What is Indexing? Indexing Methods for Vector Retrieval

Video 1/10 of the "From Beginner to Advanced LLM Developer" course by Towards AI (linked above).
The most practical and in-depth LLM Developer course out there (~90 lessons) for software developers, machine learning engineers, data scientists, aspiring founders or AI/Computer Science students. We’ve gathered everything we worked on building products and AI systems and put them into one super practical industry-focused course. Right now, this means working with Python, OpenAI, Llama 3, Gemini, Perplexity, LlamaIndex, Gradio, and many other amazing tools (we are unaffiliated and will introduce all the best LLM tool options). It also means learning many new non-technical skills and habits unique to the world of LLMs.

Learn more for free...

#ai #rag #llm

[vbridgesruiz-phd]

This is interesting to learn and compare across platforms.

Neo4j uses index-free adjacencies. Indexes are not used to traverse graph databases, but to find a starting point for the query. This means that how many nodes are in the graph does not affect the speed of retrieval!

With vector databases, indexes sound complicated. But, actually, it's just adding another layer to enable semantic search in addition to structured, deterministic search.

Ideally, we want a hybrid of the two approaches when working with text-rich tabular data!

[thomasgoodwin]

That's surely a ton of perfectly simplified heavy duty brain food. I grok!
🖖😎👍

[jacobchow]

Very easy-to-understand video. I have a question though, in 3. Inverted File Index, how do we get the centroids?

[imarri]

Narration and audio is out of sync between 1:03 and 1:17. Great video otherwise.

[aizazkhan]

Good video with simple nice explanations but I would suggest slowing down a bit. Everything was so rushed. It gives your brain a little time to stay focused.