Project Overview

In this project, I've crafted document summarization and document question answering.

Quick Links to Other readme files

• API readme: api/README.md
• Client readme: client/README.md

Running the Application with Docker

Prerequisites:

• Docker (https://www.docker.com/get-started)
• Docker Compose (https://docs.docker.com/compose/install/)

File Structure:

• api/: Python ML project
• client/: React project
• docker-compose.yml: Docker Compose configuration file

Steps:

1. Navigate to the project root directory:

   cd /path/to/project

2. Build the Docker images:

   docker-compose build

3. Start the containers:

   docker-compose up

Accessing the Application:

• API:
  • Open http://127.0.0.1:5001/
• Client:
  • Open http://127.0.0.1:3000/

Additional Notes:

• Stopping the Containers:

  docker-compose down

• Rebuilding Images:

  docker-compose build

• Viewing Container Logs:

  docker-compose logs

• Customizing Ports and Environment Variables:

  • Modify the docker-compose.yml file for any port or environment variable adjustments.

• Known possible Issues

  • Ensure that your Docker daemon is configured with a minimum of 12GB RAM. Insufficient memory might lead to application crashes.
  • The application's processing speed is influenced by the system's capabilities. For optimal performance, I recommend using a machine with sufficient processing power. In my testing, I used an Apple MacBook Pro M1 Pro with 16GB RAM and 512GB SSD. Summarizing a document typically takes a few minutes.
  • Upon the initial startup and API calls, the models from Hugging Face need to be downloaded, potentially extending the waiting time.