免责声明
仅供交流学习，不允许商用和个人商业用途，不允许转载和销售，后果自负，与本人无关,
LORA权重：0.6-1
制作使用COMFYUI，自带KS，加一个2倍放大，然后加controlnet,LORA只用了我的这个LORA，权重1，不要在问我怎么效果不一样，你自己按照这个做肯定一样,别发私信在问我了，我不可能天天的跟每个人解释,主模型用的CN，
这个LORA是某一个Virt-A-Mate作者的风格模特，我感觉主要还是偏向于脸部，因为Virt-A-Mate的色彩真的没啥可说的，所以用来做亚洲模特漫画会不错，可以尝试亚洲风格的大模型，效果应该都可以,自己多测试，暂时不做LORA了，因为BUZZ没了，每日打卡等到攒够500在做,
This LORA is a style model of a Virt-A-Mate author. I think it is mainly for the face, because the color of Virt-A-Mate is really nothing to say, so it will be good for Asian model comics. You can try a large Asian style model, and the effect should be OK. Test more by yourself. I won't make LORA for the time being, because BUZZ is gone. I will check in every day and wait until I have accumulated 500 before making it.
描述:
训练词语:
名称: VirtAMateStyle_1.safetensors
大小 (KB): 223105
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
