Overview of Radio LLM: Integrating LLMs with Meshtastic Networks

Radio LLM is a platform designed to facilitate the integration of Large Language Models (LLMs) with the Meshtastic mesh communication network. This repository, hosted on GitHub and developed by pham-tuan-binh, provides tools and functionalities that allow users on a Meshtastic network to interact with an LLM for automated, context-aware responses and to execute specific tasks through simple commands.

Key Features

• Bi-directional Communication: Enables two-way communication between the Meshtastic devices and the LLM, supporting both general broadcasts and targeted responses.
• Automatic Message Chunking: Manages long responses by automatically breaking them into chunks of up to 200 characters to fit the network's message size constraints.
• Context-Aware Interactions: Maintains a history of messages to provide context for LLM responses. It can also include node-specific information like battery level and location in the responses.
• Task Execution: Users can prompt the LLM to perform tasks such as calling emergency services, sending messages, or retrieving sensor information.

Requirements

• Python 3.8 or higher
• Meshtastic Python library
• Ollama LLM Python SDK
• PubSub library

Setup and Installation

1. Device Connection: Connect your Meshtastic device via USB or configure it for TCP access.
2. Repository Cloning: Clone the repository using git clone <repo_url>.
3. Dependency Installation: Install necessary dependencies with pip install -r requirements.txt.
4. Script Execution: Run the script using python main.py.

Usage

Radio LLM functions similarly to a chatbot on platforms like Discord. Users can send commands to configure interactions or execute specific tasks:

• LLM Chat Features:

  • /enable_llm: Activates the LLM chat feature.
  • /disable_llm: Deactivates the LLM chat feature.

• Echo Features (for testing Meshtastic):

  • /enable_echo: Activates echo.
  • /disable_echo: Deactivates echo.

Configuration

Users can modify the LLM model by updating the chat function in chat_with_llm. Adjustments can be made to the chunk size or message length limits as needed.

Custom Tools

To add custom tools:

1. Define the tool in model/tool_handler.py.
2. Register the tool in model/tool_registry.py.
3. Describe the tool in config.yaml.

Interfaces

Different interfaces for connecting to Meshtastic devices are supported, including TCP, BLE, and serial connections. Users should select the appropriate interface based on their setup and follow the Meshtastic documentation for detailed instructions.

Compliance and Licensing

The project is licensed under the GPL-3.0 license. Users are advised to ensure compliance with local laws and regulations when using Meshtastic devices and LLMs.

Contributions

Contributions to the project are welcome, and users can submit issues or pull requests on GitHub for improvements or bug fixes.

Radio LLM offers a practical solution for integrating advanced language processing capabilities with robust mesh network communications, enhancing the utility and interactivity of Meshtastic networks.