New features:

• New MLAlgos: TabM and TabICL by @dev-rinchin and @Uptimolli in c31d660 and 5a79d05
• PiecewiseLinearEmbeddings by @dev-rinchin in c31d660
• "Auto select device at inference" by @Uptimolli in cc808f7

Dependencies:

• Numpy 2.0+ and Pandas 2.0+ support by @Uptimolli and @dev-rinchin in b9f8ca1
• Replacing fasttext with fasttext-numpy2 by @Uptimolli in f48c76a

Bugfixes:

• Fix multiclass in XGBoost by @Uptimolli in 1ca9074 and 01f5fda
• Fix #173 by @dev-rinchin in bf0d518
• Fix groupby for new category by @dev-rinchin in 208ee8b
• Fix forced_features calculation by @Uptimolli in e95496c
• Fix Optuna Exception brackets bug by @dwemer8 in #174

Full Changelog: v.0.4.1...v.0.4.2

What's Changed

• Fix bug with rearranging columns and NumericRole in AutoUplift by @dev-rinchin f6ed576
• Fix the XGBoost optimization direction and custom metric passing cf2d73e
• Update blend.py by @BuldakovN in #171

Full Changelog: v.0.4.0...v.0.4.1

What's Changed

• Support for Pandas v2+ and lightgbm v4+ by @Uptimolli in a8c4e90
• Change supporting python versions: [3.8 - 3.12] and torch version by @dev-rinchin in #161
• Add XGBoost as MLAlgo by @dev-rinchin in #160
• Permutation importance bug fixed by @BELONOVSKII in #139
• Add path_to_save parameter to fit_predict by @Uptimolli in baf5be4
• Merge OptunaTuner and DLOptunaTuner by @dev-rinchin in fe4ab62
• Add fail_tolerance for OptunaTuner by @dev-rinchin in 882a9e5
• Auto adjustment CV parameter for multiclass by @Uptimolli in 526aebe
• Fix force_input using in selectors by @dev-rinchin in b5a1bba
• Add FillnaMean in torch_pipeline by @Uptimolli in 9dc9aa4
• HypEx addon moved to a separate library by @tikhomirovd in #170
• Built-in class mapping by @screengreen in #164
• Added progress bar for optuna by @screengreen in #165

Full Changelog: v0.3.8.1...v0.4.0

What's Changed

Main changes:

• Add neural networks and bug fixes by @MikhailKuz in #79

• Rel tables by @dev-rinchin in #39

• Feature/support python 3.10 by @dev-rinchin in #101

• Fix features forcing by @dev-rinchin in #63

• Add upper bound for scikit-learn version by @elineii in #73

• ts modified seqlagtransform + add difftransform by @elineii in #72

• Ts features pipeline by @elineii in #81

• time_series config and tutorial + bug fixes by @elineii in #94

• Add addon HypEx (Hypotheses and Experiments) by @dev-rinchin in #104

• Feature/groupby2 by @VaBun in #58

• Add test for presets by @dev-rinchin in #70

• CI/CD by @Cybsloth in #46

• Add pre-push rule by @Cybsloth in #51

• Feature/docs by @Cybsloth in #52

• Update docs.yml by @Cybsloth in #53

• Bugfix/docs by @Cybsloth in #54

• Bugfix/poetry deps by @Cybsloth in #55

Full Changelog: v0.3.7.3...v0.3.8