/youtube-qa

Primary LanguageJupyter Notebook

youtube-qa

Transcribing youtube video and asking Questions from the transcription.

Try it out!

Bhavyas_video__1__AdobeExpress.1.mp4

Flow

graph LR
A[Youtube Audio] -->B(openai-whisper)
B-->C(Transcription)
C -->|Query| D(Q&A Model)
D -->E[Answer]

Project Organization

├── README.md          <- The top-level README for developers using this project.
|
├── app
│   ├── app.py               <- Gradio App
│   ├── requirements.txt     <- Requirements for Gradio App
│
├── notebooks          
|   ├── Transcribe.ipynb                   <- Process of Transcription with openai-whisper
│   ├── question-answering.ipynb           <- Using Q&A Model on generated text
│
└──  requirements.txt   <- The requirements file for reproducing the analysis environment, e.g.

Run Locally

Clone the project

  git clone https://github.com/bhavya-giri/youtube-qa

Go to the project directory

  cd youtube-qa

Install dependencies

  pip install requirements.txt

Start the notebook

  jupyter notebook

or open with Juoyter Lab

 jupyter lab