heads up, this is what i call a style/lora modifier

Affects overall style and image composition, a friend of mine was using a older version of this modifier as low as .3 to change the behavior of the composition


Based on Chinese comics and Korean manhuas the Dataset evolves as go through my Ipad Screenshots and clean up each image and turn them into usuable data. this one should be trained on what was roughly 300 images and I still have another 3500 to clean up.

Use Negative Score tags: score_6, score_5, score_4, censored, 3d, monochrome, watermark,

Do not use positive Score tags it will mess with the style. Unless thats what you want. im not here to tell you what to do.