An attempt at a general "concept and style enhancer" for furry models. V1 is trained from basic Illustrious V0.1 with a dataset of almost 100 images for 10 epochs. The training data is tagged with a visual model, with no artist names. Images with text are partially filtered out but there's still a few in there to help with the negative prompt. I used a variety of different images in different styles with different species and situations to increase its variety. Also better at generating intersex characters and anatomical correct-ness! Prompt with E621 tags with spaces instead of _, and \ before parentheses! e.g. furry, female, anthro, brown fur, lying, on back, pokemon \(species\)