Unburdening Primary Healthcare

Welcome to the Unburdening Primary Healthcare: An Open-Source AI Clinician Partner Platform documentation. This project is a collaboration as part of the ClinicianFOCUS initiative with Conestoga College and Dr. Braedon Hendy. Below you’ll find documentation for the core modules, API endpoints, and utilities used in this project. Below are the pages contents:

Installation

To install and run the speech2text-container project with the latest changes, follow these instructions:

Prerequisites

Installation Steps

  1. Clone the Repository:

    First, clone the project repository by running the following command:

    git clone https://github.com/ClinicianFOCUS/speech2text-container.git

Usage

To deploy and run the speech transcription service using Docker Compose, follow the steps below.

  1. Build and run the services using Docker Compose:

    The project includes a docker-compose.yml file to simplify the setup of the speech transcription service. This file defines two services: - speech-container for the Whisper speech transcription model. - caddy for handling reverse proxy duties.

  2. Prerequisites: - Install Docker and Docker Compose on your system. - Ensure that the necessary ports (default: 2224) are available on your machine.

  3. Starting the Services:

    Run the following command to build and start the containers:

    docker-compose up --build

    This will: - Build the speech-container using the provided Dockerfile and start it with the specified WHISPER_MODEL, WHISPER_PORT, and WHISPER_HOST. - Build and run the caddy container as a reverse proxy.

  4. Service Configuration:

    The following environment variables can be customized to control the behavior of the speech-container service:

    • WHISPER_MODEL: The Whisper model to use (default: medium).

    • WHISPER_PORT: The port to expose the service (default: 2224).

    • WHISPER_HOST: The host to bind the service (default: 0.0.0.0).

    • UVICORN_WORKERS: Number of Uvicorn workers (default: 1).

    • DEBUG_MODE: Enable debug mode (default: False).

    Example:

    WHISPER_MODEL=large WHISPER_PORT=8080 WHISPER_HOST=127.0.0.1 docker-compose up --build

    This will start the transcription service with the large Whisper model on port 8080, accessible only from 127.0.0.1.

  5. Stopping the Services:

    To stop the services, use the following command:

    docker-compose down

    This will stop and remove the containers but will retain the built images.

API Reference

Below are the available API endpoints for interacting with the Whisper Server.

Whisper Audio Endpoint

URL: /whisperaudio

Method: POST

Description: This endpoint accepts an audio file (MP3 or WAV) and transcribes it using the Whisper model. Ensure that an API key is provided via the Authorization Bearer header.

Parameters:

  • file (required): The audio file (MP3 or WAV).

  • api_key (required): A valid API key for accessing the endpoint.

Response:

  • 200: Transcription text in a JSON format.

  • 400: Invalid file type.

  • 401: Unauthorized.

  • 413: Request entity too large.

  • 419: Rate limit exceeded.

  • 500: Server error.

Example:

curl -X POST "http://localhost:2224/whisperaudio" \
     -H "Authorization: Bearer <api_key>" \
     -F "file=@/path/to/audiofile.wav"

Response:

{
    "text": "Transcribed text here"
}

Python Modules

Below are the core Python modules used in this project.