사용자 LoRA
High res'd the images for better training, best to play around with which model works best for you
refizez
Dan Decarlo