logo
docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608
linux/amd64 docker.io

该镜像基于vllm框架构建,用于提供大语言模型的高效推理与服务能力,支持高吞吐量、低延迟的请求处理,可助力快速部署和运行大型语言模型相关应用。

16
浏览次数
22.92GB
镜像大小
国内镜像
swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608
源镜像
docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608
镜像ID
sha256:da1c3e883628cf4f5fcd507e8e906851f744820259393d4d1b4e13919e37f326
镜像 TAG
black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608
镜像大小
22.92GB
平台架构
linux/amd64
镜像源
docker.io
CMD
/usr/local/bin/run-kimi26-vllm
启动入口
/opt/nvidia/nvidia_entrypoint.sh
工作目录
/
OS/平台
linux/amd64
镜像创建
2026-06-08T23:40:48.217360371Z
同步时间
2026-08-30 01:03
浏览量
16 次
贡献者
⚙️ 环境变量 68
KeyValue
PATH=/opt/venv/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin 0
NVARCH=x86_64 1
NVIDIA_REQUIRE_CUDA=cuda>=13.2 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=vapps,driver>=535,driver<536 brand=vpc,driver>=535,driver<536 brand=vcs,driver>=535,driver<536 brand=vws,driver>=535,driver<536 brand=cloudgaming,driver>=535,driver<536 brand=unknown,driver>=570,driver<571 brand=grid,driver>=570,driver<571 brand=tesla,driver>=570,driver<571 brand=nvidia,driver>=570,driver<571 brand=quadro,driver>=570,driver<571 brand=quadrortx,driver>=570,driver<571 brand=nvidiartx,driver>=570,driver<571 brand=vapps,driver>=570,driver<571 brand=vpc,driver>=570,driver<571 brand=vcs,driver>=570,driver<571 brand=vws,driver>=570,driver<571 brand=cloudgaming,driver>=570,driver<571 brand=unknown,driver>=580,driver<581 brand=grid,driver>=580,driver<581 brand=tesla,driver>=580,driver<581 brand=nvidia,driver>=580,driver<581 brand=quadro,driver>=580,driver<581 brand=quadrortx,driver>=580,driver<581 brand=nvidiartx,driver>=580,driver<581 brand=vapps,driver>=580,driver<581 brand=vpc,driver>=580,driver<581 brand=vcs,driver>=580,driver<581 brand=vws,driver>=580,driver<581 brand=cloudgaming,driver>=580,driver<581 brand=unknown,driver>=590,driver<591 brand=grid,driver>=590,driver<591 brand=tesla,driver>=590,driver<591 brand=nvidia,driver>=590,driver<591 brand=quadro,driver>=590,driver<591 brand=quadrortx,driver>=590,driver<591 brand=nvidiartx,driver>=590,driver<591 brand=vapps,driver>=590,driver<591 brand=vpc,driver>=590,driver<591 brand=vcs,driver>=590,driver<591 brand=vws,driver>=590,driver<591 brand=cloudgaming,driver>=590,driver<591 2
NV_CUDA_CUDART_VERSION=13.2.75-1 3
CUDA_VERSION=13.2.1 4
LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64 5
NVIDIA_VISIBLE_DEVICES=all 6
NVIDIA_DRIVER_CAPABILITIES=compute,utility 7
NV_CUDA_LIB_VERSION=13.2.1-1 8
NV_NVTX_VERSION=13.2.75-1 9
NV_LIBNPP_VERSION=13.1.0.48-1 10
NV_LIBNPP_PACKAGE=libnpp-13-2=13.1.0.48-1 11
NV_LIBCUSPARSE_VERSION=12.7.10.1-1 12
NV_LIBCUBLAS_PACKAGE_NAME=libcublas-13-2 13
NV_LIBCUBLAS_VERSION=13.4.0.1-1 14
NV_LIBCUBLAS_PACKAGE=libcublas-13-2=13.4.0.1-1 15
NVIDIA_PRODUCT_NAME=CUDA 16
NV_CUDA_CUDART_DEV_VERSION=13.2.75-1 17
NV_NVML_DEV_VERSION=13.2.82-1 18
NV_LIBCUSPARSE_DEV_VERSION=12.7.10.1-1 19
NV_LIBNPP_DEV_VERSION=13.1.0.48-1 20
NV_LIBNPP_DEV_PACKAGE=libnpp-dev-13-2=13.1.0.48-1 21
NV_LIBCUBLAS_DEV_VERSION=13.4.0.1-1 22
NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-13-2 23
NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-13-2=13.4.0.1-1 24
NV_CUDA_NSIGHT_COMPUTE_VERSION=13.2.1-1 25
NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-13-2=13.2.1-1 26
LIBRARY_PATH=/usr/local/cuda/lib64/stubs 27
NV_CUDNN_VERSION=9.20.0.48-1 28
NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-13 29
NV_CUDNN_PACKAGE=libcudnn9-cuda-13=9.20.0.48-1 30
NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-13=9.20.0.48-1 31
NV_CUDNN_PACKAGE_DEV_HEADERS=libcudnn9-headers-cuda-13=9.20.0.48-1 32
LANG=C.UTF-8 33
LC_ALL=C.UTF-8 34
PYTHONDONTWRITEBYTECODE=1 35
PYTHONUNBUFFERED=1 36
PIP_DISABLE_PIP_VERSION_CHECK=1 37
NCCL_LOCAL_INFERENCE_PATH=/opt/libnccl-local-inference.so.2.30.4 38
NCCL_PR2127_PATH=/opt/libnccl-local-inference.so.2.30.4 39
VLLM_NCCL_SO_PATH=/opt/libnccl-local-inference.so.2.30.4 40
NCCL_GRAPH_FILE= 41
VIRTUAL_ENV=/opt/venv 42
CUDA_DEVICE_ORDER=PCI_BUS_ID 43
TORCH_CUDA_ARCH_LIST=12.0a 44
FLASHINFER_CUDA_ARCH_LIST=12.0f 45
FLASHINFER_LOCAL_VERSION=cu132 46
FLASHINFER_DISABLE_VERSION_CHECK=1 47
HF_HOME=/root/.cache/huggingface 48
TRITON_CACHE_DIR=/cache/jit/triton 49
TORCH_EXTENSIONS_DIR=/cache/jit/torch_extensions 50
VLLM_CACHE_DIR=/cache/jit/vllm 51
FLASHINFER_WORKSPACE_BASE=/cache/jit/flashinfer 52
TVM_FFI_CACHE_DIR=/cache/jit/tvm-ffi 53
XDG_CACHE_HOME=/cache/jit 54
LD_PRELOAD=/opt/libnccl-local-inference.so.2.30.4 55
NCCL_IB_DISABLE=1 56
NCCL_P2P_LEVEL=SYS 57
NCCL_PROTO=LL,LL128,Simple 58
VLLM_ENABLE_PCIE_ALLREDUCE=1 59
VLLM_PCIE_ALLREDUCE_BACKEND=cpp 60
VLLM_CPP_AR_1STAGE_NCCL_CUTOFF=56KB 61
VLLM_CPP_AR_IGNORE_CUTOFF_MAX_ROWS=0 62
VLLM_RTX6K_FUSED_ALLREDUCE_ADD=0 63
VLLM_RTX6K_FUSED_ALLREDUCE_ADD_END_BARRIER=0 64
VLLM_DISABLE_SHARED_EXPERTS_STREAM=0 65
VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=1 66
VLLM_DISABLED_KERNELS=MarlinFP8ScaledMMLinearKernel 67
🏷️ 镜像标签 22
KeyValue
9.20.0.48-1 com.nvidia.cudnn.version
refs/pull/11/head local-inference.b12x.branch
d90d89c8353adabb56cc84bd3924ef811ef8d877 local-inference.b12x.commit
source local-inference.b12x.runtime_version
13.4.1.2-1 local-inference.cublas.version
595.71.05-1ubuntu1 local-inference.cuda.compat.version
13.2.1 local-inference.cuda.version
9.22.0.52-1 local-inference.cudnn.version
refs/pull/324/head local-inference.deepgemm.branch
aced12c2c8882a945c568ace9d4a7e5778aae410 local-inference.deepgemm.commit
refs/pull/3395/head local-inference.flashinfer.branch
b41aa8dd2fb93c49b1c6134bd1953040f8089d51 local-inference.flashinfer.commit
canonical/cu132-nccl2304-amd-noxml local-inference.nccl.branch
local-inference-lab/nccl-canonical local-inference.nccl.repo
2.30.4 local-inference.nccl.version
2.12.0+cu132 local-inference.torch.version
dev/black-benediction local-inference.vllm.branch
bb6c5b7351fceb9d524e0d43b957415ffefcb981 local-inference.vllm.commit
https://github.com/local-inference-lab/vllm.git local-inference.vllm.repo
NVIDIA CORPORATION <cudatools@nvidia.com> maintainer
Clean CUDA 13.2.1 vLLM GLM/Kimi image with PyTorch 2.12 cu132, patched NCCL 2.30.4, FlashInfer git, B12X and canonical vLLM branch org.opencontainers.image.description
24.04 org.opencontainers.image.version
🛡️ 镜像安全扫描
ubuntu 24.04 Trivy 2026-08-30 01:07 查看完整报告
438
低危 LOW
3240
中危 MEDIUM
403
高危 HIGH
26
严重 CRITICAL
受影响目标 (4)
docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608 (ubuntu 24.04) ubuntu Python python-pkg opt/nvidia/nsight-compute/2026.1.1/host/target-linux-x64/plugins/efa_metrics/nic_sampler gobinary opt/vllm/rust/Cargo.lock cargo

Docker拉取命令

docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608
docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608  docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608

Containerd拉取命令

ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608
ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608  docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608

Shell快速替换命令

sed -i 's#voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608#swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608#' deployment.yaml

Ansible快速分发-Docker

#ansible k8s -m shell -a 'docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608 && docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608  docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608'

Ansible快速分发-Containerd

#ansible k8s -m shell -a 'ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608 && ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608  docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608'

镜像构建历史


# 2026-06-09 07:40:48  0.00B 添加元数据标签
LABEL org.opencontainers.image.description=Clean CUDA 13.2.1 vLLM GLM/Kimi image with PyTorch 2.12 cu132, patched NCCL 2.30.4, FlashInfer git, B12X and canonical vLLM branch local-inference.cuda.version=13.2.1 local-inference.cuda.compat.version=595.71.05-1ubuntu1 local-inference.cublas.version=13.4.1.2-1 local-inference.cudnn.version=9.22.0.52-1 local-inference.torch.version=2.12.0+cu132 local-inference.nccl.repo=local-inference-lab/nccl-canonical local-inference.nccl.branch=canonical/cu132-nccl2304-amd-noxml local-inference.nccl.version=2.30.4 local-inference.vllm.repo=https://github.com/local-inference-lab/vllm.git local-inference.vllm.branch=dev/black-benediction local-inference.vllm.commit=bb6c5b7351fceb9d524e0d43b957415ffefcb981 local-inference.b12x.branch=refs/pull/11/head local-inference.b12x.commit=d90d89c8353adabb56cc84bd3924ef811ef8d877 local-inference.b12x.runtime_version=source local-inference.flashinfer.branch=refs/pull/3395/head local-inference.flashinfer.commit=b41aa8dd2fb93c49b1c6134bd1953040f8089d51 local-inference.deepgemm.branch=refs/pull/324/head local-inference.deepgemm.commit=aced12c2c8882a945c568ace9d4a7e5778aae410
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG CUDA_COMPAT_13_2_VERSION=595.71.05-1ubuntu1
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG CUDNN_CUDA13_VERSION=9.22.0.52-1
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG CUBLAS_CUDA13_VERSION=13.4.1.2-1
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG DEEPGEMM_COMMIT=aced12c2c8882a945c568ace9d4a7e5778aae410
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG DEEPGEMM_REF=refs/pull/324/head
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG FLASHINFER_COMMIT=b41aa8dd2fb93c49b1c6134bd1953040f8089d51
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG FLASHINFER_REF=refs/pull/3395/head
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG B12X_COMMIT=d90d89c8353adabb56cc84bd3924ef811ef8d877
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG B12X_REF=refs/pull/11/head
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG VLLM_COMMIT=bb6c5b7351fceb9d524e0d43b957415ffefcb981
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG VLLM_REF=dev/black-benediction
                        
# 2026-06-09 07:40:48  0.00B 定义构建参数
ARG VLLM_REPO=https://github.com/local-inference-lab/vllm.git
                        
# 2026-06-09 07:40:48  0.00B 设置默认要执行的命令
CMD ["/usr/local/bin/run-kimi26-vllm"]
                        
# 2026-06-09 07:40:48  0.00B 设置工作目录为/
WORKDIR /
                        
# 2026-06-09 07:40:48  0.00B 执行命令并创建新的镜像层
RUN /bin/bash -euxo pipefail -c mkdir -p /cache/jit/triton /cache/jit/torch_extensions /cache/jit/vllm /cache/jit/flashinfer  && strings /opt/libnccl-local-inference.so.2.30.4 | grep -F "NCCL version 2.30.4 compiled with CUDA 13.2"  && /usr/local/bin/verify-glm-kimi-cu132 /opt/vllm  && python - <<'PY'
import torch, flashinfer, vllm
import deep_gemm
print("torch", torch.__version__, torch.version.cuda)
print("torch nccl", torch.cuda.nccl.version())
print("flashinfer", getattr(flashinfer, "__version__", "unknown"))
print("deep_gemm", getattr(deep_gemm, "__version__", "unknown"))
print("vllm", getattr(vllm, "__version__", "unknown"))
assert torch.__version__.startswith("2.12.0+cu132")
assert torch.version.cuda == "13.2"
PY # buildkit
                        
# 2026-06-09 07:40:34  0.00B 设置环境变量 VIRTUAL_ENV PATH CUDA_DEVICE_ORDER TORCH_CUDA_ARCH_LIST FLASHINFER_CUDA_ARCH_LIST FLASHINFER_LOCAL_VERSION FLASHINFER_DISABLE_VERSION_CHECK HF_HOME TRITON_CACHE_DIR TORCH_EXTENSIONS_DIR VLLM_CACHE_DIR FLASHINFER_WORKSPACE_BASE TVM_FFI_CACHE_DIR XDG_CACHE_HOME NCCL_LOCAL_INFERENCE_PATH NCCL_PR2127_PATH VLLM_NCCL_SO_PATH LD_PRELOAD NCCL_IB_DISABLE NCCL_P2P_LEVEL NCCL_PROTO VLLM_ENABLE_PCIE_ALLREDUCE VLLM_PCIE_ALLREDUCE_BACKEND VLLM_CPP_AR_1STAGE_NCCL_CUTOFF VLLM_CPP_AR_IGNORE_CUTOFF_MAX_ROWS VLLM_RTX6K_FUSED_ALLREDUCE_ADD VLLM_RTX6K_FUSED_ALLREDUCE_ADD_END_BARRIER VLLM_DISABLE_SHARED_EXPERTS_STREAM VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS VLLM_DISABLED_KERNELS
ENV VIRTUAL_ENV=/opt/venv PATH=/opt/venv/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin CUDA_DEVICE_ORDER=PCI_BUS_ID TORCH_CUDA_ARCH_LIST=12.0a FLASHINFER_CUDA_ARCH_LIST=12.0f FLASHINFER_LOCAL_VERSION=cu132 FLASHINFER_DISABLE_VERSION_CHECK=1 HF_HOME=/root/.cache/huggingface TRITON_CACHE_DIR=/cache/jit/triton TORCH_EXTENSIONS_DIR=/cache/jit/torch_extensions VLLM_CACHE_DIR=/cache/jit/vllm FLASHINFER_WORKSPACE_BASE=/cache/jit/flashinfer TVM_FFI_CACHE_DIR=/cache/jit/tvm-ffi XDG_CACHE_HOME=/cache/jit NCCL_LOCAL_INFERENCE_PATH=/opt/libnccl-local-inference.so.2.30.4 NCCL_PR2127_PATH=/opt/libnccl-local-inference.so.2.30.4 VLLM_NCCL_SO_PATH=/opt/libnccl-local-inference.so.2.30.4 LD_PRELOAD=/opt/libnccl-local-inference.so.2.30.4 NCCL_IB_DISABLE=1 NCCL_P2P_LEVEL=SYS NCCL_PROTO=LL,LL128,Simple VLLM_ENABLE_PCIE_ALLREDUCE=1 VLLM_PCIE_ALLREDUCE_BACKEND=cpp VLLM_CPP_AR_1STAGE_NCCL_CUTOFF=56KB VLLM_CPP_AR_IGNORE_CUTOFF_MAX_ROWS=0 VLLM_RTX6K_FUSED_ALLREDUCE_ADD=0 VLLM_RTX6K_FUSED_ALLREDUCE_ADD_END_BARRIER=0 VLLM_DISABLE_SHARED_EXPERTS_STREAM=0 VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=1 VLLM_DISABLED_KERNELS=MarlinFP8ScaledMMLinearKernel
                        
# 2026-06-09 07:40:34  31.77KB 执行命令并创建新的镜像层
RUN /bin/bash -euxo pipefail -c cublas="$(readlink -f /usr/lib/x86_64-linux-gnu/libcublas/13/libcublas.so.13)"  && cublas_lt="$(readlink -f /usr/lib/x86_64-linux-gnu/libcublas/13/libcublasLt.so.13)"  && cublas_name="$(basename "${cublas}")"  && cublas_lt_name="$(basename "${cublas_lt}")"  && test -f "${cublas}"  && test -f "${cublas_lt}"  && for cuda_libdir in /usr/local/cuda/targets/x86_64-linux/lib /usr/local/cuda-13.2/targets/x86_64-linux/lib; do       if [[ -d "${cuda_libdir}" ]]; then         cuda_libdir="$(readlink -f "${cuda_libdir}")";         rm -f "${cuda_libdir}"/libcublas.so "${cuda_libdir}"/libcublas.so.13 "${cuda_libdir}"/libcublas.so.13.*;         rm -f "${cuda_libdir}"/libcublasLt.so "${cuda_libdir}"/libcublasLt.so.13 "${cuda_libdir}"/libcublasLt.so.13.*;         ln -s "${cublas}" "${cuda_libdir}/${cublas_name}";         ln -s "${cublas_name}" "${cuda_libdir}/libcublas.so.13";         ln -s libcublas.so.13 "${cuda_libdir}/libcublas.so";         ln -s "${cublas_lt}" "${cuda_libdir}/${cublas_lt_name}";         ln -s "${cublas_lt_name}" "${cuda_libdir}/libcublasLt.so.13";         ln -s libcublasLt.so.13 "${cuda_libdir}/libcublasLt.so";       fi;     done  && ldconfig # buildkit
                        
# 2026-06-09 07:40:34  1.22GB 执行命令并创建新的镜像层
RUN /bin/bash -euxo pipefail -c printf '%s\n'       'torch==2.12.0+cu132'       'torchvision==0.27.0+cu132'       'cuda-bindings==13.2.0'       'cuda-python==13.2.0'       'nvidia-cuda-cccl==13.2.75'       'nvidia-cuda-crt==13.2.78'       'nvidia-cuda-nvcc==13.2.78'       'nvidia-nvjitlink==13.2.78'       'nvidia-nvvm==13.2.78'       > /tmp/torch-cu132-constraints.txt  && /opt/venv/bin/python -m pip install --upgrade       --index-url https://download.pytorch.org/whl/cu132       --extra-index-url https://pypi.org/simple       --constraint /tmp/torch-cu132-constraints.txt       "cuda-bindings==13.2.0"       "cuda-python==13.2.0"       "numba==0.65.0"       "quack-kernels>=0.3.3"       "humming-kernels[cu13]==0.1.4"       "nvidia-nvjitlink==13.2.78"       "tilelang==0.1.9"       "tokenspeed-mla==0.1.2"       "transformers>=5.5.1"       "apache-tvm-ffi==0.1.9"       "fastsafetensors>=0.3.2"  && /opt/venv/bin/python -m pip check # buildkit
                        
# 2026-06-09 07:40:20  813.00B 复制新文件或目录到容器中
COPY /usr/local/bin/verify-glm-kimi-cu132 /usr/local/bin/verify-glm-kimi-cu132 # buildkit
                        
# 2026-06-09 07:40:20  17.51KB 复制新文件或目录到容器中
COPY /usr/local/bin/serve-*.sh /usr/local/bin/ # buildkit
                        
# 2026-06-09 07:40:20  86.00B 复制新文件或目录到容器中
COPY /usr/local/bin/run-kimi26-vllm /usr/local/bin/run-kimi26-vllm # buildkit
                        
# 2026-06-09 07:40:20  7.42KB 复制新文件或目录到容器中
COPY /usr/local/bin/run-glm51-vllm /usr/local/bin/run-glm51-vllm # buildkit
                        
# 2026-06-09 07:40:20  42.36MB 复制新文件或目录到容器中
COPY /opt/vllm /opt/vllm # buildkit
                        
# 2026-06-09 07:40:20  9.48GB 复制新文件或目录到容器中
COPY /opt/venv /opt/venv # buildkit
                        
# 2026-06-09 07:40:20  0.00B 
SHELL [/bin/bash -euxo pipefail -c]
                        
# 2026-06-09 06:15:47  0.00B 添加元数据标签
LABEL org.opencontainers.image.description=Reusable CUDA 13.2.1 GLM/Kimi system base with latest CUDA 13.2-compatible cuBLAS/cuDNN overlays and patched NCCL 2.30.4 local-inference.cuda.version=13.2.1 local-inference.cuda.compat.version=595.71.05-1ubuntu1 local-inference.cublas.version=13.4.1.2-1 local-inference.cudnn.version=9.22.0.52-1 local-inference.nccl.repo=local-inference-lab/nccl-canonical local-inference.nccl.branch=canonical/cu132-nccl2304-amd-noxml local-inference.nccl.version=2.30.4
                        
# 2026-06-09 06:15:47  0.00B 设置环境变量 NCCL_LOCAL_INFERENCE_PATH NCCL_PR2127_PATH VLLM_NCCL_SO_PATH NCCL_GRAPH_FILE
ENV NCCL_LOCAL_INFERENCE_PATH=/opt/libnccl-local-inference.so.2.30.4 NCCL_PR2127_PATH=/opt/libnccl-local-inference.so.2.30.4 VLLM_NCCL_SO_PATH=/opt/libnccl-local-inference.so.2.30.4 NCCL_GRAPH_FILE=
                        
# 2026-06-09 06:15:47  224.80MB 复制新文件或目录到容器中
COPY /build/nccl/build/lib/libnccl.so.2.30.4 /opt/libnccl-local-inference.so.2.30.4 # buildkit
                        
# 2026-06-09 06:15:47  10.20KB 执行命令并创建新的镜像层
RUN |6 PYTHON_VERSION=3.12 NCCL_SONAME=libnccl.so.2.30.4 CUBLAS_CUDA13_VERSION=13.4.1.2-1 CUDNN_CUDA13_VERSION=9.22.0.52-1 CUDA_COMPAT_13_2_VERSION=595.71.05-1ubuntu1 DEBIAN_FRONTEND=noninteractive /bin/bash -euxo pipefail -c update-alternatives --install /usr/bin/python3 python3 /usr/bin/python${PYTHON_VERSION} 1  && update-alternatives --install /usr/bin/python python /usr/bin/python${PYTHON_VERSION} 1 # buildkit
                        
# 2026-06-09 06:15:47  31.77KB 执行命令并创建新的镜像层
RUN |6 PYTHON_VERSION=3.12 NCCL_SONAME=libnccl.so.2.30.4 CUBLAS_CUDA13_VERSION=13.4.1.2-1 CUDNN_CUDA13_VERSION=9.22.0.52-1 CUDA_COMPAT_13_2_VERSION=595.71.05-1ubuntu1 DEBIAN_FRONTEND=noninteractive /bin/bash -euxo pipefail -c cublas="$(readlink -f /usr/lib/x86_64-linux-gnu/libcublas/13/libcublas.so.13)"  && cublas_lt="$(readlink -f /usr/lib/x86_64-linux-gnu/libcublas/13/libcublasLt.so.13)"  && cublas_name="$(basename "${cublas}")"  && cublas_lt_name="$(basename "${cublas_lt}")"  && test -f "${cublas}"  && test -f "${cublas_lt}"  && for cuda_libdir in /usr/local/cuda/targets/x86_64-linux/lib /usr/local/cuda-${CUDA_VERSION}/targets/x86_64-linux/lib; do       if [[ -d "${cuda_libdir}" ]]; then         cuda_libdir="$(readlink -f "${cuda_libdir}")";         rm -f "${cuda_libdir}"/libcublas.so "${cuda_libdir}"/libcublas.so.13 "${cuda_libdir}"/libcublas.so.13.*;         rm -f "${cuda_libdir}"/libcublasLt.so "${cuda_libdir}"/libcublasLt.so.13 "${cuda_libdir}"/libcublasLt.so.13.*;         ln -s "${cublas}" "${cuda_libdir}/${cublas_name}";         ln -s "${cublas_name}" "${cuda_libdir}/libcublas.so.13";         ln -s libcublas.so.13 "${cuda_libdir}/libcublas.so";         ln -s "${cublas_lt}" "${cuda_libdir}/${cublas_lt_name}";         ln -s "${cublas_lt_name}" "${cuda_libdir}/libcublasLt.so.13";         ln -s libcublasLt.so.13 "${cuda_libdir}/libcublasLt.so";       fi;     done  && ldconfig # buildkit
                        
# 2026-06-09 06:15:47  4.38GB 执行命令并创建新的镜像层
RUN |6 PYTHON_VERSION=3.12 NCCL_SONAME=libnccl.so.2.30.4 CUBLAS_CUDA13_VERSION=13.4.1.2-1 CUDNN_CUDA13_VERSION=9.22.0.52-1 CUDA_COMPAT_13_2_VERSION=595.71.05-1ubuntu1 DEBIAN_FRONTEND=noninteractive /bin/bash -euxo pipefail -c for attempt in {1..5}; do       if apt-get update; then break; fi;       if [[ "${attempt}" == "5" ]]; then exit 1; fi;       rm -rf /var/lib/apt/lists/*;       sleep $((attempt * 10));     done     && apt-get install -y --allow-change-held-packages --no-install-recommends     python${PYTHON_VERSION}     python${PYTHON_VERSION}-dev     python${PYTHON_VERSION}-venv     python3-pip     git     git-lfs     wget     curl     ca-certificates     build-essential     cmake     ninja-build     ccache     pkg-config     patchelf     libssl-dev     libffi-dev     libjpeg-dev     libpng-dev     libgomp1     numactl     libnuma-dev     libibverbs-dev     pciutils     cuda-compat-13-2=${CUDA_COMPAT_13_2_VERSION}     libcublas13-cuda-13=${CUBLAS_CUDA13_VERSION}     libcublas13-dev-cuda-13=${CUBLAS_CUDA13_VERSION}     cublas-cuda-13=${CUBLAS_CUDA13_VERSION}     libcudnn9-cuda-13=${CUDNN_CUDA13_VERSION}     libcudnn9-dev-cuda-13=${CUDNN_CUDA13_VERSION}     libcudnn9-headers-cuda-13=${CUDNN_CUDA13_VERSION}     libcudnn9-static-cuda-13=${CUDNN_CUDA13_VERSION}     && ldconfig     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-06-09 06:15:47  0.00B 设置环境变量 LANG LC_ALL PYTHONDONTWRITEBYTECODE PYTHONUNBUFFERED PIP_DISABLE_PIP_VERSION_CHECK
ENV LANG=C.UTF-8 LC_ALL=C.UTF-8 PYTHONDONTWRITEBYTECODE=1 PYTHONUNBUFFERED=1 PIP_DISABLE_PIP_VERSION_CHECK=1
                        
# 2026-06-09 06:15:47  0.00B 
SHELL [/bin/bash -euxo pipefail -c]
                        
# 2026-06-09 06:15:47  0.00B 定义构建参数
ARG DEBIAN_FRONTEND=noninteractive
                        
# 2026-06-09 06:15:47  0.00B 定义构建参数
ARG CUDA_COMPAT_13_2_VERSION=595.71.05-1ubuntu1
                        
# 2026-06-09 06:15:47  0.00B 定义构建参数
ARG CUDNN_CUDA13_VERSION=9.22.0.52-1
                        
# 2026-06-09 06:15:47  0.00B 定义构建参数
ARG CUBLAS_CUDA13_VERSION=13.4.1.2-1
                        
# 2026-06-09 06:15:47  0.00B 定义构建参数
ARG NCCL_SONAME=libnccl.so.2.30.4
                        
# 2026-06-09 06:15:47  0.00B 定义构建参数
ARG PYTHON_VERSION=3.12
                        
# 2026-04-18 00:24:02  485.98MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     ${NV_CUDNN_PACKAGE}     ${NV_CUDNN_PACKAGE_DEV}     ${NV_CUDNN_PACKAGE_DEV_HEADERS}     && apt-mark hold ${NV_CUDNN_PACKAGE_NAME}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-04-18 00:24:02  0.00B 添加元数据标签
LABEL com.nvidia.cudnn.version=9.20.0.48-1
                        
# 2026-04-18 00:24:02  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2026-04-18 00:24:02  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2026-04-18 00:24:02  0.00B 设置环境变量 NV_CUDNN_PACKAGE_DEV_HEADERS
ENV NV_CUDNN_PACKAGE_DEV_HEADERS=libcudnn9-headers-cuda-13=9.20.0.48-1
                        
# 2026-04-18 00:24:02  0.00B 设置环境变量 NV_CUDNN_PACKAGE_DEV
ENV NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-13=9.20.0.48-1
                        
# 2026-04-18 00:24:02  0.00B 设置环境变量 NV_CUDNN_PACKAGE
ENV NV_CUDNN_PACKAGE=libcudnn9-cuda-13=9.20.0.48-1
                        
# 2026-04-18 00:24:02  0.00B 设置环境变量 NV_CUDNN_PACKAGE_NAME
ENV NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-13
                        
# 2026-04-18 00:24:02  0.00B 设置环境变量 NV_CUDNN_VERSION
ENV NV_CUDNN_VERSION=9.20.0.48-1
                        
# 2026-04-17 23:47:00  0.00B 设置环境变量 LIBRARY_PATH
ENV LIBRARY_PATH=/usr/local/cuda/lib64/stubs
                        
# 2026-04-17 23:47:00  393.98KB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_DEV_PACKAGE_NAME} # buildkit
                        
# 2026-04-17 23:46:59  62.00B 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c NSYS_PATH=$(find /opt/nvidia/nsight-compute -type f -name "nsys" 2>/dev/null | head -n 1);     if [ -n "$NSYS_PATH" ] && [ ! -f /usr/local/bin/nsys ]; then         ln -s "$NSYS_PATH" /usr/local/bin/nsys;     fi # buildkit
                        
# 2026-04-17 23:46:59  4.74GB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-cudart-dev-13-2=${NV_CUDA_CUDART_DEV_VERSION}     cuda-command-line-tools-13-2=${NV_CUDA_LIB_VERSION}     cuda-minimal-build-13-2=${NV_CUDA_LIB_VERSION}     cuda-libraries-dev-13-2=${NV_CUDA_LIB_VERSION}     cuda-nvml-dev-13-2=${NV_NVML_DEV_VERSION}     ${NV_LIBNPP_DEV_PACKAGE}     libcusparse-dev-13-2=${NV_LIBCUSPARSE_DEV_VERSION}     ${NV_LIBCUBLAS_DEV_PACKAGE}     ${NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-04-17 23:46:59  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2026-04-17 23:46:59  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE
ENV NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-13-2=13.2.1-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_CUDA_NSIGHT_COMPUTE_VERSION
ENV NV_CUDA_NSIGHT_COMPUTE_VERSION=13.2.1-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_LIBCUBLAS_DEV_PACKAGE
ENV NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-13-2=13.4.0.1-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_LIBCUBLAS_DEV_PACKAGE_NAME
ENV NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-13-2
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_LIBCUBLAS_DEV_VERSION
ENV NV_LIBCUBLAS_DEV_VERSION=13.4.0.1-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_LIBNPP_DEV_PACKAGE
ENV NV_LIBNPP_DEV_PACKAGE=libnpp-dev-13-2=13.1.0.48-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_LIBNPP_DEV_VERSION
ENV NV_LIBNPP_DEV_VERSION=13.1.0.48-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_LIBCUSPARSE_DEV_VERSION
ENV NV_LIBCUSPARSE_DEV_VERSION=12.7.10.1-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_NVML_DEV_VERSION
ENV NV_NVML_DEV_VERSION=13.2.82-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_CUDA_CUDART_DEV_VERSION
ENV NV_CUDA_CUDART_DEV_VERSION=13.2.75-1
                        
# 2026-04-17 23:46:59  0.00B 设置环境变量 NV_CUDA_LIB_VERSION
ENV NV_CUDA_LIB_VERSION=13.2.1-1
                        
# 2026-04-17 23:37:53  0.00B 配置容器启动时运行的命令
ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NVIDIA_PRODUCT_NAME
ENV NVIDIA_PRODUCT_NAME=CUDA
                        
# 2026-04-17 23:37:53  2.53KB 复制新文件或目录到容器中
COPY nvidia_entrypoint.sh /opt/nvidia/ # buildkit
                        
# 2026-04-17 23:37:53  3.06KB 复制新文件或目录到容器中
COPY entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit
                        
# 2026-04-17 23:37:53  241.12KB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_PACKAGE_NAME} # buildkit
                        
# 2026-04-17 23:37:53  1.83GB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-libraries-13-2=${NV_CUDA_LIB_VERSION}     ${NV_LIBNPP_PACKAGE}     cuda-nvtx-13-2=${NV_NVTX_VERSION}     libcusparse-13-2=${NV_LIBCUSPARSE_VERSION}     ${NV_LIBCUBLAS_PACKAGE}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-04-17 23:37:53  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2026-04-17 23:37:53  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_LIBCUBLAS_PACKAGE
ENV NV_LIBCUBLAS_PACKAGE=libcublas-13-2=13.4.0.1-1
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_LIBCUBLAS_VERSION
ENV NV_LIBCUBLAS_VERSION=13.4.0.1-1
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_LIBCUBLAS_PACKAGE_NAME
ENV NV_LIBCUBLAS_PACKAGE_NAME=libcublas-13-2
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_LIBCUSPARSE_VERSION
ENV NV_LIBCUSPARSE_VERSION=12.7.10.1-1
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_LIBNPP_PACKAGE
ENV NV_LIBNPP_PACKAGE=libnpp-13-2=13.1.0.48-1
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_LIBNPP_VERSION
ENV NV_LIBNPP_VERSION=13.1.0.48-1
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_NVTX_VERSION
ENV NV_NVTX_VERSION=13.2.75-1
                        
# 2026-04-17 23:37:53  0.00B 设置环境变量 NV_CUDA_LIB_VERSION
ENV NV_CUDA_LIB_VERSION=13.2.1-1
                        
# 2026-04-17 23:17:50  0.00B 设置环境变量 NVIDIA_DRIVER_CAPABILITIES
ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility
                        
# 2026-04-17 23:17:50  0.00B 设置环境变量 NVIDIA_VISIBLE_DEVICES
ENV NVIDIA_VISIBLE_DEVICES=all
                        
# 2026-04-17 23:17:50  17.29KB 复制新文件或目录到容器中
COPY NGC-DL-CONTAINER-LICENSE / # buildkit
                        
# 2026-04-17 23:17:50  0.00B 设置环境变量 LD_LIBRARY_PATH
ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64
                        
# 2026-04-17 23:17:50  0.00B 设置环境变量 PATH
ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
                        
# 2026-04-17 23:17:50  22.00B 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c echo "/usr/local/cuda/lib64" >> /etc/ld.so.conf.d/nvidia.conf # buildkit
                        
# 2026-04-17 23:17:50  0.00B 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update &&     if apt-cache policy cuda-compat-orin-13-2 2>/dev/null | grep -q "Candidate:"; then         apt-get install -y --no-install-recommends cuda-compat-orin-13-2;     fi &&     rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-04-17 23:17:44  430.90MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update &&     if apt-cache policy cuda-compat-13-2 2>/dev/null | grep -q "Candidate:"; then         apt-get install -y --no-install-recommends cuda-compat-13-2;     fi &&     rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-04-17 23:17:22  1.54MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-cudart-13-2=${NV_CUDA_CUDART_VERSION}     cuda-toolkit-13-2-config-common=${NV_CUDA_CUDART_VERSION}     cuda-toolkit-13-config-common=${NV_CUDA_CUDART_VERSION}     cuda-toolkit-config-common=${NV_CUDA_CUDART_VERSION}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-04-17 23:17:05  0.00B 设置环境变量 CUDA_VERSION
ENV CUDA_VERSION=13.2.1
                        
# 2026-04-17 23:17:05  10.27MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     gnupg2 curl ca-certificates &&     curl -fsSLO https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/${NVARCH}/cuda-keyring_1.1-1_all.deb && dpkg -i cuda-keyring_1.1-1_all.deb &&     apt-get purge --autoremove -y curl     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2026-04-17 23:17:05  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2026-04-17 23:17:05  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2026-04-17 23:17:05  0.00B 设置环境变量 NV_CUDA_CUDART_VERSION
ENV NV_CUDA_CUDART_VERSION=13.2.75-1
                        
# 2026-04-17 23:17:05  0.00B 设置环境变量 NVIDIA_REQUIRE_CUDA brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand
ENV NVIDIA_REQUIRE_CUDA=cuda>=13.2 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=vapps,driver>=535,driver<536 brand=vpc,driver>=535,driver<536 brand=vcs,driver>=535,driver<536 brand=vws,driver>=535,driver<536 brand=cloudgaming,driver>=535,driver<536 brand=unknown,driver>=570,driver<571 brand=grid,driver>=570,driver<571 brand=tesla,driver>=570,driver<571 brand=nvidia,driver>=570,driver<571 brand=quadro,driver>=570,driver<571 brand=quadrortx,driver>=570,driver<571 brand=nvidiartx,driver>=570,driver<571 brand=vapps,driver>=570,driver<571 brand=vpc,driver>=570,driver<571 brand=vcs,driver>=570,driver<571 brand=vws,driver>=570,driver<571 brand=cloudgaming,driver>=570,driver<571 brand=unknown,driver>=580,driver<581 brand=grid,driver>=580,driver<581 brand=tesla,driver>=580,driver<581 brand=nvidia,driver>=580,driver<581 brand=quadro,driver>=580,driver<581 brand=quadrortx,driver>=580,driver<581 brand=nvidiartx,driver>=580,driver<581 brand=vapps,driver>=580,driver<581 brand=vpc,driver>=580,driver<581 brand=vcs,driver>=580,driver<581 brand=vws,driver>=580,driver<581 brand=cloudgaming,driver>=580,driver<581 brand=unknown,driver>=590,driver<591 brand=grid,driver>=590,driver<591 brand=tesla,driver>=590,driver<591 brand=nvidia,driver>=590,driver<591 brand=quadro,driver>=590,driver<591 brand=quadrortx,driver>=590,driver<591 brand=nvidiartx,driver>=590,driver<591 brand=vapps,driver>=590,driver<591 brand=vpc,driver>=590,driver<591 brand=vcs,driver>=590,driver<591 brand=vws,driver>=590,driver<591 brand=cloudgaming,driver>=590,driver<591
                        
# 2026-04-17 23:17:05  0.00B 设置环境变量 NVARCH
ENV NVARCH=x86_64
                        
# 2026-04-10 14:49:18  0.00B 
/bin/sh -c #(nop)  CMD ["/bin/bash"]
                        
# 2026-04-10 14:49:17  78.14MB 
/bin/sh -c #(nop) ADD file:8ce1caf246e7c778bca84c516d02fd4e83766bb2c530a0fffa8a351b560a2728 in / 
                        
# 2026-04-10 14:49:15  0.00B 
/bin/sh -c #(nop)  LABEL org.opencontainers.image.version=24.04
                        
# 2026-04-10 14:49:15  0.00B 
/bin/sh -c #(nop)  ARG LAUNCHPAD_BUILD_ARCH
                        
# 2026-04-10 14:49:15  0.00B 
/bin/sh -c #(nop)  ARG RELEASE
                        
                    

镜像信息

{
    "Id": "sha256:da1c3e883628cf4f5fcd507e8e906851f744820259393d4d1b4e13919e37f326",
    "RepoTags": [
        "voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:black-benediction-b12xpr11-vllmbb6c5b7-b12xd90d89c-fi3395b41aa8d-dg324aced12c-cu132-20260608"
    ],
    "RepoDigests": [
        "voipmonitor/vllm@sha256:ce23a9b075bd7138ce3b12ee29609b98606e5050e2def4a29bbb917ad96e5997",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm@sha256:be32085a4382adaa19268c12dce177d58768ccc537d508e2e7d28bcb65dfd544"
    ],
    "Parent": "",
    "Comment": "buildkit.dockerfile.v0",
    "Created": "2026-06-08T23:40:48.217360371Z",
    "Container": "",
    "ContainerConfig": null,
    "DockerVersion": "",
    "Author": "",
    "Config": {
        "Hostname": "",
        "Domainname": "",
        "User": "",
        "AttachStdin": false,
        "AttachStdout": false,
        "AttachStderr": false,
        "Tty": false,
        "OpenStdin": false,
        "StdinOnce": false,
        "Env": [
            "PATH=/opt/venv/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
            "NVARCH=x86_64",
            "NVIDIA_REQUIRE_CUDA=cuda\u003e=13.2 brand=unknown,driver\u003e=535,driver\u003c536 brand=grid,driver\u003e=535,driver\u003c536 brand=tesla,driver\u003e=535,driver\u003c536 brand=nvidia,driver\u003e=535,driver\u003c536 brand=quadro,driver\u003e=535,driver\u003c536 brand=quadrortx,driver\u003e=535,driver\u003c536 brand=nvidiartx,driver\u003e=535,driver\u003c536 brand=vapps,driver\u003e=535,driver\u003c536 brand=vpc,driver\u003e=535,driver\u003c536 brand=vcs,driver\u003e=535,driver\u003c536 brand=vws,driver\u003e=535,driver\u003c536 brand=cloudgaming,driver\u003e=535,driver\u003c536 brand=unknown,driver\u003e=570,driver\u003c571 brand=grid,driver\u003e=570,driver\u003c571 brand=tesla,driver\u003e=570,driver\u003c571 brand=nvidia,driver\u003e=570,driver\u003c571 brand=quadro,driver\u003e=570,driver\u003c571 brand=quadrortx,driver\u003e=570,driver\u003c571 brand=nvidiartx,driver\u003e=570,driver\u003c571 brand=vapps,driver\u003e=570,driver\u003c571 brand=vpc,driver\u003e=570,driver\u003c571 brand=vcs,driver\u003e=570,driver\u003c571 brand=vws,driver\u003e=570,driver\u003c571 brand=cloudgaming,driver\u003e=570,driver\u003c571 brand=unknown,driver\u003e=580,driver\u003c581 brand=grid,driver\u003e=580,driver\u003c581 brand=tesla,driver\u003e=580,driver\u003c581 brand=nvidia,driver\u003e=580,driver\u003c581 brand=quadro,driver\u003e=580,driver\u003c581 brand=quadrortx,driver\u003e=580,driver\u003c581 brand=nvidiartx,driver\u003e=580,driver\u003c581 brand=vapps,driver\u003e=580,driver\u003c581 brand=vpc,driver\u003e=580,driver\u003c581 brand=vcs,driver\u003e=580,driver\u003c581 brand=vws,driver\u003e=580,driver\u003c581 brand=cloudgaming,driver\u003e=580,driver\u003c581 brand=unknown,driver\u003e=590,driver\u003c591 brand=grid,driver\u003e=590,driver\u003c591 brand=tesla,driver\u003e=590,driver\u003c591 brand=nvidia,driver\u003e=590,driver\u003c591 brand=quadro,driver\u003e=590,driver\u003c591 brand=quadrortx,driver\u003e=590,driver\u003c591 brand=nvidiartx,driver\u003e=590,driver\u003c591 brand=vapps,driver\u003e=590,driver\u003c591 brand=vpc,driver\u003e=590,driver\u003c591 brand=vcs,driver\u003e=590,driver\u003c591 brand=vws,driver\u003e=590,driver\u003c591 brand=cloudgaming,driver\u003e=590,driver\u003c591",
            "NV_CUDA_CUDART_VERSION=13.2.75-1",
            "CUDA_VERSION=13.2.1",
            "LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64",
            "NVIDIA_VISIBLE_DEVICES=all",
            "NVIDIA_DRIVER_CAPABILITIES=compute,utility",
            "NV_CUDA_LIB_VERSION=13.2.1-1",
            "NV_NVTX_VERSION=13.2.75-1",
            "NV_LIBNPP_VERSION=13.1.0.48-1",
            "NV_LIBNPP_PACKAGE=libnpp-13-2=13.1.0.48-1",
            "NV_LIBCUSPARSE_VERSION=12.7.10.1-1",
            "NV_LIBCUBLAS_PACKAGE_NAME=libcublas-13-2",
            "NV_LIBCUBLAS_VERSION=13.4.0.1-1",
            "NV_LIBCUBLAS_PACKAGE=libcublas-13-2=13.4.0.1-1",
            "NVIDIA_PRODUCT_NAME=CUDA",
            "NV_CUDA_CUDART_DEV_VERSION=13.2.75-1",
            "NV_NVML_DEV_VERSION=13.2.82-1",
            "NV_LIBCUSPARSE_DEV_VERSION=12.7.10.1-1",
            "NV_LIBNPP_DEV_VERSION=13.1.0.48-1",
            "NV_LIBNPP_DEV_PACKAGE=libnpp-dev-13-2=13.1.0.48-1",
            "NV_LIBCUBLAS_DEV_VERSION=13.4.0.1-1",
            "NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-13-2",
            "NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-13-2=13.4.0.1-1",
            "NV_CUDA_NSIGHT_COMPUTE_VERSION=13.2.1-1",
            "NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-13-2=13.2.1-1",
            "LIBRARY_PATH=/usr/local/cuda/lib64/stubs",
            "NV_CUDNN_VERSION=9.20.0.48-1",
            "NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-13",
            "NV_CUDNN_PACKAGE=libcudnn9-cuda-13=9.20.0.48-1",
            "NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-13=9.20.0.48-1",
            "NV_CUDNN_PACKAGE_DEV_HEADERS=libcudnn9-headers-cuda-13=9.20.0.48-1",
            "LANG=C.UTF-8",
            "LC_ALL=C.UTF-8",
            "PYTHONDONTWRITEBYTECODE=1",
            "PYTHONUNBUFFERED=1",
            "PIP_DISABLE_PIP_VERSION_CHECK=1",
            "NCCL_LOCAL_INFERENCE_PATH=/opt/libnccl-local-inference.so.2.30.4",
            "NCCL_PR2127_PATH=/opt/libnccl-local-inference.so.2.30.4",
            "VLLM_NCCL_SO_PATH=/opt/libnccl-local-inference.so.2.30.4",
            "NCCL_GRAPH_FILE=",
            "VIRTUAL_ENV=/opt/venv",
            "CUDA_DEVICE_ORDER=PCI_BUS_ID",
            "TORCH_CUDA_ARCH_LIST=12.0a",
            "FLASHINFER_CUDA_ARCH_LIST=12.0f",
            "FLASHINFER_LOCAL_VERSION=cu132",
            "FLASHINFER_DISABLE_VERSION_CHECK=1",
            "HF_HOME=/root/.cache/huggingface",
            "TRITON_CACHE_DIR=/cache/jit/triton",
            "TORCH_EXTENSIONS_DIR=/cache/jit/torch_extensions",
            "VLLM_CACHE_DIR=/cache/jit/vllm",
            "FLASHINFER_WORKSPACE_BASE=/cache/jit/flashinfer",
            "TVM_FFI_CACHE_DIR=/cache/jit/tvm-ffi",
            "XDG_CACHE_HOME=/cache/jit",
            "LD_PRELOAD=/opt/libnccl-local-inference.so.2.30.4",
            "NCCL_IB_DISABLE=1",
            "NCCL_P2P_LEVEL=SYS",
            "NCCL_PROTO=LL,LL128,Simple",
            "VLLM_ENABLE_PCIE_ALLREDUCE=1",
            "VLLM_PCIE_ALLREDUCE_BACKEND=cpp",
            "VLLM_CPP_AR_1STAGE_NCCL_CUTOFF=56KB",
            "VLLM_CPP_AR_IGNORE_CUTOFF_MAX_ROWS=0",
            "VLLM_RTX6K_FUSED_ALLREDUCE_ADD=0",
            "VLLM_RTX6K_FUSED_ALLREDUCE_ADD_END_BARRIER=0",
            "VLLM_DISABLE_SHARED_EXPERTS_STREAM=0",
            "VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=1",
            "VLLM_DISABLED_KERNELS=MarlinFP8ScaledMMLinearKernel"
        ],
        "Cmd": [
            "/usr/local/bin/run-kimi26-vllm"
        ],
        "ArgsEscaped": true,
        "Image": "",
        "Volumes": null,
        "WorkingDir": "/",
        "Entrypoint": [
            "/opt/nvidia/nvidia_entrypoint.sh"
        ],
        "OnBuild": null,
        "Labels": {
            "com.nvidia.cudnn.version": "9.20.0.48-1",
            "local-inference.b12x.branch": "refs/pull/11/head",
            "local-inference.b12x.commit": "d90d89c8353adabb56cc84bd3924ef811ef8d877",
            "local-inference.b12x.runtime_version": "source",
            "local-inference.cublas.version": "13.4.1.2-1",
            "local-inference.cuda.compat.version": "595.71.05-1ubuntu1",
            "local-inference.cuda.version": "13.2.1",
            "local-inference.cudnn.version": "9.22.0.52-1",
            "local-inference.deepgemm.branch": "refs/pull/324/head",
            "local-inference.deepgemm.commit": "aced12c2c8882a945c568ace9d4a7e5778aae410",
            "local-inference.flashinfer.branch": "refs/pull/3395/head",
            "local-inference.flashinfer.commit": "b41aa8dd2fb93c49b1c6134bd1953040f8089d51",
            "local-inference.nccl.branch": "canonical/cu132-nccl2304-amd-noxml",
            "local-inference.nccl.repo": "local-inference-lab/nccl-canonical",
            "local-inference.nccl.version": "2.30.4",
            "local-inference.torch.version": "2.12.0+cu132",
            "local-inference.vllm.branch": "dev/black-benediction",
            "local-inference.vllm.commit": "bb6c5b7351fceb9d524e0d43b957415ffefcb981",
            "local-inference.vllm.repo": "https://github.com/local-inference-lab/vllm.git",
            "maintainer": "NVIDIA CORPORATION \u003ccudatools@nvidia.com\u003e",
            "org.opencontainers.image.description": "Clean CUDA 13.2.1 vLLM GLM/Kimi image with PyTorch 2.12 cu132, patched NCCL 2.30.4, FlashInfer git, B12X and canonical vLLM branch",
            "org.opencontainers.image.version": "24.04"
        },
        "Shell": [
            "/bin/bash",
            "-euxo",
            "pipefail",
            "-c"
        ]
    },
    "Architecture": "amd64",
    "Os": "linux",
    "Size": 22917906023,
    "GraphDriver": {
        "Data": {
            "LowerDir": "/var/lib/docker/overlay2/b80d23f878ea550dd6e08083cc17cbc75667cca6ea69348e6c2df3325a5a5a97/diff:/var/lib/docker/overlay2/6db0da306cf32dca42ca9a081f58d88a46f659be9161d7261722371973f778f7/diff:/var/lib/docker/overlay2/8f351e3de5822f78a11b01cf941b651e339da1297c317a4eb88b68e07b9cf732/diff:/var/lib/docker/overlay2/ce3b465b46e5a9eabe89e14218f8dd60a475c2f7424de1aab2697739f728601a/diff:/var/lib/docker/overlay2/d57f8adc714480ea23863ab5acb2497201b50db0551bee51e38c144d62586fb0/diff:/var/lib/docker/overlay2/f7f7beaba0c640a05ec03f1ec01604ecab8a1b739e468bdd2f464ac15857d742/diff:/var/lib/docker/overlay2/edd9c8db12e84238d6d6780e017e3d28a664e85e0afc3e1d42b75351fae91f43/diff:/var/lib/docker/overlay2/eb52e543c0a3d557bc76bdde58ae25fc8a7836307b2455320d35ae47cb4168dd/diff:/var/lib/docker/overlay2/f218522ee0ddc60a62dccca8d3ebce3c5b348bba985b998cbd0290546dbe46cd/diff:/var/lib/docker/overlay2/8361ec0b9938e31ed41e655af4fb55fa6cbdab0512d680ad6377d28af72369da/diff:/var/lib/docker/overlay2/08746ff7311fe24f5b4bfa7a37bf30e3be99470de8aa631f52ee5b5df6c9d587/diff:/var/lib/docker/overlay2/6278d05b76816b236d5ac5fc7bae04645f2862ebf9d65524cc87502f4077d513/diff:/var/lib/docker/overlay2/9e7f3e445e269907e09af1387e298bb18e5a91909cf28338662177632e7aacff/diff:/var/lib/docker/overlay2/1cb3b17b67b8ccf34ec8357b93bc9ff695ea9c7a2837438be15e4575d3788d01/diff:/var/lib/docker/overlay2/e6bea416d712ec1aed1891c768900656a3dc5e9f6e002d8f0d14eb9750a81a50/diff:/var/lib/docker/overlay2/d6811689e8d16bbe9507401ccd654a91ed47760e67784834bcfb0e639282c443/diff:/var/lib/docker/overlay2/ed13eefcaf0a3c732324f963fe2b68dab0a5d437bc5632718ae0d5435e30ca9b/diff:/var/lib/docker/overlay2/93f8aa301b18b8a4f2d4dcefa4c95fdcf030c15f2d1e68388ac7de0662180d19/diff:/var/lib/docker/overlay2/0360a67f6858b9a25aed9e0141dbfe16fbd088c5f5d42cf68cbca528b7fdb4f7/diff:/var/lib/docker/overlay2/209c871458bc96657c2645ab80f9620df745050d58338e0e5961fb093ae1a68d/diff:/var/lib/docker/overlay2/5c72c0c31549bd3a077f0f47b189483662530e385c8efc6bc04d9af62209630a/diff:/var/lib/docker/overlay2/2b22559e6f9796def2f4eb6b20db18cf78b006d6ff2a49de75a8da79174f1a41/diff:/var/lib/docker/overlay2/f31f69485eef4e66a4ee5847dc3331fb81c46f0d4869a83ef7e2cc195e79496e/diff:/var/lib/docker/overlay2/30f1081c75243fb910f2ec21f953b7d7ab1bf79f2a767a5f3faaf970e672230c/diff:/var/lib/docker/overlay2/f0b201bf4778e8740d07c0a3808be533c34008a1f6142bb5bff81f0c1eaefce5/diff:/var/lib/docker/overlay2/192a98a6db7a905e2c5ba331021bb81ced8b34f182875139e9a04b9673f0c652/diff:/var/lib/docker/overlay2/e7a42e175a8c0c11562e50787e04e7e851a149b37804348dd6d68d45a711acd3/diff",
            "MergedDir": "/var/lib/docker/overlay2/f9edfc70fc7e9dcdf5d804ef0837d81deba8b309ea1efab4a3d97125417808c9/merged",
            "UpperDir": "/var/lib/docker/overlay2/f9edfc70fc7e9dcdf5d804ef0837d81deba8b309ea1efab4a3d97125417808c9/diff",
            "WorkDir": "/var/lib/docker/overlay2/f9edfc70fc7e9dcdf5d804ef0837d81deba8b309ea1efab4a3d97125417808c9/work"
        },
        "Name": "overlay2"
    },
    "RootFS": {
        "Type": "layers",
        "Layers": [
            "sha256:538812a4b9bd45adaac2b5e5b967daa6999aa44eb110aa32ae7c69702b906475",
            "sha256:033dd3ac1e3752546ffe96cf256b45766e15631c5876573d70f5f73069d4e7da",
            "sha256:172292358808d705cbbe333223be62c856777357bf15a3b33483182c75b60a37",
            "sha256:e3d44de88e7571b5dbe1e6448876231156b9bf77dfa63bdba78809955f43ef31",
            "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef",
            "sha256:25aa3793397fe126f13b0fa2fdb0239bfe6c55fca85dcbed58326e3401ecb9ec",
            "sha256:b523221e8b19459a04938c97406126f5d989e7c05acc49d7f5afd4cdce3100cf",
            "sha256:11f079da8fd2b879c53c386e33f47cf481d7e90ec099d7a5dd331353543bea2a",
            "sha256:5210c94dbe4ac5650a82fbf92458c74558be3406b8f466833a0d2d84649a1335",
            "sha256:49f0bdc7c42da383f8a4d02d10fd02aad14da5db9bbd02b3eda9d0b2141d6c49",
            "sha256:a783ce97bc0f45dd9c1fcd9092bea9cbf417e777f29a9fad18c2ea5c27da6a6b",
            "sha256:16a4ae14c4630c2e174fb85535addf8d9381f21dfc8d00a682e4d8286d393e8d",
            "sha256:a5ab69e1a19c8f10434501cb6dbb6e6635f13ed1966808eb2696215caff1147d",
            "sha256:7ae6ff8705b90db9846bd637b5fd07ee28013f3d465113039fbe995ced71292e",
            "sha256:3d6dccf529343be8a5d9d5cfdc881de8068a18aac1eb5821c0bd8bf6d908e041",
            "sha256:6d02419ff72157a0cbab8e8cc488f99103a62e13a4897ca03eb5421dddf6f75c",
            "sha256:0e8d975b1b9476c5c8a18f849be89c3b6bd39f69fa81f22e944736f1bf8dc969",
            "sha256:28dd8e19a2efb71c1b4bd2804e843dc1763465d286a8d57ed60ba4a930819d15",
            "sha256:c2d5058c91ddc97765690991103f6e3feef082064c6da4b7ee41b7ee91aa2fa6",
            "sha256:0684ab29d27f53045643f4197f4e58206d210ce2bbbb1e7fa9cf002a5207d55b",
            "sha256:80c7faa939511f2cc42da0317690e3196711764d74c92e200dc9a95ae7cdbe3b",
            "sha256:a2171468c884c857f7f2e97eae4321e27665452771d6c0231c4604e9387fa3a0",
            "sha256:3c8ba1c4840db19b89707c5891691a6c0bee84551ee0c2c8d2daa35d22cfd8aa",
            "sha256:6003ada1dacbc5414be11e0ec3c9dbceb554acb1f1be6d85b6e65d6ea79bfd7f",
            "sha256:77e867bcad698166694a313766a843b4492ceb933084c32e5cb3eb5d6ab6ad83",
            "sha256:6ba8c255b09388acac8a9a3580ea05e475dd2247b37de1ca2e7688a01f48c70f",
            "sha256:a573bf8ba1dac06a38526713aae690af517bb9fdd422b3f9a601273ff486eb8f",
            "sha256:009d6cdc4d0dc2f72b6bc3622dd48c816d15ed11a01529d797f52d0bb1900b8a"
        ]
    },
    "Metadata": {
        "LastTagTime": "2026-08-30T00:49:58.528364799+08:00"
    }
}
检测到您正在使用广告拦截插件,本站为公益站点,依赖广告维持运转 🙏 查看如何关闭 ×