Each experiment contains a compare.jl script that takes iter and addprocs flags.
iterdictates how many monte carlo evaluation runs to perform for each planning time and solver.addprocsdictates how many additional workers to add.
For example, if we want to reproduce the experimental results for Laser Tag, from the experiments folder we could run the following:
julia --project LaserTag/compare.jl --iter=500 --addprocs=30.
The results for experiments are stored in experiments/<experiment_name>/data/.
Hyperparameter optimization scripts are stored in experiments/<experiment_name>/opt/.
Example hyperparameter optimization results for PFT-DPW on Sub Hunt:
