ALLEGRO

References: Musaelian2023

This section specifies the input parameters for ALLEGRO potential type based on equivariant neural network potentials. Starting from the NequIP 0.6.0, one can predict stress if the config.yaml file has the StressForceOutput keyword, regardless of whether the model has been trained on the stress. Requires linking with libtorch library from https://pytorch.org/cppdocs/installing.html. [Edit on GitHub]

Keywords

Keyword descriptions

ATOMS: string

Usage: ATOMS {KIND 1} {KIND 2} .. {KIND N}

Mentions:Nequip and Allegro

Defines the atomic kinds involved in the ALLEGRO potential. Provide a list of each element, making sure that the mapping from the ATOMS list to NequIP atom types is correct. This mapping should also be consistent for the atomic coordinates as specified in the sections COORDS or TOPOLOGY. [Edit on GitHub]

PARM_FILE_NAME: string = model.pth

Aliases: PARMFILE

Usage: PARM_FILE_NAME {FILENAME}

Specifies the filename that contains the ALLEGRO model. [Edit on GitHub]

UNIT_CELL: string = angstrom

Usage: UNIT angstrom

Units of the cell vectors in the ALLEGRO model.pth file. The units of positions, energies and forces must be self-consistent: e.g. coordinates in Angstrom, energies in eV, forces in eV/Angstrom. [Edit on GitHub]

UNIT_COORDS: string = angstrom

Usage: UNIT angstrom

Units of coordinates in the ALLEGRO model.pth file. The units of positions, energies and forces must be self-consistent: e.g. coordinates in Angstrom, energies in eV, forces in eV/Angstrom. [Edit on GitHub]

UNIT_ENERGY: string = eV

Usage: UNIT hartree

Units of energy in the ALLEGRO model.pth file. The units of positions, energies and forces must be self-consistent: e.g. coordinates in Angstrom, energies in eV, forces in eV/Angstrom. [Edit on GitHub]

UNIT_FORCES: string = eV/Angstrom

Usage: UNIT hartree/bohr

Units of the forces in the ALLEGRO model.pth file. The units of positions, energies and forces must be self-consistent: e.g. coordinates in Angstrom, energies in eV, forces in eV/Angstrom. [Edit on GitHub]