voice-changer
voice-changer copied to clipboard
w-okada
Docker Model stuck on "Booting PHASE :MMVCServerSIO"
Issue Type
Bug Report
vc client version number
v.1.5.3.7 (just git cloned the repo)
OS
Gentoo Linux
GPU
AMD RX 6600
Clear setting
no
Sample model
no
Input chunk num
no
Wait for a while
yes
Voice Changer type
doesn't even load up
Model type
default with the docker model
Situation
As I mentioned in the title, upon running
USE_GPU=off bash start_docker.sh
(I use it with the GPU turned off since running with the GPU gives me an error which I will not bother fixing since AMD is not officially supported)
It keeps stuck at: Booting PHASE :MMVCServerSIO (I've left it running for about an hour)
please help me fix this as I believe this is a really interesting project!
UPDATE: Upon running with Conda I get the same results, I'm not sure if this is a me error or if this project has compatibility issues with Gentoo
I am also experiencing this at the moment running the docker container on ArchLinux
RTX2080 with Nvidia-Container-Toolkit installed
I'm having the same problem running the Docker container on Arch as well. For now, I'm currently running it through Wine, and it works there.
Same here... just having it stuck in the booting phase. I had alread installed it a couple weeks ago but now nothing works
uuu,mmm.
I test on ubuntu, debian, wsl2. But I have no environment of arch, gentoo. I have no idea.
uuu,mmm.
I test on ubuntu, debian, wsl2. But I have no environment of arch, gentoo.
I have no idea.
Try to setup an Arch vm or run on top of WSL. If it works on Arch, chances are it will also work on Gentoo.
One possibility could be that freeze_support() is causing this issue. For those who can try it on conda, could you please open the file and try to comment out the code?
https://docs.python.org/3/library/multiprocessing.html#multiprocessing.freeze_support
I may be late to this, but I tried commenting out mp.freeze_support() as you suggested, and it doesn't fix the problem.
However, using the HTTP protocol DOES work for me. I believe this should also fix the problem of the program not running from Docker. I am not sure what is causing it to fail with HTTPS, but I am running it from Anaconda.
Anyway, to fix this, change the lines
--https true
to
--https false
and type localhost:18888 in your browser's URL address. I was able to get output past the booting phase and therefore the GUI.