Pytorch geometric example for node classification using cora dataset

This repo contains the code for graph neural network implementation using pytorch geometric on the cora dataset.

Install pytorch geometric

Check pytorch version python -c "import torch; print(torch.__version__)"

Install pytorch geometric with the same pytorch version

pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html

pip install torch-sparse -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html

pip install torch-cluster -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html

pip install torch-spline-conv -f https://pytorch-geometric.com/whl/torch-1.8.0+cu101.html

pip install torch-geometric

To run the experiments:

First load data using: python load_data.py

Run the GCN model using: python node-classification-gcn.py --lr 0.01 --decay 5e-4 --hc 16 --epochs 1000

Run the GAT model using: python node-classification-gat.py --lr 0.01 --decay 5e-4 --hc 16 --epochs 1000

lr = learning rate
hc = number of neurons in the hidden layer