Data and code for training and testing a ResMLP model with experience replay for machine-learning physics parameterization

This directory contains the training data and code for training and testing a ResMLP with experience replay for creating a machine-learning physics parameterization for the Community Atmospheric Model. 

The directory is structured as follows:

1. Download training and testing data: https://portal.nersc.gov/archive/home/z/zhangtao/www/hybird_GCM_ML

2. Unzip nncam_training.zip

nncam_training

    - models

       model definition of ResMLP and other models for comparison purposes

    - dataloader 

       utility scripts to load data into pytorch dataset

    - training_scripts

       scripts to train ResMLP model with/without experience replay

    - offline_test

       scripts to perform offline test (Table 2, Figure 2)

3. Unzip nncam_coupling.zip

nncam_srcmods

     - SourceMods

          SourceMods to be used with CAM modules for coupling with neural network

     - otherfiles

          additional configuration files to setup and run SPCAM with neural network

     - pythonfiles

         python scripts to run neural network and couple with CAM

      - ClimAnalysis

          - paper_plots.ipynb

             scripts to produce online evaluation figures (Figure 1, Figure 3-10)