Go with the flow–Track your machine learning lifecycle
summary
Uses MLflow to version experiments on the California housing dataset, running a baseline Linear Regression and an XGBoost pipeline while logging hyperparameters, metrics, CSV train/test artifacts and the fitted model to a local SQLite-backed tracking server. Demonstrates the mlflow tracking UI, registers the XGBoost model into the model registry and moves it to production, then loads that production model to predict ~117,340 USD; UI reveals potential overfitting so retraining is needed.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 2:25Parkside Interactive
- 5:40Track ML lifecycle
- 12:09Introduction to MLflow and its components
- 15:07MLflow tracking basics
- 18:17Tutorial and dataset
- 21:15Exploratory data analysis and preprocessing
- 28:44Setup MLflow tracking
- 35:59Linear regression experiment
- 37:27XGBoost experiment
- 39:45MLflow UI exploration
- 43:46Register models
- 45:31Deploy to production
- 48:31Making predictions with registered models
- 50:39MLflow features
- 52:26Q&A
related talks
45mMachine Learning Lifecycle Made Easy with MLflow
57mManaging Machine Learning Lifecycle with MLflow
31mMLflow: Platform for Complete Machine Learning Lifecycle
30mMLflow: an Open Source Tool for Managing the ML Project Lifecycle
01h 25mIntroduction to MLOps with MLflow
33mOperationalizing Model Lifecycle with MLFlow
30mEnd-to-End MLOps with MLflow and Kubeflow
17mIntroducing mlflow