Discover LlamaIndex: Bottoms-Up Development With LLMs (Part 2, Documents and Metadata)

In part 2 of this series, we cover how to load data and create document objects in LlamaIndex. We explore the concept of documents and nodes, and how they are used for retrieval and question answering. We also dive into customizing documents by adding metadata and controlling which metadata is used for different parts of LlamaIndex. Finally, we demonstrate how to create DOM documents for the LlamaIndex documentation using a custom document loader for markdown files. No action is requested from the viewers in this video.

Our Bottoms Up Development series is designed to help you understand the low-level building blocks of LLM application development. It's important to play around with the basic concepts before composing more complex LLM software systems - our end goal is to help you learn all the fundamental concepts to build a chatbot over our documentation.

[meera_datey]

Super useful video series! Understanding the basics helps so much to build RAG applications

[fabianaltendorfer]

Great series! I don't know if your documentation includes pictures, but how would you deal with that? Maybe pictures include information too? thx

[johngonzalezvideo]

Hi! Excellent tutorial. I have a question. If I have a large markdown table ~100 rows. What is the best option: split it in few rows for similar category or mantain full table as a simple document? What llama index tool you recommend me for table manipulation? Thanks

[benjohnson]

Thank you, but once again this video is outdated because the locations of the modules have changed with the newer releases.

[andrewcbuensalida]

if you're doing a tutorial about llamaindex, please don't use llamaindex as an example corpus. It could get confusing.

[just-Banks]

Is this supposed to be "bottom-up" development, as in starting with the fundamentals and building up to higher concepts throughout the tutorials?

"Bottoms up" as an expression refers to turning up drink glasses to imbibe libations. And there's another dirty connotation to the term that I can think of.