First version of Navi LoRA (SDXL).

Trained with 4 high-quality images to capture the core shape, proportions, and design.

Best for standing poses / character reference / body shape design.