Trained using anyLora with official artworks and anime screen scrapes. This model should be quite faithful to the original art style. It is a flexible model that should work with all anime based checkpoint, and should be compatible with clothing and concept LoRAs.