Tries to capture the style of the artist Shiratama (monster1553). I think it tends to bring out some very nice background art from some models (despite the dataset mostly being white backgrounds, weird...) <ul><li>No trigger word.</li><li>Seems to work fine on most Illustrious/Noob-based models around 1.0 strength.</li><li>Smallish dataset focused on futas, so it tends to be good at that.</li><li>Struggles with messed-up eyes occasionally - use hi-res fix or ADetailer.</li><li>Use "mosaic censoring" in the negatives if you're getting censored naughty bits. </li></ul>

Files with SHA256: 1c33f98bf456fa44c0f0fcddc8b5bbf44c75f79bace1d127c720e206552722bf