trained with more images. rewrite training prompts.
My first practice on creating character lora; it's unfinished.
Currently, there are some issues with her eyes under 3:2.
My data needs some enhancement to solve this problem.
