Skip to content

canonical/kubeflow-single-node-dgx

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Nvidia DGX Kubeflow setup

This document contains setup instructions for running Charmed Kubeflow on Nvidia's DGX enabled hardware. You can find those instructions in the kubernetes-setup.md file.

We provide also examples on how to use Kubeflow instance with multiple GPUs. In the file gpu-pipeline.ipynb you can find basic example of Kubeflow pipeline, where we demonstrate multi GPU training on Tensorflow model. This example provides instrctions on how to deploy the trained model with MlFlow and Seldon core.

In the file gpu-notebook.ipynb you can find example Jupyter notebook with the same classiffier as in case of pipeline. Notebook also uses multi GPU setup.

Author: Michal Hucko (michal.hucko@canonical.com)

About

Guide how to deploy and test kubeflow on single node DGX system.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published