Training Settings: batch size 4, mixed res from 512px to 1216px, ~251 images per epoch, optimizer is SinkSGD_adv
Trained on images from haiz ai
haiz ai
Strength 1.25 / 1.5 might give more stylized results.
