V0.1 Important Notes:
Trained with Danboru tags only.
4000 Steps with 101 Images with Tags.
Why the hell is there so many Velina LoRAs? Where is my Norma appreciation?
(Edit: Okay i get it now, the civit bot nuked all images and almost deleted the post..)
A style LoRA for Anima (by extension its finetunes) by CircleStone Labs.
This LoRA was trained on a curated dataset of 101 images of Norma Hollowell. All images were treated to remove text, and most symbols, then batch tagged with ConvNextV2 Huge Full + EVA02 Large Patch14 448 Full and manually tagged for fixing/adding tags.
This Lora is best used at 1 strength, imo. Trigger word is very optional, but hey, its norma hollowell.
Also this was trained on my chef's kiss AMD 7900XT, (under Linux with Ostris's Ai-Toolkit❤️)
Description
Main Settings: 4000Steps, AdamW8bit, 1Batches, [768,1024], LR 4e-5, WeightDecay 1e-2, Rank16, Repeats 2, No X Axis Flip, 101 IMG
