A vector store that uses the Momento Vector Index.

Remarks

To sign up for a free Momento account, visit https://console.gomomento.com.

Hierarchy

Constructors

Methods

  • Adds vectors to the index. Generates embeddings from the documents using the Embeddings instance passed to the constructor.

    Parameters

    • documents: Document[]

      Array of Document instances to be added to the index.

    • Optional documentProps: DocumentProps

    Returns Promise<void>

    Promise that resolves when the documents have been added to the index.

  • Adds vectors to the index.

    Parameters

    • vectors: number[][]

      The vectors to add to the index.

    • documents: Document<Record<string, any>>[]

      The documents to add to the index.

    • Optional documentProps: DocumentProps

      The properties of the documents to add to the index, specifically the ids.

    Returns Promise<void | string[]>

    Promise that resolves when the vectors have been added to the index. Also returns the ids of the documents that were added.

    Remarks

    If the index does not already exist, it will be created if ensureIndexExists is true.

  • Return documents selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to the query AND diversity among selected documents.

    Parameters

    • query: string

      Text to look up documents similar to.

    • options: MaxMarginalRelevanceSearchOptions<this["FilterType"]>

    Returns Promise<Document[]>

    • List of documents selected by maximal marginal relevance.
  • Searches the index for the most similar vectors to the query vector.

    Parameters

    • query: number[]

      The query vector.

    • k: number

      The number of results to return.

    Returns Promise<[Document<Record<string, any>>, number][]>

    Promise that resolves to the documents of the most similar vectors to the query vector.

  • Stores the documents in the index.

    Converts the documents to vectors using the Embeddings instance passed.

    Parameters

    • texts: string[]

      The texts to store in the index.

    • metadatas: object | object[]

      The metadata to store in the index.

    • embeddings: EmbeddingsInterface

      The embeddings instance to use to generate embeddings from the documents.

    • dbConfig: MomentoVectorIndexLibArgs

      The configuration to use to instantiate the vector store.

    • Optional documentProps: DocumentProps

      The properties of the documents to add to the index, specifically the ids.

    Returns Promise<MomentoVectorIndex>

    Promise that resolves to the vector store.

Generated using TypeDoc