Danyalalam/End-To-End-DL-Project-With-MLOPS

Chest Cancer Classification Using Deep Learning with MLOps

Overview

This project focuses on the classification of chest cancer using deep learning techniques, particularly Convolutional Neural Networks (CNN). It provides a comprehensive pipeline from data ingestion to model deployment using MLOps practices with tools such as MLflow, DVC, and Azure CI/CD for deployment.

Key Features

• Deep Learning Model: A CNN-based model for chest cancer classification using X-ray images.
• Modular Code Structure: The project is organized in a modular fashion for better maintainability and scalability.
• Data Ingestion and Processing: Efficient data handling using Python scripts for data ingestion and preprocessing.
• Experiment Tracking: MLflow is used for tracking experiments, logging model parameters, metrics, and artifacts.
• Model Versioning and Data Management: DVC is implemented for model versioning and handling large datasets.
• Model Deployment: The model is deployed as a REST API using a Flask web application.
• CI/CD with GitHub Actions: The entire application is dockerized and deployed on Azure using a CI/CD pipeline configured via GitHub Actions.

Project Structure

Below is the simplified project structure:

├── artifacts/
│   └── training/
│       └── model.h5
├── src/
│   ├── components/
│   │   ├── data_ingestion.py
│   │   ├── data_preprocessing.py
│   │   └── model_training.py
│   ├── config/
│   │   └── configuration.py
│   └── utils/
│       └── utilities.py
├── app.py
├── Dockerfile
├── README.md
├── requirements.txt
└── .github/
    └── workflows/
        └── azure-deploy.yml