SPM
SPM copied to clipboard
Official implementation of paper "One-dimensional Adapter to Rule Them All: Concepts, Diffusion Models and Erasing Applications".
I've tested the nudity erasure model(nudity.safetensors) using the script ``` if len(os.listdir(args.img_output_path)) == 4703: print("benchmark images already generated.") else: pipe = StableDiffusionPipeline.from_pretrained( "/root/models--CompVis--stable-diffusion-v1-4/snapshots/133a221b8aa7292a167afc5127cb63fb5005638b", safety_checker=None, requires_safety_checker = False # local_files_only=~args.remote, )...
Thanks for you incredible work. I am confused by the results I get when evaluating your checkpoints. My problem includes: 1. When using the nudity spm, Does "γ = 2.0"...
What package is library==0.0.0 in requirements.txt? It doesn't seem to be recognized by pip. Is this meant to be a placeholder?
I found: parser.add_argument( "--task_args", nargs="*", help="""Extra arguments for the task. Acceptable arguments: task=general: concepts(list[str]), num_templates(optional, int, default=80), num_images_per_template(optional, int, default=10); task=artwork: datasets(list[str]); task=i2p: None. task=coco: None. """, However, I am...