Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/jw9730/tokengt
[NeurIPS'22] Tokenized Graph Transformer (TokenGT), in PyTorch
https://github.com/jw9730/tokengt
equivariance gnn graph hypergraph pytorch self-attention transformer
Last synced: 4 months ago
JSON representation
[NeurIPS'22] Tokenized Graph Transformer (TokenGT), in PyTorch
- Host: GitHub
- URL: https://github.com/jw9730/tokengt
- Owner: jw9730
- License: mit
- Created: 2022-06-29T04:30:26.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2023-04-11T05:25:29.000Z (about 1 year ago)
- Last Synced: 2024-01-16T10:44:03.275Z (5 months ago)
- Topics: equivariance, gnn, graph, hypergraph, pytorch, self-attention, transformer
- Language: Python
- Homepage:
- Size: 1.23 MB
- Stars: 281
- Watchers: 9
- Forks: 43
- Open Issues: 10
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Lists
- awesome-stars - jw9730/tokengt - [NeurIPS'22] Tokenized Graph Transformer (TokenGT), in PyTorch (Python)
README
# Tokenized Graph Transformer (PyTorch)
[**Pure Transformers are Powerful Graph Learners**](https://arxiv.org/abs/2207.02505) \
[Jinwoo Kim](https://bit.ly/jinwoo-kim), [Tien Dat Nguyen](https://github.com/tiendatnguyen-vision), [Seonwoo Min](https://scholar.google.co.kr/citations?user=dWKk68wAAAAJ&hl=en), [Sungjun Cho](https://scholar.google.com/citations?user=bEilQPMAAAAJ&hl=en), [Moontae Lee](https://moontae.people.uic.edu/), [Honglak Lee](https://web.eecs.umich.edu/~honglak/), [Seunghoon Hong](https://maga33.github.io/) \
NeurIPS 2022
## Setting up experiments
Using the provided Docker image (recommended)
```bash
docker pull jw9730/tokengt:latest
docker run -it --gpus=all --ipc=host --name=tokengt -v /home:/home jw9730/tokengt:latest bash
# upon completion, you should be at /tokengt inside the container
```
Using the provided ```Dockerfile```
```bash
git clone --recursive https://github.com/jw9730/tokengt.git /tokengt
cd tokengt
docker build --no-cache --tag tokengt:latest .
docker run -it --gpus all --ipc=host --name=tokengt -v /home:/home tokengt:latest bash
# upon completion, you should be at /tokengt inside the container
```
Using ```pip```
```bash
sudo apt-get update
sudo apt-get install python3.9
git clone --recursive https://github.com/jw9730/tokengt.git tokengt
cd tokengt
bash install.sh
```
## Running experiments
Synthetic second-order equivariant basis approximation
```bash
cd equivariant-basis-approximation/scripts
# Train and save logs, ckpts, and attention maps (--save_display)
bash [INPUT]-[NODE_IDENTIFIER]-[TYPE_IDENTIFIER].sh
# Test and save attention maps (--save_display)
bash [INPUT]-[NODE_IDENTIFIER]-[TYPE_IDENTIFIER]-test.sh
# For the visualization of saved attention maps, please see viz_multi.ipynb
```
PCQM4Mv2 large-scale graph regression
```bash
cd large-scale-regression/scripts
# TokenGT (ORF)
bash pcqv2-orf.sh
# TokenGT (Lap)
bash pcqv2-lap.sh
# TokenGT (Lap) + Performer
bash pcqv2-lap-performer-finetune.sh
# TokenGT (ablated)
bash pcqv2-ablated.sh
# Attention distance plot for TokenGT (ORF)
bash visualize-pcqv2-orf.sh
# Attention distance plot for TokenGT (Lap)
bash visualize-pcqv2-lap.sh
```
## Pre-Trained Models
We provide checkpoints of ```TokenGT (ORF)``` and ```TokenGT (Lap)```, both trained with PCQM4Mv2.
Please download ```ckpts.zip``` from [this link](https://drive.google.com/drive/folders/1mo0dV-aLxGFWbPF8xfE8phWTmOtIV1HG?usp=sharing).
Then, unzip ```ckpts``` and place it in the ```large-scale-regression/scripts``` directory, so that each trained checkpoint is located at ```large-scale-regression/scripts/ckpts/pcqv2-tokengt-[NODE_IDENTIFIER]-trained/checkpoint_best.pt```.
After that, you can resume the training from these checkpoints by adding the option ```--pretrained-model-name pcqv2-tokengt-[NODE_IDENTIFIER]-trained``` to the training scripts.
## References
Our implementation uses code from the following repositories:
- [Performer](https://github.com/lucidrains/performer-pytorch.git) for FAVOR+ attention kernel
- [Graph Transformer](https://github.com/graphdeeplearning/graphtransformer.git), [SignNet](https://github.com/cptq/SignNet-BasisNet.git), and [SAN](https://github.com/DevinKreuzer/SAN/blob/main/data/molecules.py) for Laplacian eigenvectors
- [Graphormer](https://github.com/microsoft/Graphormer.git) for PCQM4Mv2 experiment pipeline
- [timm](https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/layers/drop.py) for stochastic depth regularization
## Citation
If you find our work useful, please consider citing it:
```bib
@article{kim2022pure,
author = {Jinwoo Kim and Tien Dat Nguyen and Seonwoo Min and Sungjun Cho and Moontae Lee and Honglak Lee and Seunghoon Hong},
title = {Pure Transformers are Powerful Graph Learners},
journal = {arXiv},
volume = {abs/2207.02505},
year = {2022},
url = {https://arxiv.org/abs/2207.02505}
}
```
## Acknowledgements
The development of this open-sourced code was supported in part by the National Research Foundation of Korea (NRF) (No. 2021R1A4A3032834).