comparison between updated model and old model

as you can see, the mangio-crepe model is much better at pronunciation and doesnt mumble as much as the old harvest model

This sounds really good!

Would would an RMVPE version of this model sound like?

is RMVPE on normal RVC v2 as opposed to EasyGUI?

I assume it's available for both, but I wouldn't know because I never use the Easy GUI when I train RVC v2 models.

i dont remember seeing rmvpe on easygui

Here's my RMVPE-trained version of the model: Trained for 500 epoch with the batch size of 5.

Video removed in response to copyright claim.

oh my god, its amazing

sounds much more lifelike

Thanks!