Optimizing Retrieval-Augmented Generation: Analysis of Hyperparameter Impact on Performance and Efficiency

Adel Ammar, Anis Koubaa, Omer Nacar +1

Large language models achieve high task performance yet often hallucinate or rely on outdated knowledge. Retrieval-augmented generation (RAG) addresses these gaps by coupling gener…