YouTubeScript 🎙️

Project Description

YouTubeScript is a simple and efficient application that allows users to transcribe YouTube videos or uploaded audio files into text. It supports both original language transcription and language translation with subtitle files in .vtt format.

Features

• Transcribe YouTube videos via URL
• Transcribe uploaded audio files
• Translate transcriptions to multiple languages
• Generate .vtt subtitle files
• Summarize transcriptions using a language model
• Easy interface with clear fields for repeated use

Installation

1. Clone the Repository

git clone https://github.com/yourusername/YouTubeScript.git
cd YouTubeScript

2. Install Dependencies

pip install -r requirements.txt

3. Install ffmpeg (Required for Whisper)

• Linux: sudo apt install ffmpeg
• Windows: Download here

Required Packages

• gradio
• openai-whisper
• transformers
• yt-dlp

Usage

Run the application with:

python app.py

How to Use

1. Select input type: YouTube video or audio file.
2. Provide the YouTube link or upload an audio file.
3. Choose the mode: Original or Translate.
4. Select a language if translating.
5. Click Get YouTubeScript 🪄.
6. View the transcription summary and download the .vtt file.
7. Click **Clear Fields ** to reset the interface.

Project Structure

.
├── app.py        # Main application code
├── requirements.txt # Dependency file
├── README.md     # Project documentation
└── sub.vtt       # Generated subtitle file

Sample Commands

• To transcribe a YouTube video:
  • Provide the URL and click Get YouTubeScript 🪄.
• To transcribe an uploaded file:
  • Upload an audio file and follow the same steps.

⚠ Troubleshooting

• No transcription or empty subtitles: Make sure ffmpeg is installed.
• Failed YouTube download: Confirm yt-dlp installation.

Happy Transcribing!