用戶LoRA
Trained from 20 images in OVA video.Clothes and body shape are not trained, sorry.
sdxl-sylvia-nimrod