tencent-ailab / IP-Adapter

The image prompt adapter is designed to enable a pretrained text-to-image diffusion model to generate images with image prompt.
Apache License 2.0
4.5k stars 296 forks source link

why sdxl_plus use StableDiffusionXLPipeline while sdxl_plus_face need a custom pipeline - StableDiffusionXLCustomPipeline #280

Open akk-123 opened 5 months ago

machengcheng2016 commented 4 months ago

Hi @akk-123 , by checking the differences between StableDiffusionXLCustomPipeline and official StableDiffusionXLPipeline, i think the only thing matters is that StableDiffusionXLCustomPipeline utilizes control_guidance_start and control_guidance_end to decide which period of inference will I use image conditioning (here).