please read "about this versions"
I wrote in there that what the difference between slider and adjuster version.
Why people didn't upload like this model ?
"ahegao" and "torogao" are best expressions in this world, you know?
but what if we can't adjustment "ahegao" and "torogao"?
It means war begins between "ahegao" and "torogao" over the world.
so I made this model for peace.
Dear all of freinds who love "ahegao" and "torogao"
Please upload images on galleries by this LoRA for peace between "ahegao" and "torogao".
All of the above is a joke without "please read".
you can check about this version if you want to get info.
Description
SDXL version
slider version - pure slider (use this LoRA mainly)
adjuster version - adjuster with use prompt (use this LoRA for adjust details)
recommend strength - "-1.5~1.5"
base model - AnimagineXLV3
LoRA dataset - generated from AnimagineXLV3 images
Learning as "1024^1024"
I prepared before version and this version's XY-grid.
so you can check 2 version's difference and which is more suitable on you.
I am not sure it works on most of popular anime model without AnimagineXLV3 in civitai.
you should use this model with these prompt.
"ahegao, exhausted, pleasure, embarrassed, open mouth, tongue out, blush, sweat, saliva, fucked silly, afterglow, aftersex" and more…?
FAQ
Comments (2)
Hey, I really like you lora. But I can't use sdxl, so could you please make an sd version of lora? thank you.
I've uploaded it, but I'm not sure if it's as effective as the SDXL versions.
SD1.5 has problem in generative resolutions. 512^512 has not comfortable to draw details.
so I recommend to use this LoRA with Hires.fix or Adetailer.
and also, I am not sure how much VRAM you have.
but now we have many ways to generate SDXL with Low VRAM.
for example, Fooocus, Comfy UI or FP8 version models.
ref: https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/14031