gaianet-node

GaiaNet Node provides an out-of-the-box AI agent node that you can run locally to host and interact with a conversational AI model. The node downloads models, sets up a vector database, and exposes a web interface for chatting with the AI agent. Use the quick start flow to install, initialize, and run the node, then navigate to the provided node URL to interact with the AI. The system supports commands to initialize with a specific config, start and stop the node, and update configuration, making it suitable for local development and experimentation with different models and knowledge bases.

Once installed, you can initialize the node with gaianet init, which downloads models, vector embeddings, and a Qdrant instance. Start the node with gaianet start to bring up the WasmEdge runtime, the QA/RAG components, and the web dashboard. You can access the node's address from the console output and open it in a browser to view node information and chat with the AI agent. The node supports stopping gracefully via gaianet stop and updating configuration with gaianet config. For advanced usage, you can pass a pre-set configuration URL to gaianet init to bootstrap the node with a particular model and knowledge base, or run gaianet start --local-only to expose services only on localhost.

Prerequisites:

• curl and bash available on your system (Linux/macOS/WSL).
• A compatible environment for running WasmEdge and Qdrant if you plan to use full features.
• Optional: Docker if you prefer containerized setup.

Install steps:

1. Quick install script (default stack):

curl -sSfL 'https://github.com/GaiaNet-AI/gaianet-node/releases/latest/download/install.sh' | bash

This installs the default GaiaNet Node stack into the user’s home directory (by default under $HOME/gaianet).

2. Initialize the node (downloads models, embeddings, and QA components):

gaianet init

Output will show progress downloading models, embeddings, and Qdrant setup.

3. Start the node:

gaianet start

The command prints the node’s public or local URL where you can access the dashboard and chat with the AI agent.

4. Optional: Stop or reconfigure:

gaianet stop

gaianet config --chat-url "https://example.com/model.gguf"

If you prefer installing in an alternative base directory, re-run the install script with the --base flag:

curl -sSfL 'https://raw.githubusercontent.com/GaiaNet-AI/gaianet-node/main/install.sh' | bash -s -- --base $HOME/gaianet.alt

Tips and notes:

• The node uses WasmEdge, Qdrant, and a rag-api-server as part of its stack. Ensure you have sufficient disk space for model files and embeddings.
• The init step reads configuration from $HOME/gaianet/config.json by default; you can specify a different config with --config URL or --base to point to another installation directory.
• To run locally only (no public GaiaNet domain exposure), start with gaianet start --local-only.
• If you encounter network or download failures, re-run the install or init steps to retry downloading heavy assets. Use the provided sample outputs as references to confirm successful operations.
• You can customize the initial configuration by using gaianet init --config <config.json URL> for pre-set models and knowledge bases.