Audio research library for PyTorch
The Audiocraft library is a powerful tool for audio processing and generation using deep learning. Developed by Facebook Research, it includes the Encodec audio compressor tokenizer and MusicGen, a language model for music generation that can be controlled with text and melody. This makes it a valuable resource for developers and researchers interested in music generation and audio processing. The library is built on PyTorch and requires a GPU for operation.
facebookresearch/audiocraft repo was created 1 months ago and the last code push was 2 weeks ago.
The project is extremely popular with a mindblowing 8913 github stars!
How to Install audiocraft
You can install audiocraft using pip
pip install audiocraft
or add it to a project with poetry
poetry add audiocraft
- FAIR Speech & Audio
- MIT License
- GitHub Repo
- Scientific/Engineering/Artificial Intelligence
A list of common audiocraft errors.
Here are some
audiocraft code examples and snippets.
The audiocraft package has 110 open issues on GitHub
- Audio Embeddings
- Can't find Seed?
- FFMPEG not found?
- Version conflicts with hydra-core
- –listen - how to make it work?
- Is it possible to use multiple GPU for single generate?
- Release Date of Traning Code
- Can anyone explain how the top_p and top_k parameter affects the output of the model?
- Update Readme: Add links to Replicate deployment of MusicGen
- how to train and how much data to invest so that there is no noise?
- Non-Usable download from Gradio?
- Any chance for adjust the sample and bit rate of WAVs up?
- help by install
- How to download file using api