Reproducible material for Physics-informed neural wavefields with Gabor basis functions - Tariq Alkhalifah, Xinquan Huang
This repository is organized as follows:
- 📂 gabor2d: python library containing the training and testing pipeline;
- 📂 model_zoo: python library containing the model architecture;
- 📂 utlis: python library containing the visualization tools and other utils.
- 📂 data: folder containing data;
- 📂 conf: python library containing the configuration;
To ensure reproducibility of the results, we suggest using the pinngabor.yml file when creating an environment.
Simply run:
./install_env.sh
It will take some time, if at the end you see the word Done! on your terminal you are ready to go.
Remember to always activate the environment by typing:
conda activate pinngabor
Run
bash run.sh
Before running, you can download the data from Click here. After running, go to folder exp/results/tb in the root_path produced by the procedures, and you could use tensorboard to visualize the trainig process and predictions.
After finish the training, you could go to the <run_root>/results/tb to use tensorboard --logdir=./ to check the training metrics and testing results.
Disclaimer: All experiments have been carried on a Intel(R) Xeon(R) CPU @ 2.10GHz equipped with a single NVIDIA GEForce A6000 GPU. Different environment configurations may be required for different combinations of workstation and GPU.
@article{alkhalifah2023physics,
title={Physics-informed neural wavefields with Gabor basis functions},
author={Alkhalifah, Tariq and Huang, Xinquan},
journal={arXiv preprint arXiv:2310.10602},
year={2023}
}