Downloads · 30 days
294
0% of all-time downloads
distil-wav2vec2
OthmaneJ/distil-wav2vec2
distil-wav2vec2 is a automatic speech recognition model from OthmaneJ. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
This model is a distilled version of the wav2vec2 model (https://arxiv.org/pdf/2006.11477.pdf). This model is 45% times smaller and twice as fast as the original wav2vec2 base model.
All-time downloads
269K
Public
Repo size
1.3 GB
Likes
13
Public
Repo makeup
Click a slice to open those files.
.bin208 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- wav2vec2
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Aug 25, 2021
- SHA
- e7d24070
Datasets
Model card
From the Hugging Face model README
Distil-wav2vec2
This model is a distilled version of the wav2vec2 model (https://arxiv.org/pdf/2006.11477.pdf). This model is 45% times smaller and twice as fast as the original wav2vec2 base model.
Evaluation results
This model achieves the following results (speed is mesured for a batch size of 64):
| Model | Size | WER Librispeech-test-clean | WER Librispeech-test-other | Speed on cpu | speed on gpu |
|---|---|---|---|---|---|
| Distil-wav2vec2 | 197.9 Mb | 0.0983 | 0.2266 | 0.4006s | 0.0046s |
| wav2vec2-base | 360 Mb | 0.0389 | 0.1047 | 0.4919s | 0.0082s |
Usage
notebook (executes seamlessly on google colab) at https://github.com/OthmaneJ/distil-wav2vec2
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- License
- apache-2.0
- Languages
- en
- Created
- Mar 2, 2022
- Updated
- Aug 25, 2021