Style of artist Rocner. Made for Anima.
Even though there already exists Loras of this style for Anima made by other people and even built into the base checkpoint itself, I still wasn't satisfied. Those ones either had watermarks or didn't capture the appearance the way I liked, so this is an attempt at training my own version with no watermarks and consistent across all generations.
Recommended Weights
V1 and V2: Weigh between 0.7-1
Helpful Tips
Add the tag
lipsto force the character(s) to have a more prominent lip appearanceAnime/cartoon female characters are more likely to experience having less plump lips, so do add the tag if you want that, although it isn't always guaranteed
Version Info
V2 is retagged and retrained with much more images of new poses and unconventional characters. It is much more versatile than the last version while still remaining as faithful to the original artstyle as possible.
V1 captures the core appearance, but not perfect. It is predominantly trained on human/humanoid female subjects, thus is more likely to work with them. Furries/male subjects may be harder to generate.
DM for removal.
Have fun!
Description
Trained on 206 images, significantly more compared to last version.
Males, furries, and more poses were included into the dataset, thus making it more versatile beyond just generating humans/humanoid female characters.
Weigh between 0.7-1









