Deploy Your First LLM API on Kubernetes with vLLM

This tutorial walks through deploying the Qwen2.5-1.5B-Instruct model on a Kubernetes GPU node using vLLM as the serving engine. It covers prerequisites like GPU node setup, creating a Deployment with GPU resource requests, and exposing the model as an OpenAI-compatible API endpoint. The goal is to get from a Kubernetes cluster to a working curl request against a real LLM.