lucidrains/musiclm-pytorch
A PyTorch implementation of Google's MusicLM, a transformer-based model that generates music from text descriptions.
Not currently ranked — collecting fresh signals.
star history
This repository contains a PyTorch implementation of MusicLM, Google’s SOTA model for music generation. It combines MuLaN, a text-audio contrastive learning model, with AudioLM’s architecture to generate high-quality music from text descriptions. The system uses AudioSpectrogramTransformer and TextTransformer components trained together to align text embeddings with audio representations for conditional music synthesis.
Frequently asked
- What is lucidrains/musiclm-pytorch?
- A PyTorch implementation of Google's MusicLM, a transformer-based model that generates music from text descriptions.
- Is musiclm-pytorch open source?
- Yes — lucidrains/musiclm-pytorch is open source, released under the MIT license.
- What language is musiclm-pytorch written in?
- lucidrains/musiclm-pytorch is primarily written in Python.
- How popular is musiclm-pytorch?
- lucidrains/musiclm-pytorch has 3.3k stars on GitHub.
- Where can I find musiclm-pytorch?
- lucidrains/musiclm-pytorch is on GitHub at https://github.com/lucidrains/musiclm-pytorch.