Dataset for paper "Seismic wavefield solutions via physics-guided generative neural operator"

This file contains two main folders:

1. **code**     This folder includes MATLAB scripts for generating the training and test sets.   - To generate the training set, run `main_train.m`.   - To generate the test set, run `meta_test.m`.

2. **velmodel**     This folder contains the velocity models used for both the training and test sets.   - The training set velocity models are located in the `velmodel/train` folder.   - The test set velocity models are located in the `velmodel/test` folder.

## Dataset Description

- **Training Set:**    The training set is created by extracting 6000 velocity models from the "CurveVel-A" class of the OpenFWI dataset.

- **Test Set:**    The test set includes velocity models extracted from the following eight classes of the OpenFWI dataset:  - `FlatVel-A`  - `FlatVel-B`  - `FlatFault-A`  - `FlatFault-B`  - `CurveVel-A`  - `CurveVel-B`  - `CurveFault-A`  - `CurveFault-B`

These datasets are provided for academic and research purposes.

## Requirements

- MATLAB

## Instructions

1. Navigate to the `code` folder.2. To generate the training set, run the `main_train.m` script.3. To generate the test set, run the `meta_test.m` script.4. The generated velocity models will be stored in the corresponding subfolders under `velmodel`.

## Acknowledgments

This work utilizes the OpenFWI dataset. For more details on the dataset, please refer to the original publication.