Stanley's node_classification code for #7462

[az15240]

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: b0dba5e925a2b6cf71787d6a4e0cd4cf1b5e5c34

Build ID: 1

Status: ⚪️ CI test cancelled due to overrun.

Report path: link

Full logs path: link

[az15240]

I was using GPU on ogbn-products dataset.

Some data that I've collected:

prob: 20.44it/s (100% data)
prob: 23.74it/s (80% data)
prob: 27.71it/s (50% data)
prob: 32.93it/s (20% data)
prob: 60.71it/s (0% data)

mask: 20.67it/s (100% true)
mask: 24.16it/s (80% true)
mask: 28.42it/s (50% true)
Mask: 33.85it/s (20% true)
Mask: 67.51it/s (0% true)

None: 26.02it/s (100% true)

For example, prob_data[torch.randperm(num_edges, device=args.device)[: int(num_edges * 0.2)]] = 0.0 means that we will leave out 20% data to be 0, while the remaining 80% prob_data have a positive value.

The runtime above shows that the more edges we include in the prob/mask section, the slower it runs. It didn't show a 14x slow.

[dgl-bot]

Commit ID: 1414250

Build ID: 2

Status: ✅ CI test succeeded.

Report path: link

Full logs path: link

[dgl-bot]

Commit ID: addb4a44768e3bca260a98e605468bbd8335cf2a

Build ID: 3

Status: ✅ CI test succeeded.

Report path: link

Full logs path: link

[dgl-bot]

Commit ID: 9e6602f

Build ID: 4

Status: ⚪️ CI test cancelled due to overrun.

Report path: link

Full logs path: link

[dgl-bot]

Commit ID: b4df206

Build ID: 5

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

Report path: link

Full logs path: link