I updated the dataset and added 40 more images to improve the model and make it easier to generate! :3