arun100 commited on
Commit
c9218f3
1 Parent(s): d6f5313

End of training

Browse files
README.md CHANGED
@@ -1,21 +1,24 @@
1
  ---
 
 
2
  license: apache-2.0
3
  base_model: openai/whisper-base
4
  tags:
 
5
  - generated_from_trainer
6
  datasets:
7
- - common_voice_16_0
8
  metrics:
9
  - wer
10
  model-index:
11
- - name: openai/whisper-base
12
  results:
13
  - task:
14
  name: Automatic Speech Recognition
15
  type: automatic-speech-recognition
16
  dataset:
17
- name: common_voice_16_0
18
- type: common_voice_16_0
19
  config: hi
20
  split: test
21
  args: hi
@@ -28,9 +31,9 @@ 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
- # openai/whisper-base
32
 
33
- This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the common_voice_16_0 dataset.
34
  It achieves the following results on the evaluation set:
35
  - Loss: 0.4679
36
  - Wer: 28.6490
 
1
  ---
2
+ language:
3
+ - hi
4
  license: apache-2.0
5
  base_model: openai/whisper-base
6
  tags:
7
+ - whisper-event
8
  - generated_from_trainer
9
  datasets:
10
+ - mozilla-foundation/common_voice_16_0
11
  metrics:
12
  - wer
13
  model-index:
14
+ - name: Whisper Base Hindi
15
  results:
16
  - task:
17
  name: Automatic Speech Recognition
18
  type: automatic-speech-recognition
19
  dataset:
20
+ name: mozilla-foundation/common_voice_16_0 hi
21
+ type: mozilla-foundation/common_voice_16_0
22
  config: hi
23
  split: test
24
  args: hi
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
32
  should probably proofread and complete it, then remove this comment. -->
33
 
34
+ # Whisper Base Hindi
35
 
36
+ This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the mozilla-foundation/common_voice_16_0 hi dataset.
37
  It achieves the following results on the evaluation set:
38
  - Loss: 0.4679
39
  - Wer: 28.6490
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 65.01,
3
+ "eval_loss": 0.46787169575691223,
4
+ "eval_runtime": 334.2634,
5
+ "eval_samples_per_second": 9.295,
6
+ "eval_steps_per_second": 0.293,
7
+ "eval_wer": 28.648953267516852,
8
+ "train_loss": 0.34305178251266477,
9
+ "train_runtime": 22591.6172,
10
+ "train_samples_per_second": 14.165,
11
+ "train_steps_per_second": 0.221
12
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 65.01,
3
+ "eval_loss": 0.46787169575691223,
4
+ "eval_runtime": 334.2634,
5
+ "eval_samples_per_second": 9.295,
6
+ "eval_steps_per_second": 0.293,
7
+ "eval_wer": 28.648953267516852
8
+ }
runs/Jan19_05-40-02_ip-172-31-23-125/events.out.tfevents.1705665917.ip-172-31-23-125.20410.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad28646723f7fb75354a503b540c7c5411ebb164d48cd99c83c0c1b5ea1b8f3a
3
+ size 406
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 65.01,
3
+ "train_loss": 0.34305178251266477,
4
+ "train_runtime": 22591.6172,
5
+ "train_samples_per_second": 14.165,
6
+ "train_steps_per_second": 0.221
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,1320 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 28.648953267516852,
3
+ "best_model_checkpoint": "./checkpoint-5000",
4
+ "epoch": 65.012,
5
+ "eval_steps": 500,
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": 0.01,
13
+ "learning_rate": 4.8e-08,
14
+ "loss": 1.6929,
15
+ "step": 25
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 9.8e-08,
20
+ "loss": 1.6161,
21
+ "step": 50
22
+ },
23
+ {
24
+ "epoch": 0.01,
25
+ "learning_rate": 1.4799999999999998e-07,
26
+ "loss": 1.5665,
27
+ "step": 75
28
+ },
29
+ {
30
+ "epoch": 1.0,
31
+ "learning_rate": 1.98e-07,
32
+ "loss": 1.5875,
33
+ "step": 100
34
+ },
35
+ {
36
+ "epoch": 1.01,
37
+ "learning_rate": 2.48e-07,
38
+ "loss": 1.4633,
39
+ "step": 125
40
+ },
41
+ {
42
+ "epoch": 1.01,
43
+ "learning_rate": 2.98e-07,
44
+ "loss": 1.3517,
45
+ "step": 150
46
+ },
47
+ {
48
+ "epoch": 2.0,
49
+ "learning_rate": 3.4799999999999994e-07,
50
+ "loss": 1.3473,
51
+ "step": 175
52
+ },
53
+ {
54
+ "epoch": 2.01,
55
+ "learning_rate": 3.98e-07,
56
+ "loss": 1.2353,
57
+ "step": 200
58
+ },
59
+ {
60
+ "epoch": 2.01,
61
+ "learning_rate": 4.48e-07,
62
+ "loss": 1.1232,
63
+ "step": 225
64
+ },
65
+ {
66
+ "epoch": 3.0,
67
+ "learning_rate": 4.979999999999999e-07,
68
+ "loss": 1.0898,
69
+ "step": 250
70
+ },
71
+ {
72
+ "epoch": 3.01,
73
+ "learning_rate": 5.48e-07,
74
+ "loss": 0.9945,
75
+ "step": 275
76
+ },
77
+ {
78
+ "epoch": 3.01,
79
+ "learning_rate": 5.979999999999999e-07,
80
+ "loss": 0.8935,
81
+ "step": 300
82
+ },
83
+ {
84
+ "epoch": 4.0,
85
+ "learning_rate": 6.48e-07,
86
+ "loss": 0.8898,
87
+ "step": 325
88
+ },
89
+ {
90
+ "epoch": 4.01,
91
+ "learning_rate": 6.979999999999999e-07,
92
+ "loss": 0.8355,
93
+ "step": 350
94
+ },
95
+ {
96
+ "epoch": 4.01,
97
+ "learning_rate": 7.48e-07,
98
+ "loss": 0.7488,
99
+ "step": 375
100
+ },
101
+ {
102
+ "epoch": 5.0,
103
+ "learning_rate": 7.98e-07,
104
+ "loss": 0.7638,
105
+ "step": 400
106
+ },
107
+ {
108
+ "epoch": 5.01,
109
+ "learning_rate": 8.48e-07,
110
+ "loss": 0.7228,
111
+ "step": 425
112
+ },
113
+ {
114
+ "epoch": 5.01,
115
+ "learning_rate": 8.98e-07,
116
+ "loss": 0.6542,
117
+ "step": 450
118
+ },
119
+ {
120
+ "epoch": 6.0,
121
+ "learning_rate": 9.479999999999999e-07,
122
+ "loss": 0.6631,
123
+ "step": 475
124
+ },
125
+ {
126
+ "epoch": 6.01,
127
+ "learning_rate": 9.98e-07,
128
+ "loss": 0.6425,
129
+ "step": 500
130
+ },
131
+ {
132
+ "epoch": 6.01,
133
+ "eval_loss": 0.702507495880127,
134
+ "eval_runtime": 521.2801,
135
+ "eval_samples_per_second": 5.96,
136
+ "eval_steps_per_second": 0.188,
137
+ "eval_wer": 41.44768424787627,
138
+ "step": 500
139
+ },
140
+ {
141
+ "epoch": 6.01,
142
+ "learning_rate": 9.946666666666666e-07,
143
+ "loss": 0.5748,
144
+ "step": 525
145
+ },
146
+ {
147
+ "epoch": 7.0,
148
+ "learning_rate": 9.891111111111112e-07,
149
+ "loss": 0.5951,
150
+ "step": 550
151
+ },
152
+ {
153
+ "epoch": 7.01,
154
+ "learning_rate": 9.835555555555554e-07,
155
+ "loss": 0.5775,
156
+ "step": 575
157
+ },
158
+ {
159
+ "epoch": 7.01,
160
+ "learning_rate": 9.78e-07,
161
+ "loss": 0.5234,
162
+ "step": 600
163
+ },
164
+ {
165
+ "epoch": 8.0,
166
+ "learning_rate": 9.724444444444444e-07,
167
+ "loss": 0.5375,
168
+ "step": 625
169
+ },
170
+ {
171
+ "epoch": 8.01,
172
+ "learning_rate": 9.668888888888888e-07,
173
+ "loss": 0.5398,
174
+ "step": 650
175
+ },
176
+ {
177
+ "epoch": 8.01,
178
+ "learning_rate": 9.613333333333334e-07,
179
+ "loss": 0.479,
180
+ "step": 675
181
+ },
182
+ {
183
+ "epoch": 9.0,
184
+ "learning_rate": 9.557777777777776e-07,
185
+ "loss": 0.4986,
186
+ "step": 700
187
+ },
188
+ {
189
+ "epoch": 9.01,
190
+ "learning_rate": 9.502222222222221e-07,
191
+ "loss": 0.5009,
192
+ "step": 725
193
+ },
194
+ {
195
+ "epoch": 9.01,
196
+ "learning_rate": 9.446666666666666e-07,
197
+ "loss": 0.4526,
198
+ "step": 750
199
+ },
200
+ {
201
+ "epoch": 10.0,
202
+ "learning_rate": 9.39111111111111e-07,
203
+ "loss": 0.4678,
204
+ "step": 775
205
+ },
206
+ {
207
+ "epoch": 10.01,
208
+ "learning_rate": 9.335555555555555e-07,
209
+ "loss": 0.4692,
210
+ "step": 800
211
+ },
212
+ {
213
+ "epoch": 10.01,
214
+ "learning_rate": 9.28e-07,
215
+ "loss": 0.4286,
216
+ "step": 825
217
+ },
218
+ {
219
+ "epoch": 11.0,
220
+ "learning_rate": 9.224444444444443e-07,
221
+ "loss": 0.4415,
222
+ "step": 850
223
+ },
224
+ {
225
+ "epoch": 11.01,
226
+ "learning_rate": 9.168888888888889e-07,
227
+ "loss": 0.4503,
228
+ "step": 875
229
+ },
230
+ {
231
+ "epoch": 11.01,
232
+ "learning_rate": 9.113333333333333e-07,
233
+ "loss": 0.4035,
234
+ "step": 900
235
+ },
236
+ {
237
+ "epoch": 12.0,
238
+ "learning_rate": 9.057777777777778e-07,
239
+ "loss": 0.4175,
240
+ "step": 925
241
+ },
242
+ {
243
+ "epoch": 12.01,
244
+ "learning_rate": 9.002222222222222e-07,
245
+ "loss": 0.4333,
246
+ "step": 950
247
+ },
248
+ {
249
+ "epoch": 12.01,
250
+ "learning_rate": 8.946666666666667e-07,
251
+ "loss": 0.3904,
252
+ "step": 975
253
+ },
254
+ {
255
+ "epoch": 13.0,
256
+ "learning_rate": 8.89111111111111e-07,
257
+ "loss": 0.3973,
258
+ "step": 1000
259
+ },
260
+ {
261
+ "epoch": 13.0,
262
+ "eval_loss": 0.5367385149002075,
263
+ "eval_runtime": 333.3595,
264
+ "eval_samples_per_second": 9.32,
265
+ "eval_steps_per_second": 0.294,
266
+ "eval_wer": 33.96923462252927,
267
+ "step": 1000
268
+ },
269
+ {
270
+ "epoch": 13.01,
271
+ "learning_rate": 8.835555555555555e-07,
272
+ "loss": 0.4134,
273
+ "step": 1025
274
+ },
275
+ {
276
+ "epoch": 13.01,
277
+ "learning_rate": 8.78e-07,
278
+ "loss": 0.3698,
279
+ "step": 1050
280
+ },
281
+ {
282
+ "epoch": 14.0,
283
+ "learning_rate": 8.724444444444444e-07,
284
+ "loss": 0.3784,
285
+ "step": 1075
286
+ },
287
+ {
288
+ "epoch": 14.01,
289
+ "learning_rate": 8.668888888888889e-07,
290
+ "loss": 0.4007,
291
+ "step": 1100
292
+ },
293
+ {
294
+ "epoch": 14.01,
295
+ "learning_rate": 8.613333333333332e-07,
296
+ "loss": 0.3591,
297
+ "step": 1125
298
+ },
299
+ {
300
+ "epoch": 15.0,
301
+ "learning_rate": 8.557777777777777e-07,
302
+ "loss": 0.3615,
303
+ "step": 1150
304
+ },
305
+ {
306
+ "epoch": 15.01,
307
+ "learning_rate": 8.502222222222221e-07,
308
+ "loss": 0.3856,
309
+ "step": 1175
310
+ },
311
+ {
312
+ "epoch": 15.01,
313
+ "learning_rate": 8.446666666666666e-07,
314
+ "loss": 0.3527,
315
+ "step": 1200
316
+ },
317
+ {
318
+ "epoch": 16.0,
319
+ "learning_rate": 8.391111111111111e-07,
320
+ "loss": 0.3435,
321
+ "step": 1225
322
+ },
323
+ {
324
+ "epoch": 16.01,
325
+ "learning_rate": 8.335555555555555e-07,
326
+ "loss": 0.3771,
327
+ "step": 1250
328
+ },
329
+ {
330
+ "epoch": 16.01,
331
+ "learning_rate": 8.28e-07,
332
+ "loss": 0.3343,
333
+ "step": 1275
334
+ },
335
+ {
336
+ "epoch": 17.0,
337
+ "learning_rate": 8.224444444444444e-07,
338
+ "loss": 0.3335,
339
+ "step": 1300
340
+ },
341
+ {
342
+ "epoch": 17.01,
343
+ "learning_rate": 8.168888888888889e-07,
344
+ "loss": 0.363,
345
+ "step": 1325
346
+ },
347
+ {
348
+ "epoch": 17.01,
349
+ "learning_rate": 8.113333333333333e-07,
350
+ "loss": 0.3262,
351
+ "step": 1350
352
+ },
353
+ {
354
+ "epoch": 18.0,
355
+ "learning_rate": 8.057777777777778e-07,
356
+ "loss": 0.3186,
357
+ "step": 1375
358
+ },
359
+ {
360
+ "epoch": 18.01,
361
+ "learning_rate": 8.002222222222222e-07,
362
+ "loss": 0.3554,
363
+ "step": 1400
364
+ },
365
+ {
366
+ "epoch": 18.01,
367
+ "learning_rate": 7.946666666666666e-07,
368
+ "loss": 0.3174,
369
+ "step": 1425
370
+ },
371
+ {
372
+ "epoch": 19.0,
373
+ "learning_rate": 7.891111111111111e-07,
374
+ "loss": 0.3104,
375
+ "step": 1450
376
+ },
377
+ {
378
+ "epoch": 19.01,
379
+ "learning_rate": 7.835555555555555e-07,
380
+ "loss": 0.3405,
381
+ "step": 1475
382
+ },
383
+ {
384
+ "epoch": 19.01,
385
+ "learning_rate": 7.78e-07,
386
+ "loss": 0.3125,
387
+ "step": 1500
388
+ },
389
+ {
390
+ "epoch": 19.01,
391
+ "eval_loss": 0.4926978647708893,
392
+ "eval_runtime": 338.2106,
393
+ "eval_samples_per_second": 9.187,
394
+ "eval_steps_per_second": 0.29,
395
+ "eval_wer": 31.44580576485567,
396
+ "step": 1500
397
+ },
398
+ {
399
+ "epoch": 20.0,
400
+ "learning_rate": 7.724444444444444e-07,
401
+ "loss": 0.2975,
402
+ "step": 1525
403
+ },
404
+ {
405
+ "epoch": 20.01,
406
+ "learning_rate": 7.668888888888888e-07,
407
+ "loss": 0.335,
408
+ "step": 1550
409
+ },
410
+ {
411
+ "epoch": 20.01,
412
+ "learning_rate": 7.613333333333333e-07,
413
+ "loss": 0.3028,
414
+ "step": 1575
415
+ },
416
+ {
417
+ "epoch": 21.0,
418
+ "learning_rate": 7.557777777777777e-07,
419
+ "loss": 0.291,
420
+ "step": 1600
421
+ },
422
+ {
423
+ "epoch": 21.01,
424
+ "learning_rate": 7.502222222222222e-07,
425
+ "loss": 0.3219,
426
+ "step": 1625
427
+ },
428
+ {
429
+ "epoch": 21.01,
430
+ "learning_rate": 7.446666666666666e-07,
431
+ "loss": 0.2992,
432
+ "step": 1650
433
+ },
434
+ {
435
+ "epoch": 22.0,
436
+ "learning_rate": 7.391111111111111e-07,
437
+ "loss": 0.2821,
438
+ "step": 1675
439
+ },
440
+ {
441
+ "epoch": 22.01,
442
+ "learning_rate": 7.335555555555554e-07,
443
+ "loss": 0.3121,
444
+ "step": 1700
445
+ },
446
+ {
447
+ "epoch": 22.01,
448
+ "learning_rate": 7.28e-07,
449
+ "loss": 0.2933,
450
+ "step": 1725
451
+ },
452
+ {
453
+ "epoch": 23.0,
454
+ "learning_rate": 7.224444444444445e-07,
455
+ "loss": 0.2704,
456
+ "step": 1750
457
+ },
458
+ {
459
+ "epoch": 23.01,
460
+ "learning_rate": 7.168888888888889e-07,
461
+ "loss": 0.3079,
462
+ "step": 1775
463
+ },
464
+ {
465
+ "epoch": 23.01,
466
+ "learning_rate": 7.113333333333334e-07,
467
+ "loss": 0.284,
468
+ "step": 1800
469
+ },
470
+ {
471
+ "epoch": 24.0,
472
+ "learning_rate": 7.057777777777777e-07,
473
+ "loss": 0.2646,
474
+ "step": 1825
475
+ },
476
+ {
477
+ "epoch": 24.01,
478
+ "learning_rate": 7.002222222222222e-07,
479
+ "loss": 0.2982,
480
+ "step": 1850
481
+ },
482
+ {
483
+ "epoch": 24.01,
484
+ "learning_rate": 6.946666666666666e-07,
485
+ "loss": 0.2794,
486
+ "step": 1875
487
+ },
488
+ {
489
+ "epoch": 24.02,
490
+ "learning_rate": 6.891111111111111e-07,
491
+ "loss": 0.26,
492
+ "step": 1900
493
+ },
494
+ {
495
+ "epoch": 25.0,
496
+ "learning_rate": 6.835555555555555e-07,
497
+ "loss": 0.294,
498
+ "step": 1925
499
+ },
500
+ {
501
+ "epoch": 25.01,
502
+ "learning_rate": 6.78e-07,
503
+ "loss": 0.2774,
504
+ "step": 1950
505
+ },
506
+ {
507
+ "epoch": 25.02,
508
+ "learning_rate": 6.724444444444444e-07,
509
+ "loss": 0.2473,
510
+ "step": 1975
511
+ },
512
+ {
513
+ "epoch": 26.0,
514
+ "learning_rate": 6.668888888888888e-07,
515
+ "loss": 0.2848,
516
+ "step": 2000
517
+ },
518
+ {
519
+ "epoch": 26.0,
520
+ "eval_loss": 0.47389376163482666,
521
+ "eval_runtime": 332.7237,
522
+ "eval_samples_per_second": 9.338,
523
+ "eval_steps_per_second": 0.295,
524
+ "eval_wer": 30.103734006804284,
525
+ "step": 2000
526
+ },
527
+ {
528
+ "epoch": 26.01,
529
+ "learning_rate": 6.613333333333333e-07,
530
+ "loss": 0.2707,
531
+ "step": 2025
532
+ },
533
+ {
534
+ "epoch": 26.01,
535
+ "learning_rate": 6.557777777777777e-07,
536
+ "loss": 0.2461,
537
+ "step": 2050
538
+ },
539
+ {
540
+ "epoch": 27.0,
541
+ "learning_rate": 6.502222222222222e-07,
542
+ "loss": 0.2753,
543
+ "step": 2075
544
+ },
545
+ {
546
+ "epoch": 27.01,
547
+ "learning_rate": 6.446666666666666e-07,
548
+ "loss": 0.2666,
549
+ "step": 2100
550
+ },
551
+ {
552
+ "epoch": 27.01,
553
+ "learning_rate": 6.39111111111111e-07,
554
+ "loss": 0.2428,
555
+ "step": 2125
556
+ },
557
+ {
558
+ "epoch": 28.0,
559
+ "learning_rate": 6.335555555555556e-07,
560
+ "loss": 0.2672,
561
+ "step": 2150
562
+ },
563
+ {
564
+ "epoch": 28.01,
565
+ "learning_rate": 6.28e-07,
566
+ "loss": 0.2628,
567
+ "step": 2175
568
+ },
569
+ {
570
+ "epoch": 28.01,
571
+ "learning_rate": 6.224444444444445e-07,
572
+ "loss": 0.2378,
573
+ "step": 2200
574
+ },
575
+ {
576
+ "epoch": 29.0,
577
+ "learning_rate": 6.168888888888889e-07,
578
+ "loss": 0.2615,
579
+ "step": 2225
580
+ },
581
+ {
582
+ "epoch": 29.01,
583
+ "learning_rate": 6.113333333333333e-07,
584
+ "loss": 0.2566,
585
+ "step": 2250
586
+ },
587
+ {
588
+ "epoch": 29.01,
589
+ "learning_rate": 6.057777777777777e-07,
590
+ "loss": 0.2334,
591
+ "step": 2275
592
+ },
593
+ {
594
+ "epoch": 30.0,
595
+ "learning_rate": 6.002222222222222e-07,
596
+ "loss": 0.2507,
597
+ "step": 2300
598
+ },
599
+ {
600
+ "epoch": 30.01,
601
+ "learning_rate": 5.946666666666667e-07,
602
+ "loss": 0.2576,
603
+ "step": 2325
604
+ },
605
+ {
606
+ "epoch": 30.01,
607
+ "learning_rate": 5.891111111111111e-07,
608
+ "loss": 0.2266,
609
+ "step": 2350
610
+ },
611
+ {
612
+ "epoch": 31.0,
613
+ "learning_rate": 5.835555555555556e-07,
614
+ "loss": 0.2519,
615
+ "step": 2375
616
+ },
617
+ {
618
+ "epoch": 31.01,
619
+ "learning_rate": 5.779999999999999e-07,
620
+ "loss": 0.2478,
621
+ "step": 2400
622
+ },
623
+ {
624
+ "epoch": 31.01,
625
+ "learning_rate": 5.724444444444444e-07,
626
+ "loss": 0.2257,
627
+ "step": 2425
628
+ },
629
+ {
630
+ "epoch": 32.0,
631
+ "learning_rate": 5.668888888888888e-07,
632
+ "loss": 0.2446,
633
+ "step": 2450
634
+ },
635
+ {
636
+ "epoch": 32.01,
637
+ "learning_rate": 5.613333333333333e-07,
638
+ "loss": 0.2452,
639
+ "step": 2475
640
+ },
641
+ {
642
+ "epoch": 32.01,
643
+ "learning_rate": 5.557777777777778e-07,
644
+ "loss": 0.2201,
645
+ "step": 2500
646
+ },
647
+ {
648
+ "epoch": 32.01,
649
+ "eval_loss": 0.46750521659851074,
650
+ "eval_runtime": 335.6468,
651
+ "eval_samples_per_second": 9.257,
652
+ "eval_steps_per_second": 0.292,
653
+ "eval_wer": 29.485921813362275,
654
+ "step": 2500
655
+ },
656
+ {
657
+ "epoch": 33.0,
658
+ "learning_rate": 5.502222222222221e-07,
659
+ "loss": 0.2385,
660
+ "step": 2525
661
+ },
662
+ {
663
+ "epoch": 33.01,
664
+ "learning_rate": 5.446666666666666e-07,
665
+ "loss": 0.2409,
666
+ "step": 2550
667
+ },
668
+ {
669
+ "epoch": 33.01,
670
+ "learning_rate": 5.39111111111111e-07,
671
+ "loss": 0.2203,
672
+ "step": 2575
673
+ },
674
+ {
675
+ "epoch": 34.0,
676
+ "learning_rate": 5.335555555555556e-07,
677
+ "loss": 0.2328,
678
+ "step": 2600
679
+ },
680
+ {
681
+ "epoch": 34.01,
682
+ "learning_rate": 5.28e-07,
683
+ "loss": 0.24,
684
+ "step": 2625
685
+ },
686
+ {
687
+ "epoch": 34.01,
688
+ "learning_rate": 5.224444444444445e-07,
689
+ "loss": 0.2136,
690
+ "step": 2650
691
+ },
692
+ {
693
+ "epoch": 35.0,
694
+ "learning_rate": 5.168888888888888e-07,
695
+ "loss": 0.2286,
696
+ "step": 2675
697
+ },
698
+ {
699
+ "epoch": 35.01,
700
+ "learning_rate": 5.113333333333333e-07,
701
+ "loss": 0.2334,
702
+ "step": 2700
703
+ },
704
+ {
705
+ "epoch": 35.01,
706
+ "learning_rate": 5.057777777777778e-07,
707
+ "loss": 0.2142,
708
+ "step": 2725
709
+ },
710
+ {
711
+ "epoch": 36.0,
712
+ "learning_rate": 5.002222222222222e-07,
713
+ "loss": 0.2218,
714
+ "step": 2750
715
+ },
716
+ {
717
+ "epoch": 36.01,
718
+ "learning_rate": 4.946666666666666e-07,
719
+ "loss": 0.2342,
720
+ "step": 2775
721
+ },
722
+ {
723
+ "epoch": 36.01,
724
+ "learning_rate": 4.891111111111111e-07,
725
+ "loss": 0.2074,
726
+ "step": 2800
727
+ },
728
+ {
729
+ "epoch": 37.0,
730
+ "learning_rate": 4.835555555555555e-07,
731
+ "loss": 0.2169,
732
+ "step": 2825
733
+ },
734
+ {
735
+ "epoch": 37.01,
736
+ "learning_rate": 4.779999999999999e-07,
737
+ "loss": 0.2273,
738
+ "step": 2850
739
+ },
740
+ {
741
+ "epoch": 37.01,
742
+ "learning_rate": 4.724444444444444e-07,
743
+ "loss": 0.2087,
744
+ "step": 2875
745
+ },
746
+ {
747
+ "epoch": 38.0,
748
+ "learning_rate": 4.668888888888889e-07,
749
+ "loss": 0.2139,
750
+ "step": 2900
751
+ },
752
+ {
753
+ "epoch": 38.01,
754
+ "learning_rate": 4.613333333333333e-07,
755
+ "loss": 0.2256,
756
+ "step": 2925
757
+ },
758
+ {
759
+ "epoch": 38.01,
760
+ "learning_rate": 4.5577777777777774e-07,
761
+ "loss": 0.2044,
762
+ "step": 2950
763
+ },
764
+ {
765
+ "epoch": 39.0,
766
+ "learning_rate": 4.5022222222222225e-07,
767
+ "loss": 0.2059,
768
+ "step": 2975
769
+ },
770
+ {
771
+ "epoch": 39.01,
772
+ "learning_rate": 4.4466666666666665e-07,
773
+ "loss": 0.2257,
774
+ "step": 3000
775
+ },
776
+ {
777
+ "epoch": 39.01,
778
+ "eval_loss": 0.46374234557151794,
779
+ "eval_runtime": 331.7281,
780
+ "eval_samples_per_second": 9.366,
781
+ "eval_steps_per_second": 0.295,
782
+ "eval_wer": 28.99334182129365,
783
+ "step": 3000
784
+ },
785
+ {
786
+ "epoch": 39.01,
787
+ "learning_rate": 4.391111111111111e-07,
788
+ "loss": 0.2031,
789
+ "step": 3025
790
+ },
791
+ {
792
+ "epoch": 40.0,
793
+ "learning_rate": 4.3355555555555556e-07,
794
+ "loss": 0.205,
795
+ "step": 3050
796
+ },
797
+ {
798
+ "epoch": 40.01,
799
+ "learning_rate": 4.2799999999999997e-07,
800
+ "loss": 0.2182,
801
+ "step": 3075
802
+ },
803
+ {
804
+ "epoch": 40.01,
805
+ "learning_rate": 4.224444444444444e-07,
806
+ "loss": 0.2037,
807
+ "step": 3100
808
+ },
809
+ {
810
+ "epoch": 41.0,
811
+ "learning_rate": 4.1688888888888883e-07,
812
+ "loss": 0.1995,
813
+ "step": 3125
814
+ },
815
+ {
816
+ "epoch": 41.01,
817
+ "learning_rate": 4.113333333333333e-07,
818
+ "loss": 0.2205,
819
+ "step": 3150
820
+ },
821
+ {
822
+ "epoch": 41.01,
823
+ "learning_rate": 4.057777777777778e-07,
824
+ "loss": 0.1982,
825
+ "step": 3175
826
+ },
827
+ {
828
+ "epoch": 42.0,
829
+ "learning_rate": 4.002222222222222e-07,
830
+ "loss": 0.1954,
831
+ "step": 3200
832
+ },
833
+ {
834
+ "epoch": 42.01,
835
+ "learning_rate": 3.9466666666666665e-07,
836
+ "loss": 0.2145,
837
+ "step": 3225
838
+ },
839
+ {
840
+ "epoch": 42.01,
841
+ "learning_rate": 3.891111111111111e-07,
842
+ "loss": 0.1975,
843
+ "step": 3250
844
+ },
845
+ {
846
+ "epoch": 43.0,
847
+ "learning_rate": 3.835555555555555e-07,
848
+ "loss": 0.1911,
849
+ "step": 3275
850
+ },
851
+ {
852
+ "epoch": 43.01,
853
+ "learning_rate": 3.7799999999999997e-07,
854
+ "loss": 0.2149,
855
+ "step": 3300
856
+ },
857
+ {
858
+ "epoch": 43.01,
859
+ "learning_rate": 3.7244444444444443e-07,
860
+ "loss": 0.1959,
861
+ "step": 3325
862
+ },
863
+ {
864
+ "epoch": 44.0,
865
+ "learning_rate": 3.668888888888889e-07,
866
+ "loss": 0.1898,
867
+ "step": 3350
868
+ },
869
+ {
870
+ "epoch": 44.01,
871
+ "learning_rate": 3.6133333333333334e-07,
872
+ "loss": 0.2104,
873
+ "step": 3375
874
+ },
875
+ {
876
+ "epoch": 44.01,
877
+ "learning_rate": 3.557777777777778e-07,
878
+ "loss": 0.1947,
879
+ "step": 3400
880
+ },
881
+ {
882
+ "epoch": 45.0,
883
+ "learning_rate": 3.502222222222222e-07,
884
+ "loss": 0.1863,
885
+ "step": 3425
886
+ },
887
+ {
888
+ "epoch": 45.01,
889
+ "learning_rate": 3.4466666666666666e-07,
890
+ "loss": 0.2093,
891
+ "step": 3450
892
+ },
893
+ {
894
+ "epoch": 45.01,
895
+ "learning_rate": 3.3911111111111106e-07,
896
+ "loss": 0.1914,
897
+ "step": 3475
898
+ },
899
+ {
900
+ "epoch": 46.0,
901
+ "learning_rate": 3.335555555555555e-07,
902
+ "loss": 0.1837,
903
+ "step": 3500
904
+ },
905
+ {
906
+ "epoch": 46.0,
907
+ "eval_loss": 0.46572384238243103,
908
+ "eval_runtime": 334.5411,
909
+ "eval_samples_per_second": 9.287,
910
+ "eval_steps_per_second": 0.293,
911
+ "eval_wer": 28.914028093757178,
912
+ "step": 3500
913
+ },
914
+ {
915
+ "epoch": 46.01,
916
+ "learning_rate": 3.28e-07,
917
+ "loss": 0.2041,
918
+ "step": 3525
919
+ },
920
+ {
921
+ "epoch": 46.01,
922
+ "learning_rate": 3.2244444444444443e-07,
923
+ "loss": 0.1936,
924
+ "step": 3550
925
+ },
926
+ {
927
+ "epoch": 47.0,
928
+ "learning_rate": 3.168888888888889e-07,
929
+ "loss": 0.1808,
930
+ "step": 3575
931
+ },
932
+ {
933
+ "epoch": 47.01,
934
+ "learning_rate": 3.1133333333333334e-07,
935
+ "loss": 0.2032,
936
+ "step": 3600
937
+ },
938
+ {
939
+ "epoch": 47.01,
940
+ "learning_rate": 3.0577777777777775e-07,
941
+ "loss": 0.1924,
942
+ "step": 3625
943
+ },
944
+ {
945
+ "epoch": 48.0,
946
+ "learning_rate": 3.002222222222222e-07,
947
+ "loss": 0.1767,
948
+ "step": 3650
949
+ },
950
+ {
951
+ "epoch": 48.01,
952
+ "learning_rate": 2.9466666666666666e-07,
953
+ "loss": 0.2025,
954
+ "step": 3675
955
+ },
956
+ {
957
+ "epoch": 48.01,
958
+ "learning_rate": 2.8911111111111106e-07,
959
+ "loss": 0.1902,
960
+ "step": 3700
961
+ },
962
+ {
963
+ "epoch": 49.0,
964
+ "learning_rate": 2.8355555555555557e-07,
965
+ "loss": 0.174,
966
+ "step": 3725
967
+ },
968
+ {
969
+ "epoch": 49.01,
970
+ "learning_rate": 2.7800000000000003e-07,
971
+ "loss": 0.2009,
972
+ "step": 3750
973
+ },
974
+ {
975
+ "epoch": 49.01,
976
+ "learning_rate": 2.7244444444444443e-07,
977
+ "loss": 0.1884,
978
+ "step": 3775
979
+ },
980
+ {
981
+ "epoch": 49.02,
982
+ "learning_rate": 2.668888888888889e-07,
983
+ "loss": 0.1728,
984
+ "step": 3800
985
+ },
986
+ {
987
+ "epoch": 50.01,
988
+ "learning_rate": 2.613333333333333e-07,
989
+ "loss": 0.1984,
990
+ "step": 3825
991
+ },
992
+ {
993
+ "epoch": 50.01,
994
+ "learning_rate": 2.5577777777777775e-07,
995
+ "loss": 0.1884,
996
+ "step": 3850
997
+ },
998
+ {
999
+ "epoch": 50.02,
1000
+ "learning_rate": 2.502222222222222e-07,
1001
+ "loss": 0.1717,
1002
+ "step": 3875
1003
+ },
1004
+ {
1005
+ "epoch": 51.0,
1006
+ "learning_rate": 2.4466666666666666e-07,
1007
+ "loss": 0.1927,
1008
+ "step": 3900
1009
+ },
1010
+ {
1011
+ "epoch": 51.01,
1012
+ "learning_rate": 2.391111111111111e-07,
1013
+ "loss": 0.1909,
1014
+ "step": 3925
1015
+ },
1016
+ {
1017
+ "epoch": 51.01,
1018
+ "learning_rate": 2.3355555555555552e-07,
1019
+ "loss": 0.1698,
1020
+ "step": 3950
1021
+ },
1022
+ {
1023
+ "epoch": 52.0,
1024
+ "learning_rate": 2.28e-07,
1025
+ "loss": 0.1901,
1026
+ "step": 3975
1027
+ },
1028
+ {
1029
+ "epoch": 52.01,
1030
+ "learning_rate": 2.2244444444444443e-07,
1031
+ "loss": 0.1897,
1032
+ "step": 4000
1033
+ },
1034
+ {
1035
+ "epoch": 52.01,
1036
+ "eval_loss": 0.4658238887786865,
1037
+ "eval_runtime": 335.2394,
1038
+ "eval_samples_per_second": 9.268,
1039
+ "eval_steps_per_second": 0.292,
1040
+ "eval_wer": 28.74496462190311,
1041
+ "step": 4000
1042
+ },
1043
+ {
1044
+ "epoch": 52.01,
1045
+ "learning_rate": 2.1688888888888886e-07,
1046
+ "loss": 0.1677,
1047
+ "step": 4025
1048
+ },
1049
+ {
1050
+ "epoch": 53.0,
1051
+ "learning_rate": 2.1133333333333335e-07,
1052
+ "loss": 0.1899,
1053
+ "step": 4050
1054
+ },
1055
+ {
1056
+ "epoch": 53.01,
1057
+ "learning_rate": 2.0577777777777778e-07,
1058
+ "loss": 0.1883,
1059
+ "step": 4075
1060
+ },
1061
+ {
1062
+ "epoch": 53.01,
1063
+ "learning_rate": 2.002222222222222e-07,
1064
+ "loss": 0.1679,
1065
+ "step": 4100
1066
+ },
1067
+ {
1068
+ "epoch": 54.0,
1069
+ "learning_rate": 1.9466666666666664e-07,
1070
+ "loss": 0.1869,
1071
+ "step": 4125
1072
+ },
1073
+ {
1074
+ "epoch": 54.01,
1075
+ "learning_rate": 1.8911111111111112e-07,
1076
+ "loss": 0.1881,
1077
+ "step": 4150
1078
+ },
1079
+ {
1080
+ "epoch": 54.01,
1081
+ "learning_rate": 1.8355555555555555e-07,
1082
+ "loss": 0.1673,
1083
+ "step": 4175
1084
+ },
1085
+ {
1086
+ "epoch": 55.0,
1087
+ "learning_rate": 1.7799999999999998e-07,
1088
+ "loss": 0.1847,
1089
+ "step": 4200
1090
+ },
1091
+ {
1092
+ "epoch": 55.01,
1093
+ "learning_rate": 1.7244444444444444e-07,
1094
+ "loss": 0.1876,
1095
+ "step": 4225
1096
+ },
1097
+ {
1098
+ "epoch": 55.01,
1099
+ "learning_rate": 1.668888888888889e-07,
1100
+ "loss": 0.1667,
1101
+ "step": 4250
1102
+ },
1103
+ {
1104
+ "epoch": 56.0,
1105
+ "learning_rate": 1.6133333333333332e-07,
1106
+ "loss": 0.1807,
1107
+ "step": 4275
1108
+ },
1109
+ {
1110
+ "epoch": 56.01,
1111
+ "learning_rate": 1.5577777777777775e-07,
1112
+ "loss": 0.1851,
1113
+ "step": 4300
1114
+ },
1115
+ {
1116
+ "epoch": 56.01,
1117
+ "learning_rate": 1.5022222222222224e-07,
1118
+ "loss": 0.1705,
1119
+ "step": 4325
1120
+ },
1121
+ {
1122
+ "epoch": 57.0,
1123
+ "learning_rate": 1.4466666666666667e-07,
1124
+ "loss": 0.1773,
1125
+ "step": 4350
1126
+ },
1127
+ {
1128
+ "epoch": 57.01,
1129
+ "learning_rate": 1.391111111111111e-07,
1130
+ "loss": 0.1905,
1131
+ "step": 4375
1132
+ },
1133
+ {
1134
+ "epoch": 57.01,
1135
+ "learning_rate": 1.3355555555555555e-07,
1136
+ "loss": 0.1629,
1137
+ "step": 4400
1138
+ },
1139
+ {
1140
+ "epoch": 58.0,
1141
+ "learning_rate": 1.28e-07,
1142
+ "loss": 0.1774,
1143
+ "step": 4425
1144
+ },
1145
+ {
1146
+ "epoch": 58.01,
1147
+ "learning_rate": 1.2244444444444444e-07,
1148
+ "loss": 0.1853,
1149
+ "step": 4450
1150
+ },
1151
+ {
1152
+ "epoch": 58.01,
1153
+ "learning_rate": 1.1688888888888888e-07,
1154
+ "loss": 0.1659,
1155
+ "step": 4475
1156
+ },
1157
+ {
1158
+ "epoch": 59.0,
1159
+ "learning_rate": 1.1133333333333334e-07,
1160
+ "loss": 0.1764,
1161
+ "step": 4500
1162
+ },
1163
+ {
1164
+ "epoch": 59.0,
1165
+ "eval_loss": 0.4676055610179901,
1166
+ "eval_runtime": 335.034,
1167
+ "eval_samples_per_second": 9.274,
1168
+ "eval_steps_per_second": 0.293,
1169
+ "eval_wer": 28.717830978272215,
1170
+ "step": 4500
1171
+ },
1172
+ {
1173
+ "epoch": 59.01,
1174
+ "learning_rate": 1.0577777777777777e-07,
1175
+ "loss": 0.1844,
1176
+ "step": 4525
1177
+ },
1178
+ {
1179
+ "epoch": 59.01,
1180
+ "learning_rate": 1.0022222222222222e-07,
1181
+ "loss": 0.1682,
1182
+ "step": 4550
1183
+ },
1184
+ {
1185
+ "epoch": 60.0,
1186
+ "learning_rate": 9.466666666666665e-08,
1187
+ "loss": 0.174,
1188
+ "step": 4575
1189
+ },
1190
+ {
1191
+ "epoch": 60.01,
1192
+ "learning_rate": 8.911111111111111e-08,
1193
+ "loss": 0.1835,
1194
+ "step": 4600
1195
+ },
1196
+ {
1197
+ "epoch": 60.01,
1198
+ "learning_rate": 8.355555555555554e-08,
1199
+ "loss": 0.166,
1200
+ "step": 4625
1201
+ },
1202
+ {
1203
+ "epoch": 61.0,
1204
+ "learning_rate": 7.8e-08,
1205
+ "loss": 0.1747,
1206
+ "step": 4650
1207
+ },
1208
+ {
1209
+ "epoch": 61.01,
1210
+ "learning_rate": 7.244444444444445e-08,
1211
+ "loss": 0.1837,
1212
+ "step": 4675
1213
+ },
1214
+ {
1215
+ "epoch": 61.01,
1216
+ "learning_rate": 6.688888888888888e-08,
1217
+ "loss": 0.1673,
1218
+ "step": 4700
1219
+ },
1220
+ {
1221
+ "epoch": 62.0,
1222
+ "learning_rate": 6.133333333333333e-08,
1223
+ "loss": 0.1741,
1224
+ "step": 4725
1225
+ },
1226
+ {
1227
+ "epoch": 62.01,
1228
+ "learning_rate": 5.5777777777777777e-08,
1229
+ "loss": 0.1816,
1230
+ "step": 4750
1231
+ },
1232
+ {
1233
+ "epoch": 62.01,
1234
+ "learning_rate": 5.022222222222222e-08,
1235
+ "loss": 0.166,
1236
+ "step": 4775
1237
+ },
1238
+ {
1239
+ "epoch": 63.0,
1240
+ "learning_rate": 4.466666666666666e-08,
1241
+ "loss": 0.1725,
1242
+ "step": 4800
1243
+ },
1244
+ {
1245
+ "epoch": 63.01,
1246
+ "learning_rate": 3.9111111111111106e-08,
1247
+ "loss": 0.1849,
1248
+ "step": 4825
1249
+ },
1250
+ {
1251
+ "epoch": 63.01,
1252
+ "learning_rate": 3.355555555555555e-08,
1253
+ "loss": 0.1655,
1254
+ "step": 4850
1255
+ },
1256
+ {
1257
+ "epoch": 64.0,
1258
+ "learning_rate": 2.8e-08,
1259
+ "loss": 0.1708,
1260
+ "step": 4875
1261
+ },
1262
+ {
1263
+ "epoch": 64.01,
1264
+ "learning_rate": 2.2444444444444442e-08,
1265
+ "loss": 0.1835,
1266
+ "step": 4900
1267
+ },
1268
+ {
1269
+ "epoch": 64.01,
1270
+ "learning_rate": 1.6888888888888886e-08,
1271
+ "loss": 0.1664,
1272
+ "step": 4925
1273
+ },
1274
+ {
1275
+ "epoch": 65.0,
1276
+ "learning_rate": 1.1333333333333334e-08,
1277
+ "loss": 0.1673,
1278
+ "step": 4950
1279
+ },
1280
+ {
1281
+ "epoch": 65.01,
1282
+ "learning_rate": 5.777777777777777e-09,
1283
+ "loss": 0.1839,
1284
+ "step": 4975
1285
+ },
1286
+ {
1287
+ "epoch": 65.01,
1288
+ "learning_rate": 2.2222222222222221e-10,
1289
+ "loss": 0.1681,
1290
+ "step": 5000
1291
+ },
1292
+ {
1293
+ "epoch": 65.01,
1294
+ "eval_loss": 0.46787169575691223,
1295
+ "eval_runtime": 334.6695,
1296
+ "eval_samples_per_second": 9.284,
1297
+ "eval_steps_per_second": 0.293,
1298
+ "eval_wer": 28.648953267516852,
1299
+ "step": 5000
1300
+ },
1301
+ {
1302
+ "epoch": 65.01,
1303
+ "step": 5000,
1304
+ "total_flos": 2.06793125756928e+19,
1305
+ "train_loss": 0.34305178251266477,
1306
+ "train_runtime": 22591.6172,
1307
+ "train_samples_per_second": 14.165,
1308
+ "train_steps_per_second": 0.221
1309
+ }
1310
+ ],
1311
+ "logging_steps": 25,
1312
+ "max_steps": 5000,
1313
+ "num_input_tokens_seen": 0,
1314
+ "num_train_epochs": 9223372036854775807,
1315
+ "save_steps": 500,
1316
+ "total_flos": 2.06793125756928e+19,
1317
+ "train_batch_size": 32,
1318
+ "trial_name": null,
1319
+ "trial_params": null
1320
+ }