The ZepClient instance used to interact with Zep's API.
Adds documents to the collection. The documents are first embedded into vectors using the provided embedding model.
The documents to add.
Optional
kOrFields: number | Partial<VectorStoreRetrieverInput<ZepCloudVectorStore>>Optional
filter: string | objectOptional
callbacks: CallbacksOptional
tags: string[]Optional
metadata: Record<string, unknown>Optional
verbose: booleanDeletes documents from the collection.
The list of Zep document UUIDs to delete.
Return documents selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to the query AND diversity among selected documents.
Text to look up documents similar to.
Performs a similarity search on the Zep collection.
The query string to search for.
Optional
k: number = 4The number of results to return. Defaults to 4.
Optional
filter: string | object = undefinedAn optional set of JSONPath filters to apply to the search.
Optional
_callbacks: Callbacks = undefinedOptional callbacks. Currently not implemented.
Static
fromCreates a new ZepVectorStore instance from an array of Documents. Each Document is added to a Zep collection.
The Documents to add.
Pass FakeEmbeddings, Zep Cloud will handle text embedding for you.
The configuration object for the Zep API.
Static
fromCreates a new ZepVectorStore instance from an array of texts. Each text is converted into a Document and added to the collection.
The texts to convert into Documents.
The metadata to associate with each Document. If an array is provided, each element is associated with the corresponding Document. If an object is provided, it is associated with all Documents.
Pass FakeEmbeddings, Zep Cloud will handle text embedding for you.
The configuration object for the Zep API.
Static
init
ZepCloudVectorStore is a VectorStore implementation that uses the Zep long-term memory store as a backend.
If the collection does not exist, it will be created automatically.
Requires
@getzep/zep-cloud
to be installed: