Based on the style from the artist: TheOtherHalf / the-other-half / ToHDraws
Produces highly stylized images with a more western/cartoon vibe.
Helpful Positives: outline, simple background
(Not a generated image)
This is an update to my previous release
Description
Images made with LORA weight 0.95 - 1.20
No trigger word needed.
Sample images use highres fix. 512x768 x1.5, 0.5 denoise / ADetailer, face
Trained on 397 unique, manually edited and captioned images.
Helpful Positives: simple background, gradient background, parted lips, uncensored
Helpful negatives: wavy mouth, glowing eyes, interlocked fingers, double eyebrows, bindi
This version is very slightly less "in style" than previously, but it was trained on roughly 10x more images and much higher quality images. From what I've noticed, this results in a more flexible model and a noticeable increase in quality and detail.