Making Lt. Data - with Retrieval Augmented Generation (RAG) and the OpenAI API

This is an excerpt from our online course "Machine Learning, Data Science and Generative AI with Python" by Frank Kane. Enroll now at:

In this lesson, we cover the concepts of retrieval augmented generation (RAG) in plain language, including the use of embeddings and vector databases. Then we move on to a real example using Google CoLab to process all of Data's lines of dialog from Star Trek: The Next Generation, and augment GPT 3.5 using vector search for the lines most similar to the user's query. The result is a simulated Data that responds in ways consistent with the original scripts, and at much lower cost than using the fine tuning API (see our earlier video on that.)

We created a real AI version of a fictional AI character! How cool this that :)

[kagantimur]

Thanks for sharing it is a helpful video