This repository contains the code for Cora and Citeseer dataset classfication.
Usage: python main.py [--dataset DATASET] [--model MODEL]
Options: --dataset DATASET Specify the dataset to use. Possible values: 'Cora', 'CiteSeer'. Default is 'Cora'. --model MODEL Specify the model to use. Possible values: 'GFusion_1', 'VanillaGAT', 'VanillaGCN'. Default is 'VanillaGAT'.
Examples:
python main.py --dataset Cora --model VanillaGAT python main.py --dataset CiteSeer --model VanillaGCN
Additional Options:
--epochs EPOCHS Number of epochs to train. Default is 1001.
--lr LR Initial learning rate. Default is 0.01.
--weight_decay DECAY Weight decay (L2 loss) coefficient. Default is 5e-4.
To run the model with custom settings, include these arguments in your command line. For example: python main.py --dataset Cora --model VanillaGAT --epochs 200 --lr 0.005 --weight_decay 1e-4