se3transformers demo

[awelling2801]

Motivation

The AIOSS Ecosystems team is developing a Jupyter notebook demo to showcase its capabilities to customers. The DGL project and the se3transformers example serves as the first step toward building this interactive demonstration.

Technical Details

Most of the technical content in this notebook is based on the work contributed by @jamesETsmith

The notebook features -

1. Args, Dataset and models setup
2. An interactive module to inspect the QM9 dataset molecule
3. Training
4. Inference
5. Visualizing the Train Loss, Mean Absolute Error and Learning Rate

Submission Checklist

• Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

[mukh1l]

Looks Good. Great Job

[jamesETsmith]

This looks great thanks @awelling2801! I think it would be helpful to have a minimal markdown file explaining how to setup/run the docker container since many python people aren't necessarily docker experts.

Could you also clear the output from the jupyter notebook with something like: jupyter nbconvert --clear-output --inplace notebook.ipynb? That way the diff will be easy to read in the PR in case we need to come back to this.