Skip to content

Latest commit

 

History

History
182 lines (127 loc) · 3.6 KB

index.rst

File metadata and controls

182 lines (127 loc) · 3.6 KB

Kubeflow SDK

A Pythonic API to Run AI Workloads at Scale

Run any AI workload at any scale — without the need to learn Kubernetes. The Kubeflow SDK provides simple and consistent Python APIs across the Kubeflow ecosystem, enabling you to focus on building AI applications rather than managing infrastructure.

Quick Start

Get up and running in minutes.

pip install kubeflow
from kubeflow.trainer import TrainerClient, CustomTrainer

def train_fn():
    import torch
    model = torch.nn.Linear(10, 1)
    print("Training complete!")

client = TrainerClient()
client.train(trainer=CustomTrainer(func=train_fn, num_nodes=3))

:doc:`Full installation guide → <getting-started/installation>`

Why Kubeflow SDK?

.. grid:: 2
   :gutter: 4

   .. grid-item::

      🎯 **Unified Experience**

      Single SDK to interact with multiple Kubeflow projects through consistent Python APIs.

   .. grid-item::

      🐍 **Simplified AI Workloads**

      Abstract away Kubernetes complexity using familiar Python APIs.

   .. grid-item::

      🚀 **Built for Scale**

      From local laptop to production cluster with thousands of GPUs using the same APIs.

   .. grid-item::

      💻 **Local Development**

      First-class support for local development without a Kubernetes cluster.

Supported Projects

Project Status Description
:doc:`Trainer <train/index>` ✅ Available Train and fine-tune AI models with various frameworks
:doc:`Katib <optimize/index>` ✅ Available Hyperparameter optimization
:doc:`Model Registry <hub/index>` ✅ Available Manage model artifacts and versions
Pipelines 🚧 Planned Build, run, and track AI workflows
Spark Operator 🚧 Planned Data processing and feature engineering
Feast 🚧 Planned Feature store for machine learning

Getting Involved

Join the community and help shape the future of ML on Kubernetes.

.. grid:: 3
   :gutter: 4

   .. grid-item::

      💬 **Community**

      - `Slack <https://kubeflow.slack.com>`_
      - `Discussions <https://github.com/kubeflow/sdk/discussions>`_
      - `Meetings <https://www.kubeflow.org/docs/about/community/>`_

   .. grid-item::

      🤝 **Contribute**

      - `GitHub <https://github.com/kubeflow/sdk>`_
      - `Issues <https://github.com/kubeflow/sdk/issues>`_
      - `Contributing <https://github.com/kubeflow/sdk/blob/main/CONTRIBUTING.md>`_

   .. grid-item::

      📚 **Resources**

      - `Website <https://www.kubeflow.org>`_
      - `Blog Post <https://blog.kubeflow.org/sdk/intro/>`_
      - `DeepWiki <https://deepwiki.com/kubeflow/sdk>`_


.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Getting Started

   getting-started/installation
   getting-started/quickstart
   getting-started/local-development


.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Trainer

   train/index
   train/custom-training
   train/distributed
   train/runtimes
   train/api


.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Optimizer

   optimize/index
   optimize/search-space
   optimize/algorithms
   optimize/api


.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Model Registry

   hub/index
   hub/api


.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Examples

   examples


.. toctree::
   :maxdepth: 2
   :hidden:
   :caption: Contributing

   contributing