Run TAO training on your Kubernetes GPUs.
What it does
Submits and manages NVIDIA TAO training and inference jobs on Kubernetes with GPU scheduling and optional multi-node execution.
When to use it
Use for running TAO on Kubernetes clusters with NVIDIA GPU support or integrating it into an existing Kubernetes-native ML platform.
How to use it
Give it a TAO spec-bundle and accessible data; it submits the workload to your Kubernetes cluster and manages its lifecycle.
What you provide
- Your files
Uses
Good to know
- cannot be undone: Kubernetes
What you need · 17
Requires a reachable cluster through kubeconfig or an in-cluster service account, with permission to create Jobs in the selected namespace.
Requires a kubectl client authenticated to the cluster.
Access · 5
This skill
Kubernetes
Starts GPU container jobs (irreversible)
Creates and deletes cluster resources (irreversible)
NVIDIA GPU Operator
Installs the NVIDIA GPU Operator