Show HN: Offline Tiny-LMs 25MB for personal goals, career and pitch deck advice

Category: ai-ml

Tags: offline-ml, onnx, cli-tool, personal-advice, tiny-models

Score: 4.8/10 (Innovation: 5, Technical: 5, Documentation: 5, Utility: 4)

etiny packages compact ONNX models (25-36MB) with rule engines, regex detectors, and vector indexes to deliver offline advice scanning on personal goals, career, and pitch decks. It offers a CLI to explore local markdown/text files and bundles JSON/Python assets for use with external chat tools. The approach of distilling curated advice corpora into tiny in-browser models is a modestly novel combo, but scope and adoption remain narrow.

Target audience: individuals and knowledge-workers seeking offline AI assistance

Repository: https://github.com/shanedevane/etiny · Python · 13 stars

View on Hacker News