[Roadmap] DGL-Go
- Dominant language
- Python
- Stars
- 14.3k
- Forks
- 3.1k
- PR merge metrics
- No merged PRs in 30d
Description
DGL-Go is DGL's command line interface (CLI). See https://github.com/dmlc/dgl/tree/master/dglgo
- [ ] Add `dgl apply` command to apply a trained GNN model to a dataset.
- [ ] Add the training pipeline for graph prediction task and release:
- [ ] 2 models (GIN w/ virtual node, PNA)
- [ ] 2 datasets (ogbg-molhiv, ogbg-molpcba)
- [ ] 4 off-the-shelf recipes
- [ ] **Customizability**
- [ ] Support importing custom models to DGL-Go pipeline
- [ ] Support importing custom datasets (beyond CSV datasets)
- [ ] Support custom loggers (e.g., TensorboardLogger).
- [ ] Allow users to configure graph data transforms.
- [ ] Support loading pre-trained models for fine-tuning.
- [ ] GUI interface for `dgl configure`
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.