Can Launch automatically provision (and spin down) compute resources for me in the target environment?

Whether Launch automatically provisions and scales compute resources depends on the target environment:

Amazon SageMaker and Vertex AI: These platforms provision resources automatically.
Kubernetes: Autoscalers adjust resources based on demand. Weights & Biases solution architects can help you configure your Kubernetes infrastructure to enable retries, autoscaling, and spot instance node pools.

For more information, contact Support or use your shared Slack channel.