A LoRA trained on images of multiple angles(specifically face|ass) of people. The only needed trigger is <code>multiple views</code>. I do recommend prompting both the front and back features.

Multiple Views

Newer training method, more samples, should behave way better now. Trained on WAI-NSFW-illustrious-SDXL, so it will work better there, but should be usable with all Illustrious models.

Files with SHA256: 4989d07b546f5d0154ddefc6ece146b8ce9b462652c706c2c8bd4a0323edba3f