新的anima版本，用的模型是HomoFidelis Anima
可用提示词：(@photo:3), realistic, photorealistic,
1boy, handsome, athletic build, erotic, fit body, lean body, manly, solo, male focus, black hair, black eyes, asian, gorgeous boy,
请多多返图，这对我炼下一个更好的模型来说非常重要！！非常感谢您！
prepareing... plz give me more buzz
Please return to the drawings more often, which is very important for me to refine a better model! ! Thank you very much.
Description
FAQ
Comments (12)
关注了，在炼丹中出现黑白照频率有点高，加到负面条件里提权之后才好些；脖子有时会过长，眼睛会出现变形，手部的指甲变形，衣服穿不上身（后来发现是highres textures这个提示词），奇怪的是fullbody似乎无效，不写feet的话就怎么都不会出现脚部。
总的来说还算是很帅很好看的，腹肌和性器官上色都挺不错，不会出现断层隔离颜色的问题。另外这好像是单人效果最佳，双人多人的时候肢体会严重变形。
还有全身或者七分镜头的时候，鼻孔有时会多一个。
只用了pony V6，其他lora无，文生图。
@wjxdjj 推荐直接用DUCHAITEN REAL
@Herryomoss 大佬那个要buzz，等十多天之后我再用吧
Does not seem to work consistently. Often the men that result are not even Asian. Appears that it might require a trigger word but no trigger word is provided. The recommended CFG is also way higher than one should use for Flux. It's not clear this is really a Flux LoRA.
The model really doesn't need any prompt words, it just loads and takes effect, you can use my image thrown into comfyui to add to the workflow to see
'Asian' or country name of east asia should be the trigger word, and after activating LORA for comparison, there is a significant improvement effect on the original image.
What is the trigger word?
No keywords are needed. If you have to add them, they can be "black hair, dark eyes"
你的脸怎么都是差不多的，我生成的脸总是崩坏
我用的是comfyui，可以直接用我图片的工作流看看