Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Can't get Simswap to work after upgrading PC #452

Open
FrisbeeDog1 opened this issue Oct 22, 2023 · 7 comments
Open

Can't get Simswap to work after upgrading PC #452

FrisbeeDog1 opened this issue Oct 22, 2023 · 7 comments

Comments

@FrisbeeDog1
Copy link

Hey,

I setup Simswap about a year ago following this guide: https://youtu.be/1NWJz_rFMq0?si=SkbmKXk9gZLrabpt
It worked perfectly up until I upgraded my PC.

I've seen online that many people had issues after upgrading their GPU, I went from a 2070 Super to a 4090.

My original issue was that Simswap would get stuck at "END" when trying to process a video, but in trying to fix it I've faced many other issues.

I've tried starting from scratch many times following the instructions here: https://github.com/neuralchen/SimSwap/blob/main/docs/guidance/preparation.md
And the original YouTube guide listed above, but I always run into errors now and have no idea what I'm doing.

Is there an updated list of instructions I could use to get this working with my upgraded PC?

Thanks.

@instant-high
Copy link

Have you found a solution?
I've upgraded today from gtx1050 to rtx3060 but have not tested.
RVC working "out of the box"
Tpsmm needs to update cuda I think

@FrisbeeDog1
Copy link
Author

Unfortunately not, posting here was my last resort but no one has got back to me on this :/

@instant-high
Copy link

Next weekend I will setup the new pc.

@instant-high
Copy link

@FrisbeeDog1

Ok,
I just got another project working on RTX3060 (Thin plate spline motion model modded for face swap)
Simswap will be next...

  1. You can try creating anaconda environment version 3.8 instead of 3.6
  2. pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117

How it worked for the other project:
I have created an environment 3.8
then copied the whole existing old environment into this new folder but kept existing files!
At last pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117

I have not tried onnxruntime-gpu (in my other project I use same detection/alignment as simswap)
onnxruntime is fast enough so far

@instant-high
Copy link

@FrisbeeDog1

I've got it working on new RTX3060, maybe this will help you:

conda create -n cuda python==3.8
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117
pip install imutils==0.5.4
pip install numpy==1.21.6

I also had to update ./modules folder and arcface_checkpoint.tar

I'm not sure which version of cuda you need for 4090

Thomas

@FrisbeeDog1
Copy link
Author

Thanks for the updates!

I'll be sure to try this soon :)

@neuralchen
Copy link
Owner

What is your problem?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants