The recommended weight is 0.4-0.6, it is recommended to use HD repair to get a better experience, high weight will be close to the fandango style of painting, it is likely to iterate a new version when needed.
Trigger keyword: nayutaren,
Strongly recommended tag (please refer to the example): nayutaren,roswaal mansion maid uniform,blue hair,short hair,blue eyes, hair ornament,hair over one eye, ribbon,
If you want Rem to be bare necked without things like bows, ties, collars, you can use negative cues like: (ribbon,collar,choker,bowtie:1.5),
推荐权重为0.4-0.6,建议使用高清修复获得更好的体验，高权重会接近番剧的画风，很有可能在需要的时候迭代新的版本。
触发关键词：nayutaren,
强烈建议的tag（请参考示例）：nayutaren,roswaal mansion maid uniform,blue hair,short hair,blue eyes, hair ornament,hair over one eye, ribbon,
如果你想蕾姆是裸脖子而没有诸如蝴蝶结，领带，衣领，可以使用负面提示词例如：(ribbon,collar,choker,bowtie:1.5),
可以在这里抢先试用我的新室内模型，测试好了会发布Civitai❤️https://tusi.art/models/605426449836443140
Description
a test version
FAQ
Comments (8)
Do you have the prompt information and model used for the first 4-5 pictures? Thank you!
Hello, I'm glad you like my Lora model, but unfortunately the picture you need is my test version of the unpreserved tag (I'm used to saving jpg format), I've attached some pictures that can be referred to in the comments section. picture 4 can try to add keywords such as: "nude,nipples,onsen." picture 5 can try to add keywords such as: "on bed,lying,navel,white pantyhose,"
What model did you use to make the images?
I commonly use a few: DONKO-mix, Facebombmix, counterfeitV2.5, the sample chart uses the first two
What did you use for base model? like sd-1.5? or anything3?
Example figure using donko mix,FaceBombMix,counterfeitV2.5,training based on sd1.5.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.