Chitose Naruse from the Manga/OVA Kyochuu Rettou/Dai Kyochuu Rettou
Best girl
(TAGS)
Character - ChitoseNaruse, brown eyes
Optional - glasses
Optional - hair ornament
Choice - long hair, twin braids
Choice - long hair, hair down
Outfit (Kyochuu Rettou) - blazer, brown jacket, long sleeves, collared shirt, white shirt, neck ribbon, red ribbon, pleated skirt, red skirt, black thighhighs, loafers, brown footwear
Outfit (Dai Kyochuu Rettou) - blazer, brown jacket, tied jacket, sleeves rolled up, collared shirt, white shirt, pleated skirt, red skirt, loafers, brown footwear
Description
Entire dataset redone from scratch due to major issues in the original one
LoRA trained to the same settings I'm using on my other characters, the original model was one of my first dives into Illustrious and like most people moving over to the new IL model I didn't know what settings to use so just ran with existing Pony ones
Her Kyochuu Rettou outfit is trained using images from the OVA and her Dai Kyochuu appearance is trained using manga images, hopefully the Dai Kyochuu outfit learns the colour from the OVA one as they are the same outfits just worn differently
Her BDSM outfit was removed due to consistency issues in how it's depicted in the manga and also because it also in an image used her skirt which was crossing over into her default school uniform tags (Crossing tags between outfits isn't good as they risk generalising both which is where problems come in if the outfits are wildly different)
There weren't enough images of her in her underwear and swimsuit to train as most of them are just random illustrious at the start and end of volumes but I've left then in the dataset as they do help with body training. These outfits are just tagged to their basic tags (bra, panties, bikini, naked towel etc.)
Remove the /// when generating, they're just to help separate mutually exclusive prompts if you use LoRA metadata downloader
FAQ
Details
Files
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.