| .. | ||
| multi_client_optimization_loop.md | ||
| performance_report.md | ||
| README.md | ||
| retrieval_precision_optimization.md | ||
| storage_backends.md | ||
Benchmarks
This section centralizes public benchmark guidance and latest reported results.
Available benchmark docs
- Performance Report - canonical performance tables and release baselines.
- Storage Backends (Local vs S3-compatible) - benchmark runner for local storage compared with S3-compatible providers, including multi-client local-vs-R2 tracks.
Validation status policy
- Public benchmark docs distinguish between:
- tested paths validated in this repository, and
- supported paths that are available but not currently part of automated validation.
When in doubt, treat tested paths as production-ready defaults.
Public vs internal benchmark assets
- Public docs in this section contain scope, status, and published results.
- Internal harness scripts and run plumbing are intentionally not shipped in public docs.