Colorful, flat cartoon and anime styles built into one system. its trained on an older dataset of about 550 images with a heavy focus on portraits.

This model is trained on roughly 550 images of colorful flat cartoon and anime styles. It leans towards flat anime and gives a very powerful result. I recommend starting at lower levels and working your way up. This favors a flat anime style that tries to get away from the plastic highlights you see in many modern styles. I really like running this between 0.4 and 0.8 strength at 2.0cfg for 20 steps to get good, high-quality results. While this model does have a trigger phrase, it will trigger on most things related to anime, cartoon, or illustration. I have noticed around 0.75 strength there is a very strong shift in output.