New style lora for Z-Image, but this time with an Anime style. I've chosen a detailed and colorful screencap anime style but with more mature and realistic body proportions, trained from a dataset of 421 AI images generated specifically for this lora. I am very satisfied with the results.The recommended strength for this lora is between 0.8 and 1, and works really well even with 1 strength. Details start to shine at resolutions of 1440x1440 and above. As usual, no trigger word is needed.

Files with SHA256: b95169670b3f1b940cf8953b5525163d3e4986b7a5f52838047f2e75d46cb0a8