Namespace
lmsysorg
Image / Tag
sglang:dev-cu12-Ling-3.0-tiny
Content Digest
sha256:3076afb2abf0e927663e9b7a475277a8b1e87cbd7ae175e22a54ade33aef3e89
Details
Created

2026-08-11 07:53:06 UTC

Size

17.1 GB

Content Digest
Labels
  • ai.sglang.build.commit
    8ba213fc67550aea80dfbd5ef66f4574ec25e653
  • ai.sglang.build.url
    https://github.com/sgl-project/sglang/actions/runs/31468844644
  • ai.sglang.image.tag
    lmsysorg/sglang:dev-cu12-Ling-3.0-flash
  • com.nvidia.cudnn.version
    9.10.2.21-1
  • maintainer
    NVIDIA CORPORATION <cudatools@nvidia.com>
  • org.opencontainers.image.revision
    8ba213fc67550aea80dfbd5ef66f4574ec25e653
  • org.opencontainers.image.source
    https://github.com/sgl-project/sglang
  • org.opencontainers.image.url
    https://github.com/sgl-project/sglang/actions/runs/31468844644
  • org.opencontainers.image.version
    lmsysorg/sglang:dev-cu12-Ling-3.0-flash

Environment
CUDA_HOME

/usr/local/cuda

CUDA_VERSION

12.9.2

DEBIAN_FRONTEND

noninteractive

FLASHINFER_VERSION

0.6.15.post1

GDRCOPY_HOME

/usr/src/gdrdrv-2.5.1/

LANG

en_US.UTF-8

LANGUAGE

en_US:en

LC_ALL

en_US.UTF-8

LD_LIBRARY_PATH

/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64

LIBRARY_PATH

/usr/local/cuda/lib64/stubs

NCCL_VERSION

2.27.3-1

NVARCH

x86_64

NVIDIA_DRIVER_CAPABILITIES

compute,utility

NVIDIA_PRODUCT_NAME

CUDA

NVIDIA_REQUIRE_CUDA

cuda>=12.9 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>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=560,driver<561 brand=grid,driver>=560,driver<561 brand=tesla,driver>=560,driver<561 brand=nvidia,driver>=560,driver<561 brand=quadro,driver>=560,driver<561 brand=quadrortx,driver>=560,driver<561 brand=nvidiartx,driver>=560,driver<561 brand=vapps,driver>=560,driver<561 brand=vpc,driver>=560,driver<561 brand=vcs,driver>=560,driver<561 brand=vws,driver>=560,driver<561 brand=cloudgaming,driver>=560,driver<561 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 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

NVIDIA_VISIBLE_DEVICES

all

NV_CUDA_CUDART_DEV_VERSION

12.9.79-1

NV_CUDA_CUDART_VERSION

12.9.79-1

NV_CUDA_LIB_VERSION

12.9.2-1

NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE

cuda-nsight-compute-12-9=12.9.2-1

NV_CUDA_NSIGHT_COMPUTE_VERSION

12.9.2-1

NV_CUDNN_PACKAGE

libcudnn9-cuda-12=9.10.2.21-1

NV_CUDNN_PACKAGE_DEV

libcudnn9-dev-cuda-12=9.10.2.21-1

NV_CUDNN_PACKAGE_DEV_HEADERS

libcudnn9-headers-cuda-12=9.10.2.21-1

NV_CUDNN_PACKAGE_NAME

libcudnn9-cuda-12

NV_CUDNN_VERSION

9.10.2.21-1

NV_LIBCUBLAS_DEV_PACKAGE

libcublas-dev-12-9=12.9.2.10-1

NV_LIBCUBLAS_DEV_PACKAGE_NAME

libcublas-dev-12-9

NV_LIBCUBLAS_DEV_VERSION

12.9.2.10-1

NV_LIBCUBLAS_PACKAGE

libcublas-12-9=12.9.2.10-1

NV_LIBCUBLAS_PACKAGE_NAME

libcublas-12-9

NV_LIBCUBLAS_VERSION

12.9.2.10-1

NV_LIBCUSPARSE_DEV_VERSION

12.5.10.65-1

NV_LIBCUSPARSE_VERSION

12.5.10.65-1

NV_LIBNCCL_DEV_PACKAGE

libnccl-dev=2.27.3-1+cuda12.9

NV_LIBNCCL_DEV_PACKAGE_NAME

libnccl-dev

NV_LIBNCCL_DEV_PACKAGE_VERSION

2.27.3-1

NV_LIBNCCL_PACKAGE

libnccl2=2.27.3-1+cuda12.9

NV_LIBNCCL_PACKAGE_NAME

libnccl2

NV_LIBNCCL_PACKAGE_VERSION

2.27.3-1

NV_LIBNPP_DEV_PACKAGE

libnpp-dev-12-9=12.4.1.87-1

NV_LIBNPP_DEV_VERSION

12.4.1.87-1

NV_LIBNPP_PACKAGE

libnpp-12-9=12.4.1.87-1

NV_LIBNPP_VERSION

12.4.1.87-1

NV_NVML_DEV_VERSION

12.9.79-1

NV_NVPROF_DEV_PACKAGE

cuda-nvprof-12-9=12.9.79-1

NV_NVPROF_VERSION

12.9.79-1

NV_NVTX_VERSION

12.9.79-1

PATH

/root/.cargo/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/nvidia/bin

SGLANG_BUILD_COMMIT

8ba213fc67550aea80dfbd5ef66f4574ec25e653

SGLANG_BUILD_URL

https://github.com/sgl-project/sglang/actions/runs/31468844644

SGLANG_IMAGE_TAG

lmsysorg/sglang:dev-cu12-Ling-3.0-flash

SGLANG_TRTLLM_GEN_MOE_CUBIN_POOL

/opt/trtllm_gen_moe_cubin_pool


Layers

[#000] sha256:b40150c1c2717d324cdb17278c8efdfa4dfcd2ffe083e976f0bcedf31115f081 - 0.16% (28.4 MB)

[#001] sha256:e11d6318d8bef8a1b99dd393ac6845ebf117c9f68c81624692e00d576432f469 - 0.03% (4.93 MB)

[#002] sha256:761ae241e4c2488ad2b48982d0d7583c52b904b7ee952a117fe56567ad8aba95 - 0.0% (367 KB)

[#003] sha256:020c9394f8c3b8371d967e0fb27b856ca0cb12fb02e21bd608710fa24fa62035 - 0.56% (98.5 MB)

[#004] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#005] sha256:adae47ab9b6f583ae88e5f7946860091639ec39b96e204f851e932fcdc4c1351 - 0.0% (184 Bytes)

[#006] sha256:1dca0c9bc5f79f4cedcd68523b7682fb46a1b1546351bc67289e9440608187fd - 0.0% (6.72 KB)

[#007] sha256:21e333934e1348446c62aade4f760bd53540d289d290c498673bbf81f4e74fd4 - 12.56% (2.15 GB)

[#008] sha256:2a5703d89dc74187998c37f1e8668a158c195961801189556e347d287b6a47a9 - 0.0% (58.6 KB)

[#009] sha256:62b81882c317bac2e1b871ecc97077fe29a2b1e6c12570091270f3d43cf39542 - 0.0% (1.64 KB)

[#010] sha256:f34b8dc4eed321a3c34de1b22718a453ac047c27ab3b5e4b1da290c98417ebb8 - 0.0% (1.48 KB)

[#011] sha256:07aef1bccd93742ff5642ff536f841766136661cb7ec2cc47cf9f893153bce61 - 17.63% (3.01 GB)

[#012] sha256:d0b10745023cb53ed62a6c37d7b43be906bc510d9a62671138f5995f4640b317 - 0.0% (222 Bytes)

[#013] sha256:52598b318befda35304f718798d7db6a4c354007ba88a5195ef987499190371b - 0.0% (87.9 KB)

[#014] sha256:ffd9010dbbb8d4573374f7d4d71c1ebd021dc23e19048d0e13ce7a9738370715 - 3.88% (679 MB)

[#015] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#016] sha256:8610c6c047f723690b545250a84f39a295811b6210ebe3b5f75b6cc565d2d93d - 0.79% (138 MB)

[#017] sha256:81229a0c97162914505fc679d7f5d9b6d9b8f8572e067667caaa13afcff04973 - 2.34% (410 MB)

[#018] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#019] sha256:2ec760d502020476403fe6fc48b96c94a54e3174341435725fa5ded92a484abc - 0.01% (1.43 MB)

[#020] sha256:dd71531b08fb94880edb9467298c7fe9901a69ba1b00a44a364e2463735f5e09 - 0.0% (189 Bytes)

[#021] sha256:611be6b9e2a72caf9b51194a0a404e5facbc45a913df759f83516c9ff83a61ea - 0.0% (846 KB)

[#022] sha256:afe709a446950c5973d593783c895b00041efa96d62da5b61e9fb8c56319e35b - 0.0% (103 Bytes)

[#023] sha256:282d4d244fe8850f6905dff01cd2cc8d6764cbc1c5e340b6fb128502ad02d9cd - 1.19% (208 MB)

[#024] sha256:a47bc09311bc926a70d80a332c66deb240e8c1c761318b1eeeb5276871898938 - 2.34% (410 MB)

[#025] sha256:15395de4fb016088eab139d593c7f66931d2374787ced34382f8ffa0d9ce3b13 - 0.0% (2.86 KB)

[#026] sha256:aa82a11ad7a63c5d99948e889af4f6a6c47b6062ba906bcafc0623ab6cfce2b8 - 0.0% (20.2 KB)

[#027] sha256:729d9438482e978dfef142f696722d73c545d3e4833f72e299fd1870e7dfbbf8 - 0.0% (42.7 KB)

[#028] sha256:f23bf157571927e4a151147f7644f7346e207cae5edf6ecad93c34f60801789a - 0.0% (2.22 KB)

[#029] sha256:c4cb29201958f35a79d6f5764d56bce07dd272240fd9f74120630741bb868811 - 29.45% (5.03 GB)

[#030] sha256:f83634f26106d8cdf1677df24e1c2289ef644c09fbdadb8cd6de44453efec330 - 0.0% (66.7 KB)

[#031] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#032] sha256:729c0e5ab98be0badfa628bc6c48998141451c153264d93dd6dae7bb7d26db0c - 0.21% (36.6 MB)

[#033] sha256:bf4688f3e16bbec95c360fa2799ac47481b73509e6edcc7bdc6f0abc0b859914 - 0.21% (37.5 MB)

[#034] sha256:308507a2233934b508f4701c723c694af884292ffe3c8c076253e5b0dae093d2 - 14.51% (2.48 GB)

[#035] sha256:982f72949dfc751c3107e755f9f63ad3186a4e463da46c3bb886c007cabde79f - 0.0% (51.6 KB)

[#036] sha256:feb71d32b74f673d70065082d205cd58fd8de9c95555841b7d9f61b2dbe5ac13 - 0.0% (174 KB)

[#037] sha256:21c835c4f569fbb71355d3eb0cb0032ed4732a0cce1be205fd601d27fc595ba0 - 0.17% (30.6 MB)

[#038] sha256:67a738ed75f39576ec7558332bde564b1011bde5d2e5527aa7faaa3c4d2f4c15 - 0.0% (13.1 KB)

[#039] sha256:52e5fc0e24a20a0ede285b09dbbee553626c5e1c8a640054ff19b12df071f889 - 0.01% (1.7 MB)

[#040] sha256:d951b90db0ca37d62b30e6bbe61af4941c5bc2685a9694114a4a8132cb4cec8b - 0.46% (81 MB)

[#041] sha256:4ba7b3719fca7d179a6d920fbe69e1bb3b04fc8b9bfb52b034eeb706b56bb66d - 0.01% (957 KB)

[#042] sha256:02a60ff03c806791f6e6092cc1960a428b6e04d523e3b6e3de53a190516ff906 - 0.04% (6.92 MB)

[#043] sha256:a6a181aa12f2c213a86cb1b92c4c8225391dd065ca353109eaf97af1eda8ffea - 0.04% (7.52 MB)

[#044] sha256:4893e2ada84cbbfa6d8aac39b2b13a721bdad9d5bdbcd98f27b53885e7e9333e - 0.04% (7.1 MB)

[#045] sha256:c000adce123ef0ba09b0ec040667241ee4b04333d4d1e7ec5737b1ec92707aed - 0.01% (1.93 MB)

[#046] sha256:878980f64ff3499a2eb79a10acd77fab24c09e0cb39ce88404437d24bf8093ec - 0.01% (1.9 MB)

[#047] sha256:1396c5f81379f966995e553b9b1e3d478955a446e8fa88b30efc50434b88adaa - 0.04% (7.05 MB)

[#048] sha256:6da885b817c0338c0b435aed80ebac86136f920a1fcdcdb2faa8c9ec85066e0c - 0.37% (63.9 MB)

[#049] sha256:1a69e8d5d18caa7afdd5e5b07f0709904e69f0b294edf6547df9cce6e0d9e1a4 - 1.21% (211 MB)

[#050] sha256:3ed5cd6289042930b11a562f7ced8d03b1faf009985dad5064e471de3e535e6e - 0.3% (52 MB)

[#051] sha256:0b863105b946c8182578a859ca3ba3f2e070205b5d913fd686c431f5814764ee - 1.33% (232 MB)

[#052] sha256:85e01041e443a26989f60f1fa8757ced1e1cd0ebace083387e291cfc18830757 - 0.63% (110 MB)

[#053] sha256:de73caeaecd6cc4d6dc9e6358858ac139a98d600544e99ec610354fae7726822 - 0.49% (85.4 MB)

[#054] sha256:e8d9711ede88706351806af4ce496755e8fc3e7c311a2bba3e97e99e570918fd - 0.0% (442 Bytes)

[#055] sha256:939070748b1d2e9b0fb63ad7e32a64e3d98757a354980cbf17becae13b8b1e93 - 0.0% (722 Bytes)

[#056] sha256:122f99703bcd805f2c08e6d854b2f8752b010792ffb66c3605fd7c29fb44c569 - 0.0% (507 Bytes)

[#057] sha256:f366d0b6f95f9d2a93ff6d7140e266e9f963aa0cdabb1784a42d35e63c7bdcd3 - 0.0% (517 Bytes)

[#058] sha256:67db1caba7b6d62dd1d78e96be9cde1a23978917a96428e317db84c042a36404 - 0.0% (393 Bytes)

[#059] sha256:62b7ce6486841e0b47225a90f1862f215565fd7f0055effc1def21d8ca9ffc2e - 0.08% (14.3 MB)

[#060] sha256:c35750151d527bd5a4a294b432bfbdfdd336035291dc1a6c5fff876de21573f1 - 0.02% (3.1 MB)

[#061] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#062] sha256:cd66f44cdd87eb68b8f8a6f3dd2a4004a4b34237669df9f32f648e08cc1f9ebd - 1.9% (332 MB)

[#063] sha256:9deeba6fa11fee6213c49d1c2891d5a3b07619aab6d18b18500da14077aa8857 - 1.9% (332 MB)

[#064] sha256:95ab306e9bbeb7a6067e1d7e8ec38c7ba535260e9f13befdb826b9ee576627fc - 4.72% (825 MB)

[#065] sha256:3bdac2b02eff82a97b3c01807e5236ab1a5ac4fca49ab5c13584e2e10491de08 - 0.1% (17 MB)

[#066] sha256:cd1df4f79e4a8cfcd4ee8a4b47fe7df928d63db430e68c646b8f5fc461a755d6 - 0.1% (16.9 MB)

[#067] sha256:77116ef66a05891eddb584e3267ad72748770c261efe4dda725b32cffafbe719 - 0.15% (26.7 MB)

[#068] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)


History
2026-04-10 06:49:15 UTC

/bin/sh -c #(nop) ARG RELEASE

2026-04-10 06:49:15 UTC

/bin/sh -c #(nop) ARG LAUNCHPAD_BUILD_ARCH

2026-04-10 06:49:15 UTC

/bin/sh -c #(nop) LABEL org.opencontainers.image.version=24.04

2026-04-10 06:49:17 UTC

/bin/sh -c #(nop) ADD file:8ce1caf246e7c778bca84c516d02fd4e83766bb2c530a0fffa8a351b560a2728 in /

2026-04-10 06:49:18 UTC

/bin/sh -c #(nop) CMD ["/bin/bash"]

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

ENV NVARCH=x86_64

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_REQUIRE_CUDA=cuda>=12.9 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>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=560,driver<561 brand=grid,driver>=560,driver<561 brand=tesla,driver>=560,driver<561 brand=nvidia,driver>=560,driver<561 brand=quadro,driver>=560,driver<561 brand=quadrortx,driver>=560,driver<561 brand=nvidiartx,driver>=560,driver<561 brand=vapps,driver>=560,driver<561 brand=vpc,driver>=560,driver<561 brand=vcs,driver>=560,driver<561 brand=vws,driver>=560,driver<561 brand=cloudgaming,driver>=560,driver<561 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 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

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_CUDART_VERSION=12.9.79-1

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

ARG IMAGE_UPGRADE_ID=0

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c echo "IMAGE_UPGRADE_ID=${IMAGE_UPGRADE_ID}" && apt-get update && apt-get upgrade -y && apt-get install -y --no-install-recommends gnupg2 curl ca-certificates && curl -fsSLO https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/${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

2026-05-18 15:59:07 UTC (buildkit.dockerfile.v0)

ENV CUDA_VERSION=12.9.2

2026-05-18 15:59:18 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends cuda-cudart-12-9=${NV_CUDA_CUDART_VERSION} cuda-toolkit-12-9-config-common=${NV_CUDA_CUDART_VERSION} cuda-toolkit-12-config-common=${NV_CUDA_CUDART_VERSION} cuda-toolkit-config-common=${NV_CUDA_CUDART_VERSION} && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 15:59:42 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c apt-get update && if apt-cache policy cuda-compat-12-9 2>/dev/null | grep -q "Candidate:"; then apt-get install -y --no-install-recommends cuda-compat-12-9; fi && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 15:59:47 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c apt-get update && if apt-cache policy cuda-compat-orin-12-9 2>/dev/null | grep -q "Candidate:"; then apt-get install -y --no-install-recommends cuda-compat-orin-12-9; fi && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 15:59:47 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c echo "/usr/local/cuda/lib64" >> /etc/ld.so.conf.d/nvidia.conf # buildkit

2026-05-18 15:59:47 UTC (buildkit.dockerfile.v0)

ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin

2026-05-18 15:59:47 UTC (buildkit.dockerfile.v0)

ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64

2026-05-18 15:59:47 UTC (buildkit.dockerfile.v0)

COPY NGC-DL-CONTAINER-LICENSE / # buildkit

2026-05-18 15:59:47 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_VISIBLE_DEVICES=all

2026-05-18 15:59:47 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_LIB_VERSION=12.9.2-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_NVTX_VERSION=12.9.79-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_VERSION=12.4.1.87-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_PACKAGE=libnpp-12-9=12.4.1.87-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUSPARSE_VERSION=12.5.10.65-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_PACKAGE_NAME=libcublas-12-9

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_VERSION=12.9.2.10-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_PACKAGE=libcublas-12-9=12.9.2.10-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_PACKAGE_NAME=libnccl2

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_PACKAGE_VERSION=2.27.3-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NCCL_VERSION=2.27.3-1

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_PACKAGE=libnccl2=2.27.3-1+cuda12.9

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends cuda-libraries-12-9=${NV_CUDA_LIB_VERSION} ${NV_LIBNPP_PACKAGE} cuda-nvtx-12-9=${NV_NVTX_VERSION} libcusparse-12-9=${NV_LIBCUSPARSE_VERSION} ${NV_LIBCUBLAS_PACKAGE} ${NV_LIBNCCL_PACKAGE} && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=amd64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_PACKAGE_NAME} ${NV_LIBNCCL_PACKAGE_NAME} # buildkit

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

COPY entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

COPY nvidia_entrypoint.sh /opt/nvidia/ # buildkit

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_PRODUCT_NAME=CUDA

2026-05-18 16:26:36 UTC (buildkit.dockerfile.v0)

ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_LIB_VERSION=12.9.2-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_CUDART_DEV_VERSION=12.9.79-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_NVML_DEV_VERSION=12.9.79-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUSPARSE_DEV_VERSION=12.5.10.65-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_DEV_VERSION=12.4.1.87-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_DEV_PACKAGE=libnpp-dev-12-9=12.4.1.87-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_DEV_VERSION=12.9.2.10-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-12-9

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-12-9=12.9.2.10-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_NSIGHT_COMPUTE_VERSION=12.9.2-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-12-9=12.9.2-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_NVPROF_VERSION=12.9.79-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_NVPROF_DEV_PACKAGE=cuda-nvprof-12-9=12.9.79-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_DEV_PACKAGE_NAME=libnccl-dev

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_DEV_PACKAGE_VERSION=2.27.3-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NCCL_VERSION=2.27.3-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_DEV_PACKAGE=libnccl-dev=2.27.3-1+cuda12.9

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends cuda-cudart-dev-12-9=${NV_CUDA_CUDART_DEV_VERSION} cuda-command-line-tools-12-9=${NV_CUDA_LIB_VERSION} cuda-minimal-build-12-9=${NV_CUDA_LIB_VERSION} cuda-libraries-dev-12-9=${NV_CUDA_LIB_VERSION} cuda-nvml-dev-12-9=${NV_NVML_DEV_VERSION} ${NV_NVPROF_DEV_PACKAGE} ${NV_LIBNPP_DEV_PACKAGE} libcusparse-dev-12-9=${NV_LIBCUSPARSE_DEV_VERSION} ${NV_LIBCUBLAS_DEV_PACKAGE} ${NV_LIBNCCL_DEV_PACKAGE} ${NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE} && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

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

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=amd64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_DEV_PACKAGE_NAME} ${NV_LIBNCCL_DEV_PACKAGE_NAME} # buildkit

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV LIBRARY_PATH=/usr/local/cuda/lib64/stubs

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_VERSION=9.10.2.21-1

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-12

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE=libcudnn9-cuda-12=9.10.2.21-1

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-12=9.10.2.21-1

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE_DEV_HEADERS=libcudnn9-headers-cuda-12=9.10.2.21-1

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

LABEL com.nvidia.cudnn.version=9.10.2.21-1

2026-05-18 17:26:13 UTC (buildkit.dockerfile.v0)

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

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH=amd64

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG BRANCH_TYPE=local

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG SGL_KERNEL_VERSION=0.4.6.post1

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG SGL_VERSION

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG SGL_DEEP_GEMM_VERSION=0.1.5.post2

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG USE_LATEST_SGLANG=0

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG GDRCOPY_VERSION=2.5.1

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG PIP_DEFAULT_INDEX

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG UBUNTU_MIRROR

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG GITHUB_ARTIFACTORY=github.com

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG INSTALL_FLASHINFER_JIT_CACHE=1

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG FLASHINFER_VERSION=0.6.15.post1

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_VERSION=0.3.12.post1

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG MSCCLPP_VERSION=sglang-v0.9.1

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ENV DEBIAN_FRONTEND=noninteractive CUDA_HOME=/usr/local/cuda GDRCOPY_HOME=/usr/src/gdrdrv-2.5.1/ FLASHINFER_VERSION=0.6.15.post1 SGLANG_TRTLLM_GEN_MOE_CUBIN_POOL=/opt/trtllm_gen_moe_cubin_pool

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/nvidia/bin LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64

2026-08-11 07:26:34 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c if [ -n "$UBUNTU_MIRROR" ]; then sed -i "s|http://.*archive.ubuntu.com|$UBUNTU_MIRROR|g" /etc/apt/sources.list && sed -i "s|http://.*security.ubuntu.com|$UBUNTU_MIRROR|g" /etc/apt/sources.list; fi # buildkit

2026-08-11 07:27:07 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c apt update && apt install -y --no-install-recommends wget software-properties-common && apt install -y --no-install-recommends python3.12-full python3.12-dev && update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.12 2 && update-alternatives --set python3 /usr/bin/python3.12 && wget -q https://bootstrap.pypa.io/get-pip.py && python3 get-pip.py --break-system-packages && rm get-pip.py && python3 -m pip config set global.break-system-packages true && cd /usr/lib/python3/dist-packages/ && ln -s apt_pkg.cpython-312-*-linux-gnu.so apt_pkg.so # buildkit

2026-08-11 07:27:54 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c apt-get update && apt-get install -y --no-install-recommends ca-certificates software-properties-common netcat-openbsd kmod unzip openssh-server curl wget lsof locales build-essential cmake perl patch patchelf ccache git-lfs libopenmpi-dev libnuma1 libnuma-dev numactl ffmpeg libibverbs-dev libibverbs1 libibumad3 librdmacm1 libnl-3-200 libnl-route-3-200 libnl-route-3-dev libnl-3-dev ibverbs-providers infiniband-diags perftest libgoogle-glog-dev libgtest-dev libjsoncpp-dev libunwind-dev libboost-all-dev libssl-dev libgrpc-dev libgrpc++-dev libprotobuf-dev protobuf-compiler protobuf-compiler-grpc pybind11-dev libhiredis-dev libcurl4-openssl-dev libczmq4 libczmq-dev libfabric-dev linux-libc-dev devscripts debhelper fakeroot dkms check libsubunit0 libsubunit-dev && ln -sf /usr/bin/python3.12 /usr/bin/python && rm -rf /var/lib/apt/lists/* && apt-get clean # buildkit

2026-08-11 07:27:55 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c if [ -n "${PIP_DEFAULT_INDEX}" ]; then python3 -m pip config set global.index-url ${PIP_DEFAULT_INDEX}; fi # buildkit

2026-08-11 07:28:46 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c mkdir -p /tmp/gdrcopy && cd /tmp && curl --retry 3 --retry-delay 2 -fsSL -o v${GDRCOPY_VERSION}.tar.gz https://${GITHUB_ARTIFACTORY}/NVIDIA/gdrcopy/archive/refs/tags/v${GDRCOPY_VERSION}.tar.gz && tar -xzf v${GDRCOPY_VERSION}.tar.gz && rm v${GDRCOPY_VERSION}.tar.gz && cd gdrcopy-${GDRCOPY_VERSION}/packages && CUDA=/usr/local/cuda ./build-deb-packages.sh && dpkg -i gdrdrv-dkms_*.deb libgdrapi_*.deb gdrcopy-tests_*.deb gdrcopy_*.deb && cd / && rm -rf /tmp/gdrcopy # buildkit

2026-08-11 07:28:46 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c ln -sf /usr/lib/$(uname -m)-linux-gnu/libmlx5.so.1 /usr/lib/$(uname -m)-linux-gnu/libmlx5.so # buildkit

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c locale-gen en_US.UTF-8 # buildkit

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

ENV LANG=en_US.UTF-8 LANGUAGE=en_US:en LC_ALL=en_US.UTF-8

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

ARG CUDA_VERSION=12.9.2

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

ARG SGL_KERNEL_VERSION

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

ARG GITHUB_ARTIFACTORY

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace

2026-08-11 07:28:47 UTC (buildkit.dockerfile.v0)

ENV PATH=/root/.cargo/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/nvidia/bin

2026-08-11 07:29:16 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c curl --proto '=https' --tlsv1.2 --retry 3 --retry-delay 2 -sSf https://sh.rustup.rs | sh -s -- -y --no-modify-path --profile minimal && rustc --version && cargo --version # buildkit

2026-08-11 07:29:43 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --upgrade pip setuptools wheel html5lib six && case "$CUDA_VERSION" in 12.6.3) CUINDEX=126 ;; 12.9.2) CUINDEX=129 ;; 13.0.3) CUINDEX=130 ;; *) echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1 ;; esac && if [ "$CUDA_VERSION" = "12.6.3" ]; then python3 -m pip install https://${GITHUB_ARTIFACTORY}/sgl-project/whl/releases/download/v${SGL_KERNEL_VERSION}/sglang_kernel-${SGL_KERNEL_VERSION}+cu124-cp310-abi3-manylinux2014_$(uname -m).whl --force-reinstall --no-deps ; elif [ "$CUDA_VERSION" = "12.9.2" ]; then python3 -m pip install https://${GITHUB_ARTIFACTORY}/sgl-project/whl/releases/download/v${SGL_KERNEL_VERSION}/sglang_kernel-${SGL_KERNEL_VERSION}+cu129-cp310-abi3-manylinux2014_$(uname -m).whl --force-reinstall --no-deps ; elif [ "$CUDA_VERSION" = "13.0.3" ]; then python3 -m pip install sglang-kernel==${SGL_KERNEL_VERSION} --force-reinstall --no-deps ; else echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1 ; fi # buildkit

2026-08-11 07:29:43 UTC (buildkit.dockerfile.v0)

COPY python/pyproject.toml /tmp/sglang_deps/python/pyproject.toml # buildkit

2026-08-11 07:29:43 UTC (buildkit.dockerfile.v0)

COPY rust/sglang-grpc /tmp/sglang_deps/rust/sglang-grpc # buildkit

2026-08-11 07:29:43 UTC (buildkit.dockerfile.v0)

COPY rust/sglang-mm /tmp/sglang_deps/rust/sglang-mm # buildkit

2026-08-11 07:29:43 UTC (buildkit.dockerfile.v0)

COPY proto /tmp/sglang_deps/proto # buildkit

2026-08-11 07:34:55 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c case "$CUDA_VERSION" in 12.6.3) CUINDEX=126 ;; 12.9.2) CUINDEX=129 ;; 13.0.3) CUINDEX=130 ;; *) echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1 ;; esac && cd /tmp/sglang_deps/python && mkdir -p sglang && touch sglang/__init__.py && echo '__version__ = "0.0.0"' > sglang/version.py && touch README.md && touch LICENSE && SGL_DEEP_EP_VERSION="$(sed -n 's/^[[:space:]]*"sgl-deep-ep==\([^"]*\)",/\1/p' pyproject.toml)" && test -n "${SGL_DEEP_EP_VERSION}" && if [ "${CUDA_VERSION%%.*}" = "12" ]; then python3 -m pip install "sgl-deep-ep==${SGL_DEEP_EP_VERSION}+cu129" --index-url "https://docs.sglang.ai/whl/cu129/" --no-deps; fi && if [ "${CUDA_VERSION%%.*}" = "12" ]; then sed -i 's/cuda-python>=13\.0/cuda-python>=12,<13/' pyproject.toml && sed -i 's/flashinfer_python\[cu13\]/flashinfer_python[cu12]/' pyproject.toml && sed -i 's/nvidia-cutlass-dsl\[cu13\]/nvidia-cutlass-dsl/' pyproject.toml; fi && python3 -m pip install --extra-index-url https://download.pytorch.org/whl/cu${CUINDEX} ".[${BUILD_TYPE}]" && if [ "${CUDA_VERSION%%.*}" = "12" ]; then pip list --format=freeze | awk -F'==' '/-cu13(==|$)/ {print $1}' | xargs -r python3 -m pip uninstall -y && python3 -m pip install --index-url https://download.pytorch.org/whl/cu${CUINDEX} torch==2.13.0 torchvision==0.28.0 torchaudio==2.11.0 --force-reinstall; python3 -m pip install https://github.com/sgl-project/whl/releases/download/v${SGL_DEEP_GEMM_VERSION}/sgl_deep_gemm-${SGL_DEEP_GEMM_VERSION}+cu129-py3-none-manylinux2014_$(uname -m).whl --force-reinstall; fi && cd /sgl-workspace && rm -rf /tmp/sglang_deps && pip freeze | grep -v "^sglang==" > /sgl-workspace/constraints.txt # buildkit

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --ignore-installed --no-deps distro # buildkit

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG BRANCH_TYPE=local

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG CUDA_VERSION=12.9.2

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG SGL_VERSION

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG USE_LATEST_SGLANG

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG GITHUB_ARTIFACTORY

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_VERSION

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_COMPILE_ARG

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

ARG MSCCLPP_VERSION

2026-08-11 07:35:01 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace

2026-08-11 07:38:46 UTC (buildkit.dockerfile.v0)

COPY /wheels /tmp/wheels/hpc-ops # buildkit

2026-08-11 07:38:47 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c if ls /tmp/wheels/hpc-ops/*.whl >/dev/null 2>&1; then pip install --no-deps /tmp/wheels/hpc-ops/*.whl; fi && rm -rf /tmp/wheels/hpc-ops # buildkit

2026-08-11 07:38:59 UTC (buildkit.dockerfile.v0)

COPY /flashinfer_jit_output/ /usr/local/lib/python3.12/dist-packages/ # buildkit

2026-08-11 07:38:59 UTC (buildkit.dockerfile.v0)

COPY docker/kimi_k3/flashinfer-perkz-dcp-0.6.15.txt /tmp/flashinfer-perkz-dcp-0.6.15.txt # buildkit

2026-08-11 07:39:02 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c FLASHINFER_DCP_PATCH=/tmp/flashinfer-perkz-dcp-0.6.15.txt && FLASHINFER_SITE_PACKAGES="$(python3 -c 'from pathlib import Path; import flashinfer; print(Path(flashinfer.__file__).resolve().parent.parent)')" && sed '/^diff --git a\/tests\//,$d' "${FLASHINFER_DCP_PATCH}" | patch --dry-run --batch --forward --strip=1 --directory="${FLASHINFER_SITE_PACKAGES}" && sed '/^diff --git a\/tests\//,$d' "${FLASHINFER_DCP_PATCH}" | patch --batch --forward --strip=1 --directory="${FLASHINFER_SITE_PACKAGES}" && rm -f "${FLASHINFER_DCP_PATCH}" && rm -rf /root/.cache/flashinfer /root/.cache/pip # buildkit

2026-08-11 07:39:02 UTC (buildkit.dockerfile.v0)

COPY /opt/trtllm_gen_moe_cubin_pool /opt/trtllm_gen_moe_cubin_pool # buildkit

2026-08-11 07:39:02 UTC (buildkit.dockerfile.v0)

COPY /tools/diff-so-fancy /usr/local/bin/ # buildkit

2026-08-11 07:39:02 UTC (buildkit.dockerfile.v0)

COPY /tools/clang-format /usr/local/bin/ # buildkit

2026-08-11 07:39:02 UTC (buildkit.dockerfile.v0)

COPY /tools/clangd /usr/local/bin/ # buildkit

2026-08-11 07:39:02 UTC (buildkit.dockerfile.v0)

COPY /tools/lib /usr/local/lib/ # buildkit

2026-08-11 07:39:02 UTC (buildkit.dockerfile.v0)

COPY /tools/cmake /usr/local/bin/ # buildkit

2026-08-11 07:39:03 UTC (buildkit.dockerfile.v0)

COPY /tools/ctest /usr/local/bin/ # buildkit

2026-08-11 07:39:03 UTC (buildkit.dockerfile.v0)

COPY /tools/cpack /usr/local/bin/ # buildkit

2026-08-11 07:39:03 UTC (buildkit.dockerfile.v0)

COPY /tools/share/cmake-3.31 /usr/local/share/cmake-3.31 # buildkit

2026-08-11 07:39:03 UTC (buildkit.dockerfile.v0)

COPY /tools/just /usr/local/bin/ # buildkit

2026-08-11 07:39:03 UTC (buildkit.dockerfile.v0)

COPY /root/.oh-my-zsh /root/.oh-my-zsh # buildkit

2026-08-11 07:39:14 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends gdb ninja-build vim tmux htop zsh tree silversearcher-ag cloc pkg-config bear less rdma-core openssh-server gnuplot infiniband-diags perftest ibverbs-providers libibumad3 libibverbs1 libnl-3-200 libnl-route-3-200 librdmacm1 && rm -rf /var/lib/apt/lists/* && apt-get clean # buildkit

2026-08-11 07:39:27 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c apt update -y && apt install -y --no-install-recommends gnupg && echo "deb http://developer.download.nvidia.com/devtools/repos/ubuntu2004/$(if [ "$(uname -m)" = "aarch64" ]; then echo "arm64"; else echo "amd64"; fi) /" | tee /etc/apt/sources.list.d/nvidia-devtools.list && apt-key adv --fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/$(if [ "$(uname -m)" = "aarch64" ]; then echo "arm64"; else echo "x86_64"; fi)/7fa2af80.pub && apt update -y && apt install -y --no-install-recommends nsight-systems-cli && rm -rf /var/lib/apt/lists/* # buildkit

2026-08-11 07:39:35 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c CUDA_MAJOR="${CUDA_VERSION%%.*}" && if [ "$CUDA_MAJOR" -ge 13 ]; then python3 -m pip install mooncake-transfer-engine-cuda13==${MOONCAKE_VERSION}; else python3 -m pip install mooncake-transfer-engine==${MOONCAKE_VERSION}; fi # buildkit

2026-08-11 07:44:55 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c git clone --depth=1 --branch ${MSCCLPP_VERSION} https://${GITHUB_ARTIFACTORY}/microsoft/mscclpp.git /tmp/mscclpp && case "${CUDA_VERSION}" in 12.*) CMAKE_ARGS="-DMSCCLPP_BYPASS_GPU_CHECK=ON -DMSCCLPP_USE_CUDA=ON -DMSCCLPP_GPU_ARCHS=80,90,100,100a,103,103a" python3 -m pip install "/tmp/mscclpp[cuda12]"; ;; 13.*) CMAKE_ARGS="-DMSCCLPP_BYPASS_GPU_CHECK=ON -DMSCCLPP_USE_CUDA=ON -DMSCCLPP_GPU_ARCHS=80,90,100,100a,103,103a" python3 -m pip install "/tmp/mscclpp[cuda13]"; ;; *) echo "Unsupported CUDA version for MSCCL++: ${CUDA_VERSION}" && exit 1; ;; esac && rm -rf /tmp/mscclpp # buildkit

2026-08-11 07:45:18 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install -c /sgl-workspace/constraints.txt datamodel_code_generator pre-commit pytest black isort icdiff uv wheel scikit-build-core py-spy cubloaty google-cloud-storage pandas matplotlib tabulate termplotlib "runai-model-streamer[s3,gcs,azure]>=0.15.7" # buildkit

2026-08-11 07:45:33 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c if [ "${CUDA_VERSION%%.*}" = "12" ]; then python3 -m pip install nixl nixl-cu12 --no-deps ; python3 -m pip install "cuda-python>=12,<13" ; elif [ "${CUDA_VERSION%%.*}" = "13" ]; then python3 -m pip install nixl nixl-cu13 --no-deps ; python3 -m pip install "cuda-python>=13,<14" ; fi # buildkit

2026-08-11 07:45:33 UTC (buildkit.dockerfile.v0)

COPY --chown=root:root --chmod=755 docker/configs/yank /usr/local/bin/yank # buildkit

2026-08-11 07:45:33 UTC (buildkit.dockerfile.v0)

COPY docker/configs/opt/.vimrc /opt/sglang/.vimrc # buildkit

2026-08-11 07:45:33 UTC (buildkit.dockerfile.v0)

COPY docker/configs/opt/.tmux.conf /opt/sglang/.tmux.conf # buildkit

2026-08-11 07:45:33 UTC (buildkit.dockerfile.v0)

COPY docker/configs/opt/.gitconfig /opt/sglang/.gitconfig # buildkit

2026-08-11 07:45:33 UTC (buildkit.dockerfile.v0)

COPY docker/configs/.zshrc /root/.zshrc # buildkit

2026-08-11 07:45:39 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --upgrade "urllib3>=2.6.3" "pillow>=12.1.1" # buildkit

2026-08-11 07:45:46 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c apt-get update && apt-get install -y --only-upgrade binutils binutils-common binutils-x86-64-linux-gnu libbinutils libctf0 libctf-nobfd0 libgprofng0 libsframe1 libgnutls30t64 libpam-modules libpam-modules-bin libpam-runtime libpam0g libsqlite3-0 libtasn1-6 dpkg dpkg-dev libdpkg-perl && rm -rf /var/lib/apt/lists/* # buildkit

2026-08-11 07:45:46 UTC (buildkit.dockerfile.v0)

ARG BRANCH_TYPE=local

2026-08-11 07:45:46 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:45:46 UTC (buildkit.dockerfile.v0)

ARG CUDA_VERSION=12.9.2

2026-08-11 07:45:46 UTC (buildkit.dockerfile.v0)

ARG SGL_VERSION

2026-08-11 07:45:46 UTC (buildkit.dockerfile.v0)

ARG USE_LATEST_SGLANG

2026-08-11 07:45:46 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace

2026-08-11 07:45:47 UTC (buildkit.dockerfile.v0)

COPY /src /tmp/local_src # buildkit

2026-08-11 07:45:49 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c if [ "$BRANCH_TYPE" = "local" ]; then cp -r /tmp/local_src /sgl-workspace/sglang; elif [ "$USE_LATEST_SGLANG" = "1" ]; then git clone --depth=1 https://github.com/sgl-project/sglang.git /sgl-workspace/sglang; elif [ -z "$SGL_VERSION" ]; then echo "ERROR: SGL_VERSION must be set when USE_LATEST_SGLANG=0 and BRANCH_TYPE!=local" && exit 1; else git clone --depth=1 --branch v${SGL_VERSION} https://github.com/sgl-project/sglang.git /sgl-workspace/sglang; fi && rm -rf /tmp/local_src # buildkit

2026-08-11 07:52:55 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c cd /sgl-workspace/sglang && if [ "${CUDA_VERSION%%.*}" = "12" ]; then sed -i 's/cuda-python>=13\.0/cuda-python>=12,<13/' python/pyproject.toml && sed -i 's/flashinfer_python\[cu13\]/flashinfer_python[cu12]/' python/pyproject.toml && sed -i 's/nvidia-cutlass-dsl\[cu13\]/nvidia-cutlass-dsl/' python/pyproject.toml; fi && python3 -m pip install --no-deps -e "python[${BUILD_TYPE}]" && kernels lock python && ( success=0; if [ "$(uname -m)" = "aarch64" ]; then echo "Skipping kernels-community/sgl-flash-attn3 cubin download on aarch64 (no variants published upstream); kernels will be JIT-compiled at runtime"; success=1; else for i in 1 2 3; do echo "Attempt $i/3: downloading sgl-kernel cubins..." && kernels download python && success=1 && break; echo "sgl-kernel cubin download failed, retrying in 30s..." && sleep 30; done; if [ "$success" != "1" ]; then echo "WARNING: no matching sgl-flash-attn3 cubin variant for this torch+CUDA; kernels will be JIT-compiled at runtime"; success=1; fi; fi; [ "$success" = "1" ] ) && mkdir -p /root/.cache/huggingface /root/.cache/sglang && ( if [ -f python/kernels.lock ]; then mv python/kernels.lock /root/.cache/sglang/; fi ) && ( find /usr/local/lib/python3.12/dist-packages -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null || true ) # buildkit

2026-08-11 07:52:55 UTC (buildkit.dockerfile.v0)

COPY /build/sgl-model-gateway-bin /usr/local/bin/sgl-model-gateway # buildkit

2026-08-11 07:52:55 UTC (buildkit.dockerfile.v0)

COPY /build/gateway_wheels /tmp/gateway_wheels # buildkit

2026-08-11 07:53:06 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=amd64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --force-reinstall /tmp/gateway_wheels/*.whl && rm -rf /tmp/gateway_wheels # buildkit

2026-08-11 07:53:06 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace/sglang

2026-08-11 07:53:06 UTC (buildkit.dockerfile.v0)

ARG SGLANG_BUILD_COMMIT=8ba213fc67550aea80dfbd5ef66f4574ec25e653

2026-08-11 07:53:06 UTC (buildkit.dockerfile.v0)

ARG SGLANG_BUILD_URL=https://github.com/sgl-project/sglang/actions/runs/31468844644

2026-08-11 07:53:06 UTC (buildkit.dockerfile.v0)

ARG SGLANG_IMAGE_TAG=lmsysorg/sglang:dev-cu12-Ling-3.0-flash

2026-08-11 07:53:06 UTC (buildkit.dockerfile.v0)

ENV SGLANG_BUILD_COMMIT=8ba213fc67550aea80dfbd5ef66f4574ec25e653 SGLANG_BUILD_URL=https://github.com/sgl-project/sglang/actions/runs/31468844644 SGLANG_IMAGE_TAG=lmsysorg/sglang:dev-cu12-Ling-3.0-flash

2026-08-11 07:53:06 UTC (buildkit.dockerfile.v0)

LABEL org.opencontainers.image.source=https://github.com/sgl-project/sglang org.opencontainers.image.revision=8ba213fc67550aea80dfbd5ef66f4574ec25e653 org.opencontainers.image.version=lmsysorg/sglang:dev-cu12-Ling-3.0-flash org.opencontainers.image.url=https://github.com/sgl-project/sglang/actions/runs/31468844644 ai.sglang.build.commit=8ba213fc67550aea80dfbd5ef66f4574ec25e653 ai.sglang.build.url=https://github.com/sgl-project/sglang/actions/runs/31468844644 ai.sglang.image.tag=lmsysorg/sglang:dev-cu12-Ling-3.0-flash

Details
Created

2026-08-11 07:59:18 UTC

Size

15.6 GB

Content Digest
Labels
  • ai.sglang.build.commit
    8ba213fc67550aea80dfbd5ef66f4574ec25e653
  • ai.sglang.build.url
    https://github.com/sgl-project/sglang/actions/runs/31468844644
  • ai.sglang.image.tag
    lmsysorg/sglang:dev-cu12-Ling-3.0-flash
  • com.nvidia.cudnn.version
    9.10.2.21-1
  • maintainer
    NVIDIA CORPORATION <cudatools@nvidia.com>
  • org.opencontainers.image.revision
    8ba213fc67550aea80dfbd5ef66f4574ec25e653
  • org.opencontainers.image.source
    https://github.com/sgl-project/sglang
  • org.opencontainers.image.url
    https://github.com/sgl-project/sglang/actions/runs/31468844644
  • org.opencontainers.image.version
    lmsysorg/sglang:dev-cu12-Ling-3.0-flash

Environment
CUDA_HOME

/usr/local/cuda

CUDA_VERSION

12.9.2

DEBIAN_FRONTEND

noninteractive

FLASHINFER_VERSION

0.6.15.post1

GDRCOPY_HOME

/usr/src/gdrdrv-2.5.1/

LANG

en_US.UTF-8

LANGUAGE

en_US:en

LC_ALL

en_US.UTF-8

LD_LIBRARY_PATH

/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64

LIBRARY_PATH

/usr/local/cuda/lib64/stubs

NCCL_VERSION

2.27.3-1

NVARCH

sbsa

NVIDIA_DRIVER_CAPABILITIES

compute,utility

NVIDIA_PRODUCT_NAME

CUDA

NVIDIA_REQUIRE_CUDA

cuda>=12.9 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>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=560,driver<561 brand=grid,driver>=560,driver<561 brand=tesla,driver>=560,driver<561 brand=nvidia,driver>=560,driver<561 brand=quadro,driver>=560,driver<561 brand=quadrortx,driver>=560,driver<561 brand=nvidiartx,driver>=560,driver<561 brand=vapps,driver>=560,driver<561 brand=vpc,driver>=560,driver<561 brand=vcs,driver>=560,driver<561 brand=vws,driver>=560,driver<561 brand=cloudgaming,driver>=560,driver<561 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 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

NVIDIA_VISIBLE_DEVICES

all

NV_CUDA_CUDART_DEV_VERSION

12.9.79-1

NV_CUDA_CUDART_VERSION

12.9.79-1

NV_CUDA_LIB_VERSION

12.9.2-1

NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE

cuda-nsight-compute-12-9=12.9.2-1

NV_CUDA_NSIGHT_COMPUTE_VERSION

12.9.2-1

NV_CUDNN_PACKAGE

libcudnn9-cuda-12=9.10.2.21-1

NV_CUDNN_PACKAGE_DEV

libcudnn9-dev-cuda-12=9.10.2.21-1

NV_CUDNN_PACKAGE_DEV_HEADERS

libcudnn9-headers-cuda-12=9.10.2.21-1

NV_CUDNN_PACKAGE_NAME

libcudnn9-cuda-12

NV_CUDNN_VERSION

9.10.2.21-1

NV_LIBCUBLAS_DEV_PACKAGE

libcublas-dev-12-9=12.9.2.10-1

NV_LIBCUBLAS_DEV_PACKAGE_NAME

libcublas-dev-12-9

NV_LIBCUBLAS_DEV_VERSION

12.9.2.10-1

NV_LIBCUBLAS_PACKAGE

libcublas-12-9=12.9.2.10-1

NV_LIBCUBLAS_PACKAGE_NAME

libcublas-12-9

NV_LIBCUBLAS_VERSION

12.9.2.10-1

NV_LIBCUSPARSE_DEV_VERSION

12.5.10.65-1

NV_LIBCUSPARSE_VERSION

12.5.10.65-1

NV_LIBNCCL_DEV_PACKAGE

libnccl-dev=2.27.3-1+cuda12.9

NV_LIBNCCL_DEV_PACKAGE_NAME

libnccl-dev

NV_LIBNCCL_DEV_PACKAGE_VERSION

2.27.3-1

NV_LIBNCCL_PACKAGE

libnccl2=2.27.3-1+cuda12.9

NV_LIBNCCL_PACKAGE_NAME

libnccl2

NV_LIBNCCL_PACKAGE_VERSION

2.27.3-1

NV_LIBNPP_DEV_PACKAGE

libnpp-dev-12-9=12.4.1.87-1

NV_LIBNPP_DEV_VERSION

12.4.1.87-1

NV_LIBNPP_PACKAGE

libnpp-12-9=12.4.1.87-1

NV_LIBNPP_VERSION

12.4.1.87-1

NV_NVML_DEV_VERSION

12.9.79-1

NV_NVTX_VERSION

12.9.79-1

PATH

/root/.cargo/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/nvidia/bin

SGLANG_BUILD_COMMIT

8ba213fc67550aea80dfbd5ef66f4574ec25e653

SGLANG_BUILD_URL

https://github.com/sgl-project/sglang/actions/runs/31468844644

SGLANG_IMAGE_TAG

lmsysorg/sglang:dev-cu12-Ling-3.0-flash

SGLANG_TRTLLM_GEN_MOE_CUBIN_POOL

/opt/trtllm_gen_moe_cubin_pool


Layers

[#000] sha256:818154cda96df8bbb276b4f4339124da55756620a1037af15570bc95312850fa - 0.17% (27.5 MB)

[#001] sha256:7e13def16fdccbcc27583bc888c82ffcf2b72f080c964ca6c7c676c199950d26 - 0.03% (4.91 MB)

[#002] sha256:f51a4d08ef43d56bcc7b3c003fc50423e33a45800b725294ade5183457f4928b - 0.0% (383 KB)

[#003] sha256:3a2f4d4ac1876e0ceeef657fb5c2246cb26f9276ffdb90bbe29b696696021896 - 0.58% (92.8 MB)

[#004] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#005] sha256:8fb30479b7cb69963cec4884b9f098d7a50d0c901ad31f261b4c8414507bd1ac - 0.0% (186 Bytes)

[#006] sha256:1dca0c9bc5f79f4cedcd68523b7682fb46a1b1546351bc67289e9440608187fd - 0.0% (6.72 KB)

[#007] sha256:3f357bfb94aa4b3ca094d427d20890ef2f80152c481f25ef0292a27954b67251 - 13.41% (2.1 GB)

[#008] sha256:dc861c2179c56651f6b8d4a003ef12dd13c34c6bfdadcdc4edb514aeab17aec4 - 0.0% (58.6 KB)

[#009] sha256:4e366a28834d174dbbd109a43c47d92d99762a58404d05d627343013c79e66ec - 0.0% (1.65 KB)

[#010] sha256:80a29ee074c3c7df423662efd09cc786894fddff01beba372445021ddd834f6b - 0.0% (1.49 KB)

[#011] sha256:9c3aab2ba16f80a1823c43a809e19441ad891d9751f9bbb27c6cdc165d71edf7 - 16.94% (2.65 GB)

[#012] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#013] sha256:fb28bec4fd671d024b87083f34143fa48c980e801bb1db5c641128cd7a55f077 - 0.0% (87.8 KB)

[#014] sha256:36b40e96480be215fbe83d70fad997adb8aa743d91f564e02c493de7d0d32709 - 4.23% (678 MB)

[#015] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#016] sha256:cdcb87a434c4c3673ee7c5f212791cf1f3be7cde7f574ede11900941062d2922 - 0.87% (139 MB)

[#017] sha256:a36b3e5ccb3cc7c82fef2a785f811ffbabec5eedf7970e8474a4dc204f1b6983 - 2.08% (333 MB)

[#018] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#019] sha256:de522a50897191ccdcceaec4d3630e9fd8c37279562bd055610455b7142572bf - 0.01% (1.45 MB)

[#020] sha256:80d40994015704aaa0a3cad770cda2e1f6540fbb3a8b5b88927c2b79ed507e39 - 0.0% (188 Bytes)

[#021] sha256:7c5170f867d0cdd0a25adec8e53c64e88f539f6380dca47bcdb108f92f266ea0 - 0.01% (846 KB)

[#022] sha256:74aa934ae7a404bc19d330aca4d225eadd9a80d1574d0471ba33c76e5fa5d750 - 0.0% (103 Bytes)

[#023] sha256:e05a79b4299bab891bf3270d7b8c64d018cf0a8daa0aa99c4376f26f7cb0a284 - 1.1% (176 MB)

[#024] sha256:a932c7fa6a2ce0502ac4cfbde2c4638267289972217139cea110cc31f5445cd9 - 0.28% (44.6 MB)

[#025] sha256:7f89b7859c8f8a03812f520ee609f46201cba74af518ba9c77d93c839314ab9a - 0.0% (2.87 KB)

[#026] sha256:341a651502c507a73531f0a0178743759a097b06f3fa01bd6bd29b913bb9029d - 0.0% (20.2 KB)

[#027] sha256:97843aaf816f68d1567652b1311d50ff65c97120e4718c25ba51b687b8d33016 - 0.0% (42.7 KB)

[#028] sha256:899792c0c5526b19eb6394ce48b261e0fb3c755310674ebd1e60d02234f697d8 - 0.0% (2.23 KB)

[#029] sha256:3c0b873f393d0a44dc863c49b9d65e638b71b0f265529eaa5cf222bb284762bb - 29.83% (4.67 GB)

[#030] sha256:bb94de326685cca7e8790de29c190a6f4bbeaef52c20e91733a262ea9f2f8425 - 0.0% (66.7 KB)

[#031] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#032] sha256:26990c42d4fe441550796f73e39feea5249578b1d74427de2f3b57b5b0ae0e08 - 0.0% (143 Bytes)

[#033] sha256:a7d1058f614474132b0174621b9639c0cb5a7fbcedb9c19ccca895fc6e29136c - 0.0% (157 Bytes)

[#034] sha256:efec98312bb29d856112d17886f97034245652da6b27baf0a3a66eac7d754149 - 15.78% (2.47 GB)

[#035] sha256:f59f22212f0332464b62db0d46fa4c83656290b0d088d82cf6a0901404c8f84d - 0.0% (51.6 KB)

[#036] sha256:5101ad25353b1ed42becd5d8783bb184f231d2adf3551d558fb85ff1a8c69fcf - 0.0% (174 KB)

[#037] sha256:8d81f1c0a92c830d9396f629da783f5c2f65a4684240bcb24193f629eabe6ec9 - 0.19% (30.6 MB)

[#038] sha256:d82ad6f8bcd547e878fddb40c0edefa7175fadce0d5f249a71874cafc224e582 - 0.0% (13.1 KB)

[#039] sha256:1c3636163924e9707fe95e92d79b6e721a0f53d12b413fc3854723ebb4d00ef1 - 0.01% (1.7 MB)

[#040] sha256:8e06a781fa5e382ca4383c7e46ce684891e5766325873a762dcf9918ea1ec453 - 0.51% (81 MB)

[#041] sha256:0c12763282af5f83ec8977ece6582f34978e86eb5815a50695261966a819b955 - 0.01% (957 KB)

[#042] sha256:09efa9198ffd6c5193bf5808b91b22966b5ad38c26058b47776e8418d9be01c0 - 0.05% (7.33 MB)

[#043] sha256:aa297581873ae325540ddc89d89405c0c43b68fbcdc9840f6017b7c6d288fd35 - 0.05% (7.96 MB)

[#044] sha256:14c4676bc2bf837a30cf946900dde219da6f0ed0086c06c418f3fb9986be1ea1 - 0.05% (7.51 MB)

[#045] sha256:333ce10abec04662081b1bb6fcaabb1673bc77388bf6872b69c9eb94604358eb - 0.01% (1.93 MB)

[#046] sha256:706dc0b7d031cecf546744181d486116a6f640c78781760d835b46b4ae7b3cf9 - 0.01% (1.76 MB)

[#047] sha256:dde90cbc3709b86ea5a1ed3683d2671dcc2648150eb5af1213839d81086b4b91 - 0.04% (7.05 MB)

[#048] sha256:1c8d3de9b3b3517c707b3408231e0ee70c78dad55087184aca0f74109e5c5a71 - 0.4% (64.6 MB)

[#049] sha256:f6beedae14e4d5537b039797ce618bdfa4d6e3293a56538198788846ebde9f20 - 1.23% (197 MB)

[#050] sha256:a33daa6d14790c7584ddb321e1715c92c10516650160b4fdcc9f35708a05bc0a - 0.14% (23.1 MB)

[#051] sha256:5c8ad093628191d03babb398dc6ce229b1c2fda14bc03e7aaf96f95ac03325d9 - 1.51% (242 MB)

[#052] sha256:60ffe6d51eecc71d564ed9d397c212401c66bc2058f577669c1b2a82e3a356c3 - 0.68% (109 MB)

[#053] sha256:122b3797106a4940d5a3072f43a902219aa6638ee5a6366b806db49851556b79 - 0.52% (83.4 MB)

[#054] sha256:901ad92f5cca2c3ebeab9a1de96e0a4f3950bc598a37682a4af8be98af6602f8 - 0.0% (441 Bytes)

[#055] sha256:ece2b6b304afaa82b5e4fffabceda3556b005809ec8ffcc226f44a89955e1ef7 - 0.0% (722 Bytes)

[#056] sha256:3f6a1c18b399dc8b0b37093f00e4243cb9b45a86c78cc7f98e0469a2f880f2bd - 0.0% (507 Bytes)

[#057] sha256:bc42500503986e56e2902dd21f8334c6270ac528e77225fd6ad93b0349a471e3 - 0.0% (517 Bytes)

[#058] sha256:983994c9afe4f4de10bb80cb197104e7a3373564fa4e2c56d91ec928665b6152 - 0.0% (393 Bytes)

[#059] sha256:f1c9f546593c8447e89e34c1de8cd20b4d8aefec5798b1b683d665dfb0717b24 - 0.08% (13.1 MB)

[#060] sha256:7707cacd341838a424a93923d9d2bcbddf30eef3e81b1f5ce3c1869fe233d7a3 - 0.02% (3.12 MB)

[#061] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)

[#062] sha256:e45fc2caea196d68230b09a864f636ff9c46206a27883d837cd3a14d965a4d89 - 1.87% (299 MB)

[#063] sha256:f4a5df688f48d92a85d6b831db6065e3a19c8a99c43ace9514a9cf413ec9a2c4 - 1.87% (299 MB)

[#064] sha256:c314ab3696d24f1ca407ecdc33bbe9f34fecc5f04aff2693a71e46785f590784 - 5.05% (808 MB)

[#065] sha256:bbfe3ae9685533dd7d1842b04fe3cdf2eae1972cce18f51b5fcee42c9d503943 - 0.11% (17.4 MB)

[#066] sha256:3b9caa32dce685fa6e610712ad540bc220b911692f84b67aad486bd9efc2ba24 - 0.11% (17.3 MB)

[#067] sha256:0a917401293c7e78d4d141b752f00e2dff07c4da31cd656fb3be46e3731b866e - 0.17% (27.1 MB)

[#068] sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 - 0.0% (32 Bytes)


History
2026-04-10 06:56:52 UTC

/bin/sh -c #(nop) ARG RELEASE

2026-04-10 06:56:52 UTC

/bin/sh -c #(nop) ARG LAUNCHPAD_BUILD_ARCH

2026-04-10 06:56:52 UTC

/bin/sh -c #(nop) LABEL org.opencontainers.image.version=24.04

2026-04-10 06:56:54 UTC

/bin/sh -c #(nop) ADD file:c98b7645109cdf61ab97492b90629581b1b7cb925b9d58a5787a4aaeb719f2be in /

2026-04-10 06:56:54 UTC

/bin/sh -c #(nop) CMD ["/bin/bash"]

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

ENV NVARCH=sbsa

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_REQUIRE_CUDA=cuda>=12.9 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>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=560,driver<561 brand=grid,driver>=560,driver<561 brand=tesla,driver>=560,driver<561 brand=nvidia,driver>=560,driver<561 brand=quadro,driver>=560,driver<561 brand=quadrortx,driver>=560,driver<561 brand=nvidiartx,driver>=560,driver<561 brand=vapps,driver>=560,driver<561 brand=vpc,driver>=560,driver<561 brand=vcs,driver>=560,driver<561 brand=vws,driver>=560,driver<561 brand=cloudgaming,driver>=560,driver<561 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 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

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_CUDART_VERSION=12.9.79-1

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

ARG IMAGE_UPGRADE_ID=0

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=arm64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c echo "IMAGE_UPGRADE_ID=${IMAGE_UPGRADE_ID}" && apt-get update && apt-get upgrade -y && apt-get install -y --no-install-recommends gnupg2 curl ca-certificates && curl -fsSLO https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/${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

2026-05-18 16:10:37 UTC (buildkit.dockerfile.v0)

ENV CUDA_VERSION=12.9.2

2026-05-18 16:14:22 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=arm64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends cuda-cudart-12-9=${NV_CUDA_CUDART_VERSION} cuda-toolkit-12-9-config-common=${NV_CUDA_CUDART_VERSION} cuda-toolkit-12-config-common=${NV_CUDA_CUDART_VERSION} cuda-toolkit-config-common=${NV_CUDA_CUDART_VERSION} && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 16:20:34 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=arm64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c apt-get update && if apt-cache policy cuda-compat-12-9 2>/dev/null | grep -q "Candidate:"; then apt-get install -y --no-install-recommends cuda-compat-12-9; fi && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 16:23:23 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=arm64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c apt-get update && if apt-cache policy cuda-compat-orin-12-9 2>/dev/null | grep -q "Candidate:"; then apt-get install -y --no-install-recommends cuda-compat-orin-12-9; fi && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 16:23:23 UTC (buildkit.dockerfile.v0)

RUN |2 TARGETARCH=arm64 IMAGE_UPGRADE_ID=20260518 /bin/sh -c echo "/usr/local/cuda/lib64" >> /etc/ld.so.conf.d/nvidia.conf # buildkit

2026-05-18 16:23:23 UTC (buildkit.dockerfile.v0)

ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin

2026-05-18 16:23:23 UTC (buildkit.dockerfile.v0)

ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64

2026-05-18 16:23:23 UTC (buildkit.dockerfile.v0)

COPY NGC-DL-CONTAINER-LICENSE / # buildkit

2026-05-18 16:23:23 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_VISIBLE_DEVICES=all

2026-05-18 16:23:23 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_LIB_VERSION=12.9.2-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_NVTX_VERSION=12.9.79-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_VERSION=12.4.1.87-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_PACKAGE=libnpp-12-9=12.4.1.87-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUSPARSE_VERSION=12.5.10.65-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_PACKAGE_NAME=libcublas-12-9

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_VERSION=12.9.2.10-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_PACKAGE=libcublas-12-9=12.9.2.10-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_PACKAGE_NAME=libnccl2

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_PACKAGE_VERSION=2.27.3-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NCCL_VERSION=2.27.3-1

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_PACKAGE=libnccl2=2.27.3-1+cuda12.9

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=arm64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends cuda-libraries-12-9=${NV_CUDA_LIB_VERSION} ${NV_LIBNPP_PACKAGE} cuda-nvtx-12-9=${NV_NVTX_VERSION} libcusparse-12-9=${NV_LIBCUSPARSE_VERSION} ${NV_LIBCUBLAS_PACKAGE} ${NV_LIBNCCL_PACKAGE} && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=arm64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_PACKAGE_NAME} ${NV_LIBNCCL_PACKAGE_NAME} # buildkit

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

COPY entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

COPY nvidia_entrypoint.sh /opt/nvidia/ # buildkit

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENV NVIDIA_PRODUCT_NAME=CUDA

2026-05-18 16:33:26 UTC (buildkit.dockerfile.v0)

ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_LIB_VERSION=12.9.2-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_CUDART_DEV_VERSION=12.9.79-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_NVML_DEV_VERSION=12.9.79-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUSPARSE_DEV_VERSION=12.5.10.65-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_DEV_VERSION=12.4.1.87-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNPP_DEV_PACKAGE=libnpp-dev-12-9=12.4.1.87-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-12-9

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_DEV_VERSION=12.9.2.10-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-12-9=12.9.2.10-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_NSIGHT_COMPUTE_VERSION=12.9.2-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-12-9=12.9.2-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_DEV_PACKAGE_NAME=libnccl-dev

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_DEV_PACKAGE_VERSION=2.27.3-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NCCL_VERSION=2.27.3-1

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV NV_LIBNCCL_DEV_PACKAGE=libnccl-dev=2.27.3-1+cuda12.9

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=arm64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends cuda-cudart-dev-12-9=${NV_CUDA_CUDART_DEV_VERSION} cuda-command-line-tools-12-9=${NV_CUDA_LIB_VERSION} cuda-minimal-build-12-9=${NV_CUDA_LIB_VERSION} cuda-libraries-dev-12-9=${NV_CUDA_LIB_VERSION} cuda-nvml-dev-12-9=${NV_NVML_DEV_VERSION} ${NV_NVPROF_DEV_PACKAGE} ${NV_LIBNPP_DEV_PACKAGE} libcusparse-dev-12-9=${NV_LIBCUSPARSE_DEV_VERSION} ${NV_LIBCUBLAS_DEV_PACKAGE} ${NV_LIBNCCL_DEV_PACKAGE} ${NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE} && rm -rf /var/lib/apt/lists/* # buildkit

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=arm64 /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

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=arm64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_DEV_PACKAGE_NAME} ${NV_LIBNCCL_DEV_PACKAGE_NAME} # buildkit

2026-05-18 17:24:59 UTC (buildkit.dockerfile.v0)

ENV LIBRARY_PATH=/usr/local/cuda/lib64/stubs

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_VERSION=9.10.2.21-1

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-12

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE=libcudnn9-cuda-12=9.10.2.21-1

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-12=9.10.2.21-1

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

ENV NV_CUDNN_PACKAGE_DEV_HEADERS=libcudnn9-headers-cuda-12=9.10.2.21-1

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

LABEL com.nvidia.cudnn.version=9.10.2.21-1

2026-05-18 17:32:03 UTC (buildkit.dockerfile.v0)

RUN |1 TARGETARCH=arm64 /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

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG TARGETARCH=arm64

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG BRANCH_TYPE=local

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG SGL_KERNEL_VERSION=0.4.6.post1

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG SGL_VERSION

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG SGL_DEEP_GEMM_VERSION=0.1.5.post2

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG USE_LATEST_SGLANG=0

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG GDRCOPY_VERSION=2.5.1

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG PIP_DEFAULT_INDEX

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG UBUNTU_MIRROR

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG GITHUB_ARTIFACTORY=github.com

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG INSTALL_FLASHINFER_JIT_CACHE=1

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG FLASHINFER_VERSION=0.6.15.post1

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_VERSION=0.3.12.post1

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG MSCCLPP_VERSION=sglang-v0.9.1

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ENV DEBIAN_FRONTEND=noninteractive CUDA_HOME=/usr/local/cuda GDRCOPY_HOME=/usr/src/gdrdrv-2.5.1/ FLASHINFER_VERSION=0.6.15.post1 SGLANG_TRTLLM_GEN_MOE_CUBIN_POOL=/opt/trtllm_gen_moe_cubin_pool

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/nvidia/bin LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64

2026-08-11 07:26:45 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c if [ -n "$UBUNTU_MIRROR" ]; then sed -i "s|http://.*archive.ubuntu.com|$UBUNTU_MIRROR|g" /etc/apt/sources.list && sed -i "s|http://.*security.ubuntu.com|$UBUNTU_MIRROR|g" /etc/apt/sources.list; fi # buildkit

2026-08-11 07:27:30 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c apt update && apt install -y --no-install-recommends wget software-properties-common && apt install -y --no-install-recommends python3.12-full python3.12-dev && update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.12 2 && update-alternatives --set python3 /usr/bin/python3.12 && wget -q https://bootstrap.pypa.io/get-pip.py && python3 get-pip.py --break-system-packages && rm get-pip.py && python3 -m pip config set global.break-system-packages true && cd /usr/lib/python3/dist-packages/ && ln -s apt_pkg.cpython-312-*-linux-gnu.so apt_pkg.so # buildkit

2026-08-11 07:29:19 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c apt-get update && apt-get install -y --no-install-recommends ca-certificates software-properties-common netcat-openbsd kmod unzip openssh-server curl wget lsof locales build-essential cmake perl patch patchelf ccache git-lfs libopenmpi-dev libnuma1 libnuma-dev numactl ffmpeg libibverbs-dev libibverbs1 libibumad3 librdmacm1 libnl-3-200 libnl-route-3-200 libnl-route-3-dev libnl-3-dev ibverbs-providers infiniband-diags perftest libgoogle-glog-dev libgtest-dev libjsoncpp-dev libunwind-dev libboost-all-dev libssl-dev libgrpc-dev libgrpc++-dev libprotobuf-dev protobuf-compiler protobuf-compiler-grpc pybind11-dev libhiredis-dev libcurl4-openssl-dev libczmq4 libczmq-dev libfabric-dev linux-libc-dev devscripts debhelper fakeroot dkms check libsubunit0 libsubunit-dev && ln -sf /usr/bin/python3.12 /usr/bin/python && rm -rf /var/lib/apt/lists/* && apt-get clean # buildkit

2026-08-11 07:29:20 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c if [ -n "${PIP_DEFAULT_INDEX}" ]; then python3 -m pip config set global.index-url ${PIP_DEFAULT_INDEX}; fi # buildkit

2026-08-11 07:30:21 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c mkdir -p /tmp/gdrcopy && cd /tmp && curl --retry 3 --retry-delay 2 -fsSL -o v${GDRCOPY_VERSION}.tar.gz https://${GITHUB_ARTIFACTORY}/NVIDIA/gdrcopy/archive/refs/tags/v${GDRCOPY_VERSION}.tar.gz && tar -xzf v${GDRCOPY_VERSION}.tar.gz && rm v${GDRCOPY_VERSION}.tar.gz && cd gdrcopy-${GDRCOPY_VERSION}/packages && CUDA=/usr/local/cuda ./build-deb-packages.sh && dpkg -i gdrdrv-dkms_*.deb libgdrapi_*.deb gdrcopy-tests_*.deb gdrcopy_*.deb && cd / && rm -rf /tmp/gdrcopy # buildkit

2026-08-11 07:30:21 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c ln -sf /usr/lib/$(uname -m)-linux-gnu/libmlx5.so.1 /usr/lib/$(uname -m)-linux-gnu/libmlx5.so # buildkit

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

RUN |19 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON /bin/sh -c locale-gen en_US.UTF-8 # buildkit

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

ENV LANG=en_US.UTF-8 LANGUAGE=en_US:en LC_ALL=en_US.UTF-8

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

ARG CUDA_VERSION=12.9.2

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

ARG SGL_KERNEL_VERSION

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

ARG GITHUB_ARTIFACTORY

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace

2026-08-11 07:30:23 UTC (buildkit.dockerfile.v0)

ENV PATH=/root/.cargo/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/nvidia/bin

2026-08-11 07:30:38 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c curl --proto '=https' --tlsv1.2 --retry 3 --retry-delay 2 -sSf https://sh.rustup.rs | sh -s -- -y --no-modify-path --profile minimal && rustc --version && cargo --version # buildkit

2026-08-11 07:30:45 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --upgrade pip setuptools wheel html5lib six && case "$CUDA_VERSION" in 12.6.3) CUINDEX=126 ;; 12.9.2) CUINDEX=129 ;; 13.0.3) CUINDEX=130 ;; *) echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1 ;; esac && if [ "$CUDA_VERSION" = "12.6.3" ]; then python3 -m pip install https://${GITHUB_ARTIFACTORY}/sgl-project/whl/releases/download/v${SGL_KERNEL_VERSION}/sglang_kernel-${SGL_KERNEL_VERSION}+cu124-cp310-abi3-manylinux2014_$(uname -m).whl --force-reinstall --no-deps ; elif [ "$CUDA_VERSION" = "12.9.2" ]; then python3 -m pip install https://${GITHUB_ARTIFACTORY}/sgl-project/whl/releases/download/v${SGL_KERNEL_VERSION}/sglang_kernel-${SGL_KERNEL_VERSION}+cu129-cp310-abi3-manylinux2014_$(uname -m).whl --force-reinstall --no-deps ; elif [ "$CUDA_VERSION" = "13.0.3" ]; then python3 -m pip install sglang-kernel==${SGL_KERNEL_VERSION} --force-reinstall --no-deps ; else echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1 ; fi # buildkit

2026-08-11 07:30:46 UTC (buildkit.dockerfile.v0)

COPY python/pyproject.toml /tmp/sglang_deps/python/pyproject.toml # buildkit

2026-08-11 07:30:46 UTC (buildkit.dockerfile.v0)

COPY rust/sglang-grpc /tmp/sglang_deps/rust/sglang-grpc # buildkit

2026-08-11 07:30:46 UTC (buildkit.dockerfile.v0)

COPY rust/sglang-mm /tmp/sglang_deps/rust/sglang-mm # buildkit

2026-08-11 07:30:46 UTC (buildkit.dockerfile.v0)

COPY proto /tmp/sglang_deps/proto # buildkit

2026-08-11 07:37:28 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c case "$CUDA_VERSION" in 12.6.3) CUINDEX=126 ;; 12.9.2) CUINDEX=129 ;; 13.0.3) CUINDEX=130 ;; *) echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1 ;; esac && cd /tmp/sglang_deps/python && mkdir -p sglang && touch sglang/__init__.py && echo '__version__ = "0.0.0"' > sglang/version.py && touch README.md && touch LICENSE && SGL_DEEP_EP_VERSION="$(sed -n 's/^[[:space:]]*"sgl-deep-ep==\([^"]*\)",/\1/p' pyproject.toml)" && test -n "${SGL_DEEP_EP_VERSION}" && if [ "${CUDA_VERSION%%.*}" = "12" ]; then python3 -m pip install "sgl-deep-ep==${SGL_DEEP_EP_VERSION}+cu129" --index-url "https://docs.sglang.ai/whl/cu129/" --no-deps; fi && if [ "${CUDA_VERSION%%.*}" = "12" ]; then sed -i 's/cuda-python>=13\.0/cuda-python>=12,<13/' pyproject.toml && sed -i 's/flashinfer_python\[cu13\]/flashinfer_python[cu12]/' pyproject.toml && sed -i 's/nvidia-cutlass-dsl\[cu13\]/nvidia-cutlass-dsl/' pyproject.toml; fi && python3 -m pip install --extra-index-url https://download.pytorch.org/whl/cu${CUINDEX} ".[${BUILD_TYPE}]" && if [ "${CUDA_VERSION%%.*}" = "12" ]; then pip list --format=freeze | awk -F'==' '/-cu13(==|$)/ {print $1}' | xargs -r python3 -m pip uninstall -y && python3 -m pip install --index-url https://download.pytorch.org/whl/cu${CUINDEX} torch==2.13.0 torchvision==0.28.0 torchaudio==2.11.0 --force-reinstall; python3 -m pip install https://github.com/sgl-project/whl/releases/download/v${SGL_DEEP_GEMM_VERSION}/sgl_deep_gemm-${SGL_DEEP_GEMM_VERSION}+cu129-py3-none-manylinux2014_$(uname -m).whl --force-reinstall; fi && cd /sgl-workspace && rm -rf /tmp/sglang_deps && pip freeze | grep -v "^sglang==" > /sgl-workspace/constraints.txt # buildkit

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --ignore-installed --no-deps distro # buildkit

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG BRANCH_TYPE=local

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG CUDA_VERSION=12.9.2

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG SGL_VERSION

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG USE_LATEST_SGLANG

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG GITHUB_ARTIFACTORY

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_VERSION

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG MOONCAKE_COMPILE_ARG

2026-08-11 07:37:29 UTC (buildkit.dockerfile.v0)

ARG MSCCLPP_VERSION

2026-08-11 07:37:30 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace

2026-08-11 07:37:38 UTC (buildkit.dockerfile.v0)

COPY /wheels /tmp/wheels/hpc-ops # buildkit

2026-08-11 07:37:38 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c if ls /tmp/wheels/hpc-ops/*.whl >/dev/null 2>&1; then pip install --no-deps /tmp/wheels/hpc-ops/*.whl; fi && rm -rf /tmp/wheels/hpc-ops # buildkit

2026-08-11 07:40:20 UTC (buildkit.dockerfile.v0)

COPY /flashinfer_jit_output/ /usr/local/lib/python3.12/dist-packages/ # buildkit

2026-08-11 07:40:20 UTC (buildkit.dockerfile.v0)

COPY docker/kimi_k3/flashinfer-perkz-dcp-0.6.15.txt /tmp/flashinfer-perkz-dcp-0.6.15.txt # buildkit

2026-08-11 07:40:24 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c FLASHINFER_DCP_PATCH=/tmp/flashinfer-perkz-dcp-0.6.15.txt && FLASHINFER_SITE_PACKAGES="$(python3 -c 'from pathlib import Path; import flashinfer; print(Path(flashinfer.__file__).resolve().parent.parent)')" && sed '/^diff --git a\/tests\//,$d' "${FLASHINFER_DCP_PATCH}" | patch --dry-run --batch --forward --strip=1 --directory="${FLASHINFER_SITE_PACKAGES}" && sed '/^diff --git a\/tests\//,$d' "${FLASHINFER_DCP_PATCH}" | patch --batch --forward --strip=1 --directory="${FLASHINFER_SITE_PACKAGES}" && rm -f "${FLASHINFER_DCP_PATCH}" && rm -rf /root/.cache/flashinfer /root/.cache/pip # buildkit

2026-08-11 07:40:27 UTC (buildkit.dockerfile.v0)

COPY /opt/trtllm_gen_moe_cubin_pool /opt/trtllm_gen_moe_cubin_pool # buildkit

2026-08-11 07:40:27 UTC (buildkit.dockerfile.v0)

COPY /tools/diff-so-fancy /usr/local/bin/ # buildkit

2026-08-11 07:40:27 UTC (buildkit.dockerfile.v0)

COPY /tools/clang-format /usr/local/bin/ # buildkit

2026-08-11 07:40:28 UTC (buildkit.dockerfile.v0)

COPY /tools/clangd /usr/local/bin/ # buildkit

2026-08-11 07:40:28 UTC (buildkit.dockerfile.v0)

COPY /tools/lib /usr/local/lib/ # buildkit

2026-08-11 07:40:28 UTC (buildkit.dockerfile.v0)

COPY /tools/cmake /usr/local/bin/ # buildkit

2026-08-11 07:40:28 UTC (buildkit.dockerfile.v0)

COPY /tools/ctest /usr/local/bin/ # buildkit

2026-08-11 07:40:28 UTC (buildkit.dockerfile.v0)

COPY /tools/cpack /usr/local/bin/ # buildkit

2026-08-11 07:40:31 UTC (buildkit.dockerfile.v0)

COPY /tools/share/cmake-3.31 /usr/local/share/cmake-3.31 # buildkit

2026-08-11 07:40:31 UTC (buildkit.dockerfile.v0)

COPY /tools/just /usr/local/bin/ # buildkit

2026-08-11 07:40:32 UTC (buildkit.dockerfile.v0)

COPY /root/.oh-my-zsh /root/.oh-my-zsh # buildkit

2026-08-11 07:40:57 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends gdb ninja-build vim tmux htop zsh tree silversearcher-ag cloc pkg-config bear less rdma-core openssh-server gnuplot infiniband-diags perftest ibverbs-providers libibumad3 libibverbs1 libnl-3-200 libnl-route-3-200 librdmacm1 && rm -rf /var/lib/apt/lists/* && apt-get clean # buildkit

2026-08-11 07:41:22 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c apt update -y && apt install -y --no-install-recommends gnupg && echo "deb http://developer.download.nvidia.com/devtools/repos/ubuntu2004/$(if [ "$(uname -m)" = "aarch64" ]; then echo "arm64"; else echo "amd64"; fi) /" | tee /etc/apt/sources.list.d/nvidia-devtools.list && apt-key adv --fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/$(if [ "$(uname -m)" = "aarch64" ]; then echo "arm64"; else echo "x86_64"; fi)/7fa2af80.pub && apt update -y && apt install -y --no-install-recommends nsight-systems-cli && rm -rf /var/lib/apt/lists/* # buildkit

2026-08-11 07:41:26 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c CUDA_MAJOR="${CUDA_VERSION%%.*}" && if [ "$CUDA_MAJOR" -ge 13 ]; then python3 -m pip install mooncake-transfer-engine-cuda13==${MOONCAKE_VERSION}; else python3 -m pip install mooncake-transfer-engine==${MOONCAKE_VERSION}; fi # buildkit

2026-08-11 07:50:02 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c git clone --depth=1 --branch ${MSCCLPP_VERSION} https://${GITHUB_ARTIFACTORY}/microsoft/mscclpp.git /tmp/mscclpp && case "${CUDA_VERSION}" in 12.*) CMAKE_ARGS="-DMSCCLPP_BYPASS_GPU_CHECK=ON -DMSCCLPP_USE_CUDA=ON -DMSCCLPP_GPU_ARCHS=80,90,100,100a,103,103a" python3 -m pip install "/tmp/mscclpp[cuda12]"; ;; 13.*) CMAKE_ARGS="-DMSCCLPP_BYPASS_GPU_CHECK=ON -DMSCCLPP_USE_CUDA=ON -DMSCCLPP_GPU_ARCHS=80,90,100,100a,103,103a" python3 -m pip install "/tmp/mscclpp[cuda13]"; ;; *) echo "Unsupported CUDA version for MSCCL++: ${CUDA_VERSION}" && exit 1; ;; esac && rm -rf /tmp/mscclpp # buildkit

2026-08-11 07:50:25 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install -c /sgl-workspace/constraints.txt datamodel_code_generator pre-commit pytest black isort icdiff uv wheel scikit-build-core py-spy cubloaty google-cloud-storage pandas matplotlib tabulate termplotlib "runai-model-streamer[s3,gcs,azure]>=0.15.7" # buildkit

2026-08-11 07:50:35 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c if [ "${CUDA_VERSION%%.*}" = "12" ]; then python3 -m pip install nixl nixl-cu12 --no-deps ; python3 -m pip install "cuda-python>=12,<13" ; elif [ "${CUDA_VERSION%%.*}" = "13" ]; then python3 -m pip install nixl nixl-cu13 --no-deps ; python3 -m pip install "cuda-python>=13,<14" ; fi # buildkit

2026-08-11 07:50:35 UTC (buildkit.dockerfile.v0)

COPY --chown=root:root --chmod=755 docker/configs/yank /usr/local/bin/yank # buildkit

2026-08-11 07:50:35 UTC (buildkit.dockerfile.v0)

COPY docker/configs/opt/.vimrc /opt/sglang/.vimrc # buildkit

2026-08-11 07:50:35 UTC (buildkit.dockerfile.v0)

COPY docker/configs/opt/.tmux.conf /opt/sglang/.tmux.conf # buildkit

2026-08-11 07:50:35 UTC (buildkit.dockerfile.v0)

COPY docker/configs/opt/.gitconfig /opt/sglang/.gitconfig # buildkit

2026-08-11 07:50:35 UTC (buildkit.dockerfile.v0)

COPY docker/configs/.zshrc /root/.zshrc # buildkit

2026-08-11 07:50:38 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --upgrade "urllib3>=2.6.3" "pillow>=12.1.1" # buildkit

2026-08-11 07:50:48 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c apt-get update && apt-get install -y --only-upgrade binutils binutils-common binutils-x86-64-linux-gnu libbinutils libctf0 libctf-nobfd0 libgprofng0 libsframe1 libgnutls30t64 libpam-modules libpam-modules-bin libpam-runtime libpam0g libsqlite3-0 libtasn1-6 dpkg dpkg-dev libdpkg-perl && rm -rf /var/lib/apt/lists/* # buildkit

2026-08-11 07:50:48 UTC (buildkit.dockerfile.v0)

ARG BRANCH_TYPE=local

2026-08-11 07:50:48 UTC (buildkit.dockerfile.v0)

ARG BUILD_TYPE=all

2026-08-11 07:50:48 UTC (buildkit.dockerfile.v0)

ARG CUDA_VERSION=12.9.2

2026-08-11 07:50:48 UTC (buildkit.dockerfile.v0)

ARG SGL_VERSION

2026-08-11 07:50:48 UTC (buildkit.dockerfile.v0)

ARG USE_LATEST_SGLANG

2026-08-11 07:50:48 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace

2026-08-11 07:50:57 UTC (buildkit.dockerfile.v0)

COPY /src /tmp/local_src # buildkit

2026-08-11 07:51:02 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c if [ "$BRANCH_TYPE" = "local" ]; then cp -r /tmp/local_src /sgl-workspace/sglang; elif [ "$USE_LATEST_SGLANG" = "1" ]; then git clone --depth=1 https://github.com/sgl-project/sglang.git /sgl-workspace/sglang; elif [ -z "$SGL_VERSION" ]; then echo "ERROR: SGL_VERSION must be set when USE_LATEST_SGLANG=0 and BRANCH_TYPE!=local" && exit 1; else git clone --depth=1 --branch v${SGL_VERSION} https://github.com/sgl-project/sglang.git /sgl-workspace/sglang; fi && rm -rf /tmp/local_src # buildkit

2026-08-11 07:59:09 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c cd /sgl-workspace/sglang && if [ "${CUDA_VERSION%%.*}" = "12" ]; then sed -i 's/cuda-python>=13\.0/cuda-python>=12,<13/' python/pyproject.toml && sed -i 's/flashinfer_python\[cu13\]/flashinfer_python[cu12]/' python/pyproject.toml && sed -i 's/nvidia-cutlass-dsl\[cu13\]/nvidia-cutlass-dsl/' python/pyproject.toml; fi && python3 -m pip install --no-deps -e "python[${BUILD_TYPE}]" && kernels lock python && ( success=0; if [ "$(uname -m)" = "aarch64" ]; then echo "Skipping kernels-community/sgl-flash-attn3 cubin download on aarch64 (no variants published upstream); kernels will be JIT-compiled at runtime"; success=1; else for i in 1 2 3; do echo "Attempt $i/3: downloading sgl-kernel cubins..." && kernels download python && success=1 && break; echo "sgl-kernel cubin download failed, retrying in 30s..." && sleep 30; done; if [ "$success" != "1" ]; then echo "WARNING: no matching sgl-flash-attn3 cubin variant for this torch+CUDA; kernels will be JIT-compiled at runtime"; success=1; fi; fi; [ "$success" = "1" ] ) && mkdir -p /root/.cache/huggingface /root/.cache/sglang && ( if [ -f python/kernels.lock ]; then mv python/kernels.lock /root/.cache/sglang/; fi ) && ( find /usr/local/lib/python3.12/dist-packages -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null || true ) # buildkit

2026-08-11 07:59:09 UTC (buildkit.dockerfile.v0)

COPY /build/sgl-model-gateway-bin /usr/local/bin/sgl-model-gateway # buildkit

2026-08-11 07:59:09 UTC (buildkit.dockerfile.v0)

COPY /build/gateway_wheels /tmp/gateway_wheels # buildkit

2026-08-11 07:59:18 UTC (buildkit.dockerfile.v0)

RUN |20 TARGETARCH=arm64 BUILD_TYPE=all BRANCH_TYPE=local SGL_KERNEL_VERSION=0.4.6.post1 SGL_VERSION= SGL_DEEP_GEMM_VERSION=0.1.5.post2 USE_LATEST_SGLANG=0 GDRCOPY_VERSION=2.5.1 PIP_DEFAULT_INDEX= UBUNTU_MIRROR= GITHUB_ARTIFACTORY=github.com INSTALL_FLASHINFER_JIT_CACHE=1 FLASHINFER_VERSION=0.6.15.post1 TRTLLM_GEN_MOE_CUBIN_URL=https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip TRTLLM_GEN_MOE_CUBIN_SHA256=4900501cbe782a76b08a5858f9f07152287b97cb68114466dac286366b66c192 TRTLLM_GEN_MOE_CUBIN_ARCHIVE_ROOT=trtllm_gen_moe_cubin_pool_20260617_v0613rc1 MOONCAKE_VERSION=0.3.12.post1 MSCCLPP_VERSION=sglang-v0.9.1 MOONCAKE_COMPILE_ARG=-DUSE_HTTP=ON -DUSE_MNNVL=ON -DUSE_CUDA=ON -DWITH_EP=ON CUDA_VERSION=12.9.2 /bin/sh -c python3 -m pip install --force-reinstall /tmp/gateway_wheels/*.whl && rm -rf /tmp/gateway_wheels # buildkit

2026-08-11 07:59:18 UTC (buildkit.dockerfile.v0)

WORKDIR /sgl-workspace/sglang

2026-08-11 07:59:18 UTC (buildkit.dockerfile.v0)

ARG SGLANG_BUILD_COMMIT=8ba213fc67550aea80dfbd5ef66f4574ec25e653

2026-08-11 07:59:18 UTC (buildkit.dockerfile.v0)

ARG SGLANG_BUILD_URL=https://github.com/sgl-project/sglang/actions/runs/31468844644

2026-08-11 07:59:18 UTC (buildkit.dockerfile.v0)

ARG SGLANG_IMAGE_TAG=lmsysorg/sglang:dev-cu12-Ling-3.0-flash

2026-08-11 07:59:18 UTC (buildkit.dockerfile.v0)

ENV SGLANG_BUILD_COMMIT=8ba213fc67550aea80dfbd5ef66f4574ec25e653 SGLANG_BUILD_URL=https://github.com/sgl-project/sglang/actions/runs/31468844644 SGLANG_IMAGE_TAG=lmsysorg/sglang:dev-cu12-Ling-3.0-flash

2026-08-11 07:59:18 UTC (buildkit.dockerfile.v0)

LABEL org.opencontainers.image.source=https://github.com/sgl-project/sglang org.opencontainers.image.revision=8ba213fc67550aea80dfbd5ef66f4574ec25e653 org.opencontainers.image.version=lmsysorg/sglang:dev-cu12-Ling-3.0-flash org.opencontainers.image.url=https://github.com/sgl-project/sglang/actions/runs/31468844644 ai.sglang.build.commit=8ba213fc67550aea80dfbd5ef66f4574ec25e653 ai.sglang.build.url=https://github.com/sgl-project/sglang/actions/runs/31468844644 ai.sglang.image.tag=lmsysorg/sglang:dev-cu12-Ling-3.0-flash

Danger Zone
Delete Tag

Please be careful as this will not just delete the reference but also the actual content!

For example when you have latest and v1.2.3 both pointing to the same image
the deletion of latest will also permanently remove v1.2.3.

Delete