docs: add verified vLLM Docker image to Deploy section
#2
by zhiyucheng - opened
Adds the verified Docker image (vllm/vllm-openai:nightly-aarch64, vLLM 0.22.1rc1.dev504) and a GB300 tensor-parallel-4 serve command to the 'Deploy with vLLM' section, mirroring the DeepSeek-V4-Pro-NVFP4 card. Verified serving cleanly (SERVE_OK, coherent output, FLASHINFER_TRTLLM NvFp4 MoE, fp8_ds_mla KV).
zhiyucheng changed pull request status to merged