Strot releases open-source decision compiler that runs local models per record
Overview
Strot releases strot-ai, an MIT-licensed Python library that uses an LLM once at compile time to build a frozen artifact.
A local decision model then routes or classifies each record without further LLM calls, according to a Show HN post on Hacker News. The post reports 92.5% grounded routing accuracy on 398 records using the 144M-parameter Julia-1 model, against 74.8% for a naive baseline, at about 60 ms per record. The library installs with pip install strot-ai. These accuracy and speed figures are the author's own reported results on that test set.
Written by AI from the articles below · updated Oct 19, 2:22 AM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- Hacker News · Show HN, AI (20+ points)BlogStrot releases an open-source decision compiler that runs local models on records
AIStrot is a Python library that uses an LLM once at compile time to build a frozen artifact, which then routes or classifies each record with a local decision model. The post reports 92.5% grounded routing accuracy on 398 records with the 144M-parameter Julia-1 model, compared with 74.8% for a naive baseline, at about 60 ms per record and no marginal LLM cost. The library is MIT-licensed and installs via pip install strot-ai.
Heat Trend
Not enough continuous observations to show a trend yet.