Bijutsubu (art-club) Anime Art Model
Base: EveryLoRA v1 - a Pony/Animagine/etc hybrid (does not use pony_ tags).
LoRA compatability: Works with most SDXL and Pony-based LoRAs.
Training: Trained using 10 x A40s on a subset of ~390,000 images from the KBlueLeaf/danbooru2023-webp-4Mpixel dataset. 4 epochs + testing steps.
Data selection: Selected over 4,000 most popular artists (by average score on 100 most recent uploads), and sampled up to 100 of their most popular artworks.
Merging: This model will merge with almost anything (including Pony models) thanks to the EveryLora TE block. If you merge with Pony you will transfer over the characters and styles of this model, but will lose the score_* tags. If you do make merges or fine-tunes of this model, I kindly ask that you credit me by linking back to this model.
Credit: Thank you to the creators of Pony, Animagine and SDXL for their models. Thank you to Kohya for their training script. Thank you to Kohaku for their dataset. Big thanks to KafyaHaarukku for their training notebook and settings.
HuggingFace: advokat/bijutsubu
License: https://freedevproject.org/faipl-1.0-sd/
Typical image output (including imperfections):
Metadata:
ciloranko, ishikei,
tomoe mami, 1girl, yellow skirt, charlotte (madoka magica), blonde hair, flower, thighhighs, skirt, smile, corset, sunflower, cup, hat, holding, yellow eyes, full body, gloves, long sleeves, drill hair, striped, holding cup, breasts, puffy sleeves, twin drills, yellow flower, black headwear, fingerless gloves, shirt, looking at viewer, brown footwear, teacup, striped thighhighs, boots, closed mouth, white shirt, beret, black gloves, bow, bangs, detached sleeves, solo, pleated skirt, hair ornament, yellow ribbon, ribbon, bowtie, vertical stripes, zettai ryouiki, neck ribbon, yellow bow, twintails, magical girl, black footwear, large breasts
Negative prompt: lowres, worst quality, low quality, bad anatomy, bad hands, multiple views, nsfw
Steps: 28, Sampler: Euler a, Schedule type: Automatic, CFG scale: 6.5Description
FAQ
Comments (6)
https://civitai.com/models/751465/everylora
Could you explain in more detail where exactly the weights mentioned in the text should be applied in the supermerger? The MBW settings in the supermerger include a total of 25 layers: in00~11, m00, out00~11, and the base part. However, the numbers you provided only cover 21.
NSFW images seem to turn out well, but the fingers...
So would you say this is an improvement over the previous CashMoney model you trained?
whats the difference between this and cashmooey?
Your model is really impressive. Even without using style LoRA, it generates the art styles of many artists quite well. The only model I’ve seen that surpasses this is probably NAI3. I'm really looking forward to the next version as well!
How much money did the training cost?
