resemble-ai/chatterbox

How to fix the nice we get at the end of each model output?

Opened this issue 2 months ago · 1 comments

hp2413 commented 2 months ago

if you find this audio file, how to get read of the end noise?

rsxdalv commented 2 months ago

Try disabling alignment stream analyzer. My branch has that done in t3.py: https://github.com/rsxdalv/chatterbox/blob/faster/src/chatterbox/models/t3/t3.py#L251