Armored girls in the style of the Claymore manga by Norihiro Yagi
How to use
No trigger word, just use “1girl” and tags describing a claymore to prompt one. You can also just use “1girl” for a general style applicator.
See images for sample prompts for both.
Due to the dataset faces further away generally have to be inpainted.
Users of AOM2 might want to add “chibi,3d,realistic” to the negative tags to counteract its strict style changes. Or just use a model with better support for sketch and drawing type images.
Users of the AUTOMATIC1111 release can use the follow dynamic prompt as a general claymore generator
({upper|full} body)
(solo)
(1girl)(facing viewer)
({long|medium|short} white hair)
(({white|black} leotard)({white|black} armored skirt)({white|black} cape)(silver (pauldrons+gauntlets+greaves)))
Versions
This LoRA comes in two versions, one 8 and one 32 DIM sized. The former can be slightly more flexible and the latter slightly more detailed but the differences are fairly small.
Misc
512 resolution trained on NAI without noise offset option for best LoRA compatibility. Dataset based on 80 greyscale images from the Claymore Memorabilia Artbook.
Tested on NAI + AOM2.
All images except one or two are (slightly cherrypicked) txt2img gens for transparence, no inpainting done.
Don’t forget to review and upload your own images, it’s always fun to see what you’re creating!
Like my LoRAs and want more? Please consider a small tip towards my electricity bill and further training >.<