redhat-ai-services / ai-accelerator

The AI Accelerator is a template project for setting up Red Hat OpenShift AI using GitOps
28 stars 59 forks source link

Fix autoscaling timeout on vllm #42

Closed strangiato closed 2 months ago

strangiato commented 2 months ago

Resolves #22

Adding the progress-deadline annotation to the vllm deployment to allow the node to autoscale successfully and deploy the vLLM instance before the default timeout of 10m expires.