MLOps-Basics

This project is a basic introduction to the features of MLflow. It's designed to help me understand how MLflow works and how it can be used in machine learning projects.

Project Overview

In this project, I've set up a simple machine learning workflow to explore the features of MLflow. The goal wasn't to follow a rigorous data science workflow, but rather to get things running and see how it goes.

Features Explored

Here are some of the MLflow features I've explored in this project:

1. Tracking: I've used MLflow's tracking feature to log parameters, metrics, and artifacts. This helped me understand how MLflow can be used to keep track of different experiments in a systematic way.

2. Projects: I've set up this project as an MLflow project, which helped me understand how MLflow can be used to package machine learning code in a reusable and reproducible way.

3. Models: I've used MLflow's model feature to save and load models. This helped me understand how MLflow can be used to manage models and their versions.

Running the Project

To run this project, you need to have Python and MLflow installed on your machine. Once you have these prerequisites, you can follow these steps:

1. Clone the repository to your local machine (install requirements).
2. Navigate to the project directory in the terminal.
3. Run the main script with the command: python main.py
4. After running the script, start the MLflow tracking server with the command: mlflow ui
5. After running mlflow ui the link that appears on the terminal click on it to access the platform

Running the Project on DagsHub

To run this project on DagsHub, you need to have a DagsHub account. Once you have an account, you can follow these steps:

1. Fork the repository to your DagsHub account.
2. Set up the environment variables for MLflow tracking. You can do this by running the following commands in your terminal:

export MLFLOW_TRACKING_URI=https://dagshub.com/your_username/your_repo_name.mlflow
export MLFLOW_TRACKING_USERNAME=your_username
export MLFLOW_TRACKING_PASSWORD=your_password

3. Run the script python main.py

Please note that this project is a basic introduction to MLflow and does not follow a rigorous data science workflow.

Future Work

I plan to explore more advanced features of MLflow and use it in more complex machine learning projects. I also plan to follow a more rigorous data science workflow in my future projects.