Skip to content

Adds structural serving capabilities for external predictors (0.11.0) - #15

Merged
aaronstevenwhite merged 2 commits into
mainfrom
feat/structural-serving
Oct 25, 2026
Merged

aaronstevenwhite merged 2 commits into
mainfrom
feat/structural-serving

Conversation

@aaronstevenwhite

@aaronstevenwhite aaronstevenwhite commented Oct 25, 2026 •

Copy link
Copy Markdown
Collaborator

Description

Adds consumer-shaped runtime protocols for predictor identity, task compatibility, probabilistic prediction, native-artifact persistence, and three-way NLI. Adds a generic structural factory registry and a published conformance suite over real bead items, templates, and response encodings. Documents the new public API and releases these capabilities as bead 0.11.0.

Motivation

External packages need to serve trained predictors without inheriting bead's trainable mixed-effects hierarchy or implementing unrelated model-adapter methods. This change lets consumers request only the capabilities they use and lets third-party packages verify interoperability against bead's public contracts.

Fixes #14

Type of Change

  • Bug fix (non-breaking change that fixes an issue)
  • New feature (non-breaking change that adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to change)
  • Documentation update
  • Refactoring (no functional changes)
  • Tests (adding or updating tests)

Checklist

  • I have read the CONTRIBUTING guidelines
  • My code follows the project's style guidelines
  • I have run uv run ruff check . and uv run ruff format .
  • I have run uv run pyright with no errors
  • I have added tests that prove my fix/feature works
  • All tests pass (uv run pytest tests/)
  • I have updated documentation as needed

Testing

  • Python CI suite: 3,593 passed, 57 skipped, 257 deselected
  • Ruff lint and format checks
  • Pyright: 0 errors
  • TypeScript check, typecheck, build, gallery build, and 83 tests
  • MkDocs build
  • Version consistency tests for 0.11.0
  • Clean-tree 0.11.0 source distribution and wheel build with inspected package metadata

Screenshots (if applicable)

Not applicable.

@aaronstevenwhite aaronstevenwhite changed the title Adds structural serving capabilities for external predictors Adds structural serving capabilities for external predictors (0.11.0) Oct 25, 2026
@aaronstevenwhite
aaronstevenwhite merged commit ed9486b into main Oct 25, 2026
8 checks passed
@aaronstevenwhite
aaronstevenwhite deleted the feat/structural-serving branch October 25, 2026 19:50
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Expose structural serving capabilities for externally trained predictors

1 participant