V2==========Thanks to mikuhatsune's GPU.Fix some promblems.（but it is still not easy to get white thighhighs)Trigger Words have changed.<s>After using NAI3(SDXL) I do not want to use sd1.5 anymore which is full of drawbacks such as overfitting and poor language comprehension ability.</s>Well sd1.5 is still worth using. After all it has so many models, and is easy to use.感谢mikuhatsune的显卡。修了一些问题，但是白丝还是不好出。触发词改了。<s>曾经沧海难为水，用了NAI3(SDXL)就不想再用充满缺点的sd1.5了，各种过拟合和低下的语义理解，不想碰力.</s>嘛1.5还是值得一用的毕竟模型多好上手。V1==========I made Shoukaku (Race Queen)‘s lora long time ago. Well at that time I was really a noob who did not know how to handle with datasets and trigger words.This lora is for cosplaying Shoukaku and Zuikaku (Race Queen). If you just input "white hair, very long hair", Shoukaku might show up(Shoukaku's hairstyle can not be recognized as a tag in tagger, so you might got a tiny side ponytail sometimes. It's suggested to add "side_ponytail" and "one_side_up" to Negative Prompts if you do not want that.很久以前我做了翔鹤（赛车女郎）的lora。不过当时我是纯萌新菜鸡，不知道如何处理数据集和触发词。这个lora是用于cosplay翔鹤和瑞鹤的。如果你输入"white hair, very long hair", 翔鹤可能会出现。翔鹤的发型，tagger打不出标来，所以有时候你可能会得到一个小小的侧辫子。如果你不想要的话，建议把"side_ponytail" 和 "one_side_up" 扔进反咒里。

Files with SHA256: f23060b8b1d218fa635689e4f78625204c02acfd5dc93330f434e78699f77537