Refined the training, removed images and added new ones. Trained for more steps. I would not say that this version is "better" It's much more detailed, but not nearly as stylized.














A hypernetwork I trained on anime/painterly style post-apocalypse and cyberpunk images. I was going for a Lain/Dorohedoro/Texhnolyze kind of feel. This works great on semi-realistic models and anime models.
Recommended tags:
Location/setting: cyberpunk, cityscape, city lights, streets, alleyway, outdoors, slums, apocalypse
Colors: muted, desaturated, brown, gold, silver, gray (this is also great at doing very colorful images as well!)
Artists/style: Your favorite anime artists or anime shows/movies from the 80's/90's ;)
Best range of hypernet strength is 0.5 to 0.7 imo.
Can do ????, but no ???? was included in the training data, so use at your own risk.
Refined the training, removed images and added new ones. Trained for more steps. I would not say that this version is "better" It's much more detailed, but not nearly as stylized.
1. The rights to reposted models belong to original creators.
2. Original creators should contact SeaArt.AI staff through official channels to claim their models. We are committed to protecting every creator's rights. Click to Claim
