Kubeflow - bobbae/gcp GitHub Wiki

Kubeflow is the ML toolkit for Kubernetes.

Using the Kubeflow configuration interfaces you can specify the ML tools required for your workflow. Then you can deploy the workflow to various clouds.

Kubeflow offers several components that you can use to build your ML training, hyperparameter tuning, and serving workloads across multiple platforms.

https://www.kubeflow.org/docs/about/kubeflow/

https://www.youtube.com/watch?v=cTZArDgbIWw&list=PLIivdWyY5sqLS4lN75RPDEyBgTro_YX7x

MLOps

MLOps is the process of taking an experimental Machine Learning model into a production web system. The word is a compound of “Machine Learning” and the continuous development practice of DevOps in the software field. Machine Learning models are tested and developed in isolated experimental systems.

Deploy Kubeflow on Google Cloud

Instructions for kubectl and kpt to deploy Kubeflow on Google Cloud.

https://www.kubeflow.org/docs/distributions/gke/deploy/deploy-cli/

ML Workflow

https://www.kubeflow.org/docs/started/kubeflow-overview/#introducing-the-ml-workflow

MNIST end to end on Kubeflow on GKE

https://github.com/kubeflow/examples/blob/master/mnist/mnist_gcp.ipynb

Kubeflow pipelines

Kubeflow Pipelines is a component of Kubeflow that provides a platform for building and deploying ML workflows, called pipelines. Kubeflow Pipelines is a platform for building, deploying, and managing multi-step ML workflows based on Docker containers. Pipelines are built from self-contained sets of code called pipeline components. Kubeflow Pipelines can be deployed in AI Hub.