knowledge-rag

Live Performance Dashboard

Public transparency, updated on every push to master. Median wall-time per benchmark, grouped by category. No cherry-picked numbers — the CI runs pytest bench/ --benchmark-only on a clean Ubuntu runner and publishes the raw JSON alongside this page.

Last update Commit Version Benchmarks

Executive summary

One-line health per category. Green = within budget. Amber = watch. Red = investigate.

Loading…

Median latency by benchmark

Horizontal bars, log-scale, shortest first. Hover to see raw ops/sec.

Category breakdown

All benchmarks grouped by subsystem. Click a category to jump to the full detail table.

Loading…

Full benchmark table

Sortable, searchable. All the raw numbers behind the summary above.

Category Benchmark Median Ops/sec σ (%)
Loading benchmark data…

Methodology & disclaimers

How the numbers are produced

  • Runner: GitHub-hosted ubuntu-latest (2 vCPU, 7 GB RAM). Not a benchmark rig — treat as a floor, not a ceiling.
  • Command: pytest bench/ --benchmark-only --benchmark-min-rounds=5 against a clean checkout.
  • Data: median wall-time (nanoseconds), ops/sec, standard deviation. Warm-up rounds discarded.
  • Cadence: every push to master. The bench-pages.yml workflow rebuilds this page automatically.

How to read the numbers

  • Median < standard deviation means an unstable measurement — do not draw conclusions from a single run.
  • Log-scale bars in the chart above compress the range so you can see µs and ms benchmarks side by side.
  • Category KPIs pick the representative benchmark for that subsystem. See bench/ for the source.
  • Compared to your production hardware, these numbers will typically be 2–5× faster.