LoRA to put your subject in a harness as text-2-image or as image edit.

Works really well on Flux2-base-9b at strength 1.0 but for distilled versions of the model you may need to push the strength up.

It was trained without a trigger word but captioned with "Jasmine Harness" so that works well. e.g. "image of a woman wearing a jasmine harness..."

For image edit it helps to describe the general position of the subject in the photo you are editing, e.g., "a woman reclining with her legs spread. make her wear a Jasmine Harness." I trained it using control images so you should not need to in-paint, the lora won't change subject except to add the harness.

Feedback appreciated and have fun with it.