Faceswap Forum

In my opinion, the most likely cause of this is a conflict with an alternative python version installed Tensorflow. If you have cleared out everything as per the FAQ point (i.e. all existing python installs and temp folders) and are still getting the error, then you're going to need to find support ...

I know a lot of people have requested this, and I probably should do it.... but time/money etc.... The best way to see what has changed is look at the git history, although I appreciate that it can be overwhelming/hold a lot of information. I do try to keep it as descriptive as possible. https://git...

How are you trying to run faceswap?

This looks like a Tensorflow issue....

I would delete the faceswap folder, uninstall miniconda 3 then try running the installer again....

If you still have issues, so this FAQ:
app.php/faqpage#f1r1

If there is demand, we may consider it, but it's not a priority at the moment. Of course anyone is welcome to develop plugins and raise a Pull Request.

Honestly? Face Filter is not hugely reliable. At the moment it uses the (not very good) cv2-dnn detector/aligner to get the faces (for vram reasons) which means sometimes it just inputs junk.... It is on my (lengthy) TO DO list to look into this, but I recommend just extracting all faces and then us...

A dark theme for the app is not straight forward.

We use a GUI Tool called tkInter, so we can plug into the OS's window manager. Unfortunately different OSes handle displays differently, so creating custom coloring is tough without it looking janky.

The code which does the rotation/warping is referenced from lib.faces_detect https://faceswap.readthedocs.io/en/late ... etect.html

Json file contains where the face and landmarks are in the original frame. The cropping of the face from the frame is calculated from this information

There is a training guide coming soon. In brief: Loss The loss function to use. There are various different ways to calculate Loss, or for a NN to discern how well it is doing at training a model. I won't go into details of each of the functions available, as it would be a somewhat lengthy process a...

Yeah, unfortunately this is a limitation of OpenCV (What we use to display the windows on the screen).

A complete reworking of the manual tool with a proper GUI is on the TODO list, but it is some way off.

Hmmm, changing the batchsize is unlikely to cause shutdown issues. Lowering the mem clock speed is more likely.

However, the important thing is it works now

8 Photos is not nearly enough. See https://faceswap.dev/forum/app.php/faqpage

Secondly, you will be training on CPU, which will take weeks.

Thirdly, you have setup Faceswap wrong. You have said "Yes" to AMD somewhere along the lines, when you should have said "No".

plaidML actually should work with intel Iris. Whether it works well or not, I couldn't say, and we certainly don't directly support it. However unable to run OpenCL kernel kernel_c26_sdk_1: CL_MEM_OBJECT_ALLOCATION_FAILURE Suggests not enough VRAM available anyway. With that said, you have set up Fa...

Maybe overheating?

Honestly, it could be a myriad of things. Faceswap taxes your GPU HARD though. People do train on that card without issues, so unfortunately it will be hardware related. Beyond that, I can't really offer any more help.

If your PC shutsdown it's basically a hardware issue in some way.

This is a hardware issue. Most likely insufficient power.

NB: This guide was correct at time of writing, but things do change. I will try to keep it updated. Last edited: 23 August 2023 --- Contents Introduction What is Training? Overview Terminology Training Data Choosing a model Model configuration settings Global Settings Loss Settings Model Settings Tr...

This is an annoying inconsistent issue with Tensorflow and some GPUs.

If you update to the latest code then

GUI:
go to Settings>Extract>Global and enable allow growth

CLI:
Go to faceswap/config/extract.ini and change the allow_growth to True

This should resolve the issue.

Faceswap Forum

Search found 2702 matches

Re: Error on launch

Re: What's New ? (suggestion)

Re: crash on launch

Re: Error on launch

Re: DeepFaceLab SAEHD

Re: Trying to filter faces during extract

Re: Masking Hands over Face?

Re: Info in json files.

Re: Info in json files.

Re: Loss Function

Re: How to use manual alignments tool?

Re: PC crash

Re: ValueError: cannot reshape array of size 4050 into shape (4,4,3,5,5,3)

Re: CL_MEM_OBJECT_ALLOCATION_FAILURE

Re: PC crash

Re: PC crash

Re: PC crash

[Guide] Training in Faceswap

Re: How to use manual alignments tool?

Re: Could notExtrace: create cudnn handle: CUDNN_STATUS_ALLOC_FAILED