Making tabebm compatible with newer versions

[mumair5393]

Fix TabEBM compatibility with recent TabPFN API changes

get_preprocessed_dataset_chunks signature changed in recent TabPFN — splitter param renamed to split_fn and several required args added (max_data_size, equal_split_size, data_shuffle_seed, preprocessing_random_state). Also fixes the broken test.

Fixes #225

[CLAassistant]

All committers have signed the CLA.

[gemini-code-assist]

Code Review

This pull request updates imports and configurations to align with the new TabPFN API, including replacing ModelInterfaceConfig with InferenceConfig and using get_preprocessed_dataset_chunks. It also fixes an issue in train_test_split_allow_full_train where stratify was used without shuffling. The review feedback points out that max_data_size is hardcoded to 1000 in get_preprocessed_dataset_chunks and should instead use self.max_data_size to respect the user-configured maximum data size.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[klemens-floege]

@mumair5393 thank you for opening a PR into our repo. Plesae make sure all tests are passing then I will give it a review

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

Want higher recall? High effort reviews run extra passes and find more bugs. A team admin can switch effort levels in the Cursor dashboard.

^{Reviewed by Cursor Bugbot for commit 9d7deba. Configure here.}

[cursor]

Internal architectures import path

Medium Severity

PerformanceOptions is imported from tabpfn.architectures.interface, an internal architectures path. Project guidance prefers stable public imports such as from tabpfn.classifier / tabpfn.regressor (or avoiding the symbol if a default already works), because architecture internals have moved across TabPFN releases.

 

^{Triggered by learned rule: Import tabpfn classes from stable public API paths, not internal modules}

^{Reviewed by Cursor Bugbot for commit 9d7deba. Configure here.}

[klemens-floege]

@mumair5393 thank you for opening a PR into our repo. Plesae make sure all tests are passing then I will give it a review