wav2vec 2.0 base model pretrained on librispeech 960h

This HuggingFace repository provides all the necessary tools to extract wav2vec2 embeddings from a pretrained model. For a better experience, we encourage you to learn more about SpeechBrain. The wav2vec2 model has entirely been pretrained with SpeechBrain (not with fairseq or HuggingFace).

The performance of the model is the following:

Release	Test WER	GPUs
22-09-22	7.X	1xV100 32GB

Pipeline description

This w2v2 system is composed of 2 different but linked blocks:

A convolutional backend to extract features from the raw waveform.
A latent encoder made of a transformer network. The obtained embeddings are the output of the transformer after going through each block.

Install SpeechBrain

First of all, please install SpeechBrain with the following command:

pip install speechbrain

Please notice that we encourage you to read our tutorials and learn more about SpeechBrain.

Extracting embeddings for your own audio files

from speechbrain.inference.encoders import WaveformEncoder

ssl_model = WaveformEncoder.from_hparams(source="speechbrain/ssl-wav2vec2-base-librispeech", savedir="speechbrain/ssl-wav2vec2-base-librispeech")
ssl_model.encode_file("mywavfile.wav")

Inference on GPU

To perform inference on the GPU, add run_opts={"device":"cuda"} when calling the from_hparams method.

Training

The model was trained with SpeechBrain. To train it from scratch follow these steps:

Clone SpeechBrain:

git clone https://github.com/speechbrain/speechbrain/

Install it:

cd speechbrain
pip install -r requirements.txt
pip install -e .

Run Training:

cd recipes/LibriSpeech/self-supervised-learning/wav2vec2
python train_sb_wav2vec2.py hparams/wav2vec2_base.yaml --data_folder=your_data_folder

You can find our training results (models, logs, etc) here.

Limitations

The SpeechBrain team does not provide any warranty on the performance achieved by this model when used on other datasets.

Referencing SpeechBrain

@misc{SB2021,
    author = {Ravanelli, Mirco and Parcollet, Titouan and Rouhe, Aku and Plantinga, Peter and Rastorgueva, Elena and Lugosch, Loren and Dawalatabad, Nauman and Ju-Chieh, Chou and Heba, Abdel and Grondin, Francois and Aris, William and Liao, Chien-Feng and Cornell, Samuele and Yeh, Sung-Lin and Na, Hwidong and Gao, Yan and Fu, Szu-Wei and Subakan, Cem and De Mori, Renato and Bengio, Yoshua },
    title = {SpeechBrain},
    year = {2021},
    publisher = {GitHub},
    journal = {GitHub repository},
    howpublished = {\\\\url{https://github.com/speechbrain/speechbrain}},
  }

About SpeechBrain

SpeechBrain is an open-source and all-in-one speech toolkit. It is designed to be simple, extremely flexible, and user-friendly. Competitive or state-of-the-art performance is obtained in various domains.

Website: https://speechbrain.github.io/

GitHub: https://github.com/speechbrain/speechbrain