I trained this model on the character Mochizuki Sayoka from the visual novel Rape! Rape! Rape! The model was trained only on the standing sprites and HCGs so it might be a little finnicky or hard to work with but hey what can you do. You might need to use Adetailer and Hi-res Fix to get good results. You might have to use more emphasis to get certain images to be generated and use negatives to prevent things from showing up.