🎙 YouTube or Local Audio to Subtitles (Whisper GPU Edition)

This project lets you transcribe any audio or YouTube video into clean .srt subtitles using OpenAI Whisper via faster-whisper, accelerated by your GPU with CUDA + cuDNN.

---

⚡ Features

• 🎧 Transcribe local audio files (MP3, WAV, M4A, OGG)
• 📥 Download and transcribe YouTube videos automatically
• 💨 Uses your GPU (CUDA) for fast transcription
• 🧹 Clean, short one-line subtitles
• 💬 Outputs standard .srt subtitle file

---

🧰 Requirements

• Windows with PowerShell
• Python 3.9 or 3.10 (64-bit)
• CUDA 12.1+ (tested with 12.9.1)
• cuDNN installed in your CUDA directory
• ffmpeg (auto-included by yt-dlp or install manually)

---

🔧 Installation

1. Clone or download the repository.
2. Create and activate your virtual environment:

python -m venv venv
.\venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

---

🚀 Usage

python transcribe_from_youtube_or_file.py

You'll be prompted:

• If you choose y, paste the YouTube URL and it will auto-download the audio.
• If you choose n, a file picker will appear for you to select a local audio file.

The resulting .srt file will be saved in the same folder as the input.

---

📝 Output Example

1
00:00:00,000 --> 00:00:04,500
Welcome to this video about AI and productivity.

2
00:00:04,500 --> 00:00:08,000
Let's explore how you can save time using Whisper.

---

👨‍💻 Based on

• faster-whisper
• yt-dlp
• OpenAI Whisper

---

✅ License

MIT – use freely for educational or commercial use.