SuperGAT
Official implementation of Self-supervised Graph Attention Networks (SuperGAT). This model is presented at How to Find Your Friendly Neighborhood: Graph Attention Design with Self-Supervision, International Conference on Learning Representations (ICLR), 2021.
Notice
The documented SuperGATConv layer with an example has been merged to the PyTorch Geometric's main branch.
This repository is based on torch==1.4.0+cu100 and torch-geometric==1.4.3, which are somewhat outdated at this point (Feb 2021). If you are using recent PyTorch/CUDA/PyG, we would recommend using the PyG's. If you want to run codes in this repository, please follow #installation.
Installation
# In SuperGAT/
bash install.sh ${CUDA, default is cu100}
- If you have any trouble installing PyTorch Geometric, please install PyG's dependencies manually.
- Codes are tested with python
3.7.6andnvidia/cuda:10.0-cudnn7-devel-ubuntu16.04image. - PYG's FAQ might be helpful.
Basics
- The main train/test code is in
SuperGAT/main.py. - If you want to see the SuperGAT layer in PyTorch Geometric
MessagePassinggrammar, refer toSuperGAT/layer.py. - If you want to see hyperparameter settings, refer to
SuperGAT/args.yamlandSuperGAT/arguments.py.
Run
python3 SuperGAT/main.py \
--dataset-class Planetoid \
--dataset-name Cora \
--custom-key EV13NSO8-ES
...
## RESULTS SUMMARY ##
best_test_perf: 0.853 +- 0.003
best_test_perf_at_best_val: 0.851 +- 0.004
best_val_perf: 0.825 +- 0.003
test_perf_at_best_val: 0.849 +- 0.004
## RESULTS DETAILS ##
best_test_perf: [0.851, 0.853, 0.857, 0.852, 0.858, 0.852, 0.847]
best_test_perf_at_best_val: [0.851, 0.849, 0.855, 0.852, 0.858, 0.848, 0.844]
best_val_perf: [0.82, 0.824, 0.83, 0.826, 0.828, 0.824, 0.822]
test_perf_at_best_val: [0.851, 0.844, 0.853, 0.849, 0.857, 0.848, 0.844]
Time for runs (s): 173.85422565042973
The default setting is 7 runs with different random seeds. If you want to change this number, change num_total_runs in the main block of SuperGAT/main.py.
For ogbn-arxiv, use SuperGAT/main_ogb.py.
GPU Setting
There are three arguments for GPU settings (--num-gpus-total, --num-gpus-to-use, --gpu-deny-list). Default values are from the author's machine, so we recommend you modify these values from SuperGAT/args.yaml or by the command line.
--num-gpus-total(default 4): The total number of GPUs in your machine.--num-gpus-to-use(default 1): The number of GPUs you want to use.--gpu-deny-list(default: [1, 2, 3]): The ids of GPUs you want to not use.
If you have four GPUs and want to use the first (cuda:0),
python3 SuperGAT/main.py \
--dataset-class Planetoid \
--dataset-name Cora \
--custom-key EV13NSO8-ES \
--num-gpus-total 4 \
--gpu-deny-list 1 2 3
Model (--model-name)
| Type | Model name |
|---|---|
| GCN | GCN |
| GraphSAGE | SAGE |
| GAT | GAT |
| SuperGATGO | GAT |
| SuperGATDP | GAT |
| SuperGATSD | GAT |
| SuperGATMX | GAT |
Dataset (--dataset-class, --dataset-name)
| Dataset class | Dataset name |
|---|---|
| Planetoid | Cora |
| Planetoid | CiteSeer |
| Planetoid | PubMed |
| PPI | PPI |
| WikiCS | WikiCS |
| WebKB4Univ | WebKB4Univ |
| MyAmazon | Photo |
| MyAmazon | Computers |
| PygNodePropPredDataset | ogbn-arxiv |
| MyCoauthor | CS |
| MyCoauthor | Physics |
| MyCitationFull | Cora_ML |
| MyCitationFull | CoraFull |
| MyCitationFull | DBLP |
| Crocodile | Crocodile |
| Chameleon | Chameleon |
| Flickr | Flickr |
Custom Key (--custom-key)
| Type | Custom key (General) | Custom key (for PubMed) | Custom key (for ogbn-arxiv) |
|---|---|---|---|
| SuperGATGO | EV1O8-ES | EV1-500-ES | - |
| SuperGATDP | EV2O8-ES | EV2-500-ES | - |
| SuperGATSD | EV3O8-ES | EV3-500-ES | EV3-ES |
| SuperGATMX | EV13NSO8-ES | EV13NSO8-500-ES | EV13NS-ES |
Other Hyperparameters
See SuperGAT/args.yaml or run $ python3 SuperGAT/main.py --help.
113 Nov 27, 2022
20 Dec 31, 2021
43 Dec 7, 2022
32 Jan 2, 2023
175 Dec 1, 2022
36 Dec 12, 2022
92 Nov 30, 2022
49 Nov 23, 2022
71 Dec 20, 2022
2 Oct 31, 2022
8 Oct 15, 2021
329 Dec 30, 2022
56 Jan 2, 2023
50 Dec 9, 2022
1.4k Dec 23, 2022
19 Oct 21, 2022
189 Nov 22, 2022
12 Oct 28, 2022
