search

fumiama / Retrieval-based-Voice-Conversion-WebUI

Easily train a good VC model with voice data <= 10 mins!
GNU Affero General Public License v3.0
146 stars 20 forks source link

Add 32k for model merging. #84

Open litsa-the-dancer opened 3 months ago

litsa-the-dancer commented 3 months ago

I don't know why this hasn't been taken care of already but it's essential. image although even when merging as 40k or 48k using 2 32k weights it still outputs 32k just fine.

fumiama commented 3 months ago

It has been added in dev branch. Please try it out.