Reusing my pony lora's training data to make an Illustrious version. Still testing, this iteration shows alright results so far.
Description
Check my previews for prompting style, but can summed up with
POSITIVE: Basic Illustrious style tags, painting, traditional media, realistic ***SOLO or SOLO FOCUS***
NEGATIVE: Illustrious style tags, ***Anime*** (helps to get things to look like my preview, bad face, bad teeth (holy crap teeth are sometimes bad)
All of my images were genned in 30 steps, with the Impossiblepie models. Some with Adetailer, and some without. Please pay attention to solo vs, solo focus. Background characters can be bad quality/bad anatomy.
I was messing around with 3.5-5 CFG. But feel free to change it up.