Emmmm，我知道符华有关“兔女郎”的服装一直有着巨大的争议。但我个人是觉得阿符的这一套很好看，要是被埋没了会挺可惜的，还有想着可能有人跟我一样也挺喜欢的，所以还是加入进来了。

希望各位不要吵架，谢谢。

玄衣素裳的prompt为：

Fu Hua(Rustic Noir),shirt,white shorts,choker,black footwear,hair ornament,belt,leg rings,bracelet,black sleeve,braid,

（其中”braid,”是符华脑后的那一段小辫子，如果仅生成正面图无需加入到prompt中）

兔女郎服饰的prompt为：

Fu Hua(Bunny),leotard,rabbit ears,pantyhose,wrist cuffs,black high heels,

推荐权重为0.75-0.95（越高细节还原的越好）

此模型基于“animefull-final-pruned.ckpt”，不同大模型适配程度见图9。

你可以把“face,”也加入到关键词中去，这样能在一定程度上加强脸部的细节。

想要换装或nsfw，推荐将lora权重降到0.6左右。（可参考图7-8）

For the " Rustic Noir " prompt:

Fu Hua(Rustic Noir),shirt,white shorts,choker,black footwear,hair ornament,belt,leg rings,bracelet,black sleeve,braid,

(Please note that the “braid” refers to the intricate segment of hair cascading down Fu Hua’s nape. However, if your intention is to generate only a front view, you can omit the mention of the braid from the prompt.)

For the " Seagull's Soar" prompt:

Fu Hua(Bunny),leotard,rabbit ears,pantyhose,wrist cuffs,black high heels,

Recommended weight:0.7-0.9(The higher the weight, the better the level of detail in the restoration.)

This model is based on the "animefull-final-pruned.ckpt",and the adaptation degree of different models can be seen in Figure 9.

Including the keyword “face,” in your prompts can enhance facial details to a certain extent.

For alternate outfits or NSFW content, it is recommended to lower the Lora weight to approximately 0.6 (refer to Figures 7-8).

English translated by ChatGPT.