--- license: apache-2.0 pipeline_tag: text-to-image tags: - stable diffusion - ip adapter --- # IP Composition Adapter This adapter for Stable Diffusion 1.5 is designed to inject the general composition of an image into the model while mostly ignoring the style and content. Meaning a portrait of a person waving their left hand will result in an image of a completely different person waving with their left hand. ### SDXL SDXL variant is cooking. Follow me on here or [Twitter](https://twitter.com/ostrisai) for updates when it is avaliable. ### Thanks I want to give a special thanks to [POM](https://huggingface.co./peteromallet) with [BANODOCO](https://huggingface.co./BANODOCO). This was their idea, I just trained it. Full credit goes to them. ## Usage Use just like other IP+ adapters from [h94/IP-Adapter](https://huggingface.co./h94/IP-Adapter) for SD 1.5. Use the same CLIP vision encoder ([CLIP-H](https://huggingface.co./h94/IP-Adapter/tree/main/models/image_encoder)) ### How is it different from control nets? Control nets are more rigid. A control net will spatially align an image to nearly perfectly match the control image. The composition adapter allows the control to be more flexible. ## Examples 1 2 3 4 5 6 7 8 9 10 11 12