NOTE: USE AT 0.8 STRENGTH FOR BEST RESULTS

ALL SHOWCASE IMAGES ARE PRODUCED FROM A SINGLE TEXT2IMG PASS + ADETAILER

FEATURES:

-all dataset images were upscaled to full HD. expect good quality output

-extensively tagged captions, allowing for granular control and inclusion/exclusion of props like her blindfold, dress, fur trim, cape, staff, bird(verry), backless outfit, etc

-NSFW supported

-I INCLUDED SOME ARTIFICIALLY RENDERED FACE PORTRAITS WITHOUT THE BLINDFOLD IN THE DATASET! ENJOY :)

USAGE TIPS:

-clip skip:1

-CFG: 5-7

-strength: 0.8

use of Adetailer is recommended. use the lora in the positive prompt of adetailer along with (detailed face) and set inpaint denoising strength to 0.5 or 0.6

ADETAILER PROMPT: detailed face, realistic, white hair, black hair, two-tone hair, necklace, blindfold

make sure to include the lora in the adetailer prompt.

BASE MODEL: Hs UltraHD CG