From 46c57579ffe428ec26f8441d87b89a29f8a565d8 Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Mon, 1 Apr 2024 17:08:22 -0700
Subject: [PATCH 01/10] Remove conda from build

---
 build.py | 42 +-----------------------------------------
 1 file changed, 1 insertion(+), 41 deletions(-)

diff --git a/build.py b/build.py
index 2dcf84480c..dcc483ac1e 100755
--- a/build.py
+++ b/build.py
@@ -60,8 +60,7 @@
 #      ORT version,
 #      ORT OpenVINO version (use None to disable OpenVINO in ORT),
 #      Standalone OpenVINO version,
-#      DCGM version,
-#      Conda version
+#      DCGM version
 #     )
 #
 # Currently the OpenVINO versions used in ORT and standalone must
@@ -77,7 +76,6 @@
         "2023.3.0",  # ORT OpenVINO
         "2023.3.0",  # Standalone OpenVINO
         "3.2.6",  # DCGM version
-        "py310_23.1.0-1",  # Conda version
         "0.3.2",  # vLLM version
     )
 }
@@ -868,38 +866,6 @@ def install_dcgm_libraries(dcgm_version, target_machine):
                 dcgm_version, dcgm_version
             )
 
-
-def install_miniconda(conda_version, target_machine):
-    if target_machine == "arm64":
-        # This branch used for the case when linux container builds on MacOS with ARM chip
-        # macos arm arch names "arm64" when in linux it's names "aarch64".
-        # So we just replace the architecture to able find right conda version for Linux
-        target_machine = "aarch64"
-    if conda_version == "":
-        fail(
-            "unable to determine default repo-tag, CONDA version not known for {}".format(
-                FLAGS.version
-            )
-        )
-    miniconda_url = f"https://repo.anaconda.com/miniconda/Miniconda3-{conda_version}-Linux-{target_machine}.sh"
-    if target_machine == "x86_64":
-        sha_sum = "32d73e1bc33fda089d7cd9ef4c1be542616bd8e437d1f77afeeaf7afdb019787"
-    else:
-        sha_sum = "80d6c306b015e1e3b01ea59dc66c676a81fa30279bc2da1f180a7ef7b2191d6e"
-    return f"""
-RUN mkdir -p /opt/
-RUN wget "{miniconda_url}" -O miniconda.sh -q && \
-    echo "{sha_sum}" "miniconda.sh" > shasum && \
-    sha256sum -c ./shasum && \
-    sh miniconda.sh -b -p /opt/conda && \
-    rm miniconda.sh shasum && \
-    find /opt/conda/ -follow -type f -name '*.a' -delete && \
-    find /opt/conda/ -follow -type f -name '*.js.map' -delete && \
-    /opt/conda/bin/conda clean -afy
-ENV PATH ${{PATH}}:/opt/conda/bin
-"""
-
-
 def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
     df = """
 ARG TRITON_VERSION={}
@@ -1020,10 +986,6 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
 ENTRYPOINT []
 """
 
-    # Install miniconda required for the DALI backend.
-    if target_platform() != "windows":
-        df += install_miniconda(argmap["CONDA_VERSION"], target_machine())
-
     with open(os.path.join(ddir, dockerfile_name), "w") as dfile:
         dfile.write(df)
 
@@ -1278,8 +1240,6 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
 """
 
     if "vllm" in backends:
-        # [DLIS-5606] Build Conda environment for vLLM backend
-        # Remove Pip install once vLLM backend moves to Conda environment.
         df += """
 # vLLM needed for vLLM backend
 RUN pip3 install vllm=={}

From 5c5d85016e1c4ab3409a83115e71017f4b12bb41 Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Mon, 1 Apr 2024 17:11:26 -0700
Subject: [PATCH 02/10] Escape slash symbol

---
 build.py | 164 +++++++++++++++++++++++++++----------------------------
 1 file changed, 82 insertions(+), 82 deletions(-)

diff --git a/build.py b/build.py
index dcc483ac1e..34b56bbcc5 100755
--- a/build.py
+++ b/build.py
@@ -847,9 +847,9 @@ def install_dcgm_libraries(dcgm_version, target_machine):
             return """
 ENV DCGM_VERSION {}
 # Install DCGM. Steps from https://developer.nvidia.com/dcgm#Downloads
-RUN curl -o /tmp/cuda-keyring.deb \
-    https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/sbsa/cuda-keyring_1.0-1_all.deb \
-    && apt install /tmp/cuda-keyring.deb && rm /tmp/cuda-keyring.deb && \
+RUN curl -o /tmp/cuda-keyring.deb \\
+    https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/sbsa/cuda-keyring_1.0-1_all.deb \\
+    && apt install /tmp/cuda-keyring.deb && rm /tmp/cuda-keyring.deb && \\
     apt-get update && apt-get install -y datacenter-gpu-manager=1:{}
 """.format(
                 dcgm_version, dcgm_version
@@ -858,9 +858,9 @@ def install_dcgm_libraries(dcgm_version, target_machine):
             return """
 ENV DCGM_VERSION {}
 # Install DCGM. Steps from https://developer.nvidia.com/dcgm#Downloads
-RUN curl -o /tmp/cuda-keyring.deb \
-    https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb \
-    && apt install /tmp/cuda-keyring.deb && rm /tmp/cuda-keyring.deb && \
+RUN curl -o /tmp/cuda-keyring.deb \\
+    https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb \\
+    && apt install /tmp/cuda-keyring.deb && rm /tmp/cuda-keyring.deb && \\
     apt-get update && apt-get install -y datacenter-gpu-manager=1:{}
 """.format(
                 dcgm_version, dcgm_version
@@ -894,16 +894,16 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
 ENV DEBIAN_FRONTEND=noninteractive
 
 # Install docker docker buildx
-RUN apt-get update \
-        && apt-get install -y ca-certificates curl gnupg \
-        && install -m 0755 -d /etc/apt/keyrings \
-        && curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /etc/apt/keyrings/docker.gpg \
-        && chmod a+r /etc/apt/keyrings/docker.gpg \
-        && echo \
-            "deb [arch="$(dpkg --print-architecture)" signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \
-            "$(. /etc/os-release && echo "$VERSION_CODENAME")" stable" | \
-            tee /etc/apt/sources.list.d/docker.list > /dev/null \
-        && apt-get update \
+RUN apt-get update \\
+        && apt-get install -y ca-certificates curl gnupg \\
+        && install -m 0755 -d /etc/apt/keyrings \\
+        && curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /etc/apt/keyrings/docker.gpg \\
+        && chmod a+r /etc/apt/keyrings/docker.gpg \\
+        && echo \\
+            "deb [arch="$(dpkg --print-architecture)" signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \\
+            "$(. /etc/os-release && echo "$VERSION_CODENAME")" stable" | \\
+            tee /etc/apt/sources.list.d/docker.list > /dev/null \\
+        && apt-get update \\
         && apt-get install -y docker.io docker-buildx-plugin
 
 # libcurl4-openSSL-dev is needed for GCS
@@ -911,45 +911,45 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
 # python3-pip and libarchive-dev is needed by python backend
 # libxml2-dev is needed for Azure Storage
 # scons is needed for armnn_tflite backend build dep
-RUN apt-get update \
-    && apt-get install -y --no-install-recommends \
-            ca-certificates \
-            autoconf \
-            automake \
-            build-essential \
-            git \
-            gperf \
-            libre2-dev \
-            libssl-dev \
-            libtool \
-            libcurl4-openssl-dev \
-            libb64-dev \
-            libgoogle-perftools-dev \
-            patchelf \
-            python3-dev \
-            python3-pip \
-            python3-setuptools \
-            rapidjson-dev \
-            scons \
-            software-properties-common \
-            pkg-config \
-            unzip \
-            wget \
-            zlib1g-dev \
-            libarchive-dev \
-            libxml2-dev \
-            libnuma-dev \
-            wget \
+RUN apt-get update \\
+    && apt-get install -y --no-install-recommends \\
+            ca-certificates \\
+            autoconf \\
+            automake \\
+            build-essential \\
+            git \\
+            gperf \\
+            libre2-dev \\
+            libssl-dev \\
+            libtool \\
+            libcurl4-openssl-dev \\
+            libb64-dev \\
+            libgoogle-perftools-dev \\
+            patchelf \\
+            python3-dev \\
+            python3-pip \\
+            python3-setuptools \\
+            rapidjson-dev \\
+            scons \\
+            software-properties-common \\
+            pkg-config \\
+            unzip \\
+            wget \\
+            zlib1g-dev \\
+            libarchive-dev \\
+            libxml2-dev \\
+            libnuma-dev \\
+            wget \\
     && rm -rf /var/lib/apt/lists/*
 
-RUN pip3 install --upgrade pip && \
+RUN pip3 install --upgrade pip && \\
     pip3 install --upgrade wheel setuptools docker
 
 # Install boost version >= 1.78 for boost::span
 # Current libboost-dev apt packages are < 1.78, so install from tar.gz
-RUN wget -O /tmp/boost.tar.gz \
-        https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz && \
-    (cd /tmp && tar xzf boost.tar.gz) && \
+RUN wget -O /tmp/boost.tar.gz \\
+        https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz && \\
+    (cd /tmp && tar xzf boost.tar.gz) && \\
     mv /tmp/boost_1_80_0/boost /usr/include/boost
 
 # Server build requires recent version of CMake (FetchContent required)
@@ -1163,11 +1163,11 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
 # non-root. Make sure that this user to given ID 1000. All server
 # artifacts copied below are assign to this user.
 ENV TRITON_SERVER_USER=triton-server
-RUN userdel tensorrt-server > /dev/null 2>&1 || true && \
-    if ! id -u $TRITON_SERVER_USER > /dev/null 2>&1 ; then \
-        useradd $TRITON_SERVER_USER; \
-    fi && \
-    [ `id -u $TRITON_SERVER_USER` -eq 1000 ] && \
+RUN userdel tensorrt-server > /dev/null 2>&1 || true && \\
+    if ! id -u $TRITON_SERVER_USER > /dev/null 2>&1 ; then \\
+        useradd $TRITON_SERVER_USER; \\
+    fi && \\
+    [ `id -u $TRITON_SERVER_USER` -eq 1000 ] && \\
     [ `id -g $TRITON_SERVER_USER` -eq 1000 ]
 
 # Ensure apt-get won't prompt for selecting options
@@ -1175,22 +1175,22 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
 
 # Common dependencies. FIXME (can any of these be conditional? For
 # example libcurl only needed for GCS?)
-RUN apt-get update \
-    && apt-get install -y --no-install-recommends \
-            clang \
-            curl \
-            dirmngr \
-            git \
-            gperf \
-            libb64-0d \
-            libcurl4-openssl-dev \
-            libgoogle-perftools-dev \
-            libjemalloc-dev \
-            libnuma-dev \
-            libre2-9 \
-            software-properties-common \
-            wget \
-            {backend_dependencies} \
+RUN apt-get update \\
+    && apt-get install -y --no-install-recommends \\
+            clang \\
+            curl \\
+            dirmngr \\
+            git \\
+            gperf \\
+            libb64-0d \\
+            libcurl4-openssl-dev \\
+            libgoogle-perftools-dev \\
+            libjemalloc-dev \\
+            libnuma-dev \\
+            libre2-9 \\
+            software-properties-common \\
+            wget \\
+            {backend_dependencies} \\
     && rm -rf /var/lib/apt/lists/*
 
 # Set TCMALLOC_RELEASE_RATE for users setting LD_PRELOAD with tcmalloc
@@ -1216,9 +1216,9 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
         df += install_dcgm_libraries(argmap["DCGM_VERSION"], target_machine)
         df += """
 # Extra defensive wiring for CUDA Compat lib
-RUN 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 \
+RUN 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
 """
     else:
@@ -1228,14 +1228,14 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
     if "python" in backends:
         df += """
 # python3, python3-pip and some pip installs required for the python backend
-RUN apt-get update && \
-    apt-get install -y --no-install-recommends \
-            python3 libarchive-dev \
-            python3-pip \
-            libpython3-dev && \
-    pip3 install --upgrade pip && \
-    pip3 install --upgrade wheel setuptools && \
-    pip3 install --upgrade numpy && \
+RUN apt-get update && \\
+    apt-get install -y --no-install-recommends \\
+            python3 libarchive-dev \\
+            python3-pip \\
+            libpython3-dev && \\
+    pip3 install --upgrade pip && \\
+    pip3 install --upgrade wheel setuptools && \\
+    pip3 install --upgrade numpy && \\
     rm -rf /var/lib/apt/lists/*
 """
 
@@ -1310,7 +1310,7 @@ def add_cpu_libs_to_linux_dockerfile(backends, target_machine):
 COPY --from=min_container /usr/lib/{libs_arch}-linux-gnu/libcudnn.so.9 /usr/lib/{libs_arch}-linux-gnu/libcudnn.so.9
 
 # patchelf is needed to add deps of libcublasLt.so.12 to libtorch_cuda.so
-RUN apt-get update && \
+RUN apt-get update && \\
         apt-get install -y --no-install-recommends openmpi-bin patchelf
 
 ENV LD_LIBRARY_PATH /usr/local/cuda/targets/{cuda_arch}-linux/lib:/usr/local/cuda/lib64/stubs:${{LD_LIBRARY_PATH}}

From 9833f05ffa03eff0e4bde7acb595bc535d56c7ba Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Mon, 1 Apr 2024 17:14:38 -0700
Subject: [PATCH 03/10] Escape slash: align

---
 build.py | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/build.py b/build.py
index 34b56bbcc5..db0fb9422c 100755
--- a/build.py
+++ b/build.py
@@ -859,9 +859,11 @@ def install_dcgm_libraries(dcgm_version, target_machine):
 ENV DCGM_VERSION {}
 # Install DCGM. Steps from https://developer.nvidia.com/dcgm#Downloads
 RUN curl -o /tmp/cuda-keyring.deb \\
-    https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb \\
-    && apt install /tmp/cuda-keyring.deb && rm /tmp/cuda-keyring.deb && \\
-    apt-get update && apt-get install -y datacenter-gpu-manager=1:{}
+      https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb \\
+    && apt install /tmp/cuda-keyring.deb \\
+    && rm /tmp/cuda-keyring.deb \\
+    && apt-get update \\
+    && apt-get install -y datacenter-gpu-manager=1:{}
 """.format(
                 dcgm_version, dcgm_version
             )

From a695e62bacd5a3d73fa6a01e92feee3b0e934cee Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Mon, 1 Apr 2024 17:21:15 -0700
Subject: [PATCH 04/10] Escape slash: align

---
 build.py | 43 ++++++++++++++++++++++---------------------
 1 file changed, 22 insertions(+), 21 deletions(-)

diff --git a/build.py b/build.py
index db0fb9422c..ed4085d9a2 100755
--- a/build.py
+++ b/build.py
@@ -950,9 +950,9 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
 # Install boost version >= 1.78 for boost::span
 # Current libboost-dev apt packages are < 1.78, so install from tar.gz
 RUN wget -O /tmp/boost.tar.gz \\
-        https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz && \\
-    (cd /tmp && tar xzf boost.tar.gz) && \\
-    mv /tmp/boost_1_80_0/boost /usr/include/boost
+        https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz \\
+    && (cd /tmp && tar xzf boost.tar.gz) \\
+    && mv /tmp/boost_1_80_0/boost /usr/include/boost
 
 # Server build requires recent version of CMake (FetchContent required)
 RUN apt update -q=2 \\
@@ -1165,12 +1165,12 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
 # non-root. Make sure that this user to given ID 1000. All server
 # artifacts copied below are assign to this user.
 ENV TRITON_SERVER_USER=triton-server
-RUN userdel tensorrt-server > /dev/null 2>&1 || true && \\
-    if ! id -u $TRITON_SERVER_USER > /dev/null 2>&1 ; then \\
-        useradd $TRITON_SERVER_USER; \\
-    fi && \\
-    [ `id -u $TRITON_SERVER_USER` -eq 1000 ] && \\
-    [ `id -g $TRITON_SERVER_USER` -eq 1000 ]
+RUN userdel tensorrt-server > /dev/null 2>&1 || true \\
+    && if ! id -u $TRITON_SERVER_USER > /dev/null 2>&1 ; then \\
+         useradd $TRITON_SERVER_USER; \\
+       fi \\
+    && [ `id -u $TRITON_SERVER_USER` -eq 1000 ] \\
+    && [ `id -g $TRITON_SERVER_USER` -eq 1000 ]
 
 # Ensure apt-get won't prompt for selecting options
 ENV DEBIAN_FRONTEND=noninteractive
@@ -1219,9 +1219,9 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
         df += """
 # Extra defensive wiring for CUDA Compat lib
 RUN 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
+    && echo ${_CUDA_COMPAT_PATH}/lib > /etc/ld.so.conf.d/00-cuda-compat.conf \\
+    && ldconfig \\
+    && rm -f ${_CUDA_COMPAT_PATH}/lib
 """
     else:
         df += add_cpu_libs_to_linux_dockerfile(backends, target_machine)
@@ -1230,15 +1230,16 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
     if "python" in backends:
         df += """
 # python3, python3-pip and some pip installs required for the python backend
-RUN apt-get update && \\
-    apt-get install -y --no-install-recommends \\
-            python3 libarchive-dev \\
-            python3-pip \\
-            libpython3-dev && \\
-    pip3 install --upgrade pip && \\
-    pip3 install --upgrade wheel setuptools && \\
-    pip3 install --upgrade numpy && \\
-    rm -rf /var/lib/apt/lists/*
+RUN apt-get update \\
+    && apt-get install -y --no-install-recommends \\
+          python3 \\
+          libarchive-dev \\
+          python3-pip \\
+          libpython3-dev \\
+    && pip3 install --upgrade pip \\
+    && pip3 install --upgrade wheel setuptools \\
+    && pip3 install --upgrade numpy \\
+    && rm -rf /var/lib/apt/lists/*
 """
 
     if "vllm" in backends:

From c67efe0884efee4ca19b1850965c25b2fde3e04b Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Mon, 1 Apr 2024 17:22:16 -0700
Subject: [PATCH 05/10] Escape slash: align

---
 build.py | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/build.py b/build.py
index ed4085d9a2..ea972c5db7 100755
--- a/build.py
+++ b/build.py
@@ -944,8 +944,11 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
             wget \\
     && rm -rf /var/lib/apt/lists/*
 
-RUN pip3 install --upgrade pip && \\
-    pip3 install --upgrade wheel setuptools docker
+RUN pip3 install --upgrade pip \\
+    && pip3 install --upgrade \\
+          wheel \\
+          setuptools \\
+          docker
 
 # Install boost version >= 1.78 for boost::span
 # Current libboost-dev apt packages are < 1.78, so install from tar.gz

From 3917a9749cc880981e1e7dceac653fd7a497c538 Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Mon, 1 Apr 2024 17:27:30 -0700
Subject: [PATCH 06/10] Escape slash: align

---
 build.py | 142 ++++++++++++++++++++++++++++---------------------------
 1 file changed, 72 insertions(+), 70 deletions(-)

diff --git a/build.py b/build.py
index ea972c5db7..90f164e717 100755
--- a/build.py
+++ b/build.py
@@ -848,9 +848,11 @@ def install_dcgm_libraries(dcgm_version, target_machine):
 ENV DCGM_VERSION {}
 # Install DCGM. Steps from https://developer.nvidia.com/dcgm#Downloads
 RUN curl -o /tmp/cuda-keyring.deb \\
-    https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/sbsa/cuda-keyring_1.0-1_all.deb \\
-    && apt install /tmp/cuda-keyring.deb && rm /tmp/cuda-keyring.deb && \\
-    apt-get update && apt-get install -y datacenter-gpu-manager=1:{}
+        https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/sbsa/cuda-keyring_1.0-1_all.deb \\
+      && apt install /tmp/cuda-keyring.deb \\
+      && rm /tmp/cuda-keyring.deb \\
+      && apt-get update \\
+      && apt-get install -y datacenter-gpu-manager=1:{}
 """.format(
                 dcgm_version, dcgm_version
             )
@@ -859,11 +861,11 @@ def install_dcgm_libraries(dcgm_version, target_machine):
 ENV DCGM_VERSION {}
 # Install DCGM. Steps from https://developer.nvidia.com/dcgm#Downloads
 RUN curl -o /tmp/cuda-keyring.deb \\
-      https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb \\
-    && apt install /tmp/cuda-keyring.deb \\
-    && rm /tmp/cuda-keyring.deb \\
-    && apt-get update \\
-    && apt-get install -y datacenter-gpu-manager=1:{}
+          https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb \\
+      && apt install /tmp/cuda-keyring.deb \\
+      && rm /tmp/cuda-keyring.deb \\
+      && apt-get update \\
+      && apt-get install -y datacenter-gpu-manager=1:{}
 """.format(
                 dcgm_version, dcgm_version
             )
@@ -897,16 +899,16 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
 
 # Install docker docker buildx
 RUN apt-get update \\
-        && apt-get install -y ca-certificates curl gnupg \\
-        && install -m 0755 -d /etc/apt/keyrings \\
-        && curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /etc/apt/keyrings/docker.gpg \\
-        && chmod a+r /etc/apt/keyrings/docker.gpg \\
-        && echo \\
-            "deb [arch="$(dpkg --print-architecture)" signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \\
-            "$(. /etc/os-release && echo "$VERSION_CODENAME")" stable" | \\
-            tee /etc/apt/sources.list.d/docker.list > /dev/null \\
-        && apt-get update \\
-        && apt-get install -y docker.io docker-buildx-plugin
+      && apt-get install -y ca-certificates curl gnupg \\
+      && install -m 0755 -d /etc/apt/keyrings \\
+      && curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /etc/apt/keyrings/docker.gpg \\
+      && chmod a+r /etc/apt/keyrings/docker.gpg \\
+      && echo \\
+          "deb [arch="$(dpkg --print-architecture)" signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu \\
+          "$(. /etc/os-release && echo "$VERSION_CODENAME")" stable" | \\
+          tee /etc/apt/sources.list.d/docker.list > /dev/null \\
+      && apt-get update \\
+      && apt-get install -y docker.io docker-buildx-plugin
 
 # libcurl4-openSSL-dev is needed for GCS
 # python3-dev is needed by Torchvision
@@ -914,7 +916,7 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
 # libxml2-dev is needed for Azure Storage
 # scons is needed for armnn_tflite backend build dep
 RUN apt-get update \\
-    && apt-get install -y --no-install-recommends \\
+      && apt-get install -y --no-install-recommends \\
             ca-certificates \\
             autoconf \\
             automake \\
@@ -942,10 +944,10 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
             libxml2-dev \\
             libnuma-dev \\
             wget \\
-    && rm -rf /var/lib/apt/lists/*
+      && rm -rf /var/lib/apt/lists/*
 
 RUN pip3 install --upgrade pip \\
-    && pip3 install --upgrade \\
+      && pip3 install --upgrade \\
           wheel \\
           setuptools \\
           docker
@@ -953,18 +955,18 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
 # Install boost version >= 1.78 for boost::span
 # Current libboost-dev apt packages are < 1.78, so install from tar.gz
 RUN wget -O /tmp/boost.tar.gz \\
-        https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz \\
-    && (cd /tmp && tar xzf boost.tar.gz) \\
-    && mv /tmp/boost_1_80_0/boost /usr/include/boost
+          https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz \\
+      && (cd /tmp && tar xzf boost.tar.gz) \\
+      && mv /tmp/boost_1_80_0/boost /usr/include/boost
 
 # Server build requires recent version of CMake (FetchContent required)
 RUN apt update -q=2 \\
-    && apt install -y gpg wget \\
-    && wget -O - https://apt.kitware.com/keys/kitware-archive-latest.asc 2>/dev/null | gpg --dearmor - |  tee /usr/share/keyrings/kitware-archive-keyring.gpg >/dev/null \\
-    && . /etc/os-release \\
-    && echo "deb [signed-by=/usr/share/keyrings/kitware-archive-keyring.gpg] https://apt.kitware.com/ubuntu/ $UBUNTU_CODENAME main" | tee /etc/apt/sources.list.d/kitware.list >/dev/null \\
-    && apt-get update -q=2 \\
-    && apt-get install -y --no-install-recommends cmake=3.27.7* cmake-data=3.27.7*
+      && apt install -y gpg wget \\
+      && wget -O - https://apt.kitware.com/keys/kitware-archive-latest.asc 2>/dev/null | gpg --dearmor - |  tee /usr/share/keyrings/kitware-archive-keyring.gpg >/dev/null \\
+      && . /etc/os-release \\
+      && echo "deb [signed-by=/usr/share/keyrings/kitware-archive-keyring.gpg] https://apt.kitware.com/ubuntu/ $UBUNTU_CODENAME main" | tee /etc/apt/sources.list.d/kitware.list >/dev/null \\
+      && apt-get update -q=2 \\
+      && apt-get install -y --no-install-recommends cmake=3.27.7* cmake-data=3.27.7*
 """
 
         if FLAGS.enable_gpu:
@@ -1090,13 +1092,13 @@ def create_dockerfile_linux(
         df += """
 # Remove TRT contents that are not needed in runtime
 RUN ARCH="$(uname -i)" \\
-    && rm -fr ${TRT_ROOT}/bin ${TRT_ROOT}/targets/${ARCH}-linux-gnu/bin ${TRT_ROOT}/data \\
-    && rm -fr  ${TRT_ROOT}/doc ${TRT_ROOT}/onnx_graphsurgeon ${TRT_ROOT}/python \\
-    && rm -fr ${TRT_ROOT}/samples  ${TRT_ROOT}/targets/${ARCH}-linux-gnu/samples
+      && rm -fr ${TRT_ROOT}/bin ${TRT_ROOT}/targets/${ARCH}-linux-gnu/bin ${TRT_ROOT}/data \\
+      && rm -fr  ${TRT_ROOT}/doc ${TRT_ROOT}/onnx_graphsurgeon ${TRT_ROOT}/python \\
+      && rm -fr ${TRT_ROOT}/samples  ${TRT_ROOT}/targets/${ARCH}-linux-gnu/samples
 
 # Install required packages for TRT-LLM models
 RUN python3 -m pip install --upgrade pip \\
-    && pip3 install transformers
+      && pip3 install transformers
 
 # Uninstall unused nvidia packages
 RUN if pip freeze | grep -q "nvidia.*"; then \\
@@ -1106,7 +1108,7 @@ def create_dockerfile_linux(
 
 # Drop the static libs
 RUN ARCH="$(uname -i)" \\
-    && rm -f ${TRT_ROOT}/targets/${ARCH}-linux-gnu/lib/libnvinfer*.a \\
+      && rm -f ${TRT_ROOT}/targets/${ARCH}-linux-gnu/lib/libnvinfer*.a \\
           ${TRT_ROOT}/targets/${ARCH}-linux-gnu/lib/libnvonnxparser_*.a
 
 ENV LD_LIBRARY_PATH=/usr/local/tensorrt/lib/:/opt/tritonserver/backends/tensorrtllm:$LD_LIBRARY_PATH
@@ -1169,11 +1171,11 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
 # artifacts copied below are assign to this user.
 ENV TRITON_SERVER_USER=triton-server
 RUN userdel tensorrt-server > /dev/null 2>&1 || true \\
-    && if ! id -u $TRITON_SERVER_USER > /dev/null 2>&1 ; then \\
-         useradd $TRITON_SERVER_USER; \\
-       fi \\
-    && [ `id -u $TRITON_SERVER_USER` -eq 1000 ] \\
-    && [ `id -g $TRITON_SERVER_USER` -eq 1000 ]
+      && if ! id -u $TRITON_SERVER_USER > /dev/null 2>&1 ; then \\
+          useradd $TRITON_SERVER_USER; \\
+        fi \\
+      && [ `id -u $TRITON_SERVER_USER` -eq 1000 ] \\
+      && [ `id -g $TRITON_SERVER_USER` -eq 1000 ]
 
 # Ensure apt-get won't prompt for selecting options
 ENV DEBIAN_FRONTEND=noninteractive
@@ -1181,22 +1183,22 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
 # Common dependencies. FIXME (can any of these be conditional? For
 # example libcurl only needed for GCS?)
 RUN apt-get update \\
-    && apt-get install -y --no-install-recommends \\
-            clang \\
-            curl \\
-            dirmngr \\
-            git \\
-            gperf \\
-            libb64-0d \\
-            libcurl4-openssl-dev \\
-            libgoogle-perftools-dev \\
-            libjemalloc-dev \\
-            libnuma-dev \\
-            libre2-9 \\
-            software-properties-common \\
-            wget \\
-            {backend_dependencies} \\
-    && rm -rf /var/lib/apt/lists/*
+      && apt-get install -y --no-install-recommends \\
+              clang \\
+              curl \\
+              dirmngr \\
+              git \\
+              gperf \\
+              libb64-0d \\
+              libcurl4-openssl-dev \\
+              libgoogle-perftools-dev \\
+              libjemalloc-dev \\
+              libnuma-dev \\
+              libre2-9 \\
+              software-properties-common \\
+              wget \\
+              {backend_dependencies} \\
+      && rm -rf /var/lib/apt/lists/*
 
 # Set TCMALLOC_RELEASE_RATE for users setting LD_PRELOAD with tcmalloc
 ENV TCMALLOC_RELEASE_RATE 200
@@ -1222,9 +1224,9 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
         df += """
 # Extra defensive wiring for CUDA Compat lib
 RUN 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
+      && echo ${_CUDA_COMPAT_PATH}/lib > /etc/ld.so.conf.d/00-cuda-compat.conf \\
+      && ldconfig \\
+      && rm -f ${_CUDA_COMPAT_PATH}/lib
 """
     else:
         df += add_cpu_libs_to_linux_dockerfile(backends, target_machine)
@@ -1234,15 +1236,15 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
         df += """
 # python3, python3-pip and some pip installs required for the python backend
 RUN apt-get update \\
-    && apt-get install -y --no-install-recommends \\
-          python3 \\
-          libarchive-dev \\
-          python3-pip \\
-          libpython3-dev \\
-    && pip3 install --upgrade pip \\
-    && pip3 install --upgrade wheel setuptools \\
-    && pip3 install --upgrade numpy \\
-    && rm -rf /var/lib/apt/lists/*
+      && apt-get install -y --no-install-recommends \\
+            python3 \\
+            libarchive-dev \\
+            python3-pip \\
+            libpython3-dev \\
+      && pip3 install --upgrade pip \\
+      && pip3 install --upgrade wheel setuptools \\
+      && pip3 install --upgrade numpy \\
+      && rm -rf /var/lib/apt/lists/*
 """
 
     if "vllm" in backends:
@@ -1316,8 +1318,8 @@ def add_cpu_libs_to_linux_dockerfile(backends, target_machine):
 COPY --from=min_container /usr/lib/{libs_arch}-linux-gnu/libcudnn.so.9 /usr/lib/{libs_arch}-linux-gnu/libcudnn.so.9
 
 # patchelf is needed to add deps of libcublasLt.so.12 to libtorch_cuda.so
-RUN apt-get update && \\
-        apt-get install -y --no-install-recommends openmpi-bin patchelf
+RUN apt-get update \\
+      && apt-get install -y --no-install-recommends openmpi-bin patchelf
 
 ENV LD_LIBRARY_PATH /usr/local/cuda/targets/{cuda_arch}-linux/lib:/usr/local/cuda/lib64/stubs:${{LD_LIBRARY_PATH}}
 """.format(

From 4ded78a72f53788120efd5ff426627ca78f12e27 Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Mon, 1 Apr 2024 17:31:58 -0700
Subject: [PATCH 07/10] Install virtualenv

---
 build.py | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/build.py b/build.py
index 90f164e717..a2c6d83a9f 100755
--- a/build.py
+++ b/build.py
@@ -950,7 +950,8 @@ def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
       && pip3 install --upgrade \\
           wheel \\
           setuptools \\
-          docker
+          docker \\
+          virtualenv
 
 # Install boost version >= 1.78 for boost::span
 # Current libboost-dev apt packages are < 1.78, so install from tar.gz
@@ -1242,8 +1243,11 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
             python3-pip \\
             libpython3-dev \\
       && pip3 install --upgrade pip \\
-      && pip3 install --upgrade wheel setuptools \\
-      && pip3 install --upgrade numpy \\
+      && pip3 install --upgrade \\
+            wheel \\
+            setuptools \\
+            numpy \\
+            virtualenv \\
       && rm -rf /var/lib/apt/lists/*
 """
 

From d0114f0db491487ff8a86a400dd483ed605bca29 Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Tue, 2 Apr 2024 19:53:41 -0700
Subject: [PATCH 08/10] Fix vLLM flag

---
 build.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/build.py b/build.py
index a2c6d83a9f..4a9e0513d8 100755
--- a/build.py
+++ b/build.py
@@ -76,6 +76,7 @@
         "2023.3.0",  # ORT OpenVINO
         "2023.3.0",  # Standalone OpenVINO
         "3.2.6",  # DCGM version
+        "py310_23.1.0-1",  # Conda version
         "0.3.2",  # vLLM version
     )
 }
@@ -1256,7 +1257,7 @@ def dockerfile_prepare_container_linux(argmap, backends, enable_gpu, target_mach
 # vLLM needed for vLLM backend
 RUN pip3 install vllm=={}
 """.format(
-            TRITON_VERSION_MAP[FLAGS.version][7]
+            TRITON_VERSION_MAP[FLAGS.version][6]
         )
 
     df += """

From 3287a5ba30aa956c6d57e8fdea2b5cd0d4d48ace Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Tue, 2 Apr 2024 19:56:14 -0700
Subject: [PATCH 09/10] remove conda flag

---
 build.py | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/build.py b/build.py
index 4a9e0513d8..a1ec6774b9 100755
--- a/build.py
+++ b/build.py
@@ -76,7 +76,6 @@
         "2023.3.0",  # ORT OpenVINO
         "2023.3.0",  # Standalone OpenVINO
         "3.2.6",  # DCGM version
-        "py310_23.1.0-1",  # Conda version
         "0.3.2",  # vLLM version
     )
 }
@@ -1418,9 +1417,6 @@ def create_build_dockerfiles(
         "DCGM_VERSION": ""
         if FLAGS.version is None or FLAGS.version not in TRITON_VERSION_MAP
         else TRITON_VERSION_MAP[FLAGS.version][5],
-        "CONDA_VERSION": ""
-        if FLAGS.version is None or FLAGS.version not in TRITON_VERSION_MAP
-        else TRITON_VERSION_MAP[FLAGS.version][6],
     }
 
     # For CPU-only image we need to copy some cuda libraries and dependencies

From 9b2d3531b2645fa4a509999e24ef2a3eed5e8514 Mon Sep 17 00:00:00 2001
From: Misha Chornyi <mchornyi@nvidia.com>
Date: Wed, 3 Apr 2024 10:07:28 -0700
Subject: [PATCH 10/10] Fix code style

---
 build.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/build.py b/build.py
index a1ec6774b9..2626dd4986 100755
--- a/build.py
+++ b/build.py
@@ -870,6 +870,7 @@ def install_dcgm_libraries(dcgm_version, target_machine):
                 dcgm_version, dcgm_version
             )
 
+
 def create_dockerfile_buildbase(ddir, dockerfile_name, argmap):
     df = """
 ARG TRITON_VERSION={}