Published December 12, 2022 | Version v0.1
Software Open

GSchnet for OE62 dataset

Authors/Creators

  • 1. Westermayr
  • 2. Barrett

Description

Version of the G-SchNet code to train the OE62 data set. (https://github.com/rhyan10/G-SchNetOE62/tree/v0.1)

For tutorials and installation, please see the original repository. For training G-SchNet on the OE62 data set, use this code and follow the tutorial, but replace the QM9 data set with the OE62 data set (https://www.nature.com/articles/s41597-020-0385-y). Note: The adapted script to train OE62 is "template_data", hence add "--datset_name template_data" to the command used for training.

A tutorial how to train and use G-SchNet for OE62 can be found here: Westermayr, Julia (2022): G-SchNet for OE62. figshare. Dataset. https://doi.org/10.6084/m9.figshare.20146943.v2

Files

G-SchNetOE62-0.1.zip

Files (2.3 MB)

Name Size Download all
md5:59ba1e9b157f550101c3371f92b35be3
2.3 MB Preview Download