Hello! I am trying to run calphy example 2 using the MLIP, mattersim-v1.0.0-5M, instead of the provided EAM model. I have adjusted the input script as follows:
calculations:
- element: Fe
lattice: bcc
lattice_constant: 2.884
mass: 55.845
md:
barostat_damping: 0.1
equilibration_control: nose_hoover
thermostat_damping: 0.1
timestep: 0.001
mode: ts
n_equilibration_steps: 10000
n_iterations: 1
n_switching_steps: 25000
pair_coeff: '* * Fe'
pair_style: mliap unified ../potentials/mattersim-v1.0.0-5M-mliap.pt
pressure: 0.0
queue:
commands:
- conda activate calphy
cores: 4
scheduler: local
reference_phase: solid
repeat:
- 5
- 5
- 5
temperature:
- 100.0
- 1400.0
- element: Fe
lattice: FCC
lattice_constant: 6.0
mass: 55.845
md:
barostat_damping: 0.1
equilibration_control: nose_hoover
thermostat_damping: 0.1
timestep: 0.001
mode: ts
n_equilibration_steps: 10000
n_iterations: 1
n_switching_steps: 25000
pair_coeff: '* * Fe'
pair_style: mliap unified ../potentials/mattersim-v1.0.0-5M-mliap.pt
pressure: 0.0
queue:
commands:
- conda activate $LAMMPS_PYENV
cores: 4
scheduler: local
reference_phase: solid
repeat:
- 5
- 5
- 5
temperature:
- 100.0
- 1400.0
I am getting the following error when attempting to run:
Exception: ERROR: Loading mliappy unified module failure. (src/lammps-patch_4Jul2026/src/ML-IAP/mliap_unified.cpp:208)
Last input line: pair_style mliap unified ../potentials/mattersim-v1.0.0-5M-mliap.pt
Based on this LAMMPS documentation on using pair_stlye mliap,

I believe I need to do the following to the lammps object:
from lammps import lammps
from lammps.mliap import activate_mliappy
lmp = lammps()
activate_mliappy(lmp)
But I'm not sure how best to implement this with calphy. Any advice on how to proceed would be greatly appreciated!
Hello! I am trying to run calphy example 2 using the MLIP, mattersim-v1.0.0-5M, instead of the provided EAM model. I have adjusted the input script as follows:
calculations:
lattice: bcc
lattice_constant: 2.884
mass: 55.845
md:
barostat_damping: 0.1
equilibration_control: nose_hoover
thermostat_damping: 0.1
timestep: 0.001
mode: ts
n_equilibration_steps: 10000
n_iterations: 1
n_switching_steps: 25000
pair_coeff: '* * Fe'
pair_style: mliap unified ../potentials/mattersim-v1.0.0-5M-mliap.pt
pressure: 0.0
queue:
commands:
cores: 4
scheduler: local
reference_phase: solid
repeat:
temperature:
lattice: FCC
lattice_constant: 6.0
mass: 55.845
md:
barostat_damping: 0.1
equilibration_control: nose_hoover
thermostat_damping: 0.1
timestep: 0.001
mode: ts
n_equilibration_steps: 10000
n_iterations: 1
n_switching_steps: 25000
pair_coeff: '* * Fe'
pair_style: mliap unified ../potentials/mattersim-v1.0.0-5M-mliap.pt
pressure: 0.0
queue:
commands:
cores: 4
scheduler: local
reference_phase: solid
repeat:
temperature:
I am getting the following error when attempting to run:
Exception: ERROR: Loading mliappy unified module failure. (src/lammps-patch_4Jul2026/src/ML-IAP/mliap_unified.cpp:208)
Last input line: pair_style mliap unified ../potentials/mattersim-v1.0.0-5M-mliap.pt
Based on this LAMMPS documentation on using pair_stlye mliap,

I believe I need to do the following to the lammps object:
from lammps import lammps
from lammps.mliap import activate_mliappy
lmp = lammps()
activate_mliappy(lmp)
But I'm not sure how best to implement this with calphy. Any advice on how to proceed would be greatly appreciated!