train and HPO python scripts added for GFM testcase

ORNL / HydraGNN

Distributed PyTorch implementation of multi-headed graph convolutional neural networks

BSD 3-Clause "New" or "Revised" License

68 stars 29 forks source link

Closed allaffa closed 7 months ago

allaffa commented 7 months ago

Removed unused imports from QM9 example
Introduced iterative objects that spans all output heads in config dictionary to change num_headlayers and dim_headlayers consistently across all heads (both for graph and node types)
Introduced example that combines multidatasets for GFM and HPO