Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Convergence Acceleration in Machine Learning Potentials for Atomistic Simulations: ESI dataset

Bayerl, Dylan and Saidi, Wissam (2021) Convergence Acceleration in Machine Learning Potentials for Atomistic Simulations: ESI dataset. [Dataset] (Submitted)

[img] Archive (TGZ) (Post-review version, including training data)
Available under License Creative Commons Attribution No Derivatives.

Download (668MB)
[img] Archive (TGZ) (Original version)
Available under License Open Data Commons Attribution License (ODC-By).

Download (92kB)


Materials property datasets associated with the (submitted) manuscript entitled “Convergence Acceleration in Machine Learning Potentials for Atomistic Simulations” by Dylan Bayerl, Christopher M. Andolina, Shyam Dwaraknath, and Wissam A. Saidi. This work studied the impact of precision in Brillouin zone integrations (i.e. k-space sampling density) on stability and convergence of material properties calculated both directly with density functional theory and with machine learning potentials trained on density functional theory data. Materials include aluminum, copper, and magnesium metals in BCC, FCC, and HCP lattices. Material properties include cohesive energy and volume-per-atom at equilibrium, linear-elastic stiffness matrix elements, bulk modulus, shear modulus, Young’s modulus, Poisson ratio, vacancy formation energy, and formation energies for Oh, and Td self-interstitials. DFT calculations were performed with VASP version 5.4.1. Machine learning potentials were generated using DeepMD-kit version 1.3.2. Material property calculations were performed with DP-GEN version 0.8.2.dev4+gbd9f34b and lammps-dp version 1.3.2, which is a fork of LAMMPS compatible with DeepMD-kit.


Social Networking:
Share |


Item Type: Dataset
Status: Submitted
CreatorsEmailPitt UsernameORCID
Saidi, Wissamalsaidi@pitt.edu0000-0001-6714-4832
Date: 24 August 2021
DOI or Unique Handle: 10.18117/67ke-qg02
Schools and Programs: Swanson School of Engineering > Mechanical Engineering and Materials Science
Type of Data: Text
Copyright Holders: dylan bayerl
Additional Information: We have included a comprehensive catalog of relevant training data, output files, and validation data for reference, such as VASP input (INCAR and POSCAR) and output files (OUTCAR examples parse VASP outputs coord.raw, energies.raw, forces.raw, and virials.raw), LAMMPS input and output files, DP-GEN workflow configuration files (.json). These data and the archived ML potential files (.pb) allow for the reproduction and extension of these results by the scientific community.
Date Deposited: 27 Aug 2021 17:59
Last Modified: 03 Jan 2022 19:47


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item