torchaudio 2.4.1

An audio package for PyTorch

Stars: 2504, Watchers: 2504, Forks: 645, Open Issues: 255

The pytorch/audio repo was created 7 years ago and the last code push was 55 minutes ago.
The project is very popular with an impressive 2504 github stars!

How to Install torchaudio

You can install torchaudio using pip

pip install torchaudio

or add it to a project with poetry

poetry add torchaudio

Package Details

Author

Soumith Chintala, David Pollack, Sean Naren, Peter Goldsborough, Moto Hira, Caroline Chen, Jeff Hwang, Zhaoheng Ni, Xiaohui Zhang

License

None

Homepage

https://github.com/pytorch/audio

PyPi:

https://pypi.org/project/torchaudio/

GitHub Repo:

https://github.com/pytorch/audio

Classifiers

• Multimedia/Sound/Audio
• Scientific/Engineering/Artificial Intelligence

Related Packages

Errors

A list of common torchaudio errors.

Code Examples

Here are some torchaudio code examples and snippets.

GitHub Issues

The torchaudio package has 255 open issues on GitHub

• pitch_shift consumes excessive GPU memory which is also not cleared
• Error while using PitchShift on GPU tensor
• AttributeError: module 'torchaudio.io' has no attribute 'AudioEffector'
• Hubert Pre-Training Example : Unable to load the saved checkpoint and resume training
• Split the CTC forced aligment API tutorial into two tutorials
• Add more explanation about n_fft
• include a link to index.rst
• Usage of TimeStretch is incorrect in documentation.
• Missing labels for WAV2VEC2_XLSR53
• Update README
• Update forced_align method to only support batch Tensors
• HuggingFace Transformers integration stated to fail
• "ONLINE ASR WITH EMFORMER RNN-T" inference does not work
• Fix StreamWriter regression around RGB0/BGR0
• Introduce chroma spectrogram transform

See more issues on GitHub