pere commited on
Commit
0871114
1 Parent(s): 9f9e827

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +26 -117
README.md CHANGED
@@ -1,39 +1,38 @@
1
  ---
2
  license: apache-2.0
3
  tags:
4
- - whisper-event
5
  - generated_from_trainer
6
  datasets:
7
- - NbAiLab/NCC_S
8
  metrics:
9
  - wer
10
  model-index:
11
- - name: "Whisper Tiny Norwegian Bokm\xE5l"
12
  results:
13
  - task:
14
  name: Automatic Speech Recognition
15
  type: automatic-speech-recognition
16
  dataset:
17
- name: NbAiLab/NCC_S
18
- type: NbAiLab/NCC_S
19
  config: 'no'
20
  split: validation
21
  args: 'no'
22
  metrics:
23
  - name: Wer
24
  type: wer
25
- value: 24.878197320341048
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
- # Whisper Tiny Norwegian Bokmål
32
 
33
- This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the NbAiLab/NCC_S dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.5100
36
- - Wer: 24.8782
37
 
38
  ## Model description
39
 
@@ -53,124 +52,34 @@ More information needed
53
 
54
  The following hyperparameters were used during training:
55
  - learning_rate: 3e-06
56
- - train_batch_size: 256
57
- - eval_batch_size: 64
58
  - seed: 42
59
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
  - lr_scheduler_type: constant_with_warmup
61
- - lr_scheduler_warmup_steps: 1000
62
- - training_steps: 100000
63
  - mixed_precision_training: Native AMP
64
 
65
  ### Training results
66
 
67
- | Training Loss | Epoch | Step | Validation Loss | Wer |
68
- |:-------------:|:-----:|:------:|:---------------:|:-------:|
69
- | 1.8819 | 0.01 | 1000 | 1.1869 | 61.9671 |
70
- | 1.6425 | 0.02 | 2000 | 0.9991 | 53.6541 |
71
- | 1.548 | 0.03 | 3000 | 0.9147 | 50.2132 |
72
- | 1.4636 | 0.04 | 4000 | 0.8605 | 47.0767 |
73
- | 1.4113 | 0.05 | 5000 | 0.8253 | 45.7369 |
74
- | 1.3484 | 0.01 | 6000 | 0.7946 | 43.4531 |
75
- | 1.3127 | 0.02 | 7000 | 0.7740 | 42.2655 |
76
- | 1.2994 | 0.03 | 8000 | 0.7551 | 40.8952 |
77
- | 1.265 | 0.04 | 9000 | 0.7378 | 39.8599 |
78
- | 1.2458 | 0.05 | 10000 | 0.7257 | 39.8904 |
79
- | 1.2257 | 0.06 | 11000 | 0.7114 | 39.7990 |
80
- | 1.2126 | 0.07 | 12000 | 0.6972 | 37.8806 |
81
- | 1.1971 | 0.08 | 13000 | 0.6871 | 37.3021 |
82
- | 1.1786 | 1.01 | 14000 | 0.6786 | 37.4239 |
83
- | 1.1486 | 1.02 | 15000 | 0.6703 | 36.9976 |
84
- | 1.1505 | 1.03 | 16000 | 0.6647 | 36.3581 |
85
- | 1.1238 | 1.04 | 17000 | 0.6559 | 36.3886 |
86
- | 1.1184 | 1.05 | 18000 | 0.6509 | 36.5104 |
87
- | 1.115 | 1.06 | 19000 | 0.6452 | 35.9927 |
88
- | 1.1013 | 1.07 | 20000 | 0.6382 | 34.5006 |
89
- | 1.0969 | 1.08 | 21000 | 0.6331 | 34.3484 |
90
- | 1.0784 | 2.0 | 22000 | 0.6304 | 34.2875 |
91
- | 1.0774 | 2.01 | 23000 | 0.6249 | 34.1048 |
92
- | 1.0719 | 2.02 | 24000 | 0.6194 | 33.8307 |
93
- | 1.0638 | 2.03 | 25000 | 0.6158 | 32.9781 |
94
- | 1.0592 | 2.04 | 26000 | 0.6105 | 32.6431 |
95
- | 1.0493 | 2.05 | 27000 | 0.6041 | 32.7345 |
96
- | 1.047 | 2.06 | 28000 | 0.6040 | 32.7649 |
97
- | 1.0323 | 2.07 | 29000 | 0.5984 | 31.6078 |
98
- | 1.0189 | 3.0 | 30000 | 0.5957 | 31.3033 |
99
- | 1.0078 | 3.01 | 31000 | 0.5924 | 31.4251 |
100
- | 1.0146 | 3.02 | 32000 | 0.5940 | 31.3033 |
101
- | 1.0128 | 3.03 | 33000 | 0.5892 | 31.0292 |
102
- | 1.0025 | 3.04 | 34000 | 0.5873 | 31.1815 |
103
- | 0.999 | 3.05 | 35000 | 0.5838 | 30.6334 |
104
- | 1.0045 | 3.06 | 36000 | 0.5799 | 30.4202 |
105
- | 1.0005 | 3.07 | 37000 | 0.5770 | 30.1766 |
106
- | 1.0017 | 3.08 | 38000 | 0.5733 | 29.6590 |
107
- | 0.9878 | 4.01 | 39000 | 0.5745 | 30.2680 |
108
- | 0.9854 | 4.02 | 40000 | 0.5720 | 30.0548 |
109
- | 0.9624 | 4.03 | 41000 | 0.5703 | 29.5981 |
110
- | 0.9639 | 4.04 | 42000 | 0.5681 | 29.5067 |
111
- | 0.9569 | 4.05 | 43000 | 0.5679 | 29.6285 |
112
- | 0.9682 | 4.06 | 44000 | 0.5643 | 29.5676 |
113
- | 0.9539 | 4.07 | 45000 | 0.5601 | 29.5676 |
114
- | 0.946 | 4.08 | 46000 | 0.5562 | 29.7199 |
115
- | 0.9429 | 5.01 | 47000 | 0.5592 | 29.2935 |
116
- | 0.9462 | 5.02 | 48000 | 0.5540 | 29.0804 |
117
- | 0.9312 | 5.03 | 49000 | 0.5535 | 29.2935 |
118
- | 0.9462 | 5.04 | 50000 | 0.5536 | 28.6845 |
119
- | 0.922 | 5.05 | 51000 | 0.5539 | 28.7150 |
120
- | 0.9253 | 5.06 | 52000 | 0.5510 | 28.8368 |
121
- | 0.9065 | 0.01 | 53000 | 0.5493 | 28.5932 |
122
- | 0.9096 | 0.02 | 54000 | 0.5490 | 28.5018 |
123
- | 0.9329 | 0.03 | 55000 | 0.5483 | 28.2887 |
124
- | 0.9181 | 0.04 | 56000 | 0.5471 | 27.9842 |
125
- | 0.914 | 0.05 | 57000 | 0.5457 | 28.4105 |
126
- | 0.9149 | 0.06 | 58000 | 0.5449 | 27.5883 |
127
- | 0.9092 | 0.07 | 59000 | 0.5405 | 27.8319 |
128
- | 0.9101 | 0.08 | 60000 | 0.5402 | 27.3447 |
129
- | 0.9046 | 1.01 | 61000 | 0.5374 | 27.5579 |
130
- | 0.8917 | 1.02 | 62000 | 0.5390 | 27.7406 |
131
- | 0.8993 | 1.03 | 63000 | 0.5386 | 27.4056 |
132
- | 0.8875 | 1.04 | 64000 | 0.5361 | 26.8575 |
133
- | 0.8892 | 1.05 | 65000 | 0.5358 | 27.3447 |
134
- | 0.8929 | 1.06 | 66000 | 0.5346 | 26.7357 |
135
- | 0.8703 | 0.01 | 67000 | 0.5332 | 26.8270 |
136
- | 0.8709 | 0.02 | 68000 | 0.5336 | 26.7052 |
137
- | 0.8917 | 0.03 | 69000 | 0.5329 | 27.0706 |
138
- | 0.8867 | 0.04 | 70000 | 0.5323 | 26.3398 |
139
- | 0.8778 | 0.05 | 71000 | 0.5315 | 27.2838 |
140
- | 0.8757 | 0.06 | 72000 | 0.5317 | 26.2485 |
141
- | 0.8726 | 0.07 | 73000 | 0.5269 | 26.6443 |
142
- | 0.8792 | 0.08 | 74000 | 0.5268 | 26.1571 |
143
- | 0.8706 | 1.01 | 75000 | 0.5247 | 26.1571 |
144
- | 0.8585 | 1.02 | 76000 | 0.5265 | 26.3703 |
145
- | 0.8659 | 1.03 | 77000 | 0.5262 | 26.7357 |
146
- | 0.8551 | 1.04 | 78000 | 0.5249 | 26.0658 |
147
- | 0.8572 | 1.05 | 79000 | 0.5249 | 26.2789 |
148
- | 0.8612 | 1.06 | 80000 | 0.5235 | 25.7613 |
149
- | 0.8598 | 1.07 | 81000 | 0.5208 | 25.7004 |
150
- | 0.8686 | 1.08 | 82000 | 0.5214 | 25.7004 |
151
- | 0.8503 | 2.0 | 83000 | 0.5214 | 25.7004 |
152
- | 0.8545 | 2.01 | 84000 | 0.5215 | 28.2278 |
153
- | 0.8594 | 2.02 | 85000 | 0.5186 | 25.6699 |
154
- | 0.86 | 2.03 | 86000 | 0.5196 | 25.5786 |
155
- | 0.8514 | 2.04 | 87000 | 0.5203 | 25.1827 |
156
- | 0.8505 | 2.05 | 88000 | 0.5164 | 28.0146 |
157
- | 0.8512 | 2.06 | 89000 | 0.5174 | 25.0914 |
158
- | 0.8495 | 2.07 | 90000 | 0.5141 | 25.5481 |
159
- | 0.8381 | 3.0 | 91000 | 0.5130 | 24.9695 |
160
- | 0.8253 | 3.01 | 92000 | 0.5147 | 25.5786 |
161
- | 0.8387 | 3.02 | 93000 | 0.5168 | 24.9086 |
162
- | 0.8425 | 3.03 | 94000 | 0.5135 | 25.2436 |
163
- | 0.8339 | 3.04 | 95000 | 0.5162 | 25.6699 |
164
- | 0.8402 | 3.05 | 96000 | 0.5147 | 25.7308 |
165
- | 0.8396 | 3.06 | 97000 | 0.5143 | 25.6699 |
166
- | 0.8432 | 3.07 | 98000 | 0.5100 | 24.8782 |
167
- | 0.844 | 3.08 | 99000 | 0.5100 | 25.0609 |
168
- | 0.8333 | 4.01 | 100000 | 0.5128 | 24.9695 |
169
 
170
 
171
  ### Framework versions
172
 
173
- - Transformers 4.26.0.dev0
174
  - Pytorch 1.13.0+cu117
175
  - Datasets 2.7.1.dev0
176
  - Tokenizers 0.13.2
1
  ---
2
  license: apache-2.0
3
  tags:
 
4
  - generated_from_trainer
5
  datasets:
6
+ - ncc_s3
7
  metrics:
8
  - wer
9
  model-index:
10
+ - name: whisper-tiny-gputest
11
  results:
12
  - task:
13
  name: Automatic Speech Recognition
14
  type: automatic-speech-recognition
15
  dataset:
16
+ name: ncc_s3
17
+ type: ncc_s3
18
  config: 'no'
19
  split: validation
20
  args: 'no'
21
  metrics:
22
  - name: Wer
23
  type: wer
24
+ value: 51.37028014616322
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
28
  should probably proofread and complete it, then remove this comment. -->
29
 
30
+ # whisper-tiny-gputest
31
 
32
+ This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the ncc_s3 dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 0.9375
35
+ - Wer: 51.3703
36
 
37
  ## Model description
38
 
52
 
53
  The following hyperparameters were used during training:
54
  - learning_rate: 3e-06
55
+ - train_batch_size: 128
56
+ - eval_batch_size: 32
57
  - seed: 42
58
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
  - lr_scheduler_type: constant_with_warmup
60
+ - lr_scheduler_warmup_steps: 200
61
+ - training_steps: 2000
62
  - mixed_precision_training: Native AMP
63
 
64
  ### Training results
65
 
66
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
67
+ |:-------------:|:-----:|:----:|:---------------:|:-------:|
68
+ | 2.4574 | 0.1 | 200 | 1.4663 | 71.6504 |
69
+ | 1.9587 | 0.2 | 400 | 1.2581 | 64.7381 |
70
+ | 1.816 | 0.3 | 600 | 1.1672 | 60.9318 |
71
+ | 1.7199 | 0.4 | 800 | 1.1006 | 57.6736 |
72
+ | 1.6686 | 0.5 | 1000 | 1.0630 | 56.1815 |
73
+ | 1.621 | 0.6 | 1200 | 1.0273 | 55.4811 |
74
+ | 1.5846 | 0.7 | 1400 | 1.0017 | 53.9890 |
75
+ | 1.5482 | 0.8 | 1600 | 0.9773 | 53.0146 |
76
+ | 1.521 | 0.9 | 1800 | 0.9575 | 52.1011 |
77
+ | 1.4932 | 1.0 | 2000 | 0.9375 | 51.3703 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
78
 
79
 
80
  ### Framework versions
81
 
82
+ - Transformers 4.28.0.dev0
83
  - Pytorch 1.13.0+cu117
84
  - Datasets 2.7.1.dev0
85
  - Tokenizers 0.13.2