A character LoRA for Okusawa Misaki
Training model: Animagine XL V3，
Inference model: Animagine XL V3
Training resolution: 1024*1024
