Pocket Field Trip: a little local AI, then a little time outside
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass . What I Built Pocket Field Trip gives people taking a short break from desk work one small thing to do outside. You describe the kind of break you want, set your available time, and say whether walking is welcome. A local model chooses a mission from a small, readable collection. Then the app invites you to…
Pocket Field Trip is a project for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass. The application helps users who need a brief outdoor activity after a break from desk work. Users input their desired break type, available time, and whether walking is allowed. The app then selects a local outdoor activity from a small, pre-defined collection.
Users can close the app once they have completed their task. The AI component used is the paraphrase-MiniLM-L3-v2 sentence embedding model, converted to ONNX format and run on a CPU using ONNX Runtime and Hugging Face tokenizer. The model receives one responsibility: connecting the user’s request to an outdoor activity. The app checks outdoor conditions and stops if they are unsuitable.
The tradeoff is limited coverage, as the model can only suggest eight activities for a few useful intentions. The project emphasizes local checks, ensuring the model runs and constraints hold, without requiring internet access. All AI components are licensed under Apache-2.0.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
