Skip to content

This repo contains the scenario notebooks and pipelines.

Notifications You must be signed in to change notification settings

rhoai-mlops/jukebox

Repository files navigation

Jukebox 🎶

Welcome to Jukebox, a collection of scenario notebooks and pipelines designed to guide you through the AI/ML lifecycle—from data exploration to production-ready deployments.

Repository Structure

The content is organized into the following directories:

  • 1-data_exploration: Explore and preprocess the dataset, and set up the feature store.
  • 2-dev_datascience: Develop a simple model to test inference on the dataset.
  • 3-prod_datascience: Build production-ready pipelines for continuous training, validation, and inference.
  • 4-metrics: Set up TrustyAI to monitor data drift and model bias.
  • 5-data-versioning: Use DVC to version the dataset for reproducible training.
  • 6-advanced_deployments: Experiment with autoscaling and advanced deployment strategies.
  • 7-feature_store: Configure Feast (Feature Store) and use it for real-time inference.
  • 8-securing_ai: Explore security tools for AI/ML workloads.
  • 99-data_prep: Prepares raw data before the lab. This is not used during the lab but can be utilized by the instructor to refresh the dataset beforehand.

About

This repo contains the scenario notebooks and pipelines.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 16