This is a lora trained on an entirely synthetic dataset for Harley Quinn, particularly one specific outfit from the Birds of Prey movie, though it should still be flexible enough to do others. It should also work with other art styles, loras, and concepts.

It will not generate anything photographic though.

I specifically trained it on this outfit because I've trained other loras for the other specific ones, in order to avoid bleed by training too many things into this one.