--- language: - en pipeline_tag: unconditional-image-generation tags: - Diffusion Models - Stable Diffusion - Perturbed-Attention Guidance - PAG --- # Perturbed-Attention Guidance SDXL ![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/6601282b569b30694e67b886/27Lmuol8anwd6L6BLzyWf.jpeg) [Project](https://ku-cvlab.github.io/Perturbed-Attention-Guidance/) / [arXiv](https://arxiv.org/abs/2403.17377) / [GitHub](https://github.com/KU-CVLAB/Perturbed-Attention-Guidance) This repository is based on [Diffusers](https://huggingface.co/docs/diffusers/index). The pipeline is a modification of StableDiffusionXLPipeline to add Perturbed-Attention Guidance (PAG). The original Perturbed-Attention Guidance by [Hyoungwon Cho](https://huggingface.co/hyoungwoncho) is availiable at [hyoungwoncho/sd_perturbed_attention_guidance](https://huggingface.co/hyoungwoncho/sd_perturbed_attention_guidance) ## Quickstart Loading Custom Piepline: ```py from diffusers import StableDiffusionXLPipeline pipe = StableDiffusionXLPipeline.from_pretrained( "stabilityai/stable-diffusion-xl-base-1.0", custom_pipeline="multimodalart/sdxl_perturbed_attention_guidance", torch_dtype=torch.float16 ) device="cuda" pipe = pipe.to(device) ``` Unconditional sampling with PAG: ```py output = pipe( "", num_inference_steps=50, guidance_scale=0.0, pag_scale=5.0, pag_applied_layers=['mid'] ).images ``` Sampling with PAG and CFG: ```py output = pipe( "the spirit of a tamagotchi wandering in the city of Vienna", num_inference_steps=25, guidance_scale=4.0, pag_scale=3.0, pag_applied_layers=['mid'] ).images ``` ## Parameters `guidance_scale` : gudiance scale of CFG (ex: `7.5`) `pag_scale` : gudiance scale of PAG (ex: `4.0`) `pag_applied_layers`: layer to apply perturbation (ex: ['mid']) `pag_applied_layers_index` : index of the layer to apply perturbation (ex: ['m0']) ## Stable Diffusion XL Demo To join a demo of PAG on Stable Diffusion, run [sd_pag_demo.ipynb](https://huggingface.co/hyoungwoncho/sd_perturbed_attention_guidance/blob/main/sd_pag_demo.ipynb).