I'm bad at prompts
It's a little hit or miss on the results. Weight of 0.6-0.7 worked well with me. I was able to combine this with other loras with decent results.
Using this with anime models is possible, but far trickier for my smooth brain. I used the img2img DeepBooru Interrogator on a realistic image to come up with a prompt. Take a look at my prompts for some ideas. It's definitely more janky with anime Character Loras/ti.
Description
FAQ
Comments (9)
Too small.
Don't kick a man when he's down.
no way to dodge cum sadly due to dataset
I'd have to manually photoshop it all away.
@Shurik why would someone want to dodge cum in the first place lol. Just kidding, glad i find this one, looking promising
hopefully you can add an anime version of it.
Have you tried it with anime?
Why doesnt it work for me on sd , it gives me unexpected results like a single woman
It's an unexplainable phenomenon.













