Improve detection of scikit-learn parity regressions#6553
Merged
rapids-bot[bot] merged 8 commits intorapidsai:branch-25.06from Apr 25, 2025
Merged
Improve detection of scikit-learn parity regressions#6553rapids-bot[bot] merged 8 commits intorapidsai:branch-25.06from
rapids-bot[bot] merged 8 commits intorapidsai:branch-25.06from
Conversation
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
csadorf
added
improvement
Contributor
Author
|
/ok to test 5759457 |
Contributor
Author
|
/ok to test 8d56c54 |
Contributor
Author
|
/ok to test 1772262 |
csadorf
added a commit
to csadorf/cuml
that referenced
this pull request
Apr 18, 2025
…sai#6553) This change modifies the KMeans test to set n_init to 2, enhancing the stability of the k-means|| initialization process. Additional context can be found in issue rapidsai#5530.
rapids-bot Bot
pushed a commit
that referenced
this pull request
Apr 18, 2025
#6555) This change modifies the KMeans test to set n_init to 2, enhancing the stability of the k-means|| initialization process. Additional context can be found in issue #5530. Closes #5530. Authors: - Simon Adorf (https://github.com/csadorf) Approvers: - Divye Gala (https://github.com/divyegala) - Jim Crist-Harif (https://github.com/jcrist) URL: #6555
csadorf
force-pushed
the
ci/accel-skl-regression-testing
branch
from
1772262 to
fd2d0b2
Compare
Contributor
Author
|
/ok to test fd2d0b2 |
csadorf
force-pushed
the
ci/accel-skl-regression-testing
branch
from
fd2d0b2 to
a80327d
Compare
Contributor
Author
|
/ok to test a80327d |
Contributor
Author
|
/ok to test d5ed917 |
Contributor
Author
|
/ok to test e00dc0b |
csadorf
force-pushed
the
ci/accel-skl-regression-testing
branch
from
e00dc0b to
86819cf
Compare
Contributor
Author
|
/ok to test 86819cf |
- Move most of the related tooling from ci/ to python/cuml/cuml/accel/tests/scikit-learn/ - Add xfail-list.yaml to track known differences from scikit-learn - Add --format=xfail_list to automatically generate xfail lists from failures The infrastructure enables continuous monitoring of scikit-learn parity by: - Running scikit-learn tests with cuML acceleration - Tracking pass rates and comparing against thresholds - Automatically detecting new failures - Managing known differences via xfail lists Import pytest lazily within the pytest_collection_modifyitems function to avoid requiring it for normal cuML usage, ensuring pytest is only needed during test execution.
csadorf
force-pushed
the
ci/accel-skl-regression-testing
branch
from
86819cf to
58f24eb
Compare
csadorf
changed the title
jcrist
approved these changes
Apr 24, 2025
Member
jcrist
left a comment
jcrist
left a comment
There was a problem hiding this comment.
Two tiny nits, but otherwise this looks great! Thanks for working on this.
jameslamb
approved these changes
Apr 25, 2025
Member
jameslamb
left a comment
jameslamb
left a comment
There was a problem hiding this comment.
Approving for packaging-codeowners / ci-codeowners.
Full disclosure, I only skimmed summarize-results.py, since it looks pretty complex but also isn't user-facing.
Contributor
Author
|
/merge |
csadorf
added a commit
to csadorf/cuml
that referenced
this pull request
Apr 30, 2025
As a mitigation for rapidsaigh-6597. Partial revert of rapidsaigh-6553.
rapids-bot Bot
pushed a commit
that referenced
this pull request
Apr 30, 2025
As a mitigation for gh-6597. Partial revert of gh-6553. Authors: - Simon Adorf (https://github.com/csadorf) Approvers: - Jim Crist-Harif (https://github.com/jcrist) - Jake Awe (https://github.com/AyodeAwe) - Bradley Dice (https://github.com/bdice) URL: #6609
Ofek-Haim
pushed a commit
to Ofek-Haim/cuml
that referenced
this pull request
May 13, 2025
…sai#6553) (rapidsai#6555) This change modifies the KMeans test to set n_init to 2, enhancing the stability of the k-means|| initialization process. Additional context can be found in issue rapidsai#5530. Closes rapidsai#5530. Authors: - Simon Adorf (https://github.com/csadorf) Approvers: - Divye Gala (https://github.com/divyegala) - Jim Crist-Harif (https://github.com/jcrist) URL: rapidsai#6555
Ofek-Haim
pushed a commit
to Ofek-Haim/cuml
that referenced
this pull request
May 13, 2025
This PR enhances the infrastructure for monitoring scikit-learn parity in cuML's accelerated estimators, enabling better tracking and management of differences between scikit-learn and cuML implementations. ## Key Changes - Moved scikit-learn test tooling from `ci/` to `python/cuml/cuml/accel/tests/scikit-learn/` for better organization - Added `xfail-list.yaml` to track known differences from scikit-learn - Added tooling to _summarize-results_ script to maintain xfail-list - Added parallel test execution support via pytest-xdist for faster test runs ## Infrastructure Capabilities The new infrastructure enables continuous monitoring of scikit-learn parity by: - Running scikit-learn tests with cuML acceleration - Tracking pass rates and comparing against thresholds - Automatically detecting new failures (regressions) - Managing known differences via xfail lists ## Related Changes - Closes rapidsai#6570 Authors: - Simon Adorf (https://github.com/csadorf) Approvers: - Jim Crist-Harif (https://github.com/jcrist) - James Lamb (https://github.com/jameslamb) URL: rapidsai#6553
Ofek-Haim
pushed a commit
to Ofek-Haim/cuml
that referenced
this pull request
May 13, 2025
As a mitigation for rapidsaigh-6597. Partial revert of rapidsaigh-6553. Authors: - Simon Adorf (https://github.com/csadorf) Approvers: - Jim Crist-Harif (https://github.com/jcrist) - Jake Awe (https://github.com/AyodeAwe) - Bradley Dice (https://github.com/bdice) URL: rapidsai#6609
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR enhances the infrastructure for monitoring scikit-learn parity in cuML's accelerated estimators, enabling better tracking and management of differences between scikit-learn and cuML implementations.
Key Changes
ci/topython/cuml/cuml/accel/tests/scikit-learn/for better organizationxfail-list.yamlto track known differences from scikit-learnInfrastructure Capabilities
The new infrastructure enables continuous monitoring of scikit-learn parity by:
Related Changes