This is just a quick mix I did while experimenting around with merging Loras into checkpoints. VAE is baked in.
The goal was to make a "cleaner" model then my previous ones that goes more into the direction of simpler designs.
The base is my ChameleonAIRPG Mix merged with a bunch of Loras I trained just for this project. They are not available for download anywhere and probably will never be because I just haven't tested them enough for a proper seperate release.
DPM++ 2M Karras, Steps: 25, CFG scale: 7
Hires.fix: 4x-UltraSharp, Hires steps: 15, Denoising strength ~0.45