RFTSystems's picture
Update README.md
0bb1a6c verified
|
Raw
History Blame
2.16 kB
metadata
title: Coherent Compute Engine
emoji: 🌖
colorFrom: pink
colorTo: red
sdk: gradio
sdk_version: 6.2.0
app_file: app.py
pinned: false
license: other
short_description: Live coherence + throughput benchmark (no precomputed result
thumbnail: >-
  https://cdn-uploads.huggingface.co/production/uploads/685edcb04796127b024b4805/Sr9bUuYsRuDtVZnb-

Coherent_Compute_Engine

A real-time, on-machine benchmark that measures coherent state-update throughput and produces a tamper-evident receipt for every run.

What it does

  • Runs a coherent update kernel over N oscillators for S steps
  • Measures:
    • Throughput (B/s = billions of items per second)
    • Coherence (|C| stability proxy)
    • Mean Energy (bounded mean of E)
    • Elapsed time
  • Produces a downloadable receipt (JSON) containing inputs, environment, outputs, and a SHA-256 hash.

What an “item” is

1 item = one per-oscillator coherent state update of [Psi, E, L] per step, as implemented in this Space.

This definition is explicit so the metric is reproducible.

Verification / receipts

Each run writes a receipt JSON under results/ and exposes it via the UI. The receipt includes:

  • timestamp (UTC)
  • app version
  • platform + python
  • logical CPU count
  • inputs (N, steps, seed, scale)
  • measured outputs
  • SHA-256 hash of the canonical receipt JSON

This makes runs easy to share and tamper-evident.

Baselines (optional)

If enabled, the Space also measures:

  • live NumPy baseline (vectorised)
  • tiny Python loop baseline (hard safety caps)

If numba is available, the primary engine automatically uses a Numba JIT kernel.

Important note about “your CPU”

Hugging Face Spaces run on hosted hardware. The results reflect the Space runtime, not the visitor’s laptop CPU.

If you want your own machine’s numbers, run the repo locally.

Files

  • app.py — Gradio app + benchmark + receipt writer
  • results/ — receipts + runs.csv (created automatically)

Licence

Please read LICENCE, listed in repo

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference