HydraGNN

Distributed PyTorch implementation of multi-headed graph convolutional neural networks

Dependencies

To install required packages with only basic capability (torch, torch_geometric, and related packages) and to serialize+store the processed data for later sessions (pickle5):

pip install -r requirements.txt
pip install -r requirements-torchdep.txt

If you plan to modify the code, include packages for formatting (black) and testing (pytest) the code:

pip install -r requirements-dev.txt

Detailed dependency installation instructions are available on the Wiki

Running the code

There are two main options for running the code; both require a JSON input file for configurable options.

Training a model, including continuing from a previously trained model using configuration options:
```
import hydragnn
hydragnn.run_training("examples/configuration.json")
```

Making predictions from a previously trained model:

import hydragnn
hydragnn.run_prediction("examples/configuration.json", model)

Datasets

Built in examples are provided for testing purposes only. One source of data to create HydraGNN surrogate predictions is DFT output on the OLCF Constellation: https://doi.ccs.ornl.gov/

Detailed instructions are available on the Wiki

Configurable settings

HydraGNN uses a JSON configuration file (examples in examples/):

There are many options for HydraGNN; the dataset and model type are particularly important:

["Verbosity"]["level"]: 0, 1, 2, 3, 4
["Dataset"]["name"]: CuAu_32atoms, FePt_32atoms, FeSi_1024atoms
["NeuralNetwork"]["Architecture"]["model_type"]: PNA, MFC, GIN, GAT, CGCNN

Citations

"HydraGNN: Distributed PyTorch implementation of multi-headed graph convolutional neural networks", Copyright ID#: 81929619 https://doi.org/10.11578/dc.20211019.2

Contributing

We encourage you to contribute to HydraGNN! Please check the guidelines on how to do so.

frobnitzem / HydraGNN