Looking into the ideas and processes behind chunking, summarizing, and approaches to different problems with technical documents. Intended to be a broad overview of some cost saving and reliable ways to approach this problem with python.
Example document used: www.cisco.com/c/en/us/product...
Chapters:
00:00 - Intro
01:07 - Text Splitting & Chunking
02:45 - Token Example
03:39 - Setting Up the Code & Choosing the Document
05:14 - Loading and Splitting the Document
06:06 - Setting up LangChain
06:47 - Instantiating Model & Prompt
09:34 - Main Summarization Script
11:28 - Analyzing the Output
14:23 - Outro & Closing Thoughts