GAROM-Generative Adversarial Reduce Order Modelling
This repository contains the implementation of the GAROM model presented in https://arxiv.org/abs/2305.15881, as well as the experiments to reproduce the results. A more user-friendly implementation will be released in PINA soon.
Go into the experiments/POD folder. The programm can be lauched by:
bash perform_pod.sh
Go into the experiments/GAROM folder. The training programm can be lauched by:
bash perform_tests.sh
Once the train in finished, the results can be visualized by:
bash perform_analysis.sh
NOTE: The program lauches the simulations for regularized and not-regularized GAROM for all hidden dimensions for 5 different training. This reproduces all the results of the paper, but the overall time is considerable (~2h training for one model and one hidden dimension).