Advance RAG: LlamaParse + Reranker = Better RAG

Retrieval Augmented Generation ( RAG ) is all we want to talk about but are we trying or following the code practice. Out of many ways, in this video, I will show you how to better parse the documents and retrieve the most similar chunks from vector database based on query using reranker.
Code: github.com/sud...
------------------------------------------------------------------------------------------
☕ Buy me a Coffee: ko-fi.com/data...
✌️Patreon: / datasciencebasics
------------------------------------------------------------------------------------------
🤝 Connect with me:
📺 UA-cam: / @datasciencebasics
👔 LinkedIn: / sudarshan-koirala
🐦 Twitter: / mesudarshan
🔉Medium: / sudarshan-koirala
💼 Consulting: topmate.io/sud...
#llamaindex #llamaparser #llamacloud #rerank #openai #groq #datasciencebasics

[@htarsanam1194]

Thanks Sudarshan, you made RAG easier. Nice way of explaining and implementing. I have one question: Can you add the citation (source reference, e.g. page_number or chapter/section name) of each response?

[@adanpalma4026]

Thanks.. Just one question. You have some videos showing diferent. techniques for. rag... What is the most stable and roubust way to build a very good. RAG with higt success rate, becuase i am seen. that there are too many videos out there all say that theri techniques are the best for good RAG.

[@ag8247]

Hi, Thank you for insightful video. based on your experience is there limit to the size of RAG data after which accuracy is lost?

[@adanpalma4026]

Thanks. One question? Every time i want to ask. i have to start over and over again? You do all in memory ???

[@datasciencebasics]

You are welcome. You don’t need to start over the notebook, just ask question. But yes, once you are out of the colab notebook and runtime is not active, you need to run the notebook again.

[@Warung-AI-Channel]

Hi Sudarshan, thank you for sharing, we also did a similar project. We found that for safety reasons, AI can't directly access the document and read and provide us with the data from the PDF file. Even the data is stored in third-party like Pinecone. So the AI only can read if the data is public domain information and has already been there (Through URL). So how to make AI can access PDF files directly?

[@ZeeshanZafar-y7x]

Very informative video sir Just one request sir please provide us the code how to use other local llms in MarkdownElementNodeParser because every time they are taking openAi llm although we are changing it but still they are taking default OpenAi llm

[@ZeeshanZafar-y7x]

and one more question how can we pass our other vector db in VectorStoreIndex() instance we can easily pass storage context in VectorStoreIndex.from_documents() but not in VectorStoreIndex() how can we do that?

[@hoanimaiau449]

first