Ragas

Ragas is a framework to help you evaluate your retrieval augmented generation (RAG) applications.

Ragas and Weaviate

The Ragas metrics are:

• faithfulness
• answer_relevancy
• context_precision
• context_recall

To use Ragas, you'll need to store the question, answer, ground_truths, and contexts in a json object to send to Ragas.

Our Resources

Hands on Learning: Build your technical understanding with end-to-end tutorials.

Hands on Learning

Topic	Description	Resource
Introduction Ragas Demo	Learn how to use Weaviate and Ragas	Notebook
Ragas and LangChain	Learn how to connect Ragas and LangChain.	Notebook