Welcome to FastVPINNs’s documentation!
A robust tensor-based deep learning framework for solving PDE’s using hp-Variational Physics-Informed Neural Networks (hp-VPINNs). The framework supports handling complex geometries and uses tensor-based loss computation to accelerate the training of conventional hp-VPINNs. The framework is based on the work by FastVPINNs Paper. The framework is written using Tensorflow 2.0 and has support for handling external meshes.
Note: This framework is a highly optimised version of the the initial implementation of hp-VPINNs by Kharazmi et al. Ref hp-VPINNs(arXiv).
Variational Physics-Informed Neural Networks
Variational Physics-Informed Neural Networks is a special form of physics-informed neural networks, which uses variational form of the loss function to train the NN. A special form of hp-Variational PINNs which uses h- & p- refinement to enhance the ability of the NN to capture higher frequency solutions. For more details on the theory and implementation of hp-VPINNs, please refer to the FastVPINNs Paper and hp-VPINNs Paper.