agat 8.0.2

Atomic Graph ATtention networks for predicting atomic energies and forces.

AGAT (Atomic Graph ATtention networks)

The PyTorch backend AGAT is available now, try with pip install agat==8.*. For previous version, install with pip install agat==7.*.

Using AGAT

The documentation of AGAT API is available.

Installation

Install with conda environment

• Create a new environment

  conda create -n agat python==3.10
  

• Activate the environment

  conda activate agat
  

• Install PyTorch,
  Navigate to the installation page and choose you platform. For example (GPU):

  conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
  

• Install dgl.
  Please navigate to the Get Started page of dgl. For example (GPU):

  conda install -c dglteam/label/cu118 dgl
  

• Install AGAT package

  pip install agat
  

• Install CUDA and CUDNN [Optional].

  • For HPC, you may load CUDA by checking module av, or you can contact your administrator for help.
  • CUDA Toolkit
  • cuDNN

Quick start

Prepare VASP calculations

Run VASP calculations at this step.

Collect paths of VASP calculations

• We provided examples of VASP outputs at VASP_calculations_example.
• Find all directories containing OUTCAR file:

  find . -name OUTCAR > paths.log
  

• Remove the string 'OUTCAR' in paths.log.

  sed -i 's/OUTCAR$//g' paths.log
  

• Specify the absolute paths in paths.log.

  sed -i "s#^.#${PWD}#g" paths.log
  

Build database

from agat.data import BuildDatabase
if __name__ == '__main__':
    database = BuildDatabase(mode_of_NN='ase_dist', num_of_cores=16)
    database.build()

Train AGAT model

from agat.model import Fit
f = Fit()
f.fit()

Application (geometry optimization)

from ase.optimize import BFGS
from agat.app import AgatCalculator
from ase.io import read
from ase import Atoms

model_save_dir = 'agat_model'
graph_build_scheme_dir = 'dataset'

atoms = read('POSCAR')
calculator=AgatCalculator(model_save_dir,
                          graph_build_scheme_dir)
atoms = Atoms(atoms, calculator=calculator)
dyn = BFGS(atoms, trajectory='test.traj')
dyn.run(fmax=0.05)

Application (high-throughput prediction)

from agat.app.cata import HpAds

model_save_dir = 'agat_model'
graph_build_scheme_dir = 'dataset'
formula='NiCoFePdPt'

ha = HpAds(model_save_dir=model_save_dir, graph_build_scheme_dir=graph_build_scheme_dir)
ha.run(formula=formula)

For more custom manipulations, see our documentation page.

Some default parameters

agat/default_parameters.py; Explanations: docs/sphinx/source/Default parameters.md.

Change log

Please check Change_log.md