NVlabs/edm
PyTorch implementation of diffusion-based generative models that achieves state-of-the-art FID scores on CIFAR-10 and ImageNet.
Not currently ranked — collecting fresh signals.
star history
This repository implements a research paper that analyzes and clarifies the design space of diffusion-based generative models. It provides training and sampling code for score networks that generate images via stochastic denoising processes. The work introduces improved preconditioning, sampling schedules, and training methodologies that together achieve FID scores of 1.79 on CIFAR-10 (class-conditional) and 1.97 (unconditional), with faster convergence than prior approaches.
Frequently asked
- What is NVlabs/edm?
- PyTorch implementation of diffusion-based generative models that achieves state-of-the-art FID scores on CIFAR-10 and ImageNet.
- Is edm open source?
- Yes — NVlabs/edm is an open-source project tracked on heatdrop.
- What language is edm written in?
- NVlabs/edm is primarily written in Python.
- How popular is edm?
- NVlabs/edm has 2k stars on GitHub.
- Where can I find edm?
- NVlabs/edm is on GitHub at https://github.com/NVlabs/edm.