A LoRA trained on Shuffle aka Songdatiankong's art, including doujin book panels. Mostly trained on otoko no ko and yaoi content, but I also included some of their girl focused works in the training data.
Description
Version trained on the V-pred-test model, but works on V-pred-0.5 from what I tested. NoobAI and Illustrious in general knows Shuffle's style already, but it's in my experience weighted more towards their older work and has a yellow tint to it. This lora was trained on their more recent illustrations and doujin works and should provide a more "modern" look. This is my first attempt training on vpred, so any feedback is welcome.
Adding "pointy ears" to the negative might be helpful as the dataset consisted mostly of elf bois.
FAQ
Comments (1)
thanks for your lora, I like femboy