This is a <a rel="ugc" href="https://github.com/p1atdev/LECO">LECO</a> lora trained on SDXL's concept of "elegant". I haven't figured out if there's a way to do this without a trigger word yet, so for now "elegant" is the trigger word for this LoRA.Note that this model also does inelegance very well if you give it a negative weight, but you still need the word "elegant" in the positive prompt in order to trigger it.I've tested it successfully on people and scenery. It was trained on base SDXL 1.0, so it should be pretty compatible across the board.

Files with SHA256: ec8db95fd3807c9bf2625bde180d911c9697a43a3b150df8fbddbaa18b61d8b6