A modern, scalable, and easy to use PDE Surrogate Benchmarking Framework.

Get started
Contribute on GitHub

What is PDEArena?

An easy-to-use, scalable framework for PDE Surrogate learning:
▶️ Simple and flexible codebase
▶️ Rapid implementation and evaluation of research ideas
▶️ Large collection of baseline results and associated implementations
▶️ Supports efficient scalable distributed training, powered by PyTorch Lightning


If you find this repository useful in your work, please cite:

              title={Towards Multi-spatiotemporal-scale Generalized PDE Modeling},
              author={Gupta, Jayesh K and Brandstetter, Johannes},
              journal={arXiv preprint arXiv:2209.15616},

                  title={Clifford neural layers for PDE modeling},
                  author={Brandstetter, Johannes and Berg, Rianne van den and Welling, Max and Gupta, Jayesh K},
                  journal={arXiv preprint arXiv:2209.04934},

Also consider starring the github repo. Star