AudioCraft lets one easily generate high-quality audio and music from text.
With this, Meta is open-sourcing three models of AudioCraft — MusicGen, AudioGen and EnCodec. MusicGen was trained with Meta-owned and specifically licensed music and it generates music from text prompts, while AudioGen was trained on public sound effects and generates audio from text prompts.
By open-sourcing these models, Meta is giving researchers and practitioners access so they can train their own models with their own datasets for the first time, and help advance the field of AI-generated audio and music.
Sample AudioCraft Generated Music —
AudioCraft is a single code base that works for music, sound, compression & generation - all in the same place. It consists of three models: MusicGen, AudioGen and EnCodec. Today's release builds on our previous release of MusicGen with an improved version of our EnCodec decoder enabling higher quality music generation with fewer artifacts + pre-trained AudioGen models which can generate environmental sounds and sound effects.
Meta in a statement said, "As part of our continued investment in an open approach to today's Al, the models are available for research purposes so that researchers and practitioners can train their own models with their own datasets for the first time and help advance the state of the art. We can't wait to see what people create with AudioCraft."
The AudioCraft family of models are capable of producing high-quality audio with long-term consistency, and they’re easy to use.
Earlier this year, Google's holding firm Alphabet Inc too unveiled its own experimental audio generating AI tool called MusicLM.
Meta in a statement said, "As part of our continued investment in an open approach to today's Al, the models are available for research purposes so that researchers and practitioners can train their own models with their own datasets for the first time and help advance the state of the art. We can't wait to see what people create with AudioCraft."
The AudioCraft family of models are capable of producing high-quality audio with long-term consistency, and they’re easy to use.
Earlier this year, Google's holding firm Alphabet Inc too unveiled its own experimental audio generating AI tool called MusicLM.
Meanwhile, Artists and industry experts have raised concerns over copyright violations, as these generative AI tools used machine learning software which work by recognizing and replicating patterns from music/data scraped from the web.
Advertisements