What does llama index Query Engine use under hood for summarization? It seems like GPT , reason why I'm asking.
Want to build something were we want data to be on premise. Llama Index is great for on premise RAG. Which works well with local Mongo DB. But the summarization then can't work with GPT ..
