# openai/jukebox

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/openai-jukebox).**

8,039 stars · 1,447 forks · Python · NOASSERTION · archived

## Links

- GitHub: https://github.com/openai/jukebox
- Homepage: https://openai.com/blog/jukebox/
- awesome-repositories: https://awesome-repositories.com/repository/openai-jukebox.md

## Topics

`audio` `generative-model` `music` `paper` `pytorch` `transformer` `vq-vae`

## Description

Jukebox is a generative audio model and AI music synthesis tool designed to create high-fidelity music samples and singing voices. It functions as a deep learning system that synthesizes raw audio conditioned on genre and artist metadata, utilizing a neural audio codec to convert raw audio into discrete codes for generative modeling and reconstruction.

The system enables musical style steering and AI music composition by conditioning generation on specific artists, genres, and lyrics. It supports audio priming, allowing existing wave files to guide the creation of new musical sequences, and provides mechanisms for lyric-to-audio alignment to coordinate vocal delivery timing.

The framework covers a broad range of capabilities including music style transfer and the training of generative priors. It includes workflows for audio compression model training and the fine-tuning of pre-trained models to adapt to new musical styles or datasets.

## Tags

### Graphics & Multimedia

- [Audio Synthesis](https://awesome-repositories.com/f/graphics-multimedia/media-processing-analysis/audio-processing-systems/audio-synthesis.md) — Synthesizes high-fidelity raw audio waveforms from scratch using deep learning networks. ([source](https://cdn.jsdelivr.net/gh/openai/jukebox@master/README.md))

### Artificial Intelligence & ML

- [AI Music Composition](https://awesome-repositories.com/f/artificial-intelligence-ml/ai-music-composition.md) — Generates original musical compositions by conditioning audio on artists, genres, and lyrics.
- [Audio Generation Models](https://awesome-repositories.com/f/artificial-intelligence-ml/audio-generation-models.md) — Implements a generative audio model that predicts sequences of audio codes to create music.
- [Neural Audio Compression](https://awesome-repositories.com/f/artificial-intelligence-ml/audio-tokenization/neural-audio-compression.md) — Utilizes a neural audio codec to compress raw waveforms into discrete sequences for efficient modeling.
- [Waveform Decoders](https://awesome-repositories.com/f/artificial-intelligence-ml/audio-tokenization/waveform-decoders.md) — Transforms discrete latent codes back into high-fidelity raw audio signals via a neural synthesis network.
- [Autoregressive Models](https://awesome-repositories.com/f/artificial-intelligence-ml/autoregressive-models.md) — Implements an autoregressive model to predict sequences of audio tokens for music synthesis.
- [Discrete Audio Representations](https://awesome-repositories.com/f/artificial-intelligence-ml/discrete-audio-representations.md) — Represents continuous audio signals as discrete codes from a learned vocabulary for generative modeling.
- [Conditional Generative Modeling](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-model-training-tools/conditional-training/conditional-generative-modeling.md) — Provides conditional generative modeling to steer audio synthesis using artist, genre, and lyric labels.
- [Audio Priming](https://awesome-repositories.com/f/artificial-intelligence-ml/audio-to-audio-synthesis/audio-priming.md) — Allows existing wave files to serve as a starting point to guide the creation of new musical sequences. ([source](https://cdn.jsdelivr.net/gh/openai/jukebox@master/README.md))
- [Lyric-to-Audio Alignment](https://awesome-repositories.com/f/artificial-intelligence-ml/audio-transcription/audio-transcript-aligners/lyric-to-audio-alignment.md) — Coordinates vocal delivery timing by mapping written text to specific segments of the generated audio stream. ([source](https://cdn.jsdelivr.net/gh/openai/jukebox@master/README.md))
- [Style-Based Music Generation](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-music-agents/style-based-music-generation.md) — Conditions audio generation on specific artists and genres to match a desired musical style. ([source](https://openai.com/blog/jukebox/))
- [Music Style Transfer](https://awesome-repositories.com/f/artificial-intelligence-ml/generative-music-agents/style-based-music-generation/music-style-transfer.md) — Adapts pre-trained models to new musical styles or artists through targeted fine-tuning.
- [Hierarchical Temporal Predictions](https://awesome-repositories.com/f/artificial-intelligence-ml/hierarchical-temporal-predictions.md) — Generates audio at multiple temporal resolutions to capture high-level musical structure and fine-grained acoustic detail.

### Part of an Awesome List

- [Music And Audio Generation](https://awesome-repositories.com/f/awesome-lists/media/music-and-audio-generation.md) — Synthesizes raw audio and singing voices conditioned on genre and artist metadata.