Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 28 additions & 1 deletion CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,32 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0

## [Unreleased]

## [0.13.0] - 2026-11-06

### Added

- `ForcedChoiceModelConfig.scoring_mode` selects `joint` classification (the
existing default) or `independent` scoring. Independent scoring requires
`dual_encoder` and applies one shared scalar head to each option before
softmax, giving a Bradley-Terry choice rule for two alternatives.
- Independent scoring supports participant-specific position intercepts and
scalar scoring heads, retains unselected alternatives in the choice set, and
validates consistent option names across training and prediction.
- The argument-structure acceptability training script reads `encoder_mode` and
`scoring_mode` from its configuration. Checkpoints persist the scoring mode;
checkpoints without the field retain joint classification.

### Fixed

- Forced-choice dual encoders train through the same per-option encoding path
used for prediction, avoiding the single-embedding trainer's dimension mismatch.
- The forced-choice custom training loop creates participant parameters before
constructing its optimizer and restores the best encoder, head, and participant
effects when early stopping on held-out cross-entropy.
- Random-effect regularization accumulates penalties on the parameter device.
- Forced-choice metrics use installed scikit-learn functions without downloading
metric scripts or writing a Hugging Face metric cache.

## [0.12.0] - 2026-11-03

### Added
Expand Down Expand Up @@ -766,7 +792,8 @@ guards as type-checkers.
- CI/CD: GitHub Actions for testing, docs, PyPI publishing
- Read the Docs integration

[Unreleased]: https://github.com/FACTSlab/bead/compare/v0.12.0...HEAD
[Unreleased]: https://github.com/FACTSlab/bead/compare/v0.13.0...HEAD
[0.13.0]: https://github.com/FACTSlab/bead/compare/v0.12.0...v0.13.0
[0.12.0]: https://github.com/FACTSlab/bead/compare/v0.11.0...v0.12.0
[0.11.0]: https://github.com/FACTSlab/bead/compare/v0.10.0...v0.11.0
[0.10.0]: https://github.com/FACTSlab/bead/compare/v0.9.0...v0.10.0
Expand Down
Loading
Loading