About AudioCraft

AudioCraft is a PyTorch library for deep learning research on audio generation. It provides inference and training code for several models that produce high-quality audio. The project targets researchers and developers working on audio synthesis and processing, offering a framework to experiment with components.

The repository describes installation steps, dependencies, and optional components to enable training or watermarking. It emphasizes reproducibility and access to tools for training and inference, guiding users through setup and execution.

AudioCraft combines multiple models and utilities, including a controllable music generation model and a suite of audio processing components. It is designed to support research workflows and integration with common deep learning environments and datasets.

The project distributes under MIT license for code and CC-BY-NC 4.0 for model weights, and provides references to how to cite the framework and model-specific READMEs when reproducing work. It documents licensing and citation practices for users.

Use cases

Research on audio generation and deep learning models.

Key features

MusicGen: A state-of-the-art controllable text-to-music model.
AudioGen: A state-of-the-art text-to-sound model.
EnCodec: A high fidelity neural audio codec.
Multi Band Diffusion: An EnCodec compatible decoder using diffusion.
MAGNeT and JASCO models for music and audio tasks.
Watermarking and related tools (AudioSeal).
Training and inference code for multiple models.

Pros and cons

Pros

Licensed under MIT for code and CC-BY-NC 4.0 for weights.

Reviews

Loading reviews…

Frequently asked questions

Is the training code available?

Yes! We provide the training code for EnCodec, MusicGen, Multi Band Diffusion and JASCO.

Where are the models stored?

Hugging Face stored the model in a specific location, which can be overridden by setting the AUDIOCRAFT_CACHE_DIR environment variable for the AudioCraft models.

What licenses apply?

MIT license for code; CC-BY-NC 4.0 license for weights.

Alternatives to AudioCraft

PixExtender

PixExtender is an AI image extender that expands images to exact sizes and aspect ratios, online, with free credits to start.

Tancky AI

Tancky AI is an online AI image combiner and photo merger that merges up to 4 images into a single photorealistic scene with commercial rights.

Image2Layer

Image2Layer separates a flat image into editable layers (up to 10) and exports layered PSD, PNGs, or ZIP packages for Photoshop and other tools.

Ultralytics

Ultralytics is a leading provider of AI and machine learning solutions, specializing in open-source software development. Their flagship product, Ultralytics YOLO, offers advanced capabilities for real-time object detection, empowering users to harness the power of AI with ease. With a strong commitment to data privacy, Ultralytics ensures that users retain full ownership of their data, securely hosted on trusted platforms like Google Cloud and Amazon AWS.

1.1M

FaceSymAI

FaceSymAI analyzes facial symmetry from a single photo in a browser, using 68 landmarks and 7 feature pairs.

514.2K

Viso Now

Build any real-world vision app in minutes. From idea to live vision app, in three steps. Start with an idea. Type it in plain language or drop in a clip. No datasets, no labelling, no model selection, just say what to watch for. Then watch it build itself. Viso Now connects the feed, recognizes the scene, writes the detection logic and wires the alerts, all agentic, on auto-pilot, in front of you. Then simply refine and ship live. You can tune any rule, threshold or output, then publish. Your vision app runs live and pushes events wherever your team already works.

118.7K