dude has a pretty unique way of coloring that my skill issue couldn’t replicate, so i created a lora
v2.0: require trigger word miyajima_reiji_style, now have the drawing style, could be used with text2img
(v1.1 log:
should use with strength >= 1.1 for v1.0
around 0.8 for v1.1
sample i2i images are generated using 852_a_clone_xl ControlNet)
Description
new version to work with i2i that should use smaller lora strength
sample images are generated using 852_a_clone_xl ControlNet
p/s: i think i broke terms like 'dog' or 'ball', happy prompting :^)