How To Evaluate MLOps Tools

Показать описание

An example of how to evaluate MLOps Tooling

This talk was given at Stanford for CS 329S Machine Learning System Design on Feb 2, 2022

Links:

Hamel Husain

Рекомендации по теме

Комментарии

Fantastic talk Hamel, been using TF for years and still wondered whenever to use TFX.

Many open-source tools come from enterprise but doesn’t mean you should use them. How many companies have even 1/10 the scale of Google?

And finally tool zealots remind me of the classic engineers disease: building tools for tools sake rather than using said tools.

mrdbourke

Excellent presentation. "Complexity is just simplicity in large amounts" - this approach helps address it.

stephensimpson

A lot of great points, thanks Hamel. The religious argument is esp. dangerous. Even if the baseline at Google is higher than in your average tech company (let's assume that's the case) that still doesn't mean that the baseline at the team that developed TFX is higher than that of the competitor X. Or even if they are more talented maybe the competitor startup has way higher incentives to succeed and eventually makes a better product. It's...a complex dynamics. Shape rotators tend to overcomplicate things. Ultimately you need wordcels to build usable/readable software ;)

TheAIEpiphany

Excellent line of argument. I will use many of those points myself. Thanks!

allthingsdata

Awesome content here. Don't become a tool Zealot

jimmiemunyi

How To Evaluate MLOps Tools

How To Evaluate MLOps Tools

MLOps explained | Machine Learning Essentials

Comet Office Hours: The Best MLOps Tools for Beginners

How to Learn MLOps

MLOps tools to scale your production machine learning || Alejandro Saucedo @ FOSDEM 2019

MLOps Tutorial #5: Automated Testing for Machine Learning

Productionize Your ML Workflows with MLOps Tools

MLOps Tools update 2021

Day 3 - Data Ingestion Component | MLOPs Production Ready Deep Learning Project

Deliver high-performance ML models faster with MLOps tools

What tools do I use as an MLOps Engineer? #shorts #mlops #mlengineer #datascience #dayinthelife

Use This Way Of Training Machine Learning Models For Efficiency

Panel: MLOps Tools - Practical Considerations When Building Out Your Infrastructure

Integrating multiple MLOps tools together on Google Cloud Platform

MLFlow Tutorial Part 1: Experiment Tracking

01. Introduction To MLflow | Track Your Machine Learning Experiments | MLOps

MLflow Pipelines: Accelerating MLOps from Development to Production

MLOps (v2) - Unifying MLOps at Microsoft

What Is MLOps?

The merging of DataOps and MLOps, how does this look?

MLOps In Practice – How To Run Your Machine Learning Models In Production At Enterprise Scale

All Machine Learning Models Explained in 5 Minutes | Types of ML Models Basics

Duarte Carmo - MLOps for the rest of us- A poor man's guide to putting models in production

First Principles ML System Design, and not MLOps tools, is all you need!