Skip to content

Conversation

@tonyhoo
Copy link
Contributor

@tonyhoo tonyhoo commented Apr 21, 2025

Issue #, if available:
catboost/catboost#2671

Description of changes:
With the release of catboost 1.2.8 and resolution of the issue, update numpy support on autogluon side

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

@tonyhoo tonyhoo added this to the 1.3 Release milestone Apr 21, 2025
@tonyhoo
Copy link
Contributor Author

tonyhoo commented Apr 21, 2025

/platform_tests b76f8cf
Platform Tests Output

@tonyhoo tonyhoo closed this Apr 21, 2025
@tonyhoo tonyhoo reopened this Apr 21, 2025
@tonyhoo tonyhoo requested a review from suzhoum April 22, 2025 07:09
@tonyhoo
Copy link
Contributor Author

tonyhoo commented Apr 22, 2025

/benchmark b76f8cf module=tabular preset=tabular_best benchmark=tabular_full time_limit=8h

Benchmark Output

Benchmark Test Result - Fail Evaluation Results Path: s3://autogluon-ci-benchmark/evaluation/tabular/numpy
The dashboard website is: http://autogluon-staging.s3-website-us-west-2.amazonaws.com/benchmark-dashboard/numpy/28a242ebe8d55ba770c991b9db153ab4623c9abd/index.html

@tonyhoo
Copy link
Contributor Author

tonyhoo commented Apr 23, 2025

Noticed below list of errors from the benchmark run
results_automlbenchmark_8h_ag_bench_20250422T071618.csv

@suzhoum
Copy link
Contributor

suzhoum commented Apr 24, 2025

Noticed below list of errors from the benchmark run results_automlbenchmark_8h_ag_bench_20250422T071618.csv

There were also many failures in the benchmark but not enough error messages presented in the logs. I'll probably run a few benchmarks locally to see if I can find anything.

@suzhoum
Copy link
Contributor

suzhoum commented Apr 24, 2025

/benchmark b76f8cf module=tabular preset=tabular_best benchmark=tabular_small time_limit=1h
Benchmark Output

Benchmark Test Result - Fail Evaluation Results Path: s3://autogluon-ci-benchmark/evaluation/tabular/numpy
The dashboard website is: http://autogluon-staging.s3-website-us-west-2.amazonaws.com/benchmark-dashboard/numpy/086da85bf166228734d9687f82b4d06ea4b584f4/index.html

@suzhoum
Copy link
Contributor

suzhoum commented Apr 25, 2025

Noticed below list of errors from the benchmark run results_automlbenchmark_8h_ag_bench_20250422T071618.csv

There were also many failures in the benchmark but not enough error messages presented in the logs. I'll probably run a few benchmarks locally to see if I can find anything.

After some digging, I think the root cause is due to inconsistent numpy versions used in automlbenchmark evaluation and virtual env created for AutoGluon feature branch for training. I created a PR Innixma/automlbenchmark#12 to upgrade numpy in our custom automlbenchmark branch.

@suzhoum
Copy link
Contributor

suzhoum commented Apr 25, 2025

/benchmark b76f8cf module=tabular preset=tabular_best benchmark=tabular_full time_limit=1h
Benchmark Output

@suzhoum
Copy link
Contributor

suzhoum commented Apr 26, 2025

/benchmark b76f8cf module=tabular preset=tabular_best benchmark=tabular_full time_limit=1h
Benchmark Output

Benchmark Test Result - Pass Evaluation Results Path: s3://autogluon-ci-benchmark/evaluation/tabular/numpy
The dashboard website is: http://autogluon-staging.s3-website-us-west-2.amazonaws.com/benchmark-dashboard/numpy/b913ac86a773bbdb61b8a0573b14d664e95ec32d/index.html

@tonyhoo
Copy link
Contributor Author

tonyhoo commented Apr 28, 2025

/platform_tests 5f8bd20
Platform Tests Output

Copy link
Contributor

@suzhoum suzhoum left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM!

@tonyhoo tonyhoo merged commit 0b09703 into autogluon:master Apr 29, 2025
4 checks passed
@tonyhoo tonyhoo deleted the numpy branch April 29, 2025 16:14
@Innixma
Copy link
Contributor

Innixma commented Apr 29, 2025

@tonyhoo @suzhoum Why does this PR revert the fix I made in #5045 to support sklearn<1.6?

FireballDWF pushed a commit to FireballDWF/autogluon that referenced this pull request May 6, 2025
FireballDWF pushed a commit to FireballDWF/autogluon that referenced this pull request May 21, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants