1000

NIM_MODEL_NAME=meta/Muse-Glimmer

NIM_SERVED_MODEL_NAME=meta/Muse-Glimmer

PRODUCT_NAME=NVIDIA Inference Microservice LLM NIM

--chown=1000:1000 model_manifest_prod.yaml /opt/nim/etc/default/model_manifest.yaml

--chown=1000:1000 VERSION /opt/nim/VERSION

apt-get purge -y -qq patch

cd / &&
  patch -p0 < /tmp/patches/vllm-anthropic-client-error-status.patch &&
  patch -p0 < /tmp/patches/vllm-responses-api-store.patch &&
  patch -p0 < /tmp/patches/vllm-tool-choice-none-exclude-tools.patch &&
  patch -p0 < /tmp/patches/nim-register-management-routes.patch &&
  patch -p0 < /tmp/patches/nim-openapi-drop-unserved-paths.patch &&
  patch -p0 < /tmp/patches/nim-nginx-allow-metadata-manifest.patch &&
  patch -p0 < /tmp/patches/vllm-reasoning-effort-strength.patch &&
  patch -p0 < /tmp/patches/vllm-named-tool-choice-finish-reason.patch &&
  patch -p0 < /tmp/patches/nim-log-level-jsonl-to-vllm.patch &&
  patch -p0 < /tmp/patches/flashinfer-jit-nim-log-level.patch &&
  patch -p0 < /tmp/patches/vllm-onyx-streaming-reasoning.patch &&
  patch -p0 < /tmp/patches/vllm-atem-streaming-tool-calls.patch &&
  patch -p0 < /tmp/patches/vllm-onyx-rope-neox-layout.patch &&
  patch -p0 < /tmp/patches/nim-drop-vram-unavailable-note.patch &&
  rm -rf /tmp/patches

docker/patches/*.patch /tmp/patches/

export DEBIAN_FRONTEND=noninteractive &&
  apt-get update -qq &&
  apt-get install -y -qq --no-install-recommends patch &&
  rm -rf /var/lib/apt/lists/*