CederGroupHub / chgnet

Pretrained universal neural network potential for charge-informed atomistic modeling https://chgnet.lbl.gov
https://doi.org/10.1038/s42256-023-00716-3
Other
226 stars 61 forks source link

Report CHGNet device #3

Closed janosh closed 1 year ago

janosh commented 1 year ago

Just a few small suggestions here. I added print(f"CHGNet will run on {self.device}") as I find it helpful if the model reports what device it will be running on so I can see it in the slurm logs later. Helps me make sure I'm using GPU as expected.

Also added some literal type hints for targets: Literal["ef", "efs", "efsm"]. Modern IDEs will surface this information so you don't have to read the code or the docs to know the options.