v0.6.0

Added

• Add the EEGPT backbone (#40).

Full Changelog: v0.5.0...v0.6.0

v0.5.0

Added

• Add a required revision field on Dataset (forwarded to BaseConcatDataset.pull_from_hub) and pin every default dataset config to its current HuggingFace commit SHA for reproducibility (#37).
• Add training_required_parameters field on _BackboneBase for top-level nn.Parameter names that must remain trainable (e.g. Labram's temporal_embedding). Distinct from training_required_modules, which only accepts nn.Module names. Finetuning methods declare compatibility via the supports_training_required_parameters class variable; IA3 and OFT opt out. LoRA/AdaLoRA/DoRA forward the list to PEFT's target_parameters; Frozen/TwoStages extend their unfreeze list (#25).

Changed

• Allow custom normalization methods by subclassing Normalization, now based on exca.helpers.DiscriminatedModel. Builtin subclasses pin their pre-existing kind value to preserve cached experiment UIDs (#35).

Full Changelog: v0.4.0...v0.5.0

v0.4.0

Added

• Add ScratchBackbone for benchmarking DL models without pretrained weights. Only compatible with FullFinetune (enforced by the Experiment validator) (#30).
• Add class_weight parameter to RidgeProbingTraining ("balanced" or None); default changed to "balanced" — pass None for the previous unweighted behavior (#32).
• Add dtype parameter to RidgeProbingTraining ("float32" or "float64", default "float64"). Use "float32" only when necessary, e.g. on Apple MPS which does not support float64 (#32).

Changed

• Fill in the Zenodo concept DOI (10.5281/zenodo.19698863) in the README DOI badge and the BibTeX snippet, and add it as an identifiers entry in CITATION.cff.

Full Changelog: v0.3.0...v0.4.0

v0.3.0

Added

• Add "ridge_probe" finetuning strategy: closed-form streaming ridge regression probing on frozen backbone features. Single pass over the dataloader, eigendecomposition-based λ sweep, no hyperparameter tuning needed. For high-dimensional backbones, max_features (default 5000) triggers a Gaussian random projection seeded by Experiment.seed; multi-seed runs therefore produce different projections and enable variance estimation (#21).
• Add FlattenHead head type (used internally by ridge probing) (#21).
• Add RidgeProbingTraining config with kind="ridge" discriminator for the Training union (#21).
• Add pytest to the CI workflow to run tests on each pull request (#18).
• Add max_meta_experiments argument to helpers.run_multiple_per_node() as alternative to max_experiments_per_node (#17).
• Add training_required_modules parameter to the backbones (#20).
• Expose the preload parameter of datasets (#26).
• Add CITATION.cff metadata file and a Zenodo DOI badge in the README for archival citation via Zenodo (#27).

Changed

• Popularize the use of import open_eeg_bench as oeb via the README and documentation (#17).
• Improve import hints via the __all__ variable in __init__.py (#17).
• All backbones now use peft_target_modules="all-linear" by default for simplicity, which leads to a slight increase in the number of parameters being finetuned (#20).
• Change default LoRA.bias to "none" to match PEFT's default (#20).
• Allow disabling dropout layers of the backbone (default: True) (#20).
• Change default Experiment.seed from 42 to 0 (#21).

Fixed

• The lazy modules are now initialized before and after applying the PEFT (necessary for some PEFT methods like OFT) (#17).
• Take into account BENDR's channels projection fix in the BENDR backbone (#20).

Full Changelog: v0.2.1...v0.3.0

v0.2.1

Added

• PyPI metadata: project URLs and classifiers (#14).
• README badges: PyPI version, supported Python versions, tests status, HuggingFace Space (#14).
• New step in publish.yml to attach build artifacts (sdist + wheels) to the GitHub release (#14).

Full Changelog: v0.2.0...v0.2.1

v0.2.0

Added

• Benchopt interface for running benchmarks via the benchopt_wrappers directory (#8).
• Experiment wrapper allowing multiple experiments to run on the same node via Joblib (#10).
• NoNormalization transform to simplify serialization (#11).
• Option for collect_results to wait for pending jobs (#11).
• Experimental warnings on helper functions (#10).
• Skip experiments that are already completed (#10).
• Optional sorting in result collection (#10).

Changed

• Bumped braindecode dependency to 1.4.0 (#12).
• Set dataset preload=False by default (#11).
• Improved error parsing and infra model_dump output (#10).
• Documentation now uses the pip install command for installation instructions.

Fixed

• Corrected EEG montages handling (#9).
• Removed infra from _exclude_from_cls_uid so jobs correctly honor the retry and force modes (#10).

Full Changelog: v0.1.0...v0.2.0

v0.1.0

Initial release