Benchmarking machine learning inferencing on embedded hardware.
-
Updated
Jul 18, 2025 - Python
Benchmarking machine learning inferencing on embedded hardware.
Hetman-2b is first ai model that was developed as local laptop model and uses under 1gb of vram and runs on laptops with rtx2050+.
Offline farm evidence reports for smallholder loans. Hack-Nation x World Bank Small AI for Development, Challenge 4 (Agriculture).
Offline-first, browser-local coffee-leaf observation prototype with human review.
An editable spatial accessibility model
Kahawa, powered by Field Voice: offline Swahili voice reports for smallholder coffee farmers; AI sorts, a human officer decides. World Bank Small AI Hackathon 2026.
To associate your repository with the small-ai topic, visit your repo's landing page and select "manage topics."