Model Experimentation with MLFlow

This is a beginner-friendly project to understand how to create experiments and test model versions a simple model with MlFlow locally. The data for this project is available here https://www.kaggle.com/datasets/anshtanwar/global-data-on-sustainable-energy/data.

Table of Contents

1. [Project Overview]
2. [Installation]
3. [Project Structure]

---

Project Overview

About

Our objectives are the below:

• Testing a bagging regressor model with MLFlow.
• Creating a streamline process for data ingestion, data preparation, model training and model evaluation.
• MLFlow, Sklearn, Numpy and Pandas are mainly used, few other ancillary libraries are used as well.

Screenshots or Demos

Below screenshot of the logged model after pre-processing, training and testing steps completed with MLFlow.

Installation

Prerequisites

Below libraries needs to be installed in the virtual environment:

1. MLFlow
2. Pandas

Instructions

After cloning the repository, please follow the below steps:

1. Install all the libraries mentioned above into the virtual environment.
2. Afterwards, run 'python tracker.py'
3. Finally, run 'mlflow ui --port 5000' to view the MLFlow dashboard to view the logged model.
4. If you improve your model, you can re-run the steps again and see the improved model.

---

Project Structure

├── loading_data.py      # Ingests the data from the Kaggle website. 
├── cleaning_data.py     # Cleans the data and prepares it. 
├── model.py             # The model architecture, training and evaluation. 
├── tracker.py           # Logging all details into MlFlow for the training run and inference. 
└── README.md            # Project documentation

---