The new version is designed to be more diverse, and has been trained on more face types and jawlines to reduce the impression that all images are of the same person. Further improvements are planned
模型描述
Inspired by Suyohara style and a few other semi-realistic styles
