DRAFT: timeouts deselection on GPU deselection

[samir-nasibli]

Description

GPU timeouts deselections

• Kmeans sparse

After initializing centroids via Kmeans_Init it gets stuck here

scikit-learn-intelex/onedal/cluster/kmeans.py

Line 254 in 1c81337

result = module.train(policy, params, X_table, centroids_table)

So actually the issue on onedal backend side.

Checklist to comply with before moving PR from draft:

PR completeness and readability

• I have reviewed my changes thoroughly before submitting this pull request.
• I have commented my code, particularly in hard-to-understand areas.
• I have updated the documentation to reflect the changes or created a separate PR with update and provided its number in the description, if necessary.
• Git commit message contains an appropriate signed-off-by string (see CONTRIBUTING.md for details).
• I have added a respective label(s) to PR if I have a permission for that.
• I have resolved any merge conflicts that might occur with the base branch.

Testing

• The unit tests pass successfully.
• I have run it locally and tested the changes extensively.

Performance

• I have measured performance for affected algorithms using scikit-learn_bench and provided at least summary table with measured data, if performance change is expected.
• I have provided justification why performance has changed or why changes are not expected.

[samir-nasibli]

Just for debug; will be reverted.

[ethanglaser]

/intelci: run

[samir-nasibli]

/intelci: run

[icfaust]

Looking at the deselections: is this just all sparse tests for kmeans?

[samir-nasibli]

/intelci: run

[samir-nasibli]

Looking at the deselections: is this just all sparse tests for kmeans?

Yeah, it seems only Kmeans sparse on GPU affected with timeouts. Expecting to have green CI jobs for GPU conformance validation with this PR

[samir-nasibli]

/intelci: run

[samir-nasibli]

Interesting fact: sklearnex own tests for the KMeans doesn't cover such cases from the conformance deselected.

[samir-nasibli]

Last CI run: http://intel-ci.intel.com/ef76956a-860f-f1fb-ada5-a4bf010d0e2e

[ethanglaser]

Should be covered by oneapi-src/oneDAL#2906 (if not feel free to re-open)