This LoRA attempts to emulate Shiroi Yoru's artstyle.
The trigger is by Shiroi Yoru.
The dataset was partially uncensored. Consider positive prompting uncensored rather than uncensoring through negatives because I suspect it's going to get better results.