thesomeotherguy

can confirm this. hires fix now is more sensitive than before, I have to dial it down the upscale factor in hires fix to overcome VRAM out of memory, which...

> Can you try different `Cross attention optimization` options on the `Optimizations` settings page? If you had none specified before that's Doggettx now. it seems reinstalling a1111 from scratch to...

faceswaplabs also can do this, it can create face checkpoint from image file(s). https://gist.github.com/glucauze/4a3c458541f2278ad801f6625e5b9d3d https://github.com/glucauze/sd-webui-faceswaplab/blob/main/scripts/faceswaplab_swapping/face_checkpoints.py it improves face similarity **_significantly_**, because it can learn facial features from multiple images with...

> I just figured out a workaround for this for anyone interested. I've only tested it twice but it worked both times. Let's say you want to use 4 face...

> Have you tried with 2 completely different faces yet? The first thing I tried was mixing 2 different people and from what I remember I ended up with a...

This feature seems to be already implemented in the ReActor version of A1111. It can be found under Tools > Face Models > Blend. This feature is great because it...

> Seems to be using CPU to do all the swapping, do you know how I can switch it to GPU? it's not worth it (for now), I tried, it...

I forgot something, change this also on \stable-diffusion-webui\extensions\sd-webui-roop\scripts\swapper.py `providers = ["CUDAExecutionProvider"]` it's located after `from scripts.roop_logging import logger`, and before `@dataclass` save file, and delete `__pycache__` folder if it's exist....