search

atomistic-machine-learning / schnetpack

SchNetPack - Deep Neural Networks for Atomistic Systems
Other
774 stars 214 forks source link

qm9 needs remove_uncharacterized=False now? #587

Closed Maltimore closed 10 months ago

Maltimore commented 10 months ago

The examples in the Readme typically result in an error: One has to add data.remove_uncharacterized=False for them to run properly. I am not sure whether just the Readme should be updated, or whether this is a more fundamental problem that should be fixed elsewhere?

jnsLs commented 10 months ago

At some point the default setting in the dataloader changed. Have you tried removing your QM9.db file and running the training again?

Maltimore commented 10 months ago

Your suggestion worked, sorry for the noise!

Maltimore commented 10 months ago

Closing as this is not actually an issue.