ML-Inz-example 0.0.1

A project with Flask backend and ML service for search answers in documents.

ML-Inz-example

This project contains two main components:

1. A Flask backend serving HTML, CSS, and JavaScript.
2. A machine learning service using the nvidia/Llama3-ChatQA-1.5-8B model to answer questions based on documents.

Project Overview

Flask Backend

The Flask backend serves the frontend application, which includes HTML, CSS, and JavaScript files. It provides the user interface where users can input their question and document, and view the generated answer.

Machine Learning Service

The machine learning service utilizes the nvidia/Llama3-ChatQA-1.5-8B model. This service accepts a question and a document as input, processes them using the model, and returns an answer based on the content of the document.

Features

• Flask Backend: Serves the frontend and handles user interactions.
• ML Service: Uses a state-of-the-art language model to provide answers based on the provided documents.
• Docker Support: Both components can be easily deployed using Docker and Docker Compose.
• CORS Handling: Ensures smooth interaction between the backend and the ML service.
• Loading Indicator: Shows a loading spinner while the ML model processes the input.

Project Structure

my_project/
│
├── backend/
│   ├── app/
│   │   ├── __init__.py
│   │   ├── main.py
│   │   ├── templates/
│   │   │   └── index.html
│   │   ├── static/
│   │   │   ├── css/
│   │   │   │   └── style.css
│   │   │   └── js/
│   │   │       └── script.js
│   ├── __init__.py
│   ├── Dockerfile
│   ├── requirements.txt
│
├── ml_service/
│   ├── app/
│   │   ├── __init__.py
│   │   ├── main.py
│   │   ├── model.py
│   ├── __init__.py
│   ├── Dockerfile
│   ├── requirements.txt
│
├── tests/
│   ├── __init__.py
│   ├── test_backend.py
│   ├── test_ml_service.py
├── __init__.py
├── docker-compose.yml
├── setup.py
├── README.md
├── MANIFEST.in
└── pytest.ini

Running the Project

Using Docker Compose

To build and run the project with Docker Compose, use the following command:

docker-compose up --build

This command will build the Docker images for both the Flask backend and the ML service, and then start the containers.

Backend

The Flask backend will be accessible at http://localhost:8000. It serves the frontend application where users can input their question and document.

ML Service

The ML service will be accessible at http://localhost:8001. This service handles the processing of the question and document using the nvidia/Llama3-ChatQA-1.5-8B model and returns the answer.

Usage

1. Open your browser and navigate to http://localhost:8000.
2. Enter your question in the provided input field.
3. Enter the document text in the provided textarea.
4. Click the "Generate" button.
5. Wait for the response to be generated and displayed below the form.

Development

Prerequisites

• Docker and Docker Compose
• Python 3.11 or higher

Setting Up the Environment

1. Clone the repository:

git clone https://github.com/Appjey/ML-Inz-example.git
cd my_project

2. Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows use `venv\ScriptsActivate`

3. Install the dependencies:

pip install -r backend/requirements.txt
pip install -r ml_service/requirements.txt

Running Locally

To run the backend locally:

cd backend
python app/main.py

To run the ML service locally:

cd ml_service
python app/main.py

Running Tests

To run the tests, use the following command:

pytest

Contributing

If you would like to contribute, please open a pull request with your changes. Make sure to update the tests as appropriate.

License

This project is licensed under the MIT License. See the LICENSE file for details.