How to use this?

Step 1: generate training data by running generate_training_data.py. This will create .npz files that contain the training data. The value of parms["n_generate"] should be set in the file, which speciefies how many data points we want.

Step 2: train the model using this generated data as input by running ml_model.py.