docker.io/approachingai/ktransformers:DSV4-specific

linux/amd64 docker.io

approachingai/ktransformers 镜像描述

这是一个包含Keras Transformer层的Docker镜像。Keras Transformer提供了一组用于构建Transformer模型的预构建层，方便开发者使用Keras构建各种Transformer架构，例如BERT、GPT等。使用此镜像可以方便地进行Transformer模型的训练和推理，无需手动安装和配置相关的依赖项。

浏览次数

24.68GB

镜像大小

国内镜像

swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific

源镜像

docker.io/approachingai/ktransformers:DSV4-specific

镜像ID

sha256:9748713f873a96ae267657206a50bfcdf77210991418ec6b46b99e75ca066c10

镜像 TAG

DSV4-specific

镜像大小

24.68GB

平台架构

linux/amd64

镜像源

docker.io

项目信息

Hub 主页所有 TAG

CMD

启动入口

/usr/local/bin/entrypoint-dsv4

工作目录

/workspace

OS/平台

linux/amd64

镜像创建

2026-08-08T09:25:58.915391886Z

同步时间

2026-08-13 06:05

浏览量

15 次

贡献者

🔌 开放端口 1

30000/tcp

⚙️ 环境变量 49

Key	Value
`PATH=/opt/venv/bin:/usr/local/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin`	`0`
`NVARCH=x86_64`	`1`
NVIDIA_REQUIRE_CUDA=cuda>=12.8 brand=unknown,driver>=470,driver<471 brand=grid,driver>=470,driver<471 brand=tesla,driver>=470,driver<471 brand=nvidia,driver>=470,driver<471 brand=quadro,driver>=470,driver<471 brand=quadrortx,driver>=470,driver<471 brand=nvidiartx,driver>=470,driver<471 brand=vapps,driver>=470,driver<471 brand=vpc,driver>=470,driver<471 brand=vcs,driver>=470,driver<471 brand=vws,driver>=470,driver<471 brand=cloudgaming,driver>=470,driver<471 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	`2`
`NV_CUDA_CUDART_VERSION=12.8.90-1`	`3`
`CUDA_VERSION=12.8.1`	`4`
`LD_LIBRARY_PATH=/usr/local/cuda/lib64`	`5`
`NVIDIA_VISIBLE_DEVICES=all`	`6`
`NVIDIA_DRIVER_CAPABILITIES=compute,utility`	`7`
`NV_CUDA_LIB_VERSION=12.8.1-1`	`8`
`NV_NVTX_VERSION=12.8.90-1`	`9`
`NV_LIBNPP_VERSION=12.3.3.100-1`	`10`
`NV_LIBNPP_PACKAGE=libnpp-12-8=12.3.3.100-1`	`11`
`NV_LIBCUSPARSE_VERSION=12.5.8.93-1`	`12`
`NV_LIBCUBLAS_PACKAGE_NAME=libcublas-12-8`	`13`
`NV_LIBCUBLAS_VERSION=12.8.4.1-1`	`14`
`NV_LIBCUBLAS_PACKAGE=libcublas-12-8=12.8.4.1-1`	`15`
`NV_LIBNCCL_PACKAGE_NAME=libnccl2`	`16`
`NV_LIBNCCL_PACKAGE_VERSION=2.25.1-1`	`17`
`NCCL_VERSION=2.25.1-1`	`18`
`NV_LIBNCCL_PACKAGE=libnccl2=2.25.1-1+cuda12.8`	`19`
`NVIDIA_PRODUCT_NAME=CUDA`	`20`
`NV_CUDA_CUDART_DEV_VERSION=12.8.90-1`	`21`
`NV_NVML_DEV_VERSION=12.8.90-1`	`22`
`NV_LIBCUSPARSE_DEV_VERSION=12.5.8.93-1`	`23`
`NV_LIBNPP_DEV_VERSION=12.3.3.100-1`	`24`
`NV_LIBNPP_DEV_PACKAGE=libnpp-dev-12-8=12.3.3.100-1`	`25`
`NV_LIBCUBLAS_DEV_VERSION=12.8.4.1-1`	`26`
`NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-12-8`	`27`
`NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-12-8=12.8.4.1-1`	`28`
`NV_CUDA_NSIGHT_COMPUTE_VERSION=12.8.1-1`	`29`
`NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-12-8=12.8.1-1`	`30`
`NV_NVPROF_VERSION=12.8.90-1`	`31`
`NV_NVPROF_DEV_PACKAGE=cuda-nvprof-12-8=12.8.90-1`	`32`
`NV_LIBNCCL_DEV_PACKAGE_NAME=libnccl-dev`	`33`
`NV_LIBNCCL_DEV_PACKAGE_VERSION=2.25.1-1`	`34`
`NV_LIBNCCL_DEV_PACKAGE=libnccl-dev=2.25.1-1+cuda12.8`	`35`
`LIBRARY_PATH=/usr/local/cuda/lib64/stubs`	`36`
`NV_CUDNN_VERSION=9.8.0.87-1`	`37`
`NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-12`	`38`
`NV_CUDNN_PACKAGE=libcudnn9-cuda-12=9.8.0.87-1`	`39`
`NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-12=9.8.0.87-1`	`40`
`CUDA_HOME=/usr/local/cuda`	`41`
`DEBIAN_FRONTEND=noninteractive`	`42`
`PIP_DISABLE_PIP_VERSION_CHECK=1`	`43`
`PIP_NO_CACHE_DIR=1`	`44`
`PYTHONUNBUFFERED=1`	`45`
`SGLANG_ENABLE_HEALTH_ENDPOINT_GENERATION=false`	`46`
`MODEL_PATH=/model`	`47`
`PORT=30000`	`48`

🏷️ 镜像标签 16

Key	Value
`9.8.0.87-1`	`com.nvidia.cudnn.version`
`12.8.1`	`io.ktransformers.cuda`
`0.6.15.post1`	`io.ktransformers.flashinfer`
`25cfd305afa1f34baa2f419bca35db58c369c534ee1961662e83d9fe858ce021`	`io.ktransformers.flashinfer-cubin-sha256`
`enabled-threshold-2048-swa-0.4`	`io.ktransformers.layerwise-prefill-default`
`bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2`	`io.ktransformers.sglang-revision`
`https://github.com/kvcache-ai/sglang`	`io.ktransformers.sglang-source`
`80,86,89,90,100,120`	`io.ktransformers.sm`
`f5ee50d4cfa29f5f3e8f2af5a2fd90771c574c58`	`io.ktransformers.upstream-ktransformers-revision`
`NVIDIA CORPORATION <cudatools@nvidia.com>`	`maintainer`
`2026-08-08T09:17:45Z`	`org.opencontainers.image.created`
`ubuntu`	`org.opencontainers.image.ref.name`
`90ddf8062d7e3593ffeec5612b47b36a34c7c77d`	`org.opencontainers.image.revision`
`https://github.com/yyj6666667/ktransformers`	`org.opencontainers.image.source`
`KTransformers DeepSeek-V4-Flash`	`org.opencontainers.image.title`
`dsv4-f5ee50d-bc7f005-fi0615`	`org.opencontainers.image.version`

🛡️ 镜像安全扫描

ubuntu 22.04 Trivy 2026-08-13 06:09 查看完整报告

553

低危 LOW

4334

中危 MEDIUM

363

高危 HIGH

严重 CRITICAL

受影响目标 (3)

docker.io/approachingai/ktransformers:DSV4-specific (ubuntu 22.04) ubuntu Python python-pkg opt/nvidia/nsight-compute/2025.1.1/host/target-linux-x64/plugins/efa_metrics/nic_sampler gobinary

Chatgpt|Claude一站式支持平台提供先进的AI镜像服务，助你在国内优雅的使用Chatgpt和Claude等AI服务

无权限下载？点我修复生成 Docker Run 命令生成 Kubernetes YAML

Docker拉取命令

docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific
docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific  docker.io/approachingai/ktransformers:DSV4-specific

Containerd拉取命令

ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific
ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific  docker.io/approachingai/ktransformers:DSV4-specific

Shell快速替换命令

sed -i 's#approachingai/ktransformers:DSV4-specific#swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific#' deployment.yaml

Ansible快速分发-Docker

#ansible k8s -m shell -a 'docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific && docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific  docker.io/approachingai/ktransformers:DSV4-specific'

Ansible快速分发-Containerd

#ansible k8s -m shell -a 'ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific && ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific  docker.io/approachingai/ktransformers:DSV4-specific'

镜像构建历史


# 2026-08-08 17:25:58  0.00B 设置默认要执行的命令
CMD []
                        
# 2026-08-08 17:25:58  0.00B 配置容器启动时运行的命令
ENTRYPOINT ["/usr/local/bin/entrypoint-dsv4"]
                        
# 2026-08-08 17:25:58  0.00B 设置停止容器时发送的系统调用信号
STOPSIGNAL SIGTERM
                        
# 2026-08-08 17:25:58  0.00B 指定检查容器健康状态的命令
HEALTHCHECK &{["CMD-SHELL" "curl --fail --silent --show-error \"http://127.0.0.1:$PORT/health\" >/dev/null || exit 1"] "30s" "10s" "15m0s" "0s" '\x14'}
                        
# 2026-08-08 17:25:58  0.00B 声明容器运行时监听的端口
EXPOSE [30000/tcp]
                        
# 2026-08-08 17:25:58  0.00B 设置工作目录为/workspace
WORKDIR /workspace
                        
# 2026-08-08 17:25:58  8.70KB 执行命令并创建新的镜像层
RUN |17 IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615 VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849 SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 BUILD_DATE=2026-08-08T08:57:15Z SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers SGLANG_SOURCE=https://github.com/kvcache-ai/sglang PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple TORCH_VERSION=2.9.1 SGL_KERNEL_VERSION=0.3.21 FLASHINFER_VERSION=0.6.15.post1 TRANSFORMERS_VERSION=4.57.1 TILELANG_VERSION=0.1.10 TVM_FFI_VERSION=0.1.11 CPUINFER_CUDA_ARCHS=80;86;89;90;100;120 BUILD_JOBS=16 FLASHINFER_CUBIN_INDEX_URL=http://172.17.0.3:38151 FLASHINFER_CUBIN_TRUSTED_HOST=172.17.0.3 /bin/bash -o pipefail -c chmod 0755 /usr/local/bin/entrypoint-dsv4 # buildkit
                        
# 2026-08-08 17:25:58  8.70KB 复制新文件或目录到容器中
COPY docker/entrypoint-dsv4.sh /usr/local/bin/entrypoint-dsv4 # buildkit
                        
# 2026-08-08 17:25:58  0.00B 添加元数据标签
LABEL org.opencontainers.image.title=KTransformers DeepSeek-V4-Flash org.opencontainers.image.source=https://github.com/yyj6666667/ktransformers org.opencontainers.image.version=dsv4-f5ee50d-bc7f005-fi0615 org.opencontainers.image.revision=5383a5b3675fd155f521ffb125955abc93c93849 org.opencontainers.image.created=2026-08-08T08:57:15Z io.ktransformers.cuda=12.8.1 io.ktransformers.sm=80,86,89,90,100,120 io.ktransformers.flashinfer=0.6.15.post1 io.ktransformers.sglang-source=https://github.com/kvcache-ai/sglang io.ktransformers.sglang-revision=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 io.ktransformers.layerwise-prefill-default=enabled-threshold-2048-swa-0.4
                        
# 2026-08-08 17:25:58  6.87KB 执行命令并创建新的镜像层
RUN |17 IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615 VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849 SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 BUILD_DATE=2026-08-08T08:57:15Z SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers SGLANG_SOURCE=https://github.com/kvcache-ai/sglang PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple TORCH_VERSION=2.9.1 SGL_KERNEL_VERSION=0.3.21 FLASHINFER_VERSION=0.6.15.post1 TRANSFORMERS_VERSION=4.57.1 TILELANG_VERSION=0.1.10 TVM_FFI_VERSION=0.1.11 CPUINFER_CUDA_ARCHS=80;86;89;90;100;120 BUILD_JOBS=16 FLASHINFER_CUBIN_INDEX_URL=http://172.17.0.3:38151 FLASHINFER_CUBIN_TRUSTED_HOST=172.17.0.3 /bin/bash -o pipefail -c chmod 0755 /usr/local/bin/verify-dsv4-env     && /usr/local/bin/verify-dsv4-env --skip-cpuinfer     && rm -rf /workspace/ktransformers/kt-kernel/build # buildkit
                        
# 2026-08-08 17:25:52  6.87KB 复制新文件或目录到容器中
COPY docker/verify-dsv4-env.py /usr/local/bin/verify-dsv4-env # buildkit
                        
# 2026-08-08 17:25:52  358.04MB 执行命令并创建新的镜像层
RUN |17 IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615 VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849 SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 BUILD_DATE=2026-08-08T08:57:15Z SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers SGLANG_SOURCE=https://github.com/kvcache-ai/sglang PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple TORCH_VERSION=2.9.1 SGL_KERNEL_VERSION=0.3.21 FLASHINFER_VERSION=0.6.15.post1 TRANSFORMERS_VERSION=4.57.1 TILELANG_VERSION=0.1.10 TVM_FFI_VERSION=0.1.11 CPUINFER_CUDA_ARCHS=80;86;89;90;100;120 BUILD_JOBS=16 FLASHINFER_CUBIN_INDEX_URL=http://172.17.0.3:38151 FLASHINFER_CUBIN_TRUSTED_HOST=172.17.0.3 /bin/bash -o pipefail -c test -f /workspace/ktransformers/third_party/sglang/python/pyproject.toml     && test -f /workspace/ktransformers/third_party/pybind11/CMakeLists.txt     && export SGLANG_KT_VERSION="$(python -c "exec(open('/workspace/ktransformers/version.py').read()); print(__version__)")"     && python -m pip install --no-deps         /workspace/ktransformers/third_party/sglang/python     && cd /workspace/ktransformers/kt-kernel     && CPUINFER_BUILD_ALL_VARIANTS=1        CPUINFER_USE_CUDA=1        CPUINFER_CUDA_ARCHS="${CPUINFER_CUDA_ARCHS}"        CPUINFER_PARALLEL="${BUILD_JOBS}"        ./install.sh build     && python -m pip install --no-deps /workspace/ktransformers # buildkit
                        
# 2026-08-08 17:19:02  132.68MB 复制新文件或目录到容器中
COPY . /workspace/ktransformers # buildkit
                        
# 2026-08-08 17:19:00  6.04GB 执行命令并创建新的镜像层
RUN |17 IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615 VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849 SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 BUILD_DATE=2026-08-08T08:57:15Z SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers SGLANG_SOURCE=https://github.com/kvcache-ai/sglang PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple TORCH_VERSION=2.9.1 SGL_KERNEL_VERSION=0.3.21 FLASHINFER_VERSION=0.6.15.post1 TRANSFORMERS_VERSION=4.57.1 TILELANG_VERSION=0.1.10 TVM_FFI_VERSION=0.1.11 CPUINFER_CUDA_ARCHS=80;86;89;90;100;120 BUILD_JOBS=16 FLASHINFER_CUBIN_INDEX_URL=http://172.17.0.3:38151 FLASHINFER_CUBIN_TRUSTED_HOST=172.17.0.3 /bin/bash -o pipefail -c python -m pip install -c /tmp/constraints-dsv4.txt         "sgl-kernel==${SGL_KERNEL_VERSION}"         "flashinfer-python==${FLASHINFER_VERSION}"         "transformers==${TRANSFORMERS_VERSION}"         "tilelang==${TILELANG_VERSION}"         "apache-tvm-ffi==${TVM_FFI_VERSION}"         "cuda-python==13.2.0"         "nvidia-cutlass-dsl==4.6.1"     && python -m pip install         --index-url "${FLASHINFER_CUBIN_INDEX_URL}"         --trusted-host "${FLASHINFER_CUBIN_TRUSTED_HOST}"         --no-deps         "flashinfer-cubin==${FLASHINFER_VERSION}"     && grep -v '^quack-kernels=='         /tmp/requirements-dsv4.txt         > /tmp/requirements-dsv4-resolved.txt     && python -m pip install         -c /tmp/constraints-dsv4.txt         -r /tmp/requirements-dsv4-resolved.txt     && python -m pip install --no-deps "quack-kernels==0.2.4" # buildkit
                        
# 2026-08-08 17:05:54  0.00B 定义构建参数
ARG FLASHINFER_CUBIN_TRUSTED_HOST=172.17.0.3
                        
# 2026-08-08 17:05:54  0.00B 定义构建参数
ARG FLASHINFER_CUBIN_INDEX_URL=http://172.17.0.3:38151
                        
# 2026-08-08 17:05:54  374.00B 复制新文件或目录到容器中
COPY docker/constraints-dsv4.txt /tmp/constraints-dsv4.txt # buildkit
                        
# 2026-08-08 17:05:53  7.20GB 执行命令并创建新的镜像层
RUN |15 IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615 VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849 SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 BUILD_DATE=2026-08-08T08:57:15Z SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers SGLANG_SOURCE=https://github.com/kvcache-ai/sglang PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple TORCH_VERSION=2.9.1 SGL_KERNEL_VERSION=0.3.21 FLASHINFER_VERSION=0.6.15.post1 TRANSFORMERS_VERSION=4.57.1 TILELANG_VERSION=0.1.10 TVM_FFI_VERSION=0.1.11 CPUINFER_CUDA_ARCHS=80;86;89;90;100;120 BUILD_JOBS=16 /bin/bash -o pipefail -c python -m pip install         "torch==${TORCH_VERSION}"         "torchvision==0.24.1"         "torchaudio==${TORCH_VERSION}"         --index-url https://download.pytorch.org/whl/cu128         --extra-index-url "${PIP_INDEX_URL}"         --retries 10         --timeout 120 # buildkit
                        
# 2026-08-08 16:58:27  1.26KB 复制新文件或目录到容器中
COPY docker/requirements-dsv4.txt /tmp/requirements-dsv4.txt # buildkit
                        
# 2026-08-08 16:58:27  77.05MB 执行命令并创建新的镜像层
RUN |15 IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615 VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849 SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 BUILD_DATE=2026-08-08T08:57:15Z SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers SGLANG_SOURCE=https://github.com/kvcache-ai/sglang PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple TORCH_VERSION=2.9.1 SGL_KERNEL_VERSION=0.3.21 FLASHINFER_VERSION=0.6.15.post1 TRANSFORMERS_VERSION=4.57.1 TILELANG_VERSION=0.1.10 TVM_FFI_VERSION=0.1.11 CPUINFER_CUDA_ARCHS=80;86;89;90;100;120 BUILD_JOBS=16 /bin/bash -o pipefail -c python -m pip config set global.index-url "${PIP_INDEX_URL}"     && python -m pip install --upgrade setuptools wheel     && python -m pip install "cmake>=3.31,<4" ninja # buildkit
                        
# 2026-08-08 16:58:21  492.25MB 执行命令并创建新的镜像层
RUN |15 IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615 VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849 SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2 BUILD_DATE=2026-08-08T08:57:15Z SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers SGLANG_SOURCE=https://github.com/kvcache-ai/sglang PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple TORCH_VERSION=2.9.1 SGL_KERNEL_VERSION=0.3.21 FLASHINFER_VERSION=0.6.15.post1 TRANSFORMERS_VERSION=4.57.1 TILELANG_VERSION=0.1.10 TVM_FFI_VERSION=0.1.11 CPUINFER_CUDA_ARCHS=80;86;89;90;100;120 BUILD_JOBS=16 /bin/bash -o pipefail -c apt-get update     && apt-get install -y --no-install-recommends         build-essential         ca-certificates         curl         git         libhwloc-dev         libnuma-dev         libopenmpi-dev         numactl         pkg-config         software-properties-common     && add-apt-repository -y ppa:deadsnakes/ppa     && apt-get update     && apt-get install -y --no-install-recommends         python3.11         python3.11-dev         python3.11-venv     && rm -rf /var/lib/apt/lists/*     && python3.11 -m venv /opt/venv # buildkit
                        
# 2026-08-08 16:58:21  0.00B 设置环境变量 CUDA_HOME DEBIAN_FRONTEND PIP_DISABLE_PIP_VERSION_CHECK PIP_NO_CACHE_DIR PYTHONUNBUFFERED SGLANG_ENABLE_HEALTH_ENDPOINT_GENERATION MODEL_PATH PORT PATH
ENV CUDA_HOME=/usr/local/cuda DEBIAN_FRONTEND=noninteractive PIP_DISABLE_PIP_VERSION_CHECK=1 PIP_NO_CACHE_DIR=1 PYTHONUNBUFFERED=1 SGLANG_ENABLE_HEALTH_ENDPOINT_GENERATION=false MODEL_PATH=/model PORT=30000 PATH=/opt/venv/bin:/usr/local/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG BUILD_JOBS=16
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG CPUINFER_CUDA_ARCHS=80;86;89;90;100;120
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG TVM_FFI_VERSION=0.1.11
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG TILELANG_VERSION=0.1.10
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG TRANSFORMERS_VERSION=4.57.1
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG FLASHINFER_VERSION=0.6.15.post1
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG SGL_KERNEL_VERSION=0.3.21
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG TORCH_VERSION=2.9.1
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG SGLANG_SOURCE=https://github.com/kvcache-ai/sglang
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG SOURCE_REPOSITORY=https://github.com/yyj6666667/ktransformers
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG BUILD_DATE=2026-08-08T08:57:15Z
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG SGLANG_REF=bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG VCS_REF=5383a5b3675fd155f521ffb125955abc93c93849
                        
# 2026-08-08 16:58:21  0.00B 定义构建参数
ARG IMAGE_VERSION=dsv4-f5ee50d-bc7f005-fi0615
                        
# 2026-08-08 16:58:21  0.00B 
SHELL [/bin/bash -o pipefail -c]
                        
# 2025-03-11 07:14:56  1.05GB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     ${NV_CUDNN_PACKAGE}     ${NV_CUDNN_PACKAGE_DEV}     && apt-mark hold ${NV_CUDNN_PACKAGE_NAME}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2025-03-11 07:14:56  0.00B 添加元数据标签
LABEL com.nvidia.cudnn.version=9.8.0.87-1
                        
# 2025-03-11 07:14:56  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2025-03-11 07:14:56  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2025-03-11 07:14:56  0.00B 设置环境变量 NV_CUDNN_PACKAGE_DEV
ENV NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-12=9.8.0.87-1
                        
# 2025-03-11 07:14:56  0.00B 设置环境变量 NV_CUDNN_PACKAGE
ENV NV_CUDNN_PACKAGE=libcudnn9-cuda-12=9.8.0.87-1
                        
# 2025-03-11 07:14:56  0.00B 设置环境变量 NV_CUDNN_PACKAGE_NAME
ENV NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-12
                        
# 2025-03-11 07:14:56  0.00B 设置环境变量 NV_CUDNN_VERSION
ENV NV_CUDNN_VERSION=9.8.0.87-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 LIBRARY_PATH
ENV LIBRARY_PATH=/usr/local/cuda/lib64/stubs
                        
# 2025-03-11 06:36:52  389.48KB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_DEV_PACKAGE_NAME} ${NV_LIBNCCL_DEV_PACKAGE_NAME} # buildkit
                        
# 2025-03-11 06:36:52  5.94GB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-cudart-dev-12-8=${NV_CUDA_CUDART_DEV_VERSION}     cuda-command-line-tools-12-8=${NV_CUDA_LIB_VERSION}     cuda-minimal-build-12-8=${NV_CUDA_LIB_VERSION}     cuda-libraries-dev-12-8=${NV_CUDA_LIB_VERSION}     cuda-nvml-dev-12-8=${NV_NVML_DEV_VERSION}     ${NV_NVPROF_DEV_PACKAGE}     ${NV_LIBNPP_DEV_PACKAGE}     libcusparse-dev-12-8=${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
                        
# 2025-03-11 06:36:52  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2025-03-11 06:36:52  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBNCCL_DEV_PACKAGE
ENV NV_LIBNCCL_DEV_PACKAGE=libnccl-dev=2.25.1-1+cuda12.8
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NCCL_VERSION
ENV NCCL_VERSION=2.25.1-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBNCCL_DEV_PACKAGE_VERSION
ENV NV_LIBNCCL_DEV_PACKAGE_VERSION=2.25.1-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBNCCL_DEV_PACKAGE_NAME
ENV NV_LIBNCCL_DEV_PACKAGE_NAME=libnccl-dev
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_NVPROF_DEV_PACKAGE
ENV NV_NVPROF_DEV_PACKAGE=cuda-nvprof-12-8=12.8.90-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_NVPROF_VERSION
ENV NV_NVPROF_VERSION=12.8.90-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE
ENV NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-12-8=12.8.1-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_CUDA_NSIGHT_COMPUTE_VERSION
ENV NV_CUDA_NSIGHT_COMPUTE_VERSION=12.8.1-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBCUBLAS_DEV_PACKAGE
ENV NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-12-8=12.8.4.1-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBCUBLAS_DEV_PACKAGE_NAME
ENV NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-12-8
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBCUBLAS_DEV_VERSION
ENV NV_LIBCUBLAS_DEV_VERSION=12.8.4.1-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBNPP_DEV_PACKAGE
ENV NV_LIBNPP_DEV_PACKAGE=libnpp-dev-12-8=12.3.3.100-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBNPP_DEV_VERSION
ENV NV_LIBNPP_DEV_VERSION=12.3.3.100-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_LIBCUSPARSE_DEV_VERSION
ENV NV_LIBCUSPARSE_DEV_VERSION=12.5.8.93-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_NVML_DEV_VERSION
ENV NV_NVML_DEV_VERSION=12.8.90-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_CUDA_CUDART_DEV_VERSION
ENV NV_CUDA_CUDART_DEV_VERSION=12.8.90-1
                        
# 2025-03-11 06:36:52  0.00B 设置环境变量 NV_CUDA_LIB_VERSION
ENV NV_CUDA_LIB_VERSION=12.8.1-1
                        
# 2025-03-11 06:24:31  0.00B 配置容器启动时运行的命令
ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NVIDIA_PRODUCT_NAME
ENV NVIDIA_PRODUCT_NAME=CUDA
                        
# 2025-03-11 06:24:31  2.53KB 复制新文件或目录到容器中
COPY nvidia_entrypoint.sh /opt/nvidia/ # buildkit
                        
# 2025-03-11 06:24:31  3.06KB 复制新文件或目录到容器中
COPY entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit
                        
# 2025-03-11 06:24:31  263.00KB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_PACKAGE_NAME} ${NV_LIBNCCL_PACKAGE_NAME} # buildkit
                        
# 2025-03-11 06:24:31  3.11GB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-libraries-12-8=${NV_CUDA_LIB_VERSION}     ${NV_LIBNPP_PACKAGE}     cuda-nvtx-12-8=${NV_NVTX_VERSION}     libcusparse-12-8=${NV_LIBCUSPARSE_VERSION}     ${NV_LIBCUBLAS_PACKAGE}     ${NV_LIBNCCL_PACKAGE}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2025-03-11 06:24:31  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2025-03-11 06:24:31  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE
ENV NV_LIBNCCL_PACKAGE=libnccl2=2.25.1-1+cuda12.8
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NCCL_VERSION
ENV NCCL_VERSION=2.25.1-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE_VERSION
ENV NV_LIBNCCL_PACKAGE_VERSION=2.25.1-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE_NAME
ENV NV_LIBNCCL_PACKAGE_NAME=libnccl2
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBCUBLAS_PACKAGE
ENV NV_LIBCUBLAS_PACKAGE=libcublas-12-8=12.8.4.1-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBCUBLAS_VERSION
ENV NV_LIBCUBLAS_VERSION=12.8.4.1-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBCUBLAS_PACKAGE_NAME
ENV NV_LIBCUBLAS_PACKAGE_NAME=libcublas-12-8
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBCUSPARSE_VERSION
ENV NV_LIBCUSPARSE_VERSION=12.5.8.93-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBNPP_PACKAGE
ENV NV_LIBNPP_PACKAGE=libnpp-12-8=12.3.3.100-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_LIBNPP_VERSION
ENV NV_LIBNPP_VERSION=12.3.3.100-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_NVTX_VERSION
ENV NV_NVTX_VERSION=12.8.90-1
                        
# 2025-03-11 06:24:31  0.00B 设置环境变量 NV_CUDA_LIB_VERSION
ENV NV_CUDA_LIB_VERSION=12.8.1-1
                        
# 2025-03-11 06:19:20  0.00B 设置环境变量 NVIDIA_DRIVER_CAPABILITIES
ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility
                        
# 2025-03-11 06:19:20  0.00B 设置环境变量 NVIDIA_VISIBLE_DEVICES
ENV NVIDIA_VISIBLE_DEVICES=all
                        
# 2025-03-11 06:19:20  17.29KB 复制新文件或目录到容器中
COPY NGC-DL-CONTAINER-LICENSE / # buildkit
                        
# 2025-03-11 06:19:20  0.00B 设置环境变量 LD_LIBRARY_PATH
ENV LD_LIBRARY_PATH=/usr/local/cuda/lib64
                        
# 2025-03-11 06:19:20  0.00B 设置环境变量 PATH
ENV PATH=/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
                        
# 2025-03-11 06:19:20  22.00B 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c echo "/usr/local/cuda/lib64" >> /etc/ld.so.conf.d/nvidia.conf # buildkit
                        
# 2025-03-11 06:19:20  203.35MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-cudart-12-8=${NV_CUDA_CUDART_VERSION}     cuda-compat-12-8     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2025-03-11 06:19:05  0.00B 设置环境变量 CUDA_VERSION
ENV CUDA_VERSION=12.8.1
                        
# 2025-03-11 06:19:05  10.60MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     gnupg2 curl ca-certificates &&     curl -fsSLO https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/${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
                        
# 2025-03-11 06:19:05  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2025-03-11 06:19:05  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2025-03-11 06:19:05  0.00B 设置环境变量 NV_CUDA_CUDART_VERSION
ENV NV_CUDA_CUDART_VERSION=12.8.90-1
                        
# 2025-03-11 06:19:05  0.00B 设置环境变量 NVIDIA_REQUIRE_CUDA brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand
ENV NVIDIA_REQUIRE_CUDA=cuda>=12.8 brand=unknown,driver>=470,driver<471 brand=grid,driver>=470,driver<471 brand=tesla,driver>=470,driver<471 brand=nvidia,driver>=470,driver<471 brand=quadro,driver>=470,driver<471 brand=quadrortx,driver>=470,driver<471 brand=nvidiartx,driver>=470,driver<471 brand=vapps,driver>=470,driver<471 brand=vpc,driver>=470,driver<471 brand=vcs,driver>=470,driver<471 brand=vws,driver>=470,driver<471 brand=cloudgaming,driver>=470,driver<471 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
                        
# 2025-03-11 06:19:05  0.00B 设置环境变量 NVARCH
ENV NVARCH=x86_64
                        
# 2025-01-26 13:31:11  0.00B 
/bin/sh -c #(nop)  CMD ["/bin/bash"]
                        
# 2025-01-26 13:31:10  77.86MB 
/bin/sh -c #(nop) ADD file:1b6c8c9518be42fa2afe5e241ca31677fce58d27cdfa88baa91a65a259be3637 in / 
                        
# 2025-01-26 13:31:07  0.00B 
/bin/sh -c #(nop)  LABEL org.opencontainers.image.version=22.04
                        
# 2025-01-26 13:31:07  0.00B 
/bin/sh -c #(nop)  LABEL org.opencontainers.image.ref.name=ubuntu
                        
# 2025-01-26 13:31:07  0.00B 
/bin/sh -c #(nop)  ARG LAUNCHPAD_BUILD_ARCH
                        
# 2025-01-26 13:31:07  0.00B 
/bin/sh -c #(nop)  ARG RELEASE

镜像信息

{
    "Id": "sha256:9748713f873a96ae267657206a50bfcdf77210991418ec6b46b99e75ca066c10",
    "RepoTags": [
        "approachingai/ktransformers:DSV4-specific",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers:DSV4-specific"
    ],
    "RepoDigests": [
        "approachingai/ktransformers@sha256:60962f734b682e6e51ce884f2b39f39fbe3c7b027c16aa15d410b90721c0ac96",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/approachingai/ktransformers@sha256:c24613762ddb492c9b43244131921c0a5f0f9571f26539fdea765bff03b956c1"
    ],
    "Parent": "",
    "Comment": "buildkit.dockerfile.v0",
    "Created": "2026-08-08T09:25:58.915391886Z",
    "Container": "",
    "ContainerConfig": null,
    "DockerVersion": "",
    "Author": "",
    "Config": {
        "Hostname": "",
        "Domainname": "",
        "User": "",
        "AttachStdin": false,
        "AttachStdout": false,
        "AttachStderr": false,
        "ExposedPorts": {
            "30000/tcp": {}
        },
        "Tty": false,
        "OpenStdin": false,
        "StdinOnce": false,
        "Env": [
            "PATH=/opt/venv/bin:/usr/local/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
            "NVARCH=x86_64",
            "NVIDIA_REQUIRE_CUDA=cuda\u003e=12.8 brand=unknown,driver\u003e=470,driver\u003c471 brand=grid,driver\u003e=470,driver\u003c471 brand=tesla,driver\u003e=470,driver\u003c471 brand=nvidia,driver\u003e=470,driver\u003c471 brand=quadro,driver\u003e=470,driver\u003c471 brand=quadrortx,driver\u003e=470,driver\u003c471 brand=nvidiartx,driver\u003e=470,driver\u003c471 brand=vapps,driver\u003e=470,driver\u003c471 brand=vpc,driver\u003e=470,driver\u003c471 brand=vcs,driver\u003e=470,driver\u003c471 brand=vws,driver\u003e=470,driver\u003c471 brand=cloudgaming,driver\u003e=470,driver\u003c471 brand=unknown,driver\u003e=535,driver\u003c536 brand=grid,driver\u003e=535,driver\u003c536 brand=tesla,driver\u003e=535,driver\u003c536 brand=nvidia,driver\u003e=535,driver\u003c536 brand=quadro,driver\u003e=535,driver\u003c536 brand=quadrortx,driver\u003e=535,driver\u003c536 brand=nvidiartx,driver\u003e=535,driver\u003c536 brand=vapps,driver\u003e=535,driver\u003c536 brand=vpc,driver\u003e=535,driver\u003c536 brand=vcs,driver\u003e=535,driver\u003c536 brand=vws,driver\u003e=535,driver\u003c536 brand=cloudgaming,driver\u003e=535,driver\u003c536 brand=unknown,driver\u003e=550,driver\u003c551 brand=grid,driver\u003e=550,driver\u003c551 brand=tesla,driver\u003e=550,driver\u003c551 brand=nvidia,driver\u003e=550,driver\u003c551 brand=quadro,driver\u003e=550,driver\u003c551 brand=quadrortx,driver\u003e=550,driver\u003c551 brand=nvidiartx,driver\u003e=550,driver\u003c551 brand=vapps,driver\u003e=550,driver\u003c551 brand=vpc,driver\u003e=550,driver\u003c551 brand=vcs,driver\u003e=550,driver\u003c551 brand=vws,driver\u003e=550,driver\u003c551 brand=cloudgaming,driver\u003e=550,driver\u003c551 brand=unknown,driver\u003e=560,driver\u003c561 brand=grid,driver\u003e=560,driver\u003c561 brand=tesla,driver\u003e=560,driver\u003c561 brand=nvidia,driver\u003e=560,driver\u003c561 brand=quadro,driver\u003e=560,driver\u003c561 brand=quadrortx,driver\u003e=560,driver\u003c561 brand=nvidiartx,driver\u003e=560,driver\u003c561 brand=vapps,driver\u003e=560,driver\u003c561 brand=vpc,driver\u003e=560,driver\u003c561 brand=vcs,driver\u003e=560,driver\u003c561 brand=vws,driver\u003e=560,driver\u003c561 brand=cloudgaming,driver\u003e=560,driver\u003c561 brand=unknown,driver\u003e=565,driver\u003c566 brand=grid,driver\u003e=565,driver\u003c566 brand=tesla,driver\u003e=565,driver\u003c566 brand=nvidia,driver\u003e=565,driver\u003c566 brand=quadro,driver\u003e=565,driver\u003c566 brand=quadrortx,driver\u003e=565,driver\u003c566 brand=nvidiartx,driver\u003e=565,driver\u003c566 brand=vapps,driver\u003e=565,driver\u003c566 brand=vpc,driver\u003e=565,driver\u003c566 brand=vcs,driver\u003e=565,driver\u003c566 brand=vws,driver\u003e=565,driver\u003c566 brand=cloudgaming,driver\u003e=565,driver\u003c566",
            "NV_CUDA_CUDART_VERSION=12.8.90-1",
            "CUDA_VERSION=12.8.1",
            "LD_LIBRARY_PATH=/usr/local/cuda/lib64",
            "NVIDIA_VISIBLE_DEVICES=all",
            "NVIDIA_DRIVER_CAPABILITIES=compute,utility",
            "NV_CUDA_LIB_VERSION=12.8.1-1",
            "NV_NVTX_VERSION=12.8.90-1",
            "NV_LIBNPP_VERSION=12.3.3.100-1",
            "NV_LIBNPP_PACKAGE=libnpp-12-8=12.3.3.100-1",
            "NV_LIBCUSPARSE_VERSION=12.5.8.93-1",
            "NV_LIBCUBLAS_PACKAGE_NAME=libcublas-12-8",
            "NV_LIBCUBLAS_VERSION=12.8.4.1-1",
            "NV_LIBCUBLAS_PACKAGE=libcublas-12-8=12.8.4.1-1",
            "NV_LIBNCCL_PACKAGE_NAME=libnccl2",
            "NV_LIBNCCL_PACKAGE_VERSION=2.25.1-1",
            "NCCL_VERSION=2.25.1-1",
            "NV_LIBNCCL_PACKAGE=libnccl2=2.25.1-1+cuda12.8",
            "NVIDIA_PRODUCT_NAME=CUDA",
            "NV_CUDA_CUDART_DEV_VERSION=12.8.90-1",
            "NV_NVML_DEV_VERSION=12.8.90-1",
            "NV_LIBCUSPARSE_DEV_VERSION=12.5.8.93-1",
            "NV_LIBNPP_DEV_VERSION=12.3.3.100-1",
            "NV_LIBNPP_DEV_PACKAGE=libnpp-dev-12-8=12.3.3.100-1",
            "NV_LIBCUBLAS_DEV_VERSION=12.8.4.1-1",
            "NV_LIBCUBLAS_DEV_PACKAGE_NAME=libcublas-dev-12-8",
            "NV_LIBCUBLAS_DEV_PACKAGE=libcublas-dev-12-8=12.8.4.1-1",
            "NV_CUDA_NSIGHT_COMPUTE_VERSION=12.8.1-1",
            "NV_CUDA_NSIGHT_COMPUTE_DEV_PACKAGE=cuda-nsight-compute-12-8=12.8.1-1",
            "NV_NVPROF_VERSION=12.8.90-1",
            "NV_NVPROF_DEV_PACKAGE=cuda-nvprof-12-8=12.8.90-1",
            "NV_LIBNCCL_DEV_PACKAGE_NAME=libnccl-dev",
            "NV_LIBNCCL_DEV_PACKAGE_VERSION=2.25.1-1",
            "NV_LIBNCCL_DEV_PACKAGE=libnccl-dev=2.25.1-1+cuda12.8",
            "LIBRARY_PATH=/usr/local/cuda/lib64/stubs",
            "NV_CUDNN_VERSION=9.8.0.87-1",
            "NV_CUDNN_PACKAGE_NAME=libcudnn9-cuda-12",
            "NV_CUDNN_PACKAGE=libcudnn9-cuda-12=9.8.0.87-1",
            "NV_CUDNN_PACKAGE_DEV=libcudnn9-dev-cuda-12=9.8.0.87-1",
            "CUDA_HOME=/usr/local/cuda",
            "DEBIAN_FRONTEND=noninteractive",
            "PIP_DISABLE_PIP_VERSION_CHECK=1",
            "PIP_NO_CACHE_DIR=1",
            "PYTHONUNBUFFERED=1",
            "SGLANG_ENABLE_HEALTH_ENDPOINT_GENERATION=false",
            "MODEL_PATH=/model",
            "PORT=30000"
        ],
        "Cmd": null,
        "Healthcheck": {
            "Test": [
                "CMD-SHELL",
                "curl --fail --silent --show-error \"http://127.0.0.1:$PORT/health\" \u003e/dev/null || exit 1"
            ],
            "Interval": 30000000000,
            "Timeout": 10000000000,
            "StartPeriod": 900000000000,
            "Retries": 20
        },
        "ArgsEscaped": true,
        "Image": "",
        "Volumes": null,
        "WorkingDir": "/workspace",
        "Entrypoint": [
            "/usr/local/bin/entrypoint-dsv4"
        ],
        "OnBuild": null,
        "Labels": {
            "com.nvidia.cudnn.version": "9.8.0.87-1",
            "io.ktransformers.cuda": "12.8.1",
            "io.ktransformers.flashinfer": "0.6.15.post1",
            "io.ktransformers.flashinfer-cubin-sha256": "25cfd305afa1f34baa2f419bca35db58c369c534ee1961662e83d9fe858ce021",
            "io.ktransformers.layerwise-prefill-default": "enabled-threshold-2048-swa-0.4",
            "io.ktransformers.sglang-revision": "bc7f0058fafe6c05dc6ad52d6b1c854af2a673b2",
            "io.ktransformers.sglang-source": "https://github.com/kvcache-ai/sglang",
            "io.ktransformers.sm": "80,86,89,90,100,120",
            "io.ktransformers.upstream-ktransformers-revision": "f5ee50d4cfa29f5f3e8f2af5a2fd90771c574c58",
            "maintainer": "NVIDIA CORPORATION \u003ccudatools@nvidia.com\u003e",
            "org.opencontainers.image.created": "2026-08-08T09:17:45Z",
            "org.opencontainers.image.ref.name": "ubuntu",
            "org.opencontainers.image.revision": "90ddf8062d7e3593ffeec5612b47b36a34c7c77d",
            "org.opencontainers.image.source": "https://github.com/yyj6666667/ktransformers",
            "org.opencontainers.image.title": "KTransformers DeepSeek-V4-Flash",
            "org.opencontainers.image.version": "dsv4-f5ee50d-bc7f005-fi0615"
        },
        "StopSignal": "SIGTERM",
        "Shell": [
            "/bin/bash",
            "-o",
            "pipefail",
            "-c"
        ]
    },
    "Architecture": "amd64",
    "Os": "linux",
    "Size": 24684380426,
    "GraphDriver": {
        "Data": {
            "LowerDir": "/var/lib/docker/overlay2/1b6fabee8300b168d4250c7a1a63bf8cf2e7c7a9e1311b7af18ab3eadb002850/diff:/var/lib/docker/overlay2/77bc4c4989ad14f7c51dc1faa1a42c1b0104d813123c91d325e007c63b53357d/diff:/var/lib/docker/overlay2/1379ccf7d5ce73ea9909c7e68ff6ff69368a39f63e5e90dd6393ba6d4be8091e/diff:/var/lib/docker/overlay2/9353a4efc4fe6c52e5ba46f3341c1114d1e9ce26f3a982e418de75873a9705bd/diff:/var/lib/docker/overlay2/ce331b7ac060bba5403aba4c1101fb6f9fb72cef482ee37f4cd04f196391e929/diff:/var/lib/docker/overlay2/a88dc9df6c0b5fd9e739e1ab88f8b91a650e85de041cf8b9db002d696518dcf0/diff:/var/lib/docker/overlay2/2e6e69ed010a87406a4b7e17104cea3eba6a1aa3811296b443fcb3611d1f3532/diff:/var/lib/docker/overlay2/0635b887f0b1116dabb18ef422aebfe2f7f1290b2703e49236f188843693f037/diff:/var/lib/docker/overlay2/a7cc815685c98a510a44010a26658bc597a8e198edea61c461ff00353da59d2f/diff:/var/lib/docker/overlay2/a32d781bcaf5ae7611fc256965722a96d15b80d8ff2c70dfd965f4fc9279fffd/diff:/var/lib/docker/overlay2/7800b6de4750e70808e3e241c1107b112fe78c7ba9b72937560db83373e126b1/diff:/var/lib/docker/overlay2/e8773493643735743c6379887678034a826c07a0347a8bc891ad4e6ad1de2201/diff:/var/lib/docker/overlay2/72fd457a8fcfbb4ad104e1d131e077d89718743d7d368d9a64ae20f4c29b828f/diff:/var/lib/docker/overlay2/4c5eec5925b690e0ab5239aa50962c5c1637f4b5c08833bcf086b2d027e7a4d8/diff:/var/lib/docker/overlay2/f06a324453f0681669f2bae76f22033203934f35254752c33278e937e20addb9/diff:/var/lib/docker/overlay2/a6202db5510ec1a3916cd63bec404f041f645a6cbef5ebc7e9b484b9060fa8ae/diff:/var/lib/docker/overlay2/f5e8148fbcded40392bf280f4e11d48e712ee67c9fb52ab6a9e3ceaba685614f/diff:/var/lib/docker/overlay2/fb003b4c709b8ab2c751a50eb0fa6682b00670051f5ee283fd136c6619e9c78a/diff:/var/lib/docker/overlay2/5a243b65feeb45c9512364548d4a1c0e6c14d585b03cec76cc112a941e68b591/diff:/var/lib/docker/overlay2/e44c9b6650d154447167d6cd98a2eee7dd6765e46ae0dbaecfb913b3cc9ed7d6/diff:/var/lib/docker/overlay2/fc08216fbd3fb36c0f1f45239a0401436e6f159f92fad84601b8b4002e4784ff/diff:/var/lib/docker/overlay2/24a1de9a8d61f15729727cb00489be8c7162178bf5879c6b34239b1611cc5072/diff:/var/lib/docker/overlay2/8260ca3673e71b1f5bee5a6392897f88479a284b55ed014b7f74b9bf318d5c0f/diff:/var/lib/docker/overlay2/5250f12b22a507a7514f2ccf216bbdeff2a99b4db05b242e78cc032b4bab9a7b/diff",
            "MergedDir": "/var/lib/docker/overlay2/4cc87d70a1166518ba504451ce94cf57780a68050d80725eeaf861568f79f5af/merged",
            "UpperDir": "/var/lib/docker/overlay2/4cc87d70a1166518ba504451ce94cf57780a68050d80725eeaf861568f79f5af/diff",
            "WorkDir": "/var/lib/docker/overlay2/4cc87d70a1166518ba504451ce94cf57780a68050d80725eeaf861568f79f5af/work"
        },
        "Name": "overlay2"
    },
    "RootFS": {
        "Type": "layers",
        "Layers": [
            "sha256:270a1170e7e398434ff1b31e17e233f7d7b71aa99a40473615860068e86720af",
            "sha256:fb456a9e7760e7e4481d3d9eddee3ef62358753326d669c3efe6b0cc254ab9c3",
            "sha256:981a144d9fa2efffdbdfbdcfd3881dccb4fbf07867d4fa326a386126ca11b50f",
            "sha256:583f1d6040fc90a1380cb4d11d46b78bbbd4ad7d6274b9656ab9832ef7ba6342",
            "sha256:6020d84069cf575521faca653d2817e2195b8233b8c4e2d7bc65f3296296c2bb",
            "sha256:eb0143fcce68bd06a29bf314125d30f2a1f44d606fc419e7a54ab602bc26b2d0",
            "sha256:cb395f276e984728ca05fdee4e1dc203e1310a9e943a2d16d7aa87e5b2cf754b",
            "sha256:70a158f70e3777a129c15ffe51399bb6c226218c08ff6819948b9b1f26a46277",
            "sha256:9bb0510c7b4b74b056648f24e0a55c26f7521a1a1b766250a0776e14ce9d93f6",
            "sha256:7215ae9b9700d9530f70dea6f546b7ebdaac0098e4436b4c99e3e8063170c274",
            "sha256:acbc0d2ed199232b8f482ce41318ccd603e3c9a7731d39f0b08b88b253a6ca67",
            "sha256:f73947d6312e5c0bb71714a40dd2578d2ddd0b8bd2f6fb4075ab7252ef546a72",
            "sha256:c3ef3ecfb1637ceb491ba1f8da2a600de81b7b442faad0f5ebe78a5836ca3fe4",
            "sha256:8e666dadf0326f77b49691fa6aaac5d0c0fce53ead7bdff7429d7b707baf7480",
            "sha256:e2e5f469d8acf5ec63f0b732b1f4f5395ffb36a3595aedc3ff5224e83dadfcaf",
            "sha256:686e260d7f4e3bb64f6a18a9b7c7f30ba4695c20274b8912cbfa6e96a8d89ddd",
            "sha256:9abb6a2bd4407fbd2f56698074b0d3ea851f7b1869387111bdcf12244f18d9c7",
            "sha256:019df0c2bda08f0a9aa513adbcf7bd49ef48706321fd463c6743c574ec945900",
            "sha256:d02acd9e11d1bdcb0684874cc0bf8b793c94a899f3aef984dfdf30fabf5c9430",
            "sha256:a899f82843516060755b5aec04a90bdb36d9aca5b778bf9f3508284a4c82cde3",
            "sha256:2299ba56fd0e121c346ed3c3126350ba92ecf2321e861f7d470adf87006af2d7",
            "sha256:0cbb5ffabb929f254656ff38685268d436077e8c119557902cb68791b10a45d1",
            "sha256:c3d0e3686767e67e9e3b2916c96eccceaf5b201eb934eff159f52752d079a1c4",
            "sha256:109f04f9104aa7880bde3f045d1cc315ce922e2e05d975b3cb59d86adc384cb5",
            "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef"
        ]
    },
    "Metadata": {
        "LastTagTime": "2026-08-13T05:56:15.746153192+08:00"
    }
}

更多版本

docker.io/approachingai/ktransformers:0.2.1

linux/amd64 docker.io18.50GB2025-02-19 00:37