More speaker voice for each model

#101
by boooooiy - opened

If one listens and votes enough, it's not hard for him to tell which model generated the samples compared. Some models have only 1 or 2 voices. And I think it would lead to bias and discredit the leaderboard a bit...

TTS AGI org

Yes, I totally agree. This is something we are actively working on

Sign up or log in to comment