.π₯ liquid film π₯. π§
This model is a sort-of experimental model trained on both my Liquid Lemons and Film Photography dataset. It helps push Z-Image to a more interesting look. I haven't tried this on Turbo but theoretically it should work there, too!
No trigger words are needed but "color film photograph of..." works well! This model was trained without human subjects, so it should have high compatibility with character loras.