Improving the Life of Data Scientists - Automating ML Lifecycle through MLflow
Databricks via YouTube
Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore an insightful conference talk on automating the machine learning lifecycle using MLflow to improve data scientists' productivity. Discover Flock, an end-to-end platform designed for Enterprise Grade Machine Learning (EGML) applications. Learn how Flock leverages MLflow to simplify and automate crucial steps in the data science process, allowing professionals to focus on model improvement. Delve into Flock's features, including automatic logging, integration with relational databases, model optimizations, and support for ONNX format and runtime. Gain valuable insights into ongoing work on tracking data and ML model lineage, essential for regulated environments. Watch a demonstration of Flock's capabilities using Microsoft's Azure Data Studio and MLflow, and understand how this platform addresses challenges in data handling, model fairness, user privacy, and debuggability in enterprise applications.
Syllabus
Introduction
About Azure Systems Lab
Class of Systems
Summary
Insights
Endtoend platform
Demo
Overview
What is Onyx
Converting models to Onyx
Onyx Rebase Optimizer
Office Team Example
Integration with Onyx
Selling the Model
Store Models in Databases
Provenance
Future Work
Taught by
Databricks