Verified: a year ago
SafeTensor
Type | |
Stats | 237 7,030 1.7k |
Reviews | (35) |
Published | Mar 31, 2024 |
Base Model | |
Training | Steps: 3,240 Epochs: 20 |
Usage Tips | Clip Skip: 2 |
Hash | AutoV2 8D6345957B |
Please Read: This LoRA is trained on AutismMix Confetti. It will not work as intended on original PonyXL model. I will likely make a PonyXL version soon.
Activation word: none. Strength: 1 - 0.8. This LoRA was built for portraits, concept art and "tasteful" erotica, but go nuts regardless.
Style Description: very rich 2D. Overall style is anime, but with great attention to fine detail and visible brushstrokes of various sizes. Shadow are complex. There is also a subtle grainy noise, that makes portraits really pop.
Character LoRA compatibility: mostly fine. As long as it's not western cartoons, it seems to work. This LoRA does affect the face, so if the character LoRA is flexible enough, it may be fine.