Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
Hongzhi (Steve), Chen 704bcaf6dd | 1 year ago | |
---|---|---|
.. | ||
README.md | 1 year ago | |
main.py | 1 year ago | |
preprocessing.py | 1 year ago |
This is an example of implementing directional_GSN for graph classification in DGL.
directional_GSN is a combination of Graph Substructure Networks (GSN) with Directional Graph Networks (DGN), where we defined a vector field based on substructure encoding instead of Laplacian eigenvectors.
The script in this folder experiments directional_GSN on ogbg-molpcba dataset.
conda create --name gsn python=3.7
conda activate gsn
conda install pytorch==1.11.0 cudatoolkit=10.2 -c pytorch
pip install tqdm
pip install networkx
conda install -c conda-forge graph-tool
pip install ogb
pip install dgl-cu102 -f https://data.dgl.ai/wheels/repo.html
We fix the random seed to 41, and train the model on a single Tesla T4 GPU with 16GB memory.
train_AP | valid_AP | test_AP | #parameters | |
---|---|---|---|---|
directional_GSN | 0.4301 | 0.2598 | 0.2438 | 5142713 |
python preprocessing.py
python main.py --seed 41 --epochs 450 --hidden_dim 420 --out_dim 420 --dropout 0.2
@article{bouritsas2020improving,
title={Improving graph neural network expressivity via subgraph isomorphism counting},
author={Bouritsas, Giorgos and
Frasca, Fabrizio and
Zafeiriou, Stefanos and
Bronstein, Michael M},
journal={arXiv preprint arXiv:2006.09252},
year={2020}
}
No Description
Python C++ Jupyter Notebook Cuda Text other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》