The model is highly recommended to use hires-fix
Hires upscaler: r-esrgan 4x+
Denoising strength: 0.2～0.4
Hires steps: 10~20
It would be nice to see diferent images of the face from various angles and a smile expresssion, to know if you trained the face good with the A.I.
Why do I use your lora, often generate two Yua
mikami_yua_V1.safetensors
