SevenNetRelaxMaker

class atomate2.forcefields.jobs.SevenNetRelaxMaker(name='MLFF.SevenNet relax', calculator_kwargs=<factory>, ionic_step_data=('energy', 'forces', 'magmoms', 'stress', 'mol_or_struct'), store_trajectory=StoreTrajectoryOption.NO, tags=None, relax_cell=True, fix_symmetry=False, symprec=0.01, steps=500, relax_kwargs=<factory>, optimizer_kwargs=<factory>, force_field_name=MLFF.SevenNet, task_document_kwargs=<factory>)[source]

Bases: ForceFieldRelaxMaker

Maker to perform a relaxation using the SevenNet ML force field.

Published in https://pubs.acs.org/doi/10.1021/acs.jctc.4c00190. pip install git+https://github.com/MDIL-SNU/SevenNet

Parameters:

  • name (str) – The job name.

  • force_field_name (str or .MLFF) – The name of the force field.

  • relax_cell (bool = True) – Whether to allow the cell shape/volume to change during relaxation.

  • fix_symmetry (bool = False) – Whether to fix the symmetry during relaxation. Refines the symmetry of the initial structure.

  • symprec (float = 1e-2) – Tolerance for symmetry finding in case of fix_symmetry.

  • steps (int) – Maximum number of ionic steps allowed during relaxation.

  • relax_kwargs (dict) – Keyword arguments that will get passed to AseRelaxer.relax.

  • optimizer_kwargs (dict) – Keyword arguments that will get passed to AseRelaxer().

  • calculator_kwargs (dict) – Keyword arguments that will get passed to the ASE calculator. E.g. the “model” key configures which checkpoint to load with mace.calculators.MACECalculator(). Can be a URL starting with https://. If not set, loads the universal MACE-MP trained for Matbench Discovery on the MPtrj dataset available at https://figshare.com/articles/dataset/22715158.

  • task_document_kwargs (dict (deprecated)) – Additional keyword args passed to ForceFieldTaskDocument().

  • ionic_step_data (tuple[str, ...] | None)

  • store_trajectory (StoreTrajectoryOption)

  • tags (list[str] | None)