Style LoRA based on a set of 55 images of women by the artist kaoming -- https://danbooru.donmai.us/artists/44522.

Trained on Pony and gives best results with Pony Diffusion V6 XL. When used with Pony derived models, some elements of the style (faces, proportions) remain and others (the painterly style, blurred background) disappear -- see the picture generated with Luminaverse for an example.

Recommended weight of 0.7-0.9. Works best with minimal prompting. The training data was comprised exlusively of female portraits so it's ideal for drawing those.

Start the prompt with:

score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, 1girl,