This is a LoRA (XL) for creating DEMONIC COURTESANs for medieval rpgs.
[note on images: these were made wth ComfyUI and I thought the images would contain the generation data as they had in the past, but they appear that they don't. I believe that's because I was using some of the WAS Suite extensions and Civitai didn't know what to do with those. For future images, I'll try to keep the nodes to the vanilla Comfy UI set if I can].
The trigger words for the model are the following:
1boy, 1girl, antlers, armor, bare shoulders, beard, blonde hair, blood, blue eyes, blurry, blurry background, braid, breasts, brown eyes, brown hair, cleavage, collarbone, courtesan (demonic), crying, curled horns, curly hair, dark skin, depth of field, dress, facial hair, freckles, glowing, glowing eyes, green eyes, horns, jewelry, lips, long hair, looking at viewer, makeup, male focus, medium breasts, mole, necklace, nipples, off shoulder, parted lips, pauldrons, pointy ears, portrait, realistic, red dress, red eyes, red hair, sharp teeth, sheep horns, short hair, shoulder armor, skull, small breasts, smile, solo, tattoo, tears, teeth, upper body, wavy hair, white dress, white hair, yellow eyesDescription
initial xl version
FAQ
Comments (3)
Ash, I've seen you've been working on these XL models, but, honestly... they're about 10% the quality of your 1.5 work with 10 times the size. It's not your fault, XL is a disappointment and just plain awful.
Comparing this with the 1.5 equivalent is like night and day.
Sorry, just saying it like it is, hoping you can go back to the good stuff.
I should be done with the first round of XL models in the next week or so.
I'm keeping the 1.5 production still going until there is a wider range of checkpoints for XL and until the user population is adopting XL. Until then, I'm planning on providing both XL and 1.5 LoRAs.
One more thing....
I do agree the XL checkpoints really don't have the range of variation that I really want, but I have tried some of my XL LoRAs with normal sd 1.5 checkpoints and then just rendering out with an empty 1024x1024 image and there's a lot of variation doing that.
I do hope for some more variety in the main xl checkpoints though.










