dmlc--dgl
Mufei Li
3f6f694159
* Update * Update * Update dependencies * Update * Update * Fix ogbn-products gat * Update * Update * Reformat * Fix typo in node2vec_random_walk * Specify file encoding * Working for 6.7 * Update * Fix subgraph * Fix doc for sample_neighbors_biased * Fix hyperlink * Add example for udf cross reducer * Fix * Add example for slice_batch * Replace dgl.bipartite * Fix GATConv * Fix math rendering * Fix doc Co-authored-by: Ubuntu <ubuntu@ip-172-31-28-17.us-west-2.compute.internal> Co-authored-by: Jinjing Zhou <VoVAllen@users.noreply.github.com> Co-authored-by: Ubuntu <ubuntu@ip-172-31-22-156.us-west-2.compute.internal>
84 行
2.6 KiB
Markdown
84 行
2.6 KiB
Markdown
Inductive Representation Learning on Large Graphs (GraphSAGE)
|
|
============
|
|
|
|
- Paper link: [http://papers.nips.cc/paper/6703-inductive-representation-learning-on-large-graphs.pdf](http://papers.nips.cc/paper/6703-inductive-representation-learning-on-large-graphs.pdf)
|
|
- Author's code repo: [https://github.com/williamleif/graphsage-simple](https://github.com/williamleif/graphsage-simple). Note that the original code is
|
|
simple reference implementation of GraphSAGE.
|
|
|
|
Requirements
|
|
------------
|
|
- requests
|
|
|
|
```bash
|
|
pip install requests
|
|
```
|
|
|
|
|
|
Results
|
|
-------
|
|
|
|
### Full graph training
|
|
|
|
Run with following (available dataset: "cora", "citeseer", "pubmed")
|
|
```bash
|
|
python3 train_full.py --dataset cora --gpu 0 # full graph
|
|
```
|
|
|
|
* cora: ~0.8330
|
|
* citeseer: ~0.7110
|
|
* pubmed: ~0.7830
|
|
|
|
### Minibatch training
|
|
|
|
Train w/ mini-batch sampling (on the Reddit dataset)
|
|
```bash
|
|
python3 train_sampling.py --num-epochs 30 # neighbor sampling
|
|
python3 train_sampling.py --num-epochs 30 --inductive # inductive learning with neighbor sampling
|
|
python3 train_cv.py --num-epochs 30 # control variate sampling
|
|
```
|
|
|
|
For multi-gpu training
|
|
```bash
|
|
python3 train_sampling_multi_gpu.py --num-epochs 30 --gpu 0,1,... # neighbor sampling
|
|
python3 train_sampling_multi_gpu.py --num-epochs 30 --inductive --gpu 0,1,... # inductive learning
|
|
python3 train_cv_multi_gpu.py --num-epochs 30 --gpu 0,1,... # control variate sampling
|
|
```
|
|
|
|
Accuracy:
|
|
|
|
| Model | Accuracy |
|
|
|:---------------------:|:--------:|
|
|
| Full Graph | 0.9504 |
|
|
| Neighbor Sampling | 0.9495 |
|
|
| N.S. (Inductive) | 0.9460 |
|
|
| Control Variate | 0.9490 |
|
|
|
|
### Unsupervised training
|
|
|
|
Train w/ mini-batch sampling in an unsupervised fashion (on the Reddit dataset)
|
|
```bash
|
|
python3 train_sampling_unsupervised.py
|
|
```
|
|
|
|
Notably,
|
|
|
|
* The loss function is defined by predicting whether an edge exists between two nodes or not. This matches the official
|
|
implementation, and is equivalent to the loss defined in the paper with 1-hop random walks.
|
|
* When computing the score of `(u, v)`, the connections between node `u` and `v` are removed from neighbor sampling.
|
|
This trick increases the F1-micro score on test set by 0.02.
|
|
* The performance of the learned embeddings are measured by training a softmax regression with scikit-learn, as described
|
|
in the paper.
|
|
|
|
Micro F1 score reaches 0.9212 on test set.
|
|
|
|
### Training with PyTorch Lightning
|
|
|
|
We also provide minibatch training scripts with PyTorch Lightning in `train_lightning.py` and `train_lightning_unsupervised.py`.
|
|
|
|
Requires `pytorch_lightning` and `torchmetrics`.
|
|
|
|
```bash
|
|
python3 train_lightning.py
|
|
python3 train_lightning_unsupervised.py
|
|
```
|