KQED Codebase

This repository contains scripts for processing and evaluating podcast data as part of the KQED Bot project. It includes:

1. Chunking Script: Tools to process raw podcast transcripts into metadata-rich chunks for better handling and retrieval. You need access to our firebase to run this properly. If needed, please email [email protected] or [email protected]
2. Automated Scripting: Scripts for evaluating chatbot responses using curated queries and GPT-4O.

Folder Structure

• Chunking_Script/: Contains code for chunking podcast transcripts and formatting metadata.
• Automated_Scripting/: Includes automated testing scripts for evaluating the performance of Podbot and WikiChat.

Quick Start

1. Clone the repository:

   git clone https://github.com/your-username/KQED_Codebase.git
   cd KQED_Codebase