Problem runing lammps compiled with pair_allegro

LAMMPS LAMMPS Installation
1

Hello friends,

I compiled lammps with pair_allegro and tested the input file :

units metal
atom_style charge
dimension 3

set newton on for pair_allegro (off for pair_nequip)

newton on
boundary p p p
read_data quartz.lammps

if you want to run a larger system, simply replicate the system in space

replicate 3 3 3

allegro pair style

pair_style allegro3232
pair_coeff * * SiO-deployed.path O Si

mass 1 15.999
mass 2 28.0855

velocity all create 300.0 1234567 loop geom

neighbor 1.0 bin
neigh_modify delay 5 every 1

timestep 0.001
thermo 10

nose-hoover thermostat, 300K

fix 1 all nvt temp 300 300 $(100*dt)

compute rdf and average after some equilibration

#comm_modify cutoff 7.0
compute rdfall all rdf 1000 cutoff 5.0
compute rdfall rdf 1000 1 1 1 2 2 2 # For Si-Si, Si-O, O-O

compute rdfall all rdf 100 1 1 1 2 2 2 # For Si-Si, Si-O, O-O
#fix 2 all ave/time 1 2500 5000 c_rdfall[*] file SiO.rdf mode vector

compute myRDF all rdf 100 1 1 1 2 2 2 # For Si-Si, Si-O, O-O

Fix to output RDF data to a file, referencing the correct compute ID

fix 2 all ave/time 100 1 100 c_myRDF[1] c_myRDF[2] c_myRDF[3] file rdf_output.dat mode vector

run 5ps

run 5000

The job kept terminating with the message:

Exception: Argument passed to at() was not in the map.
Abort(1) on node 0 (rank 0 in comm 0): application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0

Please i need your help in fixing this problem.

2

This pair style is not part of the LAMMPS distribution, you have to contact its author for support.

3

@sogenyi This is probably because your Allegro model uses the old default of only having force output, through the ForceOutput keyword in the yaml file, while newer versions of pair_allegro require StressForceOutput. See Trouble Running Deployed Model: Exception: Argument passed to at() was not in the map. · mir-group/pair_allegro · Discussion #36 · GitHub

2 Likes
4

Hello Anjohan,

Exactly. This led to me reinstalling most of updated versions and with right flag in yaml
file and retrain the model again. This time around it worked.
Thanks.

1 Like