New eval/results/baseline-20260610.md: 22 queries against the prod image index rebuilt today on the 4-endpoint GPU pool with the resilient embedder (PR #8). vs the 2026-05-22 baseline: dense MRR 0.539→0.924, bm25+rerank 0.920→0.959, hybrid_rrf+rerank 0.875→0.960. Reranker legs ran against the prod jina-rerank. Provenance header in the file.
New `eval/results/baseline-20260610.md`: 22 queries against the prod image index rebuilt today on the 4-endpoint GPU pool with the resilient embedder (PR #8). vs the 2026-05-22 baseline: dense MRR 0.539→0.924, bm25+rerank 0.920→0.959, hybrid_rrf+rerank 0.875→0.960. Reranker legs ran against the prod jina-rerank. Provenance header in the file.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
22 queries against the prod image index rebuilt today on the expanded
GPU pool with the resilient embedder (PR #8): dense MRR 0.539→0.924,
bm25+rerank 0.920→0.959, hybrid_rrf+rerank 0.875→0.960 vs the
2026-05-22 baseline. No regression from mixed-provenance embeddings.
Co-Authored-By: Claude Fable 5 <[email protected]>
justin
merged commit 992d66e3d1 into main2026-06-10 20:38:25 -04:00
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
New
eval/results/baseline-20260610.md: 22 queries against the prod image index rebuilt today on the 4-endpoint GPU pool with the resilient embedder (PR #8). vs the 2026-05-22 baseline: dense MRR 0.539→0.924, bm25+rerank 0.920→0.959, hybrid_rrf+rerank 0.875→0.960. Reranker legs ran against the prod jina-rerank. Provenance header in the file.🤖 Generated with Claude Code