## ## cntk.spec -- OpenPKG RPM Package Specification ## Copyright (c) 2000-2022 OpenPKG Project ## ## Permission to use, copy, modify, and distribute this software for ## any purpose with or without fee is hereby granted, provided that ## the above copyright notice and this permission notice appear in all ## copies. ## ## THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED ## WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF ## MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. ## IN NO EVENT SHALL THE AUTHORS AND COPYRIGHT HOLDERS AND THEIR ## CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, ## SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT ## LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF ## USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ## ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, ## OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT ## OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF ## SUCH DAMAGE. ## # package information Name: cntk Summary: Computational Network ToolKit (CNTK) URL: https://github.com/Microsoft/CNTK Vendor: Microsoft Packager: OpenPKG Project Distribution: OpenPKG Community Class: EVAL Group: Algorithm License: BSD-style Version: 20170114 Release: 20170114 # package options %option with_opencv no # list of sources Source0: http://download.openpkg.org/components/versioned/cntk/CNTK-%{version}.tar.xz Source1: cntk.sh Patch0: cntk.patch # build information BuildPreReq: OpenPKG, openpkg >= 20160101, gcc, gcc::with_cxx = yes PreReq: OpenPKG, openpkg >= 20160101 BuildPreReq: openmpi, openmpi::with_cxx = yes, openblas, libzip, boost, protobuf PreReq: openmpi, openmpi::with_cxx = yes, openblas, libzip, boost, protobuf %if "%{with_opencv}" == "yes" BuildPreReq: opencv PreReq: opencv %endif %description CNTK, the Computational Network Toolkit by Microsoft Research, is a unified deep-learning toolkit that describes neural networks as a series of computational steps via a directed graph. In this directed graph, leaf nodes represent input values or network parameters, while other nodes represent matrix operations upon their inputs. CNTK allows to easily realize and combine popular model types such as feed-forward DNNs, convolutional nets (CNNs), and recurrent networks (RNNs/LSTMs). It implements stochastic gradient descent (SGD, error backpropagation) learning with automatic differentiation and parallelization across multiple GPUs and servers. %track prog cntk = { version = %{version} url = http://download.openpkg.org/components/versioned/cntk/ regex = CNTK-(__VER__)\.tar\.xz } %prep %setup -q -n CNTK %patch -p0 %build # patch the build tools %{l_shtool} subst \ -e 's;/bin/bash;%{l_bash};g' \ Tools/build-and-test \ Tools/generate_build_info \ Tools/make_binary_drop_linux # patch the build procedure case "%{l_platform -t}" in *-linux* ) ;; * ) %{l_shtool} subst -e 's; -ldl;;g' Makefile ;; esac # configure the toolkit CC="%{l_cc}" \ CXX="%{l_cxx}" \ CFLAGS="%{l_cflags -O}" \ CXXFLAGS="%{l_cxxflags -O}" \ CPPFLAGS="%{l_cppflags}" \ LDFLAGS="%{l_ldflags}" \ %{l_bash} ./configure \ --with-buildtype=release \ --with-openblas=%{l_prefix} \ %if "%{with_opencv}" == "yes" --with-opencv=%{l_prefix} \ %endif --with-libzip \ --with-code-coverage=no \ --asgd=no \ --cuda=no # build the toolkit %{l_make} %{l_mflags -O} %install # create installation hierarchy %{l_shtool} mkdir -f -p -m 755 \ $RPM_BUILD_ROOT%{l_prefix}/bin \ $RPM_BUILD_ROOT%{l_prefix}/libexec/cntk \ $RPM_BUILD_ROOT%{l_prefix}/lib/cntk # install the binary wrapper %{l_shtool} install -c -m 755 %{l_value -s -a} \ %{SOURCE cntk.sh} $RPM_BUILD_ROOT%{l_prefix}/bin/cntk # install the toolkit binaries %{l_shtool} install -c -m 755 \ bin/cntk $RPM_BUILD_ROOT%{l_prefix}/libexec/cntk/ %{l_shtool} install -c -m 644 \ bin/cntk.core.bs $RPM_BUILD_ROOT%{l_prefix}/libexec/cntk/ # install the toolkit libraries %{l_shtool} install -c -m 644 \ lib/* $RPM_BUILD_ROOT%{l_prefix}/lib/cntk/ # determine installation files %{l_rpmtool} files -v -ofiles -r$RPM_BUILD_ROOT %{l_files_std} %files -f files %clean