Retrieval Augmented Generation is a straightforward way to build customized LLM applications without investing in expensive model training.

You simply retrieve a chunk of data, and add it to the prompt along with the task for the LLM.

Most applications retrieve data by using embeddings to identify similar pieces of text.

For example, embeddings can help you identify a paragraph in a book that most closely resembles a reader’s question.

Pass the paragraph, the question and the instructions to the LLM and it will generate an answer for the reader.

For more generated AI and LLM related content - check out ‪@prolegoinc‬