Originally uploaded by Mr. Food - Reuploaded with permission
My first LoRA, but I'm pretty satisfied with the quality, it's surprisingly flexible. This LoRA was trained on 33 images made for a game called LISA: The Painful. All images were hand captioned and trained on Stable Diffusion 1.5 - Link to all images it was trained on: http://www.lisatherpg.com/gang.html - (Note: I used the high res artpack versions of these images)
Example Prompt: painfulstyle, a man wearing a poncho
512x768 - cfg 8 - clip skip: 2 - other settings to taste