Serve the GLM-5.2-NVFP4-REAP-469B model on four NVIDIA RTX PRO 6000 Blackwell GPUs using a pre-configured Docker setup with vLLM support.