Cannot use any voice model on a mac M3

[Skowizzz]

Hi there,

I really like this repo and would lik to use it, but I can't. I have installed everything and I started the program with this sell command "sh ./run.sh" (i made the pip install torch and everything). When I am on the web page I cannot load voice models as you can see :

The model selector is blank even if I have downloaded all the assets available on the RVC repo that are indicated. Even if a hit the button "refresh models" the selector remain blank. When I hit this button I have this message "2024-01-10 16:31:35 | INFO | httpx | HTTP Request: POST http://localhost:7865/api/predict "HTTP/1.1 200 OK"" that is written in the terminal. Also I am working in a proper conda environment python=3.10.

How can I use the repo with my apple m3 mac please :)

[sky-admin]

same issue with you

[Skowizzz]

Do you know how to resolve it ? I have tested on my other macs (not M3 this time, only intel), I have the same issue. I suspect not having real GPU make the code not working ...

[haseebsultankhan]

not sure if this could solve your issue. But I had the similar problem once with another project. I used Metal Api with pytorch then everything worked out fine

[Skowizzz]

@haseebsultankhan I tried your method but it did not work :// so sad I can't use RVC 😪

[Skowizzz]

Nobody have any solution?

[michelle-w-br]

same issue

[Skowizzz]

Somebody knows who contact to solve this issue?

[zhh8689]

same issue