The LoRA supports geometric styles, responding well to a variety of prompts to nudge it in the right direction. It has a curious habit of producing unexpected results, reflecting the varying geometric images on which it was trained.
Have fun, explore, and post some images to demonstrate its capabilities. Thanks for giving it a try.
Description
The LoRA was created using the raw version. It is suitable for both raw and turbo versions. All showcase images were generated using the raw model.