Toons but not all the way?
I have been messing with lora training and it has my detailed anime checkpoint as a base and I kinda like the results.
VAE is baked in
have fun with it and any feedback is appreciated
Suggested settings:
CLIP skip 2
Samplers : Eular A
CFG : 5-7
All images have highres.fix
positive: masterpiece, best quality, highres, very aesthetic, high quality, insanely detailed, beautiful, beautiful face, detailed eyes
negative: (low quality, worst quality:1.5),(bad anatomy), lowres, bad composition, fewer digits, text, username, logo, inaccurate eyes, extra digits, fewer digits, extra arms, disfigured, missing arms









