Giters

fumiama / Retrieval-based-Voice-Conversion-WebUI

Easily train a good VC model with voice data <= 10 mins!

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

mute file duplicated during training

SayanoAI opened this issue · comments

commented

what's the rationale for duplicating the mute files during training?

image

Why does including it just once not work?

commented

if_f0_3 is always true, because you can't extract without a f0 method, idk why that is there

(unrelated to OP's issue) honestly i'm extremely tired of all the weird names for vars and functions that are in RVC.
i'll start working on renaming all of the vars to be something that is easier to understand

commented

i'll start working on renaming all of the vars to be something that is easier to understand

Thanks but maybe you can start it later to aviod code conflicts because I'm changing the whole code structure recently.

As for the main topic of this issue, I will remove all unnecessary vars in train later. Now I am just working on the infer part.

commented

Thanks but maybe you can start it later to aviod code conflicts because I'm changing the whole code structure recently.

As for the main topic of this issue, I will remove all unnecessary vars in train later. Now I am just working on the infer part.

I'd recommend structuring the RVC code as a pypi package and build the gui on top of that. This will make it easier for people to create colabs and whatnot.

commented

I'd recommend structuring the RVC code as a pypi package

Yes. See the progress at https://github.com/RVC-Project/Retrieval-based-Voice-Conversion .

But now, I'm just letting the RVC code seeing more like a package so that we can migrate to package version easily.