materials
Materials Project Trajectories
Deng, Persson, et al. (CHGNet) · 2023
Constitution
| Full name | Materials Project Trajectories |
| Curator | Deng, Persson, et al. (CHGNet) |
| Size | ~1.6M VASP relaxation frames extracted from Materials Project |
| DFT level | PBE; PBE+U for select TMs; VASP |
| Released | 2023 |
| License | CC-BY 4.0 |
| DOI | https://doi.org/10.1038/s42256-023-00716-3 |
| Primary paper | Deng et al., CHGNet as a pretrained universal neural network potential for charge-informed atomistic modelling, Nat. Mach. Intell. 5:1031 (2023). |
Models trained on MPTrj
| Model |
|---|
| CHGNet |
| EquiformerV2 |
| eSEN |
| MACE-MP-0 |
| Orb-v2 |