This repository has a script demonstrating how two of LangChain's VectorStore retrievals algorithms work. Retrievers are a great tool for efficient document retrieval and prioritization, and here we try to showcase these two functionalities!

• Similarity Search: Finds documents similar to a given query by comparing their similarities.
• Maximal Marginal Relevance (MMR): Prioritizes diverse and relevant documents in search results.

1. Create a virtual enviroment

python3 -m venv .venv   
source .venv/bin/activate

2. Install dependecies

pip install -r requirements.txt

3. Set up OpenAPI key

# On a .env file:
OPENAI_API_KEY=your_key

4. Run the script!

python3 langchain_retriever_texts.py

Don't forget to have fun! Play around with the queries, add more text into the database!

This blog post was inspired by the insights and knowledge gained from the LangChain: Chat with Your Data course offered by deeplearning.ai.