Textual Inversion models for SD 1 and 2 which were trained upon all sorts of doll images.
Works with SD 1.5 and some of it's fine-tuned models
dollzy1.pt