The repository contains the dataset, the evaluation scripts, and the source code for monitors, but no script to allow easy reproduction of results in the paper. Could you please add the exact scripts to reproduce the results? It can allow further research based on the paper, and prevent hyperparameter errors in reproduction.