TensorDiffEq Overview

Official Documentation

Package Build Package Release pypi downloads python versions

TensorDiffEq is a python package built on top of Tensorflow to provide scalable and efficient PINN solvers. TensorDiffEq’s primary purpose is for scalable solving of PINNs (inference) and inverse problems (discovery).

Additionally, TensorDiffEq is the only package that fully supports and implements Self-Adaptive PINN solvers and is the only Multi-GPU PINN solution suite that is fully open-source.

Many choices for your scientific machine learning solution exist, use TensorDiffEq if you require

  • A meshless PINN solver that can distribute over multiple workers (GPUs) for forward problems (inference) and inverse problems (discovery)

  • Scalable domains - Iterated solver construction allows for N-D spatio-temporal support

    • support for N-D spatial domains with no time element is included

  • Self-Adaptive Collocation methods for forward and inverse PINNs

  • Intuitive user interface allowing for explicit definitions of variable domains, boundary conditions, initial conditions, and strong-form PDEs

What makes TensorDiffEq different?

  • Completely open-source

  • Self-Adaptive Solvers for forward and inverse problems, leading to increased accuracy of the solution and stability in training, resulting in less overall training time

  • Multi-GPU distributed training for large or fine-grain spatio-temporal domains

  • Built on top of Tensorflow 2.0 for increased support in new functionality exclusive to recent TF releases, such as XLA support, autograph for efficent graph-building, and grappler support for graph optimization* - with no chance of the source code being sunset in a further Tensorflow version release

  • Intuitive interface - defining domains, BCs, ICs, and strong-form PDEs in “plain english”

Know that some of these pages are still under construction as of February 2021, please see our repository

If you use TensorDiffEq in your work, please cite:

@article{mcclenny2021tensordiffeq,
  title={TensorDiffEq: Scalable Multi-GPU Forward and Inverse Solvers for Physics Informed Neural Networks},
  author={McClenny, Levi D and Haile, Mulugeta A and Braga-Neto, Ulisses M},
  journal={arXiv preprint arXiv:2103.16034},
  year={2021}
}

Thanks to our sponsors:

Army Research Lab Logo Texas A&M Engineering Texas A&M Institute of Data Science