Skip to content

FIX(ci-test-router): recover semantic routing after transient embed failure - #2871

Open
fl0rianr wants to merge 4 commits into
mainfrom
fl0rianr/ci_router_test_fix_v2
Open

FIX(ci-test-router): recover semantic routing after transient embed failure#2871
fl0rianr wants to merge 4 commits into
mainfrom
fl0rianr/ci_router_test_fix_v2

Conversation

@fl0rianr

@fl0rianr fl0rianr commented Aug 1, 2026

Copy link
Copy Markdown
Collaborator

Summary

Stabilizes the semantic router tests by making the embedding-model setup deterministic.

  • Loads nomic-embed-text-v1-GGUF with ctx_size=2048 on the CPU backend.
  • Verifies readiness with a real /embeddings request before testing routing.
  • Avoids persisting CI-specific load options.
  • Unloads the embedding model during cleanup.
  • Keeps routing failures visible instead of masking them with broad retries.

Validation

The updated router-onnxruntime CI job successfully loads the embedding model, completes embedding inference, and runs the semantic routing tests.

@github-actions github-actions Bot added area::ci CI / GitHub Actions / self-hosted runner infrastructure bug Something isn't working labels Aug 1, 2026
@fl0rianr
fl0rianr requested a review from ramkrishna2910 August 2, 2026 13:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

area::ci CI / GitHub Actions / self-hosted runner infrastructure bug Something isn't working

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant