个人微调的illu大模型，内含许多动漫角色，所以无需lora就可以直接在大模型中生成，但需要填写相关的动漫人物的英文提示词。

具体情况就看我给出的示例图吧。

提示词提示：前面加人物名字特征、服装、动作和环境，

后续的正面提示词如下：

depth of field,masterpiece,best quality,very aesthetic,alternate costume,alternate costume,2024,2023,perfect composition,ultra-detailed,extreme aesthetic,masterpiece,best quality,good quality,newest,year 2024,year 2023,very aesthetic,absurdres,looking_at_viewer,colorful,

负面提示词：(worst quality:1.2),(low quality:1.2),(normal quality:1.2),lowres,bad anatomy,bad hands,signature,watermarks,ugly,(cleavage:1.3),imperfect eyes,skewed eyes,unnatural face,unnatural body,error,extra limb,missing limbs,painting by bad-artist,nsfw,nsfw,poorly drawn hands,malformed hands,(missing fingers:1.3),fused fingers,too many fingers,