The answer will depend on the specifics of your system, but since the error happens right at the start when forcefield is "first applied" by Gromacs, it could be either of these things:

1) You have non-standard (or incorrectly named) atoms or residues in your starting structure (.pdb). Different forcefields use different naming conventions.

2) You have a molecule which is not parametrized by default in your specific forcefield. This could be a ligand, post-translationally modified amino acid, or something of the sort. For a way to easily parametrize your molecule if this is the case, see for example https://www.researchgate.net/post/How_can_I_parametrize_a_molecule_for_simulation_in_gromacs

Also, while I don't know about your system or research, I would highly suggest selecting either AMBER (99SB-ILDN included by default) or CHARMM (CHARMM27 included) family forcefield. These are the state-of-the-art forcefields for biomolecules. 54A7 is more than a decade old now, less accurate and IMO should not be used.

If you want to include a new forcefield in GROMACS, you simply put the directory in your simulation folder, no need for special rights on your cluster. New distributions of CHARMM36m for GROMACS can be found here https://mackerell.umaryland.edu/charmm_ff.shtml#gromacs

Happy simulating!

Try making your input files through CHARMM-GUI server with CHARMM force filed.....