Load Data from DeepLake with LlamaIndex

DeepLake Reader is a LlamaIndex integration that retrieves documents from existing DeepLake datasets using vector-based queries. It loads data by accepting query vectors and distance metrics, returning documents for downstream processing.

You provide a query vector (list of floats), the path to your DeepLake dataset, a limit on the number of results to return, and a distance metric (such as "l2"). You must also authenticate with a DeepLake API token.

You receive documents from your dataset based on the vector query.

Install via pip:

pip install llama-index-readers-deeplake

Basic usage example:

from llama_index.core.schema import Document
from llama_index.readers.deeplake import DeepLakeReader

### Initialize DeepLakeReader with the token
reader = DeepLakeReader(token="<Your DeepLake Token>")

### Load data from DeepLake
documents = reader.load_data(
    query_vector=[0.1, 0.2, 0.3],  # Query vector
    dataset_path="<Path to Dataset>",  # Path to the DeepLake dataset
    limit=4,  # Number of results to return
    distance_metric="l2",  # Distance metric
)

This loader is designed to be used as a way to load data into LlamaIndex and/or subsequently used as a Tool in a LangChain Agent.

DeepLake Reader is for developers who need to load data from existing DeepLake datasets into LlamaIndex.