YouTube Summarization Chatbot

This project encompasses a chatbot, powered by OpenAI's GPT-3.5-turbo or GPT-4 model, and hosted on a Streamlit application. This chatbot is designed to generate summaries for YouTube videos provided by the user via a textual link.

Once you've received the quick summary, you also have the option to ask follow-up questions about the video. Moreover, you can ask the chatbot to summarize another video anytime.

The chatbot interfaces with the YouTube Transcript API to retrieve video transcripts which it then summarizes, presenting a transcription that distills the key points and main content of the video.

Explore the live application here: https://ytchatbot.streamlit.app/

Installation and Setup

Prerequisites

The main libraries this project is based on are:

• Streamlit
• OpenAI
• Youtube Transcript API

Additional detailed dependencies are listed in the requirements.txt file.

Setup

Follow these steps to setup the chatbot:

Clone the Repository

Firstly, clone the repository to your local machine:

git clone https://github.com/itsBillyZee/YTChatBot.git

This will create a new directory in your current location named YTChatBot.

Setup a Virtual Environment

It's good practice to create a virtual environment for each of your projects.

• For macOS/Linux:

  cd YTChatBot                     # Move into the cloned directory
  python3 -m venv env              # Create a new virtual environment
  source env/bin/activate          # Activate the virtual environment

• For Windows:

  cd YTChatBot                     # Move into the cloned directory
  py -m venv env                   # Create a new virtual environment
  .\env\Scripts\activate           # Activate the virtual environment

Install Dependencies

After setting up the environment and activating it, install the required dependencies:

pip install -r requirements.txt

Setup Streamlit Secrets (Optional)

To enable the owner toggle and prevent entering the API keys every time, you should create a secrets file:

1. Create a new directory .streamlit if it doesn't exist already, and within it create a new file called secrets.toml.

2. Open secrets.toml with a text editor and add the following:

OWNER_PASSCODE = "your_owner_passcode"
OPENAI_API_KEY = "your_openai_api_key"

Replace "your_owner_passcode" and "your_openai_api_key" with your desired passcode and your OpenAI API key, respectively.

Running the application

You can start the application using the below command in the terminal:

streamlit run gpt.py

Open a web browser and go to http://localhost:8501 where you should see the application running.

Usage

Provide a YouTube link in the text box and receive a quick summary. You can also ask follow-up questions about the video, or ask to summarize a different video. The application is also live and can be used directly without any installation at https://ytchatbot.streamlit.app/

Research Paper

For more detailed analysis and the research behind this project, check out the research paper here.