The repository contains a simple PyTorch-based demonstration of denoising diffusion models. It just aims at providing a basic understanding of this generative modeling approach.
A short theoretical intro to the topic can be found here.
Two example applications are provided as a small experimentation playground.
First, this notebook considers the Swiss roll distribution.
Second, a DDPM for the MNIST dataset can be learned by running python train_ddpm_mnist.py.
The trained model may then be analyzed in another notebook.