Commit Graph

1 Commits (cd6efc92063047c5ba7db1752fa2d83ac61dae3a)

Author SHA1 Message Date
Hwe1916 4d960838e4 feat(eval): add retrieval and benchmark evaluation
Measures retrieval hit rate and answer quality against a held-out benchmark, plus sanity checks for interface drift.
2026-10-24 00:45:27 +08:00