Hugging Face's logo

NbAiLab
/
whisper-large-sme

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
Safetensors
Northern Sami
whisper
whisper-event
Generated from Trainer
Eval Results (legacy)
Model card Files
xet
Metrics Community
2
versae commited on
Commit
7bfb630
·
1 Parent(s): 889c748

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +134 -1
README.md CHANGED
@@ -1,3 +1,136 @@
1
  ---
2
- license: openrail
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - audiofolder
7
+ metrics:
8
+ - wer
9
+ model-index:
10
+ - name: whisper-large-sme
11
+ results:
12
+ - task:
13
+ name: Automatic Speech Recognition
14
+ type: automatic-speech-recognition
15
+ dataset:
16
+ name: audiofolder
17
+ type: audiofolder
18
+ config: default
19
+ split: test
20
+ args: default
21
+ metrics:
22
+ - name: Wer
23
+ type: wer
24
+ value: 25.257142857142856
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-large-sme
31
+
32
+ This model is a fine-tuned version of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2) on the audiofolder dataset.
33
+ It achieves the following results on the evaluation set:
34
+ - Loss: 0.5975
35
+ - Wer: 25.2571
36
+
37
+ ## Model description
38
+
39
+ More information needed
40
+
41
+ ## Intended uses & limitations
42
+
43
+ More information needed
44
+
45
+ ## Training and evaluation data
46
+
47
+ More information needed
48
+
49
+ ## Training procedure
50
+
51
+ ### Training hyperparameters
52
+
53
+ The following hyperparameters were used during training:
54
+ - learning_rate: 1e-05
55
+ - train_batch_size: 12
56
+ - eval_batch_size: 6
57
+ - seed: 42
58
+ - distributed_type: multi-GPU
59
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
+ - lr_scheduler_type: linear
61
+ - lr_scheduler_warmup_steps: 500
62
+ - training_steps: 60000
63
+ - mixed_precision_training: Native AMP
64
+
65
+ ### Training results
66
+
67
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
68
+ |:-------------:|:------:|:-----:|:---------------:|:-------:|
69
+ | 0.4665 | 58.0 | 1000 | 0.8572 | 54.5143 |
70
+ | 0.3041 | 117.0 | 2000 | 0.6711 | 44.1143 |
71
+ | 0.2671 | 176.0 | 3000 | 0.5794 | 39.7714 |
72
+ | 0.1761 | 235.0 | 4000 | 0.5357 | 35.0857 |
73
+ | 0.2089 | 294.0 | 5000 | 0.5094 | 33.6 |
74
+ | 0.1456 | 352.0 | 6000 | 0.4959 | 33.0286 |
75
+ | 0.1514 | 411.0 | 7000 | 0.4864 | 32.5714 |
76
+ | 0.1203 | 470.0 | 8000 | 0.4625 | 31.4286 |
77
+ | 0.0879 | 529.0 | 9000 | 0.4916 | 45.4857 |
78
+ | 0.0825 | 588.0 | 10000 | 0.4962 | 30.6286 |
79
+ | 0.0753 | 647.0 | 11000 | 0.4723 | 31.2 |
80
+ | 0.0812 | 705.0 | 12000 | 0.4574 | 28.6857 |
81
+ | 0.062 | 764.0 | 13000 | 0.4628 | 28.8000 |
82
+ | 0.0604 | 823.0 | 14000 | 0.4668 | 28.0000 |
83
+ | 0.0666 | 882.0 | 15000 | 0.4697 | 28.6857 |
84
+ | 0.0405 | 941.0 | 16000 | 0.4908 | 54.6286 |
85
+ | 0.0349 | 999.0 | 17000 | 0.4728 | 28.4571 |
86
+ | 0.0409 | 1058.0 | 18000 | 0.4884 | 28.4571 |
87
+ | 0.0292 | 1117.0 | 19000 | 0.4576 | 27.3143 |
88
+ | 0.0247 | 1176.0 | 20000 | 0.4734 | 28.9143 |
89
+ | 0.0229 | 1235.0 | 21000 | 0.4899 | 29.9429 |
90
+ | 0.0271 | 1294.0 | 22000 | 0.4790 | 28.1143 |
91
+ | 0.0271 | 1352.0 | 23000 | 0.5012 | 30.1714 |
92
+ | 0.0184 | 1411.0 | 24000 | 0.5008 | 27.3143 |
93
+ | 0.0211 | 1470.0 | 25000 | 0.5118 | 27.6571 |
94
+ | 0.0183 | 1529.0 | 26000 | 0.5398 | 30.0571 |
95
+ | 0.0164 | 1588.0 | 27000 | 0.5006 | 27.3143 |
96
+ | 0.0169 | 1647.0 | 28000 | 0.5059 | 27.0857 |
97
+ | 0.0147 | 1705.0 | 29000 | 0.5325 | 27.7714 |
98
+ | 0.0104 | 1764.0 | 30000 | 0.4818 | 26.1714 |
99
+ | 0.0128 | 1823.0 | 31000 | 0.5259 | 28.3429 |
100
+ | 0.0145 | 1882.0 | 32000 | 0.5299 | 26.2857 |
101
+ | 0.0075 | 1941.0 | 33000 | 0.5082 | 27.4286 |
102
+ | 0.0087 | 1999.0 | 34000 | 0.5144 | 26.6286 |
103
+ | 0.005 | 2058.0 | 35000 | 0.5590 | 27.0857 |
104
+ | 0.0099 | 2117.0 | 36000 | 0.5546 | 28.9143 |
105
+ | 0.007 | 2176.0 | 37000 | 0.5364 | 26.8571 |
106
+ | 0.0045 | 2235.0 | 38000 | 0.5574 | 27.2000 |
107
+ | 0.0064 | 2294.0 | 39000 | 0.5051 | 25.7143 |
108
+ | 0.0079 | 2352.0 | 40000 | 0.5247 | 25.9429 |
109
+ | 0.0083 | 2411.0 | 41000 | 0.5514 | 25.6 |
110
+ | 0.0101 | 2470.0 | 42000 | 0.5710 | 25.6 |
111
+ | 0.0062 | 2529.0 | 43000 | 0.5830 | 28.0000 |
112
+ | 0.0046 | 2588.0 | 44000 | 0.5828 | 26.8571 |
113
+ | 0.0053 | 2647.0 | 45000 | 0.5621 | 27.4286 |
114
+ | 0.0047 | 2705.0 | 46000 | 0.5673 | 25.9429 |
115
+ | 0.0045 | 2764.0 | 47000 | 0.5220 | 25.6 |
116
+ | 0.0065 | 2823.0 | 48000 | 0.5704 | 27.7714 |
117
+ | 0.0039 | 2882.0 | 49000 | 0.5741 | 27.7714 |
118
+ | 0.0027 | 2941.0 | 50000 | 0.5762 | 26.0571 |
119
+ | 0.0019 | 2999.0 | 51000 | 0.5559 | 24.9143 |
120
+ | 0.0015 | 3058.0 | 52000 | 0.5777 | 28.5714 |
121
+ | 0.0026 | 3117.0 | 53000 | 0.5589 | 25.2571 |
122
+ | 0.0032 | 3176.0 | 54000 | 0.6061 | 26.9714 |
123
+ | 0.0025 | 3235.0 | 55000 | 0.5776 | 25.1429 |
124
+ | 0.0046 | 3294.0 | 56000 | 0.5753 | 27.3143 |
125
+ | 0.0015 | 3352.0 | 57000 | 0.5736 | 27.2000 |
126
+ | 0.003 | 3411.0 | 58000 | 0.5933 | 25.6 |
127
+ | 0.002 | 3470.0 | 59000 | 0.6036 | 25.6 |
128
+ | 0.0007 | 58.0 | 60000 | 0.5975 | 25.2571 |
129
+
130
+
131
+ ### Framework versions
132
+
133
+ - Transformers 4.26.0.dev0
134
+ - Pytorch 1.13.0+cu117
135
+ - Datasets 2.7.1.dev0
136
+ - Tokenizers 0.11.0