33823832869529724,
679
+ "eval_runtime": 343.4666,
680
+ "eval_samples_per_second": 13.448,
681
+ "eval_steps_per_second": 0.282,
682
+ "eval_wer": 0.2601909658464928,
683
+ "step": 45135
684
+ },
685
+ {
686
+ "epoch": 46.0,
687
+ "learning_rate": 5.45619335347432e-05,
688
+ "loss": 0.084,
689
+ "step": 46138
690
+ },
691
+ {
692
+ "epoch": 46.0,
693
+ "eval_loss": 0.33022576570510864,
694
+ "eval_runtime": 342.0224,
695
+ "eval_samples_per_second": 13.505,
696
+ "eval_steps_per_second": 0.284,
697
+ "eval_wer": 0.25795691027053497,
698
+ "step": 46138
699
+ },
700
+ {
701
+ "epoch": 47.0,
702
+ "learning_rate": 5.355186304128903e-05,
703
+ "loss": 0.0815,
704
+ "step": 47141
705
+ },
706
+ {
707
+ "epoch": 47.0,
708
+ "eval_loss": 0.32415875792503357,
709
+ "eval_runtime": 344.3556,
710
+ "eval_samples_per_second": 13.413,
711
+ "eval_steps_per_second": 0.282,
712
+ "eval_wer": 0.26098665687354633,
713
+ "step": 47141
714
+ },
715
+ {
716
+ "epoch": 48.0,
717
+ "learning_rate": 5.2541792547834846e-05,
718
+ "loss": 0.0815,
719
+ "step": 48144
720
+ },
721
+ {
722
+ "epoch": 48.0,
723
+ "eval_loss": 0.3343857526779175,
724
+ "eval_runtime": 343.7802,
725
+ "eval_samples_per_second": 13.436,
726
+ "eval_steps_per_second": 0.282,
727
+ "eval_wer": 0.2582629452809401,
728
+ "step": 48144
729
+ },
730
+ {
731
+ "epoch": 49.0,
732
+ "learning_rate": 5.1532729103726084e-05,
733
+ "loss": 0.0785,
734
+ "step": 49147
735
+ },
736
+ {
737
+ "epoch": 49.0,
738
+ "eval_loss": 0.34900155663490295,
739
+ "eval_runtime": 343.1634,
740
+ "eval_samples_per_second": 13.46,
741
+ "eval_steps_per_second": 0.283,
742
+ "eval_wer": 0.25798751377157547,
743
+ "step": 49147
744
+ },
745
+ {
746
+ "epoch": 50.0,
747
+ "learning_rate": 5.052265861027191e-05,
748
+ "loss": 0.0777,
749
+ "step": 50150
750
+ },
751
+ {
752
+ "epoch": 50.0,
753
+ "eval_loss": 0.34195998311042786,
754
+ "eval_runtime": 299.1953,
755
+ "eval_samples_per_second": 15.438,
756
+ "eval_steps_per_second": 0.324,
757
+ "eval_wer": 0.2567633737299547,
758
+ "step": 50150
759
+ },
760
+ {
761
+ "epoch": 51.0,
762
+ "learning_rate": 4.951258811681773e-05,
763
+ "loss": 0.0769,
764
+ "step": 51153
765
+ },
766
+ {
767
+ "epoch": 51.0,
768
+ "eval_loss": 0.33568263053894043,
769
+ "eval_runtime": 342.9594,
770
+ "eval_samples_per_second": 13.468,
771
+ "eval_steps_per_second": 0.283,
772
+ "eval_wer": 0.25716121924348145,
773
+ "step": 51153
774
+ },
775
+ {
776
+ "epoch": 52.0,
777
+ "learning_rate": 4.8503524672708965e-05,
778
+ "loss": 0.0775,
779
+ "step": 52156
780
+ },
781
+ {
782
+ "epoch": 52.0,
783
+ "eval_loss": 0.32406118512153625,
784
+ "eval_runtime": 315.6469,
785
+ "eval_samples_per_second": 14.633,
786
+ "eval_steps_per_second": 0.307,
787
+ "eval_wer": 0.2555698371893745,
788
+ "step": 52156
789
+ },
790
+ {
791
+ "epoch": 53.0,
792
+ "learning_rate": 4.749345417925479e-05,
793
+ "loss": 0.0751,
794
+ "step": 53159
795
+ },
796
+ {
797
+ "epoch": 53.0,
798
+ "eval_loss": 0.3319285213947296,
799
+ "eval_runtime": 350.258,
800
+ "eval_samples_per_second": 13.187,
801
+ "eval_steps_per_second": 0.277,
802
+ "eval_wer": 0.25670216672787366,
803
+ "step": 53159
804
+ },
805
+ {
806
+ "epoch": 54.0,
807
+ "learning_rate": 4.6484390735146025e-05,
808
+ "loss": 0.0745,
809
+ "step": 54162
810
+ },
811
+ {
812
+ "epoch": 54.0,
813
+ "eval_loss": 0.34319552779197693,
814
+ "eval_runtime": 461.2436,
815
+ "eval_samples_per_second": 10.014,
816
+ "eval_steps_per_second": 0.21,
817
+ "eval_wer": 0.25440690414983474,
818
+ "step": 54162
819
+ },
820
+ {
821
+ "epoch": 55.0,
822
+ "learning_rate": 4.547432024169184e-05,
823
+ "loss": 0.0731,
824
+ "step": 55165
825
+ },
826
+ {
827
+ "epoch": 55.0,
828
+ "eval_loss": 0.340921550989151,
829
+ "eval_runtime": 375.608,
830
+ "eval_samples_per_second": 12.297,
831
+ "eval_steps_per_second": 0.258,
832
+ "eval_wer": 0.25575345819561757,
833
+ "step": 55165
834
+ },
835
+ {
836
+ "epoch": 56.0,
837
+ "learning_rate": 4.446424974823767e-05,
838
+ "loss": 0.0725,
839
+ "step": 56168
840
+ },
841
+ {
842
+ "epoch": 56.0,
843
+ "eval_loss": 0.35223615169525146,
844
+ "eval_runtime": 356.2173,
845
+ "eval_samples_per_second": 12.967,
846
+ "eval_steps_per_second": 0.272,
847
+ "eval_wer": 0.255661647692496,
848
+ "step": 56168
849
+ },
850
+ {
851
+ "epoch": 57.0,
852
+ "learning_rate": 4.345417925478349e-05,
853
+ "loss": 0.0712,
854
+ "step": 57171
855
+ },
856
+ {
857
+ "epoch": 57.0,
858
+ "eval_loss": 0.35607481002807617,
859
+ "eval_runtime": 336.6291,
860
+ "eval_samples_per_second": 13.721,
861
+ "eval_steps_per_second": 0.288,
862
+ "eval_wer": 0.2550495776716856,
863
+ "step": 57171
864
+ },
865
+ {
866
+ "epoch": 58.0,
867
+ "learning_rate": 4.244410876132931e-05,
868
+ "loss": 0.0703,
869
+ "step": 58174
870
+ },
871
+ {
872
+ "epoch": 58.0,
873
+ "eval_loss": 0.34993934631347656,
874
+ "eval_runtime": 338.604,
875
+ "eval_samples_per_second": 13.641,
876
+ "eval_steps_per_second": 0.286,
877
+ "eval_wer": 0.2562737177133064,
878
+ "step": 58174
879
+ },
880
+ {
881
+ "epoch": 59.0,
882
+ "learning_rate": 4.143403826787513e-05,
883
+ "loss": 0.0689,
884
+ "step": 59177
885
+ },
886
+ {
887
+ "epoch": 59.0,
888
+ "eval_loss": 0.35656341910362244,
889
+ "eval_runtime": 282.9543,
890
+ "eval_samples_per_second": 16.324,
891
+ "eval_steps_per_second": 0.343,
892
+ "eval_wer": 0.25489656016648304,
893
+ "step": 59177
894
+ },
895
+ {
896
+ "epoch": 60.0,
897
+ "learning_rate": 4.0424974823766364e-05,
898
+ "loss": 0.0667,
899
+ "step": 60180
900
+ },
901
+ {
902
+ "epoch": 60.0,
903
+ "eval_loss": 0.35375431180000305,
904
+ "eval_runtime": 291.8817,
905
+ "eval_samples_per_second": 15.825,
906
+ "eval_steps_per_second": 0.332,
907
+ "eval_wer": 0.2550495776716856,
908
+ "step": 60180
909
+ },
910
+ {
911
+ "epoch": 61.0,
912
+ "learning_rate": 3.9414904330312185e-05,
913
+ "loss": 0.0665,
914
+ "step": 61183
915
+ },
916
+ {
917
+ "epoch": 61.0,
918
+ "eval_loss": 0.3356081545352936,
919
+ "eval_runtime": 356.5563,
920
+ "eval_samples_per_second": 12.954,
921
+ "eval_steps_per_second": 0.272,
922
+ "eval_wer": 0.2545293181539968,
923
+ "step": 61183
924
+ },
925
+ {
926
+ "epoch": 62.0,
927
+ "learning_rate": 3.8405840886203424e-05,
928
+ "loss": 0.0662,
929
+ "step": 62186
930
+ },
931
+ {
932
+ "epoch": 62.0,
933
+ "eval_loss": 0.347979873418808,
934
+ "eval_runtime": 337.8069,
935
+ "eval_samples_per_second": 13.673,
936
+ "eval_steps_per_second": 0.287,
937
+ "eval_wer": 0.2541314726404701,
938
+ "step": 62186
939
+ },
940
+ {
941
+ "epoch": 63.0,
942
+ "learning_rate": 3.7395770392749245e-05,
943
+ "loss": 0.0653,
944
+ "step": 63189
945
+ },
946
+ {
947
+ "epoch": 63.0,
948
+ "eval_loss": 0.34537407755851746,
949
+ "eval_runtime": 361.1194,
950
+ "eval_samples_per_second": 12.791,
951
+ "eval_steps_per_second": 0.269,
952
+ "eval_wer": 0.2513771575468234,
953
+ "step": 63189
954
+ },
955
+ {
956
+ "epoch": 64.0,
957
+ "learning_rate": 3.6386706948640483e-05,
958
+ "loss": 0.0633,
959
+ "step": 64192
960
+ },
961
+ {
962
+ "epoch": 64.0,
963
+ "eval_loss": 0.36930787563323975,
964
+ "eval_runtime": 342.2873,
965
+ "eval_samples_per_second": 13.495,
966
+ "eval_steps_per_second": 0.283,
967
+ "eval_wer": 0.25220345207491734,
968
+ "step": 64192
969
+ },
970
+ {
971
+ "epoch": 65.0,
972
+ "learning_rate": 3.5376636455186305e-05,
973
+ "loss": 0.0632,
974
+ "step": 65195
975
+ },
976
+ {
977
+ "epoch": 65.0,
978
+ "eval_loss": 0.36128777265548706,
979
+ "eval_runtime": 361.0289,
980
+ "eval_samples_per_second": 12.794,
981
+ "eval_steps_per_second": 0.269,
982
+ "eval_wer": 0.254009058636308,
983
+ "step": 65195
984
+ },
985
+ {
986
+ "epoch": 66.0,
987
+ "learning_rate": 3.4366565961732126e-05,
988
+ "loss": 0.0615,
989
+ "step": 66198
990
+ },
991
+ {
992
+ "epoch": 66.0,
993
+ "eval_loss": 0.3568029999732971,
994
+ "eval_runtime": 343.0543,
995
+ "eval_samples_per_second": 13.464,
996
+ "eval_steps_per_second": 0.283,
997
+ "eval_wer": 0.25211164157179583,
998
+ "step": 66198
999
+ },
1000
+ {
1001
+ "epoch": 67.0,
1002
+ "learning_rate": 3.3357502517623365e-05,
1003
+ "loss": 0.0618,
1004
+ "step": 67201
1005
+ },
1006
+ {
1007
+ "epoch": 67.0,
1008
+ "eval_loss": 0.3640798032283783,
1009
+ "eval_runtime": 350.9436,
1010
+ "eval_samples_per_second": 13.162,
1011
+ "eval_steps_per_second": 0.276,
1012
+ "eval_wer": 0.2523258660790794,
1013
+ "step": 67201
1014
+ },
1015
+ {
1016
+ "epoch": 68.0,
1017
+ "learning_rate": 3.2347432024169186e-05,
1018
+ "loss": 0.0609,
1019
+ "step": 68204
1020
+ },
1021
+ {
1022
+ "epoch": 68.0,
1023
+ "eval_loss": 0.36202770471572876,
1024
+ "eval_runtime": 325.6728,
1025
+ "eval_samples_per_second": 14.183,
1026
+ "eval_steps_per_second": 0.298,
1027
+ "eval_wer": 0.25293793609988985,
1028
+ "step": 68204
1029
+ },
1030
+ {
1031
+ "epoch": 69.0,
1032
+ "learning_rate": 3.133736153071501e-05,
1033
+ "loss": 0.0594,
1034
+ "step": 69207
1035
+ },
1036
+ {
1037
+ "epoch": 69.0,
1038
+ "eval_loss": 0.3554952144622803,
1039
+ "eval_runtime": 322.2715,
1040
+ "eval_samples_per_second": 14.333,
1041
+ "eval_steps_per_second": 0.301,
1042
+ "eval_wer": 0.25205043456971477,
1043
+ "step": 69207
1044
+ },
1045
+ {
1046
+ "epoch": 70.0,
1047
+ "learning_rate": 3.0328298086606243e-05,
1048
+ "loss": 0.0594,
1049
+ "step": 70210
1050
+ },
1051
+ {
1052
+ "epoch": 70.0,
1053
+ "eval_loss": 0.35941246151924133,
1054
+ "eval_runtime": 323.1251,
1055
+ "eval_samples_per_second": 14.295,
1056
+ "eval_steps_per_second": 0.3,
1057
+ "eval_wer": 0.25208103807075527,
1058
+ "step": 70210
1059
+ },
1060
+ {
1061
+ "epoch": 71.0,
1062
+ "learning_rate": 2.9318227593152064e-05,
1063
+ "loss": 0.0598,
1064
+ "step": 71213
1065
+ },
1066
+ {
1067
+ "epoch": 71.0,
1068
+ "eval_loss": 0.35894548892974854,
1069
+ "eval_runtime": 348.1512,
1070
+ "eval_samples_per_second": 13.267,
1071
+ "eval_steps_per_second": 0.279,
1072
+ "eval_wer": 0.25067327702289144,
1073
+ "step": 71213
1074
+ },
1075
+ {
1076
+ "epoch": 72.0,
1077
+ "learning_rate": 2.8308157099697885e-05,
1078
+ "loss": 0.0579,
1079
+ "step": 72216
1080
+ },
1081
+ {
1082
+ "epoch": 72.0,
1083
+ "eval_loss": 0.34253111481666565,
1084
+ "eval_runtime": 354.641,
1085
+ "eval_samples_per_second": 13.024,
1086
+ "eval_steps_per_second": 0.274,
1087
+ "eval_wer": 0.25119353654058024,
1088
+ "step": 72216
1089
+ },
1090
+ {
1091
+ "epoch": 73.0,
1092
+ "learning_rate": 2.7299093655589124e-05,
1093
+ "loss": 0.0582,
1094
+ "step": 73219
1095
+ },
1096
+ {
1097
+ "epoch": 73.0,
1098
+ "eval_loss": 0.35133859515190125,
1099
+ "eval_runtime": 330.0079,
1100
+ "eval_samples_per_second": 13.997,
1101
+ "eval_steps_per_second": 0.294,
1102
+ "eval_wer": 0.24917370547190598,
1103
+ "step": 73219
1104
+ },
1105
+ {
1106
+ "epoch": 74.0,
1107
+ "learning_rate": 2.6289023162134945e-05,
1108
+ "loss": 0.0575,
1109
+ "step": 74222
1110
+ },
1111
+ {
1112
+ "epoch": 74.0,
1113
+ "eval_loss": 0.36137738823890686,
1114
+ "eval_runtime": 330.3739,
1115
+ "eval_samples_per_second": 13.981,
1116
+ "eval_steps_per_second": 0.294,
1117
+ "eval_wer": 0.25156077855306647,
1118
+ "step": 74222
1119
+ },
1120
+ {
1121
+ "epoch": 75.0,
1122
+ "learning_rate": 2.5278952668680767e-05,
1123
+ "loss": 0.0569,
1124
+ "step": 75225
1125
+ },
1126
+ {
1127
+ "epoch": 75.0,
1128
+ "eval_loss": 0.36236804723739624,
1129
+ "eval_runtime": 316.6665,
1130
+ "eval_samples_per_second": 14.586,
1131
+ "eval_steps_per_second": 0.306,
1132
+ "eval_wer": 0.24947974048231117,
1133
+ "step": 75225
1134
+ },
1135
+ {
1136
+ "epoch": 76.0,
1137
+ "learning_rate": 2.4269889224572005e-05,
1138
+ "loss": 0.0558,
1139
+ "step": 76228
1140
+ },
1141
+ {
1142
+ "epoch": 76.0,
1143
+ "eval_loss": 0.3634468615055084,
1144
+ "eval_runtime": 319.3864,
1145
+ "eval_samples_per_second": 14.462,
1146
+ "eval_steps_per_second": 0.304,
1147
+ "eval_wer": 0.24908189496878444,
1148
+ "step": 76228
1149
+ },
1150
+ {
1151
+ "epoch": 77.0,
1152
+ "learning_rate": 2.3259818731117823e-05,
1153
+ "loss": 0.0556,
1154
+ "step": 77231
1155
+ },
1156
+ {
1157
+ "epoch": 77.0,
1158
+ "eval_loss": 0.3610389530658722,
1159
+ "eval_runtime": 350.196,
1160
+ "eval_samples_per_second": 13.19,
1161
+ "eval_steps_per_second": 0.277,
1162
+ "eval_wer": 0.248408617945893,
1163
+ "step": 77231
1164
+ },
1165
+ {
1166
+ "epoch": 78.0,
1167
+ "learning_rate": 2.2249748237663648e-05,
1168
+ "loss": 0.0548,
1169
+ "step": 78234
1170
+ },
1171
+ {
1172
+ "epoch": 78.0,
1173
+ "eval_loss": 0.3590007722377777,
1174
+ "eval_runtime": 341.784,
1175
+ "eval_samples_per_second": 13.514,
1176
+ "eval_steps_per_second": 0.284,
1177
+ "eval_wer": 0.24657240788346188,
1178
+ "step": 78234
1179
+ },
1180
+ {
1181
+ "epoch": 79.0,
1182
+ "learning_rate": 2.123967774420947e-05,
1183
+ "loss": 0.0528,
1184
+ "step": 79237
1185
+ },
1186
+ {
1187
+ "epoch": 79.0,
1188
+ "eval_loss": 0.3674350678920746,
1189
+ "eval_runtime": 327.8795,
1190
+ "eval_samples_per_second": 14.087,
1191
+ "eval_steps_per_second": 0.296,
1192
+ "eval_wer": 0.248408617945893,
1193
+ "step": 79237
1194
+ },
1195
+ {
1196
+ "epoch": 80.0,
1197
+ "learning_rate": 2.0230614300100704e-05,
1198
+ "loss": 0.0521,
1199
+ "step": 80240
1200
+ },
1201
+ {
1202
+ "epoch": 80.0,
1203
+ "eval_loss": 0.3592623770236969,
1204
+ "eval_runtime": 329.3838,
1205
+ "eval_samples_per_second": 14.023,
1206
+ "eval_steps_per_second": 0.294,
1207
+ "eval_wer": 0.2469702533969886,
1208
+ "step": 80240
1209
+ },
1210
+ {
1211
+ "epoch": 81.0,
1212
+ "learning_rate": 1.9221550855991943e-05,
1213
+ "loss": 0.0532,
1214
+ "step": 81243
1215
+ },
1216
+ {
1217
+ "epoch": 81.0,
1218
+ "eval_loss": 0.370604932308197,
1219
+ "eval_runtime": 303.5979,
1220
+ "eval_samples_per_second": 15.214,
1221
+ "eval_steps_per_second": 0.32,
1222
+ "eval_wer": 0.24947974048231117,
1223
+ "step": 81243
1224
+ },
1225
+ {
1226
+ "epoch": 82.0,
1227
+ "learning_rate": 1.8211480362537764e-05,
1228
+ "loss": 0.0517,
1229
+ "step": 82246
1230
+ },
1231
+ {
1232
+ "epoch": 82.0,
1233
+ "eval_loss": 0.37441593408584595,
1234
+ "eval_runtime": 316.1258,
1235
+ "eval_samples_per_second": 14.611,
1236
+ "eval_steps_per_second": 0.307,
1237
+ "eval_wer": 0.24975517199167585,
1238
+ "step": 82246
1239
+ },
1240
+ {
1241
+ "epoch": 83.0,
1242
+ "learning_rate": 1.7201409869083586e-05,
1243
+ "loss": 0.0516,
1244
+ "step": 83249
1245
+ },
1246
+ {
1247
+ "epoch": 83.0,
1248
+ "eval_loss": 0.3725976347923279,
1249
+ "eval_runtime": 353.3018,
1250
+ "eval_samples_per_second": 13.074,
1251
+ "eval_steps_per_second": 0.275,
1252
+ "eval_wer": 0.2485922389521361,
1253
+ "step": 83249
1254
+ },
1255
+ {
1256
+ "epoch": 84.0,
1257
+ "learning_rate": 1.6192346424974824e-05,
1258
+ "loss": 0.0496,
1259
+ "step": 84252
1260
+ },
1261
+ {
1262
+ "epoch": 84.0,
1263
+ "eval_loss": 0.36621466279029846,
1264
+ "eval_runtime": 358.7351,
1265
+ "eval_samples_per_second": 12.876,
1266
+ "eval_steps_per_second": 0.27,
1267
+ "eval_wer": 0.24730689190843433,
1268
+ "step": 84252
1269
+ },
1270
+ {
1271
+ "epoch": 85.0,
1272
+ "learning_rate": 1.5182275931520645e-05,
1273
+ "loss": 0.0503,
1274
+ "step": 85255
1275
+ },
1276
+ {
1277
+ "epoch": 85.0,
1278
+ "eval_loss": 0.3839229941368103,
1279
+ "eval_runtime": 322.7657,
1280
+ "eval_samples_per_second": 14.311,
1281
+ "eval_steps_per_second": 0.301,
1282
+ "eval_wer": 0.24819439343860938,
1283
+ "step": 85255
1284
+ },
1285
+ {
1286
+ "epoch": 86.0,
1287
+ "learning_rate": 1.4173212487411884e-05,
1288
+ "loss": 0.0505,
1289
+ "step": 86258
1290
+ },
1291
+ {
1292
+ "epoch": 86.0,
1293
+ "eval_loss": 0.37594547867774963,
1294
+ "eval_runtime": 344.4521,
1295
+ "eval_samples_per_second": 13.41,
1296
+ "eval_steps_per_second": 0.282,
1297
+ "eval_wer": 0.24843922144693353,
1298
+ "step": 86258
1299
+ },
1300
+ {
1301
+ "epoch": 87.0,
1302
+ "learning_rate": 1.3163141993957703e-05,
1303
+ "loss": 0.0489,
1304
+ "step": 87261
1305
+ },
1306
+ {
1307
+ "epoch": 87.0,
1308
+ "eval_loss": 0.3759251832962036,
1309
+ "eval_runtime": 348.9851,
1310
+ "eval_samples_per_second": 13.236,
1311
+ "eval_steps_per_second": 0.278,
1312
+ "eval_wer": 0.24718447790427225,
1313
+ "step": 87261
1314
+ },
1315
+ {
1316
+ "epoch": 88.0,
1317
+ "learning_rate": 1.2153071500503525e-05,
1318
+ "loss": 0.0498,
1319
+ "step": 88264
1320
+ },
1321
+ {
1322
+ "epoch": 88.0,
1323
+ "eval_loss": 0.3712981045246124,
1324
+ "eval_runtime": 358.3358,
1325
+ "eval_samples_per_second": 12.89,
1326
+ "eval_steps_per_second": 0.271,
1327
+ "eval_wer": 0.2474599094136369,
1328
+ "step": 88264
1329
+ },
1330
+ {
1331
+ "epoch": 89.0,
1332
+ "learning_rate": 1.1143001007049346e-05,
1333
+ "loss": 0.0481,
1334
+ "step": 89267
1335
+ },
1336
+ {
1337
+ "epoch": 89.0,
1338
+ "eval_loss": 0.37631461024284363,
1339
+ "eval_runtime": 376.3412,
1340
+ "eval_samples_per_second": 12.273,
1341
+ "eval_steps_per_second": 0.258,
1342
+ "eval_wer": 0.2474293059125964,
1343
+ "step": 89267
1344
+ },
1345
+ {
1346
+ "epoch": 90.0,
1347
+ "learning_rate": 1.0134944612286003e-05,
1348
+ "loss": 0.0483,
1349
+ "step": 90270
1350
+ },
1351
+ {
1352
+ "epoch": 90.0,
1353
+ "eval_loss": 0.380001425743103,
1354
+ "eval_runtime": 340.9184,
1355
+ "eval_samples_per_second": 13.549,
1356
+ "eval_steps_per_second": 0.285,
1357
+ "eval_wer": 0.24651120088138084,
1358
+ "step": 90270
1359
+ },
1360
+ {
1361
+ "epoch": 91.0,
1362
+ "learning_rate": 9.124874118831823e-06,
1363
+ "loss": 0.0471,
1364
+