arun100 commited on
Commit
b71468e
·
verified ·
1 Parent(s): bb3427a

End of training

Browse files
README.md CHANGED
@@ -2,20 +2,21 @@
2
  license: apache-2.0
3
  base_model: qanastek/whisper-base-french-cased
4
  tags:
 
5
  - generated_from_trainer
6
  datasets:
7
- - fleurs
8
  metrics:
9
  - wer
10
  model-index:
11
- - name: qanastek/whisper-base-french-cased
12
  results:
13
  - task:
14
  name: Automatic Speech Recognition
15
  type: automatic-speech-recognition
16
  dataset:
17
- name: fleurs
18
- type: fleurs
19
  config: fr_fr
20
  split: test
21
  args: fr_fr
@@ -28,11 +29,11 @@ model-index:
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
29
  should probably proofread and complete it, then remove this comment. -->
30
 
31
- # qanastek/whisper-base-french-cased
32
 
33
- This model is a fine-tuned version of [qanastek/whisper-base-french-cased](https://huggingface.co/qanastek/whisper-base-french-cased) on the fleurs dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.5395
36
  - Wer: 23.7955
37
 
38
  ## Model description
 
2
  license: apache-2.0
3
  base_model: qanastek/whisper-base-french-cased
4
  tags:
5
+ - whisper-event
6
  - generated_from_trainer
7
  datasets:
8
+ - google/fleurs
9
  metrics:
10
  - wer
11
  model-index:
12
+ - name: Whisper Base French
13
  results:
14
  - task:
15
  name: Automatic Speech Recognition
16
  type: automatic-speech-recognition
17
  dataset:
18
+ name: google/fleurs fr_fr
19
+ type: google/fleurs
20
  config: fr_fr
21
  split: test
22
  args: fr_fr
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
30
  should probably proofread and complete it, then remove this comment. -->
31
 
32
+ # Whisper Base French
33
 
34
+ This model is a fine-tuned version of [qanastek/whisper-base-french-cased](https://huggingface.co/qanastek/whisper-base-french-cased) on the google/fleurs fr_fr dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.5402
37
  - Wer: 23.7955
38
 
39
  ## Model description
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 526.32,
3
+ "eval_loss": 0.5402421951293945,
4
+ "eval_runtime": 143.5733,
5
+ "eval_samples_per_second": 4.708,
6
+ "eval_steps_per_second": 0.153,
7
+ "eval_wer": 23.795498749652683,
8
+ "train_loss": 0.31816294193267824,
9
+ "train_runtime": 12909.7746,
10
+ "train_samples_per_second": 24.787,
11
+ "train_steps_per_second": 0.387
12
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 526.32,
3
+ "eval_loss": 0.5402421951293945,
4
+ "eval_runtime": 143.5733,
5
+ "eval_samples_per_second": 4.708,
6
+ "eval_steps_per_second": 0.153,
7
+ "eval_wer": 23.795498749652683
8
+ }
runs/Jan23_15-53-35_ip-172-31-23-125/events.out.tfevents.1706038421.ip-172-31-23-125.44123.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8b09084c9719e1ea7d93e9b6920be73482ec9821dd7e89efd9ee4ca86e4ac0f
3
+ size 406
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 526.32,
3
+ "train_loss": 0.31816294193267824,
4
+ "train_runtime": 12909.7746,
5
+ "train_samples_per_second": 24.787,
6
+ "train_steps_per_second": 0.387
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,1275 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 23.795498749652683,
3
+ "best_model_checkpoint": "./checkpoint-4000",
4
+ "epoch": 526.3157894736842,
5
+ "eval_steps": 1000,
6
+ "global_step": 5000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 2.63,
13
+ "learning_rate": 4.4e-09,
14
+ "loss": 0.8459,
15
+ "step": 25
16
+ },
17
+ {
18
+ "epoch": 5.26,
19
+ "learning_rate": 9.4e-09,
20
+ "loss": 0.8681,
21
+ "step": 50
22
+ },
23
+ {
24
+ "epoch": 7.89,
25
+ "learning_rate": 1.4399999999999998e-08,
26
+ "loss": 0.8349,
27
+ "step": 75
28
+ },
29
+ {
30
+ "epoch": 10.53,
31
+ "learning_rate": 1.94e-08,
32
+ "loss": 0.8471,
33
+ "step": 100
34
+ },
35
+ {
36
+ "epoch": 13.16,
37
+ "learning_rate": 2.4399999999999997e-08,
38
+ "loss": 0.8299,
39
+ "step": 125
40
+ },
41
+ {
42
+ "epoch": 15.79,
43
+ "learning_rate": 2.94e-08,
44
+ "loss": 0.8228,
45
+ "step": 150
46
+ },
47
+ {
48
+ "epoch": 18.42,
49
+ "learning_rate": 3.44e-08,
50
+ "loss": 0.8015,
51
+ "step": 175
52
+ },
53
+ {
54
+ "epoch": 21.05,
55
+ "learning_rate": 3.94e-08,
56
+ "loss": 0.8041,
57
+ "step": 200
58
+ },
59
+ {
60
+ "epoch": 23.68,
61
+ "learning_rate": 4.44e-08,
62
+ "loss": 0.797,
63
+ "step": 225
64
+ },
65
+ {
66
+ "epoch": 26.32,
67
+ "learning_rate": 4.94e-08,
68
+ "loss": 0.7771,
69
+ "step": 250
70
+ },
71
+ {
72
+ "epoch": 28.95,
73
+ "learning_rate": 5.4400000000000004e-08,
74
+ "loss": 0.7623,
75
+ "step": 275
76
+ },
77
+ {
78
+ "epoch": 31.58,
79
+ "learning_rate": 5.9399999999999996e-08,
80
+ "loss": 0.7524,
81
+ "step": 300
82
+ },
83
+ {
84
+ "epoch": 34.21,
85
+ "learning_rate": 6.44e-08,
86
+ "loss": 0.7384,
87
+ "step": 325
88
+ },
89
+ {
90
+ "epoch": 36.84,
91
+ "learning_rate": 6.939999999999999e-08,
92
+ "loss": 0.715,
93
+ "step": 350
94
+ },
95
+ {
96
+ "epoch": 39.47,
97
+ "learning_rate": 7.439999999999999e-08,
98
+ "loss": 0.6994,
99
+ "step": 375
100
+ },
101
+ {
102
+ "epoch": 42.11,
103
+ "learning_rate": 7.94e-08,
104
+ "loss": 0.6721,
105
+ "step": 400
106
+ },
107
+ {
108
+ "epoch": 44.74,
109
+ "learning_rate": 8.439999999999999e-08,
110
+ "loss": 0.676,
111
+ "step": 425
112
+ },
113
+ {
114
+ "epoch": 47.37,
115
+ "learning_rate": 8.939999999999999e-08,
116
+ "loss": 0.6614,
117
+ "step": 450
118
+ },
119
+ {
120
+ "epoch": 50.0,
121
+ "learning_rate": 9.439999999999999e-08,
122
+ "loss": 0.625,
123
+ "step": 475
124
+ },
125
+ {
126
+ "epoch": 52.63,
127
+ "learning_rate": 9.939999999999999e-08,
128
+ "loss": 0.5845,
129
+ "step": 500
130
+ },
131
+ {
132
+ "epoch": 55.26,
133
+ "learning_rate": 9.95111111111111e-08,
134
+ "loss": 0.5787,
135
+ "step": 525
136
+ },
137
+ {
138
+ "epoch": 57.89,
139
+ "learning_rate": 9.895555555555555e-08,
140
+ "loss": 0.5624,
141
+ "step": 550
142
+ },
143
+ {
144
+ "epoch": 60.53,
145
+ "learning_rate": 9.84e-08,
146
+ "loss": 0.5237,
147
+ "step": 575
148
+ },
149
+ {
150
+ "epoch": 63.16,
151
+ "learning_rate": 9.784444444444444e-08,
152
+ "loss": 0.5334,
153
+ "step": 600
154
+ },
155
+ {
156
+ "epoch": 65.79,
157
+ "learning_rate": 9.728888888888889e-08,
158
+ "loss": 0.5064,
159
+ "step": 625
160
+ },
161
+ {
162
+ "epoch": 68.42,
163
+ "learning_rate": 9.673333333333334e-08,
164
+ "loss": 0.5078,
165
+ "step": 650
166
+ },
167
+ {
168
+ "epoch": 71.05,
169
+ "learning_rate": 9.617777777777777e-08,
170
+ "loss": 0.485,
171
+ "step": 675
172
+ },
173
+ {
174
+ "epoch": 73.68,
175
+ "learning_rate": 9.562222222222221e-08,
176
+ "loss": 0.4678,
177
+ "step": 700
178
+ },
179
+ {
180
+ "epoch": 76.32,
181
+ "learning_rate": 9.506666666666666e-08,
182
+ "loss": 0.4636,
183
+ "step": 725
184
+ },
185
+ {
186
+ "epoch": 78.95,
187
+ "learning_rate": 9.45111111111111e-08,
188
+ "loss": 0.4707,
189
+ "step": 750
190
+ },
191
+ {
192
+ "epoch": 81.58,
193
+ "learning_rate": 9.395555555555555e-08,
194
+ "loss": 0.4725,
195
+ "step": 775
196
+ },
197
+ {
198
+ "epoch": 84.21,
199
+ "learning_rate": 9.34e-08,
200
+ "loss": 0.4595,
201
+ "step": 800
202
+ },
203
+ {
204
+ "epoch": 86.84,
205
+ "learning_rate": 9.284444444444444e-08,
206
+ "loss": 0.4449,
207
+ "step": 825
208
+ },
209
+ {
210
+ "epoch": 89.47,
211
+ "learning_rate": 9.228888888888889e-08,
212
+ "loss": 0.4254,
213
+ "step": 850
214
+ },
215
+ {
216
+ "epoch": 92.11,
217
+ "learning_rate": 9.173333333333333e-08,
218
+ "loss": 0.4222,
219
+ "step": 875
220
+ },
221
+ {
222
+ "epoch": 94.74,
223
+ "learning_rate": 9.117777777777777e-08,
224
+ "loss": 0.4247,
225
+ "step": 900
226
+ },
227
+ {
228
+ "epoch": 97.37,
229
+ "learning_rate": 9.062222222222222e-08,
230
+ "loss": 0.4306,
231
+ "step": 925
232
+ },
233
+ {
234
+ "epoch": 100.0,
235
+ "learning_rate": 9.006666666666665e-08,
236
+ "loss": 0.4017,
237
+ "step": 950
238
+ },
239
+ {
240
+ "epoch": 102.63,
241
+ "learning_rate": 8.95111111111111e-08,
242
+ "loss": 0.4176,
243
+ "step": 975
244
+ },
245
+ {
246
+ "epoch": 105.26,
247
+ "learning_rate": 8.895555555555555e-08,
248
+ "loss": 0.3835,
249
+ "step": 1000
250
+ },
251
+ {
252
+ "epoch": 105.26,
253
+ "eval_loss": 0.5891541838645935,
254
+ "eval_runtime": 160.0944,
255
+ "eval_samples_per_second": 4.223,
256
+ "eval_steps_per_second": 0.137,
257
+ "eval_wer": 25.423728813559322,
258
+ "step": 1000
259
+ },
260
+ {
261
+ "epoch": 107.89,
262
+ "learning_rate": 8.84e-08,
263
+ "loss": 0.4049,
264
+ "step": 1025
265
+ },
266
+ {
267
+ "epoch": 110.53,
268
+ "learning_rate": 8.784444444444445e-08,
269
+ "loss": 0.396,
270
+ "step": 1050
271
+ },
272
+ {
273
+ "epoch": 113.16,
274
+ "learning_rate": 8.72888888888889e-08,
275
+ "loss": 0.3819,
276
+ "step": 1075
277
+ },
278
+ {
279
+ "epoch": 115.79,
280
+ "learning_rate": 8.673333333333333e-08,
281
+ "loss": 0.3813,
282
+ "step": 1100
283
+ },
284
+ {
285
+ "epoch": 118.42,
286
+ "learning_rate": 8.617777777777778e-08,
287
+ "loss": 0.3786,
288
+ "step": 1125
289
+ },
290
+ {
291
+ "epoch": 121.05,
292
+ "learning_rate": 8.562222222222221e-08,
293
+ "loss": 0.3623,
294
+ "step": 1150
295
+ },
296
+ {
297
+ "epoch": 123.68,
298
+ "learning_rate": 8.506666666666666e-08,
299
+ "loss": 0.3702,
300
+ "step": 1175
301
+ },
302
+ {
303
+ "epoch": 126.32,
304
+ "learning_rate": 8.451111111111111e-08,
305
+ "loss": 0.3578,
306
+ "step": 1200
307
+ },
308
+ {
309
+ "epoch": 128.95,
310
+ "learning_rate": 8.395555555555556e-08,
311
+ "loss": 0.3559,
312
+ "step": 1225
313
+ },
314
+ {
315
+ "epoch": 131.58,
316
+ "learning_rate": 8.339999999999999e-08,
317
+ "loss": 0.3548,
318
+ "step": 1250
319
+ },
320
+ {
321
+ "epoch": 134.21,
322
+ "learning_rate": 8.284444444444444e-08,
323
+ "loss": 0.345,
324
+ "step": 1275
325
+ },
326
+ {
327
+ "epoch": 136.84,
328
+ "learning_rate": 8.228888888888889e-08,
329
+ "loss": 0.3574,
330
+ "step": 1300
331
+ },
332
+ {
333
+ "epoch": 139.47,
334
+ "learning_rate": 8.173333333333334e-08,
335
+ "loss": 0.3467,
336
+ "step": 1325
337
+ },
338
+ {
339
+ "epoch": 142.11,
340
+ "learning_rate": 8.117777777777777e-08,
341
+ "loss": 0.3511,
342
+ "step": 1350
343
+ },
344
+ {
345
+ "epoch": 144.74,
346
+ "learning_rate": 8.062222222222221e-08,
347
+ "loss": 0.3376,
348
+ "step": 1375
349
+ },
350
+ {
351
+ "epoch": 147.37,
352
+ "learning_rate": 8.006666666666666e-08,
353
+ "loss": 0.3369,
354
+ "step": 1400
355
+ },
356
+ {
357
+ "epoch": 150.0,
358
+ "learning_rate": 7.95111111111111e-08,
359
+ "loss": 0.3302,
360
+ "step": 1425
361
+ },
362
+ {
363
+ "epoch": 152.63,
364
+ "learning_rate": 7.895555555555555e-08,
365
+ "loss": 0.3339,
366
+ "step": 1450
367
+ },
368
+ {
369
+ "epoch": 155.26,
370
+ "learning_rate": 7.84e-08,
371
+ "loss": 0.33,
372
+ "step": 1475
373
+ },
374
+ {
375
+ "epoch": 157.89,
376
+ "learning_rate": 7.784444444444445e-08,
377
+ "loss": 0.3285,
378
+ "step": 1500
379
+ },
380
+ {
381
+ "epoch": 160.53,
382
+ "learning_rate": 7.728888888888888e-08,
383
+ "loss": 0.3236,
384
+ "step": 1525
385
+ },
386
+ {
387
+ "epoch": 163.16,
388
+ "learning_rate": 7.673333333333333e-08,
389
+ "loss": 0.3167,
390
+ "step": 1550
391
+ },
392
+ {
393
+ "epoch": 165.79,
394
+ "learning_rate": 7.617777777777778e-08,
395
+ "loss": 0.3117,
396
+ "step": 1575
397
+ },
398
+ {
399
+ "epoch": 168.42,
400
+ "learning_rate": 7.562222222222222e-08,
401
+ "loss": 0.3191,
402
+ "step": 1600
403
+ },
404
+ {
405
+ "epoch": 171.05,
406
+ "learning_rate": 7.506666666666666e-08,
407
+ "loss": 0.3058,
408
+ "step": 1625
409
+ },
410
+ {
411
+ "epoch": 173.68,
412
+ "learning_rate": 7.45111111111111e-08,
413
+ "loss": 0.3177,
414
+ "step": 1650
415
+ },
416
+ {
417
+ "epoch": 176.32,
418
+ "learning_rate": 7.395555555555555e-08,
419
+ "loss": 0.3017,
420
+ "step": 1675
421
+ },
422
+ {
423
+ "epoch": 178.95,
424
+ "learning_rate": 7.34e-08,
425
+ "loss": 0.3059,
426
+ "step": 1700
427
+ },
428
+ {
429
+ "epoch": 181.58,
430
+ "learning_rate": 7.284444444444444e-08,
431
+ "loss": 0.3124,
432
+ "step": 1725
433
+ },
434
+ {
435
+ "epoch": 184.21,
436
+ "learning_rate": 7.228888888888889e-08,
437
+ "loss": 0.2987,
438
+ "step": 1750
439
+ },
440
+ {
441
+ "epoch": 186.84,
442
+ "learning_rate": 7.173333333333334e-08,
443
+ "loss": 0.2981,
444
+ "step": 1775
445
+ },
446
+ {
447
+ "epoch": 189.47,
448
+ "learning_rate": 7.117777777777778e-08,
449
+ "loss": 0.3084,
450
+ "step": 1800
451
+ },
452
+ {
453
+ "epoch": 192.11,
454
+ "learning_rate": 7.062222222222221e-08,
455
+ "loss": 0.2858,
456
+ "step": 1825
457
+ },
458
+ {
459
+ "epoch": 194.74,
460
+ "learning_rate": 7.006666666666666e-08,
461
+ "loss": 0.2978,
462
+ "step": 1850
463
+ },
464
+ {
465
+ "epoch": 197.37,
466
+ "learning_rate": 6.951111111111111e-08,
467
+ "loss": 0.2818,
468
+ "step": 1875
469
+ },
470
+ {
471
+ "epoch": 200.0,
472
+ "learning_rate": 6.895555555555556e-08,
473
+ "loss": 0.2902,
474
+ "step": 1900
475
+ },
476
+ {
477
+ "epoch": 202.63,
478
+ "learning_rate": 6.84e-08,
479
+ "loss": 0.2838,
480
+ "step": 1925
481
+ },
482
+ {
483
+ "epoch": 205.26,
484
+ "learning_rate": 6.784444444444444e-08,
485
+ "loss": 0.2992,
486
+ "step": 1950
487
+ },
488
+ {
489
+ "epoch": 207.89,
490
+ "learning_rate": 6.728888888888889e-08,
491
+ "loss": 0.2774,
492
+ "step": 1975
493
+ },
494
+ {
495
+ "epoch": 210.53,
496
+ "learning_rate": 6.673333333333334e-08,
497
+ "loss": 0.2837,
498
+ "step": 2000
499
+ },
500
+ {
501
+ "epoch": 210.53,
502
+ "eval_loss": 0.5525617003440857,
503
+ "eval_runtime": 144.9276,
504
+ "eval_samples_per_second": 4.664,
505
+ "eval_steps_per_second": 0.152,
506
+ "eval_wer": 23.895526535148655,
507
+ "step": 2000
508
+ },
509
+ {
510
+ "epoch": 213.16,
511
+ "learning_rate": 6.617777777777777e-08,
512
+ "loss": 0.2805,
513
+ "step": 2025
514
+ },
515
+ {
516
+ "epoch": 215.79,
517
+ "learning_rate": 6.562222222222222e-08,
518
+ "loss": 0.2759,
519
+ "step": 2050
520
+ },
521
+ {
522
+ "epoch": 218.42,
523
+ "learning_rate": 6.506666666666665e-08,
524
+ "loss": 0.2667,
525
+ "step": 2075
526
+ },
527
+ {
528
+ "epoch": 221.05,
529
+ "learning_rate": 6.45111111111111e-08,
530
+ "loss": 0.2777,
531
+ "step": 2100
532
+ },
533
+ {
534
+ "epoch": 223.68,
535
+ "learning_rate": 6.395555555555555e-08,
536
+ "loss": 0.2672,
537
+ "step": 2125
538
+ },
539
+ {
540
+ "epoch": 226.32,
541
+ "learning_rate": 6.34e-08,
542
+ "loss": 0.2739,
543
+ "step": 2150
544
+ },
545
+ {
546
+ "epoch": 228.95,
547
+ "learning_rate": 6.284444444444445e-08,
548
+ "loss": 0.2819,
549
+ "step": 2175
550
+ },
551
+ {
552
+ "epoch": 231.58,
553
+ "learning_rate": 6.22888888888889e-08,
554
+ "loss": 0.2646,
555
+ "step": 2200
556
+ },
557
+ {
558
+ "epoch": 234.21,
559
+ "learning_rate": 6.173333333333333e-08,
560
+ "loss": 0.2618,
561
+ "step": 2225
562
+ },
563
+ {
564
+ "epoch": 236.84,
565
+ "learning_rate": 6.117777777777778e-08,
566
+ "loss": 0.2711,
567
+ "step": 2250
568
+ },
569
+ {
570
+ "epoch": 239.47,
571
+ "learning_rate": 6.062222222222221e-08,
572
+ "loss": 0.2644,
573
+ "step": 2275
574
+ },
575
+ {
576
+ "epoch": 242.11,
577
+ "learning_rate": 6.006666666666666e-08,
578
+ "loss": 0.2494,
579
+ "step": 2300
580
+ },
581
+ {
582
+ "epoch": 244.74,
583
+ "learning_rate": 5.951111111111111e-08,
584
+ "loss": 0.2601,
585
+ "step": 2325
586
+ },
587
+ {
588
+ "epoch": 247.37,
589
+ "learning_rate": 5.895555555555556e-08,
590
+ "loss": 0.2538,
591
+ "step": 2350
592
+ },
593
+ {
594
+ "epoch": 250.0,
595
+ "learning_rate": 5.8399999999999994e-08,
596
+ "loss": 0.2551,
597
+ "step": 2375
598
+ },
599
+ {
600
+ "epoch": 252.63,
601
+ "learning_rate": 5.784444444444444e-08,
602
+ "loss": 0.2494,
603
+ "step": 2400
604
+ },
605
+ {
606
+ "epoch": 255.26,
607
+ "learning_rate": 5.7288888888888884e-08,
608
+ "loss": 0.2518,
609
+ "step": 2425
610
+ },
611
+ {
612
+ "epoch": 257.89,
613
+ "learning_rate": 5.673333333333333e-08,
614
+ "loss": 0.2455,
615
+ "step": 2450
616
+ },
617
+ {
618
+ "epoch": 260.53,
619
+ "learning_rate": 5.617777777777778e-08,
620
+ "loss": 0.2606,
621
+ "step": 2475
622
+ },
623
+ {
624
+ "epoch": 263.16,
625
+ "learning_rate": 5.5622222222222215e-08,
626
+ "loss": 0.2533,
627
+ "step": 2500
628
+ },
629
+ {
630
+ "epoch": 265.79,
631
+ "learning_rate": 5.5066666666666663e-08,
632
+ "loss": 0.2436,
633
+ "step": 2525
634
+ },
635
+ {
636
+ "epoch": 268.42,
637
+ "learning_rate": 5.4511111111111105e-08,
638
+ "loss": 0.2346,
639
+ "step": 2550
640
+ },
641
+ {
642
+ "epoch": 271.05,
643
+ "learning_rate": 5.3955555555555553e-08,
644
+ "loss": 0.2456,
645
+ "step": 2575
646
+ },
647
+ {
648
+ "epoch": 273.68,
649
+ "learning_rate": 5.34e-08,
650
+ "loss": 0.2462,
651
+ "step": 2600
652
+ },
653
+ {
654
+ "epoch": 276.32,
655
+ "learning_rate": 5.284444444444445e-08,
656
+ "loss": 0.24,
657
+ "step": 2625
658
+ },
659
+ {
660
+ "epoch": 278.95,
661
+ "learning_rate": 5.2288888888888885e-08,
662
+ "loss": 0.2466,
663
+ "step": 2650
664
+ },
665
+ {
666
+ "epoch": 281.58,
667
+ "learning_rate": 5.173333333333333e-08,
668
+ "loss": 0.24,
669
+ "step": 2675
670
+ },
671
+ {
672
+ "epoch": 284.21,
673
+ "learning_rate": 5.1177777777777775e-08,
674
+ "loss": 0.2419,
675
+ "step": 2700
676
+ },
677
+ {
678
+ "epoch": 286.84,
679
+ "learning_rate": 5.0622222222222223e-08,
680
+ "loss": 0.2366,
681
+ "step": 2725
682
+ },
683
+ {
684
+ "epoch": 289.47,
685
+ "learning_rate": 5.0066666666666665e-08,
686
+ "loss": 0.232,
687
+ "step": 2750
688
+ },
689
+ {
690
+ "epoch": 292.11,
691
+ "learning_rate": 4.9511111111111107e-08,
692
+ "loss": 0.235,
693
+ "step": 2775
694
+ },
695
+ {
696
+ "epoch": 294.74,
697
+ "learning_rate": 4.8955555555555555e-08,
698
+ "loss": 0.234,
699
+ "step": 2800
700
+ },
701
+ {
702
+ "epoch": 297.37,
703
+ "learning_rate": 4.8399999999999997e-08,
704
+ "loss": 0.2368,
705
+ "step": 2825
706
+ },
707
+ {
708
+ "epoch": 300.0,
709
+ "learning_rate": 4.7844444444444445e-08,
710
+ "loss": 0.2231,
711
+ "step": 2850
712
+ },
713
+ {
714
+ "epoch": 302.63,
715
+ "learning_rate": 4.7288888888888887e-08,
716
+ "loss": 0.224,
717
+ "step": 2875
718
+ },
719
+ {
720
+ "epoch": 305.26,
721
+ "learning_rate": 4.673333333333333e-08,
722
+ "loss": 0.2285,
723
+ "step": 2900
724
+ },
725
+ {
726
+ "epoch": 307.89,
727
+ "learning_rate": 4.6177777777777777e-08,
728
+ "loss": 0.2265,
729
+ "step": 2925
730
+ },
731
+ {
732
+ "epoch": 310.53,
733
+ "learning_rate": 4.562222222222222e-08,
734
+ "loss": 0.2238,
735
+ "step": 2950
736
+ },
737
+ {
738
+ "epoch": 313.16,
739
+ "learning_rate": 4.5066666666666667e-08,
740
+ "loss": 0.2224,
741
+ "step": 2975
742
+ },
743
+ {
744
+ "epoch": 315.79,
745
+ "learning_rate": 4.451111111111111e-08,
746
+ "loss": 0.2323,
747
+ "step": 3000
748
+ },
749
+ {
750
+ "epoch": 315.79,
751
+ "eval_loss": 0.5431821942329407,
752
+ "eval_runtime": 144.0249,
753
+ "eval_samples_per_second": 4.694,
754
+ "eval_steps_per_second": 0.153,
755
+ "eval_wer": 24.012225618227284,
756
+ "step": 3000
757
+ },
758
+ {
759
+ "epoch": 318.42,
760
+ "learning_rate": 4.395555555555555e-08,
761
+ "loss": 0.2231,
762
+ "step": 3025
763
+ },
764
+ {
765
+ "epoch": 321.05,
766
+ "learning_rate": 4.34e-08,
767
+ "loss": 0.2132,
768
+ "step": 3050
769
+ },
770
+ {
771
+ "epoch": 323.68,
772
+ "learning_rate": 4.2844444444444446e-08,
773
+ "loss": 0.2275,
774
+ "step": 3075
775
+ },
776
+ {
777
+ "epoch": 326.32,
778
+ "learning_rate": 4.228888888888889e-08,
779
+ "loss": 0.2149,
780
+ "step": 3100
781
+ },
782
+ {
783
+ "epoch": 328.95,
784
+ "learning_rate": 4.173333333333333e-08,
785
+ "loss": 0.214,
786
+ "step": 3125
787
+ },
788
+ {
789
+ "epoch": 331.58,
790
+ "learning_rate": 4.117777777777778e-08,
791
+ "loss": 0.2181,
792
+ "step": 3150
793
+ },
794
+ {
795
+ "epoch": 334.21,
796
+ "learning_rate": 4.062222222222222e-08,
797
+ "loss": 0.2153,
798
+ "step": 3175
799
+ },
800
+ {
801
+ "epoch": 336.84,
802
+ "learning_rate": 4.006666666666667e-08,
803
+ "loss": 0.2266,
804
+ "step": 3200
805
+ },
806
+ {
807
+ "epoch": 339.47,
808
+ "learning_rate": 3.951111111111111e-08,
809
+ "loss": 0.2154,
810
+ "step": 3225
811
+ },
812
+ {
813
+ "epoch": 342.11,
814
+ "learning_rate": 3.895555555555555e-08,
815
+ "loss": 0.2088,
816
+ "step": 3250
817
+ },
818
+ {
819
+ "epoch": 344.74,
820
+ "learning_rate": 3.84e-08,
821
+ "loss": 0.2104,
822
+ "step": 3275
823
+ },
824
+ {
825
+ "epoch": 347.37,
826
+ "learning_rate": 3.784444444444445e-08,
827
+ "loss": 0.2115,
828
+ "step": 3300
829
+ },
830
+ {
831
+ "epoch": 350.0,
832
+ "learning_rate": 3.728888888888888e-08,
833
+ "loss": 0.2103,
834
+ "step": 3325
835
+ },
836
+ {
837
+ "epoch": 352.63,
838
+ "learning_rate": 3.673333333333333e-08,
839
+ "loss": 0.209,
840
+ "step": 3350
841
+ },
842
+ {
843
+ "epoch": 355.26,
844
+ "learning_rate": 3.617777777777777e-08,
845
+ "loss": 0.216,
846
+ "step": 3375
847
+ },
848
+ {
849
+ "epoch": 357.89,
850
+ "learning_rate": 3.562222222222222e-08,
851
+ "loss": 0.2125,
852
+ "step": 3400
853
+ },
854
+ {
855
+ "epoch": 360.53,
856
+ "learning_rate": 3.506666666666667e-08,
857
+ "loss": 0.2161,
858
+ "step": 3425
859
+ },
860
+ {
861
+ "epoch": 363.16,
862
+ "learning_rate": 3.4511111111111105e-08,
863
+ "loss": 0.2148,
864
+ "step": 3450
865
+ },
866
+ {
867
+ "epoch": 365.79,
868
+ "learning_rate": 3.395555555555555e-08,
869
+ "loss": 0.2093,
870
+ "step": 3475
871
+ },
872
+ {
873
+ "epoch": 368.42,
874
+ "learning_rate": 3.34e-08,
875
+ "loss": 0.215,
876
+ "step": 3500
877
+ },
878
+ {
879
+ "epoch": 371.05,
880
+ "learning_rate": 3.284444444444444e-08,
881
+ "loss": 0.2142,
882
+ "step": 3525
883
+ },
884
+ {
885
+ "epoch": 373.68,
886
+ "learning_rate": 3.2288888888888885e-08,
887
+ "loss": 0.2069,
888
+ "step": 3550
889
+ },
890
+ {
891
+ "epoch": 376.32,
892
+ "learning_rate": 3.173333333333333e-08,
893
+ "loss": 0.2067,
894
+ "step": 3575
895
+ },
896
+ {
897
+ "epoch": 378.95,
898
+ "learning_rate": 3.1177777777777775e-08,
899
+ "loss": 0.2084,
900
+ "step": 3600
901
+ },
902
+ {
903
+ "epoch": 381.58,
904
+ "learning_rate": 3.062222222222222e-08,
905
+ "loss": 0.209,
906
+ "step": 3625
907
+ },
908
+ {
909
+ "epoch": 384.21,
910
+ "learning_rate": 3.006666666666667e-08,
911
+ "loss": 0.2139,
912
+ "step": 3650
913
+ },
914
+ {
915
+ "epoch": 386.84,
916
+ "learning_rate": 2.951111111111111e-08,
917
+ "loss": 0.2019,
918
+ "step": 3675
919
+ },
920
+ {
921
+ "epoch": 389.47,
922
+ "learning_rate": 2.8955555555555555e-08,
923
+ "loss": 0.1983,
924
+ "step": 3700
925
+ },
926
+ {
927
+ "epoch": 392.11,
928
+ "learning_rate": 2.8399999999999996e-08,
929
+ "loss": 0.2043,
930
+ "step": 3725
931
+ },
932
+ {
933
+ "epoch": 394.74,
934
+ "learning_rate": 2.784444444444444e-08,
935
+ "loss": 0.2044,
936
+ "step": 3750
937
+ },
938
+ {
939
+ "epoch": 397.37,
940
+ "learning_rate": 2.728888888888889e-08,
941
+ "loss": 0.2025,
942
+ "step": 3775
943
+ },
944
+ {
945
+ "epoch": 400.0,
946
+ "learning_rate": 2.673333333333333e-08,
947
+ "loss": 0.2036,
948
+ "step": 3800
949
+ },
950
+ {
951
+ "epoch": 402.63,
952
+ "learning_rate": 2.6177777777777776e-08,
953
+ "loss": 0.1942,
954
+ "step": 3825
955
+ },
956
+ {
957
+ "epoch": 405.26,
958
+ "learning_rate": 2.5622222222222224e-08,
959
+ "loss": 0.1977,
960
+ "step": 3850
961
+ },
962
+ {
963
+ "epoch": 407.89,
964
+ "learning_rate": 2.5066666666666663e-08,
965
+ "loss": 0.2042,
966
+ "step": 3875
967
+ },
968
+ {
969
+ "epoch": 410.53,
970
+ "learning_rate": 2.451111111111111e-08,
971
+ "loss": 0.1933,
972
+ "step": 3900
973
+ },
974
+ {
975
+ "epoch": 413.16,
976
+ "learning_rate": 2.3955555555555553e-08,
977
+ "loss": 0.1988,
978
+ "step": 3925
979
+ },
980
+ {
981
+ "epoch": 415.79,
982
+ "learning_rate": 2.34e-08,
983
+ "loss": 0.1943,
984
+ "step": 3950
985
+ },
986
+ {
987
+ "epoch": 418.42,
988
+ "learning_rate": 2.2844444444444443e-08,
989
+ "loss": 0.2033,
990
+ "step": 3975
991
+ },
992
+ {
993
+ "epoch": 421.05,
994
+ "learning_rate": 2.2288888888888888e-08,
995
+ "loss": 0.1961,
996
+ "step": 4000
997
+ },
998
+ {
999
+ "epoch": 421.05,
1000
+ "eval_loss": 0.5402421951293945,
1001
+ "eval_runtime": 143.5894,
1002
+ "eval_samples_per_second": 4.708,
1003
+ "eval_steps_per_second": 0.153,
1004
+ "eval_wer": 23.795498749652683,
1005
+ "step": 4000
1006
+ },
1007
+ {
1008
+ "epoch": 423.68,
1009
+ "learning_rate": 2.1733333333333333e-08,
1010
+ "loss": 0.2009,
1011
+ "step": 4025
1012
+ },
1013
+ {
1014
+ "epoch": 426.32,
1015
+ "learning_rate": 2.1177777777777778e-08,
1016
+ "loss": 0.1961,
1017
+ "step": 4050
1018
+ },
1019
+ {
1020
+ "epoch": 428.95,
1021
+ "learning_rate": 2.0622222222222223e-08,
1022
+ "loss": 0.1987,
1023
+ "step": 4075
1024
+ },
1025
+ {
1026
+ "epoch": 431.58,
1027
+ "learning_rate": 2.0066666666666664e-08,
1028
+ "loss": 0.1951,
1029
+ "step": 4100
1030
+ },
1031
+ {
1032
+ "epoch": 434.21,
1033
+ "learning_rate": 1.9533333333333333e-08,
1034
+ "loss": 0.1914,
1035
+ "step": 4125
1036
+ },
1037
+ {
1038
+ "epoch": 436.84,
1039
+ "learning_rate": 1.8977777777777775e-08,
1040
+ "loss": 0.1955,
1041
+ "step": 4150
1042
+ },
1043
+ {
1044
+ "epoch": 439.47,
1045
+ "learning_rate": 1.8422222222222223e-08,
1046
+ "loss": 0.1967,
1047
+ "step": 4175
1048
+ },
1049
+ {
1050
+ "epoch": 442.11,
1051
+ "learning_rate": 1.7866666666666665e-08,
1052
+ "loss": 0.2011,
1053
+ "step": 4200
1054
+ },
1055
+ {
1056
+ "epoch": 444.74,
1057
+ "learning_rate": 1.731111111111111e-08,
1058
+ "loss": 0.1953,
1059
+ "step": 4225
1060
+ },
1061
+ {
1062
+ "epoch": 447.37,
1063
+ "learning_rate": 1.6755555555555554e-08,
1064
+ "loss": 0.1927,
1065
+ "step": 4250
1066
+ },
1067
+ {
1068
+ "epoch": 450.0,
1069
+ "learning_rate": 1.62e-08,
1070
+ "loss": 0.1918,
1071
+ "step": 4275
1072
+ },
1073
+ {
1074
+ "epoch": 452.63,
1075
+ "learning_rate": 1.5644444444444444e-08,
1076
+ "loss": 0.195,
1077
+ "step": 4300
1078
+ },
1079
+ {
1080
+ "epoch": 455.26,
1081
+ "learning_rate": 1.5088888888888886e-08,
1082
+ "loss": 0.1939,
1083
+ "step": 4325
1084
+ },
1085
+ {
1086
+ "epoch": 457.89,
1087
+ "learning_rate": 1.4533333333333333e-08,
1088
+ "loss": 0.1916,
1089
+ "step": 4350
1090
+ },
1091
+ {
1092
+ "epoch": 460.53,
1093
+ "learning_rate": 1.3977777777777778e-08,
1094
+ "loss": 0.1968,
1095
+ "step": 4375
1096
+ },
1097
+ {
1098
+ "epoch": 463.16,
1099
+ "learning_rate": 1.3422222222222221e-08,
1100
+ "loss": 0.193,
1101
+ "step": 4400
1102
+ },
1103
+ {
1104
+ "epoch": 465.79,
1105
+ "learning_rate": 1.2866666666666668e-08,
1106
+ "loss": 0.1888,
1107
+ "step": 4425
1108
+ },
1109
+ {
1110
+ "epoch": 468.42,
1111
+ "learning_rate": 1.2311111111111111e-08,
1112
+ "loss": 0.1858,
1113
+ "step": 4450
1114
+ },
1115
+ {
1116
+ "epoch": 471.05,
1117
+ "learning_rate": 1.1755555555555554e-08,
1118
+ "loss": 0.1872,
1119
+ "step": 4475
1120
+ },
1121
+ {
1122
+ "epoch": 473.68,
1123
+ "learning_rate": 1.12e-08,
1124
+ "loss": 0.1919,
1125
+ "step": 4500
1126
+ },
1127
+ {
1128
+ "epoch": 476.32,
1129
+ "learning_rate": 1.0644444444444443e-08,
1130
+ "loss": 0.1919,
1131
+ "step": 4525
1132
+ },
1133
+ {
1134
+ "epoch": 478.95,
1135
+ "learning_rate": 1.0088888888888888e-08,
1136
+ "loss": 0.1901,
1137
+ "step": 4550
1138
+ },
1139
+ {
1140
+ "epoch": 481.58,
1141
+ "learning_rate": 9.533333333333334e-09,
1142
+ "loss": 0.1947,
1143
+ "step": 4575
1144
+ },
1145
+ {
1146
+ "epoch": 484.21,
1147
+ "learning_rate": 8.977777777777778e-09,
1148
+ "loss": 0.189,
1149
+ "step": 4600
1150
+ },
1151
+ {
1152
+ "epoch": 486.84,
1153
+ "learning_rate": 8.422222222222223e-09,
1154
+ "loss": 0.1872,
1155
+ "step": 4625
1156
+ },
1157
+ {
1158
+ "epoch": 489.47,
1159
+ "learning_rate": 7.866666666666666e-09,
1160
+ "loss": 0.1848,
1161
+ "step": 4650
1162
+ },
1163
+ {
1164
+ "epoch": 492.11,
1165
+ "learning_rate": 7.311111111111111e-09,
1166
+ "loss": 0.1876,
1167
+ "step": 4675
1168
+ },
1169
+ {
1170
+ "epoch": 494.74,
1171
+ "learning_rate": 6.755555555555554e-09,
1172
+ "loss": 0.1911,
1173
+ "step": 4700
1174
+ },
1175
+ {
1176
+ "epoch": 497.37,
1177
+ "learning_rate": 6.199999999999999e-09,
1178
+ "loss": 0.1882,
1179
+ "step": 4725
1180
+ },
1181
+ {
1182
+ "epoch": 500.0,
1183
+ "learning_rate": 5.644444444444444e-09,
1184
+ "loss": 0.1872,
1185
+ "step": 4750
1186
+ },
1187
+ {
1188
+ "epoch": 502.63,
1189
+ "learning_rate": 5.088888888888888e-09,
1190
+ "loss": 0.1895,
1191
+ "step": 4775
1192
+ },
1193
+ {
1194
+ "epoch": 505.26,
1195
+ "learning_rate": 4.533333333333333e-09,
1196
+ "loss": 0.1895,
1197
+ "step": 4800
1198
+ },
1199
+ {
1200
+ "epoch": 507.89,
1201
+ "learning_rate": 3.9777777777777776e-09,
1202
+ "loss": 0.1834,
1203
+ "step": 4825
1204
+ },
1205
+ {
1206
+ "epoch": 510.53,
1207
+ "learning_rate": 3.422222222222222e-09,
1208
+ "loss": 0.1923,
1209
+ "step": 4850
1210
+ },
1211
+ {
1212
+ "epoch": 513.16,
1213
+ "learning_rate": 2.8666666666666667e-09,
1214
+ "loss": 0.1884,
1215
+ "step": 4875
1216
+ },
1217
+ {
1218
+ "epoch": 515.79,
1219
+ "learning_rate": 2.311111111111111e-09,
1220
+ "loss": 0.1834,
1221
+ "step": 4900
1222
+ },
1223
+ {
1224
+ "epoch": 518.42,
1225
+ "learning_rate": 1.7555555555555556e-09,
1226
+ "loss": 0.1964,
1227
+ "step": 4925
1228
+ },
1229
+ {
1230
+ "epoch": 521.05,
1231
+ "learning_rate": 1.2e-09,
1232
+ "loss": 0.1923,
1233
+ "step": 4950
1234
+ },
1235
+ {
1236
+ "epoch": 523.68,
1237
+ "learning_rate": 6.444444444444445e-10,
1238
+ "loss": 0.1866,
1239
+ "step": 4975
1240
+ },
1241
+ {
1242
+ "epoch": 526.32,
1243
+ "learning_rate": 8.888888888888889e-11,
1244
+ "loss": 0.1863,
1245
+ "step": 5000
1246
+ },
1247
+ {
1248
+ "epoch": 526.32,
1249
+ "eval_loss": 0.5395069122314453,
1250
+ "eval_runtime": 143.5193,
1251
+ "eval_samples_per_second": 4.71,
1252
+ "eval_steps_per_second": 0.153,
1253
+ "eval_wer": 23.795498749652683,
1254
+ "step": 5000
1255
+ },
1256
+ {
1257
+ "epoch": 526.32,
1258
+ "step": 5000,
1259
+ "total_flos": 1.97317080317952e+19,
1260
+ "train_loss": 0.31816294193267824,
1261
+ "train_runtime": 12909.7746,
1262
+ "train_samples_per_second": 24.787,
1263
+ "train_steps_per_second": 0.387
1264
+ }
1265
+ ],
1266
+ "logging_steps": 25,
1267
+ "max_steps": 5000,
1268
+ "num_input_tokens_seen": 0,
1269
+ "num_train_epochs": 556,
1270
+ "save_steps": 1000,
1271
+ "total_flos": 1.97317080317952e+19,
1272
+ "train_batch_size": 32,
1273
+ "trial_name": null,
1274
+ "trial_params": null
1275
+ }