Trained on 2 Large Datasets with captions, One was Real Anime Art, the other was Generative Anime Art
