Also trained on the runwayml Stable Diffusion 1.5 checkpoint. This Lora is trained on three times as many images, but is a little less strong than either of the previous versions
This model helps to create scenes where people are getting arrested in handcuffs. I had a hard time creating these kinds of scenes. It works best in my opinion at about 0.6-0.7 strength. It was trained using RealisticV30 so probably gives the best results with that model.