I was trying to help out with this problem too - its running our our cluster at Northwestern.
The pour example doesn’t have a create atoms command. Instead “atoms” are added by the fix … pour command.
I wonder about the compatibility of the GRANULAR package with compute stress/atom. Per the documentation granular particles have angular momentum - something not treated by the virial stress theorem. To be honest, I’d be more suspicious if it did work - would it even be correct?
I do think lammps should spit out a more meaningful warning in this case. When I tried running it… I just a seg fault.