This repository provides training data for Home Assistant's local voice control, enabling users to interact with their smart home devices using natural language. It targets Home Assistant users seeking to enhance their voice command capabilities with a focus on privacy and local processing. The project facilitates community contributions to expand language support and intent recognition.

How It Works

The project structures training data for voice intents into YAML files, categorized by language and intent domain. It uses a defined format for sentences, responses, and tests, allowing for systematic development and validation. A custom Python script, script.intentfest, is provided to parse sentences, validate data, sample potential voice commands, and even generate LLM prompts to aid in translation efforts.

Quick Start & Requirements

• Install/Run: python3 -m script.intentfest <command> (e.g., validate, parse, sample)
• Prerequisites: Python 3.x, pytest (for testing). Devcontainers are recommended for development.
• Links: Documentation

Highlighted Details

• Supports community-driven expansion of languages and intents.
• Includes tools for data validation, sentence parsing, and sampling.
• Offers LLM prompt generation to assist with translations for new languages.
• Repository layout clearly separates languages, intents, responses, and tests.

Maintenance & Community

• Community-driven project with a focus on contributions.
• Development is facilitated via devcontainers.

Licensing & Compatibility

• License: Not explicitly stated in the README.
• Compatibility: Designed for use with Home Assistant.

Limitations & Caveats

The specific license is not detailed, which may impact commercial use or integration into closed-source projects. The project relies on community contributions for language and intent expansion.