The RAG Visual Breakdown - The Ultimate guide to building powerful LLM pipelines!

In this video, we talk about Retrieval Augmented Generation. The idea of RAGs are pretty simple – suppose you want to ask a question to a LLM, instead of just relying on the LLM's pre-trained knowledge, you first retrieve relevant information from an external knowledge base. This retrieved information is then provided to the LLM along with the question, allowing it to generate a more informed and up-to-date response. In this video, we are going to start with the most basic barebones RAG pipeline – and identify how individual components of this pipeline works and how modern frameworks have made it ultra-powerful and ultra-reliable.

To get access to the write-up, slides, and other files produced for every video in the channel, check out our Patreon.

#ai #largelanguagemodels #machinelearning

Resources :

Timestamps:
0:00 - Intro
1:19 - Retrieval Augmented Generation Blueprint
4:00 - Chunking and Contextual Chunking
6:54 - Data Conversion - Language Model Embeddings
8:29 - Data Conversion - TF-IDF and BM-25
10:54 - Vector and Graph Databases
13:00 - Query Rewriting
14:21 - Contextual Query Rewriting, HYDE
15:24 - Post Retrieval
16:00 - Reciprocal Rank Fusion
17:00 - Outro

[hailrider]

These are some of the best videos available on the topics you are discussing. Nicely done!

[thanX]

I found you through your article on medium (suggested on the stream of my browser on mobile), I read it in its entirety and you already have my respect and esteem. Now I'll watch the video...

[aumvyas]

Surfed the entire YouTube but couldn't find a more visual and easier explaination for RAGs. I would like you to continue this RAG video series and explore more on the subdomains of Agentic RAG and RAGs in production.

Big ups brother🎉

[foramjoshi]

I love the clarity of the video, and the diagrams really helped me understand the concept better!!! Thank you, AVB, and please keep the videos coming.

[andydataguy]

Great channel man! Thanks for this video 🙌🏾💜

[sivi]

Looking forward to the RAG series considering POC RAG is easier and Production quality grade RAG is a whole different animal. Can't agree more on a separate video just for the research on pure vector database like Milvus vs RDMS vector feature enabled like PostgreSQL PGvector Vs no-sql vector feature enabled like elastic search vs graph database like neo4j!!!

I am grateful to find your channel. Keep up your clarity in the diagrams and the choice of words in your explanation!!! Awesome work!

[TheShreyas]

Quite an insightful video, would be glad if you create a playlist explaining all these tools and methods individually if that works or in a better fashion. However, a highly insightful video, thanks :)

[TP-ctqm]

You explain really well man! Kudos and thanks 🙏

[OmKale-hr]

Good work bro. Keep it up 💪


I would suggest you to create proper playlist of these

[blancanthony]

Excatement ce que je cherchais, merci !!!!

[archiee]

How to implement some of that techniques, such as multiple retrieval and ranking consolidation / rank fusion, or embeddings+bm25 retrieval. Do u have videos on that? :)

[mechwar]

What's the design software you use for those diagrams?