Founding Engineer
Creator of PyG (PyTorch Geometric) - Founding Engineer @ kumo.ai - PhD from TU Dortmund University - Interested in Graph Representation Learning
- Dortmund, Germany
- https://rusty1s.github.io
- @rusty1s
Highlights
- Pro
Block or Report
Block or report rusty1s
Report abuse
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePinned
-
-
snap-stanford/ogb Public
Benchmark datasets, data loaders, and evaluators for graph machine learning
-
-
pytorch_sparse Public
PyTorch Extension Library of Optimized Autograd Sparse Matrix Operations
-
pyg_autoscale Public
Implementation of "GNNAutoScale: Scalable and Expressive Graph Neural Networks via Historical Embeddings" in PyTorch
-
Implementation of "Deep Graph Matching Consensus" in PyTorch
5,143 contributions in the last year
Less
More
Contribution activity
March 2023
Created a pull request in pyg-team/pytorch_geometric that received 7 comments
Opened 37 other pull requests in 6 repositories
pyg-team/pytorch_geometric
30
merged
-
Evaluate
torch.compileagainst custom GNN layer -
Added
HeteroData.set_value_dictfunctionality - Add PyTorch 1.13 CI tests
-
Disable
torch.compilelogging in tests -
Add support for sparse CSR tensors in
gcn_norm -
Added
to_torch_csc_tensorfunctionality -
torch.sparse.mmbenchmark script -
Introduce
index2ptrandptr2indexhelper functions -
torch.sparse_csrsupport inMessagePassing - Test minimal installation
-
Fix PyTorch 2.0 warnings and add
torch.sparse_csrutilities - Fixx nightly tests on Windows
- PyTorch 2.0 test support
- PyTorch Lightning 2.0 integration
-
Bring back
EllipticBitcoinDatasettest -
Fix
DBLPdownload, moveget_edge_indextotestingpackage -
Add note in documentation about interface change in
sort_edge_index,coalesceandto_undirected - Remove remaining warnings during full CI test
-
Fix
DegreeScalerto not divide by zero -
Changed the interface of
sort_edge_indexto only return singleedge_indexinformation in case theedge_attrargument is not specified - Fix GraphGym tests on Windows
- Fix example CI path
-
Use
tmp_pathinpytest -
Added the
HeterophilousGraphDatasetsuite -
Fix tests in case
pyg-libis not installed - Some pull requests not shown.
rusty1s/pytorch_sparse
2
merged
pyg-team/pyg-lib
2
merged
rusty1s/pytorch_cluster
1
merged
rusty1s/pytorch_scatter
1
merged
rusty1s/pytorch_spline_conv
1
merged
Reviewed 30 pull requests in 5 repositories
pyg-team/pytorch_geometric
25 pull requests
-
feat: add
DistMultKGE model -
Pytorch Sparse Tensor support:
PointConv,PointGNNConv,PointTransformerConv,PPFConv, andResGatedGraphConv -
PyTorch Sparse Tensor support:
HANConv,GATv2Conv,HGTConv,GMMConv,GPSConv, andRGATConv -
Fix typo in
AddRandomWalkPEdocumentation -
feat: Add
ComplExKGE -
Pytorch Sparse tensor support:
GATConv,GatedGraphConv,GCN2Conv, andGENConv -
Enable
filter-per-workeras argument in benchmarks - Fix distributed batching example
-
Fix
cugraphtests -
Small typo in the
LightGCNdocumentation - Add option to save benchmark results to csv
-
Respect tuple of
dataobjects inDataset.num_classes -
Fix failing
PyGModelHubMixintests -
Pytorch Sparse tensor support:
ClusterGCN,SAGEConv,AGNNConv,APPNP, andFeaStConv -
Solve
cugraphtest failures -
Added
meanreduction toutils.spmmfor PyTorch Sparse Tensor -
pyg-libintegration inHGTConv -
Fix
to_heterowhen using an uninitialized submodule without implementingreset_parameters -
Add
HeteroLayerNormandHeteroBatchNorm - Restructure examples folder (1/n)
-
Enable
persistent_workersin GraphGym dataloaders - Add sparse tensor support in full batch mode
-
Enable
NeighborLoaderto return number of sampled nodes and edges per hop -
[Code Coverage]
explain/pg_explainer.py - Update benchmark README
- Some pull request reviews not shown.
rusty1s/pytorch_cluster
1 pull request
rusty1s/pytorch_sparse
1 pull request
openai/triton
1 pull request
pyg-team/pyg-lib
1 pull request
Answered 7 discussions in 2 repositories
pyg-team/pytorch_geometric
pyg-team/pytorch_geometric
-
How to train graph classification on IMDB-BINARY?
This contribution was made on Mar 16
-
Best practice for dividing/multiplying hidden channels?
This contribution was made on Mar 16
-
Getting weight of a MLP Model
This contribution was made on Mar 16
-
Second InMemoryDataset in the same notebook comes out as a list
This contribution was made on Mar 11
-
How to make a GNN model scalable?
This contribution was made on Mar 6
-
How to compute the Floating point operations (FLOPs) for a given model?
This contribution was made on Mar 2
pyg-team/pyg-lib
pyg-team/pyg-lib
-
What are the differences between `segment_matmul` and `grouped_matmul` ?
This contribution was made on Mar 2
112
contributions
in private repositories
Mar 1 – Mar 18






