sumanth107 / PINNs-TF2.4.0

Contains implementation of PINN using Tensorflow 2.4.0

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

PINNs-TF2.4.0

This repo contains implementation in Tensorflow 2.4.0 of the Burger's Equation example put together by Raissi et al. in their original publication about Physics Informed Neural Networks (PINNs)

burger.JPG

The code uses a custom loss function that accounts for the physics as well (by leveraging the automatic differentiation ability of Neural Networks), apart from minimizing the error between predicted and training data.

Key Feature

Mini-Batch Gradient Descent is adopted along with Adam optimizer (contrary to LBFGS used in the original paper).

Link to the original publication : https://maziarraissi.github.io/PINNs/

About

Contains implementation of PINN using Tensorflow 2.4.0

License:MIT License


Languages

Language:Jupyter Notebook 100.0%