Skip to content

Document MS MARCO 1M SQLite reuse result#106

Merged
SonAIengine merged 1 commit into
mainfrom
codex/msmarco-1m-reuse-result
Jul 2, 2026
Merged

Document MS MARCO 1M SQLite reuse result#106
SonAIengine merged 1 commit into
mainfrom
codex/msmarco-1m-reuse-result

Conversation

@SonAIengine

Copy link
Copy Markdown
Contributor

Summary

  • document the local persistent MS MARCO 1M SQLite DB build
  • record the reuse-mode rerun that skips the ingest/index phase
  • note the gitignored DB and sidecar artifact sizes

Results

  • persistent build: 1,000,000 docs, 50 queries, MRR@10 0.462, Hit@10 30/50, build 2184.3s, search 71.0s
  • persistent reuse: 1,000,000 docs, 50 queries, MRR@10 0.462, Hit@10 30/50, build 0.0s, search 70.1s
  • local artifacts: tests/benchmark/data/msmarco_1m.db (1.2 GB), tests/benchmark/data/msmarco_1m.db.tier1.json (535 bytes)

Checks

  • git diff --check
  • inspected updated markdown section

No code changes in this PR.

@SonAIengine SonAIengine merged commit 760164d into main Jul 2, 2026
2 checks passed
@SonAIengine SonAIengine deleted the codex/msmarco-1m-reuse-result branch July 2, 2026 03:56
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