Hacker News,

Training custom wake words like "Hey Alexa" is often a resource-intensive task, demanding powerful hardware and complex manual tuning.

NanoWakeWord is an open-source framework designed to solve this. It features an intelligent engine that automates the ML pipeline, making it possible to build high-performance, production-ready wake word models with minimal effort.

What makes it different:

Train Anywhere, On Anything: The core architecture is built for extreme efficiency. You can train on massive, terabyte-scale datasets using a standard laptop or even a low-spec machine, all without needing a GPU. This is achieved through memory-mapped files that stream data directly from disk, eliminating RAM limitations.

Intelligent Automation: The framework analyzes your data to automatically configure an optimal model architecture, learning schedule, and training parameters. It removes the guesswork from building a robust model.

Total Flexibility and Control: While it automates everything, it also offers deep customization. You can choose from 11+ built-in architectures (from lightweight DNNs to SOTA Conformers) or easily extend the framework to add your own custom architecture. Every single parameter generated by the engine can be manually overridden for full control.

Smarter Data Processing: It moves beyond generic negatives. The system performs phonetic analysis on your wake word to synthesize acoustically confusing counter-examples, which drastically reduces the false positive rate in real-world use.

Ready for the Edge: Models are exported to the standard ONNX format. The framework also includes a lightweight, stateful streaming inference engine designed for low-latency performance on devices like the Raspberry Pi.

Try It in Your Browser (No Install Needed):

This single Google Colab notebook is a playground to train your first model. Inside, you can select and experiment with any of the available architectures with just a few clicks.

Launch the Training Notebook: https://colab.research.google.com/github/arcosoph/nanowakewo...

The goal is to produce models with an extremely low false positive rate (tests show less than one false activation every 16-28 hours on average).

The project is actively developed by Arcosoph, and all feedback or questions are highly welcome!

Key Links:

GitHub Repo: https://github.com/arcosoph/nanowakeword

PyPI Package: https://pypi.org/project/nanowakeword/

Pre-trained Models: https://huggingface.co/arcosoph/nanowakeword-models#pre-trai...

Discord Community: https://discord.gg/rYfShVvacB

Every Reason to Hate Cars

Whisqr – Ethical Investment Platform – Invest with Your Values

Studies Suggest Red Meat May Help Prevent Alzheimer's

Wordle as a Two-Player Game

3D-Printing a Trombone

Let's Take Esoteric Programming Languages Seriously [video]

The Quest for Oral GLP-1s

Show HN: Covalence – Cross-Client Memory for Claude, Cursor, and MCP (macOS)

CISA spikes CyberCorps internships amid shutdown

GitHub Copilot is serving Opus 4.7 at 7.5x multiplier until April 30th

Replace the standard deduction with a standard credit to start UBI

Zigfrak: Shoot-and-Loot Action RPG

An MCP control plane that sits between your agent and every tool it can reach

From five optional fields to a discriminated union: CLI parsing with Optique 1.0

Boycott of major AI conference exposes a growing US–China divide

Show HN: NanoWakeWord – Open-source wake word training for any device

Anthropic's AI downgrade stings power users

Show HN: I built a music theory course with games and spaced repetition

rlvrbook

Ask HN: Does ChatGPT's data export feature work?

#1 Tip for Developers with No Idea What to Build [video]

Locally Uncensored v2.3.3 – local AI with Qwen3.6 and remote access

We improved OneDrive folder automount

RunReveal in Kubernetes for On-Prem and in VPC SIEM

Health care can't be the only job in town – but it is

Going Headless Is Hard

Let's Kill Webhooks

Turning RAG pipelines into enterprise-grade Data Subscriptions

Launch HN: Kampala (YC W26) – Reverse-Engineer Apps into APIs

15 Ways I'm Using AI to Manage My Small Business

Show HN: Git-chain – Simplify branch dependency management in Git

Comments