A note taking assistant that helps you summarize educational YouTube videos into concise study notes.
Project created during Hack Your Portfolio hackathon event hosted by MLH (Major League Hacking).
- Download Ollama from ollama.com/download following the installation instructions provided by Ollama's website.
- Once ollama is installed, open terminal or command prompt and pull the gemma2:9b model:
ollama pull gemma2:9b
- Start the local ollama server:
ollama serve
- Clone the repository:
git clone https://github.com/RobH0/StudyBot.git
- Create Python3 virtual environment and activate it:
# Linux create venv python3 -m venv env # Linux activate venv source env/bin/activate # Windows create venv python -m venv env # Windows activate venv venv\Scripts\activate
- Install python3 dependancies:
pip install -r requirements.txt
- Make sure Ollama server is running:
ollama serve
- Execute the study_bot.py script:
# Linux python3 study_bot.py # Windows python study_bot.py
- Respond to prompts displayed in your terminal.
A user is prompted to provide a URL to a YouTube video and pick a filename that will be used when the generated notes are saved to disk.
Once the user enters the prompted information into the CLI, the 'study_bot.py' script retrieves the YouTube video's title and English language transcript.
The YouTube video title and its transcript are used in a prompt that instructs the Gemma2 LLM (Large Language Model), run by Ollama, to summarize the transcript of the video into a concise study note document.
Once Gemma2 has finished generating the study note, it is saved as a Markdown file on disk, allowing the user to view it whenever required.