Shakker.ai’s IP Adapter (style transfer)

[evelryu]

Can this be implemented as a controlnet feature or other kind of implementation?

https://www.reddit.com/r/StableDiffusion/comments/1gr0d6w/just_tried_shakkerais_ip_adapter_customizing/

[AnabelleKM]

Style transfer is already possible by using ControlNet. Example for SDXL:

ControlNet:		IP-ADAPTER
Preprocessor:		CLIP-ViT-bigG
Model:			ip-adapter_xl [4209e9f7]

Control weight: 	1	(How aggressive you want the style transfer to show up in your image. Adjust to your liking)
Timestep Range:		0-1	(0 % to 100 %. / Start of image to end of image generation: At what percentage of image generation the style transfer should should take ffect. Adjust to your liking)

On this page is a download link to the file CLIP-Vit-bigG.safetensors
(Be sure to name it like that when downloading)
https://github.com/cubiq/ComfyUI_IPAdapter_plus/blob/main/README.md
Download to folder \webui\models\ControlNetPreprocessor

You also need this file "ip-adapter_sdxl.safetensors"
(https://huggingface.co/h94/IP-Adapter/resolve/main/sdxl_models/ip-adapter_sdxl.safetensors).
Download to folder \webui\models\ControlNet

Start / Restart Forge to see it in the list of Preprocessors.