Optimizing for relevance using MongoDB, LlamaIndex and Together.ai

In this notebook, we will explore and tune different retrieval options in MongoDB's LlamaIndex integration using Together.ai to get the most relevant results.

Step 1: Install libraries

• pymongo: Python package to interact with MongoDB databases and collections

- **llama-index**: Python package for the LlamaIndex LLM framework

- **llama-index-llms-together**: Python package to use TogetherAI models via their LlamaIndex integration

- **llama-index-vector-stores-mongodb**: Python package for MongoDB’s LlamaIndex integration

Step 2: Setup prerequisites

• Set the MongoDB connection string: Follow the steps here to get the connection string from the Atlas UI.

• Set the Together.ai API key: Steps to obtain an API key as here

Step 3: Load and process the dataset

Step 4: Define the LLM and Embedding Model

Step 5: Create MongoDB Atlas Vector store

Create Atlas Vector Index

Create Atlas Search Index

Step 6: Get movie recommendations

Query

Aggregate

Full-text search

Vector search

Hybrid search

Combining metadata filters with search