Vector Database Configuration

DeepSearcher uses vector databases to store and retrieve document embeddings for efficient semantic search.

📝 Basic Configuration

config.set_provider_config("vector_db", "(VectorDBName)", "(Arguments dict)")

Currently supported vector databases: - Milvus (including Milvus Lite and Zilliz Cloud)

🔍 Milvus Configuration

config.set_provider_config("vector_db", "Milvus", {"uri": "./milvus.db", "token": ""})

Deployment Options

Local Storage with Milvus Lite

Setting the uri as a local file (e.g., ./milvus.db) automatically utilizes Milvus Lite to store all data in this file. This is the most convenient method for development and smaller datasets.

config.set_provider_config("vector_db", "Milvus", {"uri": "./milvus.db", "token": ""})

Standalone Milvus Server

For larger datasets, you can set up a more performant Milvus server using Docker or Kubernetes. In this setup, use the server URI as your uri parameter:

config.set_provider_config("vector_db", "Milvus", {"uri": "http://localhost:19530", "token": ""})

Zilliz Cloud (Managed Service)

Zilliz Cloud provides a fully managed cloud service for Milvus. To use Zilliz Cloud, adjust the uri and token according to the Public Endpoint and API Key:

config.set_provider_config("vector_db", "Milvus", {
    "uri": "https://your-instance-id.api.gcp-us-west1.zillizcloud.com", 
    "token": "your_api_key"
})