The model I think is fairly self explanatory.

Works well at a strength of 1.

It's a bit messy to use 'just the style', since some of the character elements like the hat seem to pop up even if you disregard the style tag and use the lora incognito without it, but it should be simple enough to put a few minor things like that in the negative while using the lora without the style tag to manifest the style without the character.

*v2 version solves the above issues from 1.0.*

Previews made with novaFurryXL_illustriousV40.