The RAG pipeline
The way to make a model answer with knowledge it does not have. And why bad RAG is usually the fault of retrieval rather than of the LLM.
2 steps
110 XP
A free account is needed
Start the lesson →
Sources
- Lewis, P. et al. 2020 · Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks · NeurIPS 2020
- Gao, Y. et al. 2024 · Retrieval-Augmented Generation for Large Language Models: A Survey · arXiv:2312.10997
- Liu, N. et al. 2024 · Lost in the Middle: How Language Models Use Long Contexts · TACL 2024
- Nogueira, R. & Cho, K. 2019 · Passage Re-ranking with BERT · arXiv:1901.04085
ML Academy · an interactive machine learning course that runs in your browser ·
All lessons