Works with SD 2.0, SD 2.1, SD 2 Depth model for Image2Image, 512x512 and 768x768 models
Textual Inversion models for SD 1 and 2 which were trained upon all sorts of doll images.
