Meta Platforms (NASDAQ:META) introduced its open-source AI tool called AudioCraft which enables users to generate audio and music from text prompts.
Read More:– Google Assistant May Soon Be ‘Supercharged’ With Bard-Like AI
AudioCraft consists of three models — MusicGen, AudioGen and EnCodec. The company said is open-sourcing the three models 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 field of AI-generated audio.
The tech giant added that MusicGen was trained using company-owned and specifically licensed music, while AudioGen was trained on public sound effects. Both these models generate audio from text prompts.
In addition, Meta released an improved version of its EnCodec decoder, which allows higher quality music generation with fewer artifacts.
Read More:– Today’s ‘Wordle’ #771 Answer, Hints and Clues for Sunday, July 30 Game
The company also released its pre-trained AudioGen models, which let users generate environmental sounds and sound effects like a dog barking, cars honking or footsteps on a wooden floor.
Meta noted that it is sharing all of the AudioCraft model weights and code.
The company said that while they have seen a lot of excitement around generative AI for images, video, and text, audio has seemed to lag a bit behind.
Read More:– Today’s ‘Wordle’ #771 Answer, Hints and Clues for Sunday, July 30 Game
In July, Meta introduced Llama 2, in collaboration with Microsoft (MSFT), for free research and commercial use. Meta made Llama 2 available in the Azure AI model catalog, allowing developers using Microsoft Azure to build with it and use their cloud-native tools for content filtering and safety features.