a realistic LORA model for yushuang, a LPL host,
her smile, grin, mouth opening for chilloutmix
used in forms of full portrait,three quarters portrait, or close-up shot
0.7-0.8 weights will be fine,
it might includes many watermarks from training images set, which means you need to offset by adding appropriate negative prompt
余霜的真实LORA模型，
可以生成露齿笑，微笑，微微张嘴
全身，半身，特写
权重大概0.7-0.8
生成的图片需要特别地用去水印的负面tag来抵消，这是我制作训练图片时遗留下来的错误
描述:
this is the first version of this model, still has some extra noise generated from training sets.
maybe I will refine it some day
这是余霜的第一版模型，仍然有来自于原始训练集图片的多余干扰
训练词语:
名称: ys_test.safetensors
大小 (KB): 147578
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
