FLAML Files

This release adds Python 3.13 support, fixes several bugs in search, configuration handling, and estimator preprocessing, and improves overall robustness. It also includes major documentation updates covering metrics, estimators, best practices, and hierarchical search spaces. Most of the work was contributed by @Copilot.

What's Changed

• Fix Best Practices not shown by @thinkall in https://github.com/microsoft/FLAML/pull/1483
• Fix isinstance usage issues by @thinkall in https://github.com/microsoft/FLAML/pull/1488
• Add support to python 3.13 by @thinkall in https://github.com/microsoft/FLAML/pull/1486
• Add objective parameter to LGBMEstimator search space by @Copilot in https://github.com/microsoft/FLAML/pull/1474
• Fix BlendSearch OptunaSearch warning for non-hierarchical spaces with Ray Tune domains by @Copilot in https://github.com/microsoft/FLAML/pull/1477
• Fix: Preserve FLAML_sample_size in best_config_per_estimator by @Copilot in https://github.com/microsoft/FLAML/pull/1475
• Bump version to 2.5.0 by @thinkall in https://github.com/microsoft/FLAML/pull/1492
• Add comprehensive metric documentation and URL reference to AutoML docstrings by @Copilot in https://github.com/microsoft/FLAML/pull/1471
• Expand docs to include all flamlized estimators by @Copilot in https://github.com/microsoft/FLAML/pull/1472
• Fix indents in Task-Oriented-AutoML.md by @thinkall in https://github.com/microsoft/FLAML/pull/1493
• Fix nested dictionary merge in SearchThread losing sampled hyperparameters by @Copilot in https://github.com/microsoft/FLAML/pull/1494
• Add GitHub Copilot instructions for FLAML repository by @Copilot in https://github.com/microsoft/FLAML/pull/1502
• Fix eval_set preprocessing for XGBoost estimators with categorical features by @Copilot in https://github.com/microsoft/FLAML/pull/1470
• Document missing value handling behavior for AutoML estimators by @Copilot in https://github.com/microsoft/FLAML/pull/1473
• Document that final_estimator parameters in ensemble are not auto-tuned by @Copilot in https://github.com/microsoft/FLAML/pull/1499
• Add validation and clear error messages for custom_metric parameter by @Copilot in https://github.com/microsoft/FLAML/pull/1500
• Use scientific notation for best error in logger output by @Copilot in https://github.com/microsoft/FLAML/pull/1498
• Add configurable label overlap handling for classification holdout strategy by @Copilot in https://github.com/microsoft/FLAML/pull/1491
• Clarify period parameter and automatic label lagging in time series forecasting by @Copilot in https://github.com/microsoft/FLAML/pull/1495
• Expose task-level and estimator-level preprocessors as public API by @Copilot in https://github.com/microsoft/FLAML/pull/1497
• Add hierarchical search space documentation by @Copilot in https://github.com/microsoft/FLAML/pull/1496
• Update Copilot Instructions by @thinkall in https://github.com/microsoft/FLAML/pull/1505

Full Changelog: https://github.com/microsoft/FLAML/compare/v2.4.1...v2.5.0