This LoRA is kind of an experiment of what comes out if you train on low-quality art - the artist we're taking on this time is s u, and they post sketches only.

The trigger is by s u. I did try to tag spanking concepts like over the knee, caned and holding cane, but I doubt they work particularly well. If you want to get something closer to the original input, prompt as sketch.

This dataset wasn't really censored - it's more like the overall quality is so lowres that you don't even see the explicit bits properly.