Resemblyzer
Resemblyzer copied to clipboard
Using VoiceEncoder as loss with time domain models
Is there a way we could use the VoiceEncoder as the loss function with time-domain models?