MLflow Started Code

Welcome to the MLflow Started Code repository! This repository provides a hands-on example of how to use MLflow for tracking experiments, comparing models, and managing machine learning workflows.

Project Overview

This project demonstrates how to:

• Create and manage MLflow experiments.
• Train and evaluate multiple machine learning models (Decision Tree and Random Forest).
• Log metrics, parameters, feature importances, and predictions.
• Save and load models using MLflow.

Repository Structure

• run_experiments.py: Script to train models, log metrics, and save artifacts.
• requirements.txt: Dependencies for the project.
• results/: Directory where logs, model artifacts, and predictions will be saved.

Source: MLflow Documentation

Setup Instructions

1. Clone the Repository

git clone https://github.com/palbha/mlflow_started_code.git
cd mlflow_started_code

2. Install Dependencies

Make sure you have Python 3.6+ installed. Then, install the required packages:

pip install -r requirements.txt

3.Usage

Running the Experiment Script Execute the run_experiments.py script to start the MLflow experiment

python run_experiments.py 

Once the file ran completely fine, take a look at the mlflow UI to see the results from your experiments

mlflow server --host 127.0.0.1 --port 8080

Open your browser & go to http://127.0.0.1:8080/ & You can see th experiments

Click on any experiments & take a look at artifacts to analyse & see the output further

One can also download the details of each run to create their own custom graphs & share results with stakeholders