updated MLFC for better parallelization by reducing size of functions shipped to different ray workers