[dev] update minimum torch version to 2.1.0

[Rhett-Ying]

Description

Checklist

Please feel free to remove inapplicable items for your PR.

• The PR title starts with [$CATEGORY] (such as [NN], [Model], [Doc], [Feature]])
• I've leverage the tools to beautify the python and c++ code.
• The PR is complete and small, read the Google eng practice (CL equals to PR) to understand more about small PR. In DGL, we consider PRs with less than 200 lines of core code change are small (example, test and documentation could be exempted).
• All changes have test coverage
• Code is well-documented
• To the best of my knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change
• Related issue is referred in this PR
• If the PR is for a new model/paper, I've updated the example index here.

Changes

[dgl-bot]

To trigger regression tests:

• @dgl-bot run [instance-type] [which tests] [compare-with-branch];
  For example: @dgl-bot run g4dn.4xlarge all dmlc/master or @dgl-bot run c5.9xlarge kernel,api dmlc/master

[dgl-bot]

Commit ID: 9165048

Build ID: 1

Status: ❌ CI test failed in Stage [PyTorch Cugraph GPU Unit test].

Report path: link

Full logs path: link

[Rhett-Ying]

@TristonC

We're updating the minimum pytorch version from 2.0.0 to 2.1.0 and CI is blocked by cugraph docker image: rapidsai/cugraph_stable_torch-cuda:11.8-base-ubuntu20.04-py3.10-pytorch2.0.0-rapids23.04.

On the web page: https://hub.docker.com/r/rapidsai/cugraph_stable_torch-cuda, I cannot find any newer ones.

So do you have ideas about how to handle this?

[frozenbugs]

@dgl-bot

[dgl-bot]

Commit ID: 9165048

Build ID: 2

Status: ❌ CI test failed in Stage [PyTorch Cugraph GPU Unit test].

Report path: link

Full logs path: link

[Rhett-Ying]

@TristonC

We're updating the minimum pytorch version from 2.0.0 to 2.1.0 and CI is blocked by cugraph docker image: rapidsai/cugraph_stable_torch-cuda:11.8-base-ubuntu20.04-py3.10-pytorch2.0.0-rapids23.04.

On the web page: https://hub.docker.com/r/rapidsai/cugraph_stable_torch-cuda, I cannot find any newer ones.

So do you have ideas about how to handle this?

@chang-l Hi, do you have any suggestion on this? Should we bypass the test on cugraph or upgrade docker image?

[chang-l]

@Rhett-Ying, can you please try switch to NVIDIA's public NGC DGL container for your CI/CD?(https://catalog.ngc.nvidia.com/orgs/nvidia/containers/dgl) docker pull nvcr.io/nvidia/dgl:24.04-py3

It includes latest PyTorch and latest RAPIDS (e.g., cugraph/wholegraph) for graph analysis.

[TristonC]

@Rhett-Ying We can pause (bypass) the cugraph test for now. cugraph is in the process of refactoring. We can add the test back once its ready. Or as @chang-l suggested to use NVIDIA DGL container.

[Rhett-Ying]

@Rhett-Ying We can pause (bypass) the cugraph test for now. cugraph is in the process of refactoring. We can add the test back once its ready. Or as @chang-l suggested to use NVIDIA DGL container.

Sure. Let's pause the cugraph test for now.

And is there any ticket link to track the status of cugraph refactoring? we could link it here.

[dgl-bot]

Commit ID: cb19e6bc4730f67057122ee707d31fdf9513b338

Build ID: 3

Status: ✅ CI test succeeded.

Report path: link

Full logs path: link