The model I think is fairly self explanatory.

Works well at a strength of 1. Sort of. Experimenting with negative prompts of monochrome and greyscale to force color, for some reason has issues with this one, might be because there was basically no color pages in the dataset, but even at a strength of .8 in an attempt to correct for it doesn't seem to help, sort of a 50/50 chance to generate an odd 'red' image, an image that is mostly red, not sure why that happens or how to control for it, but if not trying to generate color images it works pretty good.

Previews made with novaFurryXL_illustriousV5.0