How from-kishore-kumar/end-to-end-ml-project-with-mlflow works
Project type and purpose
• This is an end-to-end machine-learning application and web service for wine-quality prediction.
• It implements a reusable ML pipeline (data ingestion, validation, transformation, training, evaluation), experiment tracking (MLflow, DagsHub), and a Flask-based web front end. It also has CI/CD via GitHub Actions and Docker/AWS deployment.
File structure
• Top level:
• app.py: Flask application
• main.py: orchestrates full pipeline run
• config/: YAML configuration
• params.yaml, schema.yaml: param and schema definitions
• Dockerfile, .github/workflows/main.yaml: container build & CI/CD
• research/: exploratory notebooks
• src/mlProject/: core pipeline code organized by components, config, constants, entities, pipeline stages, and utils
• templates/: HTML templates for the Flask UI
• Patterns: clear pipeline stages (stage_01…05), modular components under src/mlProject/components.
README for additional insights
• Describes MLflow tracking, AWS ECR/ECS deployment, Flask UI, and CI/CD steps.
• Workflow: update configs → run pipeline → serve model via Flask → monitor with MLflow & DagsHub.