Tingyu Wang

[Feature] Create an accelerated RGCN model using cugraph-ops

11

## Description This PR adds an accelerated R-GCN model that uses the aggregation functions in `pylibcugraphops`, a Python wrapper of `cugraph-ops`. Closes https://github.com/rapidsai/cugraph-ops/issues/179. Closes #4669. Link to internal reviews https://github.com/rapidsai/dgl/pull/15....

## 🚀 Feature This integration creates an accelerated R-GCN model that uses the aggregation functions from `pylibcugraphops`, a Python wrapper around cugraph-ops aggregators. Per our previous discussion, this new model...

Update cugraph-pyg models for PyG 2.5

2

Support [`torch_geometric.EdgeIndex`](https://pytorch-geometric.readthedocs.io/en/latest/generated/torch_geometric.edge_index.EdgeIndex.html) in models. Breaking changes: The `csc` argument in `model.forward()` has been renamed to `edge_index` to align with upstream models. For now, users can still pass in CSC tuples...

Fix TensorProductConv test and improve docs

2

Closes https://github.com/rapidsai/cugraph/issues/4459

`cupy` wheel and `torch` wheel link to different NCCL shared libraries in RAPIDS CI containers

5

We have experienced the same issue several times in CI wheel-tests workflows when using `cupy` and `torch>=2.2` together: ``` torch = import_optional("torch") /pyenv/versions/3.9.19/lib/python3.9/site-packages/cugraph/utilities/utils.py:455: in import_optional return importlib.import_module(mod) /pyenv/versions/3.9.19/lib/python3.9/site-packages/torch/__init__.py:237: in from...