Azure CosmosDB MongoDB Vector Store

In this notebook we are going to show how to use Azure Cosmosdb Mongodb vCore to perform vector searches in LlamaIndex. We will create the embedding using Azure Open AI.

If you’re opening this Notebook on colab, you will probably need to install LlamaIndex 🦙.

%pip install llama-index-embeddings-openai
%pip install llama-index-vector-stores-azurecosmosmongo
%pip install llama-index-llms-azure-openai

!pip install llama-index

import os
import json
import openai
from llama_index.llms.azure_openai import AzureOpenAI
from llama_index.embeddings.openai import OpenAIEmbedding
from llama_index.core import VectorStoreIndex, SimpleDirectoryReader

Setup Azure OpenAI

The first step is to configure the models. They will be used to create embeddings for the documents loaded into the db and for llm completions.

import os

# Set up the AzureOpenAI instance
llm = AzureOpenAI(
    model_name=os.getenv("OPENAI_MODEL_COMPLETION"),
    deployment_name=os.getenv("OPENAI_MODEL_COMPLETION"),
    api_base=os.getenv("OPENAI_API_BASE"),
    api_key=os.getenv("OPENAI_API_KEY"),
    api_type=os.getenv("OPENAI_API_TYPE"),
    api_version=os.getenv("OPENAI_API_VERSION"),
    temperature=0,
)

# Set up the OpenAIEmbedding instance
embed_model = OpenAIEmbedding(
    model=os.getenv("OPENAI_MODEL_EMBEDDING"),
    deployment_name=os.getenv("OPENAI_DEPLOYMENT_EMBEDDING"),
    api_base=os.getenv("OPENAI_API_BASE"),
    api_key=os.getenv("OPENAI_API_KEY"),
    api_type=os.getenv("OPENAI_API_TYPE"),
    api_version=os.getenv("OPENAI_API_VERSION"),
)

from llama_index.core import Settings

Settings.llm = llm
Settings.embed_model = embed_model

Download Data

!mkdir -p 'data/paul_graham/'
!wget 'https://raw.githubusercontent.com/run-llama/llama_index/main/docs/docs/examples/data/paul_graham/paul_graham_essay.txt' -O 'data/paul_graham/paul_graham_essay.txt'

Loading documents

Load the documents stored in the data/paul_graham/ using the SimpleDirectoryReader

documents = SimpleDirectoryReader("./data/paul_graham/").load_data()

print("Document ID:", documents[0].doc_id)

Document ID: c432ff1c-61ea-4c91-bd89-62be29078e79

Create the index

Here we establish the connection to an Azure Cosmosdb mongodb vCore cluster and create an vector search index.

import pymongo
from llama_index.vector_stores.azurecosmosmongo import (
    AzureCosmosDBMongoDBVectorSearch,
)
from llama_index.core import VectorStoreIndex
from llama_index.core import StorageContext
from llama_index.core import SimpleDirectoryReader

connection_string = os.environ.get("AZURE_COSMOSDB_MONGODB_URI")
mongodb_client = pymongo.MongoClient(connection_string)
store = AzureCosmosDBMongoDBVectorSearch(
    mongodb_client=mongodb_client,
    db_name="demo_vectordb",
    collection_name="paul_graham_essay",
)
storage_context = StorageContext.from_defaults(vector_store=store)
index = VectorStoreIndex.from_documents(
    documents, storage_context=storage_context
)

Query the index

We can now ask questions using our index.

query_engine = index.as_query_engine()
response = query_engine.query("What did the author love working on?")

import textwrap

print(textwrap.fill(str(response), 100))

The author loved working on multiple projects that were not their thesis while in grad school,
including Lisp hacking and writing On Lisp. They eventually wrote a dissertation on applications of
continuations in just 5 weeks to graduate. Afterward, they applied to art schools and were accepted
into the BFA program at RISD.

response = query_engine.query("What did he/she do in summer of 2016?")

print(textwrap.fill(str(response), 100))

The person moved to England with their family in the summer of 2016.