First taste of using Vertex AI Generative AI with LangChain
I don’ t think I’m alone in feeling like the world of large language models (LLMs) is hard to keep up with these days. Every week there are new announcements and new potential directions. With so many LLM / generative AI themes competing for my attention, it’s hard to decide where to invest my time to learn something new. LangChain, a framework for developing LLM-powered applications, is one development in the world of LLMs that looks like it has staying power. This weekend I told myself I would tune out some of the overall LLM chatter and focus on trying out LangChain. As I describe in this article, it was easy to get a first taste of LangChain, and I am eager to learn more.
Step 1: Flowise
To ease myself in, I started with this excellent video on Flowise, a no-code front-end for LangChain. Following the step-by-step instructions in the video, I was able to set up a basic Flowise Chatflow to generate embeddings for a PDF version of my book , Deep Learning with Structured Data, load the embeddings into a vector database, and answer questions about the book in a chat interface.
