The experimental family designed to train LLMs to understand sound natively.
Convert text to audio and vice versa
Note You can demo the model live here!