llm-d-incubation/workload-variant-autoscaler

Variant optimization autoscaler for distributed inference workloads