TinyLLaMA-1.1B Home Assistant Voice Model

This model, developed by premrajreddy, is a specialized fine-tuned version of the TinyLlama/TinyLlama-1.1B-Chat-v1.0 base model. It has been specifically trained using the acon96/Home-Assistant-Requests dataset to function as an efficient voice-controlled smart home assistant.

Key Capabilities

• Natural Language to Home Assistant Commands: Converts spoken natural language instructions directly into executable Home Assistant service calls.
• Structured Output: Generates structured JSON service commands, ensuring compatibility and ease of integration with Home Assistant automations.
• User-Friendly Confirmations: Provides friendly, conversational confirmations to the user after processing a command.
• Lightweight and Efficient: With only 1.1 billion parameters, this model is designed for efficient operation across various hardware, including CPUs and GPUs, and supports quantization for deployment via tools like Ollama.

Good For

• Developers building local, privacy-focused smart home voice assistants.
• Integrating natural language control into Home Assistant setups.
• Applications requiring a small, fast language model for command interpretation in home automation contexts.