fix(vllm-router): allow using prefill-decode for subset of models (by checking labels) and add a fallback routing strategy#3

nejch merged 1 commit into

deploysiemens/production-stack:deployfrom

fix/allow-using-prefill-decode-for-subset-of-models-by-checking-labels-in-vllm-routersiemens/production-stack:fix/allow-using-prefill-decode-for-subset-of-models-by-checking-labels-in-vllm-routerCopy head branch name to clipboard