用戶LoRA
Trained on couple hundred images of inkplasm / welwraith style, creates pretty flat images. Can use lineless to generate more vector like images. If you are having issues with it generating masked characters, would suggest adding human to prompt.
