Hugging Face's logo

birgermoell
/
wav2vec2-swedish-common-voice

Automatic Speech Recognition
Transformers
PyTorch
JAX
Swedish
wav2vec2
audio
speech
xlsr-fine-tuning-week
Eval Results
Inference Endpoints
Model card Files Community
birgermoell commited on
Commit
29cf512
•
1 Parent(s): 989b43e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -2
README.md CHANGED
@@ -1,7 +1,12 @@
 
 
1
  Wav2vec2-swedish trained on the common voice dataset.
2
 
 
3
  WER 0.511916 on a dataset of 402mb.
4
 
5
- Link to notebook for training.
 
6
 
7
- https://colab.research.google.com/drive/1KkD4PeZwnIwxxxOP1bUE7XTZMK7-SzRj?usp=sharing
 
 
1
+ ## wav2vec2-swedish-common-voice
2
+ This model is fine-tuned from Facebook's wav2vec2-large-xlsr-53 using Mozilla's common voice dataset.
3
  Wav2vec2-swedish trained on the common voice dataset.
4
 
5
+ ## Model metric
6
  WER 0.511916 on a dataset of 402mb.
7
 
8
+ ## Model Fine-Tuning-Colab
9
+ https://colab.research.google.com/drive/1KkD4PeZwnIwxxxOP1bUE7XTZMK7-SzRj?usp=sharing
10
 
11
+ Based on this great blogpost by Patrick von Platen
12
+ https://huggingface.co/blog/fine-tune-xlsr-wav2vec2