Add linux support #475
Comments
|
I believe that Linux users can use the Anaconda environment. |
|
I followed the installation guide, however, I used venv instead of conda. So i tried degrading the protobuf package to |
|
how about Set PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python |
You mean as an environment variable? |
|
yes error log said
|
|
ye i get that anaconda and docker is available, but its very hit or miss especially considering the video tutorial is made for wsl and the cloned binary does not give an amd version (linux users use amd alot) i did try to get it running, but anaconda was a bust because pyworld wouldn't build the wheel when installing requirements.txt, and docker did load up a web Gui but that is really as far as i was able to get since other than settings changing audio wouldnt record and the list for voice models wasn't visible |
|
I am trying to set it up to work and it is pain. So far I have been manually installing python packages for last 15 minutes, bcs for some reason they are not included in requirements.txt. It just crashes on runtime that packages are missing so I need to install them by myself. After spending some time figuring out that module socketio is not socketio but python-socketio, I finally got it run. |
|
I got it working today with the As far as setup is concerned, yeah lacking Linux support is sad, and although i was able to run the docker container, there were simular errors and i wasn't able to get to a working condition with it. |
How did you set the ENV variable? Im currently stuck in the webui. It failes to initialize with "no error message" |
Humm... Not sure what the problem with your machine might be. |
I'm having the exact same issues, it downloaded all the models fine I have it in it's own conda environment and installed all the dependencies as asked but it get's stuck after completing boot with no errors and the website gives the same error as theirs. PC Specs:Arch linux kernel 6.4.3-zen, RTX 2080 TI + GTX 1060 6GB, Ryzen 5900x, 32GB 3600MHZ ram with Nvidia proprietary drivers. Log with command run to launch it: |
Simply bump pyworld to version 0.3.4 in the requirements file. It works that way. I didn't have to set any protobuf environment variables. But server mode on Linux felt really slow though when I tried it, this might be a setup issue on my end though. |
Same issue even after updating pyworld |
|
I tried commenting out mp.freeze_support() as well as setting https to false as suggested in #377 with no luck on Conda, what's strange is before the error page I see the actual webui load for a split second then switch. I also tried running it with PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python to no avail. |
|
After running git pull to update it I am still experiencing the issue but this time with an error message on the site! The bottom error |
|
After switching browsers as seen in #587 to a chrome based one the voice changer works! I'm unsure if it was due to a repo update in Arch or something Okada did, but it seems to be working for me using Anaconda on arch now. For anyone else experiencing this issue Arch steps (for conda atleast): |
|
I followed that guide but still idk where to launch it, build version (like windows and mac os) would be awesome |
|
|
@w-okada since no one else is commenting and the issue doesn't seem to appear anymore do you want to close this issue? |
|
Yes, I'll close this issue but I'm truly, truly grateful for everyone's discussion. And Linux is the best. |
|
For anyone that is trying to install this on linux with pyenv, I ran into an issue related to pyworld which is solved by using This is how I was able to set it up with the fish shell: pyenv install 3.10
pyenv global 3.10
python3 -m venv venv
source ./venv/bin/activate.fish
python3 -m pip install -U pip numpy wheel
cd server
python3 -m pip install -r requirements.txt --no-build-isolation |
|
Hello. I'm using Anaconda. It is running without any errors, but I cannot select my GPU. I am using an AMD 6700 XT and I am hoping to make use of the ROCm support on Linux. Any ideas? I can only see my CPU. Additionally, I get the following error after selecting a model: I am on Arch Linux. |
|
So after installing fairseq and pyworld using pip, I am able to use the voice changer client without issue except that it still doesn't show my GPU. |
Issue Type
Question
vc client version number
It doesn't have linux support
OS
Kinux
GPU
Linux
Clear setting
no
Sample model
no
Input chunk num
no
Wait for a while
The GUI successfully launched.
read tutorial
no
Extract files to a new folder.
no
Voice Changer type
Linux
Model type
Linux
Situation
I am writing to request that you add Linux support for our application. I have been a Linux user for many years and am a big fan of our application. However, I am disappointed that there is no Linux support.
I know that there are many Linux users who would use our application if it were available to them. Linux is a popular platform and the number of Linux users is growing every year.
Adding Linux support would be a great move for our business. It would allow us to reach a wider audience and grow our user base.
Thank you for your time.
Sincerely,
Raul Popescu
The text was updated successfully, but these errors were encountered: