Autogrow4 performance on slurm

[TommyBiomodellingAU]

hi autogrow4 developers
I am having trouble with performance issues when using autogrow4 on slurm. The time it takes for an autogrow4 run is dependent on the input parameters (number of molecules, mutations etc), but which settings do you recommended when running the program on slurm?
How is autogrow4 run most optimally on a HPC node?

Best,
Tommy

[TommyBiomodellingAU]

As a follow-up clarification:
When running Autogrow in multiprocessing mode, the command I have used in slurm is:

mpirun -n 4 python -m mpi4py ./autogrow4/RunAutogrow.py --json ${json_input}

Autogrow4, however, seems to just create duplicate runs of autogrow. Is this the actual multiprocessing?