触发词还是和ill版的lora一样的,并没有使用自然语言。在描写场景和姿势的时候还是建议用自然语言描述更好一些。
置顶提示词
aemeath,solo,1girl,头部描写
ahoge,multicolored hair,pink hair,ponytail,sidelocks,very long hair,hair ornament,heart,halo,symbol-shaped pupils,+ +,yellow eyes,躯干和衣物描写
breasts,medium breasts,cleavage,detached sleeves,detached collar,leotard,gloves,covered navel,elbow gloves,strapless,white dress,white gloves,white leotard,下半身描写
white thighhighs,thigh strap,white footwear,high heels,single thighhigh, boots推荐的参数:
底模/checkpoint:WAI-ANIMA
步数/Steps:30
CFG:4
权重/LoRA Weights:1.0
采样算法/Sampler_name:er_sde(normal/beta57)
放大算法:R-ESRGAN 4x+ Anime6B
降噪/denoise:1(高清放大,面部修复为0.5 High-resolution upscaling enabled, facial restoration set to 0.5.)
分辨率/resolution:768*1536、832*1536、1280*2048都可以
建议:
推荐画风lora:
Anime Shiny Skin Style - Anima & Illustrious
https://civarchive.com/models/2597738/anime-shiny-skin-style-anima-and-illustrious
这个画风lora可以让anima原本画风变得柔和、不过饱和,真心推荐
Description
对于眼睛这点的还原程度,anima模型的确要比ill模型要精确一些,可能是我对anima理解不够深刻,个人感觉色彩过饱和了,还是ill风格看着舒服不少,最好搭配其他风格的lora使用






