Meta, formerly known as Facebook, has introduced ‘MusicGen,’ an advanced AI music generator. This innovative tool can transform text descriptions and melodies into high-quality audio compositions. In addition to its release, Meta has made the code and models accessible to foster open research, reproducibility, and collaboration within the music community.
Felix Kreuk, a Research Engineer at Meta AI research, shared on Twitter, “MusicGen is a simple and controllable music generation model that can be prompted by both text and melody. We have released the code (MIT) and models (CC-BY NC) to facilitate open research and reproducibility.”
Built on the EnCodec audio tokenizer, MusicGen sets itself apart with its single-stage transformerLM architecture, eliminating the need for cascading multiple models. The training data for MusicGen includes an extensive collection of 20,000 hours of music, comprising 10,000 licensed high-quality tracks and 390,000 instrument-only tracks from the ShutterStock and Pond5 stock media libraries.
While Meta joins the AI music generation landscape, it follows in the footsteps of Google, which introduced its own AI tool called ‘MusicLM’ capable of generating high-fidelity music based on textual descriptions. Meta’s MusicGen offers a distinct approach and is part of the company’s broader vision to integrate generative AI tools across its social media platforms and services.
Furthermore, Meta’s foray into AI-powered music generation follows in the footsteps of other tech giants. Google, for instance, unveiled ‘MusicLM’ earlier this year—an experimental AI tool capable of generating high-quality music in various genres based on text descriptions. Currently available to the public through the AI Test Kitchen app, MusicLM has garnered attention and sparked interest among music enthusiasts.
With Meta’s latest release of MusicGen, the company aims to carve its own path in the realm of AI music generation, fostering creativity and pushing the boundaries of what’s possible in the field. As Meta continues to explore the integration of generative AI tools, we can anticipate exciting developments and innovations across its social media platforms, including the potential introduction of an AI chatbot akin to Snapchat on Instagram, enriching the user experience and sparking new possibilities in the world of digital content creation.