ADM priors

ADM (guided-diffusion UNet) priors trained on the same data as the matching IHDM priors in Lin-Wang-NUS/IHDM, used as the hot-diffusion baselines (DPS, ΠGDM, DDRM, DAPS, ...). One subfolder per training set; *_ema.pt holds the EMA weights.

Subfolder	Training data	Resolution	Files
`imagenet100_128/`	ImageNet-100	128	`imagenet100_128.pth`, `imagenet100_128_ema.pt`
`fastmri_knee_320/`	fastMRI knee (magnitude)	320	`fastmri_knee320_100m.pth`, `fastmri_knee320_100m_ema.pt` (~~100M); `fastmri_knee320_50m.pth`, `fastmri_knee320_50m_ema.pt` (~~50M)

from huggingface_hub import hf_hub_download
path = hf_hub_download("Lin-Wang-NUS/ADM", "imagenet100_128/imagenet100_128_ema.pt")

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support