Good practices when using Numba in a cluster e.g. using SLURM

Hi, I would like to know if there are any good practices to follow when using Numba in scripts that are executed via SLURM.

I ask this because I often receive the following error that leads to a crash:

OMP: Error #179: Function Can't open SHM failed:
OMP: System error #0 Sucess

I apologize in advance if this error is not related to Numba. The reason why I think it is related to Numba is due to its use of OpenMP to parallelize computations (although I am not using parallel=True in the signatures, I think this can stem from SIMD vectorizations when using fastmath=True).

I think this error specifically occurs when two or more processes, all using Numba, are being executed.

Numba v: 0.56.4
Python v: 3.10.

I would highly appreciate any recommendations that I could follow to avoid such errors.

Thank you for this amazing library.