Speech-to-Text Service

Overview

Speech-to-Text service using gRPC and the Whisper model from Hugging Face's Transformers library. The service allows users to send audio data and receive transcribed text in response.

Features

• gRPC-based communication for efficient data transfer.
• Utilizes the Whisper model for high-quality speech recognition.
• Includes a client for sending audio data and receiving transcriptions.
• Unit tests to ensure the functionality of the service.

Requirements

• Python 3.10 or higher

Setup Instructions

1. Clone the Repository

   git clone https://github.com/your-username/speech-to-text-service.git
   cd speech-to-text-service

2. Install Dependencies Make sure to install the required packages as mentioned above:

   pip install -r requirements.txt
   pip install -r requirements-dev.txt

3. Install Pre-commit To install pre-commit, run:

   pip install pre-commit

4. Environment Variables Create a .env file in the root directory and set the following variables for local environment:

   ENVIRONMENT=local
   PORT=50051
   

   For non-local environments (development or production), set the following variables:

   ENVIRONMENT=development  # or production
   SSL_PRIVATE_KEY_PATH=<path-to-private-key>
   SSL_CERTIFICATE_CHAIN_PATH=<path-to-certificate-chain>
   

5. Run the gRPC Server Start the server by running:

   python main.py

6. Run the Client In a separate terminal, run the client to send audio data:

   python src/client/client.py

7. Run Pre-commit Hooks To ensure code quality, install and run pre-commit hooks:

   pre-commit run --all-files

8. Code Formatting and Linting Use ruff for linting and black for code formatting:

   ruff check .
   black .

Testing

To run the unit tests, execute the following command:

python -m unittest discover -s src/tests

Audio Data

Place your test audio files in the data directory. The client will look for test_audio.wav by default.

Contributing

Feel free to submit issues or pull requests. Contributions are welcome!

License

This project is licensed under the MIT License.