search

RVC-Project / Retrieval-based-Voice-Conversion-WebUI

Easily train a good VC model with voice data <= 10 mins!
MIT License
22.6k stars 3.41k forks source link

A bug in preprocess pipeline: ValueError: mode must be 'r', 'w', or None, got: rb #2116

Open colstone opened 2 months ago

colstone commented 2 months ago

used system and hardware: ubuntu 20.04 and 4*3090

hi, I'm find a bug in RVC1006Nvidia zip file. The trackback is "ValueError: mode must be 'r', 'w', or None, got: rb".

i'm changed a part of code on "RVC1006Nvidia/infer/lib/audio.py":

original is:

def audio2(i, o, format, sr):
    inp = av.open(i, "rb")
    out = av.open(o, "wb", format=format)

i'm fixed it is:

def audio2(i, o, format, sr):
    inp = av.open(i, "r")
    out = av.open(o, "w", format=format)

so this is a small bug, but i hope this bug are fix it well.

Boyishdude commented 1 month ago

I was having the same issue after finally getting the WebUi working in a weird way (I don't wanna talk about the troubleshooting and installations rabbit hole I went down), and this manual fix appears to have worked for me. I can pre-process my audio now! (I'm using Linux Mint 21).

MagicD3VIL commented 3 weeks ago

Thank you, this fixed the latest release for me on Linux.