logo
docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24
linux/amd64 docker.io 修 贡献

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。

15
浏览次数
25.19GB
镜像大小
国内镜像
swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24
源镜像
docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24
镜像ID
sha256:dc0bc459b8c1d59f84e945a4b77f65ea474778b58f4a95d3e3d1c97632daeb1d
镜像 TAG
gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24
镜像大小
25.19GB
平台架构
linux/amd64
镜像源
docker.io
项目信息
Hub 主页 所有 TAG
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
2026-08-03T02:10:27.848351156Z
同步时间
2026-08-31 00:22
浏览量
15 次
贡献者
avatar 修
⚙️ 环境变量 89
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
VIRTUAL_ENV=/opt/venv 41
CUDA_DEVICE_ORDER=PCI_BUS_ID 42
TORCH_CUDA_ARCH_LIST=12.0a 43
FLASHINFER_CUDA_ARCH_LIST=12.0f 44
FLASHINFER_LOCAL_VERSION=cu132 45
FLASHINFER_DISABLE_VERSION_CHECK=1 46
HF_HOME=/root/.cache/huggingface 47
LOCAL_INFERENCE_CACHE_FINGERPRINT=vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174 48
XDG_CACHE_HOME=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174 49
VLLM_CACHE_ROOT=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/vllm 50
VLLM_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/vllm 51
TRITON_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/triton 52
TORCHINDUCTOR_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/torchinductor 53
TORCH_EXTENSIONS_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/torch_extensions 54
FLASHINFER_WORKSPACE_BASE=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/flashinfer 55
VLLM_FLASHINFER_AUTOTUNE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/flashinfer-autotune 56
TVM_FFI_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tvm-ffi 57
TVM_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tvm 58
TILELANG_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tilelang 59
TILELANG_TMP_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tilelang/tmp 60
CUTE_DSL_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cute-dsl 61
B12X_CUTE_COMPILE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/b12x-cute 62
SPARKINFER_COMPILE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/sparkinfer/compile 63
DG_JIT_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/deep-gemm 64
MM_SPARSE_ATTN_AOT_CACHE=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/minfer/mm_sparse_attn 65
MINFER_FMHA_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/minfer/fmha_sm100 66
CUDA_CACHE_PATH=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cuda 67
CUPY_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cupy 68
NUMBA_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/numba 69
VLLM_EXL3_EXT_PATH=/opt/exllamav3 70
VLLM_EXL3_ENCODER_SOURCE=/opt/exllamav3-python/exllamav3 71
VLLM_EXL3_ENCODER_REVISION=704aefd743b390af4bd0fb429d1906f9b964c7d8 72
VLLM_EXL3_ONLINE_CACHE_DIR=/cache/exl3-online 73
VLLM_EXL3_ONLINE_CACHE_MODE=readwrite 74
LD_PRELOAD=/opt/libnccl-local-inference.so.2.30.4 75
NCCL_IB_DISABLE=1 76
NCCL_P2P_LEVEL=SYS 77
NCCL_PROTO=LL,LL128,Simple 78
VLLM_ENABLE_PCIE_ALLREDUCE=1 79
VLLM_PCIE_ALLREDUCE_BACKEND=cpp 80
VLLM_CPP_AR_1STAGE_NCCL_CUTOFF=56KB 81
VLLM_CPP_AR_IGNORE_CUTOFF_MAX_ROWS=0 82
VLLM_RTX6K_FUSED_ALLREDUCE_ADD=0 83
VLLM_RTX6K_FUSED_ALLREDUCE_ADD_END_BARRIER=0 84
VLLM_DISABLE_SHARED_EXPERTS_STREAM=0 85
VLLM_MEMORY_PROFILE_INCLUDE_ATTN=1 86
VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=1 87
VLLM_DISABLED_KERNELS=MarlinFP8ScaledMMLinearKernel 88
🏷️ 镜像标签 73
KeyValue
9.20.0.48-1 com.nvidia.cudnn.version
master local-inference.b12x.branch
59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.b12x.commit
59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.b12x.integration.base_commit
a6511d517e3d0a8d59afd10d4b339eb7faad29bc878bc922d4cf740217926d78 local-inference.b12x.integration.lock_sha256
106@35588fb570d23713bcb2a81749d0fdf3c4396ace,113@f3308fbdb6a5613dd3ba18bf6d482458c10684d2,112@fee1d784a798439de4e3462b0f0fc831491b5c7f local-inference.b12x.integration.prs
2b9bf2a4d15770c0c23e19cc13a75843f2f0a995 local-inference.b12x.integration.tree
releases/gilded-gnosis-v20-r24/sparkinfer/integration.patch local-inference.b12x.patch_file
5fa88526cf7f468efdedcb484ed6b6a4ace065e3eb7d04a69bd46ed2570fd4b1 local-inference.b12x.patch_sha256
source local-inference.b12x.runtime_version
vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174 local-inference.cache.fingerprint
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
4.6.0 local-inference.cutlass_dsl.version
a6b593d2826719dcf4892609af7b84ee23aaf32a local-inference.deepgemm.branch
a6b593d2826719dcf4892609af7b84ee23aaf32a local-inference.deepgemm.commit
a1-retile-sm120 local-inference.exllamav3.branch
704aefd743b390af4bd0fb429d1906f9b964c7d8 local-inference.exllamav3.commit
https://github.com/brandonmmusic-max/exllamav3.git local-inference.exllamav3.repo
codex/sm120-dspark-stack-20260711 local-inference.flashinfer.branch
801d57a08958c13d375ddbb6be3be4808f48a708 local-inference.flashinfer.commit
25b3f268ea95b76bd03c825a1681872c9b615428 local-inference.instanttensor.branch
25b3f268ea95b76bd03c825a1681872c9b615428 local-inference.instanttensor.commit
https://github.com/voipmonitor/InstantTensor.git local-inference.instanttensor.repo
76b1cf0350709910208e61285adc955e34655136 local-inference.launcher.branch
76b1cf0350709910208e61285adc955e34655136 local-inference.launcher.commit
serve-gilded-gnosis.sh serve-fathomless-firmament.sh serve-glm52-v16.sh serve-glm52-v18.sh serve-glm52-v19.sh serve-glm52-hybrid-v17.sh serve-glm52-hybrid-v18.sh serve-glm52-hybrid-v19.sh glm52-dcp-prefill-policy.sh glm52-pcie-runtime-env.sh glm52-pcie-calibration.py glm52-lmcache-wrapper.sh local-inference.launchers
9cebd405d0caf4bebe01d694b5a8bf4e3e354314 local-inference.lmcache.commit
9cebd405d0caf4bebe01d694b5a8bf4e3e354314 local-inference.lmcache.integration.base_commit
573209e9258c4706b5f1f1e77c2e8a6de504a8068cfe80b6fcc76f5b21a56d79 local-inference.lmcache.integration.lock_sha256
7@7b7583aef55e98ba05a6c199e71601d78552e794,8@31c4175d2134518e5b43fe8f4a7d072df6043a13,9@c3b73de74d855f6263deea8b6c6c4dde87e6e5b9,10@59186798bb5859b7add7bcac81dff9f2fd4037ab,11@2cee5a2c1ef7e80028e94b1bc4bfbfc02e2c2a8d,12@baf1345106ebe568a116a35fdcee36bfec73da57,13@d4dce69bf90362953cf9ac9a86cccc186b3219dc,14@502b3c60f2e61625d6e32a31d0d725eedc620792,15@7729a9c75f2c29f14a27ea1e84ee6970950197ce,16@e6c2708fe8ae8a96d3c66e83bf4b548c734fbb13,17@f9639e2094f70955a1145bfd7219b6fdd39ee143 local-inference.lmcache.integration.prs
9a05c8818bae48d15b79c7e876418bb813c08cd0 local-inference.lmcache.integration.tree
e9f5369ee134cbc697e77d9fd9d77219aceea5fabf2fe90e3b74542bbeb200ca local-inference.lmcache.patch_sha256
release/v0.5.2-glm52-dcp-base local-inference.lmcache.ref
https://github.com/local-inference-lab/LMCache.git local-inference.lmcache.repo
0.5.2+glm52dcp.4 local-inference.lmcache.version
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
master local-inference.sparkinfer.branch
59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.sparkinfer.commit
59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.sparkinfer.integration.base_commit
a6511d517e3d0a8d59afd10d4b339eb7faad29bc878bc922d4cf740217926d78 local-inference.sparkinfer.integration.lock_sha256
106@35588fb570d23713bcb2a81749d0fdf3c4396ace,113@f3308fbdb6a5613dd3ba18bf6d482458c10684d2,112@fee1d784a798439de4e3462b0f0fc831491b5c7f local-inference.sparkinfer.integration.prs
2b9bf2a4d15770c0c23e19cc13a75843f2f0a995 local-inference.sparkinfer.integration.tree
releases/gilded-gnosis-v20-r24/sparkinfer/integration.patch local-inference.sparkinfer.patch_file
5fa88526cf7f468efdedcb484ed6b6a4ace065e3eb7d04a69bd46ed2570fd4b1 local-inference.sparkinfer.patch_sha256
source local-inference.sparkinfer.runtime_version
0.1.8 local-inference.tokenspeed_mla.version
2.12.0+cu132 local-inference.torch.version
local-inference.triton_kernels.branch
local-inference.triton_kernels.commit
0.1.10 local-inference.tvm_ffi.version
dev/gilded-gnosis local-inference.vllm.branch
30038602b71395f481ef4a6edfe4fcf8551d9c15 local-inference.vllm.commit
30038602b71395f481ef4a6edfe4fcf8551d9c15 local-inference.vllm.integration.base_commit
3de67581d0353e44240725ee8a0cd756c8d2a424e37120333bb1e91acc352ecc local-inference.vllm.integration.lock_sha256
145@99bf3f13f13012b72d6d67e49572d686b45b2833,229@672aec40dc664de014bd1e92d9a2885d7658c37f,213@5f144274d6e092237f4f4f437b7e9d3aaa427157,214@e63a190ee1b48d051ccebce485de343dbe2b3505,217@c7ac22d5bfe923803cc923f2293085faac7d1dd0,218@3b89c7a113f92522aa07cc3271422c04e458d771,216@52201611719edf67f18bc0cbec17802880a4afb7,228@1702ed2d28342d84551539f8d7e2de637f7ca04e,230@9401f9d8916f3bdf34123d2d493a918b748c8132 local-inference.vllm.integration.prs
f5981f14b4d39979bc0d799c020d42002b707257 local-inference.vllm.integration.tree
releases/gilded-gnosis-v20-r24/vllm/integration.patch local-inference.vllm.patch_file
11e6c2b80b25d6433f01229bf9f0a0cafbfee8bd7125332a708398f93d8d77e9 local-inference.vllm.patch_sha256
local-inference.vllm.patch_url
https://github.com/local-inference-lab/vllm.git local-inference.vllm.repo
2ea71da4ccb997a06928c9fb69b99f330da56697 local-inference.xgrammar.commit
v0.2.5 local-inference.xgrammar.ref
https://github.com/mlc-ai/xgrammar.git local-inference.xgrammar.repo
1 local-inference.xgrammar.transformers5_compat
0.2.5 local-inference.xgrammar.version
NVIDIA CORPORATION <cudatools@nvidia.com> maintainer
Clean CUDA 13.2.1 vLLM+SparkInfer image with PyTorch 2.12 cu132, patched NCCL 2.30.4, FlashInfer git, DeepGEMM, SparkInfer and selected vLLM branch org.opencontainers.image.description
24.04 org.opencontainers.image.version
🛡️ 镜像安全扫描
ubuntu 24.04 Trivy 2026-08-31 00:26 查看完整报告
415
低危 LOW
3233
中危 MEDIUM
389
高危 HIGH
26
严重 CRITICAL
受影响目标 (4)
docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24 (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
阿里云阿里云高校计划(￥300代金券)领取 阿里云阿里云2核2G高性能云服务器 99元/年❤️ 华为云华为云2核4G2M云服务器，惊爆价99元1年 腾讯云腾讯云2核2G云服务器，惊爆价99元1年 UcloudUcloud 2核2G4M 59元/年❤️ 香港专线400G流量/月服务器 93元/年❤️ 低成本编程模型Opencode-Go低成本编程模型 chatgptGPT人工智能60G学习资料❤️ 云端GPU大模型AI训练️
无权限下载？点我修复 生成 Docker Run 命令 生成 Kubernetes YAML

Docker拉取命令 

docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24
docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24  docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24

Containerd拉取命令 

ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24
ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24  docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24

Shell快速替换命令 

sed -i 's#voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24#swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24#' 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:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24 && docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24  docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24'

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:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24 && ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24  docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24'

镜像构建历史 


# 2026-08-03 10:10:27  0.00B 添加元数据标签
LABEL org.opencontainers.image.description=Clean CUDA 13.2.1 vLLM+SparkInfer image with PyTorch 2.12 cu132, patched NCCL 2.30.4, FlashInfer git, DeepGEMM, SparkInfer and selected 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.cutlass_dsl.version=4.6.0 local-inference.tokenspeed_mla.version=0.1.8 local-inference.tvm_ffi.version=0.1.10 local-inference.cache.fingerprint=vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174 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/gilded-gnosis local-inference.vllm.commit=30038602b71395f481ef4a6edfe4fcf8551d9c15 local-inference.vllm.patch_url= local-inference.vllm.patch_sha256=11e6c2b80b25d6433f01229bf9f0a0cafbfee8bd7125332a708398f93d8d77e9 local-inference.vllm.patch_file=releases/gilded-gnosis-v20-r24/vllm/integration.patch local-inference.vllm.integration.base_commit=30038602b71395f481ef4a6edfe4fcf8551d9c15 local-inference.vllm.integration.tree=f5981f14b4d39979bc0d799c020d42002b707257 local-inference.vllm.integration.prs=145@99bf3f13f13012b72d6d67e49572d686b45b2833,229@672aec40dc664de014bd1e92d9a2885d7658c37f,213@5f144274d6e092237f4f4f437b7e9d3aaa427157,214@e63a190ee1b48d051ccebce485de343dbe2b3505,217@c7ac22d5bfe923803cc923f2293085faac7d1dd0,218@3b89c7a113f92522aa07cc3271422c04e458d771,216@52201611719edf67f18bc0cbec17802880a4afb7,228@1702ed2d28342d84551539f8d7e2de637f7ca04e,230@9401f9d8916f3bdf34123d2d493a918b748c8132 local-inference.vllm.integration.lock_sha256=3de67581d0353e44240725ee8a0cd756c8d2a424e37120333bb1e91acc352ecc local-inference.b12x.branch=master local-inference.b12x.commit=59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.b12x.patch_sha256=5fa88526cf7f468efdedcb484ed6b6a4ace065e3eb7d04a69bd46ed2570fd4b1 local-inference.b12x.patch_file=releases/gilded-gnosis-v20-r24/sparkinfer/integration.patch local-inference.b12x.integration.base_commit=59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.b12x.integration.tree=2b9bf2a4d15770c0c23e19cc13a75843f2f0a995 local-inference.b12x.integration.prs=106@35588fb570d23713bcb2a81749d0fdf3c4396ace,113@f3308fbdb6a5613dd3ba18bf6d482458c10684d2,112@fee1d784a798439de4e3462b0f0fc831491b5c7f local-inference.b12x.integration.lock_sha256=a6511d517e3d0a8d59afd10d4b339eb7faad29bc878bc922d4cf740217926d78 local-inference.b12x.runtime_version=source local-inference.sparkinfer.branch=master local-inference.sparkinfer.commit=59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.sparkinfer.patch_sha256=5fa88526cf7f468efdedcb484ed6b6a4ace065e3eb7d04a69bd46ed2570fd4b1 local-inference.sparkinfer.patch_file=releases/gilded-gnosis-v20-r24/sparkinfer/integration.patch local-inference.sparkinfer.integration.base_commit=59216fa25f3d5fc9d4df2d052e02d05f763906e9 local-inference.sparkinfer.integration.tree=2b9bf2a4d15770c0c23e19cc13a75843f2f0a995 local-inference.sparkinfer.integration.prs=106@35588fb570d23713bcb2a81749d0fdf3c4396ace,113@f3308fbdb6a5613dd3ba18bf6d482458c10684d2,112@fee1d784a798439de4e3462b0f0fc831491b5c7f local-inference.sparkinfer.integration.lock_sha256=a6511d517e3d0a8d59afd10d4b339eb7faad29bc878bc922d4cf740217926d78 local-inference.sparkinfer.runtime_version=source local-inference.flashinfer.branch=codex/sm120-dspark-stack-20260711 local-inference.flashinfer.commit=801d57a08958c13d375ddbb6be3be4808f48a708 local-inference.deepgemm.branch=a6b593d2826719dcf4892609af7b84ee23aaf32a local-inference.deepgemm.commit=a6b593d2826719dcf4892609af7b84ee23aaf32a local-inference.triton_kernels.branch= local-inference.triton_kernels.commit= local-inference.instanttensor.repo=https://github.com/voipmonitor/InstantTensor.git local-inference.instanttensor.branch=25b3f268ea95b76bd03c825a1681872c9b615428 local-inference.instanttensor.commit=25b3f268ea95b76bd03c825a1681872c9b615428 local-inference.exllamav3.repo=https://github.com/brandonmmusic-max/exllamav3.git local-inference.exllamav3.branch=a1-retile-sm120 local-inference.exllamav3.commit=704aefd743b390af4bd0fb429d1906f9b964c7d8 local-inference.lmcache.repo=https://github.com/local-inference-lab/LMCache.git local-inference.lmcache.ref=release/v0.5.2-glm52-dcp-base local-inference.lmcache.commit=9cebd405d0caf4bebe01d694b5a8bf4e3e354314 local-inference.lmcache.patch_sha256=e9f5369ee134cbc697e77d9fd9d77219aceea5fabf2fe90e3b74542bbeb200ca local-inference.lmcache.version=0.5.2+glm52dcp.4 local-inference.lmcache.integration.base_commit=9cebd405d0caf4bebe01d694b5a8bf4e3e354314 local-inference.lmcache.integration.tree=9a05c8818bae48d15b79c7e876418bb813c08cd0 local-inference.lmcache.integration.prs=7@7b7583aef55e98ba05a6c199e71601d78552e794,8@31c4175d2134518e5b43fe8f4a7d072df6043a13,9@c3b73de74d855f6263deea8b6c6c4dde87e6e5b9,10@59186798bb5859b7add7bcac81dff9f2fd4037ab,11@2cee5a2c1ef7e80028e94b1bc4bfbfc02e2c2a8d,12@baf1345106ebe568a116a35fdcee36bfec73da57,13@d4dce69bf90362953cf9ac9a86cccc186b3219dc,14@502b3c60f2e61625d6e32a31d0d725eedc620792,15@7729a9c75f2c29f14a27ea1e84ee6970950197ce,16@e6c2708fe8ae8a96d3c66e83bf4b548c734fbb13,17@f9639e2094f70955a1145bfd7219b6fdd39ee143 local-inference.lmcache.integration.lock_sha256=573209e9258c4706b5f1f1e77c2e8a6de504a8068cfe80b6fcc76f5b21a56d79 local-inference.xgrammar.repo=https://github.com/mlc-ai/xgrammar.git local-inference.xgrammar.ref=v0.2.5 local-inference.xgrammar.commit=2ea71da4ccb997a06928c9fb69b99f330da56697 local-inference.xgrammar.version=0.2.5 local-inference.xgrammar.transformers5_compat=1 local-inference.launcher.branch=76b1cf0350709910208e61285adc955e34655136 local-inference.launcher.commit=76b1cf0350709910208e61285adc955e34655136 local-inference.launchers=serve-gilded-gnosis.sh serve-fathomless-firmament.sh serve-glm52-v16.sh serve-glm52-v18.sh serve-glm52-v19.sh serve-glm52-hybrid-v17.sh serve-glm52-hybrid-v18.sh serve-glm52-hybrid-v19.sh glm52-dcp-prefill-policy.sh glm52-pcie-runtime-env.sh glm52-pcie-calibration.py glm52-lmcache-wrapper.sh
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG XGRAMMAR_TRANSFORMERS5_COMPAT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG XGRAMMAR_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG XGRAMMAR_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG XGRAMMAR_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG XGRAMMAR_REPO
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_INTEGRATION_LOCK_SHA256
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_INTEGRATION_PRS
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_INTEGRATION_TREE
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_INTEGRATION_BASE_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_BUILD_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_PATCH_SHA256
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LMCACHE_REPO
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG EXLLAMAV3_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG EXLLAMAV3_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG EXLLAMAV3_REPO
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG TVM_FFI_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG TOKENSPEED_MLA_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG CUTLASS_DSL_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG CUDA_COMPAT_13_2_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG CUDNN_CUDA13_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG CUBLAS_CUDA13_VERSION
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_REQUIRED_LAUNCHERS
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LAUNCHER_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG LAUNCHER_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG INSTANTTENSOR_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG INSTANTTENSOR_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG INSTANTTENSOR_REPO
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG TRITON_KERNELS_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG TRITON_KERNELS_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG DEEPGEMM_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG DEEPGEMM_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG FLASHINFER_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG FLASHINFER_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_INTEGRATION_LOCK_SHA256
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_INTEGRATION_PRS
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_INTEGRATION_TREE
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_INTEGRATION_BASE_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_PATCH_FILE
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_PATCH_SHA256
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG B12X_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_INTEGRATION_LOCK_SHA256
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_INTEGRATION_PRS
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_INTEGRATION_TREE
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_INTEGRATION_BASE_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_PATCH_FILE
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_PATCH_SHA256
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_PATCH_URL
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_COMMIT
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_REF
                        
# 2026-08-03 10:10:27  0.00B 定义构建参数
ARG VLLM_REPO
                        
# 2026-08-03 10:10:27  0.00B 设置默认要执行的命令
CMD ["/usr/local/bin/run-kimi26-vllm"]
                        
# 2026-08-03 10:10:27  0.00B 设置工作目录为/
WORKDIR /
                        
# 2026-08-03 10:10:27  0.00B 执行命令并创建新的镜像层
RUN |11 HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.10 VLLM_RUNTIME_EXTRA_PACKAGES=nvtx==0.2.15 PyNvVideoCodec==2.0.4 nccl4py==0.3.1 INSTANTTENSOR_REPO=https://github.com/voipmonitor/InstantTensor.git INSTANTTENSOR_REF=25b3f268ea95b76bd03c825a1681872c9b615428 INSTANTTENSOR_COMMIT=25b3f268ea95b76bd03c825a1681872c9b615428 CUTLASS_DSL_VERSION=4.6.0 TOKENSPEED_MLA_VERSION=0.1.8 TVM_FFI_VERSION=0.1.10 LMCACHE_BUILD_VERSION=0.5.2+glm52dcp.4 CACHE_FINGERPRINT=vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174 EXLLAMAV3_COMMIT=704aefd743b390af4bd0fb429d1906f9b964c7d8 /bin/bash -euxo pipefail -c mkdir -p       "${XDG_CACHE_HOME}"       "${VLLM_CACHE_ROOT}"       "${TRITON_CACHE_DIR}"       "${TORCHINDUCTOR_CACHE_DIR}"       "${TORCH_EXTENSIONS_DIR}"       "${FLASHINFER_WORKSPACE_BASE}"       "${VLLM_FLASHINFER_AUTOTUNE_CACHE_DIR}"       "${TVM_FFI_CACHE_DIR}"       "${TVM_CACHE_DIR}"       "${TILELANG_CACHE_DIR}"       "${TILELANG_TMP_DIR}"       "${CUTE_DSL_CACHE_DIR}"       "${B12X_CUTE_COMPILE_CACHE_DIR}"       "${SPARKINFER_COMPILE_CACHE_DIR}"       "${DG_JIT_CACHE_DIR}"       "${MM_SPARSE_ATTN_AOT_CACHE}"       "${MINFER_FMHA_CACHE_DIR}"       "${CUDA_CACHE_PATH}"       "${CUPY_CACHE_DIR}"       "${NUMBA_CACHE_DIR}"       "${VLLM_EXL3_ONLINE_CACHE_DIR}"  && 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-vllm-b12x-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
                        
# 2026-08-03 10:10:12  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 LOCAL_INFERENCE_CACHE_FINGERPRINT XDG_CACHE_HOME VLLM_CACHE_ROOT VLLM_CACHE_DIR TRITON_CACHE_DIR TORCHINDUCTOR_CACHE_DIR TORCH_EXTENSIONS_DIR FLASHINFER_WORKSPACE_BASE VLLM_FLASHINFER_AUTOTUNE_CACHE_DIR TVM_FFI_CACHE_DIR TVM_CACHE_DIR TILELANG_CACHE_DIR TILELANG_TMP_DIR CUTE_DSL_CACHE_DIR B12X_CUTE_COMPILE_CACHE_DIR SPARKINFER_COMPILE_CACHE_DIR DG_JIT_CACHE_DIR MM_SPARSE_ATTN_AOT_CACHE MINFER_FMHA_CACHE_DIR CUDA_CACHE_PATH CUPY_CACHE_DIR NUMBA_CACHE_DIR NCCL_LOCAL_INFERENCE_PATH NCCL_PR2127_PATH VLLM_NCCL_SO_PATH VLLM_EXL3_EXT_PATH VLLM_EXL3_ENCODER_SOURCE VLLM_EXL3_ENCODER_REVISION VLLM_EXL3_ONLINE_CACHE_DIR VLLM_EXL3_ONLINE_CACHE_MODE 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_PROFILE_INCLUDE_ATTN 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 LOCAL_INFERENCE_CACHE_FINGERPRINT=vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174 XDG_CACHE_HOME=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174 VLLM_CACHE_ROOT=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/vllm VLLM_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/vllm TRITON_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/triton TORCHINDUCTOR_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/torchinductor TORCH_EXTENSIONS_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/torch_extensions FLASHINFER_WORKSPACE_BASE=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/flashinfer VLLM_FLASHINFER_AUTOTUNE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/flashinfer-autotune TVM_FFI_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tvm-ffi TVM_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tvm TILELANG_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tilelang TILELANG_TMP_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tilelang/tmp CUTE_DSL_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cute-dsl B12X_CUTE_COMPILE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/b12x-cute SPARKINFER_COMPILE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/sparkinfer/compile DG_JIT_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/deep-gemm MM_SPARSE_ATTN_AOT_CACHE=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/minfer/mm_sparse_attn MINFER_FMHA_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/minfer/fmha_sm100 CUDA_CACHE_PATH=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cuda CUPY_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cupy NUMBA_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/numba 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 VLLM_EXL3_EXT_PATH=/opt/exllamav3 VLLM_EXL3_ENCODER_SOURCE=/opt/exllamav3-python/exllamav3 VLLM_EXL3_ENCODER_REVISION=704aefd743b390af4bd0fb429d1906f9b964c7d8 VLLM_EXL3_ONLINE_CACHE_DIR=/cache/exl3-online VLLM_EXL3_ONLINE_CACHE_MODE=readwrite 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_PROFILE_INCLUDE_ATTN=1 VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=1 VLLM_DISABLED_KERNELS=MarlinFP8ScaledMMLinearKernel
                        
# 2026-08-03 10:10:12  0.00B 定义构建参数
ARG EXLLAMAV3_COMMIT
                        
# 2026-08-03 10:10:12  0.00B 定义构建参数
ARG CACHE_FINGERPRINT=unversioned
                        
# 2026-08-03 10:10:12  32.29KB 执行命令并创建新的镜像层
RUN |9 HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.10 VLLM_RUNTIME_EXTRA_PACKAGES=nvtx==0.2.15 PyNvVideoCodec==2.0.4 nccl4py==0.3.1 INSTANTTENSOR_REPO=https://github.com/voipmonitor/InstantTensor.git INSTANTTENSOR_REF=25b3f268ea95b76bd03c825a1681872c9b615428 INSTANTTENSOR_COMMIT=25b3f268ea95b76bd03c825a1681872c9b615428 CUTLASS_DSL_VERSION=4.6.0 TOKENSPEED_MLA_VERSION=0.1.8 TVM_FFI_VERSION=0.1.10 LMCACHE_BUILD_VERSION=0.5.2+glm52dcp.4 /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
                        
# 2026-08-03 10:10:11  686.15MB 执行命令并创建新的镜像层
RUN |9 HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.10 VLLM_RUNTIME_EXTRA_PACKAGES=nvtx==0.2.15 PyNvVideoCodec==2.0.4 nccl4py==0.3.1 INSTANTTENSOR_REPO=https://github.com/voipmonitor/InstantTensor.git INSTANTTENSOR_REF=25b3f268ea95b76bd03c825a1681872c9b615428 INSTANTTENSOR_COMMIT=25b3f268ea95b76bd03c825a1681872c9b615428 CUTLASS_DSL_VERSION=4.6.0 TOKENSPEED_MLA_VERSION=0.1.8 TVM_FFI_VERSION=0.1.10 LMCACHE_BUILD_VERSION=0.5.2+glm52dcp.4 /bin/bash -euxo pipefail -c torch_nccl=/opt/venv/lib/python3.12/site-packages/nvidia/nccl/lib/libnccl.so.2  && local_nccl=/opt/libnccl-local-inference.so.2.30.4  && normal_nccl=/opt/libnccl.so.2.30.4  && test -e "${torch_nccl}"  && test -e "${local_nccl}"  && if [[ ! -e "${normal_nccl}" ]]; then cp -aL "${local_nccl}" "${normal_nccl}"; fi  && if [[ ! -e "${local_nccl}.orig" ]]; then cp -aL "${local_nccl}" "${local_nccl}.orig"; fi  && if [[ ! -e "${torch_nccl}.torch-bundled-2.29.7" ]]; then cp -aL "${torch_nccl}" "${torch_nccl}.torch-bundled-2.29.7"; fi  && readelf -d "${normal_nccl}" | grep -F 'Library soname: [libnccl.so.2]'  && ln -sfn "${normal_nccl}" "${local_nccl}"  && ln -sfn "${normal_nccl}" /opt/libnccl.so.2  && ln -sfn "${normal_nccl}" "${torch_nccl}"  && ldconfig  && /opt/venv/bin/python - <<'PY'
import torch

paths = []
with open("/proc/self/maps", encoding="utf-8") as maps:
    for line in maps:
        if "libnccl" in line:
            paths.append(line.split()[-1])

unique = sorted(set(paths))
print("torch nccl api", torch.cuda.nccl.version())
print("mapped libnccl paths", unique)
assert unique == ["/opt/libnccl.so.2.30.4"], unique
PY # buildkit
                        
# 2026-08-03 10:10:09  311.45MB 执行命令并创建新的镜像层
RUN |9 HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.10 VLLM_RUNTIME_EXTRA_PACKAGES=nvtx==0.2.15 PyNvVideoCodec==2.0.4 nccl4py==0.3.1 INSTANTTENSOR_REPO=https://github.com/voipmonitor/InstantTensor.git INSTANTTENSOR_REF=25b3f268ea95b76bd03c825a1681872c9b615428 INSTANTTENSOR_COMMIT=25b3f268ea95b76bd03c825a1681872c9b615428 CUTLASS_DSL_VERSION=4.6.0 TOKENSPEED_MLA_VERSION=0.1.8 TVM_FFI_VERSION=0.1.10 LMCACHE_BUILD_VERSION=0.5.2+glm52dcp.4 /bin/bash -euxo pipefail -c /opt/venv/bin/python -m pip install --force-reinstall --no-deps       "nvidia-cutlass-dsl-libs-cu13==${CUTLASS_DSL_VERSION}"  && CUTLASS_DSL_VERSION="${CUTLASS_DSL_VERSION}" /opt/venv/bin/python - <<'PY'
import hashlib
import importlib.metadata as md
import os
from pathlib import Path

import cutlass.cute as cute

version = os.environ["CUTLASS_DSL_VERSION"]
mma = Path(cute.__file__).parent / "nvgpu" / "warp" / "mma.py"
assert md.version("nvidia-cutlass-dsl") == version
assert md.version("nvidia-cutlass-dsl-libs-base") == version
assert md.version("nvidia-cutlass-dsl-libs-cu13") == version
assert hasattr(cute.nvgpu.warp, "MmaMXF8Op")
if version == "4.5.3":
    assert hashlib.sha256(mma.read_bytes()).hexdigest() == (
        "cccf48864bae9554acdf708fd66a7b2fe729948ca3768205df3e251c8dc71fd2"
    )
print("CUTLASS DSL CUDA 13 overlay: PASS", mma)
PY # buildkit
                        
# 2026-08-03 10:10:04  22.65MB 执行命令并创建新的镜像层
RUN |9 HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.10 VLLM_RUNTIME_EXTRA_PACKAGES=nvtx==0.2.15 PyNvVideoCodec==2.0.4 nccl4py==0.3.1 INSTANTTENSOR_REPO=https://github.com/voipmonitor/InstantTensor.git INSTANTTENSOR_REF=25b3f268ea95b76bd03c825a1681872c9b615428 INSTANTTENSOR_COMMIT=25b3f268ea95b76bd03c825a1681872c9b615428 CUTLASS_DSL_VERSION=4.6.0 TOKENSPEED_MLA_VERSION=0.1.8 TVM_FFI_VERSION=0.1.10 LMCACHE_BUILD_VERSION=0.5.2+glm52dcp.4 /bin/bash -euxo pipefail -c if [[ -n "${INSTANTTENSOR_REF}" ]]; then       git clone --recursive "${INSTANTTENSOR_REPO}" /tmp/instanttensor-src;       cd /tmp/instanttensor-src;       if [[ "${INSTANTTENSOR_REF}" =~ ^[0-9a-f]{40}$ ]]; then         git fetch --depth=1 origin "${INSTANTTENSOR_REF}";         git checkout FETCH_HEAD;       else         git checkout "${INSTANTTENSOR_REF}";       fi;       git submodule update --init --recursive;       if [[ -n "${INSTANTTENSOR_COMMIT}" ]]; then [[ "$(git rev-parse HEAD)" = "${INSTANTTENSOR_COMMIT}" ]] || { echo "ERROR: INSTANTTENSOR_COMMIT mismatch: HEAD=$(git rev-parse HEAD) expected=${INSTANTTENSOR_COMMIT}" >&2; exit 1; }; fi;       /opt/venv/bin/python -m pip install --no-build-isolation --no-deps --force-reinstall .;       cd /;       rm -rf /tmp/instanttensor-src;     fi  && /opt/venv/bin/python - <<'PY'
import importlib.metadata as md
import instanttensor

print("instanttensor", md.version("instanttensor"), instanttensor.__file__)
PY # buildkit
                        
# 2026-08-03 10:09:24  255.84MB 执行命令并创建新的镜像层
RUN |9 HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.10 VLLM_RUNTIME_EXTRA_PACKAGES=nvtx==0.2.15 PyNvVideoCodec==2.0.4 nccl4py==0.3.1 INSTANTTENSOR_REPO=https://github.com/voipmonitor/InstantTensor.git INSTANTTENSOR_REF=25b3f268ea95b76bd03c825a1681872c9b615428 INSTANTTENSOR_COMMIT=25b3f268ea95b76bd03c825a1681872c9b615428 CUTLASS_DSL_VERSION=4.6.0 TOKENSPEED_MLA_VERSION=0.1.8 TVM_FFI_VERSION=0.1.10 LMCACHE_BUILD_VERSION=0.5.2+glm52dcp.4 /bin/bash -euxo pipefail -c /opt/venv/bin/python -m pip install       sortedcontainers==2.4.0       aiofile==3.11.1       aiofiles==25.1.0       cupy-cuda13x==13.6.0       redis==8.0.1       opentelemetry-exporter-prometheus==0.65b0  && /opt/venv/bin/python -m pip install --no-deps --force-reinstall       /tmp/lmcache-wheels/lmcache-*.whl  && LMCACHE_BUILD_VERSION="${LMCACHE_BUILD_VERSION}" /opt/venv/bin/python - <<'PY'
import importlib.metadata as md
import os

import cupy
from lmcache.integration.vllm.vllm_multi_process_adapter import ParallelStrategy

assert md.version("lmcache") == os.environ["LMCACHE_BUILD_VERSION"]
assert md.version("cupy-cuda13x") == "13.6.0"
strategy = ParallelStrategy(
    use_mla=True,
    vllm_world_size=8,
    vllm_worker_id=5,
    tp_size=8,
    pp_size=1,
    n_servers=1,
    dcp_size=4,
)
assert strategy.kv_world_size == 4
assert strategy.kv_worker_id == 1
assert strategy.kv_tp_size == 2
assert strategy.kv_readers_per_object == 2
print("lmcache", md.version("lmcache"))
PY # buildkit
                        
# 2026-08-03 10:07:25  1.35GB 执行命令并创建新的镜像层
RUN |9 HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.10 VLLM_RUNTIME_EXTRA_PACKAGES=nvtx==0.2.15 PyNvVideoCodec==2.0.4 nccl4py==0.3.1 INSTANTTENSOR_REPO=https://github.com/voipmonitor/InstantTensor.git INSTANTTENSOR_REF=25b3f268ea95b76bd03c825a1681872c9b615428 INSTANTTENSOR_COMMIT=25b3f268ea95b76bd03c825a1681872c9b615428 CUTLASS_DSL_VERSION=4.6.0 TOKENSPEED_MLA_VERSION=0.1.8 TVM_FFI_VERSION=0.1.10 LMCACHE_BUILD_VERSION=0.5.2+glm52dcp.4 /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-cuda-tileiras==13.2.78'       'nvidia-nvjitlink==13.2.78'       'nvidia-nvvm==13.2.78'       "nvidia-cutlass-dsl==${CUTLASS_DSL_VERSION}"       "nvidia-cutlass-dsl-libs-base==${CUTLASS_DSL_VERSION}"       "nvidia-cutlass-dsl-libs-cu13==${CUTLASS_DSL_VERSION}"       > /tmp/torch-cu132-constraints.txt  && /opt/venv/bin/python -m pip install --upgrade       --index-url https://download.pytorch.org/whl/cu132       --extra-index-url https://pypi.org/simple       --constraint /tmp/torch-cu132-constraints.txt       "cuda-bindings==13.2.0"       "cuda-python==13.2.0"       "numba==0.65.0"       "quack-kernels>=0.3.3"       "${HUMMING_KERNELS_SPEC}"       "nvidia-cuda-tileiras==13.2.78"       "nvidia-nvjitlink==13.2.78"       "cuda-tile>=1.4.0"       "tilelang==0.1.9"       "tokenspeed-mla==${TOKENSPEED_MLA_VERSION}"       "nvidia-cutlass-dsl[cu13]==${CUTLASS_DSL_VERSION}"       "transformers>=5.5.1"       "apache-tvm-ffi==${TVM_FFI_VERSION}"       "fastsafetensors>=0.3.2"       "pytest==8.4.1"       "tblib==3.2.2"  && if [[ -n "${VLLM_RUNTIME_EXTRA_PACKAGES}" ]]; then       /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         ${VLLM_RUNTIME_EXTRA_PACKAGES};     fi  && /opt/venv/bin/python -m pip check # buildkit
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG LMCACHE_BUILD_VERSION
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG TVM_FFI_VERSION
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG TOKENSPEED_MLA_VERSION
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG CUTLASS_DSL_VERSION
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG INSTANTTENSOR_COMMIT
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG INSTANTTENSOR_REF
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG INSTANTTENSOR_REPO
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG VLLM_RUNTIME_EXTRA_PACKAGES=
                        
# 2026-08-03 10:07:07  0.00B 定义构建参数
ARG HUMMING_KERNELS_SPEC=humming-kernels[cu13]==0.1.4
                        
# 2026-08-03 10:07:07  8.24MB 复制新文件或目录到容器中
COPY /opt/exllamav3-python /opt/exllamav3-python # buildkit
                        
# 2026-08-03 10:07:07  95.62MB 复制新文件或目录到容器中
COPY /opt/exllamav3 /opt/exllamav3 # buildkit
                        
# 2026-08-03 10:07:07  1.48KB 复制新文件或目录到容器中
COPY /usr/local/bin/verify-vllm-b12x-cu132 /usr/local/bin/verify-vllm-b12x-cu132 # buildkit
                        
# 2026-08-03 10:07:07  19.59KB 复制新文件或目录到容器中
COPY /usr/local/bin/glm52-pcie-calibration.py /usr/local/bin/glm52-pcie-calibration.py # buildkit
                        
# 2026-08-03 10:07:07  12.92KB 复制新文件或目录到容器中
COPY /usr/local/bin/glm52-*.sh /usr/local/bin/ # buildkit
                        
# 2026-08-03 10:07:07  159.58KB 复制新文件或目录到容器中
COPY /usr/local/bin/serve-*.sh /usr/local/bin/ # buildkit
                        
# 2026-08-03 10:07:07  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
                        
# 2026-08-03 10:07:07  7.39KB 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
                        
# 2026-08-03 10:07:07  1.64GB 复制新文件或目录到容器中
COPY /opt/vllm /opt/vllm # buildkit
                        
# 2026-08-03 10:07:06  8.62GB 复制新文件或目录到容器中
COPY /opt/venv /opt/venv # buildkit
                        
# 2026-08-03 10:07:06  0.00B 
SHELL [/bin/bash -euxo pipefail -c]
                        
# 2026-06-28 01:59:28  0.00B 添加元数据标签
LABEL org.opencontainers.image.description=Reusable CUDA 13.2.1 vLLM+B12X 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
                        
# 2026-06-28 01:59:28  0.00B 设置环境变量 NCCL_LOCAL_INFERENCE_PATH NCCL_PR2127_PATH VLLM_NCCL_SO_PATH
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
                        
# 2026-06-28 01:59:28  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
                        
# 2026-06-28 01:57:19  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
                        
# 2026-06-28 01:57:18  32.29KB 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
                        
# 2026-06-28 01:57:18  4.40GB 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     libprotobuf-dev     protobuf-compiler     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
                        
# 2026-06-28 01:57:18  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
                        
# 2026-06-28 01:57:18  0.00B 
SHELL [/bin/bash -euxo pipefail -c]
                        
# 2026-06-28 01:57:18  0.00B 定义构建参数
ARG DEBIAN_FRONTEND=noninteractive
                        
# 2026-06-28 01:57:18  0.00B 定义构建参数
ARG CUDA_COMPAT_13_2_VERSION
                        
# 2026-06-28 01:57:18  0.00B 定义构建参数
ARG CUDNN_CUDA13_VERSION
                        
# 2026-06-28 01:57:18  0.00B 定义构建参数
ARG CUBLAS_CUDA13_VERSION
                        
# 2026-06-28 01:57:18  0.00B 定义构建参数
ARG NCCL_SONAME
                        
# 2026-06-28 01:57:18  0.00B 定义构建参数
ARG PYTHON_VERSION
                        
# 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:dc0bc459b8c1d59f84e945a4b77f65ea474778b58f4a95d3e3d1c97632daeb1d",
    "RepoTags": [
        "voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24"
    ],
    "RepoDigests": [
        "voipmonitor/vllm@sha256:64b94299abdd3bcf5bb5050ca91b378f9ee4e0b0eff4748375b95352371d7cb2",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/voipmonitor/vllm@sha256:5e6b4ccd3e6416f15966cb25e5a67cecd218a0b16f51fc8e2c61718b0b80247d"
    ],
    "Parent": "",
    "Comment": "buildkit.dockerfile.v0",
    "Created": "2026-08-03T02:10:27.848351156Z",
    "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",
            "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",
            "LOCAL_INFERENCE_CACHE_FINGERPRINT=vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174",
            "XDG_CACHE_HOME=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174",
            "VLLM_CACHE_ROOT=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/vllm",
            "VLLM_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/vllm",
            "TRITON_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/triton",
            "TORCHINDUCTOR_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/torchinductor",
            "TORCH_EXTENSIONS_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/torch_extensions",
            "FLASHINFER_WORKSPACE_BASE=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/flashinfer",
            "VLLM_FLASHINFER_AUTOTUNE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/flashinfer-autotune",
            "TVM_FFI_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tvm-ffi",
            "TVM_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tvm",
            "TILELANG_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tilelang",
            "TILELANG_TMP_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/tilelang/tmp",
            "CUTE_DSL_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cute-dsl",
            "B12X_CUTE_COMPILE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/b12x-cute",
            "SPARKINFER_COMPILE_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/sparkinfer/compile",
            "DG_JIT_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/deep-gemm",
            "MM_SPARSE_ATTN_AOT_CACHE=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/minfer/mm_sparse_attn",
            "MINFER_FMHA_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/minfer/fmha_sm100",
            "CUDA_CACHE_PATH=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cuda",
            "CUPY_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/cupy",
            "NUMBA_CACHE_DIR=/cache/jit/vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174/numba",
            "VLLM_EXL3_EXT_PATH=/opt/exllamav3",
            "VLLM_EXL3_ENCODER_SOURCE=/opt/exllamav3-python/exllamav3",
            "VLLM_EXL3_ENCODER_REVISION=704aefd743b390af4bd0fb429d1906f9b964c7d8",
            "VLLM_EXL3_ONLINE_CACHE_DIR=/cache/exl3-online",
            "VLLM_EXL3_ONLINE_CACHE_MODE=readwrite",
            "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_PROFILE_INCLUDE_ATTN=1",
            "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": "master",
            "local-inference.b12x.commit": "59216fa25f3d5fc9d4df2d052e02d05f763906e9",
            "local-inference.b12x.integration.base_commit": "59216fa25f3d5fc9d4df2d052e02d05f763906e9",
            "local-inference.b12x.integration.lock_sha256": "a6511d517e3d0a8d59afd10d4b339eb7faad29bc878bc922d4cf740217926d78",
            "local-inference.b12x.integration.prs": "106@35588fb570d23713bcb2a81749d0fdf3c4396ace,113@f3308fbdb6a5613dd3ba18bf6d482458c10684d2,112@fee1d784a798439de4e3462b0f0fc831491b5c7f",
            "local-inference.b12x.integration.tree": "2b9bf2a4d15770c0c23e19cc13a75843f2f0a995",
            "local-inference.b12x.patch_file": "releases/gilded-gnosis-v20-r24/sparkinfer/integration.patch",
            "local-inference.b12x.patch_sha256": "5fa88526cf7f468efdedcb484ed6b6a4ace065e3eb7d04a69bd46ed2570fd4b1",
            "local-inference.b12x.runtime_version": "source",
            "local-inference.cache.fingerprint": "vllm30038602b7-b12x59216fa25f-30dc52e3e06aa174",
            "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.cutlass_dsl.version": "4.6.0",
            "local-inference.deepgemm.branch": "a6b593d2826719dcf4892609af7b84ee23aaf32a",
            "local-inference.deepgemm.commit": "a6b593d2826719dcf4892609af7b84ee23aaf32a",
            "local-inference.exllamav3.branch": "a1-retile-sm120",
            "local-inference.exllamav3.commit": "704aefd743b390af4bd0fb429d1906f9b964c7d8",
            "local-inference.exllamav3.repo": "https://github.com/brandonmmusic-max/exllamav3.git",
            "local-inference.flashinfer.branch": "codex/sm120-dspark-stack-20260711",
            "local-inference.flashinfer.commit": "801d57a08958c13d375ddbb6be3be4808f48a708",
            "local-inference.instanttensor.branch": "25b3f268ea95b76bd03c825a1681872c9b615428",
            "local-inference.instanttensor.commit": "25b3f268ea95b76bd03c825a1681872c9b615428",
            "local-inference.instanttensor.repo": "https://github.com/voipmonitor/InstantTensor.git",
            "local-inference.launcher.branch": "76b1cf0350709910208e61285adc955e34655136",
            "local-inference.launcher.commit": "76b1cf0350709910208e61285adc955e34655136",
            "local-inference.launchers": "serve-gilded-gnosis.sh serve-fathomless-firmament.sh serve-glm52-v16.sh serve-glm52-v18.sh serve-glm52-v19.sh serve-glm52-hybrid-v17.sh serve-glm52-hybrid-v18.sh serve-glm52-hybrid-v19.sh glm52-dcp-prefill-policy.sh glm52-pcie-runtime-env.sh glm52-pcie-calibration.py glm52-lmcache-wrapper.sh",
            "local-inference.lmcache.commit": "9cebd405d0caf4bebe01d694b5a8bf4e3e354314",
            "local-inference.lmcache.integration.base_commit": "9cebd405d0caf4bebe01d694b5a8bf4e3e354314",
            "local-inference.lmcache.integration.lock_sha256": "573209e9258c4706b5f1f1e77c2e8a6de504a8068cfe80b6fcc76f5b21a56d79",
            "local-inference.lmcache.integration.prs": "7@7b7583aef55e98ba05a6c199e71601d78552e794,8@31c4175d2134518e5b43fe8f4a7d072df6043a13,9@c3b73de74d855f6263deea8b6c6c4dde87e6e5b9,10@59186798bb5859b7add7bcac81dff9f2fd4037ab,11@2cee5a2c1ef7e80028e94b1bc4bfbfc02e2c2a8d,12@baf1345106ebe568a116a35fdcee36bfec73da57,13@d4dce69bf90362953cf9ac9a86cccc186b3219dc,14@502b3c60f2e61625d6e32a31d0d725eedc620792,15@7729a9c75f2c29f14a27ea1e84ee6970950197ce,16@e6c2708fe8ae8a96d3c66e83bf4b548c734fbb13,17@f9639e2094f70955a1145bfd7219b6fdd39ee143",
            "local-inference.lmcache.integration.tree": "9a05c8818bae48d15b79c7e876418bb813c08cd0",
            "local-inference.lmcache.patch_sha256": "e9f5369ee134cbc697e77d9fd9d77219aceea5fabf2fe90e3b74542bbeb200ca",
            "local-inference.lmcache.ref": "release/v0.5.2-glm52-dcp-base",
            "local-inference.lmcache.repo": "https://github.com/local-inference-lab/LMCache.git",
            "local-inference.lmcache.version": "0.5.2+glm52dcp.4",
            "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.sparkinfer.branch": "master",
            "local-inference.sparkinfer.commit": "59216fa25f3d5fc9d4df2d052e02d05f763906e9",
            "local-inference.sparkinfer.integration.base_commit": "59216fa25f3d5fc9d4df2d052e02d05f763906e9",
            "local-inference.sparkinfer.integration.lock_sha256": "a6511d517e3d0a8d59afd10d4b339eb7faad29bc878bc922d4cf740217926d78",
            "local-inference.sparkinfer.integration.prs": "106@35588fb570d23713bcb2a81749d0fdf3c4396ace,113@f3308fbdb6a5613dd3ba18bf6d482458c10684d2,112@fee1d784a798439de4e3462b0f0fc831491b5c7f",
            "local-inference.sparkinfer.integration.tree": "2b9bf2a4d15770c0c23e19cc13a75843f2f0a995",
            "local-inference.sparkinfer.patch_file": "releases/gilded-gnosis-v20-r24/sparkinfer/integration.patch",
            "local-inference.sparkinfer.patch_sha256": "5fa88526cf7f468efdedcb484ed6b6a4ace065e3eb7d04a69bd46ed2570fd4b1",
            "local-inference.sparkinfer.runtime_version": "source",
            "local-inference.tokenspeed_mla.version": "0.1.8",
            "local-inference.torch.version": "2.12.0+cu132",
            "local-inference.triton_kernels.branch": "",
            "local-inference.triton_kernels.commit": "",
            "local-inference.tvm_ffi.version": "0.1.10",
            "local-inference.vllm.branch": "dev/gilded-gnosis",
            "local-inference.vllm.commit": "30038602b71395f481ef4a6edfe4fcf8551d9c15",
            "local-inference.vllm.integration.base_commit": "30038602b71395f481ef4a6edfe4fcf8551d9c15",
            "local-inference.vllm.integration.lock_sha256": "3de67581d0353e44240725ee8a0cd756c8d2a424e37120333bb1e91acc352ecc",
            "local-inference.vllm.integration.prs": "145@99bf3f13f13012b72d6d67e49572d686b45b2833,229@672aec40dc664de014bd1e92d9a2885d7658c37f,213@5f144274d6e092237f4f4f437b7e9d3aaa427157,214@e63a190ee1b48d051ccebce485de343dbe2b3505,217@c7ac22d5bfe923803cc923f2293085faac7d1dd0,218@3b89c7a113f92522aa07cc3271422c04e458d771,216@52201611719edf67f18bc0cbec17802880a4afb7,228@1702ed2d28342d84551539f8d7e2de637f7ca04e,230@9401f9d8916f3bdf34123d2d493a918b748c8132",
            "local-inference.vllm.integration.tree": "f5981f14b4d39979bc0d799c020d42002b707257",
            "local-inference.vllm.patch_file": "releases/gilded-gnosis-v20-r24/vllm/integration.patch",
            "local-inference.vllm.patch_sha256": "11e6c2b80b25d6433f01229bf9f0a0cafbfee8bd7125332a708398f93d8d77e9",
            "local-inference.vllm.patch_url": "",
            "local-inference.vllm.repo": "https://github.com/local-inference-lab/vllm.git",
            "local-inference.xgrammar.commit": "2ea71da4ccb997a06928c9fb69b99f330da56697",
            "local-inference.xgrammar.ref": "v0.2.5",
            "local-inference.xgrammar.repo": "https://github.com/mlc-ai/xgrammar.git",
            "local-inference.xgrammar.transformers5_compat": "1",
            "local-inference.xgrammar.version": "0.2.5",
            "maintainer": "NVIDIA CORPORATION \u003ccudatools@nvidia.com\u003e",
            "org.opencontainers.image.description": "Clean CUDA 13.2.1 vLLM+SparkInfer image with PyTorch 2.12 cu132, patched NCCL 2.30.4, FlashInfer git, DeepGEMM, SparkInfer and selected vLLM branch",
            "org.opencontainers.image.version": "24.04"
        },
        "Shell": [
            "/bin/bash",
            "-euxo",
            "pipefail",
            "-c"
        ]
    },
    "Architecture": "amd64",
    "Os": "linux",
    "Size": 25194001123,
    "GraphDriver": {
        "Data": {
            "LowerDir": "/var/lib/docker/overlay2/dac893a6d8aec3653098fe15e4c7d1ff0c3719c7b3151914f56f9f68f1cb2a1f/diff:/var/lib/docker/overlay2/5e961b835efccd2279a4e11694a9578791a5b5732fd6cf77f5dffbbb75185ae8/diff:/var/lib/docker/overlay2/ee66fc55846a12fb6656ed0f36c87b9c7ddb3a7c16c756391c6fa2ecb82738e6/diff:/var/lib/docker/overlay2/09eac470231c683722f5d5998ae2c8193543aab7c3f859d89f24a24b2f0275f1/diff:/var/lib/docker/overlay2/9bd05b9578bbcff3c179b133b3761b7af7f24ee8423ffe0ed3c5dab4b9403df0/diff:/var/lib/docker/overlay2/068d7a4ecd11b6b5b6745f0e56883498092ba303286d2b7354b93a8948ef7733/diff:/var/lib/docker/overlay2/b55021685047a3c15726c27a73c9aabac3084d044cf1b8312b22041b5a95319a/diff:/var/lib/docker/overlay2/eaf5f7d1649b70b7f4b29a7c86c08cd4542c3c0d98d0c5294c4884eae7941945/diff:/var/lib/docker/overlay2/77f2b70f0325e6316da9dc2f622cc47c6b6edaf761380ccb9f489369ae8bb878/diff:/var/lib/docker/overlay2/ea8e0bbea82e55c21c05cf6072bb25ffdc3068f356c8de493e4b31388ae6ea7e/diff:/var/lib/docker/overlay2/327384095f1c5138845115754dd2766a0e5b77639e4fe8613ab2e79d55e175c5/diff:/var/lib/docker/overlay2/6885ede81475defb265eda94e5a1bcc73271211caa07d7f03dc8c905dedd5955/diff:/var/lib/docker/overlay2/59d92afbfbf677a2a09ba43bc7085b3fb6dfaec8800cf75802bd8f791ddb334d/diff:/var/lib/docker/overlay2/83494ca397a16ef51d0c02c15a7cc376a555e0649e8d82f4f1cb5c40f2f9d917/diff:/var/lib/docker/overlay2/607902f687cf3d49b5466252d7f2716b2c47f435e106a914731ef26416be65f9/diff:/var/lib/docker/overlay2/5d0c071cf8ddaadf7e159d0aed4c3b24457bc6f8705fbb5fc2cba4a52f2d6eb2/diff:/var/lib/docker/overlay2/6f1003b7b1698d884dc2ede69ee76a7511d4b28abd43481aea1d083a9bbb539a/diff:/var/lib/docker/overlay2/9a73602ceec68bd1ddf133254aef8bf9e4fca2498d6d5be5daa1706df2bcfd69/diff:/var/lib/docker/overlay2/36fb6536fa85403457c7454a10c16292ecbfdcc6e1c1ae32a62f16a264fa2de6/diff:/var/lib/docker/overlay2/a0d9f150a5d51447e80e49e6e1a209675043fc9e310f42f6c974ff5c65a28bf4/diff:/var/lib/docker/overlay2/e13fa4342f41320e3b6613967ba414c9aae585f83b33e9049155acf3eb356602/diff:/var/lib/docker/overlay2/f0b7f22126e3cc2af36d7fa4086730982b59de2cda5bca0aa73bc089c8c7e9f8/diff:/var/lib/docker/overlay2/13802effd34ff11ddedf6dd3aa4e3edf947a4f03cdb8033a34561ec04909355c/diff:/var/lib/docker/overlay2/3dc5c8f1b8c49503f982c8faec928cc768495999ac68ec12a6d40b07120ddd53/diff:/var/lib/docker/overlay2/6be5189ffbbcce9b496a5d8d578325a0080353643049806f28cd62f99bec004c/diff:/var/lib/docker/overlay2/468e2425fa37c06bd375b9cdb4c98c51bb5182deff1bedd4544594c52dc372fc/diff:/var/lib/docker/overlay2/6d434aa2a63e84ea1991e69bc8f06ba8e99f655c6b86f92e28844758fe9fb94e/diff:/var/lib/docker/overlay2/fee6d41ad9013144f11382ef4587592c043f29b829fe11e9fbabe7a082abe83e/diff:/var/lib/docker/overlay2/61c75c473ffa989f17a74002178e640d5d8f57939065fbbc0340cccea23e6cd0/diff:/var/lib/docker/overlay2/5912198e582f191f37d26b9d65d8a85effe5094ee0780e677d1e7b6424266f07/diff:/var/lib/docker/overlay2/09af4e8ec4b8b9a0978623511fa8bd8f08ecbb6185ea78fb80306c60e5b1e3a0/diff:/var/lib/docker/overlay2/26df83c67415cf6a7759262284402e1f534edc38d6cd373e2b7658dba0f4b069/diff:/var/lib/docker/overlay2/68cd58ba704a89aa2f7dec33982931753e11b3738fc4b353f504fcbc020bfb21/diff:/var/lib/docker/overlay2/d38838908087793cbe52e352d4b2c9be5960f3df7a611c66c923685084a6be8f/diff:/var/lib/docker/overlay2/e7a42e175a8c0c11562e50787e04e7e851a149b37804348dd6d68d45a711acd3/diff",
            "MergedDir": "/var/lib/docker/overlay2/8d9371bc661d5a2474d59344d7c06a6de6caf8d2481f43d622b40c0ded7cd56c/merged",
            "UpperDir": "/var/lib/docker/overlay2/8d9371bc661d5a2474d59344d7c06a6de6caf8d2481f43d622b40c0ded7cd56c/diff",
            "WorkDir": "/var/lib/docker/overlay2/8d9371bc661d5a2474d59344d7c06a6de6caf8d2481f43d622b40c0ded7cd56c/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",
            "sha256:5f96e3ccf36c8c73c0c1e7f54e909e5b939967fdd1a842448b28ecbf22a712ca",
            "sha256:3405f007b9a97b4802dd53f5beac42c639c4de1bc4f6b852627dba530151b7ec",
            "sha256:3c93f297d3d3498f3875c891a6f43c5b9a2d0ad790351f207fc0d09b6c68945b",
            "sha256:452c3195deca8ea93f9f2bf065d5a8f390ec9020f304ddb185a49a8a1b5ad6bf",
            "sha256:b41aaa4835dd84a9c21f44359dece0e32f4deeea537b5f353a48191f83faeaa4",
            "sha256:fc56b182fb5d403da02482665c464d0bb562566b7e0f25df9836b92049eee953",
            "sha256:a7dd39a1251787cebebda0c2384ae54a1da0d9772228c8f600f13a6fb0f65689",
            "sha256:c7c7b4df56999ff7e37056fa5b9f88e69591897f1f9fe48227b56a04d5266755",
            "sha256:6893ffef4feff8204bef178b062ed446da293ae986a2f7de3d84dd9631726a1c",
            "sha256:50dde3cae645066251d236374e39d1bfdf6ea11cb42e555e9456d238753fd117",
            "sha256:776ce556238893d3018307f8f2a4f99ac6262d7cb1f1a9285d534e3cc8d8433a",
            "sha256:9e68a9eb3d002d5a718e66708fb5e364cfb22ea8818e79c64552efe916801686",
            "sha256:58028e3a70605a276f771402ea7b5a0a8230422abc60d47deb6d5a65f5c70111",
            "sha256:5fa688a8eefd9d7aa64636a0131d21611fe124f3bf1d0ab312e8765a286c6d6c",
            "sha256:9f0ac29e53919bd539cba1c75adae968be54e5c82cf0647e7efb10ecd36b6d27",
            "sha256:9287d01902e202c60c054344120092f771864bde30b3d7af607cf0636fb2262b",
            "sha256:2b9c0d2d347b03eac36e114801b2c7bbdce081b59a7046a4f45ee0d7d5f86c46",
            "sha256:6a6fbc513b1bc392e0fed98b20aec8fd4ca4251c879524366c2c2d965214bc09",
            "sha256:d0985ce11992b61e777507ddb30e3070d9a427d1d0d55da5adb321d5b0343c2e",
            "sha256:d8f6cb85a26a7633969a702dc4c2d17bdf17c7e9b6caf0da37baadbad6b9145d",
            "sha256:0da1a591ed9ff91871e2e70a42981669ddeb2fafbd684896bdfef46e248d0939"
        ]
    },
    "Metadata": {
        "LastTagTime": "2026-08-31T00:11:02.078269274+08:00"
    }
}

更多版本

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

linux/amd64 docker.io22.92GB2026-08-30 01:03
26

docker.io/voipmonitor/vllm:gilded-gnosis-v20-vllmf5981f1-si2b9bf2a-fi801d57a-cu132-20260803-r24

linux/amd64 docker.io25.19GB2026-08-31 00:22
14
0 点赞 0 收藏镜像 Run 生成 Docker Run 命令 K8s 生成 Kubernetes YAML 回到顶部
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 ×