In memory of Kim Kwang Seok.
Description
Test dataset.
Since datas are low resolution and most of them are side-view(stage footage), result wasn't satisfying.
Will be fine-tuned more later.
FAQ
Comments (7)
Can you tell me your learning setting? and some tips for learning. I would like to try learning
109906Feb 19, 2023
CivitAI
Thanks for nice model. Do you train the model with default SD 1.5? I was observing that trained faces are cartoonized with other models (such as chilloutmix), but yours still retains the face with chilloutmix model. I wonder how you solved this problem.
I have done nothing special. Just used chilloutmix.
Never used SD 1.5.
Why dont you put 'realistic' tag on dataset caption files?
(Not sure if chilloutmix or chilloutmix-Ni.)
109906Feb 19, 2023
@kicktest Thanks. I will try it!