Skip to content
Nappai Documentation

Milvus

Milvus is a component in Nappai that helps you find similar items in your data much faster than traditional search methods. It’s especially useful when dealing with data that’s not easily described with simple keywords, like images, audio, or complex text. It does this by representing your data as “vectors” – mathematical representations that capture the essence of your data. Milvus then uses these vectors to find the most similar items based on their vector representations.

Relationship with Milvus

This component uses the Milvus vector database to store and search your data. Milvus is a powerful and scalable database specifically designed for handling vector data, making searches incredibly fast and efficient, even with massive datasets.

Inputs

  • Collection Name: The name you give to your collection of data within Milvus. Think of it like a folder name for your data. Defaults to “langflow”.
  • Collection Description: A short description of your data collection. This is optional.
  • Connection URI: The address where your Milvus database is located. This is usually provided by your Milvus administrator. Defaults to “http://localhost:19530”.
  • Connection Password: The password for your Milvus database (if required). Leave this blank if your Milvus database doesn’t need a password.
  • Primary Field Name: The name of the unique identifier for each item in your collection. Defaults to “pk”.
  • Text Field Name: The name of the field containing the text description of each item. Defaults to “text”.
  • Vector Field Name: The name of the field containing the vector representation of each item. Defaults to “vector”.
  • Search Query: The text you use to search for similar items.
  • Ingest Data: The data you want to add to your Milvus collection. This will be a list of items.
  • Embedding: This refers to the method used to convert your data (like text or images) into vectors. Nappai will handle this for you.
  • Number of Results: How many of the most similar items you want Milvus to return in your search results. Defaults to 4.

Advanced Inputs (Optional):

  • Connection Arguments: Additional settings for connecting to your Milvus database. Only use this if you have specific instructions from your Milvus administrator.
  • Consistency Level: Controls how consistent the data is across the system. Usually, the default (“Session”) is fine. Choose other options only if you have specific requirements.
  • Index Parameters: Advanced settings for how Milvus indexes your data for faster searching. Leave this at the default unless you have specific needs.
  • Search Parameters: Advanced settings for how Milvus performs searches. Leave this at the default unless you have specific needs.
  • Drop Old Collection: If checked, this will delete any existing collection with the same name before creating a new one. Use with caution!
  • Timeout: How long to wait for a response from Milvus before giving up.

Outputs

The Milvus component doesn’t have explicit outputs in the traditional sense. Instead, it returns the results of your search query as a list of items that are most similar to your search. These items will be presented in a user-friendly format within the Nappai dashboard.

Usage Example

Let’s say you have a collection of product descriptions. You can use Milvus to find products similar to a specific product. You would input the description of the target product as your Search Query, and Milvus would return a list of the most similar products from your collection.

Templates

[List of templates where the component can be seen and its configuration] (This section requires information not provided in the prompt)

[Links to other related components and a brief description of each] (This section requires information not provided in the prompt)

Tips and Best Practices

  • Make sure your Milvus database is running and accessible before using this component.
  • Use descriptive names for your collections.
  • Start with the default settings and only adjust advanced options if necessary.
  • Test your search queries to ensure they return the expected results.

Security Considerations

  • Protect your Milvus database credentials (URI and password). Do not hardcode these values directly into your Nappai workflows; use secure configuration methods provided by Nappai.
  • Regularly review and update your Milvus database security settings.