I had trouble getting images like this without img2img support, so I decided to train a model for it. LORA comes in two flavors: Dom (holding leash) and Sub (on leash).
Description
FAQ
Comments (5)
Hi, there seems to be some issues generating on dom pov where:
the leash isn't connected to the collar
that the girl is holding it, not the viewer
both of the above
(rarely) there's no leash at all
sub pov just isn't working at all. I get weird pictures with sometimes lowres, leash not in hand, or with multileash. I even copy your generation data on pic 2(with pink-haired elf) and get weird stuff. Maybe i am doing it wrong?
and you are using not <lora:leashPOV_subPOV:1>, you are using "leashed_pov" but it is lora not TI. And you don't leave your model. Please explain how to get smth like that i just give up
Models used are Anything V4 and AOM2. It doesn't appear in the prompt because I was using the Additional Networks plugin.
@ricketysplit well i tried same promts on anything v4 and aom 2 hard and nsfw and again i can't get smth like your pics. Even change clip to 2 and back to 1 and nothing changes. Did i miss something?
Is there anything on whips, like better models of bullwhips, 9tails, horseriding whips....?