Sloth Inbox: let a tiny open model sort the notes you keep avoiding
Описание и применение
Sloth Inbox использует открытый embedding‑модель all‑MiniLM‑L6‑v2, работающую в браузере, чтобы автоматически сортировать заметки по пользовательским категориям. Приложение загружает модель (~23 МБ), генерирует предложения, а пользователь может отредактировать их и экспортировать в Markdown. В тесте 12 заметок все правильно классифицированы, но 4 остаются в режиме проверки из‑за низкой схожести.
Задача
Organizing loose notes into sensible buckets is tedious.
Результат
The app correctly categorized all twelve synthetic notes, automatically assigning eight and leaving four for review due to low similarity scores. Processing time after model loading was 1.24 s.
Рабочий процесс
- Paste a messy list of notes
- Model downloads (≈23 MB) and runs in a web worker
- App suggests shelves for each note
- User reviews and edits suggestions
- Preview Markdown
- Download or copy the organized list
Доказательства
I built Sloth Inbox to take one small chore off that list: putting loose notes into sensible buckets.
The AI runs in the browser. There is no account, API key, or server waiting to read your notes.
I tested the deployed example with twelve notes. Its top shelf suggestion matched the intended category for all twelve, but the app assigned only eight automatically.
The displayed processing time was 1.24 seconds after model loading in that run.
The core is all-MiniLM-L6-v2, an open embedding model, running with Transformers.js 3.8.1 and WebAssembly.
- Дата записи: 2026-10-21T20:11:13.088Z