RAG This is the RAG Sequence Model of the the paper Retrieval Augmented Generation for Knowledge Intensive NLP Tasks by Patrick Lewis, Ethan Perez, Aleksandara Piktus et al. The model is a uncased model, which means that capital letters are simply converted to lower case letters. The model consits of a question encoder , retriever and a generator . The retriever extracts relevant passages from the wiki dpr train datasets, which is linked above. The question encoder and retriever are based on facebook/dpr question encoder single nq base and facebook/bart large , which were jointly finetuned on on the wiki dpr QA dataset in an end to end fashion. Usage: Note : In the usage example below only the dummy retriever of wiki dpr is used because the complete lecagy index requires over 75 GB of RAM. The model can generate answers to any factoid question as follows:
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy