zuazo commited on
Commit
fc93831
1 Parent(s): ddb5c15

Model save

Browse files
README.md ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: openai/whisper-base
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - common_voice_16_1
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_1
18
+ type: common_voice_16_1
19
+ config: eu
20
+ split: test
21
+ args: eu
22
+ metrics:
23
+ - name: Wer
24
+ type: wer
25
+ value: 16.431530404877286
26
+ ---
27
+
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_1 dataset.
34
+ It achieves the following results on the evaluation set:
35
+ - Loss: 0.6085
36
+ - Wer: 16.4315
37
+
38
+ ## Model description
39
+
40
+ More information needed
41
+
42
+ ## Intended uses & limitations
43
+
44
+ More information needed
45
+
46
+ ## Training and evaluation data
47
+
48
+ More information needed
49
+
50
+ ## Training procedure
51
+
52
+ ### Training hyperparameters
53
+
54
+ The following hyperparameters were used during training:
55
+ - learning_rate: 2.5e-05
56
+ - train_batch_size: 128
57
+ - eval_batch_size: 64
58
+ - seed: 42
59
+ - gradient_accumulation_steps: 2
60
+ - total_train_batch_size: 256
61
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
+ - lr_scheduler_type: linear
63
+ - lr_scheduler_warmup_steps: 500
64
+ - training_steps: 40000
65
+ - mixed_precision_training: Native AMP
66
+
67
+ ### Training results
68
+
69
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
70
+ |:-------------:|:-----:|:-----:|:---------------:|:-------:|
71
+ | 0.0225 | 10.0 | 1000 | 0.3059 | 19.0812 |
72
+ | 0.0037 | 20.0 | 2000 | 0.3530 | 18.4618 |
73
+ | 0.0012 | 30.0 | 3000 | 0.3724 | 17.9332 |
74
+ | 0.0004 | 40.0 | 4000 | 0.4025 | 17.8951 |
75
+ | 0.0002 | 50.0 | 5000 | 0.4245 | 17.8951 |
76
+ | 0.0001 | 60.0 | 6000 | 0.4459 | 17.9772 |
77
+ | 0.0001 | 70.0 | 7000 | 0.4665 | 18.0163 |
78
+ | 0.0 | 80.0 | 8000 | 0.4882 | 18.1081 |
79
+ | 0.0003 | 90.0 | 9000 | 0.3803 | 16.3807 |
80
+ | 0.0001 | 100.0 | 10000 | 0.4047 | 16.2293 |
81
+ | 0.0001 | 110.0 | 11000 | 0.4207 | 16.2420 |
82
+ | 0.0001 | 120.0 | 12000 | 0.4353 | 16.2879 |
83
+ | 0.0 | 130.0 | 13000 | 0.4502 | 16.3700 |
84
+ | 0.0 | 140.0 | 14000 | 0.4653 | 16.5087 |
85
+ | 0.0 | 150.0 | 15000 | 0.4805 | 16.4393 |
86
+ | 0.0 | 160.0 | 16000 | 0.4964 | 16.4941 |
87
+ | 0.0 | 170.0 | 17000 | 0.5128 | 16.5107 |
88
+ | 0.0 | 180.0 | 18000 | 0.5285 | 16.6377 |
89
+ | 0.0 | 190.0 | 19000 | 0.5457 | 16.6572 |
90
+ | 0.0102 | 200.0 | 20000 | 0.4229 | 18.1902 |
91
+ | 0.0 | 210.0 | 21000 | 0.4498 | 16.2117 |
92
+ | 0.0 | 220.0 | 22000 | 0.4646 | 16.2146 |
93
+ | 0.0 | 230.0 | 23000 | 0.4754 | 16.1961 |
94
+ | 0.0 | 240.0 | 24000 | 0.4853 | 16.1863 |
95
+ | 0.0 | 250.0 | 25000 | 0.4946 | 16.1912 |
96
+ | 0.0 | 260.0 | 26000 | 0.5038 | 16.1765 |
97
+ | 0.0 | 270.0 | 27000 | 0.5133 | 16.2215 |
98
+ | 0.0 | 280.0 | 28000 | 0.5228 | 16.2224 |
99
+ | 0.0 | 290.0 | 29000 | 0.5326 | 16.2557 |
100
+ | 0.0 | 300.0 | 30000 | 0.5427 | 16.2420 |
101
+ | 0.0 | 310.0 | 31000 | 0.5525 | 16.2635 |
102
+ | 0.0 | 320.0 | 32000 | 0.5624 | 16.2957 |
103
+ | 0.0 | 330.0 | 33000 | 0.5706 | 16.3299 |
104
+ | 0.0 | 340.0 | 34000 | 0.5798 | 16.3534 |
105
+ | 0.0 | 350.0 | 35000 | 0.5880 | 16.3495 |
106
+ | 0.0 | 360.0 | 36000 | 0.5948 | 16.3622 |
107
+ | 0.0 | 370.0 | 37000 | 0.6005 | 16.3934 |
108
+ | 0.0 | 380.0 | 38000 | 0.6045 | 16.3876 |
109
+ | 0.0 | 390.0 | 39000 | 0.6074 | 16.4325 |
110
+ | 0.0 | 400.0 | 40000 | 0.6085 | 16.4315 |
111
+
112
+
113
+ ### Framework versions
114
+
115
+ - Transformers 4.37.2
116
+ - Pytorch 2.2.0+cu121
117
+ - Datasets 2.16.1
118
+ - Tokenizers 0.15.1
generation_config.json ADDED
@@ -0,0 +1,256 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 3,
5
+ 1
6
+ ],
7
+ [
8
+ 4,
9
+ 2
10
+ ],
11
+ [
12
+ 4,
13
+ 3
14
+ ],
15
+ [
16
+ 4,
17
+ 7
18
+ ],
19
+ [
20
+ 5,
21
+ 1
22
+ ],
23
+ [
24
+ 5,
25
+ 2
26
+ ],
27
+ [
28
+ 5,
29
+ 4
30
+ ],
31
+ [
32
+ 5,
33
+ 6
34
+ ]
35
+ ],
36
+ "begin_suppress_tokens": [
37
+ 220,
38
+ 50257
39
+ ],
40
+ "bos_token_id": 50257,
41
+ "decoder_start_token_id": 50258,
42
+ "eos_token_id": 50257,
43
+ "forced_decoder_ids": [
44
+ [
45
+ 1,
46
+ null
47
+ ],
48
+ [
49
+ 2,
50
+ 50359
51
+ ]
52
+ ],
53
+ "is_multilingual": true,
54
+ "lang_to_id": {
55
+ "<|af|>": 50327,
56
+ "<|am|>": 50334,
57
+ "<|ar|>": 50272,
58
+ "<|as|>": 50350,
59
+ "<|az|>": 50304,
60
+ "<|ba|>": 50355,
61
+ "<|be|>": 50330,
62
+ "<|bg|>": 50292,
63
+ "<|bn|>": 50302,
64
+ "<|bo|>": 50347,
65
+ "<|br|>": 50309,
66
+ "<|bs|>": 50315,
67
+ "<|ca|>": 50270,
68
+ "<|cs|>": 50283,
69
+ "<|cy|>": 50297,
70
+ "<|da|>": 50285,
71
+ "<|de|>": 50261,
72
+ "<|el|>": 50281,
73
+ "<|en|>": 50259,
74
+ "<|es|>": 50262,
75
+ "<|et|>": 50307,
76
+ "<|eu|>": 50310,
77
+ "<|fa|>": 50300,
78
+ "<|fi|>": 50277,
79
+ "<|fo|>": 50338,
80
+ "<|fr|>": 50265,
81
+ "<|gl|>": 50319,
82
+ "<|gu|>": 50333,
83
+ "<|haw|>": 50352,
84
+ "<|ha|>": 50354,
85
+ "<|he|>": 50279,
86
+ "<|hi|>": 50276,
87
+ "<|hr|>": 50291,
88
+ "<|ht|>": 50339,
89
+ "<|hu|>": 50286,
90
+ "<|hy|>": 50312,
91
+ "<|id|>": 50275,
92
+ "<|is|>": 50311,
93
+ "<|it|>": 50274,
94
+ "<|ja|>": 50266,
95
+ "<|jw|>": 50356,
96
+ "<|ka|>": 50329,
97
+ "<|kk|>": 50316,
98
+ "<|km|>": 50323,
99
+ "<|kn|>": 50306,
100
+ "<|ko|>": 50264,
101
+ "<|la|>": 50294,
102
+ "<|lb|>": 50345,
103
+ "<|ln|>": 50353,
104
+ "<|lo|>": 50336,
105
+ "<|lt|>": 50293,
106
+ "<|lv|>": 50301,
107
+ "<|mg|>": 50349,
108
+ "<|mi|>": 50295,
109
+ "<|mk|>": 50308,
110
+ "<|ml|>": 50296,
111
+ "<|mn|>": 50314,
112
+ "<|mr|>": 50320,
113
+ "<|ms|>": 50282,
114
+ "<|mt|>": 50343,
115
+ "<|my|>": 50346,
116
+ "<|ne|>": 50313,
117
+ "<|nl|>": 50271,
118
+ "<|nn|>": 50342,
119
+ "<|no|>": 50288,
120
+ "<|oc|>": 50328,
121
+ "<|pa|>": 50321,
122
+ "<|pl|>": 50269,
123
+ "<|ps|>": 50340,
124
+ "<|pt|>": 50267,
125
+ "<|ro|>": 50284,
126
+ "<|ru|>": 50263,
127
+ "<|sa|>": 50344,
128
+ "<|sd|>": 50332,
129
+ "<|si|>": 50322,
130
+ "<|sk|>": 50298,
131
+ "<|sl|>": 50305,
132
+ "<|sn|>": 50324,
133
+ "<|so|>": 50326,
134
+ "<|sq|>": 50317,
135
+ "<|sr|>": 50303,
136
+ "<|su|>": 50357,
137
+ "<|sv|>": 50273,
138
+ "<|sw|>": 50318,
139
+ "<|ta|>": 50287,
140
+ "<|te|>": 50299,
141
+ "<|tg|>": 50331,
142
+ "<|th|>": 50289,
143
+ "<|tk|>": 50341,
144
+ "<|tl|>": 50348,
145
+ "<|tr|>": 50268,
146
+ "<|tt|>": 50351,
147
+ "<|uk|>": 50280,
148
+ "<|ur|>": 50290,
149
+ "<|uz|>": 50337,
150
+ "<|vi|>": 50278,
151
+ "<|yi|>": 50335,
152
+ "<|yo|>": 50325,
153
+ "<|zh|>": 50260
154
+ },
155
+ "max_initial_timestamp_index": 50,
156
+ "max_length": 448,
157
+ "no_timestamps_token_id": 50363,
158
+ "pad_token_id": 50257,
159
+ "prev_sot_token_id": 50361,
160
+ "return_timestamps": false,
161
+ "suppress_tokens": [
162
+ 1,
163
+ 2,
164
+ 7,
165
+ 8,
166
+ 9,
167
+ 10,
168
+ 14,
169
+ 25,
170
+ 26,
171
+ 27,
172
+ 28,
173
+ 29,
174
+ 31,
175
+ 58,
176
+ 59,
177
+ 60,
178
+ 61,
179
+ 62,
180
+ 63,
181
+ 90,
182
+ 91,
183
+ 92,
184
+ 93,
185
+ 359,
186
+ 503,
187
+ 522,
188
+ 542,
189
+ 873,
190
+ 893,
191
+ 902,
192
+ 918,
193
+ 922,
194
+ 931,
195
+ 1350,
196
+ 1853,
197
+ 1982,
198
+ 2460,
199
+ 2627,
200
+ 3246,
201
+ 3253,
202
+ 3268,
203
+ 3536,
204
+ 3846,
205
+ 3961,
206
+ 4183,
207
+ 4667,
208
+ 6585,
209
+ 6647,
210
+ 7273,
211
+ 9061,
212
+ 9383,
213
+ 10428,
214
+ 10929,
215
+ 11938,
216
+ 12033,
217
+ 12331,
218
+ 12562,
219
+ 13793,
220
+ 14157,
221
+ 14635,
222
+ 15265,
223
+ 15618,
224
+ 16553,
225
+ 16604,
226
+ 18362,
227
+ 18956,
228
+ 20075,
229
+ 21675,
230
+ 22520,
231
+ 26130,
232
+ 26161,
233
+ 26435,
234
+ 28279,
235
+ 29464,
236
+ 31650,
237
+ 32302,
238
+ 32470,
239
+ 36865,
240
+ 42863,
241
+ 47425,
242
+ 49870,
243
+ 50254,
244
+ 50258,
245
+ 50358,
246
+ 50359,
247
+ 50360,
248
+ 50361,
249
+ 50362
250
+ ],
251
+ "task_to_id": {
252
+ "transcribe": 50359,
253
+ "translate": 50358
254
+ },
255
+ "transformers_version": "4.37.2"
256
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e4592fb59c48d12313a96695723a715cf6e4fc707b723b8ef1c2e081844309e0
3
  size 290403936
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b0150bcb926139cc36ff51eed3f67837912d564bfceae3946f6b2ee7a9c07ce
3
  size 290403936
runs/Apr06_12-25-48_hyperion-255/events.out.tfevents.1712401186.hyperion-255 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a699266568473622e17c9eaa920d82e89a963528226edb19282dbf95ed7686ff
3
- size 264906
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e173ace33318aa678cd7f5b354e5ce7a47c49ebfa46c96887aa4945c22d83b14
3
+ size 272314