STT (Speech-to-Text with Transformers)

This project implements a Speech-to-Text (STT) system built using Transformers.
It provides modular components for audio preprocessing, model inference, and pipeline integration, making it easy to extend and adapt to different speech recognition tasks.

---

🚀 Features

• Speech-to-Text transcription using Transformer-based models.
• Modular architecture:
  • components/ – Core building blocks
  • models/ – Transformer-based speech recognition models
  • pipeline/ – End-to-end pipeline for transcription
  • utils/ – Helper functions
• Configurable settings in configuration/.
• Logging and exception handling built-in.

---

📂 Project Structure

STT/
│── cloud_storage/ # Cloud storage utilities
│── components/ # Core components of the STT system
│── configuration/ # Configuration files
│── constants/ # Constant values
│── entity/ # Data/entity definitions
│── exceptions/ # Custom exception handling
│── logger/ # Logging setup
│── models/ # Transformer models for STT
│── pipeline/ # Transcription pipeline
│── utils/ # Utility functions
│── templates/ # Templates (if needed for deployment/UI)
│── app.py # Main entry point
│── requirements.txt # Python dependencies
│── setup.py # Setup script

---

⚙️ Installation & Setup

1. Clone the repo:

   git clone https://github.com/mohamedabbouda/STT.git
cd STT

3. Install dependencies:

 pip install -r requirements.txt

⚙️ Run pipeline (generate feedback logs)

 python pipeline.py

⚙️ Run web app

python app.py