Skip to content
FindOpenSource

Kubeflow

Machine learning platform for Kubernetes that bundles pipelines, model training, hyperparameter tuning, and serving into one toolkit.

Overview

Kubeflow packages together the pieces a data science team needs to run ML workflows on Kubernetes - pipeline orchestration, distributed training jobs, hyperparameter tuning, model serving, and notebook environments - so a platform team doesn't have to assemble and integrate each piece separately. It's built specifically to bring ML workloads into the same Kubernetes infrastructure a team already operates.

Kubeflow is a fit for platform teams standardizing ML infrastructure on Kubernetes, a broader end-to-end MLOps platform than more focused tools like MLflow (also in this catalog), which handles experiment tracking and model management specifically.

Categories
AI & Machine Learning
Keywords
mlopskubernetes-nativeml-pipelineshyperparameter-tuning
Languages
Python, Go
License
Apache-2.0

Spotted an error? Suggest an edit on GitHub.