hyperdb-python 0.1.1

A hyper-fast local vector database for use with LLM Agents.

HyperDB

A hyper-fast local vector database for use with LLM Agents. Now accepting SAFEs ($35M cap minimum.)

Installation

Install the package from PyPI:

pip install hyperdb-python

Usage

Here's an example of using HyperDB to store and query documents:

import json
from hyperdb import HyperDB

# Load documents from the JSONL file
documents = []

with open("demo/pokemon.jsonl", "r") as f:
    for line in f:
        documents.append(json.loads(line))

# Instantiate HyperDB with the list of documents and the key "description"
db = HyperDB(documents, key="info.description")

# Save the HyperDB instance to a file
db.save("demo/pokemon_hyperdb.json")

# Load the HyperDB instance from the save file
db.load("demo/pokemon_hyperdb.json")

# Query the HyperDB instance with a text input
results = db.query("Likes to sleep.", top_k=5)