Faceswap Forum

You should edit the video or images before extract, NOT the extracted faces.

If you're doing images, then you can't use the frame-range tool. Instead just create a folder with just the images you want to convert as the input.

No, DeepFaceLabs is a different project and is not compatible with FaceSwap or it's models.

Faceswap does not multiply your batch size. That's up to you. If you set the Batch Size to 16 and have 4 GPUs, it'll split that so that each GPU has a batch size of 4.

No, this is just AR and some clever transforms.

frankr69 wrote: ↑Thu Jan 27, 2022 11:41 am

Hi, is it productive to buy a new RTX 3050 8gb?

Should be fine (if you can get ahold of one).

You can reuse extractions. Just specify the folder containing the already-extracted faces on the next training.

lhchsqwct wrote: ↑Fri Jan 14, 2022 1:32 am

It's solved, I removed all the .decode("utf-8") field and it works now, I think it's trying to decode the already decoded part.

Don't do this, it prevents updates.

This however should now be fixed. Torzdf had pushed a fix. Just reinstall to pull it and it should be fixed.

The problem is some lib has changed or you have installed the wrong one. We need to see what's causing the problem, so please give that sysinfo.

This seems to have just started hapenning. Please try the following to get us the information we need to troubleshoot this issue.

From inside your virtual environment, inside your faceswap folder, run:

python -c "from lib.sysinfo import sysinfo ;  print(sysinfo)"

My recommendation for large videos is EEN=10 or so. I only ever extract every image when I'm preping convert.

It sounds like it's not using the GPU.

app.php/faqpage#f1r1

Yes Faceswap can handle images instead of videos.

Quality issues are almost always with data. Unfortunately it's impossible to guess what might be wrong without examining your data or results. If you can, please post a training preview so we can evaluate your data and results.

viewtopic.php?p=6454#p6454

It looks like it was successful. See if everything runs correctly.

That is entirely a network issue. If your network is stable and you can access the anaconda.com you should be able to download without problem.

If you keep getting problems there, that's an anaconda issue and it's not anything we can fix here.