Warning: Eyes don't generate properly unless it is really up close
Trained in Anything V4.5 but should work fine in any anime model.
Lora weight: 0.4 ~ .75
Upscaler
Highres: 4x animeSharp
Hires steps: 20
Denoising strength .55 ~ .65
Trigger word not needed.
Description
Warning: Eyes don't generate properly unless really upclose.
About this version
Trained with about 31 images. Trained in Anything 4.5V model.
for weight, .04 ~ .75
Enjoy ^^