Use

The following guides cover common tasks and use cases for using Charmed Kubeflow (CKF).

Set up scheduling depending on your use case:

• Configure advanced scheduling

Learn how to customize the link configuration of the Kubeflow Central Dashboard:

• Customise link configuration

Use PodDefaults to configure Kubernetes Pods:

• Leverage PodDefaults

Leverage NVIDIA with CKF:

• Use NVIDIA GPUs
• Deploy NVIDIA NIMs
• Launch NVIDIA NGC notebooks
• Serve a model using Triton Inference Server

Deploy KServe, Knative, and Istio charms on their own using the Autoscaling model serving:

• Deploy Autoscaling model serving

Find out how to perform inference using programmatic access tokens:

• Perform inference on ISVCs using access tokens

Last updated 3 days ago.