This is a style model that does art like the artist Modan. Modan style has changed quite a bit over the time of their work this was trained on their images that were more of a realistic anime style, rather than their artsy, or high realism images. Tends to work best at 0.6. This model was trained on illustrious-xl and generated in Illustrious pencil-xl. Made with the local Kohya tool for AI training. Any suggestions to improve or recommended things for the data set would be appreciated.
Artist twitter:
https://x.com/0_modan?lang=en
Description
I wasn't planning on doing another style so soon, but I have tried to do this style for the artist in the past but their weren't enough images for training. Its still on the small side but I learned their was enough now. So after collecting the new new data and revising my old data I made this. This lora is made from 71 normal images and 94 cropped heads. Optimized for Illustrious 1.0.