Clean up examples and improve cuGraph docs

[puririshi98]

The examples/multi_gpu folder is stale. The data parallel examples are all stale in lieu of new updates from cugraph making a single unified api that handles any scale of compute seamlessly. (single gpu, single node multigpu, and multinode multigpu)
It is too much effort to mantain mirrored cugraph examples in pyg when we can just point directly to the cugraph examples. @alexbarghi-nv to review.

Model parralel is best recommended through PyT as opposed to mantaining our own third party examples

https://github.com/rapidsai/cugraph-gnn/tree/branch-25.12/python/cugraph-pyg/cugraph_pyg/examples

[alexbarghi-nv]

👍

[puririshi98]

@akihironitta @rusty1s @wsad1 ready for merge

[akihironitta]

@puririshi98 Mind doing a favor before merging this PR?

• List all the up-to-date cuGraph examples in the PR description so that we can guide PyG users to the up-to-date examples?
• Make sure there're no reference in the codebase or docs to these examples being deleted.

@alexbarghi-nv @puririshi98 Are there any considerations/implications of using cuGraph? It seems performant, but for example, I wonder how hard it'd be for users to try cuGraph because real use cases usually have many other (old) dependencies that may conflict with cuGraph's dependencies (, which was my experience with cudf in the past). Because of that, I'm slightly in favor of keeping these pure PyTorch examples as they are.

[akihironitta]

Here're some changes I made:

• Reverted changes to kuzu and graphlearn examples because they're mainly for showcasing how one can integrate 3rd-party libraries with PyG.
• Reverted all changes to examples/mulit_gpu/. I think I needed to decouple the changes from this PR because it removed more than this PR should in my opinion. I will submit a PR and ping you for your review regarding this.
• Added references to https://pytorch-geometric.readthedocs.io/en/latest/install/installation.html#accelerating-pyg-with-nvidia-cugraph-gnn which I added to the installation page in our docs the other day in Document more on cuGraph GNN #10603.