Skip to content

CompileML v0.1.0 — Initial Release

Choose a tag to compare

View all tags
@orgoca orgoca released this 16 Aug 16:39
· 8 commits to main since this release
v0.1.0
76e237a
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

CompileML’s first public release.

CompileML turns trained tree-ensemble models into deterministic, auditable decision artifacts. The resulting artifact keeps the score, calibration, risk bands, reason codes, and per-feature integer impacts together in one portable JSON representation.

Highlights:

  • Integer-exact decision waterfalls whose impacts reconcile to the final score
  • Support for compiling tree-ensemble models into portable artifacts
  • Driver-fidelity validation for reason codes and feature impacts
  • Built-in calibration and risk-band handling
  • Structural verification and artifact hashing
  • Eight-check validation framework
  • CLI commands for compiling, inspecting, verifying, scoring, validating, and exporting artifacts
  • Dependency-free Python runtime for executing compiled artifacts

Install it with:

pip install compileml

This is the initial alpha release. Feedback, bug reports, and contributions are very welcome.

Assets 2
Loading