v2023.01.09

Added init.py ff , else cannot import ALIGNN-FF.

v2023.01.08

Added supercon model, added FF dataset for elements from Mlearn, early_stopping fix version update.

v2022.11.06

• Fix pytorch version to match torchvision, which is necessary for atomvision
• https://github.com/pytorch/vision/blob/cb60e97a778ad6626980b67c456984e5fcf1d507/README.rst#L22

v2022.11.05

• ALIGNN-FF and ALIGNN-Evac model added.
• Updated README

v2022.10.23

• Added number of classes in alignn.

v2022.09.22

• ALIGNN PhononDOS model added.

v2022.01.25

• get_multiple_predictions added for high throughput predictions with trained ALIGNN models
• Added note about QM9 results issue in README

v2021.12.27

• Python 3.7 requirement in setup.py to make colab work.

v2021.12.26

• README update
• Added QETB database training script
• Removed dgl-cu101 from setup.py as some mac systems are having issue to directly using setup.py. They should install using pip install dgl-cuXYZ command.

v2021.11.16

• Minor fixes to avoid printing model params while loading pretrained model
• Added pepy.tech downloads