whispy commited on
Commit
c48a1a5
1 Parent(s): 339c8c3

upload model checkpoint-900

Browse files
config.json ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "architectures": [
6
+ "WhisperForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "begin_suppress_tokens": [
10
+ 220,
11
+ 50257
12
+ ],
13
+ "bos_token_id": 50257,
14
+ "d_model": 768,
15
+ "decoder_attention_heads": 12,
16
+ "decoder_ffn_dim": 3072,
17
+ "decoder_layerdrop": 0.0,
18
+ "decoder_layers": 12,
19
+ "decoder_start_token_id": 50258,
20
+ "dropout": 0.0,
21
+ "encoder_attention_heads": 12,
22
+ "encoder_ffn_dim": 3072,
23
+ "encoder_layerdrop": 0.0,
24
+ "encoder_layers": 12,
25
+ "eos_token_id": 50257,
26
+ "forced_decoder_ids": null,
27
+ "init_std": 0.02,
28
+ "is_encoder_decoder": true,
29
+ "max_length": 448,
30
+ "max_source_positions": 1500,
31
+ "max_target_positions": 448,
32
+ "model_type": "whisper",
33
+ "num_hidden_layers": 12,
34
+ "num_mel_bins": 80,
35
+ "pad_token_id": 50257,
36
+ "scale_embedding": false,
37
+ "suppress_tokens": [],
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.26.0.dev0",
40
+ "use_cache": true,
41
+ "vocab_size": 51865
42
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40285f874317ca24b1dad2933725bf78b294f2c7a23d08b771adb4393c1e0984
3
+ size 1934158597
preprocessor_config.json ADDED
The diff for this file is too large to render. See raw diff
 
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe2a621d57dc3ddf3b296c0f461236886b8b365fe29b8fec793e4cc42baa1f22
3
+ size 967099139
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2b268c6a53db14a554efa21f4ee380b823db290f3a21b3061e381cc25abbd00
3
+ size 14503
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e058e61ed5c50e9f9a93d8a8b8ace71ae0475455378aaed9bfedf1e0a4403f77
3
+ size 559
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58de88b5045fa6a92be68eec5fa808830f00d77b11cf3eda52c1abf1758c16d2
3
+ size 623
trainer_state.json ADDED
@@ -0,0 +1,313 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 17.391605006569392,
3
+ "best_model_checkpoint": "/content/drive/MyDrive/whisper_it/checkpoint-900",
4
+ "epoch": 0.9433962264150944,
5
+ "global_step": 900,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 4.2000000000000006e-07,
13
+ "loss": 3.6389,
14
+ "step": 25
15
+ },
16
+ {
17
+ "epoch": 0.05,
18
+ "learning_rate": 9.200000000000001e-07,
19
+ "loss": 3.019,
20
+ "step": 50
21
+ },
22
+ {
23
+ "epoch": 0.08,
24
+ "learning_rate": 1.42e-06,
25
+ "loss": 2.2988,
26
+ "step": 75
27
+ },
28
+ {
29
+ "epoch": 0.1,
30
+ "learning_rate": 1.9200000000000003e-06,
31
+ "loss": 1.4744,
32
+ "step": 100
33
+ },
34
+ {
35
+ "epoch": 0.1,
36
+ "eval_loss": 1.1852223873138428,
37
+ "eval_runtime": 687.2583,
38
+ "eval_samples_per_second": 2.183,
39
+ "eval_steps_per_second": 0.274,
40
+ "eval_wer": 117.60597469054699,
41
+ "step": 100
42
+ },
43
+ {
44
+ "epoch": 0.13,
45
+ "learning_rate": 2.42e-06,
46
+ "loss": 1.062,
47
+ "step": 125
48
+ },
49
+ {
50
+ "epoch": 0.16,
51
+ "learning_rate": 2.92e-06,
52
+ "loss": 0.931,
53
+ "step": 150
54
+ },
55
+ {
56
+ "epoch": 0.18,
57
+ "learning_rate": 3.4200000000000007e-06,
58
+ "loss": 0.8408,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 0.21,
63
+ "learning_rate": 3.920000000000001e-06,
64
+ "loss": 0.7241,
65
+ "step": 200
66
+ },
67
+ {
68
+ "epoch": 0.21,
69
+ "eval_loss": 0.7452345490455627,
70
+ "eval_runtime": 856.828,
71
+ "eval_samples_per_second": 1.751,
72
+ "eval_steps_per_second": 0.219,
73
+ "eval_wer": 79.73860728856926,
74
+ "step": 200
75
+ },
76
+ {
77
+ "epoch": 0.24,
78
+ "learning_rate": 4.42e-06,
79
+ "loss": 0.6557,
80
+ "step": 225
81
+ },
82
+ {
83
+ "epoch": 0.26,
84
+ "learning_rate": 4.92e-06,
85
+ "loss": 0.5752,
86
+ "step": 250
87
+ },
88
+ {
89
+ "epoch": 0.29,
90
+ "learning_rate": 5.420000000000001e-06,
91
+ "loss": 0.5109,
92
+ "step": 275
93
+ },
94
+ {
95
+ "epoch": 0.31,
96
+ "learning_rate": 5.92e-06,
97
+ "loss": 0.3321,
98
+ "step": 300
99
+ },
100
+ {
101
+ "epoch": 0.31,
102
+ "eval_loss": 0.32159581780433655,
103
+ "eval_runtime": 644.4221,
104
+ "eval_samples_per_second": 2.328,
105
+ "eval_steps_per_second": 0.292,
106
+ "eval_wer": 21.04971993638061,
107
+ "step": 300
108
+ },
109
+ {
110
+ "epoch": 0.34,
111
+ "learning_rate": 6.42e-06,
112
+ "loss": 0.2987,
113
+ "step": 325
114
+ },
115
+ {
116
+ "epoch": 0.37,
117
+ "learning_rate": 6.92e-06,
118
+ "loss": 0.2577,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 0.39,
123
+ "learning_rate": 7.420000000000001e-06,
124
+ "loss": 0.2928,
125
+ "step": 375
126
+ },
127
+ {
128
+ "epoch": 0.42,
129
+ "learning_rate": 7.92e-06,
130
+ "loss": 0.293,
131
+ "step": 400
132
+ },
133
+ {
134
+ "epoch": 0.42,
135
+ "eval_loss": 0.3030182719230652,
136
+ "eval_runtime": 618.4321,
137
+ "eval_samples_per_second": 2.425,
138
+ "eval_steps_per_second": 0.304,
139
+ "eval_wer": 20.212986653758385,
140
+ "step": 400
141
+ },
142
+ {
143
+ "epoch": 0.45,
144
+ "learning_rate": 8.42e-06,
145
+ "loss": 0.2734,
146
+ "step": 425
147
+ },
148
+ {
149
+ "epoch": 0.47,
150
+ "learning_rate": 8.920000000000001e-06,
151
+ "loss": 0.2863,
152
+ "step": 450
153
+ },
154
+ {
155
+ "epoch": 0.5,
156
+ "learning_rate": 9.42e-06,
157
+ "loss": 0.2847,
158
+ "step": 475
159
+ },
160
+ {
161
+ "epoch": 0.52,
162
+ "learning_rate": 9.920000000000002e-06,
163
+ "loss": 0.2698,
164
+ "step": 500
165
+ },
166
+ {
167
+ "epoch": 0.52,
168
+ "eval_loss": 0.2989492118358612,
169
+ "eval_runtime": 616.9209,
170
+ "eval_samples_per_second": 2.431,
171
+ "eval_steps_per_second": 0.305,
172
+ "eval_wer": 19.76350183251504,
173
+ "step": 500
174
+ },
175
+ {
176
+ "epoch": 0.55,
177
+ "learning_rate": 9.537444933920706e-06,
178
+ "loss": 0.2828,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 0.58,
183
+ "learning_rate": 8.986784140969164e-06,
184
+ "loss": 0.2609,
185
+ "step": 550
186
+ },
187
+ {
188
+ "epoch": 0.6,
189
+ "learning_rate": 8.436123348017622e-06,
190
+ "loss": 0.242,
191
+ "step": 575
192
+ },
193
+ {
194
+ "epoch": 0.63,
195
+ "learning_rate": 7.88546255506608e-06,
196
+ "loss": 0.2453,
197
+ "step": 600
198
+ },
199
+ {
200
+ "epoch": 0.63,
201
+ "eval_loss": 0.28984662890434265,
202
+ "eval_runtime": 632.3877,
203
+ "eval_samples_per_second": 2.372,
204
+ "eval_steps_per_second": 0.297,
205
+ "eval_wer": 19.009750363045434,
206
+ "step": 600
207
+ },
208
+ {
209
+ "epoch": 0.66,
210
+ "learning_rate": 7.334801762114538e-06,
211
+ "loss": 0.2461,
212
+ "step": 625
213
+ },
214
+ {
215
+ "epoch": 0.68,
216
+ "learning_rate": 6.784140969162997e-06,
217
+ "loss": 0.2962,
218
+ "step": 650
219
+ },
220
+ {
221
+ "epoch": 0.71,
222
+ "learning_rate": 6.233480176211454e-06,
223
+ "loss": 0.2454,
224
+ "step": 675
225
+ },
226
+ {
227
+ "epoch": 0.73,
228
+ "learning_rate": 5.682819383259912e-06,
229
+ "loss": 0.2338,
230
+ "step": 700
231
+ },
232
+ {
233
+ "epoch": 0.73,
234
+ "eval_loss": 0.27684447169303894,
235
+ "eval_runtime": 636.9783,
236
+ "eval_samples_per_second": 2.355,
237
+ "eval_steps_per_second": 0.295,
238
+ "eval_wer": 18.705483714819167,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 0.76,
243
+ "learning_rate": 5.1321585903083705e-06,
244
+ "loss": 0.2568,
245
+ "step": 725
246
+ },
247
+ {
248
+ "epoch": 0.79,
249
+ "learning_rate": 4.581497797356828e-06,
250
+ "loss": 0.2672,
251
+ "step": 750
252
+ },
253
+ {
254
+ "epoch": 0.81,
255
+ "learning_rate": 4.030837004405287e-06,
256
+ "loss": 0.2256,
257
+ "step": 775
258
+ },
259
+ {
260
+ "epoch": 0.84,
261
+ "learning_rate": 3.4801762114537445e-06,
262
+ "loss": 0.2402,
263
+ "step": 800
264
+ },
265
+ {
266
+ "epoch": 0.84,
267
+ "eval_loss": 0.26460355520248413,
268
+ "eval_runtime": 635.6494,
269
+ "eval_samples_per_second": 2.36,
270
+ "eval_steps_per_second": 0.296,
271
+ "eval_wer": 18.221423138095567,
272
+ "step": 800
273
+ },
274
+ {
275
+ "epoch": 0.86,
276
+ "learning_rate": 2.9295154185022028e-06,
277
+ "loss": 0.2636,
278
+ "step": 825
279
+ },
280
+ {
281
+ "epoch": 0.89,
282
+ "learning_rate": 2.378854625550661e-06,
283
+ "loss": 0.2354,
284
+ "step": 850
285
+ },
286
+ {
287
+ "epoch": 0.92,
288
+ "learning_rate": 1.828193832599119e-06,
289
+ "loss": 0.2205,
290
+ "step": 875
291
+ },
292
+ {
293
+ "epoch": 0.94,
294
+ "learning_rate": 1.2775330396475772e-06,
295
+ "loss": 0.234,
296
+ "step": 900
297
+ },
298
+ {
299
+ "epoch": 0.94,
300
+ "eval_loss": 0.2581839859485626,
301
+ "eval_runtime": 635.7247,
302
+ "eval_samples_per_second": 2.36,
303
+ "eval_steps_per_second": 0.296,
304
+ "eval_wer": 17.391605006569392,
305
+ "step": 900
306
+ }
307
+ ],
308
+ "max_steps": 954,
309
+ "num_train_epochs": 1,
310
+ "total_flos": 4.155629764608e+18,
311
+ "trial_name": null,
312
+ "trial_params": null
313
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de07bb980ecf2d0d78a66d6e6fa60edc01c0b8b01e18d5480fb9f15a331f3628
3
+ size 3567