What is RAG and How it Works - step by step visualisation
Retrieval Augmented Generation (RAG), with the help of RAG you can overcome the limitations of llms and chat with your own documents such as pdf, markdown files.
Timestamps -
0:00 Introduction of RAG
2:38 RAG vs LLMs
3:37 Understanding the architecture of RAG
4:41 Overview of Streamlit, Langchain and Huggingface
6:16 Streamlit python app ui overview
7:46 Step1 - Extracting text from pdf
9:56 Step2 - Creating text chunks
12:24 Step 3&4 - Creating embeddings and storing in vector database
17:50 Understanding similarity search and context
21:15 Final steps - Augmenting the LLM with context and query
24:20 Getting response from chatbot
27:47 Outro
To know more about langchain and document loaders -
https://python.langchain.com/docs/mod...
Steamlit
https://streamlit.io/
Huggingface
https://huggingface.co/
Thanks for watching!
Like, Share and Subscribe