On-the-Fly Training

OTF is the on-the-fly training module for ASE, WITHOUT molecular dynamics engine. It needs to be used adjointly with ASE MD engine.

class flare.learners.otf.OTF(atoms, dt, number_of_steps, dft_calc, md_engine, md_kwargs, flare_calc=None, trajectory=None, prev_pos_init: ndarray = None, rescale_steps: List[int] = [], rescale_temps: List[int] = [], write_model: int = 0, force_only: bool = True, std_tolerance_factor: float = 1, skip: int = 0, init_atoms: List[int] = None, output_name: str = 'otf_run', max_atoms_added: int = 1, train_hyps: tuple = (0, 1), min_steps_with_model: int = 0, update_style: str = 'add_n', update_threshold: float = None, dft_kwargs=None, store_dft_output: Tuple[Union[str, List[str]], str] = None, build_mode='bayesian', wandb_log=None, **kwargs)
Trains a Gaussian process force field on the fly during

molecular dynamics.

Parameters

  • atoms (ASE Atoms) – the ASE Atoms object for the on-the-fly MD run.

  • flare_calc – ASE calculator. Must have “get_uncertainties” method implemented.

  • dt – the timestep in MD, in the units of pico-second.

  • number_of_steps (int) – the total number of steps for MD.

  • dft_calc (ASE Calculator) – any ASE calculator is supported, e.g. Espresso, VASP etc.

  • md_engine (str) – the name of MD thermostat, only VelocityVerlet, NVTBerendsen, NPTBerendsen, NPT and Langevin, NoseHoover are supported.

  • md_kwargs (dict) – Specify the args for MD as a dictionary, the args are as required by the ASE MD modules consistent with the md_engine.

  • trajectory (ASE Trajectory) – default None, not recommended, currently in experiment.

The following arguments are for on-the-fly training, the user can also refer to flare.otf.OTF

Parameters

  • prev_pos_init ([type], optional) – Previous positions. Defaults to None.

  • rescale_steps (List[int], optional) – List of frames for which the velocities of the atoms are rescaled. Defaults to [].

  • rescale_temps (List[int], optional) – List of rescaled temperatures. Defaults to [].

  • write_model (int, optional) – If 0, write never. If 1, write at end of run. If 2, write after each training and end of run. If 3, write after each time atoms are added and end of run. If 4, write after each training and end of run, and back up after each write.

  • force_only (bool, optional) – If True, only use forces for training. Default to False, use forces, energy and stress for training.

  • std_tolerance_factor (float, optional) – Threshold that determines when DFT is called. Specifies a multiple of the current noise hyperparameter. If the epistemic uncertainty on a force component exceeds this value, DFT is called. Defaults to 1.

  • skip (int, optional) – Number of frames that are skipped when dumping to the output file. Defaults to 0.

  • init_atoms (List[int], optional) – List of atoms from the input structure whose local environments and force components are used to train the initial GP model. If None is specified, all atoms are used to train the initial GP. Defaults to None.

  • output_name (str, optional) – Name of the output file. Defaults to ‘otf_run’.

  • max_atoms_added (int, optional) – Number of atoms added each time DFT is called. Defaults to 1.

  • train_hyps (tuple, optional) – Specifies the range of steps the hyperparameters of the GP are optimized. If the number of DFT calls is in this range, the hyperparameters are frozen. Defaults to (None, None) which means always training.

  • min_steps_with_model (int, optional) – Minimum number of steps the model takes in between calls to DFT. Defaults to 0.

  • dft_kwargs ([type], optional) – Additional arguments which are passed when DFT is called; keyword arguments vary based on the program (e.g. ESPRESSO vs. VASP). Defaults to None.

  • store_dft_output (Tuple[Union[str,List[str]],str], optional) – After DFT calculations are called, copy the file or files specified in the first element of the tuple to a directory specified as the second element of the tuple. Useful when DFT calculations are expensive and want to be kept for later use. The first element of the tuple can either be a single file name, or a list of several. Copied files will be prepended with the date and time with the format ‘Year.Month.Day:Hour:Minute:Second:’.

  • build_mode (str) – default “bayesian”, run on-the-fly training. “direct” mode constructs GP model from a given list of frames, with FakeMD and FakeDFT. Each frame needs to have a global property called “target_atoms” specifying a list of atomic environments added to the GP model.

compute_properties()
Compute energies, forces, stresses, and their uncertainties with

the FLARE ASE calcuator, and write the results to the OTF structure object.

md_step()

Get new position in molecular dynamics based on the forces predicted by FLARE_Calculator or DFT calculator

rescale_temperature(new_pos: ndarray)

Change the previous positions to update the temperature

Parameters

new_pos (np.ndarray) – Positions of atoms in the next MD frame.

run()

Performs an on-the-fly training run.

If OTF has store_dft_output set, then the specified DFT files will be copied with the current date and time prepended in the format ‘Year.Month.Day:Hour:Minute:Second:’.

run_dft()

Calculates DFT forces on atoms in the current structure.

If OTF has store_dft_output set, then the specified DFT files will be copied with the current date and time prepended in the format ‘Year.Month.Day:Hour:Minute:Second:’.

Calculates DFT forces on atoms in the current structure.

train_gp()

Optimizes the hyperparameters of the current GP model.

update_gp(train_atoms: List[int], dft_frcs: ndarray, dft_energy: float = None, dft_stress: ndarray = None)

Updates the current GP model.

Parameters

  • train_atoms (List[int]) – List of atoms whose local environments will be added to the training set.

  • dft_frcs (np.ndarray) – DFT forces on all atoms in the structure.

update_temperature()

Updates the instantaneous temperatures of the system.