ChemLit-QA: a human evaluated dataset for chemistry RAG tasks
Retrieval-Augmented Generation (RAG) is a widely used strategy in Large-Language Models (LLMs) to extrapolate beyond the inherent pre-trained knowledge.Hence, RAG is crucial when working in data-sparse fields such as Chemistry.The evaluation of RAG systems is Astaxanthin commonly conducted using specialized datasets.However, existing datasets, typi