Course Outline

Introduction

Overview of Kubeflow Features and Components

  • Containers, manifests, etc.

Overview of a Machine Learning Pipeline

  • Training, testing, tuning, deploying, etc.

Deploying Kubeflow to a Kubernetes Cluster

  • Preparing the execution environment (training cluster, production cluster, etc.)
  • Downloading, installing and customizing.

Running a Machine Learning Pipeline on Kubernetes

  • Building a TensorFlow pipeline.
  • Building a PyTorch pipleline.

Visualizing the Results

  • Exporting and visualizing pipeline metrics

Customizing the Execution Environment

  • Customizing the stack for diverse infrastructures
  • Upgrading a Kubeflow deployment

Running Kubeflow on Public Clouds

  • AWS, Microsoft Azure, Google Cloud Platform

Managing Production Workflows

  • Running with GitOps methodology
  • Scheduling jobs
  • Spawning Jupyter notebooks

Troubleshooting

Summary and Conclusion

Requirements

  • Familiarity with Python syntax 
  • Experience with Tensorflow, PyTorch, or other machine learning framework
  • A public cloud provider account (optional) 

Audience

  • Developers
  • Data scientists
  28 Hours
 

Number of participants


Starts

Ends


Dates are subject to availability and take place between 09:30 and 16:30.
Open Training Courses require 5+ participants.

Related Courses

Related Categories