Meta Launches Tool To Create Music Using AI

Meta has launched AudioCraft, an open-source AI tool that allows users to create audio and music from simple text prompts. The tool consists of three models: MusicGen, AudioGen, and EnCodec. MusicGen generates music from text inputs using Meta's music library, while AudioGen produces audio based on text inputs using public sound effects. The EnCodec decoder has been enhanced to improve music generation quality with fewer artifacts. Meta is making the pre-trained AudioGen models available for generating environmental sounds and sound effects. The company aims to fill the gap in generative AI for audio and encourages researchers to train their own models using AudioCraft.

‍