I used 28 second reference audio, and some of the audio is cut in the end of speech.Is this the behavior of model? if yes, is fine-tuning with the correct EOS will do the job?
· Sign up or log in to comment
