Introduction
RAG, or Retrieval-Augmented Generation, is a cutting-edge AI technology developed by Facebook AI that combines information retrieval and knowledge databases with a sequence-to-sequence model. RAG introduces a new dynamic to traditional NLP (Natural Language Processing) techniques by enabling the model to refer to an external knowledge base during the generation process and thus provide more detailed and informative responses.
The RAG implementation in the SERMAS toolkit can ingest data from: