Retrieval Augmented Generation
Last Updated on October 31, 2024 by Editorial Team
Author(s): Derrick Mwiti
Originally published on Towards AI.
This member-only story is on us. Upgrade to access all of Medium.
Photo by ZHENYU LUO on UnsplashRetrieving documents from external sources gives language models access to data that they didnβt see during training. Particularly for question-answering systems where you want the model to answer questions based on specific documents and not from its training corpus. This is important because you can have the model answer questions from private data sources.
The process works by retrieving the relevant documents and having the model answer questions from them. Having language models help in this process is time-saving because it would be quite cumbersome to search for answers from hundreds of documents.
involves:
Loading relevant documentsCreating embeddings for the documents using an embedding modelStoring the documents and embeddings in a vector databaseQuerying the vector database using a retriever to obtain the relevant documents based on certain criteria, such as cosine similarityPassing relevant documents to the language model for question-answering
Augmenting the language models with external data is critical in ensuring that its responses are up-to-date. For example, a model that was trained before COVID-19 will make up information when asked about COVID-19. Since the language models are also trained on general domain corpora, they may not… Read the full blog for free on Medium.
Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming aΒ sponsor.
Published via Towards AI