Skip to content

Run each memory model's tests in its own CI job - #79

Merged
ASDAlexander77 merged 1 commit into
mainfrom
ci-tests-per-model
Nov 9, 2026
Merged

ASDAlexander77 merged 1 commit into
mainfrom
ci-tests-per-model

Conversation

@ASDAlexander77

Copy link
Copy Markdown
Owner

Summary

  • test_windows and test_linux in tests.yml each become a matrix over gc, rc, none, own (fail-fast: false), so the 4 models run as 8 parallel jobs instead of one after another.
  • Each job downloads the compiler, builds the library (1–2 min) and runs only its model; it uploads its result as results-<os>-<model>. The badges job already merges results-*, so it is unchanged.
  • Per-job timeout lowered from 180 to 120 min.

Before: Windows ~21 min (19 min of tests for the 4 models), Ubuntu ~14 min. Expected after: about 8 min, bounded by the slowest model.

Test plan

  • The 8 Test (<os>, -mm=<model>) jobs run on this PR and pass
  • After merge, the push run on main publishes all 8 badges

🤖 Generated with Claude Code

The Windows and Ubuntu test jobs each become a matrix over gc, rc, none
and own, so the models run side by side instead of one after another.
Each job builds the library itself (a minute or two) and uploads its
result as results-<os>-<model>, which the badges job already collects.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@ASDAlexander77
ASDAlexander77 merged commit 906be39 into main Nov 9, 2026
10 checks passed
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.

1 participant