justindav1s / ai-on-openshift

Using Openshift to train and serve AI/ML models

Configuring Openshift 4 to use GPUs

Start here :

https://docs.nvidia.com/datacenter/cloud-native/openshift/introduction.html

Building Jupyter notebooks capable of running AI/ML workloads using Tensorflow and GPUs

https://developer.nvidia.com/blog/streamlining-nvidia-driver-deployment-on-rhel-8-with-modularity-streams/

Dploying Jupyter Notebooks at scale

Developing AI/ML models with Openshift and Jupyter Notebooks

Deploying and using AI/ML models on OPenshift

Sources

https://github.com/kubeflow/kubeflow

https://www.lightbend.com/blog/how-to-deploy-kubeflow-on-lightbend-platform-openshift-introduction

https://github.com/rbo/RedHatHackathon-KubeFlow-Demo/blob/master/deployer/kubeflow-deploy.sh

GPU stuff

Install CUDA, this will also install an appropriate driver

https://blog.openshift.com/how-to-use-gpus-with-deviceplugin-in-openshift-3-10/

for OCP 3.11 start here :

http://docs.h2o.ai/driverless-ai/latest-stable/docs/userguide/install/rhel.html

https://nvidia.github.io/nvidia-docker/