Hugging Face's logo

thisisHJLee
/
wav2vec2-large-xls-r-300m-korean-g

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
1
thisisHJLee commited on
Commit
a05ccb5
•
1 Parent(s): 07c3fbb

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +14 -15
README.md CHANGED
@@ -14,8 +14,8 @@ should probably proofread and complete it, then remove this comment. -->
14
 
15
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the None dataset.
16
  It achieves the following results on the evaluation set:
17
- - Loss: 0.9226
18
- - Cer: 0.1638
19
 
20
  ## Model description
21
 
@@ -42,26 +42,25 @@ The following hyperparameters were used during training:
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 | Cer |
50
  |:-------------:|:-----:|:----:|:---------------:|:------:|
51
- | 4.8333 | 3.25 | 500 | 3.4624 | 0.9560 |
52
- | 1.243 | 6.49 | 1000 | 1.0049 | 0.2488 |
53
- | 0.3657 | 9.74 | 1500 | 0.8749 | 0.2087 |
54
- | 0.2104 | 12.99 | 2000 | 0.8799 | 0.1909 |
55
- | 0.1508 | 16.23 | 2500 | 0.9321 | 0.1845 |
56
- | 0.1245 | 19.48 | 3000 | 0.8778 | 0.1744 |
57
- | 0.1048 | 22.73 | 3500 | 0.9793 | 0.1808 |
58
- | 0.0922 | 25.97 | 4000 | 0.9464 | 0.1697 |
59
- | 0.0801 | 29.22 | 4500 | 0.9226 | 0.1638 |
60
 
61
 
62
  ### Framework versions
63
 
64
- - Transformers 4.17.0
65
- - Pytorch 1.10.0+cu113
66
  - Datasets 1.18.3
67
- - Tokenizers 0.13.0
 
14
 
15
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the None dataset.
16
  It achieves the following results on the evaluation set:
17
+ - Loss: 0.9504
18
+ - Cer: 0.1673
19
 
20
  ## Model description
21
 
 
42
  - lr_scheduler_type: linear
43
  - lr_scheduler_warmup_steps: 1000
44
  - num_epochs: 30
 
45
 
46
  ### Training results
47
 
48
  | Training Loss | Epoch | Step | Validation Loss | Cer |
49
  |:-------------:|:-----:|:----:|:---------------:|:------:|
50
+ | 4.7279 | 3.25 | 500 | 3.5098 | 0.9857 |
51
+ | 1.2914 | 6.49 | 1000 | 0.9970 | 0.2487 |
52
+ | 0.3791 | 9.74 | 1500 | 0.9036 | 0.2060 |
53
+ | 0.2147 | 12.99 | 2000 | 0.8963 | 0.1983 |
54
+ | 0.1607 | 16.23 | 2500 | 0.9386 | 0.1848 |
55
+ | 0.1281 | 19.48 | 3000 | 0.8897 | 0.1736 |
56
+ | 0.1094 | 22.73 | 3500 | 0.9474 | 0.1754 |
57
+ | 0.0933 | 25.97 | 4000 | 0.9557 | 0.1711 |
58
+ | 0.0845 | 29.22 | 4500 | 0.9504 | 0.1673 |
59
 
60
 
61
  ### Framework versions
62
 
63
+ - Transformers 4.30.2
64
+ - Pytorch 2.0.1+cu118
65
  - Datasets 1.18.3
66
+ - Tokenizers 0.13.3