Explore Kubeflow's potential to revolutionize Machine Learning DevOps in this 46-minute conference talk from GOTO Chicago 2019. Discover how this cutting-edge technology enables seamless transitions from development to production for data scientists and software engineers. Learn about Kubeflow's integration with Kubernetes clusters, secure Jupiter Notebooks, and TF Job implementation. Follow along with a comprehensive demo of Kubeflow's architecture, user interface, and pipeline dashboard. Gain insights into running and coding pipelines, as well as making predictions. Delve into the challenges faced in implementing Kubeflow and explore solutions to common issues. Perfect for those interested in advancing their knowledge of machine learning workflows and DevOps practices in the context of modern data science.
Accelerating Machine Learning DevOps with Kubeflow
Overview
Syllabus
Introduction
Show of hands
Agenda
Who am I
City Scholars
The Problem
The Data
The Doctrine Community
Other stuff to do
Docker
Challenges
Kubeflow doesnt know
Everything starts wide open
What is Kubeflow
Secure Jupiter Notebooks
TF Job
Fraying Attributes
MiniKF
Demo Architecture
User Interface
Pipeline Dashboard
Pipeline Demo
Running a Pipeline Demo
Pipeline Code Walkthrough
Prediction Walkthrough
Additional Content
Taught by
GOTO Conferences
Related Courses
-
Machine Learning Using Kubeflow and Kubernetes
-
Kubeflow 1.0 Update by a Kubeflow Community Product Manager
-
Introduction to Kubeflow and Machine Learning on Kubernetes
-
Kubeflow for Machine Learning
-
Machine Learning Made Easy on Kubernetes - DevOps for Data Scientists
-
Introducing Kubeflow for Machine Learning and AI on Kubernetes
