LearnNewsExamplesServices
Frontmatter
id9755
titleTest: Correct embeddingModel to qwen3-embedding in SessionSummarization
stateClosed
labels
aitesting
assigneestobiu
createdAtApr 7, 2026, 3:11 PM
updatedAtApr 7, 2026, 3:11 PM
githubUrlhttps://github.com/neomjs/neo/issues/9755
authortobiu
commentsCount0
parentIssuenull
subIssues[]
subIssuesCompleted0
subIssuesTotal0
blockedBy[]
blocking[]
closedAtApr 7, 2026, 3:11 PM

Test: Correct embeddingModel to qwen3-embedding in SessionSummarization

Closedaitesting
tobiu
tobiu commented on Apr 7, 2026, 3:11 PM

Description

The SessionSummarization.spec.mjs Playwright test incorrectly overrides SDK.Memory_Config.data.ollama.embeddingModel to use gemma4. Since gemma4 is the core inferencer (LLM) and not an embedding model, this is an architectural drift compared to the local native config (qwen3-embedding). Tracking the fix to correct the test configuration.

tobiu added the ai label on Apr 7, 2026, 3:11 PM
tobiu added the testing label on Apr 7, 2026, 3:11 PM
tobiu referenced in commit f3dc18c - "test: fix embedding mock to use native qwen3-embedding config (#9755)" on Apr 7, 2026, 3:11 PM
tobiu assigned to @tobiu on Apr 7, 2026, 3:11 PM
tobiu closed this issue on Apr 7, 2026, 3:11 PM