This is just a test of a lora for vore. I'm still training and adjusting it, so it can break easily and isn't working correctly yet, but it's already usable.
trained using 167 images from a dataset.
internal pov_rank16_fp16-step01400.safetensors