Hugging Face's logo

toshiouchiyama
/
whisper-small-ja

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
Japanese
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
toshiouchiyama commited on
Commit
a495983
•
1 Parent(s): 9d90235

Model save

Browse files
Files changed (3) hide show
  1. pytorch_model.bin +1 -1
  2. runs/Jan31_08-35-01_FSK-1041/events.out.tfevents.1675121736.FSK-1041.11896.0 +2 -2
  3. trainer_state.json +253 -0
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b6421a593428442cf4bba6b9ffb3e594c6799c18d3abe8be71df14dc48fecc0c
3
  size 967095881
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:651aefc63d1fc1ad4db97c9e0b0a5ea94bc3c37abe812db7c20f3ddc30ad2475
3
  size 967095881
runs/Jan31_08-35-01_FSK-1041/events.out.tfevents.1675121736.FSK-1041.11896.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1cae5db5ccca69c83d8cd0340aaabeb1fe63e927cf903448dfc231b1c0801d6
3
- size 8469
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:088d253162d82e6dc9ce9e9ace28e08ce1e6471bf8fcfcc326795197ac8d9a8b
3
+ size 12304
trainer_state.json ADDED
@@ -0,0 +1,253 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 28.62849533954727,
3
+ "best_model_checkpoint": "./whisper-small-ja\\checkpoint-170",
4
+ "epoch": 6.0606060606060606,
5
+ "global_step": 200,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.3,
12
+ "eval_loss": 1.0626804828643799,
13
+ "eval_runtime": 108.3983,
14
+ "eval_samples_per_second": 0.185,
15
+ "eval_steps_per_second": 0.028,
16
+ "eval_wer": 123.16910785619174,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.61,
21
+ "eval_loss": 0.6694555878639221,
22
+ "eval_runtime": 138.6792,
23
+ "eval_samples_per_second": 0.144,
24
+ "eval_steps_per_second": 0.022,
25
+ "eval_wer": 123.43541944074568,
26
+ "step": 20
27
+ },
28
+ {
29
+ "epoch": 0.76,
30
+ "learning_rate": 8.974358974358976e-06,
31
+ "loss": 1.1477,
32
+ "step": 25
33
+ },
34
+ {
35
+ "epoch": 0.91,
36
+ "eval_loss": 0.5782793164253235,
37
+ "eval_runtime": 133.633,
38
+ "eval_samples_per_second": 0.15,
39
+ "eval_steps_per_second": 0.022,
40
+ "eval_wer": 102.6631158455393,
41
+ "step": 30
42
+ },
43
+ {
44
+ "epoch": 1.21,
45
+ "eval_loss": 0.5438617467880249,
46
+ "eval_runtime": 132.1171,
47
+ "eval_samples_per_second": 0.151,
48
+ "eval_steps_per_second": 0.023,
49
+ "eval_wer": 100.0,
50
+ "step": 40
51
+ },
52
+ {
53
+ "epoch": 1.52,
54
+ "learning_rate": 7.692307692307694e-06,
55
+ "loss": 0.5121,
56
+ "step": 50
57
+ },
58
+ {
59
+ "epoch": 1.52,
60
+ "eval_loss": 0.49029508233070374,
61
+ "eval_runtime": 130.7034,
62
+ "eval_samples_per_second": 0.153,
63
+ "eval_steps_per_second": 0.023,
64
+ "eval_wer": 100.0,
65
+ "step": 50
66
+ },
67
+ {
68
+ "epoch": 1.82,
69
+ "eval_loss": 0.45088639855384827,
70
+ "eval_runtime": 130.4494,
71
+ "eval_samples_per_second": 0.153,
72
+ "eval_steps_per_second": 0.023,
73
+ "eval_wer": 100.0,
74
+ "step": 60
75
+ },
76
+ {
77
+ "epoch": 2.12,
78
+ "eval_loss": 0.4300904870033264,
79
+ "eval_runtime": 210.4853,
80
+ "eval_samples_per_second": 0.095,
81
+ "eval_steps_per_second": 0.014,
82
+ "eval_wer": 100.0,
83
+ "step": 70
84
+ },
85
+ {
86
+ "epoch": 2.27,
87
+ "learning_rate": 6.410256410256412e-06,
88
+ "loss": 0.3039,
89
+ "step": 75
90
+ },
91
+ {
92
+ "epoch": 2.42,
93
+ "eval_loss": 0.40350860357284546,
94
+ "eval_runtime": 132.7429,
95
+ "eval_samples_per_second": 0.151,
96
+ "eval_steps_per_second": 0.023,
97
+ "eval_wer": 111.31824234354195,
98
+ "step": 80
99
+ },
100
+ {
101
+ "epoch": 2.73,
102
+ "eval_loss": 0.38837021589279175,
103
+ "eval_runtime": 134.6115,
104
+ "eval_samples_per_second": 0.149,
105
+ "eval_steps_per_second": 0.022,
106
+ "eval_wer": 106.52463382157124,
107
+ "step": 90
108
+ },
109
+ {
110
+ "epoch": 3.03,
111
+ "learning_rate": 5.128205128205128e-06,
112
+ "loss": 0.1777,
113
+ "step": 100
114
+ },
115
+ {
116
+ "epoch": 3.03,
117
+ "eval_loss": 0.3667179048061371,
118
+ "eval_runtime": 135.322,
119
+ "eval_samples_per_second": 0.148,
120
+ "eval_steps_per_second": 0.022,
121
+ "eval_wer": 117.31025299600533,
122
+ "step": 100
123
+ },
124
+ {
125
+ "epoch": 3.33,
126
+ "eval_loss": 0.3598465919494629,
127
+ "eval_runtime": 130.8128,
128
+ "eval_samples_per_second": 0.153,
129
+ "eval_steps_per_second": 0.023,
130
+ "eval_wer": 114.51398135818908,
131
+ "step": 110
132
+ },
133
+ {
134
+ "epoch": 3.64,
135
+ "eval_loss": 0.36537379026412964,
136
+ "eval_runtime": 137.4183,
137
+ "eval_samples_per_second": 0.146,
138
+ "eval_steps_per_second": 0.022,
139
+ "eval_wer": 107.0572569906791,
140
+ "step": 120
141
+ },
142
+ {
143
+ "epoch": 3.79,
144
+ "learning_rate": 3.846153846153847e-06,
145
+ "loss": 0.0953,
146
+ "step": 125
147
+ },
148
+ {
149
+ "epoch": 3.94,
150
+ "eval_loss": 0.35574963688850403,
151
+ "eval_runtime": 129.8707,
152
+ "eval_samples_per_second": 0.154,
153
+ "eval_steps_per_second": 0.023,
154
+ "eval_wer": 85.88548601864181,
155
+ "step": 130
156
+ },
157
+ {
158
+ "epoch": 4.24,
159
+ "eval_loss": 0.3518906533718109,
160
+ "eval_runtime": 132.4888,
161
+ "eval_samples_per_second": 0.151,
162
+ "eval_steps_per_second": 0.023,
163
+ "eval_wer": 54.32756324900133,
164
+ "step": 140
165
+ },
166
+ {
167
+ "epoch": 4.55,
168
+ "learning_rate": 2.564102564102564e-06,
169
+ "loss": 0.0589,
170
+ "step": 150
171
+ },
172
+ {
173
+ "epoch": 4.55,
174
+ "eval_loss": 0.36332806944847107,
175
+ "eval_runtime": 134.4235,
176
+ "eval_samples_per_second": 0.149,
177
+ "eval_steps_per_second": 0.022,
178
+ "eval_wer": 43.67509986684421,
179
+ "step": 150
180
+ },
181
+ {
182
+ "epoch": 4.85,
183
+ "eval_loss": 0.3542396128177643,
184
+ "eval_runtime": 181.4645,
185
+ "eval_samples_per_second": 0.11,
186
+ "eval_steps_per_second": 0.017,
187
+ "eval_wer": 43.54194407456725,
188
+ "step": 160
189
+ },
190
+ {
191
+ "epoch": 5.15,
192
+ "eval_loss": 0.35989412665367126,
193
+ "eval_runtime": 113.4241,
194
+ "eval_samples_per_second": 0.176,
195
+ "eval_steps_per_second": 0.026,
196
+ "eval_wer": 28.62849533954727,
197
+ "step": 170
198
+ },
199
+ {
200
+ "epoch": 5.3,
201
+ "learning_rate": 1.282051282051282e-06,
202
+ "loss": 0.0373,
203
+ "step": 175
204
+ },
205
+ {
206
+ "epoch": 5.45,
207
+ "eval_loss": 0.3669775724411011,
208
+ "eval_runtime": 115.4134,
209
+ "eval_samples_per_second": 0.173,
210
+ "eval_steps_per_second": 0.026,
211
+ "eval_wer": 28.62849533954727,
212
+ "step": 180
213
+ },
214
+ {
215
+ "epoch": 5.76,
216
+ "eval_loss": 0.3680533766746521,
217
+ "eval_runtime": 114.6467,
218
+ "eval_samples_per_second": 0.174,
219
+ "eval_steps_per_second": 0.026,
220
+ "eval_wer": 32.35685752330226,
221
+ "step": 190
222
+ },
223
+ {
224
+ "epoch": 6.06,
225
+ "learning_rate": 0.0,
226
+ "loss": 0.0271,
227
+ "step": 200
228
+ },
229
+ {
230
+ "epoch": 6.06,
231
+ "eval_loss": 0.3673397898674011,
232
+ "eval_runtime": 189.6433,
233
+ "eval_samples_per_second": 0.105,
234
+ "eval_steps_per_second": 0.016,
235
+ "eval_wer": 32.35685752330226,
236
+ "step": 200
237
+ },
238
+ {
239
+ "epoch": 6.06,
240
+ "step": 200,
241
+ "total_flos": 4.5481059090432e+17,
242
+ "train_loss": 0.29501178592443467,
243
+ "train_runtime": 14912.6939,
244
+ "train_samples_per_second": 0.107,
245
+ "train_steps_per_second": 0.013
246
+ }
247
+ ],
248
+ "max_steps": 200,
249
+ "num_train_epochs": 7,
250
+ "total_flos": 4.5481059090432e+17,
251
+ "trial_name": null,
252
+ "trial_params": null
253
+ }