Skip to content
Nappai Documentation

Chroma DB

Chroma DB is a component in the Nappai system that acts as a vector store, allowing users to store and retrieve vectorized documents efficiently. It is particularly useful for applications that require similarity-based searches, making it easier to find relevant documents quickly.

Relationship with Chroma

Chroma DB utilizes the Chroma technology to provide vector storage and search capabilities. This integration allows users to perform efficient searches based on document similarity or Maximal Marginal Relevance (MMR), leveraging Chroma’s powerful vector processing capabilities.

Inputs

  • Collection Name: The name of the collection in the vector store where documents are stored.
  • Persist Directory: The directory where data is saved for persistence.
  • Search Query: The search query used to retrieve documents from the vector store.
  • Ingest Data: The data to be added to the vector store.
  • Embedding: The embedding function used to convert documents into vectors.
  • Server CORS Allow Origins: Specifies allowed origins for CORS on the server (advanced setting).
  • Server Host: The host address of the Chroma server (advanced setting).
  • Server HTTP Port: The HTTP port of the Chroma server (advanced setting).
  • Server gRPC Port: The gRPC port of the Chroma server (advanced setting).
  • Server SSL Enabled: Indicates if SSL is enabled on the server (advanced setting).
  • Allow Duplicates: Specifies if duplicates are allowed in the vector store (advanced setting).
  • Search Type: The type of search to perform, either Similarity or MMR (advanced setting).
  • Number of Results: The number of results to return from a search (advanced setting).
  • Limit: The limit of records to compare when duplicates are not allowed (advanced setting).

Outputs

The component produces a Chroma object representing the constructed vector store. This object can be used to perform searches and manage documents within the vector store, facilitating efficient document retrieval based on vector similarity.

Usage Example

Imagine you have a large collection of research papers and you want to find papers similar to a specific topic. You can use Chroma DB to store these papers as vectors and perform a similarity search to quickly find relevant documents.

Templates

Currently, there are no specific templates where this component is used.

  • LCVectorStoreComponent: The parent component from which Chroma DB inherits, providing foundational vector store functionalities.

Tips and Best Practices

  • Ensure that the Persist Directory is correctly set to avoid data loss.
  • Use the Search Type setting to tailor searches to your specific needs, choosing between Similarity and MMR.
  • Regularly update the Ingest Data to keep your vector store current and relevant.

Security Considerations

  • When enabling SSL (Server SSL Enabled), ensure that your server configuration is secure to protect data in transit.
  • Carefully manage Server CORS Allow Origins to prevent unauthorized access to your vector store.

By following these guidelines, you can effectively use Chroma DB within the Nappai system to enhance your document management and retrieval processes.