Hugging Face's logo

202015004
/
wav2vec2-base-timit-demo-colab

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Model card Files
xet
Metrics Community
2
202015004 commited on
Commit
b8137a4
·
1 Parent(s): 131609b

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +117 -0
README.md ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ model-index:
6
+ - name: wav2vec2-base-timit-demo-colab
7
+ results: []
8
+ ---
9
+
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
+
13
+ # wav2vec2-base-timit-demo-colab
14
+
15
+ This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 0.6259
18
+ - Wer: 0.3544
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 0.0001
38
+ - train_batch_size: 4
39
+ - eval_batch_size: 8
40
+ - seed: 42
41
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
42
+ - lr_scheduler_type: linear
43
+ - lr_scheduler_warmup_steps: 1000
44
+ - num_epochs: 30
45
+ - mixed_precision_training: Native AMP
46
+
47
+ ### Training results
48
+
49
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
50
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
51
+ | 3.6744 | 0.5 | 500 | 2.9473 | 1.0 |
52
+ | 1.4535 | 1.01 | 1000 | 0.7774 | 0.6254 |
53
+ | 0.7376 | 1.51 | 1500 | 0.6923 | 0.5712 |
54
+ | 0.5848 | 2.01 | 2000 | 0.5445 | 0.5023 |
55
+ | 0.4492 | 2.51 | 2500 | 0.5148 | 0.4958 |
56
+ | 0.4006 | 3.02 | 3000 | 0.5283 | 0.4781 |
57
+ | 0.3319 | 3.52 | 3500 | 0.5196 | 0.4628 |
58
+ | 0.3424 | 4.02 | 4000 | 0.5285 | 0.4551 |
59
+ | 0.2772 | 4.52 | 4500 | 0.5060 | 0.4532 |
60
+ | 0.2724 | 5.03 | 5000 | 0.5216 | 0.4422 |
61
+ | 0.2375 | 5.53 | 5500 | 0.5376 | 0.4443 |
62
+ | 0.2279 | 6.03 | 6000 | 0.6051 | 0.4308 |
63
+ | 0.2091 | 6.53 | 6500 | 0.5084 | 0.4423 |
64
+ | 0.2029 | 7.04 | 7000 | 0.5083 | 0.4242 |
65
+ | 0.1784 | 7.54 | 7500 | 0.6123 | 0.4297 |
66
+ | 0.1774 | 8.04 | 8000 | 0.5749 | 0.4339 |
67
+ | 0.1542 | 8.54 | 8500 | 0.5110 | 0.4033 |
68
+ | 0.1638 | 9.05 | 9000 | 0.6324 | 0.4318 |
69
+ | 0.1493 | 9.55 | 9500 | 0.6100 | 0.4152 |
70
+ | 0.1591 | 10.05 | 10000 | 0.5508 | 0.4022 |
71
+ | 0.1304 | 10.55 | 10500 | 0.5090 | 0.4054 |
72
+ | 0.1234 | 11.06 | 11000 | 0.6282 | 0.4093 |
73
+ | 0.1218 | 11.56 | 11500 | 0.5817 | 0.3941 |
74
+ | 0.121 | 12.06 | 12000 | 0.5741 | 0.3999 |
75
+ | 0.1073 | 12.56 | 12500 | 0.5818 | 0.4149 |
76
+ | 0.104 | 13.07 | 13000 | 0.6492 | 0.3953 |
77
+ | 0.0934 | 13.57 | 13500 | 0.5393 | 0.4083 |
78
+ | 0.0961 | 14.07 | 14000 | 0.5510 | 0.3919 |
79
+ | 0.0965 | 14.57 | 14500 | 0.5896 | 0.3992 |
80
+ | 0.0921 | 15.08 | 15000 | 0.5554 | 0.3947 |
81
+ | 0.0751 | 15.58 | 15500 | 0.6312 | 0.3934 |
82
+ | 0.0805 | 16.08 | 16000 | 0.6732 | 0.3948 |
83
+ | 0.0742 | 16.58 | 16500 | 0.5990 | 0.3884 |
84
+ | 0.0708 | 17.09 | 17000 | 0.6186 | 0.3869 |
85
+ | 0.0679 | 17.59 | 17500 | 0.5837 | 0.3848 |
86
+ | 0.072 | 18.09 | 18000 | 0.5831 | 0.3775 |
87
+ | 0.0597 | 18.59 | 18500 | 0.6562 | 0.3843 |
88
+ | 0.0612 | 19.1 | 19000 | 0.6298 | 0.3756 |
89
+ | 0.0514 | 19.6 | 19500 | 0.6746 | 0.3720 |
90
+ | 0.061 | 20.1 | 20000 | 0.6236 | 0.3788 |
91
+ | 0.054 | 20.6 | 20500 | 0.6012 | 0.3718 |
92
+ | 0.0521 | 21.11 | 21000 | 0.6053 | 0.3778 |
93
+ | 0.0494 | 21.61 | 21500 | 0.6154 | 0.3772 |
94
+ | 0.0468 | 22.11 | 22000 | 0.6052 | 0.3747 |
95
+ | 0.0413 | 22.61 | 22500 | 0.5877 | 0.3716 |
96
+ | 0.0424 | 23.12 | 23000 | 0.5786 | 0.3658 |
97
+ | 0.0403 | 23.62 | 23500 | 0.5828 | 0.3658 |
98
+ | 0.0391 | 24.12 | 24000 | 0.5913 | 0.3685 |
99
+ | 0.0312 | 24.62 | 24500 | 0.5850 | 0.3625 |
100
+ | 0.0316 | 25.13 | 25000 | 0.6029 | 0.3611 |
101
+ | 0.0282 | 25.63 | 25500 | 0.6312 | 0.3624 |
102
+ | 0.0328 | 26.13 | 26000 | 0.6312 | 0.3621 |
103
+ | 0.0258 | 26.63 | 26500 | 0.5891 | 0.3581 |
104
+ | 0.0256 | 27.14 | 27000 | 0.6259 | 0.3546 |
105
+ | 0.0255 | 27.64 | 27500 | 0.6315 | 0.3587 |
106
+ | 0.0249 | 28.14 | 28000 | 0.6547 | 0.3579 |
107
+ | 0.025 | 28.64 | 28500 | 0.6237 | 0.3565 |
108
+ | 0.0228 | 29.15 | 29000 | 0.6187 | 0.3559 |
109
+ | 0.0209 | 29.65 | 29500 | 0.6259 | 0.3544 |
110
+
111
+
112
+ ### Framework versions
113
+
114
+ - Transformers 4.11.3
115
+ - Pytorch 1.10.0+cu102
116
+ - Datasets 1.18.3
117
+ - Tokenizers 0.10.3