Support OGB datasets in GRACE.
<username>@node02:~/projects/GRACE$ pip list | grep torch
torch 1.12.0+cu116
torch-cluster 1.6.1
torch-geometric 2.0.4
torch-scatter 2.1.1
torch-sparse 0.6.17
torch-spline-conv 1.2.2
torchaudio 0.12.0+cu116
torchvision 0.13.0+cu116
Install other necessary dependencies with
pip install -r requirements.txt
Train and evaluate the model by executing
python train.py --gpu_id 2 --dataset ogbn-arxiv
The --dataset
argument should be one of [ Cora, CiteSeer, PubMed, DBLP, ogbn-arxiv, ogbn-products].