ComfyUI
ComfyUI copied to clipboard
comfyanonymous
how can i use fooocus_inpaint model in my inpaint workflow
The inpainting functionality of fooocus seems better than comfyui's inpainting, both in using VAE encoding for inpainting and in setting latent noise masks
inpaint foocus patch is just a lora, Modify the correct lora loading method, just copy the way fooocus loaded. for inpainting head im not sure.
inpaint foocus patch is just a lora, Modify the correct lora loading method, just copy the way fooocus loaded. for inpainting head im not sure.
Have you tested it yet? Can it be used?
not yet,I read the codes of both and they are basically the same, but I don't have a good way to modify the code. Fooocus redefines a lot of the underlying code.
It's a somewhat custom Lora format, so the loader is modified, but also ComfyUI's ModelPatcher needs to apply the weights differently than other Lora. Makes it a bit ugly to implement, but here is a first version: https://github.com/Acly/comfyui-inpaint-nodes
Works well so far, but still experimenting.
