| ARG RELEASE |
| ARG LAUNCHPAD_BUILD_ARCH |
| LABEL org.opencontainers.image.ref.name=ubuntu |
| LABEL org.opencontainers.image.version=24.04 |
| ADD file:bcebbf0fddcba5b864d5d267b68dd23bcfb01275e6ec7bcab69bf8b56af14804 in / |
| CMD ["/bin/bash"] |
| RUN /bin/sh -c export DEBIAN_FRONTEND=noninteractive && apt-get update && apt-get install -y --no-install-recommends apt-utils build-essential ca-certificates curl libncurses6 libncursesw6 patch wget rsync unzip jq gnupg libtcmalloc-minimal4 && rm -rf /var/lib/apt/lists/* && echo "hsts=0" > /root/.wgetrc # buildkit |
| ARG CUDA_VERSION=12.8.0.038 |
| ARG CUDA_DRIVER_VERSION=570.86.10 |
| ARG JETPACK_HOST_MOUNTS= |
| ENV CUDA_VERSION=12.8.0.038 CUDA_DRIVER_VERSION=570.86.10 CUDA_CACHE_DISABLE=1 NVIDIA_REQUIRE_JETPACK_HOST_MOUNTS= |
| RUN |3 CUDA_VERSION=12.8.0.038 CUDA_DRIVER_VERSION=570.86.10 JETPACK_HOST_MOUNTS= /bin/sh -c if [ -n "${JETPACK_HOST_MOUNTS}" ]; then echo "/usr/lib/aarch64-linux-gnu/tegra" > /etc/ld.so.conf.d/nvidia-tegra.conf && echo "/usr/lib/aarch64-linux-gnu/tegra-egl" >> /etc/ld.so.conf.d/nvidia-tegra.conf; fi # buildkit |
| RUN |3 CUDA_VERSION=12.8.0.038 CUDA_DRIVER_VERSION=570.86.10 JETPACK_HOST_MOUNTS= /bin/sh -c /nvidia/build-scripts/installCUDA.sh # buildkit |
| RUN |3 CUDA_VERSION=12.8.0.038 CUDA_DRIVER_VERSION=570.86.10 JETPACK_HOST_MOUNTS= /bin/sh -c cp -vprd /nvidia/. / && patch -p0 < /etc/startup_scripts.patch && rm -f /etc/startup_scripts.patch # buildkit |
| ENV _CUDA_COMPAT_PATH=/usr/local/cuda/compat ENV=/etc/shinit_v2 BASH_ENV=/etc/bash.bashrc SHELL=/bin/bash NVIDIA_REQUIRE_CUDA=cuda>=9.0 |
| LABEL com.nvidia.volumes.needed=nvidia_driver com.nvidia.cuda.version=9.0 |
| ARG NCCL_VERSION=2.25.1 |
| ARG CUBLAS_VERSION=12.8.3.14 |
| ARG CUFFT_VERSION=11.3.3.41 |
| ARG CURAND_VERSION=10.3.9.55 |
| ARG CUSPARSE_VERSION=12.5.7.53 |
| ARG CUSOLVER_VERSION=11.7.2.55 |
| ARG CUTENSOR_VERSION=2.1.0.9 |
| ARG NPP_VERSION=12.3.3.65 |
| ARG NVJPEG_VERSION=12.3.5.57 |
| ARG CUFILE_VERSION=1.13.0.11 |
| ARG NVJITLINK_VERSION=12.8.61 |
| ARG CUDNN_VERSION=9.7.0.66 |
| ARG CUDNN_FRONTEND_VERSION=1.9.0 |
| ARG TRT_VERSION=10.8.0.43 |
| ARG TRTOSS_VERSION= |
| ARG NSIGHT_SYSTEMS_VERSION=2024.6.2.225 |
| ARG NSIGHT_COMPUTE_VERSION=2025.1.0.14 |
| ARG CUSPARSELT_VERSION=0.6.3.2 |
| ENV NCCL_VERSION=2.25.1 CUBLAS_VERSION=12.8.3.14 CUFFT_VERSION=11.3.3.41 CURAND_VERSION=10.3.9.55 CUSPARSE_VERSION=12.5.7.53 CUSPARSELT_VERSION=0.6.3.2 CUSOLVER_VERSION=11.7.2.55 CUTENSOR_VERSION=2.1.0.9 NPP_VERSION=12.3.3.65 NVJPEG_VERSION=12.3.5.57 CUFILE_VERSION=1.13.0.11 NVJITLINK_VERSION=12.8.61 CUDNN_VERSION=9.7.0.66 CUDNN_FRONTEND_VERSION=1.9.0 TRT_VERSION=10.8.0.43 TRTOSS_VERSION= NSIGHT_SYSTEMS_VERSION=2024.6.2.225 NSIGHT_COMPUTE_VERSION=2025.1.0.14 |
| RUN |21 CUDA_VERSION=12.8.0.038 CUDA_DRIVER_VERSION=570.86.10 JETPACK_HOST_MOUNTS= NCCL_VERSION=2.25.1 CUBLAS_VERSION=12.8.3.14 CUFFT_VERSION=11.3.3.41 CURAND_VERSION=10.3.9.55 CUSPARSE_VERSION=12.5.7.53 CUSOLVER_VERSION=11.7.2.55 CUTENSOR_VERSION=2.1.0.9 NPP_VERSION=12.3.3.65 NVJPEG_VERSION=12.3.5.57 CUFILE_VERSION=1.13.0.11 NVJITLINK_VERSION=12.8.61 CUDNN_VERSION=9.7.0.66 CUDNN_FRONTEND_VERSION=1.9.0 TRT_VERSION=10.8.0.43 TRTOSS_VERSION= NSIGHT_SYSTEMS_VERSION=2024.6.2.225 NSIGHT_COMPUTE_VERSION=2025.1.0.14 CUSPARSELT_VERSION=0.6.3.2 /bin/sh -c /nvidia/build-scripts/installLIBS.sh && /nvidia/build-scripts/installCUDNN.sh && /nvidia/build-scripts/installTRT.sh && /nvidia/build-scripts/installNSYS.sh && /nvidia/build-scripts/installNCU.sh && /nvidia/build-scripts/installCUTENSOR.sh && /nvidia/build-scripts/installCUSPARSELT.sh && if [ -z "${JETPACK_HOST_MOUNTS}" ]; then /nvidia/build-scripts/installNCCL.sh; fi; # buildkit |
| LABEL com.nvidia.nccl.version=2.25.1 com.nvidia.cublas.version=12.8.3.14 com.nvidia.cufft.version=11.3.3.41 com.nvidia.curand.version=10.3.9.55 com.nvidia.cusparse.version=12.5.7.53 com.nvidia.cusparselt.version=0.6.3.2 com.nvidia.cusolver.version=11.7.2.55 com.nvidia.cutensor.version=2.1.0.9 com.nvidia.npp.version=12.3.3.65 com.nvidia.nvjpeg.version=12.3.5.57 com.nvidia.cudnn.version=9.7.0.66 com.nvidia.tensorrt.version=10.8.0.43 com.nvidia.tensorrtoss.version= com.nvidia.nsightsystems.version=2024.6.2.225 com.nvidia.nsightcompute.version=2025.1.0.14 |
| ARG DALI_VERSION=1.45.0 |
| ARG DALI_BUILD= |
| ARG DALI_URL_SUFFIX=120 |
| ARG POLYGRAPHY_VERSION=0.49.16 |
| ARG TRANSFORMER_ENGINE_VERSION=1.14 |
| ARG MODEL_OPT_VERSION=0.21.0 |
| ENV DALI_VERSION=1.45.0 DALI_BUILD= DALI_URL_SUFFIX=120 POLYGRAPHY_VERSION=0.49.16 TRANSFORMER_ENGINE_VERSION=1.14 MODEL_OPT_VERSION=0.21.0 |
| ADD docs.tgz / # buildkit |
| RUN |27 CUDA_VERSION=12.8.0.038 CUDA_DRIVER_VERSION=570.86.10 JETPACK_HOST_MOUNTS= NCCL_VERSION=2.25.1 CUBLAS_VERSION=12.8.3.14 CUFFT_VERSION=11.3.3.41 CURAND_VERSION=10.3.9.55 CUSPARSE_VERSION=12.5.7.53 CUSOLVER_VERSION=11.7.2.55 CUTENSOR_VERSION=2.1.0.9 NPP_VERSION=12.3.3.65 NVJPEG_VERSION=12.3.5.57 CUFILE_VERSION=1.13.0.11 NVJITLINK_VERSION=12.8.61 CUDNN_VERSION=9.7.0.66 CUDNN_FRONTEND_VERSION=1.9.0 TRT_VERSION=10.8.0.43 TRTOSS_VERSION= NSIGHT_SYSTEMS_VERSION=2024.6.2.225 NSIGHT_COMPUTE_VERSION=2025.1.0.14 CUSPARSELT_VERSION=0.6.3.2 DALI_VERSION=1.45.0 DALI_BUILD= DALI_URL_SUFFIX=120 POLYGRAPHY_VERSION=0.49.16 TRANSFORMER_ENGINE_VERSION=1.14 MODEL_OPT_VERSION=0.21.0 /bin/sh -c echo "/usr/local/nvidia/lib" >> /etc/ld.so.conf.d/nvidia.conf && echo "/usr/local/nvidia/lib64" >> /etc/ld.so.conf.d/nvidia.conf # buildkit |
| ARG _LIBPATH_SUFFIX= |
| ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin LD_LIBRARY_PATH=/usr/local/cuda/compat/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64 NVIDIA_VISIBLE_DEVICES=all NVIDIA_DRIVER_CAPABILITIES=compute,utility,video |
| COPY entrypoint/ /opt/nvidia/ # buildkit |
| ENV NVIDIA_PRODUCT_NAME=CUDA |
| ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"] |
| COPY NVIDIA_Deep_Learning_Container_License.pdf /workspace/ # buildkit |
| RUN /bin/sh -c export DEBIAN_FRONTEND=noninteractive && apt-get update && apt-get install -y --no-install-recommends build-essential git libglib2.0-0 less libhwloc15 libnl-route-3-200 libnl-3-dev libnl-route-3-dev libnuma-dev libnuma1 libpmi2-0-dev nano numactl openssh-client vim wget && rm -rf /var/lib/apt/lists/* # buildkit |
| ARG GDRCOPY_VERSION=2.4.1 |
| ARG HPCX_VERSION=2.21 |
| ARG RDMACORE_VERSION=39.0 |
| ARG MOFED_VERSION=5.4-rdmacore39.0 |
| ARG OPENUCX_VERSION=1.18.0 |
| ARG OPENMPI_VERSION=4.1.7 |
| ARG EFA_VERSION=1.34.0 |
| ARG AWS_OFI_NCCL_VERSION=1.12.1 |
| ENV GDRCOPY_VERSION=2.4.1 HPCX_VERSION=2.21 MOFED_VERSION=5.4-rdmacore39.0 OPENUCX_VERSION=1.18.0 OPENMPI_VERSION=4.1.7 RDMACORE_VERSION=39.0 EFA_VERSION=1.34.0 AWS_OFI_NCCL_VERSION=1.12.1 |
| ARG TARGETARCH=amd64 |
| RUN |9 GDRCOPY_VERSION=2.4.1 HPCX_VERSION=2.21 RDMACORE_VERSION=39.0 MOFED_VERSION=5.4-rdmacore39.0 OPENUCX_VERSION=1.18.0 OPENMPI_VERSION=4.1.7 EFA_VERSION=1.34.0 AWS_OFI_NCCL_VERSION=1.12.1 TARGETARCH=amd64 /bin/sh -c cd /nvidia && ( export DEBIAN_FRONTEND=noninteractive && apt-get update && apt-get install -y --no-install-recommends libibverbs1 libibverbs-dev librdmacm1 librdmacm-dev libibumad3 libibumad-dev ibverbs-utils ibverbs-providers && rm -rf /var/lib/apt/lists/* && rm $(dpkg-query -L libibverbs-dev librdmacm-dev libibumad-dev | grep "\(\.so\|\.a\)$") ) && ( cd opt/gdrcopy/ && dpkg -i libgdrapi_*.deb ) && ( cp -r opt/hpcx /opt/ && cp etc/ld.so.conf.d/hpcx.conf /etc/ld.so.conf.d/ && ln -sf /opt/hpcx/ompi /usr/local/mpi && ln -sf /opt/hpcx/ucx /usr/local/ucx && sed -i 's/^\(hwloc_base_binding_policy\) = core$/\1 = none/' /opt/hpcx/ompi/etc/openmpi-mca-params.conf && sed -i 's/^\(btl = self\)$/#\1/' /opt/hpcx/ompi/etc/openmpi-mca-params.conf ) && ( if [ ! -f /etc/ld.so.conf.d/nvidia-tegra.conf ]; then cd opt/amazon/efa/ && dpkg -i libfabric*.deb && rm /opt/amazon/efa/lib/libfabric.a && echo "/opt/amazon/efa/lib" > /etc/ld.so.conf.d/efa.conf; fi ) && ldconfig # buildkit |
| ENV OPAL_PREFIX=/opt/hpcx/ompi PATH=/usr/local/mpi/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/ucx/bin:/opt/amazon/efa/bin |
| ENV OMPI_MCA_coll_hcoll_enable=0 |
| ENV NCCL_NVLS_ENABLE=0 |
| COPY cuda-*.patch /tmp # buildkit |
| RUN |9 GDRCOPY_VERSION=2.4.1 HPCX_VERSION=2.21 RDMACORE_VERSION=39.0 MOFED_VERSION=5.4-rdmacore39.0 OPENUCX_VERSION=1.18.0 OPENMPI_VERSION=4.1.7 EFA_VERSION=1.34.0 AWS_OFI_NCCL_VERSION=1.12.1 TARGETARCH=amd64 /bin/sh -c export DEVEL=1 BASE=0 && /nvidia/build-scripts/installNCU.sh && /nvidia/build-scripts/installCUDA.sh && /nvidia/build-scripts/installLIBS.sh && if [ ! -f /etc/ld.so.conf.d/nvidia-tegra.conf ]; then /nvidia/build-scripts/installNCCL.sh; fi && /nvidia/build-scripts/installCUDNN.sh && /nvidia/build-scripts/installCUTENSOR.sh && /nvidia/build-scripts/installTRT.sh && /nvidia/build-scripts/installNSYS.sh && /nvidia/build-scripts/installCUSPARSELT.sh && if [ -f "/tmp/cuda-${_CUDA_VERSION_MAJMIN}.patch" ]; then patch -p0 < /tmp/cuda-${_CUDA_VERSION_MAJMIN}.patch; fi && rm -f /tmp/cuda-*.patch # buildkit |
| ENV LIBRARY_PATH=/usr/local/cuda/lib64/stubs: |
| COPY /opt/amazon/aws-ofi-nccl /opt/amazon/aws-ofi-nccl # buildkit |
| RUN |9 GDRCOPY_VERSION=2.4.1 HPCX_VERSION=2.21 RDMACORE_VERSION=39.0 MOFED_VERSION=5.4-rdmacore39.0 OPENUCX_VERSION=1.18.0 OPENMPI_VERSION=4.1.7 EFA_VERSION=1.34.0 AWS_OFI_NCCL_VERSION=1.12.1 TARGETARCH=amd64 /bin/sh -c if [ ! -f /etc/ld.so.conf.d/nvidia-tegra.conf ]; then echo "/opt/amazon/aws-ofi-nccl/lib" > /etc/ld.so.conf.d/aws-ofi-nccl.conf && ldconfig; fi # buildkit |
| ARG NVIDIA_TRITON_SERVER_BASE_VERSION |
| ARG NVIDIA_BUILD_ID |
| ENV NVIDIA_TRITON_SERVER_BASE_VERSION=25.01 |
| ENV NVIDIA_BUILD_ID=136230209 |
| ENV NVIDIA_PRODUCT_NAME=Triton Server Base |
| COPY docker/entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit |
| RUN |2 NVIDIA_TRITON_SERVER_BASE_VERSION=25.01 NVIDIA_BUILD_ID=136230209 /bin/sh -c sed 's/Server/Server Base/' /opt/nvidia/entrypoint.d/10-banner.txt | sed 's/^===/========/' > /opt/nvidia/entrypoint.d/10-banner.new && mv /opt/nvidia/entrypoint.d/10-banner.new /opt/nvidia/entrypoint.d/10-banner.txt # buildkit |
| ARG DEBIAN_FRONTEND=noninteractive |
| ARG TARGET_ARCH=x86_64 # valid values: x86_64, aarch64 |
| ARG TARGET_OS=linux # valid values: linux, l4t, qnx |
| ARG BUILD_MODE=release # valid values: release, debug |
| ARG TRITON_RELEASE_VERSION=2.54.0 |
| ARG TRITON_RELEASE_VERSION_L4T=2.50.0 |
| ARG DCGM_VERSION=3.2.6 |
| ARG TARGET_CUDA_VERSION=12.8 |
| WORKDIR /opt/tritonserver |
| RUN |24 DEBIAN_FRONTEND=noninteractive TARGET_ARCH=x86_64 #= valid= values:= x86_64,= aarch64= TARGET_OS=linux #= valid= values:= linux,= l4t,= qnx= BUILD_MODE=release #= valid= values:= release,= debug= TRITON_RELEASE_VERSION=2.54.0 TRITON_RELEASE_VERSION_L4T=2.50.0 DCGM_VERSION=3.2.6 TARGET_CUDA_VERSION=12.8 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends software-properties-common python3-dev python3-pip wget && rm -rf /var/lib/apt/lists/* # buildkit |
| RUN |24 DEBIAN_FRONTEND=noninteractive TARGET_ARCH=x86_64 #= valid= values:= x86_64,= aarch64= TARGET_OS=linux #= valid= values:= linux,= l4t,= qnx= BUILD_MODE=release #= valid= values:= release,= debug= TRITON_RELEASE_VERSION=2.54.0 TRITON_RELEASE_VERSION_L4T=2.50.0 DCGM_VERSION=3.2.6 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ "$TARGET_OS" != "l4t" ] && [ "$TARGET_OS" != "qnx" ]; then cp -R /opt2/tritonserver/bin . && cp -R /opt2/tritonserver/lib . && cp -R /opt2/tritonserver/include . && cp /opt2/tritonserver/TRITON_VERSION . && mkdir /opt/tritonserver/backends && cp -R /opt2/tritonserver/backends/tensorrt backends/tensorrt && cp -R /opt2/tritonserver/backends/onnxruntime backends/onnxruntime && cp -R /opt2/tritonserver/backends/python backends/python && cp -R /opt2/tritonserver/backends/pytorch backends/pytorch; else cd /opt && wget -nv https://github.com/triton-inference-server/server/releases/download/v${TRITON_RELEASE_VERSION_L4T}/tritonserver-${TRITON_RELEASE_VERSION_L4T}-igpu.tar.gz && tar -zxf tritonserver-${TRITON_RELEASE_VERSION_L4T}-igpu.tar.gz tritonserver/bin tritonserver/lib tritonserver/include tritonserver/TRITON_VERSION tritonserver/backends/onnxruntime tritonserver/backends/tensorrt tritonserver/backends/python tritonserver/backends/pytorch && rm tritonserver-${TRITON_RELEASE_VERSION_L4T}-igpu.tar.gz; fi # buildkit |
| RUN |24 DEBIAN_FRONTEND=noninteractive TARGET_ARCH=x86_64 #= valid= values:= x86_64,= aarch64= TARGET_OS=linux #= valid= values:= linux,= l4t,= qnx= BUILD_MODE=release #= valid= values:= release,= debug= TRITON_RELEASE_VERSION=2.54.0 TRITON_RELEASE_VERSION_L4T=2.50.0 DCGM_VERSION=3.2.6 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ "$TARGET_OS" != "l4t" ] && [ "$TARGET_OS" != "qnx" ]; then cp -R /pytorch_lib/python3.12/dist-packages/torch/include /opt/tritonserver/backends/pytorch/include; else cp -R /pytorch_lib/python3.10/dist-packages/torch/include /opt/tritonserver/backends/pytorch/include; fi # buildkit |
| SHELL [/bin/bash -c] |
| RUN |24 DEBIAN_FRONTEND=noninteractive TARGET_ARCH=x86_64 #= valid= values:= x86_64,= aarch64= TARGET_OS=linux #= valid= values:= linux,= l4t,= qnx= BUILD_MODE=release #= valid= values:= release,= debug= TRITON_RELEASE_VERSION=2.54.0 TRITON_RELEASE_VERSION_L4T=2.50.0 DCGM_VERSION=3.2.6 TARGET_CUDA_VERSION=12.8 /bin/bash -c if [ "$TARGET_OS" != "l4t" ] && [ "$TARGET_OS" != "qnx" ]; then wget -nv https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/${TARGET_ARCH//aarch64/sbsa}/cuda-ubuntu2204.pin && mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600 && wget -nv https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/${TARGET_ARCH//aarch64/sbsa}/cuda-keyring_1.0-1_all.deb && dpkg -i cuda-keyring_1.0-1_all.deb && add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/${TARGET_ARCH//aarch64/sbsa}/ /" && apt-get update && apt-get install -y datacenter-gpu-manager=1:${DCGM_VERSION} && rm -f cuda-keyring_1.0-1_all.deb; fi && apt-get update && apt-get install -y --no-install-recommends libboost-program-options1.74.0 libboost-system1.74.0 libboost-thread1.74.0 libboost-locale1.74.0 && wget -nv -O /tmp/boost.tar.gz https://archives.boost.io/release/1.79.0/source/boost_1_79_0.tar.gz && tar xzf /tmp/boost.tar.gz -C /tmp/ && cd /tmp/boost_1_79_0 && ./bootstrap.sh --prefix=/usr && ./b2 install && rm -rf /tmp/boost* && wget -nv -O /tmp/boost.tar.gz https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz && tar xzf /tmp/boost.tar.gz -C /tmp/ && cd /tmp/boost_1_80_0 && ./bootstrap.sh --prefix=/usr && ./b2 install && mv /tmp/boost_1_80_0/boost /usr/include/boost && rm -rf /tmp/boost* && cd /tmp && wget -nv https://urm.nvidia.com/artifactory/sw-ai-app-bazel/sw-ai-app-jarvis/tritonserver-be/onnxruntime-24.06/onnxruntime-backend-${TARGET_ARCH//x86_64/amd64}.tar && rm -rf /opt/tritonserver/backends/onnxruntime && tar -xvf /tmp/onnxruntime-backend-${TARGET_ARCH//x86_64/amd64}.tar -C /opt/tritonserver/backends && rm -rf /tmp/onnxruntime* && apt-get install -y libcudnn8-dev # buildkit |
| SHELL [/bin/sh -c] |
| RUN |24 DEBIAN_FRONTEND=noninteractive TARGET_ARCH=x86_64 #= valid= values:= x86_64,= aarch64= TARGET_OS=linux #= valid= values:= linux,= l4t,= qnx= BUILD_MODE=release #= valid= values:= release,= debug= TRITON_RELEASE_VERSION=2.54.0 TRITON_RELEASE_VERSION_L4T=2.50.0 DCGM_VERSION=3.2.6 TARGET_CUDA_VERSION=12.8 /bin/sh -c python3 -m pip config set global.break-system-packages true # buildkit |
| RUN |24 DEBIAN_FRONTEND=noninteractive TARGET_ARCH=x86_64 #= valid= values:= x86_64,= aarch64= TARGET_OS=linux #= valid= values:= linux,= l4t,= qnx= BUILD_MODE=release #= valid= values:= release,= debug= TRITON_RELEASE_VERSION=2.54.0 TRITON_RELEASE_VERSION_L4T=2.50.0 DCGM_VERSION=3.2.6 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ "$TARGET_ARCH" = "x86_64" ]; then wget -nv https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.8.0/tars/TensorRT-10.8.0.43.Linux.x86_64-gnu.cuda-12.8.tar.gz && tar xf TensorRT-10.8.0.43.Linux.x86_64-gnu.cuda-12.8.tar.gz && pip3 install --no-cache-dir --force-reinstall TensorRT-10.8.0.43/python/tensorrt-10.8.0.43-cp312-none-linux_x86_64.whl && rm -rf TensorRT*; elif [ "$TARGET_ARCH" = "aarch64" ] && [ "$TARGET_OS" = "linux" ]; then wget -nv https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.8.0/tars/TensorRT-10.8.0.43.Linux.aarch64-gnu.cuda-12.8.tar.gz && tar xf TensorRT-10.8.0.43.Linux.aarch64-gnu.cuda-12.8.tar.gz && pip3 install --no-cache-dir --force-reinstall TensorRT-10.8.0.43/python/tensorrt-10.8.0.43-cp312-none-linux_aarch64.whl && rm -rf TensorRT*; elif [ "$TARGET_ARCH" = "aarch64" ] && [ "$TARGET_OS" = "l4t" ]; then wget -nv https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.6.0/tars/TensorRT-10.6.0.26.l4t.aarch64-gnu.cuda-12.6.tar.gz && tar xf TensorRT-10.6.0.26.l4t.aarch64-gnu.cuda-12.6.tar.gz && pip3 install --no-cache-dir --force-reinstall TensorRT-10.6.0.26/python/tensorrt-10.6.0-cp310-none-linux_aarch64.whl && wget -nv https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/arm64/cuda-keyring_1.1-1_all.deb && dpkg -i cuda-keyring_1.1-1_all.deb && apt-get update && apt-get install -y libcudnn9-dev libnvinfer-dev && rm -rf TensorRT* cuda-keyring*; elif [ "$TARGET_ARCH" = "aarch64" ] && [ "$TARGET_OS" = "qnx" ]; then apt-get purge -y libcudnn8* libnvinfer* && wget -nv http://cuda-repo/release-candidates/kitpicks/cudnn-v8-4-cuda-11-4-orin/8.4.1.88/001/redist/cudnn/cudnn/linux-aarch64/cudnn-linux-aarch64-8.4.1.88_cuda11.4-archive.tar.xz && tar xvf cudnn-linux-aarch64-8.4.1.88_cuda11.4-archive.tar.xz && mv cudnn-linux-aarch64-8.4.1.88_cuda11.4-archive cudnn && rm -rf /lib/aarch64-linux-gnu/libcudnn* /usr/include/aarch64-linux-gnu/cudnn* && cp -r cudnn/lib/* /lib/aarch64-linux-gnu/ && cp -r cudnn/include/* /usr/include/aarch64-linux-gnu/ && rm -rf cudnn cudnn-linux-aarch64-8.4.1.88_cuda11.4-archive.tar.xz && wget -nv http://cuda-repo.nvidia.com/release-candidates/Libraries/TensorRT/v8.4/8.4.11.6-3306d99a/11.4-r470/d6l-aarch64/tar/TensorRT-8.4.11.6.Ubuntu-20.04.aarch64-gnu.cuda-11.4.cudnn8.4.tar.gz && tar xvf TensorRT-8.4.11.6.Ubuntu-20.04.aarch64-gnu.cuda-11.4.cudnn8.4.tar.gz && mv TensorRT-8.4.11.6 TensorRT && cp -r TensorRT/lib/* /lib/aarch64-linux-gnu/ && cp -r TensorRT/include/* /usr/include/aarch64-linux-gnu/ && rm -rf TensorRT TensorRT-8.4.11.6.Ubuntu-20.04.aarch64-gnu.cuda-11.4.cudnn8.4.tar.gz; fi # buildkit |
| RUN |24 DEBIAN_FRONTEND=noninteractive TARGET_ARCH=x86_64 #= valid= values:= x86_64,= aarch64= TARGET_OS=linux #= valid= values:= linux,= l4t,= qnx= BUILD_MODE=release #= valid= values:= release,= debug= TRITON_RELEASE_VERSION=2.54.0 TRITON_RELEASE_VERSION_L4T=2.50.0 DCGM_VERSION=3.2.6 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ "$TARGET_ARCH" = "x86_64" ]; then pip3 install --no-cache-dir https://urm.nvidia.com/artifactory/sw-ai-app-bazel-local/sw-ai-app-jarvis/tools/torch-2.6.0a0+ecf3bae40a.nv25.1-22181419-cp312-cp312-linux_x86_64.whl && pip3 install --no-cache-dir https://gitlab-master.nvidia.com/api/v4/projects/100660/packages/pypi/files/5fab58d043583f9b59df5c36bb797fd147cc488fff1c79fc11e974721845ea9e/torchvision-0.20.0a0+ecf3bae40a.nv25.1-22932737-cp312-cp312-manylinux_2_28_x86_64.whl#sha256=5fab58d043583f9b59df5c36bb797fd147cc488fff1c79fc11e974721845ea9e && pip3 install --no-cache-dir tensorrt_cu12==10.8.0.43 --pre --extra-index-url https://pypi.nvidia.com && pip3 install --no-cache-dir https://urm.nvidia.com/artifactory/sw-ai-app-bazel-local/sw-ai-app-jarvis/tools/tensorrt_llm-0.17.0-cp312-cp312-linux_x86_64.whl pynvml==11.5.3 && pip3 install --no-cache-dir https://download.pytorch.org/whl/nightly/pytorch_triton-3.1.0%2Bcf34004b8a-cp312-cp312-linux_x86_64.whl && pip3 install --no-cache-dir nemo_toolkit['asr']==2.0.0rc1 && pip3 install --no-cache-dir huggingface-hub==0.24.0 && pip3 install --no-cache-dir pycuda==2024.1.2 && pip3 install --no-cache-dir tiktoken==0.7.0 && pip3 install --no-cache-dir transformers==4.48.0; else pip3 install --no-cache-dir torch==2.4.0; fi # buildkit |
| ARG TARGET_OS |
| ARG TARGET_ARCH |
| ARG TARGET_CUDA_VERSION |
| ENV DEBIAN_FRONTEND=noninteractive |
| ENV PATH=/opt/riva/bin:/opt/tritonserver/bin:/usr/local/mpi/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/ucx/bin:/opt/amazon/efa/bin |
| ENV LD_LIBRARY_PATH=/opt/riva/lib:/usr/local/cuda/compat/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64 |
| ENV LD_LIBRARY_PATH=/opt/tritonserver/lib:/opt/riva/lib:/usr/local/cuda/compat/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64 |
| ENV LD_LIBRARY_PATH=/opt/tritonserver/backends/pytorch:/opt/tritonserver/lib:/opt/riva/lib:/usr/local/cuda/compat/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64 |
| ENV RIVA_SERVER_USER=riva-server |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c userdel $RIVA_SERVER_USER > /dev/null 2>&1 || true && if ! id -u $RIVA_SERVER_USER > /dev/null 2>&1 ; then useradd $RIVA_SERVER_USER; fi && if [ "$TARGET_OS" != "l4t" ]; then usermod -u 1002 ubuntu && groupmod -g 1002 ubuntu && usermod -u 1000 $RIVA_SERVER_USER && groupmod -g 1000 $RIVA_SERVER_USER; fi && [ `id -u $RIVA_SERVER_USER` -eq 1000 ] && [ `id -g $RIVA_SERVER_USER` -eq 1000 ] && mkdir -p /home/$RIVA_SERVER_USER/ && chmod 777 /home/$RIVA_SERVER_USER/ # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends libb64-0d bc libcurl4-openssl-dev libexpat1 linux-libc-dev libc6 libc-bin libcurl3-gnutls libnghttp2-14 flac wget libatlas-base-dev libgoogle-glog-dev libflac++-dev libarchive-dev libogg0 libnghttp2-14 libtool locales policykit-1 python3-pip && apt-get install -y --no-install-recommends openssl libssl3 libssl-dev portaudio19-dev python3-pyaudio alsa-utils && if [ "$TARGET_OS" = "l4t" ]; then apt-get install -y libopenblas-dev libopencc-dev libopenmpi3 git curl vim unzip && cp /usr/lib/aarch64-linux-gnu/lapack/liblapack.so.3.10.0 /usr/lib/aarch64-linux-gnu/liblapack.so.3 && cp /usr/lib/aarch64-linux-gnu/blas/libblas.so.3.10.0 /usr/lib/aarch64-linux-gnu/libblas.so.3; fi && rm -rf /var/lib/apt/lists/* # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c apt-get remove -y --purge libslurm*; apt-get remove -y --purge libpmi*; apt-get remove --purge -y openssh* || true # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ -f "/opt/hpcx/ompi/etc/openmpi-mca-params.conf" ]; then echo "plm_rsh_agent = false" >> /opt/hpcx/ompi/etc/openmpi-mca-params.conf; fi # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c touch /usr/share/locale/locale.alias && sed -i '/en_US.UTF-8/s/^# //g' /etc/locale.gen && locale-gen # buildkit |
| WORKDIR /opt/tritonserver |
| COPY /opt/riva/backends/model.py /opt/riva/backends/model.py # buildkit |
| COPY /opt/riva/backends/megatron_model.py /opt/riva/backends/megatron_model.py # buildkit |
| COPY /opt/riva/backends/whisper.py /opt/riva/backends/whisper.py # buildkit |
| COPY /opt/riva/backends/canary.py /opt/riva/backends/canary.py # buildkit |
| COPY /opt/riva/backends/silero.py /opt/riva/backends/silero.py # buildkit |
| COPY riva/cbe/python-asr/nemo_offline_asr_model.py /opt/riva/backends/nemo_offline_asr_model.py # buildkit |
| COPY /opt/riva/backends/nemo_asr_model.py /opt/riva/backends/nemo_asr_model.py # buildkit |
| COPY /opt/riva/backends/cache_aware_model.py /opt/riva/backends/cache_aware_model.py # buildkit |
| COPY /opt/riva/backends/clustering_model.py /opt/riva/backends/clustering_model.py # buildkit |
| COPY /opt/riva/backends/nmesc_clustering.py /opt/riva/backends/nmesc_clustering.py # buildkit |
| COPY /opt/riva/backends/magpie_tts_model.py /opt/riva/backends/magpie_tts_model.py # buildkit |
| COPY /opt/riva/backends/sortformer_model.py /opt/riva/backends/sortformer_model.py # buildkit |
| COPY /opt/riva/backends/sortformer_modules.py /opt/riva/backends/sortformer_modules.py # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c pip install --no-cache-dir cython==3.0.11 fastapi==0.115.3 "uvicorn[standard]==0.32.0" tritonclient==${TRITON_RELEASE_VERSION} packaging==24.1 pydantic_settings==2.5.2 python-multipart==0.0.12 gunicorn==23.0.0 # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ "$TARGET_ARCH" = "x86_64" ]; then pip install --no-cache-dir opencc==1.1.7; else pip install --no-cache-dir opencc==0.2; fi # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c pip install --no-cache-dir sacremoses==0.1.1 numpy==1.26.4 sentencepiece==0.2.0 youtokentome==1.0.6 jieba==0.42.1 pangu==4.0.6.1 # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ "$TARGET_ARCH" = "x86_64" ]; then pip install --no-cache-dir cupy-cuda12x; elif [ "$TARGET_ARCH" = "aarch64" ]; then pip install --no-cache-dir cupy-cuda12x -f https://pip.cupy.dev/aarch64; fi # buildkit |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/sh -c if [ "$TARGET_OS" != "l4t" ]; then ln -sf ${_CUDA_COMPAT_PATH}/lib.real ${_CUDA_COMPAT_PATH}/lib && echo ${_CUDA_COMPAT_PATH}/lib > /etc/ld.so.conf.d/00-cuda-compat.conf && ldconfig && rm -f ${_CUDA_COMPAT_PATH}/lib; fi # buildkit |
| SHELL [/bin/bash -c] |
| RUN |3 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 /bin/bash -c wget -nv -O /bin/grpc_health_probe https://urm.nvidia.com/artifactory/sw-ai-app-bazel/sw-ai-app-jarvis/tools/grpc-health-probe-bca6287/grpc_health_probe_$TARGET_ARCH && chmod +x /bin/grpc_health_probe; # buildkit |
| SHELL [/bin/sh -c] |
| WORKDIR /opt/riva |
| ARG KALDI_DIR=submodules/kaldi |
| COPY /workspace/submodules/kaldi/tools/openfst-1.7.2/bin /opt/kaldi/tools/openfst-1.7.2/bin # buildkit |
| COPY /workspace/submodules/kaldi/tools/openfst-1.7.2/lib /opt/kaldi/tools/openfst-1.7.2/lib # buildkit |
| COPY /workspace/submodules/kaldi/src/fstbin /opt/kaldi/src/fstbin # buildkit |
| COPY /workspace/submodules/kaldi/src/lmbin /opt/kaldi/src/lmbin # buildkit |
| COPY /opt/riva/kenlm/build/bin/ /usr/local/bin/ # buildkit |
| COPY ./scripts/generate_LM_binary.sh utils/generate_LM_binary.sh # buildkit |
| ARG TOOLS_DIR=arpa2fst |
| COPY tools/asr/arpa2fst utils/arpa2fst # buildkit |
| ENV PATH=/opt/riva/bin:/opt/tritonserver/bin:/usr/local/mpi/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/ucx/bin:/opt/amazon/efa/bin:/opt/kaldi/tools/openfst-1.7.2/bin:/opt/kaldi/src/fstbin:/opt/kaldi/src/lmbin |
| ENV LD_LIBRARY_PATH=/opt/tritonserver/backends/pytorch:/opt/tritonserver/lib:/opt/riva/lib:/usr/local/cuda/compat/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/opt/kaldi/tools/openfst-1.7.2/lib:/opt/kaldi/src/lib |
| COPY VERSION / # buildkit |
| COPY python/servicemaker/requirements.txt /tmp/ # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c pip3 install --no-cache-dir --upgrade nvidia-pyindex==1.0.6 && pip3 install --no-cache-dir --default-timeout=100 librosa==0.10.2.post1 onnxruntime==1.17.0 onnx-graphsurgeon==0.3.27 pycuda==2024.1.2 sentencepiece==0.2.0 omegaconf==2.3.0 "sqlparse>=0.5" polygraphy==0.49.9 pytest==8.3.3 pillow==10.4.0 mock==5.1.0 deepdiff==8.0.1 python-multipart==0.0.18 setuptools==70.0.0 pypinyin==0.49.0 nltk==3.9.1 # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c if [ "$TARGET_ARCH" = "x86_64" ]; then pip3 install --no-cache-dir onnxruntime-gpu==1.17.0 --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/; fi # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c cd /tmp && pip3 install --no-cache-dir --default-timeout=100 -r requirements.txt --extra-index-url=https://anonymous@urm.nvidia.com/artifactory/api/pypi/sw-eff-pypi-local/simple # buildkit |
| COPY python/servicemaker /tmp/servicemaker/ # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c cd /tmp/servicemaker/ && pip3 install --no-cache-dir .[all] # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c pip3 uninstall -y mistune # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c apt-get remove -y --purge libslurm*; apt-get remove -y --purge libpmi*; apt-get remove --purge -y openssh* || true # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c apt-get update && apt-get install --only-upgrade rsync # buildkit |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c if [ -f "/opt/hpcx/ompi/etc/openmpi-mca-params.conf" ]; then echo "plm_rsh_agent = false" >> /opt/hpcx/ompi/etc/openmpi-mca-params.conf; fi # buildkit |
| COPY python/download_ngc_models /usr/local/bin # buildkit |
| COPY python/download_and_deploy_ngc_models /usr/local/bin # buildkit |
| COPY python/deploy_all_models /usr/local/bin # buildkit |
| ENV PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python |
| ENV NUMBA_CACHE_DIR=/tmp |
| RUN |5 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst /bin/sh -c mkdir -p /opt/riva && tar xf /tmp/riva/riva_server_tar_distribution.tgz -C /opt/riva && tar xf /tmp/riva/riva_asrlib_tar_distribution.tgz -C /opt/riva # buildkit |
| ARG RIVA_NEW_TRITON_BACKENDS=riva_tts_chunker riva_tts_preprocessor riva_tts_postprocessor riva_asr_decoder riva_asr_endpointing riva_nlp_tokenizer riva_asr_features riva_nlp_qa riva_asr_lattices riva_nlp_seqlabel riva_nlp_detokenizer riva_nlp_pipeline riva_asr_pipeline riva_asr_ensemble_pipeline riva_tts_pipeline |
| RUN |6 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst RIVA_NEW_TRITON_BACKENDS=riva_tts_chunker riva_tts_preprocessor riva_tts_postprocessor riva_asr_decoder riva_asr_endpointing riva_nlp_tokenizer riva_asr_features riva_nlp_qa riva_asr_lattices riva_nlp_seqlabel riva_nlp_detokenizer riva_nlp_pipeline riva_asr_pipeline riva_asr_ensemble_pipeline riva_tts_pipeline /bin/sh -c for backend in ${RIVA_NEW_TRITON_BACKENDS}; do mkdir -p /opt/tritonserver/backends/${backend} && ln -s /opt/riva/backends/libtriton_${backend}.so /opt/tritonserver/backends/${backend}/; done # buildkit |
| ENV NVIDIA_PRODUCT_NAME=Riva Speech Server |
| COPY docker/entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit |
| ARG RIVA_SPEECH_VERSION=devel |
| ARG RIVA_SPEECH_CONTAINER_VERSION=devel |
| ENV RIVA_SPEECH_VERSION=2.19.0 |
| ENV NVIDIA_RIVA_SPEECH_SERVER_VERSION=25.02 |
| LABEL com.nvidia.riva.speech.version=2.19.0 |
| LABEL com.nvidia.riva.speech.container.version=25.02 |
| ARG NVIDIA_BUILD_ID |
| ENV NVIDIA_BUILD_ID=151443007 |
| LABEL com.nvidia.build.id=151443007 |
| ARG NVIDIA_BUILD_REF |
| LABEL com.nvidia.build.ref=9fc022890a4183d9105c373028a3217dab6dddde |
| SHELL [/bin/bash -c] |
| RUN |10 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst RIVA_NEW_TRITON_BACKENDS=riva_tts_chunker riva_tts_preprocessor riva_tts_postprocessor riva_asr_decoder riva_asr_endpointing riva_nlp_tokenizer riva_asr_features riva_nlp_qa riva_asr_lattices riva_nlp_seqlabel riva_nlp_detokenizer riva_nlp_pipeline riva_asr_pipeline riva_asr_ensemble_pipeline riva_tts_pipeline RIVA_SPEECH_VERSION=2.19.0 RIVA_SPEECH_CONTAINER_VERSION=25.02 NVIDIA_BUILD_ID=151443007 NVIDIA_BUILD_REF=9fc022890a4183d9105c373028a3217dab6dddde /bin/bash -c if [ "${RIVA_SPEECH_VERSION:(-4)}" = "-dev" ] || [ "${RIVA_SPEECH_VERSION}" = "devel" ]; then github_branch="main"; else github_branch="release/${RIVA_SPEECH_VERSION}"; fi && mkdir -p /tmp/client && cd /tmp/client/ && git clone -b $github_branch https://github.com/nvidia-riva/python-clients.git && cd python-clients && python3 -m pip install --no-cache-dir -r requirements.txt && cd scripts && find . -name "*.py" | grep -v __init__.py | xargs -i cp {} /opt/riva/examples && pip install --no-cache-dir git+https://github.com/nvidia-riva/python-clients/@$github_branch && rm -rf /tmp/client/ # buildkit |
| SHELL [/bin/sh -c] |
| ENV PATH=/opt/riva/clients:/usr/local/cuda-12.8/bin:/opt/riva/bin:/opt/tritonserver/bin:/usr/local/mpi/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/ucx/bin:/opt/amazon/efa/bin:/opt/kaldi/tools/openfst-1.7.2/bin:/opt/kaldi/src/fstbin:/opt/kaldi/src/lmbin |
| ENV LD_LIBRARY_PATH=/usr/local/cuda-12.8/lib64:/opt/tritonserver/backends/pytorch:/opt/tritonserver/lib:/opt/riva/lib:/usr/local/cuda/compat/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/opt/kaldi/tools/openfst-1.7.2/lib:/opt/kaldi/src/lib |
| COPY scripts/calc_wer.py utils/calc_wer.py # buildkit |
| COPY test_files/asr/public wav # buildkit |
| COPY test_files/asr/wav/test/1272-135031-0000.alaw /opt/riva/examples/asr_lib/ # buildkit |
| COPY test_files/asr/wav/test/1272-135031-0000.mulaw /opt/riva/examples/asr_lib/ # buildkit |
| COPY test_files/asr/wav/test/1272-135031-0000.wav /opt/riva/examples/asr_lib/ # buildkit |
| RUN |10 TARGET_OS=linux TARGET_ARCH=x86_64 TARGET_CUDA_VERSION=12.8 KALDI_DIR=submodules/kaldi TOOLS_DIR=arpa2fst RIVA_NEW_TRITON_BACKENDS=riva_tts_chunker riva_tts_preprocessor riva_tts_postprocessor riva_asr_decoder riva_asr_endpointing riva_nlp_tokenizer riva_asr_features riva_nlp_qa riva_asr_lattices riva_nlp_seqlabel riva_nlp_detokenizer riva_nlp_pipeline riva_asr_pipeline riva_asr_ensemble_pipeline riva_tts_pipeline RIVA_SPEECH_VERSION=2.19.0 RIVA_SPEECH_CONTAINER_VERSION=25.02 NVIDIA_BUILD_ID=151443007 NVIDIA_BUILD_REF=9fc022890a4183d9105c373028a3217dab6dddde /bin/sh -c mkdir /data/ && chmod 777 /data/ # buildkit |
