Softwares

A library for scientific machine learning and physics-informed learning.

>700,000 downloads, >2,500 GitHub Stars, >70 Contributors. Used in hundreds of papers published by a diverse range of scientists from >200 universities, national labs, and industry.

A library for benchmarking model-based and learning-based PDE control methods. Currently, it includes three PDE control tasks: 1D hyperbolic PDEs(Burger's Equation), 1D parabolic PDEs (Reaction-Diffusion/Heat Equations), and 2-D Navier Stokes PDEs, that are all wrapped under a gym environment designed for PDE boundary control problems. 

See the full paper presented at L4DC 2024 for more details.