Levi McClenny – Research Assistant, Department of Electrical Engineering
Jian Tao – Assistant Professor, Department of Visualization/TAMIDS/HPRC
Background and Objectives
TensorDiffEq is a software package designed and developed by members of the TAMU SciML community to implement collocation-based neuralPDE solvers, data assimilation solvers, as well as parameter inference and PDE discovery. Based on the distributed training strategies implemented in TensorFlow 2.x, TensorDiffEq can utilize multiple GPUs for faster solutions in large or complex domains in which traditional iterative solvers are slow or demonstrate convergence issues. Additionally, TensorDiffEq implements Self-Adaptive Solvers, which have experimentally demonstrated the ability to solve semi-stiff PDEs such as the Allen-Cahn equation using collocation-based solvers far more effectively than the baseline NeuralPDE solvers. This self-adaptive framework opens the possibility of solving more problems using the NeuralPDE methodology.
This workshop aims to promote scientific machine learning methods within the A&M research community and get more A&M researchers started in this exciting field. Meanwhile, we are interested in engaging a local community in the future development of TensorDiffEq.
The recorded video can be found at https://youtu.be/zV0DM67K4IA.
|1:00 PM – 1:25 PM||Introduction to SciML methods and Physics-Informed Neural Networks (Jian Tao) (Slides)|
|1:25 PM – 1:50 PM||Introduction to TensorDiffEq (Levi McClenny)|
|1:50 PM – 2:00 PM||Break|
|2:00 PM – 3:00 PM||Hands-on session with TensorDiffEq (Levi McClenny, Jian Tao) (Jupyter notebooks)|