V2 is available now!

Trained with higher quality dataset and more consistent round face.

Version 2 does have tendency to be too glossy on the skin (my fault), so you might have to play around with prompting or inpainting to get the glossiness out.

Character Lora depicting generic female mannequin (can do males too). Can be used for SFW or NSFW purposes.

Adjust as needed to your preferences. It was trained without explicit reproductive anatomy, but you can change that if you want.

Made with WAI-illustrious-SDXL checkpoint and Facelerss ILXL Lora for body and face during training.

V2 Positive Prompt:

femalemannequin, female, white body, bald, no face, no eyes, no nose, no mouth

or if you want male:

femalemannequin, male, white body, muscular, bald, no face, no eyes, no nose, no mouth

V1 Positive prompt:

femalemannequin, solo, completely white body, smooth, lines around limbs, breasts, faceless, no eyes,

V1 Negative prompt:

female flesh,

V1 - This Lora has a strong tendency to include other mannequins in the scene. Include this in the negative prompt:

another mannequin in scene,

By doing that, you alleviate the frequency of having more mannequins in the background/foreground and if you only want one mannequin as the main subject.

May be more improvements in the future.