You know, with the LoRA Creator, this is actually doable. I'll need to recreate the dataset so as to have an accurate guideline and once that's done, load the images and tags into the maker.