SD v2.1-base with Self-Perceptual Objective

This is the official model in Diffusion Model with Perceptual Loss paper.

This model is trained using the self-perceptual objective. It no longer needs classifier-free guidance to produce sensible images.

This model is trained using zero terminal SNR schedule following Common Diffusion Noise Schedules and Sample Steps are Flawed paper on LAION aesthetic 6+ data.

This model is finetuned from stabilityai/stable-diffusion-2-1-base.

This model is meant for research demonstration, not for production use.

Usage

from diffusers import StableDiffusionPipeline
prompt = "A young girl smiling"
pipe = StableDiffusionPipeline.from_pretrained("ByteDance/sd2.1-base-zsnr-laionaes6-perceptual").to("cuda")
pipe(prompt, guidance_scale=0).images[0].save("out.jpg") # No need for CFG!

Related Models

Cite as

@misc{lin2024diffusion,
      title={Diffusion Model with Perceptual Loss}, 
      author={Shanchuan Lin and Xiao Yang},
      year={2024},
      eprint={2401.00110},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

@misc{lin2023common,
      title={Common Diffusion Noise Schedules and Sample Steps are Flawed}, 
      author={Shanchuan Lin and Bingchen Liu and Jiashi Li and Xiao Yang},
      year={2023},
      eprint={2305.08891},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}