Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Build dataset

See original GitHub issue

Hi, Is it possible to build my own dataset via torchani? Is it necessary to use the built-in ANI dataset to train models? I did not find any information in the document.

Thanks.

Issue Analytics

State:
Created 4 years ago
Comments:5 (3 by maintainers)

Top GitHub Comments

1reaction

zasdfgbnmcommented, Nov 5, 2019

@njzjz You can have your own dataset in any format. But our dataset loader only supports the hdf5 we are using, as @farhadrgh mentioned, they should have have the same key as our format. But using our data loader is not mandatory, and you can always load your data in your favorite way, convert it to PyTorch tensors, and feed it to your training pipeline.

1reaction

farhadrghcommented, Nov 5, 2019

Yes, you can create your dataset and use it for training (not with TorchANI though) as long as each entry has the same keys as the sample HDF5 datasets. This should be done using libraries like h5py.

The required keys for compatible datasets are: coordinates, species, energies, and forces.