Chenxi-Chelsea-Liu commited on
Commit
110e8e5
1 Parent(s): 50792b6

Model save

Browse files
Files changed (3) hide show
  1. README.md +121 -0
  2. generation_config.json +264 -0
  3. model.safetensors +1 -1
README.md ADDED
@@ -0,0 +1,121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: openai/whisper-small
4
+ tags:
5
+ - generated_from_trainer
6
+ metrics:
7
+ - wer
8
+ model-index:
9
+ - name: whisper-small-noisy-hindi-10dB
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # whisper-small-noisy-hindi-10dB
17
+
18
+ This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the None dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 0.7442
21
+ - Wer: 41.8554
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 1e-05
41
+ - train_batch_size: 64
42
+ - eval_batch_size: 32
43
+ - seed: 42
44
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
+ - lr_scheduler_type: linear
46
+ - lr_scheduler_warmup_steps: 500
47
+ - training_steps: 3000
48
+ - mixed_precision_training: Native AMP
49
+
50
+ ### Training results
51
+
52
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
53
+ |:-------------:|:-----:|:----:|:---------------:|:-------:|
54
+ | 1.6146 | 0.61 | 50 | 1.3244 | 85.2585 |
55
+ | 0.8209 | 1.22 | 100 | 0.7607 | 55.4556 |
56
+ | 0.6434 | 1.83 | 150 | 0.6184 | 51.5822 |
57
+ | 0.5053 | 2.44 | 200 | 0.5191 | 46.7404 |
58
+ | 0.409 | 3.05 | 250 | 0.4271 | 41.9938 |
59
+ | 0.265 | 3.66 | 300 | 0.3151 | 39.4778 |
60
+ | 0.1786 | 4.27 | 350 | 0.2965 | 37.3076 |
61
+ | 0.1617 | 4.88 | 400 | 0.2826 | 36.2355 |
62
+ | 0.103 | 5.49 | 450 | 0.2877 | 35.5957 |
63
+ | 0.0907 | 6.1 | 500 | 0.2929 | 35.3450 |
64
+ | 0.0595 | 6.71 | 550 | 0.3032 | 34.8262 |
65
+ | 0.0338 | 7.32 | 600 | 0.3186 | 34.7743 |
66
+ | 0.0365 | 7.93 | 650 | 0.3303 | 34.3853 |
67
+ | 0.021 | 8.54 | 700 | 0.3414 | 34.3420 |
68
+ | 0.0174 | 9.15 | 750 | 0.3561 | 34.1605 |
69
+ | 0.0129 | 9.76 | 800 | 0.3619 | 34.3247 |
70
+ | 0.009 | 10.37 | 850 | 0.3681 | 33.9703 |
71
+ | 0.0082 | 10.98 | 900 | 0.3802 | 34.2469 |
72
+ | 0.006 | 11.59 | 950 | 0.3817 | 33.4083 |
73
+ | 0.0052 | 12.2 | 1000 | 0.4054 | 34.4112 |
74
+ | 0.005 | 12.8 | 1050 | 0.4113 | 34.2123 |
75
+ | 0.0041 | 13.41 | 1100 | 0.4139 | 33.8060 |
76
+ | 0.0043 | 14.02 | 1150 | 0.4161 | 32.9500 |
77
+ | 0.0028 | 14.63 | 1200 | 0.4284 | 33.0192 |
78
+ | 0.0027 | 15.24 | 1250 | 0.4349 | 33.1229 |
79
+ | 0.0027 | 15.85 | 1300 | 0.4253 | 32.7598 |
80
+ | 0.0022 | 16.46 | 1350 | 0.4419 | 33.1143 |
81
+ | 0.0023 | 17.07 | 1400 | 0.4453 | 32.9154 |
82
+ | 0.002 | 17.68 | 1450 | 0.4457 | 32.5696 |
83
+ | 0.0014 | 18.29 | 1500 | 0.4592 | 32.8809 |
84
+ | 0.0014 | 18.9 | 1550 | 0.4757 | 32.8290 |
85
+ | 0.001 | 19.51 | 1600 | 0.4767 | 33.4169 |
86
+ | 0.0008 | 20.12 | 1650 | 0.4876 | 32.4831 |
87
+ | 0.0008 | 20.73 | 1700 | 0.4905 | 32.9760 |
88
+ | 0.0011 | 21.34 | 1750 | 0.4876 | 32.7252 |
89
+ | 0.0007 | 21.95 | 1800 | 0.4992 | 33.0105 |
90
+ | 0.0003 | 22.56 | 1850 | 0.5190 | 32.3102 |
91
+ | 0.0007 | 23.17 | 1900 | 0.5240 | 32.6734 |
92
+ | 0.0005 | 23.78 | 1950 | 0.5315 | 32.8809 |
93
+ | 0.0003 | 24.39 | 2000 | 0.5333 | 32.7771 |
94
+ | 0.0002 | 25.0 | 2050 | 0.5441 | 32.1200 |
95
+ | 0.0001 | 25.61 | 2100 | 0.5626 | 32.4313 |
96
+ | 0.0001 | 26.22 | 2150 | 0.5690 | 32.1546 |
97
+ | 0.0001 | 26.83 | 2200 | 0.5861 | 32.1978 |
98
+ | 0.0001 | 27.44 | 2250 | 0.6071 | 32.0163 |
99
+ | 0.0 | 28.05 | 2300 | 0.6214 | 32.6388 |
100
+ | 0.0001 | 28.66 | 2350 | 0.6333 | 32.7512 |
101
+ | 0.0 | 29.27 | 2400 | 0.6525 | 32.5782 |
102
+ | 0.0 | 29.88 | 2450 | 0.6627 | 32.6647 |
103
+ | 0.0 | 30.49 | 2500 | 0.6759 | 32.5523 |
104
+ | 0.0 | 31.1 | 2550 | 0.6960 | 33.3737 |
105
+ | 0.0 | 31.71 | 2600 | 0.7087 | 34.1864 |
106
+ | 0.0 | 32.32 | 2650 | 0.7228 | 34.4544 |
107
+ | 0.0 | 32.93 | 2700 | 0.7274 | 35.1634 |
108
+ | 0.0 | 33.54 | 2750 | 0.7327 | 35.7254 |
109
+ | 0.0 | 34.15 | 2800 | 0.7369 | 37.0569 |
110
+ | 0.0 | 34.76 | 2850 | 0.7405 | 38.2155 |
111
+ | 0.0 | 35.37 | 2900 | 0.7433 | 40.8871 |
112
+ | 0.0 | 35.98 | 2950 | 0.7441 | 41.6739 |
113
+ | 0.0 | 36.59 | 3000 | 0.7442 | 41.8554 |
114
+
115
+
116
+ ### Framework versions
117
+
118
+ - Transformers 4.37.0.dev0
119
+ - Pytorch 1.12.1
120
+ - Datasets 2.16.1
121
+ - Tokenizers 0.15.0
generation_config.json ADDED
@@ -0,0 +1,264 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 5,
5
+ 3
6
+ ],
7
+ [
8
+ 5,
9
+ 9
10
+ ],
11
+ [
12
+ 8,
13
+ 0
14
+ ],
15
+ [
16
+ 8,
17
+ 4
18
+ ],
19
+ [
20
+ 8,
21
+ 7
22
+ ],
23
+ [
24
+ 8,
25
+ 8
26
+ ],
27
+ [
28
+ 9,
29
+ 0
30
+ ],
31
+ [
32
+ 9,
33
+ 7
34
+ ],
35
+ [
36
+ 9,
37
+ 9
38
+ ],
39
+ [
40
+ 10,
41
+ 5
42
+ ]
43
+ ],
44
+ "begin_suppress_tokens": [
45
+ 220,
46
+ 50257
47
+ ],
48
+ "bos_token_id": 50257,
49
+ "decoder_start_token_id": 50258,
50
+ "eos_token_id": 50257,
51
+ "forced_decoder_ids": [
52
+ [
53
+ 1,
54
+ null
55
+ ],
56
+ [
57
+ 2,
58
+ 50359
59
+ ]
60
+ ],
61
+ "is_multilingual": true,
62
+ "lang_to_id": {
63
+ "<|af|>": 50327,
64
+ "<|am|>": 50334,
65
+ "<|ar|>": 50272,
66
+ "<|as|>": 50350,
67
+ "<|az|>": 50304,
68
+ "<|ba|>": 50355,
69
+ "<|be|>": 50330,
70
+ "<|bg|>": 50292,
71
+ "<|bn|>": 50302,
72
+ "<|bo|>": 50347,
73
+ "<|br|>": 50309,
74
+ "<|bs|>": 50315,
75
+ "<|ca|>": 50270,
76
+ "<|cs|>": 50283,
77
+ "<|cy|>": 50297,
78
+ "<|da|>": 50285,
79
+ "<|de|>": 50261,
80
+ "<|el|>": 50281,
81
+ "<|en|>": 50259,
82
+ "<|es|>": 50262,
83
+ "<|et|>": 50307,
84
+ "<|eu|>": 50310,
85
+ "<|fa|>": 50300,
86
+ "<|fi|>": 50277,
87
+ "<|fo|>": 50338,
88
+ "<|fr|>": 50265,
89
+ "<|gl|>": 50319,
90
+ "<|gu|>": 50333,
91
+ "<|haw|>": 50352,
92
+ "<|ha|>": 50354,
93
+ "<|he|>": 50279,
94
+ "<|hi|>": 50276,
95
+ "<|hr|>": 50291,
96
+ "<|ht|>": 50339,
97
+ "<|hu|>": 50286,
98
+ "<|hy|>": 50312,
99
+ "<|id|>": 50275,
100
+ "<|is|>": 50311,
101
+ "<|it|>": 50274,
102
+ "<|ja|>": 50266,
103
+ "<|jw|>": 50356,
104
+ "<|ka|>": 50329,
105
+ "<|kk|>": 50316,
106
+ "<|km|>": 50323,
107
+ "<|kn|>": 50306,
108
+ "<|ko|>": 50264,
109
+ "<|la|>": 50294,
110
+ "<|lb|>": 50345,
111
+ "<|ln|>": 50353,
112
+ "<|lo|>": 50336,
113
+ "<|lt|>": 50293,
114
+ "<|lv|>": 50301,
115
+ "<|mg|>": 50349,
116
+ "<|mi|>": 50295,
117
+ "<|mk|>": 50308,
118
+ "<|ml|>": 50296,
119
+ "<|mn|>": 50314,
120
+ "<|mr|>": 50320,
121
+ "<|ms|>": 50282,
122
+ "<|mt|>": 50343,
123
+ "<|my|>": 50346,
124
+ "<|ne|>": 50313,
125
+ "<|nl|>": 50271,
126
+ "<|nn|>": 50342,
127
+ "<|no|>": 50288,
128
+ "<|oc|>": 50328,
129
+ "<|pa|>": 50321,
130
+ "<|pl|>": 50269,
131
+ "<|ps|>": 50340,
132
+ "<|pt|>": 50267,
133
+ "<|ro|>": 50284,
134
+ "<|ru|>": 50263,
135
+ "<|sa|>": 50344,
136
+ "<|sd|>": 50332,
137
+ "<|si|>": 50322,
138
+ "<|sk|>": 50298,
139
+ "<|sl|>": 50305,
140
+ "<|sn|>": 50324,
141
+ "<|so|>": 50326,
142
+ "<|sq|>": 50317,
143
+ "<|sr|>": 50303,
144
+ "<|su|>": 50357,
145
+ "<|sv|>": 50273,
146
+ "<|sw|>": 50318,
147
+ "<|ta|>": 50287,
148
+ "<|te|>": 50299,
149
+ "<|tg|>": 50331,
150
+ "<|th|>": 50289,
151
+ "<|tk|>": 50341,
152
+ "<|tl|>": 50348,
153
+ "<|tr|>": 50268,
154
+ "<|tt|>": 50351,
155
+ "<|uk|>": 50280,
156
+ "<|ur|>": 50290,
157
+ "<|uz|>": 50337,
158
+ "<|vi|>": 50278,
159
+ "<|yi|>": 50335,
160
+ "<|yo|>": 50325,
161
+ "<|zh|>": 50260
162
+ },
163
+ "max_initial_timestamp_index": 50,
164
+ "max_length": 448,
165
+ "no_timestamps_token_id": 50363,
166
+ "pad_token_id": 50257,
167
+ "prev_sot_token_id": 50361,
168
+ "return_timestamps": false,
169
+ "suppress_tokens": [
170
+ 1,
171
+ 2,
172
+ 7,
173
+ 8,
174
+ 9,
175
+ 10,
176
+ 14,
177
+ 25,
178
+ 26,
179
+ 27,
180
+ 28,
181
+ 29,
182
+ 31,
183
+ 58,
184
+ 59,
185
+ 60,
186
+ 61,
187
+ 62,
188
+ 63,
189
+ 90,
190
+ 91,
191
+ 92,
192
+ 93,
193
+ 359,
194
+ 503,
195
+ 522,
196
+ 542,
197
+ 873,
198
+ 893,
199
+ 902,
200
+ 918,
201
+ 922,
202
+ 931,
203
+ 1350,
204
+ 1853,
205
+ 1982,
206
+ 2460,
207
+ 2627,
208
+ 3246,
209
+ 3253,
210
+ 3268,
211
+ 3536,
212
+ 3846,
213
+ 3961,
214
+ 4183,
215
+ 4667,
216
+ 6585,
217
+ 6647,
218
+ 7273,
219
+ 9061,
220
+ 9383,
221
+ 10428,
222
+ 10929,
223
+ 11938,
224
+ 12033,
225
+ 12331,
226
+ 12562,
227
+ 13793,
228
+ 14157,
229
+ 14635,
230
+ 15265,
231
+ 15618,
232
+ 16553,
233
+ 16604,
234
+ 18362,
235
+ 18956,
236
+ 20075,
237
+ 21675,
238
+ 22520,
239
+ 26130,
240
+ 26161,
241
+ 26435,
242
+ 28279,
243
+ 29464,
244
+ 31650,
245
+ 32302,
246
+ 32470,
247
+ 36865,
248
+ 42863,
249
+ 47425,
250
+ 49870,
251
+ 50254,
252
+ 50258,
253
+ 50358,
254
+ 50359,
255
+ 50360,
256
+ 50361,
257
+ 50362
258
+ ],
259
+ "task_to_id": {
260
+ "transcribe": 50359,
261
+ "translate": 50358
262
+ },
263
+ "transformers_version": "4.37.0.dev0"
264
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a77f1bc60b881cf1fe1a4a44f4efe6f78c3c03c9af689e73b95be0a136d6103d
3
  size 966995080
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2fea3cfad83c8f46fdeef487bea90b3247d7525a87bf22196428c9b7b68560f
3
  size 966995080