tensorjackal
tensorjackal
Any updates here?
Interesting. I was trying to add instruction following to the model, and it started to bleed the instruction in the output audio itself. Very weird.
My bad! I meant laugh, breathe, cry like tags. If the model doesn't support them, how can we add them to it? Or maybe add instructions to model like: "Speak...
Can you share how can we add instructions as well? I tried to append it before the prompt text, but it started bleeding it into the generated audio.