Running Evals on a Bloated RAG Pipeline | Towards Data Science

Comparing metrics across datasets and models