Segmentation Fault in LAMMPS Simulation with hybrid classical LJ and NequIP Potentials

[HasnainMSajid]

Dear NequIP Community, This is Hasnain

I seek advice on how to fix a "segmentation fault" issue encountered during my Nequip-pairwise LAMMPS simulations. My setup involves a porous framework with guest molecules, where the framework atoms are modelled using NequIP potentials and the guest molecules are treated with classical LJ potentials. The guest@framework structure was generated through a GCMC simulation, and subsequent MD simulations are intended to analyse RDF and/or MSD.
However, when I attempt to run my simulation using the provided lmp.input file, it terminates with a segmentation fault. Attached is the input script I am using.

My trained Nequip potential is trained only on framework atoms and it did work for an empty framework.

Please provide suggestions or troubleshooting steps to resolve this issue. Any insights or recommendations would be greatly appreciated.

Thank you so much for your help.
in_rdf.txt