AudioGen is a text-to-sound model, based on Metas Audiocraft library, the same on which MusicGen is based.

sepal
/
audiogen
Generate sounds from a text prompt