Skip to content

InferRSI v0.2.1

Pre-release
Pre-release

Choose a tag to compare

View all tags
@github-actions github-actions released this 14 Nov 21:56
· 8 commits to main since this release
v0.2.1
8c32100
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Source: https://github.com/InferMatrixLab/InferRSI/tree/8c321005d50c2ecef61bf20132409d88474a6450

These notes describe software changes. Release builds and scaffold checks do not establish campaign PASS, model quality or service performance. Any such claim requires its acceptance issue and complete frozen evidence.

AI-generated summary

Features

  • The skill evaluation four-file module family is now split by function owner, moving benchmark-related code into src/inferrsi/benchmark and introducing separate improvement and engineering modules for clearer separation of concerns 8c32100.
  • A new artifact sandbox and atomic JSON helper are added to the harness for safer artifact handling 8c32100.
  • The README now includes a release news section for readers 37220b1.

Fixes

  • Skill evaluation, improvement and cloud replay tooling are updated to align with the new module layout and function-owner boundaries 8c32100.

Compatibility

  • Documentation contracts (current design, function owners, module layout, gate policy, isolation errors and CLI) are revised to reflect the reorganized modules 8c32100.
  • Local, cloud and comparison skill evaluation docs are updated for the new structure, and import-gate tests are adjusted accordingly 8c32100.

What's Changed

Other changes

  • docs: add important release news to README by @hsliuustc0106 in #445
  • skill-eval: split the four-file family by function owner and drop the residual (S01) by @Dong1017 in #437

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

Contributors

Dong1017 and hsliuustc0106
Assets 7
Loading