The sample images are baffling: they don't use the trigger words, the prompts don't match the images, and—despite the base model being Illustrious—SDXL was used instead. These are all remixes. I also changed the models and experimented with using—or not using—trigger words.