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

Easily train a good VC model with voice data <= 10 mins!
MIT License
24.17k stars 3.58k forks source link

How do I use pretrained model? #363

Open DepressiveDude opened 1 year ago

DepressiveDude commented 1 year ago

Sorry if it is a dumb question, but I can't seem to figure out which models are which. I followed the instructions for installing, which includes downloading from huggingface. However, the inference voice list is empty, even if I restarted the program. I am new to this so I might be missing something, any help is appreciated.

johndpope commented 1 year ago

this is too hard -

Suggest updating ui to say

Inferencing voice: (load models to ./pretrained/1111.pth)

but doesn't work

am using a extracted zip file from here https://huggingface.co/QuickWick/Music-AI-Voices/tree/main

UPDATE - some progress. (load models to ./weights/1111.pth)

UPDATE 2 make sure you got all the files https://github.com/w-okada/rvc-trainer-docker

Ma5onic commented 1 year ago

https://youtu.be/-JcvdDErkAU You need to download the appropriate .7z (use 7-zip to extract), those zip archives contain some pretrained models.

kronkinatorix commented 1 year ago

Of course, why write anything down or have meaningful documentation I can search when I can waste even more time sitting through a fifteen minute youtube video./

Youtube is a grotesque metastasized adenoid greedily eating the internet's utility.