We believe in open-source code development and hence all contributions that helps to grow and build IGB are most welcome.
Please open a github issue request for posting questions, issues or feedbacks. We try to respond as early as possible.
Please create pull request for all contributions. We currently have not enabled CI/CD pipeline on IGB codebase and hence all code will be going through manual review process. If anyone is interested in building CI/CD pipeline, your contributions are most welcome.
Each PR will have an assignee and will engage with the contributor to merge the PR. Please actively ping assignees after you have addresses the comments requested.
This list describes the activities were are actively looking for collaborators to help enhance the IGB dataset.
- Setting up CI/CD pipelines
- Adding GNN models apart from the ones that is currently provided.
- Ranking
- @Elnifio (NVIDIA) (train_multi_hetero.py, mlperf pyg/glt script)
- @kaixuanliu (Alibaba) (fixes to the download script)