GitHub - NTUDDSNLab/SRGCL: Self-Reinforced Graph Contrastive Learning (2025 arxiv)

Requirement

python 3.9.16
torch 2.4.1
torchvision 0.19.1

Project Directory Structure

  • autogcl/
  • data/ - Directory for storing datasets
  • graphcl/
  • .gitignore
  • README.md

Note: Place your data in the current directory.

GraphCL

Run

Please check the README in graphcl/ for more details.

cd graphcl
CUDA_VISIBLE_DEVICES=0 python gcl_gcl_method6_temperature.py --DS='MUTAG' --aug='dnodes'

AutoGCL

Run

Please check the README in autogcl/ for more details.

cd autogcl/unsupervised
CUDA_VISIBLE_DEVICES=0 python us_main_temperature_hybrid2.py --dataset='MUTAG'