Princess Luo Shiyu from "I’m Actually a Cultivation Bigshot" for Illustrious

Be sure to use low CFG scale (2.0 - 3.0) for best results

Training info:

Trained on a little over 90 cropped and adjusted images. Throughout the webtoon the style differs a bit, starting very simplified and, after hundered chapters, improving in quality, and this character is introduced at the very beginning and doesn't really show up later, unfortunately. But there were plenty of detailed shots with different poses and emotions, so the LoRA can do a lot and presents her consistently enough (mainly from the higher quality shots from later chapters).

Generation info:

For some of the showcase pictures I partially tried to recreate backgrounds from the webtoon (e.g. simple drawings of nature, non-ideal outlines everywhere), so some prompts are set to generate not so high resolution high quality nature, architecture, etc., like it is in the original (not a slander, just enjoying the character of this style in generating :)). Long story short - wanted to recreate the style, not sure if needed that many descripting prompts to get it though. You can follow my efforts further from the generation data, or try "best quality" and lower LoRA strengths for less faithful but more detailed and free to manipulate image. Also, don't forget to increase denoising strength if it's set low, so it generates less simplified image. Also ADetailer for face, in which I also included some custom prompts that should carry over after copying generation data.

Enjoy and feel free to post your pictures!