patrickvonplaten commited on
Commit
5dadc39
1 Parent(s): 8d88236

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -1
README.md CHANGED
@@ -19,6 +19,7 @@ The base model pretrained and fine-tuned on 960 hours of Librispeech.
19
  Authors: Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli
20
 
21
  **Abstract**
 
22
  We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/other test sets. When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 4.8/8.2 WER. This demonstrates the feasibility of speech recognition with limited amounts of labeled data.
23
 
24
  The original model can be found under https://github.com/pytorch/fairseq/tree/master/examples/wav2vec#wav2vec-20.
@@ -26,7 +27,7 @@ The original model can be found under https://github.com/pytorch/fairseq/tree/ma
26
 
27
  # Usage
28
 
29
- The model can be used as follows to classify some speech input
30
 
31
  ```python
32
  from transformers import Wav2Vec2Tokenizer, Wav2Vec2Model
19
  Authors: Alexei Baevski, Henry Zhou, Abdelrahman Mohamed, Michael Auli
20
 
21
  **Abstract**
22
+
23
  We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/other test sets. When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 4.8/8.2 WER. This demonstrates the feasibility of speech recognition with limited amounts of labeled data.
24
 
25
  The original model can be found under https://github.com/pytorch/fairseq/tree/master/examples/wav2vec#wav2vec-20.
27
 
28
  # Usage
29
 
30
+ To transcribe audio files the model can be used as a standalone acoustic model as follows:
31
 
32
  ```python
33
  from transformers import Wav2Vec2Tokenizer, Wav2Vec2Model