Cadet in Neural Network
Implementation of PINN from Raissi in Pytorch. Continuous Time Inference of Burgers' Equation. Cuda version and CPU version.
Cuda version updated, bugs fixed. Model parameters saved in .tar. - 08/06/2021 21:36
Existing Issues:
#A. Loss not converging well, especially the part on boundaries.
#B. Currently the optimizer L-BFGS-B as stated in Raissi's paper is not available.
Issues solved, Cuda version could reproduce Raissi's result quite well, at least visually. :)))
Discussion and Critics are welcomed.
Main Reference:
https://maziarraissi.github.io/PINNs/
Raissi, Maziar, Paris Perdikaris, and George E. Karniadakis. "Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations." Journal of Computational Physics 378 (2019): 686-707.
https://github.com/rodsveiga/PINNs (Helps rewriting Raissi's code into Pytorch)