{% extends "base.html" %} {% block title %}Stress test · mneme{% endblock %} {% block content %}

Stress test

Run all {{ corpus_size }} seed messages through the classifier in a deterministic shuffled order. Watch the cumulative hit rate climb from 0% on a cold cache to over 70% as paraphrases start landing on entries the LLM already classified.

LLM endpoint: {{ spark_url }} · model: {{ model }}

0 / {{ corpus_size }} · hit rate: 0% · idle

Cumulative cache hit rate

Live tail

#LayerLatencyQuery→ Intent
{% endblock %} {% block scripts %} {% endblock %}