Overview of Kubeflow

Below is an overview of Kubeflow.

*Double-tap to expand/collapse an item. Left-tap & drag to move. Mouse-wheel/pinch to zoom.
Knowledge Graph |  Text |  Top
Kubeflow Overview An open-source project that packages machine learning code for running on Kubernetes Kubeflow Pipelines A platform for composing, deploying and managing end-to-end ML workflows - promoting reusability and portability Concepts Reusability Portability ML Engineer can create a pipeline component/step for feature engineering, modelling and hyper-parameter tuning, while the Data Engineer can use that step as part of their pipeline as part of their data engineering solution. The solution can then appear as a service used by a Data Analyst to derive business insights. Features UI for managing and tracking experiments, jobs and runs, and sharing pipelines Engine for scheduling multi-step ML workflows SDK for defining and manipulating pipelines and components Notebooks for interacting with the system using the SDK