PINA: Physics-Informed Neural networks for Advance modeling

PINA is a Python package providing an easy interface to deal with physics-informed neural networks (PINN) for the approximation of (differential, nonlinear, ...) functions. Based on Pytorch, PINA offers a simple and intuitive way to formalize a specific problem and solve it using PINN.

PINA is freely available under the MIT License.

  • Download the source code at
  • Check the GitHub page to see how to cite it and the last works made possible by the package

PINA is currently developed and maintained at SISSA mathLab by

under the supervision of Prof. Gianluigi Rozza.

Contact us by email for further information or questions about PINA, or suggest pull requests.