infinitejoy commited on
Commit
29b40dc
1 Parent(s): 27d68b0

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_loss": 0.16141469776630402,
4
+ "eval_runtime": 414.7266,
5
+ "eval_samples": 9184,
6
+ "eval_samples_per_second": 22.145,
7
+ "eval_steps_per_second": 1.384,
8
+ "eval_wer": 0.2906979143028077,
9
+ "train_loss": 1.0648517145022656,
10
+ "train_runtime": 95151.9589,
11
+ "train_samples": 30002,
12
+ "train_samples_per_second": 15.765,
13
+ "train_steps_per_second": 0.493
14
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_loss": 0.16141469776630402,
4
+ "eval_runtime": 414.7266,
5
+ "eval_samples": 9184,
6
+ "eval_samples_per_second": 22.145,
7
+ "eval_steps_per_second": 1.384,
8
+ "eval_wer": 0.2906979143028077
9
+ }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1224588e5573446179d17c967c4a0dca9945f01750ae759c783317806d47dc3c
3
  size 1262112241
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07818b226b1b6fecb327f658c084ce757214f2f0892a7af368eae1c7b5e6c1fd
3
  size 1262112241
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "train_loss": 1.0648517145022656,
4
+ "train_runtime": 95151.9589,
5
+ "train_samples": 30002,
6
+ "train_samples_per_second": 15.765,
7
+ "train_steps_per_second": 0.493
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,400 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 50.0,
5
+ "global_step": 46900,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 1.07,
12
+ "learning_rate": 1.74475e-05,
13
+ "loss": 7.5225,
14
+ "step": 1000
15
+ },
16
+ {
17
+ "epoch": 2.13,
18
+ "learning_rate": 3.49475e-05,
19
+ "loss": 2.9185,
20
+ "step": 2000
21
+ },
22
+ {
23
+ "epoch": 3.2,
24
+ "learning_rate": 5.2447499999999995e-05,
25
+ "loss": 1.5385,
26
+ "step": 3000
27
+ },
28
+ {
29
+ "epoch": 4.26,
30
+ "learning_rate": 6.99475e-05,
31
+ "loss": 1.2881,
32
+ "step": 4000
33
+ },
34
+ {
35
+ "epoch": 4.26,
36
+ "eval_loss": 0.37639039754867554,
37
+ "eval_runtime": 433.5071,
38
+ "eval_samples_per_second": 21.185,
39
+ "eval_steps_per_second": 1.324,
40
+ "eval_wer": 0.6460503785612609,
41
+ "step": 4000
42
+ },
43
+ {
44
+ "epoch": 5.33,
45
+ "learning_rate": 6.837482517482516e-05,
46
+ "loss": 1.2081,
47
+ "step": 5000
48
+ },
49
+ {
50
+ "epoch": 6.4,
51
+ "learning_rate": 6.674475524475524e-05,
52
+ "loss": 1.1501,
53
+ "step": 6000
54
+ },
55
+ {
56
+ "epoch": 7.46,
57
+ "learning_rate": 6.511305361305361e-05,
58
+ "loss": 1.1115,
59
+ "step": 7000
60
+ },
61
+ {
62
+ "epoch": 8.53,
63
+ "learning_rate": 6.348135198135198e-05,
64
+ "loss": 1.0767,
65
+ "step": 8000
66
+ },
67
+ {
68
+ "epoch": 8.53,
69
+ "eval_loss": 0.26572874188423157,
70
+ "eval_runtime": 423.456,
71
+ "eval_samples_per_second": 21.688,
72
+ "eval_steps_per_second": 1.356,
73
+ "eval_wer": 0.5164111979865079,
74
+ "step": 8000
75
+ },
76
+ {
77
+ "epoch": 9.59,
78
+ "learning_rate": 6.185128205128204e-05,
79
+ "loss": 1.0506,
80
+ "step": 9000
81
+ },
82
+ {
83
+ "epoch": 10.66,
84
+ "learning_rate": 6.021958041958042e-05,
85
+ "loss": 1.0261,
86
+ "step": 10000
87
+ },
88
+ {
89
+ "epoch": 11.73,
90
+ "learning_rate": 5.858787878787878e-05,
91
+ "loss": 1.0029,
92
+ "step": 11000
93
+ },
94
+ {
95
+ "epoch": 12.79,
96
+ "learning_rate": 5.695617715617715e-05,
97
+ "loss": 0.9841,
98
+ "step": 12000
99
+ },
100
+ {
101
+ "epoch": 12.79,
102
+ "eval_loss": 0.23304273188114166,
103
+ "eval_runtime": 424.6539,
104
+ "eval_samples_per_second": 21.627,
105
+ "eval_steps_per_second": 1.352,
106
+ "eval_wer": 0.4444742433932292,
107
+ "step": 12000
108
+ },
109
+ {
110
+ "epoch": 13.86,
111
+ "learning_rate": 5.532447552447552e-05,
112
+ "loss": 0.9694,
113
+ "step": 13000
114
+ },
115
+ {
116
+ "epoch": 14.93,
117
+ "learning_rate": 5.369603729603729e-05,
118
+ "loss": 0.9593,
119
+ "step": 14000
120
+ },
121
+ {
122
+ "epoch": 15.99,
123
+ "learning_rate": 5.206433566433566e-05,
124
+ "loss": 0.9428,
125
+ "step": 15000
126
+ },
127
+ {
128
+ "epoch": 17.06,
129
+ "learning_rate": 5.0432634032634027e-05,
130
+ "loss": 0.9274,
131
+ "step": 16000
132
+ },
133
+ {
134
+ "epoch": 17.06,
135
+ "eval_loss": 0.213353231549263,
136
+ "eval_runtime": 424.1743,
137
+ "eval_samples_per_second": 21.651,
138
+ "eval_steps_per_second": 1.353,
139
+ "eval_wer": 0.39289913972727086,
140
+ "step": 16000
141
+ },
142
+ {
143
+ "epoch": 18.12,
144
+ "learning_rate": 4.8800932400932396e-05,
145
+ "loss": 0.9171,
146
+ "step": 17000
147
+ },
148
+ {
149
+ "epoch": 19.19,
150
+ "learning_rate": 4.7170862470862466e-05,
151
+ "loss": 0.9057,
152
+ "step": 18000
153
+ },
154
+ {
155
+ "epoch": 20.26,
156
+ "learning_rate": 4.5539160839160836e-05,
157
+ "loss": 0.8881,
158
+ "step": 19000
159
+ },
160
+ {
161
+ "epoch": 21.32,
162
+ "learning_rate": 4.3909090909090906e-05,
163
+ "loss": 0.8781,
164
+ "step": 20000
165
+ },
166
+ {
167
+ "epoch": 21.32,
168
+ "eval_loss": 0.19452312588691711,
169
+ "eval_runtime": 418.381,
170
+ "eval_samples_per_second": 21.951,
171
+ "eval_steps_per_second": 1.372,
172
+ "eval_wer": 0.3886287211437295,
173
+ "step": 20000
174
+ },
175
+ {
176
+ "epoch": 22.39,
177
+ "learning_rate": 4.2277389277389276e-05,
178
+ "loss": 0.8633,
179
+ "step": 21000
180
+ },
181
+ {
182
+ "epoch": 23.45,
183
+ "learning_rate": 4.064731934731934e-05,
184
+ "loss": 0.8537,
185
+ "step": 22000
186
+ },
187
+ {
188
+ "epoch": 24.52,
189
+ "learning_rate": 3.901561771561771e-05,
190
+ "loss": 0.8462,
191
+ "step": 23000
192
+ },
193
+ {
194
+ "epoch": 25.59,
195
+ "learning_rate": 3.738391608391608e-05,
196
+ "loss": 0.8381,
197
+ "step": 24000
198
+ },
199
+ {
200
+ "epoch": 25.59,
201
+ "eval_loss": 0.184038907289505,
202
+ "eval_runtime": 425.8794,
203
+ "eval_samples_per_second": 21.565,
204
+ "eval_steps_per_second": 1.348,
205
+ "eval_wer": 0.3736719410806016,
206
+ "step": 24000
207
+ },
208
+ {
209
+ "epoch": 26.65,
210
+ "learning_rate": 3.575221445221445e-05,
211
+ "loss": 0.8237,
212
+ "step": 25000
213
+ },
214
+ {
215
+ "epoch": 27.72,
216
+ "learning_rate": 3.412214452214452e-05,
217
+ "loss": 0.819,
218
+ "step": 26000
219
+ },
220
+ {
221
+ "epoch": 28.78,
222
+ "learning_rate": 3.249207459207459e-05,
223
+ "loss": 0.8114,
224
+ "step": 27000
225
+ },
226
+ {
227
+ "epoch": 29.85,
228
+ "learning_rate": 3.086037296037296e-05,
229
+ "loss": 0.8054,
230
+ "step": 28000
231
+ },
232
+ {
233
+ "epoch": 29.85,
234
+ "eval_loss": 0.17562612891197205,
235
+ "eval_runtime": 420.9127,
236
+ "eval_samples_per_second": 21.819,
237
+ "eval_steps_per_second": 1.364,
238
+ "eval_wer": 0.3523198481628948,
239
+ "step": 28000
240
+ },
241
+ {
242
+ "epoch": 30.92,
243
+ "learning_rate": 2.9228671328671326e-05,
244
+ "loss": 0.7997,
245
+ "step": 29000
246
+ },
247
+ {
248
+ "epoch": 31.98,
249
+ "learning_rate": 2.7598601398601396e-05,
250
+ "loss": 0.7867,
251
+ "step": 30000
252
+ },
253
+ {
254
+ "epoch": 33.05,
255
+ "learning_rate": 2.5966899766899766e-05,
256
+ "loss": 0.7764,
257
+ "step": 31000
258
+ },
259
+ {
260
+ "epoch": 34.12,
261
+ "learning_rate": 2.4336829836829836e-05,
262
+ "loss": 0.7763,
263
+ "step": 32000
264
+ },
265
+ {
266
+ "epoch": 34.12,
267
+ "eval_loss": 0.17449165880680084,
268
+ "eval_runtime": 413.6966,
269
+ "eval_samples_per_second": 22.2,
270
+ "eval_steps_per_second": 1.387,
271
+ "eval_wer": 0.3299156231304025,
272
+ "step": 32000
273
+ },
274
+ {
275
+ "epoch": 35.18,
276
+ "learning_rate": 2.2705128205128205e-05,
277
+ "loss": 0.7676,
278
+ "step": 33000
279
+ },
280
+ {
281
+ "epoch": 36.25,
282
+ "learning_rate": 2.1075058275058275e-05,
283
+ "loss": 0.7571,
284
+ "step": 34000
285
+ },
286
+ {
287
+ "epoch": 37.31,
288
+ "learning_rate": 1.944335664335664e-05,
289
+ "loss": 0.7556,
290
+ "step": 35000
291
+ },
292
+ {
293
+ "epoch": 38.38,
294
+ "learning_rate": 1.7814918414918412e-05,
295
+ "loss": 0.7474,
296
+ "step": 36000
297
+ },
298
+ {
299
+ "epoch": 38.38,
300
+ "eval_loss": 0.1677420288324356,
301
+ "eval_runtime": 414.2329,
302
+ "eval_samples_per_second": 22.171,
303
+ "eval_steps_per_second": 1.386,
304
+ "eval_wer": 0.30740824789057825,
305
+ "step": 36000
306
+ },
307
+ {
308
+ "epoch": 39.45,
309
+ "learning_rate": 1.6183216783216782e-05,
310
+ "loss": 0.7411,
311
+ "step": 37000
312
+ },
313
+ {
314
+ "epoch": 40.51,
315
+ "learning_rate": 1.455151515151515e-05,
316
+ "loss": 0.7369,
317
+ "step": 38000
318
+ },
319
+ {
320
+ "epoch": 41.58,
321
+ "learning_rate": 1.2919813519813518e-05,
322
+ "loss": 0.7309,
323
+ "step": 39000
324
+ },
325
+ {
326
+ "epoch": 42.64,
327
+ "learning_rate": 1.1288111888111887e-05,
328
+ "loss": 0.7298,
329
+ "step": 40000
330
+ },
331
+ {
332
+ "epoch": 42.64,
333
+ "eval_loss": 0.1649060845375061,
334
+ "eval_runtime": 413.307,
335
+ "eval_samples_per_second": 22.221,
336
+ "eval_steps_per_second": 1.389,
337
+ "eval_wer": 0.2962886555401976,
338
+ "step": 40000
339
+ },
340
+ {
341
+ "epoch": 43.71,
342
+ "learning_rate": 9.65967365967366e-06,
343
+ "loss": 0.7201,
344
+ "step": 41000
345
+ },
346
+ {
347
+ "epoch": 44.78,
348
+ "learning_rate": 8.027972027972028e-06,
349
+ "loss": 0.72,
350
+ "step": 42000
351
+ },
352
+ {
353
+ "epoch": 45.84,
354
+ "learning_rate": 6.396270396270396e-06,
355
+ "loss": 0.7092,
356
+ "step": 43000
357
+ },
358
+ {
359
+ "epoch": 46.91,
360
+ "learning_rate": 4.764568764568764e-06,
361
+ "loss": 0.7125,
362
+ "step": 44000
363
+ },
364
+ {
365
+ "epoch": 46.91,
366
+ "eval_loss": 0.161702960729599,
367
+ "eval_runtime": 419.2465,
368
+ "eval_samples_per_second": 21.906,
369
+ "eval_steps_per_second": 1.369,
370
+ "eval_wer": 0.29307036907144185,
371
+ "step": 44000
372
+ },
373
+ {
374
+ "epoch": 47.97,
375
+ "learning_rate": 3.1344988344988342e-06,
376
+ "loss": 0.7093,
377
+ "step": 45000
378
+ },
379
+ {
380
+ "epoch": 49.04,
381
+ "learning_rate": 1.5044289044289044e-06,
382
+ "loss": 0.7049,
383
+ "step": 46000
384
+ },
385
+ {
386
+ "epoch": 50.0,
387
+ "step": 46900,
388
+ "total_flos": 2.4630771835403823e+20,
389
+ "train_loss": 1.0648517145022656,
390
+ "train_runtime": 95151.9589,
391
+ "train_samples_per_second": 15.765,
392
+ "train_steps_per_second": 0.493
393
+ }
394
+ ],
395
+ "max_steps": 46900,
396
+ "num_train_epochs": 50,
397
+ "total_flos": 2.4630771835403823e+20,
398
+ "trial_name": null,
399
+ "trial_params": null
400
+ }