Skip to content

P
Paddle

BaiXuePrincess / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle
提交 a9ea99d7 编写于 作者: S sneaxiy
浏览文件
操作

merge develop

上级 eac5a0aa dd1c7ee6
展开全部 隐藏空白更改
内联 并排
Showing with 34959 addition and 41055 deletion

要显示的变更太多。

To preserve performance only 1000 of 1000+ files are displayed.
文本差异 电子邮件补丁
.pre-commit-config.yaml
浏览文件 @ a9ea99d7
...@@ -42,12 +42,6 @@ repos: ...@@ -42,12 +42,6 @@ repos:
entry: bash ./tools/codestyle/pylint_pre_commit.hook entry: bash ./tools/codestyle/pylint_pre_commit.hook
language: system language: system
files: \.(py)$ files: \.(py)$
- repo: https://github.com/PaddlePaddle/pre-commit-golang
sha: 8337620115c25ff8333f1b1a493bd031049bd7c0
hooks:
- id: go-fmt
types:
- go
- repo: local - repo: local
hooks: hooks:
- id: copyright_checker - id: copyright_checker
......
.travis.yml
浏览文件 @ a9ea99d7
...@@ -4,7 +4,6 @@ cache: ...@@ -4,7 +4,6 @@ cache:
- $HOME/.ccache - $HOME/.ccache
- $HOME/.cache/pip - $HOME/.cache/pip
- $TRAVIS_BUILD_DIR/build/third_party - $TRAVIS_BUILD_DIR/build/third_party
- $TRAVIS_BUILD_DIR/build_android/third_party
sudo: required sudo: required
dist: trusty dist: trusty
services: services:
...@@ -13,7 +12,6 @@ os: ...@@ -13,7 +12,6 @@ os:
- linux - linux
env: env:
- JOB=check_style - JOB=check_style
- JOB=build_android
addons: addons:
ssh_known_hosts: 13.229.163.131 ssh_known_hosts: 13.229.163.131
before_install: before_install:
......
AUTHORS.md
浏览文件 @ a9ea99d7
...@@ -44,6 +44,7 @@ ...@@ -44,6 +44,7 @@
| qingqing01 | Qing-Qing Dang | | qingqing01 | Qing-Qing Dang |
| reyoung | Yang Yu | | reyoung | Yang Yu |
| Sand3r- | Michal Gallus | | Sand3r- | Michal Gallus |
| sfraczek | Sylwester Fraczek |
| Superjom | Chun-Wei Yan | | Superjom | Chun-Wei Yan |
| tensor-tang | Jian Tang | | tensor-tang | Jian Tang |
| tianbingsz | Tian-Bing Xu | | tianbingsz | Tian-Bing Xu |
...@@ -54,6 +55,7 @@ ...@@ -54,6 +55,7 @@
| wangyang59 | Yang Wang | | wangyang59 | Yang Wang |
| wangzhen-nlp | Zhen Wang | | wangzhen-nlp | Zhen Wang |
| wen-bo-yang | Wen-Bo Yang | | wen-bo-yang | Wen-Bo Yang |
| wojtuss | Wojciech Uss |
| wwhu | Wei-Wei Hu | | wwhu | Wei-Wei Hu |
| xinghai-sun | Xing-Hai Sun | | xinghai-sun | Xing-Hai Sun |
| Xreki | Yi-Qun Liu | | Xreki | Yi-Qun Liu |
......
CMakeLists.txt
浏览文件 @ a9ea99d7
...@@ -25,17 +25,21 @@ message(STATUS "CXX compiler: ${CMAKE_CXX_COMPILER}, version: " ...@@ -25,17 +25,21 @@ message(STATUS "CXX compiler: ${CMAKE_CXX_COMPILER}, version: "
message(STATUS "C compiler: ${CMAKE_C_COMPILER}, version: " message(STATUS "C compiler: ${CMAKE_C_COMPILER}, version: "
"${CMAKE_C_COMPILER_ID} ${CMAKE_C_COMPILER_VERSION}") "${CMAKE_C_COMPILER_ID} ${CMAKE_C_COMPILER_VERSION}")
if(WIN32) if(WIN32)
set(CMAKE_SUPPRESS_REGENERATION ON)
set(CMAKE_STATIC_LIBRARY_PREFIX lib) set(CMAKE_STATIC_LIBRARY_PREFIX lib)
add_definitions("/DGOOGLE_GLOG_DLL_DECL=") add_definitions("/DGOOGLE_GLOG_DLL_DECL=")
set(CMAKE_C_FLAGS_DEBUG "${CMAKE_C_FLAGS_DEBUG} /bigobj /MTd") set(CMAKE_C_FLAGS_DEBUG "${CMAKE_C_FLAGS_DEBUG} /bigobj /MTd")
set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} /bigobj /MT") set(CMAKE_C_FLAGS_RELEASE "${CMAKE_C_FLAGS_RELEASE} /bigobj /MT")
set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /bigobj /MTd") set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /bigobj /MTd")
set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /bigobj /MT") set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /bigobj /MT")
add_compile_options(/wd4068 /wd4129 /wd4244 /wd4267 /wd4297 /wd4530 /wd4577 /wd4819 /wd4838)
set(PADDLE_LINK_FLAGS "/IGNORE:4006 /IGNORE:4098 /IGNORE:4217 /IGNORE:4221")
set(CMAKE_STATIC_LINKER_FLAGS "${CMAKE_STATIC_LINKER_FLAGS} ${PADDLE_LINK_FLAGS}")
set(CMAKE_SHARED_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS} ${PADDLE_LINK_FLAGS}")
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} ${PADDLE_LINK_FLAGS}")
endif(WIN32) endif(WIN32)
if(NOT CMAKE_CROSSCOMPILING) find_package(CUDA QUIET)
find_package(CUDA QUIET)
endif(NOT CMAKE_CROSSCOMPILING)
find_package(Git REQUIRED) find_package(Git REQUIRED)
find_package(Threads REQUIRED) find_package(Threads REQUIRED)
...@@ -49,27 +53,13 @@ option(WITH_MKL "Compile PaddlePaddle with MKL support." ${AVX_FO ...@@ -49,27 +53,13 @@ option(WITH_MKL "Compile PaddlePaddle with MKL support." ${AVX_FO
option(WITH_NGRAPH "Compile PaddlePaddle with nGraph support." OFF) option(WITH_NGRAPH "Compile PaddlePaddle with nGraph support." OFF)
option(WITH_DSO "Compile PaddlePaddle with dynamic linked CUDA" ON) option(WITH_DSO "Compile PaddlePaddle with dynamic linked CUDA" ON)
option(WITH_TESTING "Compile PaddlePaddle with unit testing" OFF) option(WITH_TESTING "Compile PaddlePaddle with unit testing" OFF)
option(WITH_SWIG_PY "Compile PaddlePaddle with inference api" ON)
option(WITH_PYTHON "Compile PaddlePaddle with python interpreter" ON) option(WITH_PYTHON "Compile PaddlePaddle with python interpreter" ON)
option(WITH_DOUBLE "Compile PaddlePaddle with double precision" OFF)
option(WITH_RDMA "Compile PaddlePaddle with RDMA support" OFF)
option(WITH_TIMER "Compile PaddlePaddle with stats timer" OFF)
option(WITH_PROFILER "Compile PaddlePaddle with GPU profiler and gperftools" OFF) option(WITH_PROFILER "Compile PaddlePaddle with GPU profiler and gperftools" OFF)
option(WITH_JEMALLOC "Compile PaddlePaddle with jemalloc" OFF) option(WITH_JEMALLOC "Compile PaddlePaddle with jemalloc" OFF)
option(WITH_DOC "Compile PaddlePaddle with documentation" OFF)
option(WITH_COVERAGE "Compile PaddlePaddle with code coverage" OFF) option(WITH_COVERAGE "Compile PaddlePaddle with code coverage" OFF)
option(COVERALLS_UPLOAD "Package code coverage data to coveralls" OFF) option(COVERALLS_UPLOAD "Package code coverage data to coveralls" OFF)
option(ON_TRAVIS "Exclude special unit test on Travis CI" OFF)
option(WITH_C_API "Compile PaddlePaddle with C-API(Prediction)" OFF)
option(WITH_FLUID_ONLY "Compile PaddlePaddle fluid only" OFF)
option(WITH_GOLANG "Compile PaddlePaddle with GOLANG" OFF)
option(GLIDE_INSTALL "Download and install go dependencies " ON)
option(USE_NNPACK "Compile PaddlePaddle with NNPACK library" OFF)
option(WITH_DISTRIBUTE "Compile with distributed support" OFF) option(WITH_DISTRIBUTE "Compile with distributed support" OFF)
option(WITH_PSLIB "Compile with pslib support" OFF) option(WITH_PSLIB "Compile with pslib support" OFF)
option(USE_EIGEN_FOR_BLAS "Use matrix multiplication in Eigen" OFF)
option(EIGEN_USE_THREADS "Compile with multi-threaded Eigen" OFF)
option(WITH_ARM_FP16 "Use half precision support on armv8.2-a cpu" OFF)
option(WITH_CONTRIB "Compile the third-party contributation" OFF) option(WITH_CONTRIB "Compile the third-party contributation" OFF)
option(REPLACE_ENFORCE_GLOG "Replace PADDLE_ENFORCE with glog/CHECK for better debug." OFF) option(REPLACE_ENFORCE_GLOG "Replace PADDLE_ENFORCE with glog/CHECK for better debug." OFF)
option(WITH_ANAKIN "Compile with Anakin library" OFF) option(WITH_ANAKIN "Compile with Anakin library" OFF)
...@@ -96,37 +86,6 @@ if(NOT CMAKE_BUILD_TYPE) ...@@ -96,37 +86,6 @@ if(NOT CMAKE_BUILD_TYPE)
FORCE) FORCE)
endif() endif()
if(ANDROID OR IOS)
if(ANDROID)
if(${CMAKE_SYSTEM_VERSION} VERSION_LESS "16")
message(FATAL_ERROR "Unsupport standalone toolchains with Android API level lower than 16")
endif()
endif()
set(WITH_GPU OFF CACHE STRING
"Disable GPU when cross-compiling for Android and iOS" FORCE)
set(WITH_AVX OFF CACHE STRING
"Disable AVX when cross-compiling for Android and iOS" FORCE)
set(WITH_PYTHON OFF CACHE STRING
"Disable PYTHON when cross-compiling for Android and iOS" FORCE)
set(WITH_RDMA OFF CACHE STRING
"Disable RDMA when cross-compiling for Android and iOS" FORCE)
set(WITH_MKL OFF CACHE STRING
"Disable MKL when cross-compiling for Android and iOS" FORCE)
set(WITH_NGRAPH OFF CACHE STRING
"Disable nGraph when cross-compiling for Android and iOS" FORCE)
set(WITH_GOLANG OFF CACHE STRING
"Disable golang when cross-compiling for Android and iOS" FORCE)
# Compile PaddlePaddle mobile inference library
if (NOT WITH_C_API)
set(WITH_C_API ON CACHE STRING
"Always compile the C_API when cross-compiling for Android and iOS" FORCE)
endif()
set(MOBILE_INFERENCE ON)
add_definitions(-DPADDLE_MOBILE_INFERENCE)
endif()
if (APPLE) if (APPLE)
set(WITH_MKL OFF CACHE STRING set(WITH_MKL OFF CACHE STRING
"Disable MKL for building on mac" FORCE) "Disable MKL for building on mac" FORCE)
...@@ -135,10 +94,6 @@ endif() ...@@ -135,10 +94,6 @@ endif()
if (WIN32) if (WIN32)
set(WITH_DISTRIBUTE OFF CACHE STRING set(WITH_DISTRIBUTE OFF CACHE STRING
"Disable DISTRIBUTE when compiling for Windows" FORCE) "Disable DISTRIBUTE when compiling for Windows" FORCE)
set(WITH_C_API OFF CACHE STRING
"Disable C_API when compiling for Windows" FORCE)
set(WITH_FLUID_ONLY ON CACHE STRING
"Enable FLUID_ONLY when compiling for Windows" FORCE)
endif() endif()
set(THIRD_PARTY_PATH "${CMAKE_BINARY_DIR}/third_party" CACHE STRING set(THIRD_PARTY_PATH "${CMAKE_BINARY_DIR}/third_party" CACHE STRING
...@@ -150,21 +105,7 @@ set(FLUID_INSTALL_DIR "${CMAKE_BINARY_DIR}/fluid_install_dir" CACHE STRING ...@@ -150,21 +105,7 @@ set(FLUID_INSTALL_DIR "${CMAKE_BINARY_DIR}/fluid_install_dir" CACHE STRING
set(FLUID_INFERENCE_INSTALL_DIR "${CMAKE_BINARY_DIR}/fluid_inference_install_dir" CACHE STRING set(FLUID_INFERENCE_INSTALL_DIR "${CMAKE_BINARY_DIR}/fluid_inference_install_dir" CACHE STRING
"A path setting fluid inference shared and static libraries") "A path setting fluid inference shared and static libraries")
if (WITH_C_API AND WITH_PYTHON) set(THIRD_PARTY_BUILD_TYPE Release)
message(WARNING "It is suggest not embedded a python interpreter in Paddle "
"when using C-API. It will give an unpredictable behavior when using a "
"different Python interpreter from compiling.")
endif()
if (WITH_C_API)
set(WITH_FLUID_ONLY OFF CACHE STRING "Disable install fluid when compile the C_API" FORCE)
endif()
if(MOBILE_INFERENCE)
set(THIRD_PARTY_BUILD_TYPE MinSizeRel)
else()
set(THIRD_PARTY_BUILD_TYPE Release)
endif()
set(WITH_MKLML ${WITH_MKL}) set(WITH_MKLML ${WITH_MKL})
if (NOT DEFINED WITH_MKLDNN) if (NOT DEFINED WITH_MKLDNN)
...@@ -193,9 +134,7 @@ include(external/python) # download, build, install python ...@@ -193,9 +134,7 @@ include(external/python) # download, build, install python
include(external/openblas) # download, build, install openblas include(external/openblas) # download, build, install openblas
include(external/mkldnn) # download, build, install mkldnn include(external/mkldnn) # download, build, install mkldnn
include(external/ngraph) # download, build, install nGraph include(external/ngraph) # download, build, install nGraph
include(external/swig) # download, build, install swig
include(external/boost) # download boost include(external/boost) # download boost
include(external/any) # download libn::any
include(external/eigen) # download eigen3 include(external/eigen) # download eigen3
include(external/pybind11) # download pybind11 include(external/pybind11) # download pybind11
include(external/cares) include(external/cares)
...@@ -265,76 +204,25 @@ endif() ...@@ -265,76 +204,25 @@ endif()
if (WITH_JEMALLOC) if (WITH_JEMALLOC)
find_package(JeMalloc REQUIRED) find_package(JeMalloc REQUIRED)
include_directories(${JEMALLOC_INCLUDE_DIR}) include_directories(${JEMALLOC_INCLUDE_DIR})
add_definitions(-DWITH_JEMALLOC) add_definitions(-DPADDLE_WITH_JEMALLOC)
endif() endif()
include(generic) # simplify cmake module include(generic) # simplify cmake module
include(package) # set paddle packages include(package) # set paddle packages
include(ccache) # set ccache for compilation include(ccache) # set ccache for compilation
include(util) # set unittest and link libs include(util) # set unittest and link libs
include(rdma) # set rdma libraries
include(version) # set PADDLE_VERSION include(version) # set PADDLE_VERSION
include(coveralls) # set code coverage include(coveralls) # set code coverage
include(inference_lib) # add paddle fluid inference libraries include(inference_lib) # add paddle fluid inference libraries
include_directories("${PADDLE_SOURCE_DIR}") include_directories("${PADDLE_SOURCE_DIR}")
include_directories("${PADDLE_SOURCE_DIR}/paddle/legacy/cuda/include")
include_directories("${CMAKE_CURRENT_BINARY_DIR}/proto")
include_directories("${CMAKE_CURRENT_BINARY_DIR}/go/pserver/client/c")
set(EXTERNAL_LIBS
gflags
glog
${CBLAS_LIBRARIES}
protobuf
zlib
${PYTHON_LIBRARIES}
)
if(WITH_PSLIB)
list(APPEND EXTERNAL_LIBS pslib)
list(APPEND EXTERNAL_LIBS pslib_brpc)
list(APPEND EXTERNAL_LIBS libmct)
endif(WITH_PSLIB)
if(WITH_AMD_GPU) if(WITH_AMD_GPU)
find_package(HIP) find_package(HIP)
include(hip) include(hip)
endif(WITH_AMD_GPU) endif(WITH_AMD_GPU)
if(WITH_MKLML)
list(APPEND EXTERNAL_LIBS ${MKLML_IOMP_LIB})
endif()
if(WITH_LIBXSMM)
list(APPEND EXTERNAL_LIBS ${LIBXSMM_LIBS})
endif()
if(WITH_MKLDNN)
list(APPEND EXTERNAL_LIBS ${MKLDNN_LIB})
endif()
if(USE_NNPACK)
include(external/nnpack)
list(APPEND EXTERNAL_LIBS ${NNPACK_LIBS})
endif(USE_NNPACK)
add_subdirectory(proto)
if(NOT MOBILE_INFERENCE AND NOT WITH_FLUID_ONLY)
# "add_subdirectory(go)" should be placed after the following loine,
# because it depends on paddle/optimizer.
add_subdirectory(paddle/legacy/optimizer)
endif()
# "add_subdirectory(paddle)" and "add_subdirectory(python)" should be
# placed after this block, because they depends on it.
if(WITH_GOLANG)
enable_language(Go)
add_subdirectory(go)
endif(WITH_GOLANG)
set(PADDLE_PYTHON_BUILD_DIR "${CMAKE_CURRENT_BINARY_DIR}/python/build") set(PADDLE_PYTHON_BUILD_DIR "${CMAKE_CURRENT_BINARY_DIR}/python/build")
set(CMAKE_CXX_FLAGS_RELWITHDEBINFO "-O3 -g -DNDEBUG") set(CMAKE_CXX_FLAGS_RELWITHDEBINFO "-O3 -g -DNDEBUG")
...@@ -352,9 +240,3 @@ add_subdirectory(paddle) ...@@ -352,9 +240,3 @@ add_subdirectory(paddle)
if(WITH_PYTHON) if(WITH_PYTHON)
add_subdirectory(python) add_subdirectory(python)
endif() endif()
if(WITH_DOC)
find_package(Sphinx REQUIRED)
find_python_module(recommonmark REQUIRED)
add_subdirectory(doc)
endif()
CONTRIBUTING.md
浏览文件 @ a9ea99d7
...@@ -128,7 +128,7 @@ Please install pre-commit, which automatically reformat the changes to C/C++ and ...@@ -128,7 +128,7 @@ Please install pre-commit, which automatically reformat the changes to C/C++ and
Please remember to add related unit tests. Please remember to add related unit tests.
- For C/C++ code, please follow [`google-test` Primer](https://github.com/google/googletest/blob/master/googletest/docs/Primer.md). - For C/C++ code, please follow [`google-test` Primer](https://github.com/google/googletest/blob/master/googletest/docs/primer.md) .
- For Python code, please use [Python's standard `unittest` package](http://pythontesting.net/framework/unittest/unittest-introduction/). - For Python code, please use [Python's standard `unittest` package](http://pythontesting.net/framework/unittest/unittest-introduction/).
......
Dockerfile
浏览文件 @ a9ea99d7
...@@ -11,12 +11,10 @@ RUN /bin/bash -c 'if [[ -n ${UBUNTU_MIRROR} ]]; then sed -i 's#http://archive.ub ...@@ -11,12 +11,10 @@ RUN /bin/bash -c 'if [[ -n ${UBUNTU_MIRROR} ]]; then sed -i 's#http://archive.ub
# ENV variables # ENV variables
ARG WITH_GPU ARG WITH_GPU
ARG WITH_AVX ARG WITH_AVX
ARG WITH_DOC
ENV WOBOQ OFF ENV WOBOQ OFF
ENV WITH_GPU=${WITH_GPU:-ON} ENV WITH_GPU=${WITH_GPU:-ON}
ENV WITH_AVX=${WITH_AVX:-ON} ENV WITH_AVX=${WITH_AVX:-ON}
ENV WITH_DOC=${WITH_DOC:-OFF}
ENV HOME /root ENV HOME /root
# Add bash enhancements # Add bash enhancements
......
Dockerfile.android 已删除 100644 → 0
浏览文件 @ eac5a0aa
FROM ubuntu:16.04
MAINTAINER PaddlePaddle Authors <paddle-dev@baidu.com>
ARG UBUNTU_MIRROR
RUN /bin/bash -c 'if [[ -n ${UBUNTU_MIRROR} ]]; then sed -i 's#http://archive.ubuntu.com/ubuntu#${UBUNTU_MIRROR}#g' /etc/apt/sources.list; fi'
# ENV variables
ARG ANDROID_ABI
ARG ANDROID_API
ENV ANDROID_ABI=${ANDROID_ABI:-"armeabi-v7a"}
ENV ANDROID_API=${ANDROID_API:-21}
ENV HOME=/root \
ANDROID_NDK_HOME=/opt/android-ndk-linux \
ANDROID_TOOLCHAINS_DIR=/opt/toolchains
RUN apt-get update && \
apt-get install -y \
git python-dev python-pip python-numpy \
wget curl tar unzip gcc g++ locales clang-format-3.8 swig cmake && \
apt-get clean -y
# git credential to skip password typing
RUN git config --global credential.helper store
# Fix locales to en_US.UTF-8
RUN localedef -i en_US -f UTF-8 en_US.UTF-8
RUN pip install --upgrade pip==9.0.3 && \
pip install -U 'protobuf==3.1.0' && \
pip install -U wheel sphinx && \
pip install pre-commit
# Android NDK
RUN mkdir -p ${ANDROID_TOOLCHAINS_DIR} && \
mkdir -p /opt/android-ndk-tmp && \
cd /opt/android-ndk-tmp && \
wget -q https://dl.google.com/android/repository/android-ndk-r14b-linux-x86_64.zip && \
unzip -q android-ndk-r14b-linux-x86_64.zip && \
mv android-ndk-r14b ${ANDROID_NDK_HOME} && \
rm -rf /opt/android-ndk-tmp
README.md
浏览文件 @ a9ea99d7
# PaddlePaddle # PaddlePaddle
English | [简体中文](./README_cn.md)
[![Build Status](https://travis-ci.org/PaddlePaddle/Paddle.svg?branch=develop)](https://travis-ci.org/PaddlePaddle/Paddle) [![Build Status](https://travis-ci.org/PaddlePaddle/Paddle.svg?branch=develop)](https://travis-ci.org/PaddlePaddle/Paddle)
[![Documentation Status](https://img.shields.io/badge/docs-latest-brightgreen.svg?style=flat)](http://paddlepaddle.org/documentation/docs/en/1.2/getstarted/index_en.html) [![Documentation Status](https://img.shields.io/badge/docs-latest-brightgreen.svg?style=flat)](http://paddlepaddle.org/documentation/docs/en/1.3/beginners_guide/index_en.html)
[![Documentation Status](https://img.shields.io/badge/中文文档-最新-brightgreen.svg)](http://paddlepaddle.org/documentation/docs/zh/1.2/beginners_guide/index.html) [![Documentation Status](https://img.shields.io/badge/中文文档-最新-brightgreen.svg)](http://paddlepaddle.org/documentation/docs/zh/1.3/beginners_guide/index.html)
[![Release](https://img.shields.io/github/release/PaddlePaddle/Paddle.svg)](https://github.com/PaddlePaddle/Paddle/releases) [![Release](https://img.shields.io/github/release/PaddlePaddle/Paddle.svg)](https://github.com/PaddlePaddle/Paddle/releases)
[![License](https://img.shields.io/badge/license-Apache%202-blue.svg)](LICENSE) [![License](https://img.shields.io/badge/license-Apache%202-blue.svg)](LICENSE)
Welcome to the PaddlePaddle GitHub. Welcome to the PaddlePaddle GitHub.
PaddlePaddle (PArallel Distributed Deep LEarning) is an easy-to-use, PaddlePaddle (PArallel Distributed Deep LEarning) is an easy-to-use,
...@@ -18,17 +18,7 @@ learning to many products at Baidu. ...@@ -18,17 +18,7 @@ learning to many products at Baidu.
Our vision is to enable deep learning for everyone via PaddlePaddle. Our vision is to enable deep learning for everyone via PaddlePaddle.
Please refer to our [release announcement](https://github.com/PaddlePaddle/Paddle/releases) to track the latest feature of PaddlePaddle. Please refer to our [release announcement](https://github.com/PaddlePaddle/Paddle/releases) to track the latest feature of PaddlePaddle.
### Latest PaddlePaddle Release: [Fluid 1.3.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.3)
欢迎来到 PaddlePaddle GitHub
PaddlePaddle (PArallel Distributed Deep LEarning) 是一个简单易用、高效灵活、可扩展的深度学习平台，最初由百度科学家和工程师共同开发，目的是将深度学习技术应用到百度的众多产品中。
我们的愿景是让每个人都能通过PaddlePaddle接触深度学习
跟进PaddlePaddle最新特性请参考我们的[版本说明](https://github.com/PaddlePaddle/Paddle/releases)
### Latest PaddlePaddle Release: [Fluid 1.2.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.2)
### Install Latest Stable Release: ### Install Latest Stable Release:
``` ```
# Linux CPU # Linux CPU
...@@ -36,30 +26,13 @@ pip install paddlepaddle ...@@ -36,30 +26,13 @@ pip install paddlepaddle
# Linux GPU cuda9cudnn7 # Linux GPU cuda9cudnn7
pip install paddlepaddle-gpu pip install paddlepaddle-gpu
# Linux GPU cuda8cudnn7 # Linux GPU cuda8cudnn7
pip install paddlepaddle-gpu==1.2.0.post87 pip install paddlepaddle-gpu==1.3.0.post87
# Linux GPU cuda8cudnn5 # Linux GPU cuda8cudnn5
pip install paddlepaddle-gpu==1.2.0.post85 pip install paddlepaddle-gpu==1.3.0.post85
# For installation on other platform, refer to http://paddlepaddle.org/ # For installation on other platform, refer to http://paddlepaddle.org/
``` ```
### PaddlePaddle最新版本: [Fluid 1.2.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.2)
### 安装最新稳定版本:
```
# Linux CPU
pip install paddlepaddle
# Linux GPU cuda9cudnn7
pip install paddlepaddle-gpu
# Linux GPU cuda8cudnn7
pip install paddlepaddle-gpu==1.2.0.post87
# Linux GPU cuda8cudnn5
pip install paddlepaddle-gpu==1.2.0.post85
# 其他平台上的安装指引请参考 http://paddlepaddle.org/
```
## Features ## Features
- **Flexibility** - **Flexibility**
...@@ -100,90 +73,34 @@ pip install paddlepaddle-gpu==1.2.0.post85 ...@@ -100,90 +73,34 @@ pip install paddlepaddle-gpu==1.2.0.post85
Baidu and it has achieved a significant impact. We hope you can also explore Baidu and it has achieved a significant impact. We hope you can also explore
the capability of PaddlePaddle to make an impact on your product. the capability of PaddlePaddle to make an impact on your product.
## 特点
- **灵活性**
PaddlePaddle支持丰富的神经网络架构和优化算法。易于配置复杂模型，例如带有注意力机制或复杂记忆连接的神经网络机器翻译模型。
- **高效性**
为了高效使用异步计算资源，PaddlePaddle对框架的不同层进行优化，包括计算、存储、架构和通信。下面是一些样例：
- 通过SSE/AVX 内置函数、BLAS库(例如MKL、OpenBLAS、cuBLAS)或定制的CPU/GPU内核优化数学操作。
- 通过MKL-DNN库优化CNN网络
- 高度优化循环网络，无需执行 `padding` 操作即可处理 **变长** 序列
- 针对高维稀疏数据模型，优化了局部和分布式训练。
- **稳定性**
有了 PaddlePaddle，使得利用各种CPU/GPU和机器来加速训练变得简单。PaddlePaddle 通过优化通信可以实现巨大吞吐量和快速执行。
- **连接产品**
另外，PaddlePaddle 的设计也易于部署。在百度，PaddlePaddle 已经部署到含有巨大用户量的产品和服务上，包括广告点击率（CTR）预测、大规模图像分类、光学字符识别（OCR）、搜索排序，计算机病毒检测、推荐系统等等。PaddlePaddle广泛应用于百度产品中，产生了非常重要的影响。我们希望您也能探索 PaddlePaddle 的能力，为您的产品创造新的影响力和效果。
## Installation ## Installation
It is recommended to read [this doc](http://paddlepaddle.org/documentation/docs/zh/1.2/beginners_guide/install/index_cn.html) on our website. It is recommended to read [this doc](http://paddlepaddle.org/documentation/docs/en/1.3/beginners_guide/index_en.html) on our website.
## 安装
推荐阅读官网上的[安装说明](http://paddlepaddle.org/documentation/docs/zh/1.2/beginners_guide/install/index_cn.html)
## Documentation ## Documentation
We provide [English](http://paddlepaddle.org/documentation/docs/en/1.2/getstarted/index_en.html) and We provide [English](http://paddlepaddle.org/documentation/docs/en/1.3/beginners_guide/index_en.html) and
[Chinese](http://paddlepaddle.org/documentation/docs/zh/1.2/beginners_guide/index.html) documentation. [Chinese](http://paddlepaddle.org/documentation/docs/zh/1.3/beginners_guide/index.html) documentation.
- [Deep Learning 101](https://github.com/PaddlePaddle/book) - [Deep Learning 101](https://github.com/PaddlePaddle/book)
You might want to start from this online interactive book that can run in a Jupyter Notebook. You might want to start from this online interactive book that can run in a Jupyter Notebook.
- [Distributed Training](http://paddlepaddle.org/documentation/docs/zh/1.2/user_guides/howto/training/cluster_howto.html) - [Distributed Training](http://paddlepaddle.org/documentation/docs/en/1.3/user_guides/howto/training/multi_node_en.html)
You can run distributed training jobs on MPI clusters. You can run distributed training jobs on MPI clusters.
- [Python API](http://paddlepaddle.org/documentation/docs/zh/1.2/api_cn/index_cn.html) - [Python API](http://paddlepaddle.org/documentation/docs/en/1.3/api/index_en.html)
Our new API enables much shorter programs. Our new API enables much shorter programs.
- [How to Contribute](http://paddlepaddle.org/documentation/docs/zh/1.2/advanced_usage/development/contribute_to_paddle/index_cn.html) - [How to Contribute](http://paddlepaddle.org/documentation/docs/en/1.3/advanced_usage/development/contribute_to_paddle/index_en.html)
We appreciate your contributions! We appreciate your contributions!
## 文档
我们提供[英文](http://paddlepaddle.org/documentation/docs/en/1.2/getstarted/index_en.html)和
[中文](http://paddlepaddle.org/documentation/docs/zh/1.2/beginners_guide/index.html) 文档
- [深度学习101](https://github.com/PaddlePaddle/book)
或许您想从这个在线交互式书籍开始，可以在Jupyter Notebook中运行
- [分布式训练](http://paddlepaddle.org/documentation/docs/zh/1.2/user_guides/howto/training/cluster_howto.html)
可以在MPI集群上运行分布式训练任务
- [Python API](http://paddlepaddle.org/documentation/docs/zh/1.2/api_cn/index_cn.html)
新的API支持代码更少更简洁的程序
- [贡献方式](http://paddlepaddle.org/documentation/docs/zh/1.2/advanced_usage/development/contribute_to_paddle/index_cn.html)
欢迎您的贡献!
## Ask Questions ## Ask Questions
You are welcome to submit questions and bug reports as [Github Issues](https://github.com/PaddlePaddle/Paddle/issues). You are welcome to submit questions and bug reports as [Github Issues](https://github.com/PaddlePaddle/Paddle/issues).
## 答疑
欢迎您将问题和bug报告以[Github Issues](https://github.com/PaddlePaddle/Paddle/issues)的形式提交
## Copyright and License ## Copyright and License
PaddlePaddle is provided under the [Apache-2.0 license](LICENSE). PaddlePaddle is provided under the [Apache-2.0 license](LICENSE).
## 版权和许可证
PaddlePaddle由[Apache-2.0 license](LICENSE)提供
README_cn.md 0 → 100644
浏览文件 @ a9ea99d7
# PaddlePaddle
[English](./README.md) | 简体中文
[![Build Status](https://travis-ci.org/PaddlePaddle/Paddle.svg?branch=develop)](https://travis-ci.org/PaddlePaddle/Paddle)
[![Documentation Status](https://img.shields.io/badge/docs-latest-brightgreen.svg?style=flat)](http://paddlepaddle.org/documentation/docs/en/1.3/beginners_guide/index_en.html)
[![Documentation Status](https://img.shields.io/badge/中文文档-最新-brightgreen.svg)](http://paddlepaddle.org/documentation/docs/zh/1.3/beginners_guide/index.html)
[![Release](https://img.shields.io/github/release/PaddlePaddle/Paddle.svg)](https://github.com/PaddlePaddle/Paddle/releases)
[![License](https://img.shields.io/badge/license-Apache%202-blue.svg)](LICENSE)
欢迎来到 PaddlePaddle GitHub
PaddlePaddle (PArallel Distributed Deep LEarning) 是一个简单易用、高效灵活、可扩展的深度学习平台，最初由百度科学家和工程师共同开发，目的是将深度学习技术应用到百度的众多产品中。
我们的愿景是让每个人都能通过PaddlePaddle接触深度学习
跟进PaddlePaddle最新特性请参考我们的[版本说明](https://github.com/PaddlePaddle/Paddle/releases)
### PaddlePaddle最新版本: [Fluid 1.3.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.3)
### 安装最新稳定版本:
```
# Linux CPU
pip install paddlepaddle
# Linux GPU cuda9cudnn7
pip install paddlepaddle-gpu
# Linux GPU cuda8cudnn7
pip install paddlepaddle-gpu==1.3.0.post87
# Linux GPU cuda8cudnn5
pip install paddlepaddle-gpu==1.3.0.post85
# 其他平台上的安装指引请参考 http://paddlepaddle.org/
```
## 特性
- **灵活性**
PaddlePaddle支持丰富的神经网络架构和优化算法。易于配置复杂模型，例如带有注意力机制或复杂记忆连接的神经网络机器翻译模型。
- **高效性**
为了高效使用异步计算资源，PaddlePaddle对框架的不同层进行优化，包括计算、存储、架构和通信。下面是一些样例：
- 通过SSE/AVX 内置函数、BLAS库(例如MKL、OpenBLAS、cuBLAS)或定制的CPU/GPU内核优化数学操作。
- 通过MKL-DNN库优化CNN网络
- 高度优化循环网络，无需执行 `padding` 操作即可处理 **变长** 序列
- 针对高维稀疏数据模型，优化了局部和分布式训练。
- **稳定性**
有了 PaddlePaddle，使得利用各种CPU/GPU和机器来加速训练变得简单。PaddlePaddle 通过优化通信可以实现巨大吞吐量和快速执行。
- **与产品相连**
另外，PaddlePaddle 的设计也易于部署。在百度，PaddlePaddle 已经部署到含有巨大用户量的产品和服务上，包括广告点击率（CTR）预测、大规模图像分类、光学字符识别（OCR）、搜索排序，计算机病毒检测、推荐系统等等。PaddlePaddle广泛应用于百度产品中，产生了非常重要的影响。我们希望您也能探索 PaddlePaddle 的能力，为您的产品创造新的影响力和效果。
## 安装
推荐阅读官网上的[安装说明](http://paddlepaddle.org/documentation/docs/zh/1.3/beginners_guide/install/index_cn.html)
## 文档
我们提供[英文](http://paddlepaddle.org/documentation/docs/en/1.3/beginners_guide/index_en.html)和
[中文](http://paddlepaddle.org/documentation/docs/zh/1.3/beginners_guide/index.html) 文档
- [深度学习101](https://github.com/PaddlePaddle/book)
或许您想从这个在线交互式书籍开始，可以在Jupyter Notebook中运行
- [分布式训练](http://paddlepaddle.org/documentation/docs/zh/1.3/user_guides/howto/training/multi_node.html)
可以在MPI集群上运行分布式训练任务
- [Python API](http://paddlepaddle.org/documentation/docs/zh/1.3/api_cn/index_cn.html)
新的API支持代码更少更简洁的程序
- [贡献方式](http://paddlepaddle.org/documentation/docs/zh/1.3/advanced_usage/development/contribute_to_paddle/index_cn.html)
欢迎您的贡献!
## 答疑
欢迎您将问题和bug报告以[Github Issues](https://github.com/PaddlePaddle/Paddle/issues)的形式提交
## 版权和许可证
PaddlePaddle由[Apache-2.0 license](LICENSE)提供
benchmark/IntelOptimizedPaddle.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Benchmark
Machine:
- Server: Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz, 2 Sockets, 20 Cores per socket
- Laptop: TBD
System: CentOS release 6.3 (Final), Docker 1.12.1.
PaddlePaddle:
- paddlepaddle/paddle:0.11.0 (for MKLML and MKL-DNN)
- MKL-DNN tag v0.11
- MKLML 2018.0.1.20171007
- paddlepaddle/paddle:0.11.0-openblas (for OpenBLAS)
- OpenBLAS v0.2.20
On each machine, we will test and compare the performance of training on single node using MKL-DNN / MKLML / OpenBLAS respectively.
## Benchmark Model
### Server
#### Training
Test on batch size 64, 128, 256 on Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz
Pay attetion that the speed below includes forward, backward and parameter update time. So we can not directly compare the data with the benchmark of caffe `time` [command](https://github.com/PaddlePaddle/Paddle/blob/develop/benchmark/caffe/image/run.sh#L9), which only contain forward and backward. The updating time of parameter would become very heavy when the weight size are large, especially on alexnet.
Input image size - 3 * 224 * 224, Time: images/second
- VGG-19
| BatchSize | 64 | 128 | 256 |
|--------------|-------| -----| --------|
| OpenBLAS | 7.80 | 9.00 | 10.80 |
| MKLML | 12.12 | 13.70 | 16.18 |
| MKL-DNN | 28.46 | 29.83 | 30.44 |
<img src="figs/vgg-cpu-train.png" width="500">
- ResNet-50
| BatchSize | 64 | 128 | 256 |
|--------------|-------| ------| -------|
| OpenBLAS | 25.22 | 25.68 | 27.12 |
| MKLML | 32.52 | 31.89 | 33.12 |
| MKL-DNN | 81.69 | 82.35 | 84.08 |
<img src="figs/resnet-cpu-train.png" width="500">
- GoogLeNet
| BatchSize | 64 | 128 | 256 |
|--------------|-------| ------| -------|
| OpenBLAS | 89.52 | 96.97 | 108.25 |
| MKLML | 128.46| 137.89| 158.63 |
| MKL-DNN     | 250.46| 264.83| 269.50 |
<img src="figs/googlenet-cpu-train.png" width="500">
- AlexNet
| BatchSize | 64 | 128 | 256 |
|--------------|--------| ------ | -------|
| OpenBLAS | 45.62 | 72.79 | 107.22 |
| MKLML | 66.37 | 105.60 | 144.04 |
| MKL-DNN | 399.00 | 498.94 | 626.53 |
<img src="figs/alexnet-cpu-train.png" width="500">
#### Inference
Test on batch size 1, 2, 4, 8, 16 on Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz
- VGG-19
| BatchSize | 1 | 2 | 4 | 8 | 16 |
|-----------|-------|-------|-------|-------|-------|
| OpenBLAS | 1.10 | 1.96 | 3.62 | 3.63 | 2.25 |
| MKLML | 5.58 | 9.80 | 15.15 | 21.21 | 28.67 |
| MKL-DNN | 75.07 | 88.64 | 82.58 | 92.29 | 96.75 |
<img src="figs/vgg-cpu-infer.png" width="500">
- ResNet-50
| BatchSize | 1 | 2 | 4 | 8 | 16 |
|-----------|-------|--------|--------|--------|--------|
| OpenBLAS | 3.31 | 6.72 | 11.59 | 13.17 | 9.27 |
| MKLML | 6.33 | 12.02 | 22.88 | 40.53 | 63.09 |
| MKL-DNN | 107.83| 148.84 | 177.78 | 189.35 | 217.69 |
<img src="figs/resnet-cpu-infer.png" width="500">
- GoogLeNet
| BatchSize | 1 | 2 | 4 | 8 | 16 |
|-----------|--------|--------|--------|--------|--------|
| OpenBLAS | 12.06 | 23.56 | 34.48 | 36.45 | 23.12 |
| MKLML | 22.74 | 41.56 | 81.22 | 133.47 | 210.53 |
| MKL-DNN | 175.10 | 272.92 | 450.70 | 512.00 | 600.94 |
<img src="figs/googlenet-cpu-infer.png" width="500">
- AlexNet
| BatchSize | 1 | 2 | 4 | 8 | 16 |
|-----------|--------|--------|--------|--------|--------|
| OpenBLAS | 3.53 | 6.23 | 15.04 | 26.06 | 31.62 |
| MKLML | 21.32 | 36.55 | 73.06 | 131.15 | 192.77 |
| MKL-DNN | 442.91 | 656.41 | 719.10 | 847.68 | 850.51 |
<img src="figs/alexnet-cpu-infer.png" width="500">
### Laptop
TBD
benchmark/README.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Benchmark
Machine:
- CPU: 12-core Intel(R) Xeon(R) CPU E5-2620 v2 @2.10GHz
- GPU: Tesla K40m
- cuDNN: v5.1
- system: Docker 1.12.1, all platforms are tested in docker environment.
Platforms:
- PaddlePaddle: paddledev/paddle:gpu-devel-v0.9.0a0
- Tensorflow: gcr.io/tensorflow/tensorflow:0.11.0rc0-gpu
- Caffe: kaixhin/cuda-caffe
Several convolutional neural networks and recurrent neural networks are used to test.
## Image
### Benchmark Model
AlexNet, GoogleNet and a small network used in Caffe.
- [AlexNet](https://github.com/BVLC/caffe/tree/master/models/bvlc_alexnet): but the group size is one.
- [GoogleNet](https://github.com/BVLC/caffe/tree/master/models/bvlc_googlenet): but remove loss1 and loss2 when testing benchmark.
- [SmallNet](https://github.com/BVLC/caffe/blob/master/examples/cifar10/cifar10\_quick\_train\_test.prototxt)
### Single-GPU
- AlexNet: input - 3 * 227 * 227, Time: ms/batch
| BatchSize | 64 | 128 | 256 | 512 |
|--------------|-----| -----| ------| -----|
| PaddlePaddle | 195 | 334 | 602 | 1629 |
| TensorFlow | 223 | 364 | 645 | 1235 |
| Caffe | 324 | 627 | 1232 | 2513 |
**Notation**
All platforms use cuDNN-v5.1. We see that caffe is slower in this experiment, because its workspace limit size of cuDNN-conv interface is 8 * 1024 * 1024, which is smaller in PaddlePaddle and TensorFlow. Note that Caffe will be faster if increasing the workspace limit size.
- GoogletNet: input - 3 * 224 * 224, Time: ms/batch
| BatchSize | 64 | 128 | 256 |
|--------------|-------| -------| --------|
| PaddlePaddle | 613 | 1149 | 2348 |
| TensorFlow | 644 | 1176 | 2219 |
| Caffe | 694 | 1364 | out of memory |
- SmallNet: input - 3 * 32 * 32, Time ms/batch
| BatchSize | 64 | 128 | 256 | 512 |
|--------------|--------| -------- | --------|---------|
| PaddlePaddle | 10.463 | 18.184 | 33.113 | 63.039 |
| TensorFlow | 9 | 15 | 28 | 59 |
| Caffe | 9.373 | 16.6606 | 31.4797 | 59.719 |
**Notation**
All the single-GPU experiments in caffe use `caffe time` to calculate elapsed time, which does not include parameter updating time. However, both PaddlePaddle and TensorFlow experiments contain the parameter updating time. As compared with the total time, this part is relatively little on single machine, we can ignore it.
In Tensorflow, they implement algorithm searching method instead of using the algorithm searching interface in cuDNN.
### Multi-GPU: 4 GPUs
- AlexNet, ms / batch
| total-BatchSize | 128 * 4 | 256 * 4 |
|------------------|----------| -----------|
| PaddlePaddle | 347 | 622 |
| TensorFlow | 377 | 675 |
| Caffe | 1229 | 2435 |
For example, if `total-BatchSize = 128 * 4`, the speedup ratio is calculated by
```
time_at_1gpu_batch_128 * 4 / time_at_4gpu_total_batch_512
= (334 * 4)/347
= 3.85
```
<img src="figs/alexnet-4gpu.png" width="420">
- GoogleNet, ms / batch
| total-BatchSize | 128 * 4 | 256 * 4 |
|-------------------|--------------| ----------- |
| PaddlePaddle | 1178 | 2367 |
| TensorFlow | 1210 | 2292 |
| Caffe | 2007 | out of memory |
<img src="figs/googlenet-4gpu.png" width="420">
## RNN
We use lstm network for text classfication to test benchmark.
### Dataset
- [IMDB](http://www.iro.umontreal.ca/~lisa/deep/data/imdb.pkl)
- Sequence length is 100. In fact, PaddlePaddle supports training with variable-length sequence, but TensorFlow needs to pad. Thus, we also pad sequence length to 100 in PaddlePaddle in order to compare.
- Dictionary size=30000
- Peephole connection is used in `lstmemory` by default in PaddlePaddle. It is also configured in TensorFlow.
### Single-GPU
#### LSTM in Text Classification
Testing `2 lstm layer + fc` network with different hidden size and batch size.
- Batch size = 64, ms / batch
| hidden_size | 256 | 512 | 1280 |
|--------------|-------| -------| --------|
| PaddlePaddle | 83 | 184 | 641 |
| TensorFlow | 175 | 280 | 818 |
- Batch size = 128, ms / batch
| hidden_size | 256 | 512 | 1280 |
|--------------|------- | -------| --------|
| PaddlePaddle | 110 | 261 | 1007 |
| TensorFlow | 181 | 361 | 1237 |
- Batch size = 256, ms / batch
| hidden_size | 256 | 512 | 1280 |
|--------------|-------| -------| --------|
| PaddlePaddle | 170 | 414 | 1655 |
| TensorFlow | 238 | 536 | 1905 |
<img src="figs/rnn_lstm_cls.png" width="600">
#### Seq2Seq
The benchmark of sequence-to-sequence network will be added later.
### Multi GPU: 4 GPUs
#### LSTM in Text Classification
- hidden_size = 256, ms / batch
| batch_size | 256 | 512 |
|--------------| -------| --------|
| PaddlePaddle | 90 | 118 |
| TensorFlow | 226 | 118 |
- hidden_size = 512, ms / batch
| batch_size | 256 | 512 |
|--------------| -------| --------|
| PaddlePaddle | 189 | 268 |
| TensorFlow | 297 | 383 |
<img src="figs/rnn_lstm_4gpus.png" width="420">
#### Seq2Seq
The benchmark of sequence-to-sequence network will be added later.
benchmark/fluid/Dockerfile
浏览文件 @ a9ea99d7
...@@ -15,9 +15,6 @@ RUN ln -s /usr/lib/x86_64-linux-gnu/libcudnn.so.7 /usr/lib/libcudnn.so && ln -s ...@@ -15,9 +15,6 @@ RUN ln -s /usr/lib/x86_64-linux-gnu/libcudnn.so.7 /usr/lib/libcudnn.so && ln -s
RUN pip install -U pip RUN pip install -U pip
RUN pip install -U kubernetes paddlepaddle RUN pip install -U kubernetes paddlepaddle
RUN sh -c 'echo "import paddle.v2 as paddle\npaddle.dataset.cifar.train10()\npaddle.dataset.flowers.fetch()" | python'
RUN sh -c 'echo "import paddle.v2 as paddle\npaddle.dataset.mnist.train()\npaddle.dataset.mnist.test()\npaddle.dataset.imdb.fetch()" | python'
RUN sh -c 'echo "import paddle.v2 as paddle\npaddle.dataset.imikolov.fetch()" | python'
RUN pip uninstall -y paddlepaddle && mkdir /workspace RUN pip uninstall -y paddlepaddle && mkdir /workspace
ADD https://raw.githubusercontent.com/PaddlePaddle/cloud/develop/docker/paddle_k8s /usr/bin ADD https://raw.githubusercontent.com/PaddlePaddle/cloud/develop/docker/paddle_k8s /usr/bin
......
benchmark/paddle/image/alexnet.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from paddle.trainer_config_helpers import *
height = 227
width = 227
num_class = 1000
batch_size = get_config_arg('batch_size', int, 128)
gp = get_config_arg('layer_num', int, 1)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list" if not is_infer else None,
"test.list" if is_infer else None,
module="provider",
obj="process",
args=args)
settings(
batch_size=batch_size,
learning_rate=0.01 / batch_size,
learning_method=MomentumOptimizer(0.9),
regularization=L2Regularization(0.0005 * batch_size))
# conv1
net = data_layer('data', size=height * width * 3)
net = img_conv_layer(
input=net,
filter_size=11,
num_channels=3,
num_filters=96,
stride=4,
padding=1)
net = img_cmrnorm_layer(input=net, size=5, scale=0.0001, power=0.75)
net = img_pool_layer(input=net, pool_size=3, stride=2)
# conv2
net = img_conv_layer(
input=net, filter_size=5, num_filters=256, stride=1, padding=2, groups=gp)
net = img_cmrnorm_layer(input=net, size=5, scale=0.0001, power=0.75)
net = img_pool_layer(input=net, pool_size=3, stride=2)
# conv3
net = img_conv_layer(
input=net, filter_size=3, num_filters=384, stride=1, padding=1)
# conv4
net = img_conv_layer(
input=net, filter_size=3, num_filters=384, stride=1, padding=1, groups=gp)
# conv5
net = img_conv_layer(
input=net, filter_size=3, num_filters=256, stride=1, padding=1, groups=gp)
net = img_pool_layer(input=net, pool_size=3, stride=2)
net = fc_layer(
input=net,
size=4096,
act=ReluActivation(),
layer_attr=ExtraAttr(drop_rate=0.5))
net = fc_layer(
input=net,
size=4096,
act=ReluActivation(),
layer_attr=ExtraAttr(drop_rate=0.5))
net = fc_layer(input=net, size=1000, act=SoftmaxActivation())
if is_infer:
outputs(net)
else:
lab = data_layer('label', num_class)
loss = cross_entropy(input=net, label=lab)
outputs(loss)
benchmark/paddle/image/googlenet.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
#!/usr/bin/env python
from paddle.trainer_config_helpers import *
height = 224
width = 224
num_class = 1000
batch_size = get_config_arg('batch_size', int, 128)
use_gpu = get_config_arg('use_gpu', bool, True)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list" if not is_infer else None,
"test.list" if is_infer else None,
module="provider",
obj="process",
args=args)
settings(
batch_size=batch_size,
learning_rate=0.01 / batch_size,
learning_method=MomentumOptimizer(0.9),
regularization=L2Regularization(0.0005 * batch_size))
conv_projection = conv_projection if use_gpu else img_conv_layer
def inception2(name, input, channels, \
filter1,
filter3R, filter3,
filter5R, filter5,
proj):
conv1 = name + '_1'
conv3r = name + '_3r'
conv3 = name + '_3'
conv5r = name + '_5r'
conv5 = name + '_5'
maxpool = name + '_max'
convproj = name + '_proj'
cov1 = img_conv_layer(
name=conv1,
input=input,
filter_size=1,
num_channels=channels,
num_filters=filter1,
stride=1,
padding=0)
cov3r = img_conv_layer(
name=conv3r,
input=input,
filter_size=1,
num_channels=channels,
num_filters=filter3R,
stride=1,
padding=0)
cov3 = img_conv_layer(
name=conv3,
input=cov3r,
filter_size=3,
num_filters=filter3,
stride=1,
padding=1)
cov5r = img_conv_layer(
name=conv5r,
input=input,
filter_size=1,
num_channels=channels,
num_filters=filter5R,
stride=1,
padding=0)
cov5 = img_conv_layer(
name=conv5,
input=cov5r,
filter_size=5,
num_filters=filter5,
stride=1,
padding=2)
pool1 = img_pool_layer(
name=maxpool,
input=input,
pool_size=3,
num_channels=channels,
stride=1,
padding=1)
covprj = img_conv_layer(
name=convproj,
input=pool1,
filter_size=1,
num_filters=proj,
stride=1,
padding=0)
cat = concat_layer(name=name, input=[cov1, cov3, cov5, covprj])
return cat
def inception(name, input, channels, \
filter1,
filter3R, filter3,
filter5R, filter5,
proj):
cov1 = conv_projection(
input=input,
filter_size=1,
num_channels=channels,
num_filters=filter1,
stride=1,
padding=0)
cov3r = img_conv_layer(
name=name + '_3r',
input=input,
filter_size=1,
num_channels=channels,
num_filters=filter3R,
stride=1,
padding=0)
cov3 = conv_projection(
input=cov3r, filter_size=3, num_filters=filter3, stride=1, padding=1)
cov5r = img_conv_layer(
name=name + '_5r',
input=input,
filter_size=1,
num_channels=channels,
num_filters=filter5R,
stride=1,
padding=0)
cov5 = conv_projection(
input=cov5r, filter_size=5, num_filters=filter5, stride=1, padding=2)
pool1 = img_pool_layer(
name=name + '_max',
input=input,
pool_size=3,
num_channels=channels,
stride=1,
padding=1)
covprj = conv_projection(
input=pool1, filter_size=1, num_filters=proj, stride=1, padding=0)
cat = concat_layer(
name=name,
input=[cov1, cov3, cov5, covprj],
bias_attr=True if use_gpu else False,
act=ReluActivation())
return cat
data = data_layer(name="input", size=3 * height * width)
# stage 1
conv1 = img_conv_layer(
name="conv1",
input=data,
filter_size=7,
num_channels=3,
num_filters=64,
stride=2,
padding=3)
pool1 = img_pool_layer(
name="pool1", input=conv1, pool_size=3, num_channels=64, stride=2)
# stage 2
conv2_1 = img_conv_layer(
name="conv2_1",
input=pool1,
filter_size=1,
num_filters=64,
stride=1,
padding=0)
conv2_2 = img_conv_layer(
name="conv2_2",
input=conv2_1,
filter_size=3,
num_filters=192,
stride=1,
padding=1)
pool2 = img_pool_layer(
name="pool2", input=conv2_2, pool_size=3, num_channels=192, stride=2)
# stage 3
ince3a = inception("ince3a", pool2, 192, 64, 96, 128, 16, 32, 32)
ince3b = inception("ince3b", ince3a, 256, 128, 128, 192, 32, 96, 64)
pool3 = img_pool_layer(
name="pool3", input=ince3b, num_channels=480, pool_size=3, stride=2)
# stage 4
ince4a = inception("ince4a", pool3, 480, 192, 96, 208, 16, 48, 64)
ince4b = inception("ince4b", ince4a, 512, 160, 112, 224, 24, 64, 64)
ince4c = inception("ince4c", ince4b, 512, 128, 128, 256, 24, 64, 64)
ince4d = inception("ince4d", ince4c, 512, 112, 144, 288, 32, 64, 64)
ince4e = inception("ince4e", ince4d, 528, 256, 160, 320, 32, 128, 128)
pool4 = img_pool_layer(
name="pool4", input=ince4e, num_channels=832, pool_size=3, stride=2)
# stage 5
ince5a = inception("ince5a", pool4, 832, 256, 160, 320, 32, 128, 128)
ince5b = inception("ince5b", ince5a, 832, 384, 192, 384, 48, 128, 128)
pool5 = img_pool_layer(
name="pool5",
input=ince5b,
num_channels=1024,
pool_size=7,
stride=7,
pool_type=AvgPooling())
# We remove loss1 and loss2 for all system when testing benchmark
# output 1
# pool_o1 = img_pool_layer(name="pool_o1", input=ince4a, num_channels=512, pool_size=5, stride=3, pool_type=AvgPooling())
# conv_o1 = img_conv_layer(name="conv_o1", input=pool_o1, filter_size=1, num_filters=128, stride=1, padding=0)
# fc_o1 = fc_layer(name="fc_o1", input=conv_o1, size=1024, layer_attr=ExtraAttr(drop_rate=0.7), act=ReluActivation())
# out1 = fc_layer(name="output1", input=fc_o1, size=1000, act=SoftmaxActivation())
# loss1 = cross_entropy(name='loss1', input=out1, label=lab, coeff=0.3)
# output 2
#pool_o2 = img_pool_layer(name="pool_o2", input=ince4d, num_channels=528, pool_size=5, stride=3, pool_type=AvgPooling())
#conv_o2 = img_conv_layer(name="conv_o2", input=pool_o2, filter_size=1, num_filters=128, stride=1, padding=0)
#fc_o2 = fc_layer(name="fc_o2", input=conv_o2, size=1024, layer_attr=ExtraAttr(drop_rate=0.7), act=ReluActivation())
#out2 = fc_layer(name="output2", input=fc_o2, size=1000, act=SoftmaxActivation())
#loss2 = cross_entropy(name='loss2', input=out2, label=lab, coeff=0.3)
# output 3
dropout = dropout_layer(name="dropout", input=pool5, dropout_rate=0.4)
out3 = fc_layer(
name="output3", input=dropout, size=1000, act=SoftmaxActivation())
if is_infer:
outputs(out3)
else:
lab = data_layer(name="label", size=num_class)
loss3 = cross_entropy(name='loss3', input=out3, label=lab)
outputs(loss3)
benchmark/paddle/image/plotlog.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import sys
import argparse
import matplotlib.pyplot as plt
def parse_args():
parser = argparse.ArgumentParser('Parse Log')
parser.add_argument(
'--file_path', '-f', type=str, help='the path of the log file')
parser.add_argument(
'--sample_rate',
'-s',
type=float,
default=1.0,
help='the rate to take samples from log')
parser.add_argument(
'--log_period', '-p', type=int, default=1, help='the period of log')
args = parser.parse_args()
return args
def parse_file(file_name):
loss = []
error = []
with open(file_name) as f:
for i, line in enumerate(f):
line = line.strip()
if not line.startswith('pass'):
continue
line_split = line.split(' ')
if len(line_split) != 5:
continue
loss_str = line_split[2][:-1]
cur_loss = float(loss_str.split('=')[-1])
loss.append(cur_loss)
err_str = line_split[3][:-1]
cur_err = float(err_str.split('=')[-1])
error.append(cur_err)
accuracy = [1.0 - err for err in error]
return loss, accuracy
def sample(metric, sample_rate):
interval = int(1.0 / sample_rate)
if interval > len(metric):
return metric[:1]
num = len(metric) / interval
idx = [interval * i for i in range(num)]
metric_sample = [metric[id] for id in idx]
return metric_sample
def plot_metric(metric,
batch_id,
graph_title,
line_style='b-',
line_label='y',
line_num=1):
plt.figure()
plt.title(graph_title)
if line_num == 1:
plt.plot(batch_id, metric, line_style, label=line_label)
else:
for i in range(line_num):
plt.plot(batch_id, metric[i], line_style[i], label=line_label[i])
plt.xlabel('batch')
plt.ylabel(graph_title)
plt.legend()
plt.savefig(graph_title + '.jpg')
plt.close()
def main():
args = parse_args()
assert args.sample_rate > 0. and args.sample_rate <= 1.0, "The sample rate should in the range (0, 1]."
loss, accuracy = parse_file(args.file_path)
batch = [args.log_period * i for i in range(len(loss))]
batch_sample = sample(batch, args.sample_rate)
loss_sample = sample(loss, args.sample_rate)
accuracy_sample = sample(accuracy, args.sample_rate)
plot_metric(loss_sample, batch_sample, 'loss', line_label='loss')
plot_metric(
accuracy_sample,
batch_sample,
'accuracy',
line_style='g-',
line_label='accuracy')
if __name__ == '__main__':
main()
benchmark/paddle/image/provider.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import io, os
import random
import numpy as np
from paddle.trainer.PyDataProvider2 import *
def initHook(settings, height, width, color, num_class, **kwargs):
settings.height = height
settings.width = width
settings.color = color
settings.num_class = num_class
if settings.color:
settings.data_size = settings.height * settings.width * 3
else:
settings.data_size = settings.height * settings.width
settings.is_infer = kwargs.get('is_infer', False)
settings.num_samples = kwargs.get('num_samples', 2560)
if settings.is_infer:
settings.slots = [dense_vector(settings.data_size)]
else:
settings.slots = [dense_vector(settings.data_size), integer_value(1)]
@provider(
init_hook=initHook, min_pool_size=-1, cache=CacheType.CACHE_PASS_IN_MEM)
def process(settings, file_list):
for i in xrange(settings.num_samples):
img = np.random.rand(1, settings.data_size).reshape(-1, 1).flatten()
if settings.is_infer:
yield img.astype('float32')
else:
lab = random.randint(0, settings.num_class - 1)
yield img.astype('float32'), int(lab)
benchmark/paddle/image/resnet.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
#!/usr/bin/env python
from paddle.trainer_config_helpers import *
height = 224
width = 224
num_class = 1000
batch_size = get_config_arg('batch_size', int, 64)
layer_num = get_config_arg("layer_num", int, 50)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list" if not is_infer else None,
"test.list" if is_infer else None,
module="provider",
obj="process",
args=args)
settings(
batch_size=batch_size,
learning_rate=0.01 / batch_size,
learning_method=MomentumOptimizer(0.9),
regularization=L2Regularization(0.0005 * batch_size))
#######################Network Configuration #############
def conv_bn_layer(name,
input,
filter_size,
num_filters,
stride,
padding,
channels=None,
active_type=ReluActivation()):
"""
A wrapper for conv layer with batch normalization layers.
Note:
conv layer has no activation.
"""
tmp = img_conv_layer(
name=name + "_conv",
input=input,
filter_size=filter_size,
num_channels=channels,
num_filters=num_filters,
stride=stride,
padding=padding,
act=LinearActivation(),
bias_attr=False)
return batch_norm_layer(
name=name + "_bn",
input=tmp,
act=active_type,
use_global_stats=is_infer)
def bottleneck_block(name, input, num_filters1, num_filters2):
"""
A wrapper for bottlenect building block in ResNet.
Last conv_bn_layer has no activation.
Addto layer has activation of relu.
"""
last_name = conv_bn_layer(
name=name + '_branch2a',
input=input,
filter_size=1,
num_filters=num_filters1,
stride=1,
padding=0)
last_name = conv_bn_layer(
name=name + '_branch2b',
input=last_name,
filter_size=3,
num_filters=num_filters1,
stride=1,
padding=1)
last_name = conv_bn_layer(
name=name + '_branch2c',
input=last_name,
filter_size=1,
num_filters=num_filters2,
stride=1,
padding=0,
active_type=LinearActivation())
return addto_layer(
name=name + "_addto", input=[input, last_name], act=ReluActivation())
def mid_projection(name, input, num_filters1, num_filters2, stride=2):
"""
A wrapper for middile projection in ResNet.
projection shortcuts are used for increasing dimensions,
and other shortcuts are identity
branch1: projection shortcuts are used for increasing
dimensions, has no activation.
branch2x: bottleneck building block, shortcuts are identity.
"""
# stride = 2
branch1 = conv_bn_layer(
name=name + '_branch1',
input=input,
filter_size=1,
num_filters=num_filters2,
stride=stride,
padding=0,
active_type=LinearActivation())
last_name = conv_bn_layer(
name=name + '_branch2a',
input=input,
filter_size=1,
num_filters=num_filters1,
stride=stride,
padding=0)
last_name = conv_bn_layer(
name=name + '_branch2b',
input=last_name,
filter_size=3,
num_filters=num_filters1,
stride=1,
padding=1)
last_name = conv_bn_layer(
name=name + '_branch2c',
input=last_name,
filter_size=1,
num_filters=num_filters2,
stride=1,
padding=0,
active_type=LinearActivation())
return addto_layer(
name=name + "_addto", input=[branch1, last_name], act=ReluActivation())
img = data_layer(name='image', size=height * width * 3)
def deep_res_net(res2_num=3, res3_num=4, res4_num=6, res5_num=3):
"""
A wrapper for 50,101,152 layers of ResNet.
res2_num: number of blocks stacked in conv2_x
res3_num: number of blocks stacked in conv3_x
res4_num: number of blocks stacked in conv4_x
res5_num: number of blocks stacked in conv5_x
"""
# For ImageNet
# conv1: 112x112
tmp = conv_bn_layer(
"conv1",
input=img,
filter_size=7,
channels=3,
num_filters=64,
stride=2,
padding=3)
tmp = img_pool_layer(name="pool1", input=tmp, pool_size=3, stride=2)
# conv2_x: 56x56
tmp = mid_projection(
name="res2_1", input=tmp, num_filters1=64, num_filters2=256, stride=1)
for i in xrange(2, res2_num + 1, 1):
tmp = bottleneck_block(
name="res2_" + str(i), input=tmp, num_filters1=64, num_filters2=256)
# conv3_x: 28x28
tmp = mid_projection(
name="res3_1", input=tmp, num_filters1=128, num_filters2=512)
for i in xrange(2, res3_num + 1, 1):
tmp = bottleneck_block(
name="res3_" + str(i),
input=tmp,
num_filters1=128,
num_filters2=512)
# conv4_x: 14x14
tmp = mid_projection(
name="res4_1", input=tmp, num_filters1=256, num_filters2=1024)
for i in xrange(2, res4_num + 1, 1):
tmp = bottleneck_block(
name="res4_" + str(i),
input=tmp,
num_filters1=256,
num_filters2=1024)
# conv5_x: 7x7
tmp = mid_projection(
name="res5_1", input=tmp, num_filters1=512, num_filters2=2048)
for i in xrange(2, res5_num + 1, 1):
tmp = bottleneck_block(
name="res5_" + str(i),
input=tmp,
num_filters1=512,
num_filters2=2048)
tmp = img_pool_layer(
name='avgpool',
input=tmp,
pool_size=7,
stride=1,
pool_type=AvgPooling())
return fc_layer(input=tmp, size=num_class, act=SoftmaxActivation())
if layer_num == 50:
resnet = deep_res_net(3, 4, 6, 3)
elif layer_num == 101:
resnet = deep_res_net(3, 4, 23, 3)
elif layer_num == 152:
resnet = deep_res_net(3, 8, 36, 3)
else:
print("Wrong layer number.")
if is_infer:
outputs(resnet)
else:
lbl = data_layer(name="label", size=num_class)
loss = cross_entropy(name='loss', input=resnet, label=lbl)
outputs(loss)
benchmark/paddle/image/run.sh 已删除 100755 → 0
浏览文件 @ eac5a0aa
#!/bin/bash
set -e
function train() {
cfg=$1
thread=$2
bz=$3
args="batch_size=$3"
prefix=$4
paddle train --job=time \
--config=$cfg \
--use_gpu=True \
--trainer_count=$thread \
--log_period=10 \
--test_period=100 \
--config_args=$args \
> logs/$prefix-${thread}gpu-$bz.log 2>&1
}
if [ ! -d "train.list" ]; then
echo " " > train.list
fi
if [ ! -d "logs" ]; then
mkdir logs
fi
#========single-gpu=========#
# alexnet
train alexnet.py 1 64 alexnet
train alexnet.py 1 128 alexnet
train alexnet.py 1 256 alexnet
train alexnet.py 1 512 alexnet
# googlenet
train googlenet.py 1 64 googlenet
train googlenet.py 1 128 googlenet
train googlenet.py 1 256 googlenet
# smallnet
train smallnet_mnist_cifar.py 1 64 smallnet
train smallnet_mnist_cifar.py 1 128 smallnet
train smallnet_mnist_cifar.py 1 256 smallnet
train smallnet_mnist_cifar.py 1 512 smallnet
############################
#========multi-gpus=========#
train alexnet.py 4 512 alexnet
train alexnet.py 4 1024 alexnet
train googlenet.py 4 512 googlenet
train googlenet.py 4 1024 googlenet
benchmark/paddle/image/run_mkl_infer.sh 已删除 100755 → 0
浏览文件 @ eac5a0aa
#!/bin/bash
set -e
function clock_to_seconds() {
hours=`echo $1 | awk -F ':' '{print $1}'`
mins=`echo $1 | awk -F ':' '{print $2}'`
secs=`echo $1 | awk -F ':' '{print $3}'`
echo `awk 'BEGIN{printf "%.2f",('$secs' + '$mins' * 60 + '$hours' * 3600)}'`
}
function infer() {
unset OMP_NUM_THREADS MKL_NUM_THREADS OMP_DYNAMIC KMP_AFFINITY
topology=$1
layer_num=$2
bs=$3
use_mkldnn=$4
if [ $4 == "True" ]; then
thread=1
log="logs/infer-${topology}-${layer_num}-mkldnn-${bs}.log"
elif [ $4 == "False" ]; then
thread=`nproc`
if [ $thread -gt $bs ]; then
thread=$bs
fi
log="logs/infer-${topology}-${layer_num}-${thread}mklml-${bs}.log"
else
echo "Wrong input $4, use True or False."
exit 0
fi
models_in="models/${topology}-${layer_num}/pass-00000/"
if [ ! -d $models_in ]; then
echo "Training model ${topology}_${layer_num}"
paddle train --job=train \
--config="${topology}.py" \
--use_mkldnn=True \
--use_gpu=False \
--trainer_count=1 \
--num_passes=1 \
--save_dir="models/${topology}-${layer_num}" \
--config_args="batch_size=128,layer_num=${layer_num},num_samples=256" \
> /dev/null 2>&1
echo "Done"
fi
log_period=$((256 / bs))
paddle train --job=test \
--config="${topology}.py" \
--use_mkldnn=$use_mkldnn \
--use_gpu=False \
--trainer_count=$thread \
--log_period=$log_period \
--config_args="batch_size=${bs},layer_num=${layer_num},is_infer=True" \
--init_model_path=$models_in \
2>&1 | tee ${log}
# calculate the last 5 logs period time of 1280 samples,
# the time before are burning time.
start=`tail ${log} -n 7 | head -n 1 | awk -F ' ' '{print $2}' | xargs`
end=`tail ${log} -n 2 | head -n 1 | awk -F ' ' '{print $2}' | xargs`
start_sec=`clock_to_seconds $start`
end_sec=`clock_to_seconds $end`
fps=`awk 'BEGIN{printf "%.2f",(1280 / ('$end_sec' - '$start_sec'))}'`
echo "Last 1280 samples start: ${start}(${start_sec} sec), end: ${end}(${end_sec} sec;" >> ${log}
echo "FPS: $fps images/sec" 2>&1 | tee -a ${log}
}
if [ ! -f "train.list" ]; then
echo " " > train.list
fi
if [ ! -f "test.list" ]; then
echo " " > test.list
fi
if [ ! -d "logs" ]; then
mkdir logs
fi
if [ ! -d "models" ]; then
mkdir -p models
fi
# inference benchmark
for use_mkldnn in True False; do
for batchsize in 1 2 4 8 16; do
infer vgg 19 $batchsize $use_mkldnn
infer resnet 50 $batchsize $use_mkldnn
infer googlenet v1 $batchsize $use_mkldnn
infer alexnet 2 $batchsize $use_mkldnn
done
done
benchmark/paddle/image/run_mkl_train.sh 已删除 100755 → 0
浏览文件 @ eac5a0aa
#!/bin/bash
set -e
function train() {
unset OMP_NUM_THREADS MKL_NUM_THREADS OMP_DYNAMIC KMP_AFFINITY
topology=$1
layer_num=$2
bs=$3
use_mkldnn=$4
if [ $4 == "True" ]; then
thread=1
log="logs/train-${topology}-${layer_num}-mkldnn-${bs}.log"
elif [ $4 == "False" ]; then
thread=`nproc`
# each trainer_count use only 1 core to avoid conflict
log="logs/train-${topology}-${layer_num}-${thread}mklml-${bs}.log"
else
echo "Wrong input $4, use True or False."
exit 0
fi
args="batch_size=${bs},layer_num=${layer_num}"
config="${topology}.py"
paddle train --job=time \
--config=$config \
--use_mkldnn=$use_mkldnn \
--use_gpu=False \
--trainer_count=$thread \
--log_period=10 \
--test_period=100 \
--config_args=$args \
2>&1 | tee ${log}
avg_time=`tail ${log} -n 1 | awk -F ' ' '{print $8}' | sed 's/avg=//'`
fps=`awk 'BEGIN{printf "%.2f",('$bs' / '$avg_time' * 1000)}'`
echo "FPS: $fps images/sec" 2>&1 | tee -a ${log}
}
if [ ! -f "train.list" ]; then
echo " " > train.list
fi
if [ ! -d "logs" ]; then
mkdir logs
fi
# training benchmark
for use_mkldnn in True False; do
for batchsize in 64 128 256; do
train vgg 19 $batchsize $use_mkldnn
train resnet 50 $batchsize $use_mkldnn
train googlenet v1 $batchsize $use_mkldnn
train alexnet 2 $batchsize $use_mkldnn
done
done
benchmark/paddle/image/run_openblas_infer.sh 已删除 100755 → 0
浏览文件 @ eac5a0aa
#!/bin/bash
set -e
function clock_to_seconds() {
hours=`echo $1 | awk -F ':' '{print $1}'`
mins=`echo $1 | awk -F ':' '{print $2}'`
secs=`echo $1 | awk -F ':' '{print $3}'`
echo `awk 'BEGIN{printf "%.2f",('$secs' + '$mins' * 60 + '$hours' * 3600)}'`
}
function infer() {
export OPENBLAS_MAIN_FREE=1
topology=$1
layer_num=$2
bs=$3
trainers=`nproc`
if [ $trainers -gt $bs ]; then
trainers=$bs
fi
log="logs/infer-${topology}-${layer_num}-${trainers}openblas-${bs}.log"
threads=$((`nproc` / trainers))
if [ $threads -eq 0 ]; then
threads=1
fi
export OPENBLAS_NUM_THREADS=$threads
models_in="models/${topology}-${layer_num}/pass-00000/"
if [ ! -d $models_in ]; then
echo "./run_mkl_infer.sh to save the model first"
exit 0
fi
log_period=$((32 / bs))
paddle train --job=test \
--config="${topology}.py" \
--use_mkldnn=False \
--use_gpu=False \
--trainer_count=$trainers \
--log_period=$log_period \
--config_args="batch_size=${bs},layer_num=${layer_num},is_infer=True,num_samples=256" \
--init_model_path=$models_in \
2>&1 | tee ${log}
# calculate the last 5 logs period time of 160(=32*5) samples,
# the time before are burning time.
start=`tail ${log} -n 7 | head -n 1 | awk -F ' ' '{print $2}' | xargs`
end=`tail ${log} -n 2 | head -n 1 | awk -F ' ' '{print $2}' | xargs`
start_sec=`clock_to_seconds $start`
end_sec=`clock_to_seconds $end`
fps=`awk 'BEGIN{printf "%.2f",(160 / ('$end_sec' - '$start_sec'))}'`
echo "Last 160 samples start: ${start}(${start_sec} sec), end: ${end}(${end_sec} sec;" >> ${log}
echo "FPS: $fps images/sec" 2>&1 | tee -a ${log}
}
if [ ! -f "train.list" ]; then
echo " " > train.list
fi
if [ ! -f "test.list" ]; then
echo " " > test.list
fi
if [ ! -d "logs" ]; then
mkdir logs
fi
# inference benchmark
for batchsize in 1 2 4 8 16; do
infer vgg 19 $batchsize
infer resnet 50 $batchsize
infer googlenet v1 $batchsize
infer alexnet 2 $batchsize
done
benchmark/paddle/image/run_openblas_train.sh 已删除 100755 → 0
浏览文件 @ eac5a0aa
#!/bin/bash
set -e
function train() {
export OPENBLAS_NUM_THREADS=1
topology=$1
layer_num=$2
bs=$3
thread=`nproc`
# each trainer_count use only 1 core to avoid conflict
log="logs/train-${topology}-${layer_num}-${thread}openblas-${bs}.log"
args="batch_size=${bs},layer_num=${layer_num}"
config="${topology}.py"
paddle train --job=time \
--config=$config \
--use_mkldnn=False \
--use_gpu=False \
--trainer_count=$thread \
--log_period=3 \
--test_period=30 \
--config_args=$args \
2>&1 | tee ${log}
avg_time=`tail ${log} -n 1 | awk -F ' ' '{print $8}' | sed 's/avg=//'`
fps=`awk 'BEGIN{printf "%.2f",('$bs' / '$avg_time' * 1000)}'`
echo "FPS: $fps images/sec" 2>&1 | tee -a ${log}
}
if [ ! -f "train.list" ]; then
echo " " > train.list
fi
if [ ! -d "logs" ]; then
mkdir logs
fi
# training benchmark
for batchsize in 64 128 256; do
train vgg 19 $batchsize
train resnet 50 $batchsize
train googlenet v1 $batchsize
train alexnet 2 $batchsize
done
benchmark/paddle/image/smallnet_mnist_cifar.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
#!/usr/bin/env python
from paddle.trainer_config_helpers import *
height = 32
width = 32
num_class = 10
batch_size = get_config_arg('batch_size', int, 128)
args = {'height': height, 'width': width, 'color': True, 'num_class': num_class}
define_py_data_sources2(
"train.list", None, module="provider", obj="process", args=args)
settings(
batch_size=batch_size,
learning_rate=0.01 / batch_size,
learning_method=MomentumOptimizer(0.9),
regularization=L2Regularization(0.0005 * batch_size))
# conv1
net = data_layer('data', size=height * width * 3)
net = img_conv_layer(
input=net,
filter_size=5,
num_channels=3,
num_filters=32,
stride=1,
padding=2)
net = img_pool_layer(input=net, pool_size=3, stride=2, padding=1)
# conv2
net = img_conv_layer(
input=net, filter_size=5, num_filters=32, stride=1, padding=2)
net = img_pool_layer(
input=net, pool_size=3, stride=2, padding=1, pool_type=AvgPooling())
# conv3
net = img_conv_layer(
input=net, filter_size=3, num_filters=64, stride=1, padding=1)
net = img_pool_layer(
input=net, pool_size=3, stride=2, padding=1, pool_type=AvgPooling())
net = fc_layer(input=net, size=64, act=ReluActivation())
net = fc_layer(input=net, size=10, act=SoftmaxActivation())
lab = data_layer('label', num_class)
loss = classification_cost(input=net, label=lab)
outputs(loss)
benchmark/paddle/image/vgg.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
#!/usr/bin/env python
from paddle.trainer_config_helpers import *
height = 224
width = 224
num_class = 1000
batch_size = get_config_arg('batch_size', int, 64)
layer_num = get_config_arg('layer_num', int, 19)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list" if not is_infer else None,
"test.list" if is_infer else None,
module="provider",
obj="process",
args=args)
settings(
batch_size=batch_size,
learning_rate=0.001 / batch_size,
learning_method=MomentumOptimizer(0.9),
regularization=L2Regularization(0.0005 * batch_size))
img = data_layer(name='image', size=height * width * 3)
def vgg_network(vgg_num=3):
tmp = img_conv_group(
input=img,
num_channels=3,
conv_padding=1,
conv_num_filter=[64, 64],
conv_filter_size=3,
conv_act=ReluActivation(),
pool_size=2,
pool_stride=2,
pool_type=MaxPooling())
tmp = img_conv_group(
input=tmp,
conv_num_filter=[128, 128],
conv_padding=1,
conv_filter_size=3,
conv_act=ReluActivation(),
pool_stride=2,
pool_type=MaxPooling(),
pool_size=2)
channels = []
for i in range(vgg_num):
channels.append(256)
tmp = img_conv_group(
input=tmp,
conv_num_filter=channels,
conv_padding=1,
conv_filter_size=3,
conv_act=ReluActivation(),
pool_stride=2,
pool_type=MaxPooling(),
pool_size=2)
channels = []
for i in range(vgg_num):
channels.append(512)
tmp = img_conv_group(
input=tmp,
conv_num_filter=channels,
conv_padding=1,
conv_filter_size=3,
conv_act=ReluActivation(),
pool_stride=2,
pool_type=MaxPooling(),
pool_size=2)
tmp = img_conv_group(
input=tmp,
conv_num_filter=channels,
conv_padding=1,
conv_filter_size=3,
conv_act=ReluActivation(),
pool_stride=2,
pool_type=MaxPooling(),
pool_size=2)
tmp = fc_layer(
input=tmp,
size=4096,
act=ReluActivation(),
layer_attr=ExtraAttr(drop_rate=0.5))
tmp = fc_layer(
input=tmp,
size=4096,
act=ReluActivation(),
layer_attr=ExtraAttr(drop_rate=0.5))
return fc_layer(input=tmp, size=num_class, act=SoftmaxActivation())
if layer_num == 16:
vgg = vgg_network(3)
elif layer_num == 19:
vgg = vgg_network(4)
else:
print("Wrong layer number.")
if is_infer:
outputs(vgg)
else:
lab = data_layer('label', num_class)
loss = cross_entropy(input=vgg, label=lab)
outputs(loss)
benchmark/paddle/rnn/imdb.py 已删除 100755 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import print_function
import six.moves.cPickle as pickle
import gzip
import os
import numpy
def get_dataset_file(dataset, default_dataset, origin):
data_dir, data_file = os.path.split(dataset)
if (not os.path.isfile(dataset)) and data_file == default_dataset:
from six.moves import urllib
print('Downloading data from %s' % origin)
urllib.request.urlretrieve(origin, dataset)
return dataset
def create_data(path="imdb.pkl"):
if (not os.path.isfile('imdb.train.pkl')):
path = get_dataset_file(
path, "imdb.pkl",
"http://www.iro.umontreal.ca/~lisa/deep/data/imdb.pkl")
if path.endswith(".gz"):
f = gzip.open(path, 'rb')
else:
f = open(path, 'rb')
train_set = pickle.load(f)
test_set = pickle.load(f)
f.close()
pickle.dump(train_set, open('imdb.train.pkl', 'wb'))
pickle.dump(test_set, open('imdb.test.pkl', 'wb'))
if (not os.path.isfile('train.list')):
file('train.list', 'w').write('imdb.train.pkl\n')
def main():
create_data('imdb.pkl')
if __name__ == "__main__":
main()
benchmark/paddle/rnn/provider.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import io, os
import random
import numpy as np
import six.moves.cPickle as pickle
from paddle.trainer.PyDataProvider2 import *
def remove_unk(x, n_words):
return [[1 if w >= n_words else w for w in sen] for sen in x]
# ==============================================================
# tensorflow uses fixed length, but PaddlePaddle can process
# variable-length. Padding is used in benchmark in order to
# compare with other platform.
# ==============================================================
def pad_sequences(sequences,
maxlen=None,
dtype='int32',
padding='post',
truncating='post',
value=0.):
lengths = [len(s) for s in sequences]
nb_samples = len(sequences)
if maxlen is None:
maxlen = np.max(lengths)
x = (np.ones((nb_samples, maxlen)) * value).astype(dtype)
for idx, s in enumerate(sequences):
if len(s) == 0:
continue # empty list was found
if truncating == 'pre':
trunc = s[-maxlen:]
elif truncating == 'post':
trunc = s[:maxlen]
else:
raise ValueError("Truncating type '%s' not understood" % padding)
if padding == 'post':
x[idx, :len(trunc)] = trunc
elif padding == 'pre':
x[idx, -len(trunc):] = trunc
else:
raise ValueError("Padding type '%s' not understood" % padding)
return x
def initHook(settings, vocab_size, pad_seq, maxlen, **kwargs):
settings.vocab_size = vocab_size
settings.pad_seq = pad_seq
settings.maxlen = maxlen
settings.input_types = [
integer_value_sequence(vocab_size), integer_value(2)
]
@provider(
init_hook=initHook, min_pool_size=-1, cache=CacheType.CACHE_PASS_IN_MEM)
def process(settings, file):
f = open(file, 'rb')
train_set = pickle.load(f)
f.close()
x, y = train_set
# remove unk, namely remove the words out of dictionary
x = remove_unk(x, settings.vocab_size)
if settings.pad_seq:
x = pad_sequences(x, maxlen=settings.maxlen, value=0.)
for i in range(len(y)):
yield map(int, x[i]), int(y[i])
benchmark/paddle/rnn/rnn.py 已删除 100755 → 0
浏览文件 @ eac5a0aa
#!/usr/bin/env python
from paddle.trainer_config_helpers import *
import imdb
num_class = 2
vocab_size = 30000
fixedlen = 100
batch_size = get_config_arg('batch_size', int, 128)
lstm_num = get_config_arg('lstm_num', int, 1)
hidden_size = get_config_arg('hidden_size', int, 128)
# whether to pad sequence into fixed length
pad_seq = get_config_arg('pad_seq', bool, True)
imdb.create_data('imdb.pkl')
args = {'vocab_size': vocab_size, 'pad_seq': pad_seq, 'maxlen': fixedlen}
define_py_data_sources2(
"train.list", None, module="provider", obj="process", args=args)
settings(
batch_size=batch_size,
learning_rate=2e-3,
learning_method=AdamOptimizer(),
regularization=L2Regularization(8e-4),
gradient_clipping_threshold=25)
net = data_layer('data', size=vocab_size)
net = embedding_layer(input=net, size=128)
for i in xrange(lstm_num):
net = simple_lstm(input=net, size=hidden_size)
net = last_seq(input=net)
net = fc_layer(input=net, size=2, act=SoftmaxActivation())
lab = data_layer('label', num_class)
loss = classification_cost(input=net, label=lab)
outputs(loss)
benchmark/paddle/rnn/run.sh 已删除 100755 → 0
浏览文件 @ eac5a0aa
#!/bin/bash
set -e
function train() {
cfg=$1
thread=$2
args="lstm_num=${3},seq_pad=${4},hidden_size=${5},batch_size=${6}"
paddle train --job=time \
--config=$cfg \
--use_gpu=1 \
--trainer_count=$thread \
--log_period=10 \
--test_period=100 \
--num_passes=1 \
--feed_data=1 \
--config_args=$args \
>logs/rnn-pad${4}-${thread}gpu-lstm${3}-batch${6}-hid${5}.log 2>&1
}
if [ ! -d "logs" ]; then
mkdir logs
fi
## padding, single gpu
#-----config--gpu--lstm_num--padding--hidden_size--batch_size
## lstm_num=2, batch_size=64
train rnn.py 1 2 1 256 64
train rnn.py 1 2 1 512 64
train rnn.py 1 2 1 1280 64
## lstm_num=2, batch_size=128
train rnn.py 1 2 1 256 128
train rnn.py 1 2 1 512 128
train rnn.py 1 2 1 1280 128
## lstm_num=4, batch_size=256
train rnn.py 1 2 1 256 256
train rnn.py 1 2 1 512 256
train rnn.py 1 2 1 1280 256
#==================multi gpus=====================#
# hidden_size=256, lstm_num=2, different batch size
train rnn.py 4 2 1 256 128
train rnn.py 4 2 1 256 256
train rnn.py 4 2 1 256 512
# hidden_size=512, lstm_num=4, different batch size
train rnn.py 4 2 1 512 128
train rnn.py 4 2 1 512 256
train rnn.py 4 2 1 512 512
benchmark/tensorflow/machine_translation.py
浏览文件 @ a9ea99d7
...@@ -35,8 +35,6 @@ import os ...@@ -35,8 +35,6 @@ import os
import argparse import argparse
import time import time
import paddle.v2 as paddle
parser = argparse.ArgumentParser(description=__doc__) parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument( parser.add_argument(
"--embedding_dim", "--embedding_dim",
......
benchmark/tensorflow/mnist.py
浏览文件 @ a9ea99d7
...@@ -21,7 +21,6 @@ import time ...@@ -21,7 +21,6 @@ import time
import numpy as np import numpy as np
import tensorflow as tf import tensorflow as tf
import paddle.v2 as paddle
DTYPE = tf.float32 DTYPE = tf.float32
......
benchmark/tensorflow/resnet.py
浏览文件 @ a9ea99d7
...@@ -27,7 +27,6 @@ import argparse ...@@ -27,7 +27,6 @@ import argparse
import time import time
import numpy as np import numpy as np
import paddle.v2 as paddle
import tensorflow as tf import tensorflow as tf
DTYPE = tf.float32 DTYPE = tf.float32
......
benchmark/tensorflow/stacked_dynamic_lstm.py
浏览文件 @ a9ea99d7
...@@ -21,8 +21,6 @@ import argparse ...@@ -21,8 +21,6 @@ import argparse
import time import time
import tensorflow as tf import tensorflow as tf
import paddle.v2 as paddle
def parse_args(): def parse_args():
parser = argparse.ArgumentParser("LSTM model benchmark.") parser = argparse.ArgumentParser("LSTM model benchmark.")
......
benchmark/tensorflow/vgg.py
浏览文件 @ a9ea99d7
...@@ -13,7 +13,6 @@ ...@@ -13,7 +13,6 @@
# limitations under the License. # limitations under the License.
"""VGG16 benchmark in TensorFlow""" """VGG16 benchmark in TensorFlow"""
import tensorflow as tf import tensorflow as tf
import paddle.v2 as paddle
import numpy as np import numpy as np
import argparse import argparse
import time import time
......
cmake/FindSphinx.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# - This module looks for Sphinx
# Find the Sphinx documentation generator
#
# This modules defines
# SPHINX_EXECUTABLE
# SPHINX_FOUND
find_program(SPHINX_EXECUTABLE
NAMES sphinx-build
PATHS
/usr/bin
/usr/local/bin
/opt/local/bin
DOC "Sphinx documentation generator"
)
if( NOT SPHINX_EXECUTABLE )
set(_Python_VERSIONS
2.7 2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.6 1.5
)
foreach( _version ${_Python_VERSIONS} )
set( _sphinx_NAMES sphinx-build-${_version} )
find_program( SPHINX_EXECUTABLE
NAMES ${_sphinx_NAMES}
PATHS
/usr/bin
/usr/local/bin
/opt/loca/bin
DOC "Sphinx documentation generator"
)
endforeach()
endif()
include(FindPackageHandleStandardArgs)
find_package_handle_standard_args(Sphinx DEFAULT_MSG
SPHINX_EXECUTABLE
)
option( SPHINX_HTML_OUTPUT "Build a single HTML with the whole content." ON )
option( SPHINX_DIRHTML_OUTPUT "Build HTML pages, but with a single directory per document." OFF )
option( SPHINX_HTMLHELP_OUTPUT "Build HTML pages with additional information for building a documentation collection in htmlhelp." OFF )
option( SPHINX_QTHELP_OUTPUT "Build HTML pages with additional information for building a documentation collection in qthelp." OFF )
option( SPHINX_DEVHELP_OUTPUT "Build HTML pages with additional information for building a documentation collection in devhelp." OFF )
option( SPHINX_EPUB_OUTPUT "Build HTML pages with additional information for building a documentation collection in epub." OFF )
option( SPHINX_LATEX_OUTPUT "Build LaTeX sources that can be compiled to a PDF document using pdflatex." OFF )
option( SPHINX_MAN_OUTPUT "Build manual pages in groff format for UNIX systems." OFF )
option( SPHINX_TEXT_OUTPUT "Build plain text files." OFF )
mark_as_advanced(
SPHINX_EXECUTABLE
SPHINX_HTML_OUTPUT
SPHINX_DIRHTML_OUTPUT
SPHINX_HTMLHELP_OUTPUT
SPHINX_QTHELP_OUTPUT
SPHINX_DEVHELP_OUTPUT
SPHINX_EPUB_OUTPUT
SPHINX_LATEX_OUTPUT
SPHINX_MAN_OUTPUT
SPHINX_TEXT_OUTPUT
)
function( Sphinx_add_target target_name builder conf cache source destination )
add_custom_target( ${target_name} ALL
COMMAND ${SPHINX_EXECUTABLE} -b ${builder}
-d ${cache}
-c ${conf}
${source}
${destination}
COMMENT "Generating sphinx documentation: ${builder}"
COMMAND cd ${destination} && ln -sf ./index_*.html index.html
)
set_property(
DIRECTORY APPEND PROPERTY
ADDITIONAL_MAKE_CLEAN_FILES
${destination}
)
endfunction()
# Target dependencies can be optionally listed at the end.
function( Sphinx_add_targets target_base_name conf source base_destination )
set( _dependencies )
foreach( arg IN LISTS ARGN )
set( _dependencies ${_dependencies} ${arg} )
endforeach()
if( ${SPHINX_HTML_OUTPUT} )
Sphinx_add_target( ${target_base_name}_html html ${conf} ${source} ${base_destination}/html )
add_dependencies( ${target_base_name}_html ${_dependencies} )
endif()
if( ${SPHINX_DIRHTML_OUTPUT} )
Sphinx_add_target( ${target_base_name}_dirhtml dirhtml ${conf} ${source} ${base_destination}/dirhtml )
add_dependencies( ${target_base_name}_dirhtml ${_dependencies} )
endif()
if( ${SPHINX_QTHELP_OUTPUT} )
Sphinx_add_target( ${target_base_name}_qthelp qthelp ${conf} ${source} ${base_destination}/qthelp )
add_dependencies( ${target_base_name}_qthelp ${_dependencies} )
endif()
if( ${SPHINX_DEVHELP_OUTPUT} )
Sphinx_add_target( ${target_base_name}_devhelp devhelp ${conf} ${source} ${base_destination}/devhelp )
add_dependencies( ${target_base_name}_devhelp ${_dependencies} )
endif()
if( ${SPHINX_EPUB_OUTPUT} )
Sphinx_add_target( ${target_base_name}_epub epub ${conf} ${source} ${base_destination}/epub )
add_dependencies( ${target_base_name}_epub ${_dependencies} )
endif()
if( ${SPHINX_LATEX_OUTPUT} )
Sphinx_add_target( ${target_base_name}_latex latex ${conf} ${source} ${base_destination}/latex )
add_dependencies( ${target_base_name}_latex ${_dependencies} )
endif()
if( ${SPHINX_MAN_OUTPUT} )
Sphinx_add_target( ${target_base_name}_man man ${conf} ${source} ${base_destination}/man )
add_dependencies( ${target_base_name}_man ${_dependencies} )
endif()
if( ${SPHINX_TEXT_OUTPUT} )
Sphinx_add_target( ${target_base_name}_text text ${conf} ${source} ${base_destination}/text )
add_dependencies( ${target_base_name}_text ${_dependencies} )
endif()
if( ${BUILD_TESTING} )
sphinx_add_target( ${target_base_name}_linkcheck linkcheck ${conf} ${source} ${base_destination}/linkcheck )
add_dependencies( ${target_base_name}_linkcheck ${_dependencies} )
endif()
endfunction()
cmake/cblas.cmake
浏览文件 @ a9ea99d7
...@@ -64,24 +64,18 @@ endif() ...@@ -64,24 +64,18 @@ endif()
## Then find the reference-cblas. www.netlib.org/blas/ ## Then find the reference-cblas. www.netlib.org/blas/
set(REFERENCE_CBLAS_ROOT $ENV{REFERENCE_CBLAS_ROOT} CACHE PATH set(REFERENCE_CBLAS_ROOT $ENV{REFERENCE_CBLAS_ROOT} CACHE PATH
"Folder contains reference-cblas") "Folder contains reference-cblas")
if(NOT CMAKE_CROSSCOMPILING) set(REFERENCE_CBLAS_INCLUDE_SEARCH_PATHS
set(REFERENCE_CBLAS_INCLUDE_SEARCH_PATHS ${REFERENCE_CBLAS_ROOT}/include
${REFERENCE_CBLAS_ROOT}/include /usr/include
/usr/include /usr/include/cblas
/usr/include/cblas )
)
set(REFERENCE_CBLAS_LIB_SEARCH_PATHS set(REFERENCE_CBLAS_LIB_SEARCH_PATHS
${REFERENCE_CBLAS_ROOT}/lib ${REFERENCE_CBLAS_ROOT}/lib
/usr/lib /usr/lib
/usr/lib/blas/reference/ /usr/lib/blas/reference/
/usr/lib/reference/ /usr/lib/reference/
) )
else()
# Disable the finding of reference cblas under host's system path
set(REFERENCE_CBLAS_INCLUDE_SEARCH_PATHS ${REFERENCE_CBLAS_ROOT}/include)
set(REFERENCE_CBLAS_LIB_SEARCH_PATHS ${REFERENCE_CBLAS_ROOT}/lib)
endif()
if(WITH_SYSTEM_BLAS) if(WITH_SYSTEM_BLAS)
find_path(REFERENCE_CBLAS_INCLUDE_DIR NAMES cblas.h PATHS find_path(REFERENCE_CBLAS_INCLUDE_DIR NAMES cblas.h PATHS
...@@ -98,10 +92,3 @@ if(WITH_SYSTEM_BLAS) ...@@ -98,10 +92,3 @@ if(WITH_SYSTEM_BLAS)
message(STATUS "Found reference-cblas (include: ${CBLAS_INC_DIR}, library: ${CBLAS_LIBRARIES})") message(STATUS "Found reference-cblas (include: ${CBLAS_INC_DIR}, library: ${CBLAS_LIBRARIES})")
endif() endif()
endif() endif()
if(IOS_USE_VECLIB_FOR_BLAS AND VECLIB_FOUND)
set(CBLAS_FOUND ON)
set(CBLAS_PROVIDER vecLib)
set(CBLAS_INC_DIR ${VECLIB_INC_DIR})
add_definitions(-DPADDLE_USE_VECLIB)
endif()
cmake/configure.cmake
浏览文件 @ a9ea99d7
...@@ -20,41 +20,18 @@ if(WITH_DSO) ...@@ -20,41 +20,18 @@ if(WITH_DSO)
add_definitions(-DPADDLE_USE_DSO) add_definitions(-DPADDLE_USE_DSO)
endif(WITH_DSO) endif(WITH_DSO)
if(WITH_DOUBLE)
add_definitions(-DPADDLE_TYPE_DOUBLE)
endif(WITH_DOUBLE)
if(WITH_ARM_FP16)
add_definitions(-DPADDLE_ARM_FP16)
add_definitions("-march=armv8.2-a+fp16+simd")
endif(WITH_ARM_FP16)
if(WITH_TESTING) if(WITH_TESTING)
add_definitions(-DPADDLE_WITH_TESTING) add_definitions(-DPADDLE_WITH_TESTING)
endif(WITH_TESTING) endif(WITH_TESTING)
if(NOT WITH_TIMER)
add_definitions(-DPADDLE_DISABLE_TIMER)
endif(NOT WITH_TIMER)
if(USE_EIGEN_FOR_BLAS)
add_definitions(-DPADDLE_USE_EIGEN_FOR_BLAS)
endif(USE_EIGEN_FOR_BLAS)
if(EIGEN_USE_THREADS)
add_definitions(-DEIGEN_USE_THREADS)
endif(EIGEN_USE_THREADS)
if(NOT WITH_PROFILER) if(NOT WITH_PROFILER)
add_definitions(-DPADDLE_DISABLE_PROFILER) add_definitions(-DPADDLE_DISABLE_PROFILER)
endif(NOT WITH_PROFILER) endif(NOT WITH_PROFILER)
if(NOT CMAKE_CROSSCOMPILING) if(WITH_AVX AND AVX_FOUND)
if(WITH_AVX AND AVX_FOUND) set(SIMD_FLAG ${AVX_FLAG})
set(SIMD_FLAG ${AVX_FLAG}) elseif(SSE3_FOUND)
elseif(SSE3_FOUND) set(SIMD_FLAG ${SSE3_FLAG})
set(SIMD_FLAG ${SSE3_FLAG})
endif()
endif() endif()
if(WIN32) if(WIN32)
...@@ -80,10 +57,6 @@ if(WIN32) ...@@ -80,10 +57,6 @@ if(WIN32)
endif(NOT MSVC) endif(NOT MSVC)
endif(WIN32) endif(WIN32)
if(NOT WITH_GOLANG)
add_definitions(-DPADDLE_WITHOUT_GOLANG)
endif(NOT WITH_GOLANG)
if(WITH_PSLIB) if(WITH_PSLIB)
add_definitions(-DPADDLE_WITH_PSLIB) add_definitions(-DPADDLE_WITH_PSLIB)
endif() endif()
...@@ -154,7 +127,12 @@ endif() ...@@ -154,7 +127,12 @@ endif()
if (WITH_MKLML AND MKLML_IOMP_LIB) if (WITH_MKLML AND MKLML_IOMP_LIB)
message(STATUS "Enable Intel OpenMP with ${MKLML_IOMP_LIB}") message(STATUS "Enable Intel OpenMP with ${MKLML_IOMP_LIB}")
set(OPENMP_FLAGS "-fopenmp") if(WIN32)
# openmp not support well for now on windows
set(OPENMP_FLAGS "")
else(WIN32)
set(OPENMP_FLAGS "-fopenmp")
endif(WIN32)
set(CMAKE_C_CREATE_SHARED_LIBRARY_FORBIDDEN_FLAGS ${OPENMP_FLAGS}) set(CMAKE_C_CREATE_SHARED_LIBRARY_FORBIDDEN_FLAGS ${OPENMP_FLAGS})
set(CMAKE_CXX_CREATE_SHARED_LIBRARY_FORBIDDEN_FLAGS ${OPENMP_FLAGS}) set(CMAKE_CXX_CREATE_SHARED_LIBRARY_FORBIDDEN_FLAGS ${OPENMP_FLAGS})
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${OPENMP_FLAGS}") set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${OPENMP_FLAGS}")
...@@ -168,55 +146,6 @@ if(WITH_DISTRIBUTE) ...@@ -168,55 +146,6 @@ if(WITH_DISTRIBUTE)
add_definitions(-DPADDLE_WITH_DISTRIBUTE) add_definitions(-DPADDLE_WITH_DISTRIBUTE)
endif() endif()
if(WITH_GOLANG)
# we need to symlink Paddle directory into GOPATH. If we
# don't do it and we have code that depends on Paddle, go
# get ./... will download a new Paddle repo from Github,
# without the changes in our current Paddle repo that we
# want to build.
set(GOPATH "${CMAKE_CURRENT_BINARY_DIR}/go")
file(MAKE_DIRECTORY ${GOPATH})
set(PADDLE_IN_GOPATH "${GOPATH}/src/github.com/PaddlePaddle/Paddle")
file(MAKE_DIRECTORY "${PADDLE_IN_GOPATH}")
set(PADDLE_GO_PATH "${CMAKE_SOURCE_DIR}/go")
add_custom_target(go_path)
add_custom_command(TARGET go_path
# Symlink Paddle directory into GOPATH
COMMAND mkdir -p ${PADDLE_IN_GOPATH}
COMMAND rm -rf ${PADDLE_IN_GOPATH}
COMMAND ln -sf ${CMAKE_SOURCE_DIR} ${PADDLE_IN_GOPATH}
# Automatically get all dependencies specified in the source code
# We can't run `go get -d ./...` for every target, because
# multiple `go get` can not run concurrently, but make need to be
# able to run with multiple jobs.
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}
)
if (GLIDE_INSTALL)
if(EXISTS $ENV{GOPATH}/bin/glide)
set(GLIDE "$ENV{GOPATH}/bin/glide")
else()
message(FATAL_ERROR "no glide executeble found: $ENV{GOPATH}/bin/glide")
endif()
# this command will only run when the file it depends is missing
# or has changed, or the output is missing.
add_custom_command(OUTPUT ${CMAKE_BINARY_DIR}/glide
COMMAND env GOPATH=${GOPATH} ${GLIDE} install
COMMAND touch ${CMAKE_BINARY_DIR}/glide
DEPENDS ${PADDLE_SOURCE_DIR}/go/glide.lock
WORKING_DIRECTORY "${PADDLE_IN_GOPATH}/go"
)
# depends on the custom command which outputs
# ${CMAKE_BINARY_DIR}/glide, the custom command does not need to
# run every time this target is built.
add_custom_target(go_vendor DEPENDS ${CMAKE_BINARY_DIR}/glide go_path)
endif()
endif(WITH_GOLANG)
if(WITH_GRPC) if(WITH_GRPC)
add_definitions(-DPADDLE_WITH_GRPC) add_definitions(-DPADDLE_WITH_GRPC)
endif(WITH_GRPC) endif(WITH_GRPC)
......
cmake/cross_compiling/android.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# This is a toolchain file for cross-compiling for Android, and the
# configuration refers to the open-source resposity:
# https://github.com/taka-no-me/android-cmake
# Most of the variables are compatible with that used in
# https://developer.android.com/ndk/guides/cmake.html
# The supported variables are listed belows:
#
# ANDROID_STANDALONE_TOOLCHAIN
# ANDROID_TOOLCHAIN
# ANDROID_ABI
# ANDROID_NATIVE_API_LEVEL
# ANDROID_ARM_MODE
# ANDROID_ARM_NEON
#
# For CMake >= 3.7.0, all the settings will be delivered to CMake system
# variables to let CMake do the cross-compiling configurations itself.
# More detail of cross-compiling settings
# https://cmake.org/cmake/help/v3.7/manual/cmake-toolchains.7.html
IF(NOT ANDROID)
return()
ENDIF()
# check the exist of android standalone toolchain
IF(NOT DEFINED ANDROID_STANDALONE_TOOLCHAIN)
SET(ANDROID_STANDALONE_TOOLCHAIN $ENV{ANDROID_STANDALONE_TOOLCHAIN}
CACHE PATH "Folder holds the standalone toolchain of Android NDK")
ENDIF()
IF(NOT ANDROID_STANDALONE_TOOLCHAIN)
MESSAGE(WARNING "It is recommended to set ANDROID_STANDALONE_TOOLCHAIN to "
"use a standalone toolchain.\n"
"To cross-compile for Android, you need to:\n"
"1. Download an Android NDK from"
" https://developer.android.com/ndk/downloads/index.html\n"
"2. Setup a standalone toolchain"
"https://developer.android.google.cn/ndk/guides/standalone_toolchain.html?hl=zh-cn\n")
ENDIF()
IF(NOT DEFINED CMAKE_SYSTEM_VERSION AND ANDROID_NATIVE_API_LEVEL)
IF(ANDROID_NATIVE_API_LEVEL MATCHES "^android-[0-9]+$")
STRING(REPLACE "android-" "" CMAKE_SYSTEM_VERSION "${CMAKE_MATCH_0}")
ELSEIF(ANDROID_NATIVE_API_LEVEL MATCHES "^[0-9]+$")
SET(CMAKE_SYSTEM_VERSION ${ANDROID_NATIVE_API_LEVEL})
ENDIF()
ENDIF()
IF(NOT DEFINED ANDROID_TOOLCHAIN)
SET(ANDROID_TOOLCHAIN clang)
ENDIF()
IF(NOT DEFINED ANDROID_ABI)
SET(ANDROID_ABI "armeabi-v7a")
ENDIF()
IF(NOT DEFINED ANDROID_ARM_MODE)
SET(ANDROID_ARM_MODE ON)
ENDIF()
IF(ANDROID_ARM_MODE)
SET(ANDROID_ARM_MODE_NAME "arm")
ELSE(ANDROID_ARM_MODE)
SET(ANDROID_ARM_MODE_NAME "thumb")
ENDIF(ANDROID_ARM_MODE)
IF(NOT DEFINED ANDROID_ARM_NEON)
SET(ANDROID_ARM_NEON ON)
ENDIF()
IF("${CMAKE_VERSION}" VERSION_LESS "3.7.0")
IF("${CMAKE_VERSION}" VERSION_LESS "3.1.0")
SET(CMAKE_SYSTEM_NAME "Linux")
ENDIF()
MESSAGE(WARNING "It is recommended to use CMake >= 3.7.0 (current version: "
"${CMAKE_VERSION}), when cross-compiling for Android.")
IF(ANDROID_STANDALONE_TOOLCHAIN)
# Use standalone toolchain
SET(CMAKE_SYSROOT "${ANDROID_STANDALONE_TOOLCHAIN}/sysroot")
IF(NOT CMAKE_SYSTEM_VERSION)
SET(ANDROID_STANDALONE_TOOLCHAIN_API "")
SET(ANDROID_API_LEVEL_H_REGEX "^[\t ]*#[\t ]*define[\t ]+__ANDROID_API__[\t ]+([0-9]+)")
FILE(STRINGS "${ANDROID_STANDALONE_TOOLCHAIN}/sysroot/usr/include/android/api-level.h"
ANDROID_API_LEVEL_H_CONTENT REGEX "${ANDROID_API_LEVEL_H_REGEX}")
IF(ANDROID_API_LEVEL_H_CONTENT MATCHES "${ANDROID_API_LEVEL_H_REGEX}")
SET(ANDROID_STANDALONE_TOOLCHAIN_API "${CMAKE_MATCH_1}")
ENDIF()
SET(CMAKE_SYSTEM_VERSION ${ANDROID_STANDALONE_TOOLCHAIN_API})
ENDIF()
# Toolchain
SET(ANDROID_TOOLCHAIN_ROOT ${ANDROID_STANDALONE_TOOLCHAIN})
ELSE(ANDROID_NDK)
# TODO: use android ndk
ENDIF()
IF(ANDROID_ABI MATCHES "^armeabi(-v7a)?$")
SET(ANDROID_TOOLCHAIN_NAME arm-linux-androideabi)
IF(ANDROID_ABI STREQUAL "armeabi")
SET(CMAKE_SYSTEM_PROCESSOR armv5te)
SET(ANDROID_CLANG_TRIPLE armv5te-none-linux-androideabi)
ELSEIF(ANDROID_ABI STREQUAL "armeabi-v7a")
SET(CMAKE_SYSTEM_PROCESSOR armv7-a)
SET(ANDROID_CLANG_TRIPLE armv7-none-linux-androideabi)
ENDIF()
ELSEIF(ANDROID_ABI STREQUAL "arm64-v8a")
SET(ANDROID_TOOLCHAIN_NAME aarch64-linux-android)
SET(CMAKE_SYSTEM_PROCESSOR aarch64)
SET(ANDROID_CLANG_TRIPLE aarch64-none-linux-android)
ELSE()
MESSAGE(FATAL_ERROR "Invalid Android ABI: ${ANDROID_ABI}.")
ENDIF()
SET(ANDROID_TOOLCHAIN_PREFIX "${ANDROID_TOOLCHAIN_ROOT}/bin/${ANDROID_TOOLCHAIN_NAME}-")
IF(ANDROID_TOOLCHAIN STREQUAL clang)
SET(ANDROID_C_COMPILER_NAME clang)
SET(ANDROID_CXX_COMPILER_NAME clang++)
SET(CMAKE_C_COMPILER_TARGET ${ANDROID_CLANG_TRIPLE})
SET(CMAKE_CXX_COMPILER_TARGET ${ANDROID_CLANG_TRIPLE})
ELSEIF(ANDROID_TOOLCHAIN STREQUAL gcc)
SET(ANDROID_C_COMPILER_NAME gcc)
SET(ANDROID_CXX_COMPILER_NAME g++)
ELSE()
MESSAGE(FATAL_ERROR "Invalid Android toolchain: ${ANDROID_TOOLCHAIN}")
ENDIF()
# C compiler
IF(NOT CMAKE_C_COMPILER)
SET(ANDROID_C_COMPILER "${ANDROID_TOOLCHAIN_PREFIX}${ANDROID_C_COMPILER_NAME}")
ELSE()
GET_FILENAME_COMPONENT(ANDROID_C_COMPILER ${CMAKE_C_COMPILER} PROGRAM)
ENDIF()
IF(NOT EXISTS ${ANDROID_C_COMPILER})
MESSAGE(FATAL_ERROR "Cannot find C compiler: ${ANDROID_C_COMPILER}")
ENDIF()
# CXX compiler
IF(NOT CMAKE_CXX_COMPILER)
SET(ANDROID_CXX_COMPILER "${ANDROID_TOOLCHAIN_PREFIX}${ANDROID_CXX_COMPILER_NAME}")
ELSE()
GET_FILENAME_COMPONENT(ANDROID_CXX_COMPILER ${CMAKE_CXX_COMPILER} PROGRAM)
ENDIF()
IF(NOT EXISTS ${ANDROID_CXX_COMPILER})
MESSAGE(FATAL_ERROR "Cannot find CXX compiler: ${ANDROID_CXX_COMPILER}")
ENDIF()
SET(CMAKE_C_COMPILER ${ANDROID_C_COMPILER} CACHE PATH "C compiler" FORCE)
SET(CMAKE_CXX_COMPILER ${ANDROID_CXX_COMPILER} CACHE PATH "CXX compiler" FORCE)
# Toolchain and ABI specific flags.
SET(ANDROID_COMPILER_FLAGS "-ffunction-sections -fdata-sections")
SET(ANDROID_LINKER_FLAGS "-Wl,--gc-sections")
IF(ANDROID_ABI STREQUAL "armeabi")
LIST(APPEND ANDROID_COMPILER_FLAGS
-march=armv5te
-mtune=xscale
-msoft-float)
ELSEIF(ANDROID_ABI STREQUAL "armeabi-v7a")
LIST(APPEND ANDROID_COMPILER_FLAGS
-march=armv7-a
-mfloat-abi=softfp)
IF(ANDROID_ARM_NEON)
LIST(APPEND ANDROID_COMPILER_FLAGS -mfpu=neon)
ELSE()
LIST(APPEND ANDROID_COMPILER_FLAGS -mfpu=vfpv3-d16)
ENDIF()
LIST(APPEND ANDROID_LINKER_FLAGS -Wl,--fix-cortex-a8)
ELSEIF(ANDROID_ABI STREQUAL "arm64-v8a")
LIST(APPEND ANDROID_COMPILER_FLAGS -march=armv8-a)
ENDIF()
IF(ANDROID_ABI MATCHES "^armeabi(-v7a)?$")
IF(ANDROID_ARM_MODE)
LIST(APPEND ANDROID_COMPILER_FLAGS -marm)
ELSE()
LIST(APPEND ANDROID_COMPILER_FLAGS -mthumb)
ENDIF()
IF(ANDROID_TOOLCHAIN STREQUAL clang)
# Disable integrated-as for better compatibility.
LIST(APPEND ANDROID_COMPILER_FLAGS -fno-integrated-as)
ENDIF()
ENDIF()
IF(ANDROID_TOOLCHAIN STREQUAL clang)
# CMake automatically forwards all compiler flags to the linker,
# and clang doesn't like having -Wa flags being used for linking.
# To prevent CMake from doing this would require meddling with
# the CMAKE_<LANG>_COMPILE_OBJECT rules, which would get quite messy.
LIST(APPEND ANDROID_LINKER_FLAGS -Qunused-arguments)
ENDIF()
STRING(REPLACE ";" " " ANDROID_COMPILER_FLAGS "${ANDROID_COMPILER_FLAGS}")
STRING(REPLACE ";" " " ANDROID_LINKER_FLAGS "${ANDROID_LINKER_FLAGS}")
SET(CMAKE_C_FLAGS "${ANDROID_COMPILER_FLAGS} ${CMAKE_C_FLAGS}"
CACHE STRING "C flags")
SET(CMAKE_CXX_FLAGS "${ANDROID_COMPILER_FLAGS} ${CMAKE_CXX_FLAGS}"
CACHE STRING "CXX flags")
SET(CMAKE_SHARED_LINKER_FLAGS "${ANDROID_LINKER_FLAGS} ${CMAKE_SHARED_LINKER_FLAGS}"
CACHE STRING "shared linker flags")
SET(CMAKE_POSITION_INDEPENDENT_CODE TRUE)
SET(CMAKE_EXE_LINKER_FLAGS "-pie -fPIE ${ANDROID_LINKER_FLAGS} ${CMAKE_EXE_LINKER_FLAGS}"
CACHE STRING "executable linker flags")
MESSAGE(STATUS "Android: Targeting API '${CMAKE_SYSTEM_VERSION}' "
"with architecture '${ANDROID_ARM_MODE_NAME}', "
"ABI '${ANDROID_ABI}', and processor '${CMAKE_SYSTEM_PROCESSOR}'")
MESSAGE(STATUS "System CMAKE_C_FLAGS: " ${CMAKE_C_FLAGS})
MESSAGE(STATUS "System CMAKE_CXX_FLAGS: " ${CMAKE_CXX_FLAGS})
ELSE()
IF(ANDROID_STANDALONE_TOOLCHAIN)
SET(CMAKE_ANDROID_STANDALONE_TOOLCHAIN ${ANDROID_STANDALONE_TOOLCHAIN})
ENDIF()
SET(CMAKE_ANDROID_ARCH_ABI ${ANDROID_ABI})
IF(ANDROID_ABI MATCHES "^armeabi(-v7a)?$")
SET(CMAKE_ANDROID_ARM_MODE ${ANDROID_ARM_MODE})
IF(ANDROID_ABI STREQUAL "armeabi-v7a")
SET(CMAKE_ANDROID_ARM_NEON ${ANDROID_ARM_NEON})
ENDIF()
ENDIF()
ENDIF()
cmake/cross_compiling/host.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# find host C compiler
IF(HOST_C_COMPILER)
SET(HOST_C_COMPILER_NAME ${HOST_C_COMPILER})
ELSEIF(NOT $ENV{CC} STREQUAL "")
SET(HOST_C_COMPILER_NAME $ENV{CC})
ELSE()
SET(HOST_C_COMPILER_NAME cc)
ENDIF()
GET_FILENAME_COMPONENT(HOST_C_COMPILER_PATH ${HOST_C_COMPILER_NAME} PROGRAM)
IF(NOT HOST_C_COMPILER_PATH OR NOT EXISTS ${HOST_C_COMPILER_PATH})
MESSAGE(FATAL_ERROR "Cannot find host C compiler, set host C compiler:\n"
"\tcmake .. -DHOST_C_COMPILER=...")
ENDIF()
# find host CXX compiler
IF(HOST_CXX_COMPILER)
SET(HOST_CXX_COMPILER_NAME ${HOST_CXX_COMPILER})
ELSEIF(NOT $ENV{CXX} STREQUAL "")
SET(HOST_CXX_COMPILER_NAME $ENV{CXX})
ELSE()
SET(HOST_CXX_COMPILER_NAME c++)
ENDIF()
GET_FILENAME_COMPONENT(HOST_CXX_COMPILER_PATH ${HOST_CXX_COMPILER_NAME} PROGRAM)
IF(NOT HOST_CXX_COMPILER_PATH OR NOT EXISTS ${HOST_CXX_COMPILER_PATH})
MESSAGE(FATAL_ERROR "Cannot find host CXX compiler, set host CXX compiler:\n"
"\tcmake .. -DHOST_CXX_COMPILER=...")
ENDIF()
SET(HOST_C_COMPILER ${HOST_C_COMPILER_PATH} CACHE PATH "Host C compiler")
SET(HOST_CXX_COMPILER ${HOST_CXX_COMPILER_PATH} CACHE PATH "Host CXX compiler")
MESSAGE(STATUS "Found host C compiler: " ${HOST_C_COMPILER})
MESSAGE(STATUS "Found host CXX compiler: " ${HOST_CXX_COMPILER})
cmake/cross_compiling/ios.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# This is a toolchain file for cross-compiling for iOS, and the
# configuration largely refers to public toolchain file:
# https://raw.githubusercontent.com/leetal/ios-cmake/master/ios.toolchain.cmake
# and
# https://github.com/cristeab/ios-cmake
#
# Supports options:
# IOS_PLATFORM = OS (default) or SIMULATOR
# This decides if SDKS will be selected from the iPhoneOS.platform or iPhoneSimulator.platform folders
# OS - the default, used to build for iPhone and iPad physical devices, which have an arm arch.
# SIMULATOR - used to build for the Simulator platforms, which have an x86 arch.
# IOS_ARCH
# The archectures wanted to support, such "arm64", "armv7;arm64"
# IOS_DEPLOYMENT_TARGET
# The minimum iOS deployment version, such as "7.0"
# IOS_ENABLE_BITCODE = ON (default) or OFF
# IOS_USE_VECLIB_FOR_BLAS = OFF (default) or ON
# IOS_DEVELOPER_ROOT = automatic(default) or /path/to/platform/Developer folder
# By default this location is automatcially chosen based on the IOS_PLATFORM value above.
# If set manually, it will override the default location and force the user of a particular Developer Platform
# IOS_SDK_ROOT = automatic(default) or /path/to/platform/Developer/SDKs/SDK folder
# By default this location is automatcially chosen based on the IOS_DEVELOPER_ROOT value.
# In this case it will always be the most up-to-date SDK found in the IOS_DEVELOPER_ROOT path.
# If set manually, this will force the use of a specific SDK version
# Macros:
# set_xcode_property (TARGET XCODE_PROPERTY XCODE_VALUE)
# A convenience macro for setting xcode specific properties on targets
# example: set_xcode_property (myioslib IPHONEOS_DEPLOYMENT_TARGET "3.1")
# find_host_package (PROGRAM ARGS)
# A macro used to find executable programs on the host system, not within the iOS environment.
# Thanks to the android-cmake project for providing the command
if(NOT IOS)
return()
endif()
set(CMAKE_SYSTEM_NAME Darwin)
# Get the Xcode version being used.
execute_process(COMMAND xcodebuild -version
OUTPUT_VARIABLE XCODE_VERSION
RESULT_VARIABLE XCODE_VERSION_RESULT
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
if(NOT ${XCODE_VERSION_RESULT})
string(REGEX MATCH "Xcode [0-9\\.]+" XCODE_VERSION "${XCODE_VERSION}")
string(REGEX REPLACE "Xcode ([0-9\\.]+)" "\\1" XCODE_VERSION "${XCODE_VERSION}")
message(STATUS "Building with Xcode version: ${XCODE_VERSION}")
else()
message(FATAL_ERROR "Cannot execute xcodebuild, please check whether xcode is installed.")
endif()
# Required as of cmake 2.8.10
set(CMAKE_OSX_DEPLOYMENT_TARGET "" CACHE STRING "Force unset of the deployment target for iOS" FORCE)
# Setup iOS platform unless specified manually with IOS_PLATFORM
if(NOT DEFINED IOS_PLATFORM)
set(IOS_PLATFORM "OS")
endif()
set(IOS_PLATFORM ${IOS_PLATFORM} CACHE STRING "Type of iOS Platform")
# Set the architecture for iOS
if(NOT DEFINED IOS_ARCH)
if(IOS_PLATFORM STREQUAL "OS")
set(IOS_ARCH "armv7;armv7s;arm64")
elseif(IOS_PLATFORM STREQUAL "SIMULATOR")
set(IOS_ARCH "i386;x86_64")
endif()
endif()
set(CMAKE_OSX_ARCHITECTURES ${IOS_ARCH} CACHE string "Build architecture for iOS")
# Specify minimum iOS deployment version
if(NOT DEFINED IOS_DEPLOYMENT_TARGET)
set(IOS_DEPLOYMENT_TARGET "7.0")
endif()
set(IOS_DEPLOYMENT_TARGET ${IOS_DEPLOYMENT_TARGET} CACHE STRING "Minimum iOS version")
# Whether to enable bitcode
if(NOT DEFINED IOS_ENABLE_BITCODE)
set(IOS_ENABLE_BITCODE ON)
endif()
set(IOS_ENABLE_BITCODE ${IOS_ENABLE_BITCODE} CACHE BOOL "Whether to enable bitcode")
if(NOT DEFINED IOS_USE_VECLIB_FOR_BLAS)
set(IOS_USE_VECLIB_FOR_BLAS OFF)
endif()
set(IOS_USE_VECLIB_FOR_BLAS ${IOS_UES_VECLIB_FOR_BLAS} CACHE BOOL "Whether to use veclib")
# Check the platform selection and setup for developer root
if(${IOS_PLATFORM} STREQUAL "OS")
set(IOS_PLATFORM_LOCATION "iPhoneOS.platform")
set(XCODE_IOS_PLATFORM iphoneos)
# This causes the installers to properly locate the output libraries
set(CMAKE_XCODE_EFFECTIVE_PLATFORMS "-iphoneos")
elseif(${IOS_PLATFORM} STREQUAL "SIMULATOR")
set(IOS_PLATFORM_LOCATION "iPhoneSimulator.platform")
set(XCODE_IOS_PLATFORM iphonesimulator)
# This causes the installers to properly locate the output libraries
set(CMAKE_XCODE_EFFECTIVE_PLATFORMS "-iphonesimulator")
elseif(${IOS_PLATFORM} STREQUAL "WATCHOS")
set(IOS_PLATFORM_LOCATION "WatchOS.platform")
set(XCODE_IOS_PLATFORM watchos)
# This causes the installers to properly locate the output libraries
set(CMAKE_XCODE_EFFECTIVE_PLATFORMS "-watchos")
else(${IOS_PLATFORM} STREQUAL "OS")
message(FATAL_ERROR "Unsupported IOS_PLATFORM value selected. Please set to\n"
"\t OS, SIMULATOR, or WATCHOS.")
endif()
# Check iOS developer toolchain
if(NOT DEFINED IOS_DEVELOPER_ROOT)
# Setup iOS developer location
execute_process(COMMAND xcode-select -print-path
OUTPUT_VARIABLE XCODE_DEVELOPER_DIR
RESULT_VARIABLE XCODE_DEVELOPER_DIR_RESULT
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
# Xcode 4.3 changed the installation location, choose the most recent one available
if(${XCODE_VERSION} VERSION_LESS "4.3.0")
set(IOS_DEVELOPER_ROOT "/Developer/Platforms/${IOS_PLATFORM_LOCATION}/Developer")
else()
set(IOS_DEVELOPER_ROOT "${XCODE_DEVELOPER_DIR}/Platforms/${IOS_PLATFORM_LOCATION}/Developer")
endif()
endif()
if(EXISTS ${IOS_DEVELOPER_ROOT})
set(IOS_DEVELOPER_ROOT ${IOS_DEVELOPER_ROOT} CACHE PATH "Location of iOS Platform")
else()
message(FATAL_ERROR "Invalid IOS_DEVELOPER_ROOT: ${IOS_DEVELOPER_ROOT} does not exist.")
endif()
# Check iOS SDK
if(NOT DEFINED IOS_SDK_ROOT)
# Find and use the most recent iOS sdk
file(GLOB IOS_SDK_LISTS "${IOS_DEVELOPER_ROOT}/SDKs/*")
if(IOS_SDK_LISTS)
list(SORT IOS_SDK_LISTS)
list(REVERSE IOS_SDK_LISTS)
list(GET IOS_SDK_LISTS 0 IOS_SDK_ROOT)
else(IOS_SDK_LISTS)
message(FATAL_ERROR "No iOS SDK's found in default search path ${IOS_DEVELOPER_ROOT}."
" Please manually set IOS_SDK_ROOT or install the iOS SDK.")
endif(IOS_SDK_LISTS)
endif()
if(EXISTS ${IOS_SDK_ROOT})
set(IOS_SDK_ROOT ${IOS_SDK_ROOT} CACHE PATH "Location of the selected iOS SDK")
message(STATUS "iOS toolchain: ${IOS_SDK_ROOT}")
else()
message(FATAL_ERROR "Invalid IOS_SDK_ROOT: ${IOS_SDK_ROOT} does not exist.")
endif()
# Set the sysroot default to the most recent SDK
set(CMAKE_OSX_SYSROOT ${IOS_SDK_ROOT} CACHE PATH "Sysroot used for iOS support")
# Get version of iOS SDK
execute_process(COMMAND xcodebuild -sdk ${CMAKE_OSX_SYSROOT} -version SDKVersion
OUTPUT_VARIABLE IOS_SDK_VERSION
RESULT_VARIABLE IOS_SDK_VERSION_RESULT
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
if(${IOS_SDK_VERSION_RESULT})
string(REGEX MATCH "(([0-9]+)\\.)+([0-9]+)" IOS_SDK_VERSION "${IOS_SDK_ROOT}")
endif()
if(NOT IOS_SDK_VERSION)
message(WARNING "Cannot get SDK's version.")
set(IOS_SDK_VERSION 1)
endif()
set(CMAKE_SYSTEM_VERSION ${IOS_SDK_VERSION})
# Find the C & C++ compilers for the specified SDK.
if(NOT CMAKE_C_COMPILER)
# Default to use clang
execute_process(COMMAND xcrun -sdk ${CMAKE_OSX_SYSROOT} -find clang
OUTPUT_VARIABLE IOS_C_COMPILER
RESULT_VARIABLE IOS_C_COMPILER_RESULT
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
if(${IOS_C_COMPILER_RESULT})
get_filename_component(IOS_C_COMPILER clang PROGRAM)
endif()
else(NOT CMAKE_C_COMPILER)
# User can set it in cmake command
get_filename_component(IOS_C_COMPILER ${CMAKE_C_COMPILER} PROGRAM)
endif(NOT CMAKE_C_COMPILER)
if(NOT EXISTS ${IOS_C_COMPILER})
message(FATAL_ERROR "Cannot find C compiler: ${IOS_C_COMPILER}")
endif()
if(NOT CMAKE_CXX_COMPILER)
# Default to use clang++
execute_process(COMMAND xcrun -sdk ${CMAKE_OSX_SYSROOT} -find clang++
OUTPUT_VARIABLE IOS_CXX_COMPILER
RESULT_VARIABLE IOS_CXX_COMPILER_RESULT
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
if(${IOS_CXX_COMPILER_RESULT})
get_filename_component(IOS_CXX_COMPILER clang++ PROGRAM)
endif()
else(NOT CMAKE_CXX_COMPILER)
# User can set it in cmake command
get_filename_component(IOS_CXX_COMPILER ${CMAKE_CXX_COMPILER} PROGRAM)
endif(NOT CMAKE_CXX_COMPILER)
if(NOT EXISTS ${IOS_CXX_COMPILER})
message(FATAL_ERROR "Cannot find CXX compiler: ${IOS_CXX_COMPILER}")
endif()
set(CMAKE_C_COMPILER ${IOS_C_COMPILER} CACHE PATH "C compiler" FORCE)
set(CMAKE_CXX_COMPILER ${IOS_CXX_COMPILER} CACHE PATH "CXX compiler" FORCE)
set(CMAKE_C_OSX_COMPATIBILITY_VERSION_FLAG "-compatibility_version ")
set(CMAKE_C_OSX_CURRENT_VERSION_FLAG "-current_version ")
set(CMAKE_CXX_OSX_COMPATIBILITY_VERSION_FLAG "${CMAKE_C_OSX_COMPATIBILITY_VERSION_FLAG}")
set(CMAKE_CXX_OSX_CURRENT_VERSION_FLAG "${CMAKE_C_OSX_CURRENT_VERSION_FLAG}")
# Set iOS specific C/C++ flags
if(IOS_PLATFORM STREQUAL "OS")
if(XCODE_VERSION VERSION_LESS "7.0")
set(XCODE_IOS_PLATFORM_VERSION_FLAGS "-mios-version-min=${IOS_DEPLOYMENT_TARGET}")
else()
# Xcode 7.0+ uses flags we can build directly from XCODE_IOS_PLATFORM.
set(XCODE_IOS_PLATFORM_VERSION_FLAGS "-m${XCODE_IOS_PLATFORM}-version-min=${IOS_DEPLOYMENT_TARGET}")
endif()
else()
set(XCODE_IOS_FLATFORM_VERSION_FLAGS "-mios-simulator-version-min=${IOS_DEPLOYMENT_TARGET}")
endif()
if(IOS_ENABLE_BITCODE)
set(XCODE_IOS_BITCODE_FLAGS "${IOS_COMPILER_FLAGS} -fembed-bitcode")
else()
set(XCODE_IOS_BITCODE_FLAGS "")
endif()
set(IOS_COMPILER_FLAGS "${XCODE_IOS_PLATFORM_VERSION_FLAGS} ${XCODE_IOS_BITCODE_FLAGS}")
# Hidden visibilty is required for cxx on iOS
set(CMAKE_C_FLAGS "${IOS_COMPILER_FLAGS} ${CMAKE_C_FLAGS}" CACHE STRING "C flags")
set(CMAKE_CXX_FLAGS "${IOS_COMPILER_FLAGS} -fvisibility=hidden -fvisibility-inlines-hidden ${CMAKE_CXX_FLAGS}" CACHE STRING "CXX flags")
set(IOS_LINK_FLAGS "${XCODE_IOS_PLATFORM_VERSION_FLAGS} -Wl,-search_paths_first")
if(IOS_USE_VECLIB_FOR_BLAS)
# Find vecLib for iOS
set(VECLIB_SEARCH_DIRS
${IOS_SDK_ROOT}/System/Library/Frameworks/Accelerate.framework/Versions/Current/Frameworks
${IOS_SDK_ROOT}/System/Library/Frameworks/Accelerate.framework/Frameworks
)
find_path(VECLIB_INC_DIR vecLib.h PATHS ${VECLIB_SEARCH_DIRS}/vecLib.framework/Headers)
include(FindPackageHandleStandardArgs)
find_package_handle_standard_args(vecLib DEFAULT_MSG VECLIB_INC_DIR)
if(VECLIB_FOUND)
if(VECLIB_INC_DIR MATCHES "^/System/Library/Frameworks/vecLib.framework.*")
set(IOS_LINK_FLAGS ${IOS_LINK_FLAGS} -lcblas "-framework vecLib")
message(STATUS "Found standalone vecLib.framework")
else()
set(IOS_LINK_FLAGS ${IOS_LINK_FLAGS} -lcblas "-framework Accelerate")
message(STATUS "Found vecLib as part of Accelerate.framework")
endif()
endif()
endif()
set(CMAKE_C_LINK_FLAGS "${IOS_LINK_FLAGS} ${CMAKE_C_LINK_FLAGS}")
set(CMAKE_CXX_LINK_FLAGS "${IOS_LINK_FLAGS} ${CMAKE_CXX_LINK_FLAGS}")
set(CMAKE_PLATFORM_HAS_INSTALLNAME 1)
if(NOT IOS_ENABLE_BITCODE)
set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS "-dynamiclib -headerpad_max_install_names")
set(CMAKE_SHARED_MODULE_CREATE_C_FLAGS "-bundle -headerpad_max_install_names")
else()
set(CMAKE_SHARED_LIBRARY_CREATE_C_FLAGS "-dynamiclib")
set(CMAKE_SHARED_MODULE_CREATE_C_FLAGS "-bundle")
endif()
set(CMAKE_SHARED_MODULE_LOADER_C_FLAG "-Wl,-bundle_loader,")
set(CMAKE_SHARED_MODULE_LOADER_CXX_FLAG "-Wl,-bundle_loader,")
set(CMAKE_FIND_LIBRARY_SUFFIXES ".dylib" ".so" ".a")
# hack: if a new cmake (which uses CMAKE_INSTALL_NAME_TOOL) runs on an old build tree
# (where install_name_tool was hardcoded) and where CMAKE_INSTALL_NAME_TOOL isn't in the cache
# and still cmake didn't fail in CMakeFindBinUtils.cmake (because it isn't rerun)
# hardcode CMAKE_INSTALL_NAME_TOOL here to install_name_tool, so it behaves as it did before, Alex
if(NOT DEFINED CMAKE_INSTALL_NAME_TOOL)
find_program(CMAKE_INSTALL_NAME_TOOL install_name_tool)
endif()
# Set the find root to the iOS developer roots and to user defined paths
set(CMAKE_FIND_ROOT_PATH ${IOS_DEVELOPER_ROOT} ${IOS_SDK_ROOT} ${CMAKE_PREFIX_PATH}
CACHE string "iOS find search path root")
# default to searching for frameworks first
set(CMAKE_FIND_FRAMEWORK FIRST)
# set up the default search directories for frameworks
set(CMAKE_SYSTEM_FRAMEWORK_PATH
${IOS_SDK_ROOT}/System/Library/Frameworks
${IOS_SDK_ROOT}/System/Library/PrivateFrameworks
${IOS_SDK_ROOT}/Developer/Library/Frameworks
)
# only search the iOS sdks, not the remainder of the host filesystem
set(CMAKE_FIND_ROOT_PATH_MODE_PROGRAM NEVER)
set(CMAKE_FIND_ROOT_PATH_MODE_LIBRARY ONLY)
set(CMAKE_FIND_ROOT_PATH_MODE_INCLUDE ONLY)
message(STATUS "iOS: Targeting iOS '${CMAKE_SYSTEM_VERSION}', "
"building for '${IOS_PLATFORM}' platform, with architecture '${CMAKE_OSX_ARCHITECTURES}'")
message(STATUS "System CMAKE_C_FLAGS: ${CMAKE_C_FLAGS}")
message(STATUS "System CMAKE_CXX_FLAGS: ${CMAKE_CXX_FLAGS}")
# Used in ExternalProject command
string(REPLACE ";" "\\$<SEMICOLON>" EXTERNAL_IOS_ARCHITECTURES "${CMAKE_OSX_ARCHITECTURES}")
set(EXTERNAL_OPTIONAL_ARGS
-DCMAKE_OSX_SYSROOT=${CMAKE_OSX_SYSROOT}
-DCMAKE_OSX_ARCHITECTURES=${EXTERNAL_IOS_ARCHITECTURES})
# This little macro lets you set any XCode specific property
macro(set_xcode_property TARGET XCODE_PROPERTY XCODE_VALUE)
set_property (TARGET ${TARGET} PROPERTY XCODE_ATTRIBUTE_${XCODE_PROPERTY} ${XCODE_VALUE})
endmacro(set_xcode_property)
# This macro lets you find executable programs on the host system
macro(find_host_package)
set(CMAKE_FIND_ROOT_PATH_MODE_PROGRAM NEVER)
set(CMAKE_FIND_ROOT_PATH_MODE_LIBRARY NEVER)
set(CMAKE_FIND_ROOT_PATH_MODE_INCLUDE NEVER)
set(IOS FALSE)
find_package(${ARGN})
set(IOS TRUE)
set(CMAKE_FIND_ROOT_PATH_MODE_PROGRAM ONLY)
set(CMAKE_FIND_ROOT_PATH_MODE_LIBRARY ONLY)
set(CMAKE_FIND_ROOT_PATH_MODE_INCLUDE ONLY)
endmacro(find_host_package)
cmake/cross_compiling/raspberry_pi.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# This is a toolchain file for cross-compiling for Raspberry Pi.
# The supported variables are listed belows:
#
# RPI_TOOLCHAIN
# RPI_ARM_NEON
#
# Also you can set CMAKE_C/CXX_COMPILER yourself, through cmake arguments.
IF(NOT RPI)
return()
ENDIF()
SET(CMAKE_SYSTEM_NAME Linux)
SET(CMAKE_SYSTEM_VERSION 1)
SET(CMAKE_SYSTEM_PROCESSOR arm)
# check the exist of raspberry pi toolchain
IF(NOT DEFINED RPI_TOOLCHAIN)
SET(RPI_TOOLCHAIN $ENV{RPI_TOOLCHAIN}
CACHE PATH "Folder holds the toolchain of Raspberr Pi")
ENDIF()
IF(NOT RPI_TOOLCHAIN)
MESSAGE(WARNING "It is recommended to set RPI_TOOLCHAIN to use toolchain.\n"
"To cross-compile for Raspberry Pi, you need to download the tools using:\n"
" git clone https://github.com/raspberrypi/tools\n")
ENDIF()
IF(NOT DEFINED RPI_ARM_NEON)
SET(RPI_ARM_NEON ON)
ENDIF()
IF(RPI_TOOLCHAIN)
SET(RPI_TOOLCHAIN_ROOT ${RPI_TOOLCHAIN})
IF(RPI_TOOLCHAIN_ROOT MATCHES "gcc-linaro-arm-linux-gnueabihf-raspbian(-x64)?$")
# gcc-linaro-arm-linux-gnueabihf-raspbian
# gcc-linaro-arm-linux-gnueabihf-raspbian-x64
SET(RPI_TOOLCHAIN_NAME arm-linux-gnueabihf)
ENDIF()
SET(RPI_TOOLCHAIN_PREFIX "${RPI_TOOLCHAIN_ROOT}/bin/${RPI_TOOLCHAIN_NAME}-")
ENDIF()
# C compiler
IF(NOT CMAKE_C_COMPILER)
SET(RPI_C_COMPILER "${RPI_TOOLCHAIN_PREFIX}gcc")
ELSE()
GET_FILENAME_COMPONENT(RPI_C_COMPILER ${CMAKE_C_COMPILER} PROGRAM)
ENDIF()
IF(NOT EXISTS ${RPI_C_COMPILER})
MESSAGE(FATAL_ERROR "Cannot find C compiler: ${RPI_C_COMPILER}")
ENDIF()
# CXX compiler
IF(NOT CMAKE_CXX_COMPILER)
SET(RPI_CXX_COMPILER "${RPI_TOOLCHAIN_PREFIX}g++")
ELSE()
GET_FILENAME_COMPONENT(RPI_CXX_COMPILER ${CMAKE_CXX_COMPILER} PROGRAM)
ENDIF()
IF(NOT EXISTS ${RPI_CXX_COMPILER})
MESSAGE(FATAL_ERROR "Cannot find CXX compiler: ${RPI_CXX_COMPILER}")
ENDIF()
SET(CMAKE_C_COMPILER ${RPI_C_COMPILER} CACHE PATH "C compiler" FORCE)
SET(CMAKE_CXX_COMPILER ${RPI_CXX_COMPILER} CACHE PATH "CXX compiler" FORCE)
IF(RPI_ARM_NEON)
SET(RPI_C_FLAGS "${RPI_C_FLAGS} -mfpu=neon")
ENDIF()
SET(CMAKE_C_FLAGS "${RPI_C_FLAGS} ${CMAKE_C_FLAGS}" CACHE STRING "C flags")
SET(CMAKE_CXX_FLAGS "${RPI_C_FLAGS} ${CMAKE_CXX_FLAGS}" CACHE STRING "CXX flags")
cmake/cuda.cmake
浏览文件 @ a9ea99d7
...@@ -63,9 +63,7 @@ function(select_nvcc_arch_flags out_variable) ...@@ -63,9 +63,7 @@ function(select_nvcc_arch_flags out_variable)
# List of arch names # List of arch names
set(archs_names "Kepler" "Maxwell" "Pascal" "Volta" "Turing" "All" "Manual") set(archs_names "Kepler" "Maxwell" "Pascal" "Volta" "Turing" "All" "Manual")
set(archs_name_default "All") set(archs_name_default "All")
if(NOT CMAKE_CROSSCOMPILING) list(APPEND archs_names "Auto")
list(APPEND archs_names "Auto")
endif()
# set CUDA_ARCH_NAME strings (so it will be seen as dropbox in CMake-Gui) # set CUDA_ARCH_NAME strings (so it will be seen as dropbox in CMake-Gui)
set(CUDA_ARCH_NAME ${archs_name_default} CACHE STRING "Select target NVIDIA GPU achitecture.") set(CUDA_ARCH_NAME ${archs_name_default} CACHE STRING "Select target NVIDIA GPU achitecture.")
...@@ -170,10 +168,7 @@ elseif (${CUDA_VERSION} LESS 11.0) # CUDA 10.x ...@@ -170,10 +168,7 @@ elseif (${CUDA_VERSION} LESS 11.0) # CUDA 10.x
endif() endif()
include_directories(${CUDA_INCLUDE_DIRS}) include_directories(${CUDA_INCLUDE_DIRS})
list(APPEND EXTERNAL_LIBS ${CUDA_LIBRARIES} ${CUDA_rt_LIBRARY})
if(NOT WITH_DSO) if(NOT WITH_DSO)
# TODO(panyx0718): CUPTI only allows DSO?
list(APPEND EXTERNAL_LIBS ${CUDNN_LIBRARY} ${CUPTI_LIBRARY} ${CUDA_CUBLAS_LIBRARIES} ${CUDA_curand_LIBRARY} ${NCCL_LIBRARY})
if(WIN32) if(WIN32)
set_property(GLOBAL PROPERTY CUDA_MODULES ${CUDNN_LIBRARY} ${CUDA_CUBLAS_LIBRARIES} ${CUDA_curand_LIBRARY}) set_property(GLOBAL PROPERTY CUDA_MODULES ${CUDNN_LIBRARY} ${CUDA_CUBLAS_LIBRARIES} ${CUDA_curand_LIBRARY})
endif(WIN32) endif(WIN32)
...@@ -205,25 +200,26 @@ list(APPEND CUDA_NVCC_FLAGS "-w") ...@@ -205,25 +200,26 @@ list(APPEND CUDA_NVCC_FLAGS "-w")
list(APPEND CUDA_NVCC_FLAGS "--expt-relaxed-constexpr") list(APPEND CUDA_NVCC_FLAGS "--expt-relaxed-constexpr")
if (NOT WIN32) if (NOT WIN32)
if(CMAKE_BUILD_TYPE STREQUAL "Debug") if(CMAKE_BUILD_TYPE STREQUAL "Debug")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_DEBUG}) list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_DEBUG})
elseif(CMAKE_BUILD_TYPE STREQUAL "Release") elseif(CMAKE_BUILD_TYPE STREQUAL "Release")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE}) list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE})
elseif(CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo") elseif(CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELWITHDEBINFO}) list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELWITHDEBINFO})
elseif(CMAKE_BUILD_TYPE STREQUAL "MinSizeRel") elseif(CMAKE_BUILD_TYPE STREQUAL "MinSizeRel")
# nvcc 9 does not support -Os. Use Release flags instead # nvcc 9 does not support -Os. Use Release flags instead
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE}) list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE})
endif() endif()
else(NOT WIN32) else(NOT WIN32)
list(APPEND CUDA_NVCC_FLAGS "--compiler-options;/bigobj") list(APPEND CUDA_NVCC_FLAGS "-Xcompiler \"/wd 4244 /wd 4267 /wd 4819\"")
if(CMAKE_BUILD_TYPE STREQUAL "Debug") list(APPEND CUDA_NVCC_FLAGS "--compiler-options;/bigobj")
list(APPEND CUDA_NVCC_FLAGS "-g -G") if(CMAKE_BUILD_TYPE STREQUAL "Debug")
# match the cl's _ITERATOR_DEBUG_LEVEL list(APPEND CUDA_NVCC_FLAGS "-g -G")
list(APPEND CUDA_NVCC_FLAGS "-D_DEBUG") # match the cl's _ITERATOR_DEBUG_LEVEL
elseif(CMAKE_BUILD_TYPE STREQUAL "Release") list(APPEND CUDA_NVCC_FLAGS "-D_DEBUG")
list(APPEND CUDA_NVCC_FLAGS "-O3 -DNDEBUG") elseif(CMAKE_BUILD_TYPE STREQUAL "Release")
else() list(APPEND CUDA_NVCC_FLAGS "-O3 -DNDEBUG")
else()
message(FATAL "Windows only support Release or Debug build now. Please set visual studio build type to Release/Debug, x64 build.") message(FATAL "Windows only support Release or Debug build now. Please set visual studio build type to Release/Debug, x64 build.")
endif() endif()
endif(NOT WIN32) endif(NOT WIN32)
......
cmake/external/anakin.cmake
浏览文件 @ a9ea99d7
...@@ -74,5 +74,3 @@ add_dependencies(anakin_shared extern_anakin) ...@@ -74,5 +74,3 @@ add_dependencies(anakin_shared extern_anakin)
add_library(anakin_saber SHARED IMPORTED GLOBAL) add_library(anakin_saber SHARED IMPORTED GLOBAL)
set_property(TARGET anakin_saber PROPERTY IMPORTED_LOCATION ${ANAKIN_SABER_LIB}) set_property(TARGET anakin_saber PROPERTY IMPORTED_LOCATION ${ANAKIN_SABER_LIB})
add_dependencies(anakin_saber extern_anakin) add_dependencies(anakin_saber extern_anakin)
list(APPEND external_project_dependencies anakin_shared anakin_saber)
cmake/external/any.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
INCLUDE(ExternalProject)
SET(ANY_SOURCE_DIR ${THIRD_PARTY_PATH}/any)
INCLUDE_DIRECTORIES(${ANY_SOURCE_DIR}/src/extern_lib_any)
ExternalProject_Add(
extern_lib_any
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/PaddlePaddle/any.git"
GIT_TAG "15595d8324be9e8a9a80d9ae442fdd12bd66df5d"
PREFIX ${ANY_SOURCE_DIR}
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
BUILD_COMMAND ""
INSTALL_COMMAND ""
TEST_COMMAND ""
)
if (${CMAKE_VERSION} VERSION_LESS "3.3.0")
set(dummyfile ${CMAKE_CURRENT_BINARY_DIR}/lib_any_dummy.c)
file(WRITE ${dummyfile} "const char * dummy_any = \"${dummyfile}\";")
add_library(lib_any STATIC ${dummyfile})
else()
add_library(lib_any INTERFACE)
endif()
add_dependencies(lib_any extern_lib_any)
add_definitions(-DANY_IMPL_ANY_CAST_MOVEABLE)
LIST(APPEND external_project_dependencies lib_any)
cmake/external/boost.cmake
浏览文件 @ a9ea99d7
...@@ -57,5 +57,4 @@ else() ...@@ -57,5 +57,4 @@ else()
endif() endif()
add_dependencies(boost ${BOOST_PROJECT}) add_dependencies(boost ${BOOST_PROJECT})
list(APPEND external_project_dependencies boost)
set(Boost_INCLUDE_DIR ${BOOST_INCLUDE_DIR}) set(Boost_INCLUDE_DIR ${BOOST_INCLUDE_DIR})
cmake/external/brpc.cmake
浏览文件 @ a9ea99d7
...@@ -69,5 +69,3 @@ SET_PROPERTY(TARGET brpc PROPERTY IMPORTED_LOCATION ${BRPC_LIBRARIES}) ...@@ -69,5 +69,3 @@ SET_PROPERTY(TARGET brpc PROPERTY IMPORTED_LOCATION ${BRPC_LIBRARIES})
ADD_DEPENDENCIES(brpc extern_brpc) ADD_DEPENDENCIES(brpc extern_brpc)
add_definitions(-DBRPC_WITH_GLOG) add_definitions(-DBRPC_WITH_GLOG)
LIST(APPEND external_project_dependencies brpc)
cmake/external/cares.cmake
浏览文件 @ a9ea99d7
...@@ -13,7 +13,7 @@ ...@@ -13,7 +13,7 @@
# limitations under the License. # limitations under the License.
# #
IF(MOBILE_INFERENCE OR NOT WITH_DISTRIBUTE) IF(NOT WITH_DISTRIBUTE)
return() return()
ENDIF() ENDIF()
......
cmake/external/cub.cmake
浏览文件 @ a9ea99d7
...@@ -31,5 +31,3 @@ else() ...@@ -31,5 +31,3 @@ else()
endif() endif()
add_dependencies(cub extern_cub) add_dependencies(cub extern_cub)
LIST(APPEND external_project_dependencies cub)
cmake/external/dlpack.cmake
浏览文件 @ a9ea99d7
...@@ -27,5 +27,3 @@ else() ...@@ -27,5 +27,3 @@ else()
endif() endif()
add_dependencies(dlpack extern_dlpack) add_dependencies(dlpack extern_dlpack)
LIST(APPEND external_project_dependencies dlpack)
cmake/external/eigen.cmake
浏览文件 @ a9ea99d7
...@@ -52,5 +52,3 @@ else() ...@@ -52,5 +52,3 @@ else()
endif() endif()
add_dependencies(eigen3 extern_eigen3) add_dependencies(eigen3 extern_eigen3)
LIST(APPEND external_project_dependencies eigen3)
cmake/external/gflags.cmake
浏览文件 @ a9ea99d7
...@@ -61,8 +61,6 @@ ADD_LIBRARY(gflags STATIC IMPORTED GLOBAL) ...@@ -61,8 +61,6 @@ ADD_LIBRARY(gflags STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET gflags PROPERTY IMPORTED_LOCATION ${GFLAGS_LIBRARIES}) SET_PROPERTY(TARGET gflags PROPERTY IMPORTED_LOCATION ${GFLAGS_LIBRARIES})
ADD_DEPENDENCIES(gflags extern_gflags) ADD_DEPENDENCIES(gflags extern_gflags)
LIST(APPEND external_project_dependencies gflags)
# On Windows (including MinGW), the Shlwapi library is used by gflags if available. # On Windows (including MinGW), the Shlwapi library is used by gflags if available.
if (WIN32) if (WIN32)
include(CheckIncludeFileCXX) include(CheckIncludeFileCXX)
...@@ -71,13 +69,3 @@ if (WIN32) ...@@ -71,13 +69,3 @@ if (WIN32)
set_property(GLOBAL PROPERTY OS_DEPENDENCY_MODULES shlwapi.lib) set_property(GLOBAL PROPERTY OS_DEPENDENCY_MODULES shlwapi.lib)
endif(HAVE_SHLWAPI) endif(HAVE_SHLWAPI)
endif (WIN32) endif (WIN32)
IF(WITH_C_API)
INSTALL(DIRECTORY ${GFLAGS_INCLUDE_DIR} DESTINATION third_party/gflags)
IF(ANDROID)
INSTALL(FILES ${GFLAGS_LIBRARIES} DESTINATION third_party/gflags/lib/${ANDROID_ABI})
ELSE()
INSTALL(FILES ${GFLAGS_LIBRARIES} DESTINATION third_party/gflags/lib)
ENDIF()
ENDIF()
cmake/external/glog.cmake
浏览文件 @ a9ea99d7
...@@ -20,20 +20,16 @@ SET(GLOG_INCLUDE_DIR "${GLOG_INSTALL_DIR}/include" CACHE PATH "glog include dire ...@@ -20,20 +20,16 @@ SET(GLOG_INCLUDE_DIR "${GLOG_INSTALL_DIR}/include" CACHE PATH "glog include dire
IF(WIN32) IF(WIN32)
SET(GLOG_LIBRARIES "${GLOG_INSTALL_DIR}/lib/libglog.lib" CACHE FILEPATH "glog library." FORCE) SET(GLOG_LIBRARIES "${GLOG_INSTALL_DIR}/lib/libglog.lib" CACHE FILEPATH "glog library." FORCE)
SET(GLOG_CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /wd4267 /wd4530")
ELSE(WIN32) ELSE(WIN32)
SET(GLOG_LIBRARIES "${GLOG_INSTALL_DIR}/lib/libglog.a" CACHE FILEPATH "glog library." FORCE) SET(GLOG_LIBRARIES "${GLOG_INSTALL_DIR}/lib/libglog.a" CACHE FILEPATH "glog library." FORCE)
SET(GLOG_CMAKE_CXX_FLAGS ${CMAKE_CXX_FLAGS})
ENDIF(WIN32) ENDIF(WIN32)
INCLUDE_DIRECTORIES(${GLOG_INCLUDE_DIR}) INCLUDE_DIRECTORIES(${GLOG_INCLUDE_DIR})
IF(ANDROID AND ${CMAKE_SYSTEM_VERSION} VERSION_LESS "21") SET(GLOG_REPOSITORY "https://github.com/google/glog.git")
# Using the unofficial glog for Android API < 21 SET(GLOG_TAG "v0.3.5")
SET(GLOG_REPOSITORY "https://github.com/Xreki/glog.git")
SET(GLOG_TAG "8a547150548b284382ccb6582408e9140ff2bea8")
ELSE()
SET(GLOG_REPOSITORY "https://github.com/google/glog.git")
SET(GLOG_TAG "v0.3.5")
ENDIF()
ExternalProject_Add( ExternalProject_Add(
extern_glog extern_glog
...@@ -45,7 +41,7 @@ ExternalProject_Add( ...@@ -45,7 +41,7 @@ ExternalProject_Add(
UPDATE_COMMAND "" UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER} CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER} -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS} -DCMAKE_CXX_FLAGS=${GLOG_CMAKE_CXX_FLAGS}
-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE} -DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG} -DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS} -DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
...@@ -76,14 +72,3 @@ ADD_LIBRARY(glog STATIC IMPORTED GLOBAL) ...@@ -76,14 +72,3 @@ ADD_LIBRARY(glog STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET glog PROPERTY IMPORTED_LOCATION ${GLOG_LIBRARIES}) SET_PROPERTY(TARGET glog PROPERTY IMPORTED_LOCATION ${GLOG_LIBRARIES})
ADD_DEPENDENCIES(glog extern_glog gflags) ADD_DEPENDENCIES(glog extern_glog gflags)
LINK_LIBRARIES(glog gflags) LINK_LIBRARIES(glog gflags)
LIST(APPEND external_project_dependencies glog)
IF(WITH_C_API)
INSTALL(DIRECTORY ${GLOG_INCLUDE_DIR} DESTINATION third_party/glog)
IF(ANDROID)
INSTALL(FILES ${GLOG_LIBRARIES} DESTINATION third_party/glog/lib/${ANDROID_ABI})
ELSE()
INSTALL(FILES ${GLOG_LIBRARIES} DESTINATION third_party/glog/lib)
ENDIF()
ENDIF()
cmake/external/grpc.cmake
浏览文件 @ a9ea99d7
...@@ -13,7 +13,7 @@ ...@@ -13,7 +13,7 @@
# limitations under the License. # limitations under the License.
# #
IF(MOBILE_INFERENCE OR NOT WITH_DISTRIBUTE) IF(NOT WITH_DISTRIBUTE)
return() return()
ENDIF() ENDIF()
......
cmake/external/gtest.cmake
浏览文件 @ a9ea99d7
...@@ -79,5 +79,4 @@ IF(WITH_TESTING OR (WITH_DISTRIBUTE AND NOT WITH_GRPC)) ...@@ -79,5 +79,4 @@ IF(WITH_TESTING OR (WITH_DISTRIBUTE AND NOT WITH_GRPC))
SET_PROPERTY(TARGET gtest_main PROPERTY IMPORTED_LOCATION ${GTEST_MAIN_LIBRARIES}) SET_PROPERTY(TARGET gtest_main PROPERTY IMPORTED_LOCATION ${GTEST_MAIN_LIBRARIES})
ADD_DEPENDENCIES(gtest_main extern_gtest) ADD_DEPENDENCIES(gtest_main extern_gtest)
LIST(APPEND external_project_dependencies gtest gtest_main)
ENDIF(WITH_TESTING OR (WITH_DISTRIBUTE AND NOT WITH_GRPC)) ENDIF(WITH_TESTING OR (WITH_DISTRIBUTE AND NOT WITH_GRPC))
cmake/external/gzstream.cmake
浏览文件 @ a9ea99d7
...@@ -13,10 +13,6 @@ ...@@ -13,10 +13,6 @@
# limitations under the License. # limitations under the License.
# #
IF(MOBILE_INFERENCE)
return()
ENDIF()
include (ExternalProject) include (ExternalProject)
# NOTE: gzstream is needed when linking with ctr reader. # NOTE: gzstream is needed when linking with ctr reader.
......
cmake/external/leveldb.cmake
浏览文件 @ a9ea99d7
...@@ -39,6 +39,3 @@ ADD_DEPENDENCIES(extern_leveldb snappy) ...@@ -39,6 +39,3 @@ ADD_DEPENDENCIES(extern_leveldb snappy)
ADD_LIBRARY(leveldb STATIC IMPORTED GLOBAL) ADD_LIBRARY(leveldb STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET leveldb PROPERTY IMPORTED_LOCATION ${LEVELDB_LIBRARIES}) SET_PROPERTY(TARGET leveldb PROPERTY IMPORTED_LOCATION ${LEVELDB_LIBRARIES})
ADD_DEPENDENCIES(leveldb extern_leveldb) ADD_DEPENDENCIES(leveldb extern_leveldb)
LIST(APPEND external_project_dependencies leveldb)
cmake/external/libmct.cmake
浏览文件 @ a9ea99d7
...@@ -72,7 +72,4 @@ else() ...@@ -72,7 +72,4 @@ else()
add_library(libmct INTERFACE) add_library(libmct INTERFACE)
endif() endif()
#ADD_LIBRARY(libmct SHARED IMPORTED GLOBAL)
ADD_DEPENDENCIES(libmct ${LIBMCT_PROJECT}) ADD_DEPENDENCIES(libmct ${LIBMCT_PROJECT})
LIST(APPEND external_project_dependencies libmct)
cmake/external/libxsmm.cmake
浏览文件 @ a9ea99d7
...@@ -19,8 +19,8 @@ IF(NOT WITH_LIBXSMM) ...@@ -19,8 +19,8 @@ IF(NOT WITH_LIBXSMM)
return() return()
ENDIF() ENDIF()
IF(WIN32 OR APPLE OR ANDROID OR IOS) IF(WIN32 OR APPLE)
MESSAGE(WARNING "Windows, Mac or Mobile are not supported with libxsmm in Paddle yet.") MESSAGE(WARNING "Windows, Mac are not supported with libxsmm in Paddle yet.")
SET(WITH_LIBXSMM OFF CACHE STRING "Disable LIBXSMM" FORCE) SET(WITH_LIBXSMM OFF CACHE STRING "Disable LIBXSMM" FORCE)
return() return()
ENDIF() ENDIF()
...@@ -53,5 +53,3 @@ MESSAGE(STATUS "Libxsmm library: ${LIBXSMM_LIBS}") ...@@ -53,5 +53,3 @@ MESSAGE(STATUS "Libxsmm library: ${LIBXSMM_LIBS}")
include_directories(${LIBXSMM_INCLUDE_DIR}) include_directories(${LIBXSMM_INCLUDE_DIR})
ADD_DEFINITIONS(-DPADDLE_WITH_LIBXSMM) ADD_DEFINITIONS(-DPADDLE_WITH_LIBXSMM)
ADD_DEPENDENCIES(libxsmm extern_libxsmm) ADD_DEPENDENCIES(libxsmm extern_libxsmm)
LIST(APPEND external_project_dependencies libxsmm)
cmake/external/mkldnn.cmake
浏览文件 @ a9ea99d7
...@@ -49,6 +49,8 @@ IF(NOT WIN32) ...@@ -49,6 +49,8 @@ IF(NOT WIN32)
SET(MKLDNN_FLAG "${MKLDNN_FLAG} -Wno-unused-result -Wno-unused-value") SET(MKLDNN_FLAG "${MKLDNN_FLAG} -Wno-unused-result -Wno-unused-value")
SET(MKLDNN_CFLAG "${CMAKE_C_FLAGS} ${MKLDNN_FLAG}") SET(MKLDNN_CFLAG "${CMAKE_C_FLAGS} ${MKLDNN_FLAG}")
SET(MKLDNN_CXXFLAG "${CMAKE_CXX_FLAGS} ${MKLDNN_FLAG}") SET(MKLDNN_CXXFLAG "${CMAKE_CXX_FLAGS} ${MKLDNN_FLAG}")
ELSE()
SET(MKLDNN_CXXFLAG "${CMAKE_CXX_FLAGS} /EHsc")
ENDIF(NOT WIN32) ENDIF(NOT WIN32)
ExternalProject_Add( ExternalProject_Add(
...@@ -61,7 +63,6 @@ ExternalProject_Add( ...@@ -61,7 +63,6 @@ ExternalProject_Add(
UPDATE_COMMAND "" UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER} CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
CMAKE_ARGS -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER} CMAKE_ARGS -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
CMAKE_ARGS -DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
CMAKE_ARGS -DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE} CMAKE_ARGS -DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
CMAKE_ARGS -DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG} CMAKE_ARGS -DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
CMAKE_ARGS -DCMAKE_C_FLAGS=${CMAKE_C_FLAGS} CMAKE_ARGS -DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
...@@ -88,7 +89,6 @@ SET_PROPERTY(TARGET shared_mkldnn PROPERTY IMPORTED_LOCATION ${MKLDNN_LIB}) ...@@ -88,7 +89,6 @@ SET_PROPERTY(TARGET shared_mkldnn PROPERTY IMPORTED_LOCATION ${MKLDNN_LIB})
ADD_DEPENDENCIES(shared_mkldnn ${MKLDNN_PROJECT}) ADD_DEPENDENCIES(shared_mkldnn ${MKLDNN_PROJECT})
MESSAGE(STATUS "MKLDNN library: ${MKLDNN_LIB}") MESSAGE(STATUS "MKLDNN library: ${MKLDNN_LIB}")
add_definitions(-DPADDLE_WITH_MKLDNN) add_definitions(-DPADDLE_WITH_MKLDNN)
LIST(APPEND external_project_dependencies shared_mkldnn)
# generate a static dummy target to track mkldnn dependencies # generate a static dummy target to track mkldnn dependencies
# for cc_library(xxx SRCS xxx.c DEPS mkldnn) # for cc_library(xxx SRCS xxx.c DEPS mkldnn)
...@@ -110,7 +110,3 @@ else(WIN32) ...@@ -110,7 +110,3 @@ else(WIN32)
endif(WIN32) endif(WIN32)
ADD_CUSTOM_TARGET(mkldnn_shared_lib ALL DEPENDS ${MKLDNN_SHARED_LIB}) ADD_CUSTOM_TARGET(mkldnn_shared_lib ALL DEPENDS ${MKLDNN_SHARED_LIB})
ADD_DEPENDENCIES(mkldnn_shared_lib ${MKLDNN_PROJECT} mkldnn) ADD_DEPENDENCIES(mkldnn_shared_lib ${MKLDNN_PROJECT} mkldnn)
IF(WITH_C_API)
INSTALL(FILES ${MKLDNN_SHARED_LIB} DESTINATION lib)
ENDIF()
cmake/external/mklml.cmake
浏览文件 @ a9ea99d7
...@@ -39,8 +39,10 @@ IF(WIN32) ...@@ -39,8 +39,10 @@ IF(WIN32)
SET(MKLML_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5md.lib) SET(MKLML_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5md.lib)
SET(MKLML_SHARED_LIB ${MKLML_LIB_DIR}/mklml.dll) SET(MKLML_SHARED_LIB ${MKLML_LIB_DIR}/mklml.dll)
SET(MKLML_SHARED_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5md.dll) SET(MKLML_SHARED_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5md.dll)
ELSE() ELSE()
SET(MKLML_VER "mklml_lnx_${TIME_VERSION}" CACHE STRING "" FORCE) #TODO(intel-huying):
# Now enable Erf function in mklml library temporarily, it will be updated as offical version later.
SET(MKLML_VER "Glibc225_vsErf_mklml_lnx_${TIME_VERSION}" CACHE STRING "" FORCE)
SET(MKLML_URL "http://paddlepaddledeps.cdn.bcebos.com/${MKLML_VER}.tgz" CACHE STRING "" FORCE) SET(MKLML_URL "http://paddlepaddledeps.cdn.bcebos.com/${MKLML_VER}.tgz" CACHE STRING "" FORCE)
SET(MKLML_LIB ${MKLML_LIB_DIR}/libmklml_intel.so) SET(MKLML_LIB ${MKLML_LIB_DIR}/libmklml_intel.so)
SET(MKLML_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5.so) SET(MKLML_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5.so)
...@@ -73,8 +75,3 @@ INCLUDE_DIRECTORIES(${MKLML_INC_DIR}) ...@@ -73,8 +75,3 @@ INCLUDE_DIRECTORIES(${MKLML_INC_DIR})
ADD_LIBRARY(mklml SHARED IMPORTED GLOBAL) ADD_LIBRARY(mklml SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET mklml PROPERTY IMPORTED_LOCATION ${MKLML_LIB}) SET_PROPERTY(TARGET mklml PROPERTY IMPORTED_LOCATION ${MKLML_LIB})
ADD_DEPENDENCIES(mklml ${MKLML_PROJECT}) ADD_DEPENDENCIES(mklml ${MKLML_PROJECT})
LIST(APPEND external_project_dependencies mklml)
IF(WITH_C_API)
INSTALL(FILES ${MKLML_LIB} ${MKLML_IOMP_LIB} DESTINATION lib)
ENDIF()
cmake/external/ngraph.cmake
浏览文件 @ a9ea99d7
...@@ -37,7 +37,7 @@ INCLUDE(GNUInstallDirs) ...@@ -37,7 +37,7 @@ INCLUDE(GNUInstallDirs)
INCLUDE(ExternalProject) INCLUDE(ExternalProject)
SET(NGRAPH_PROJECT "extern_ngraph") SET(NGRAPH_PROJECT "extern_ngraph")
SET(NGRAPH_GIT_TAG "20bd8bbc79ae3a81c57313846a2be7313e5d1dab") SET(NGRAPH_GIT_TAG "a444f7a959b7d87f2c117c9b57a4c387759e481e")
SET(NGRAPH_SOURCES_DIR ${THIRD_PARTY_PATH}/ngraph) SET(NGRAPH_SOURCES_DIR ${THIRD_PARTY_PATH}/ngraph)
SET(NGRAPH_INSTALL_DIR ${THIRD_PARTY_PATH}/install/ngraph) SET(NGRAPH_INSTALL_DIR ${THIRD_PARTY_PATH}/install/ngraph)
SET(NGRAPH_INC_DIR ${NGRAPH_INSTALL_DIR}/include) SET(NGRAPH_INC_DIR ${NGRAPH_INSTALL_DIR}/include)
...@@ -69,7 +69,7 @@ ExternalProject_Add( ...@@ -69,7 +69,7 @@ ExternalProject_Add(
CMAKE_ARGS -DNGRAPH_DEX_ONLY=TRUE CMAKE_ARGS -DNGRAPH_DEX_ONLY=TRUE
CMAKE_ARGS -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE} CMAKE_ARGS -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE}
CMAKE_ARGS -DMKLDNN_INCLUDE_DIR=${MKLDNN_INC_DIR} CMAKE_ARGS -DMKLDNN_INCLUDE_DIR=${MKLDNN_INC_DIR}
CMAKE_ARGS -DMKLDNN_LIB_DIR=${MKLDNN_INSTALL_DIR}/lib CMAKE_ARGS -DMKLDNN_LIB_DIR=${MKLDNN_INSTALL_DIR}/${CMAKE_INSTALL_LIBDIR}
CMAKE_ARGS -DMKLML_LIB_DIR=${MKLML_INSTALL_DIR}/lib CMAKE_ARGS -DMKLML_LIB_DIR=${MKLML_INSTALL_DIR}/lib
) )
...@@ -77,4 +77,3 @@ add_dependencies(ngraph ${NGRAPH_PROJECT}) ...@@ -77,4 +77,3 @@ add_dependencies(ngraph ${NGRAPH_PROJECT})
target_compile_definitions(ngraph INTERFACE -DPADDLE_WITH_NGRAPH) target_compile_definitions(ngraph INTERFACE -DPADDLE_WITH_NGRAPH)
target_include_directories(ngraph INTERFACE ${NGRAPH_INC_DIR}) target_include_directories(ngraph INTERFACE ${NGRAPH_INC_DIR})
target_link_libraries(ngraph INTERFACE ${NGRAPH_SHARED_LIB}) target_link_libraries(ngraph INTERFACE ${NGRAPH_SHARED_LIB})
LIST(APPEND external_project_dependencies ngraph)
cmake/external/nnpack.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Find the NNPACK library
# NNPACK_ROOT - where to find NNPACK include and library.
#
set(NNPACK_FOUND OFF)
set(NNPACK_ROOT $ENV{NNPACK_ROOT} CACHE PATH "Folder contains NNPACK")
find_path(NNPACK_INC_DIR nnpack.h PATHS ${NNPACK_ROOT}/include)
find_library(NNPACK_LIB NAMES nnpack PATHS ${NNPACK_ROOT}/lib)
find_library(PTHREADPOOL_LIB NAMES pthreadpool PATHS ${NNPACK_ROOT}/lib)
find_library(NNPACK_UKERNELS_LIB NAMES nnpack_ukernels PATHS ${NNPACK_ROOT}/lib)
find_library(NNPACK_CPUFEATURES_LIB NAMES cpufeatures PATHS ${NNPACK_ROOT}/lib)
if(NNPACK_INC_DIR AND NNPACK_LIB AND PTHREADPOOL_LIB)
set(NNPACK_FOUND ON)
INCLUDE_DIRECTORIES(${NNPACK_INC_DIR})
set(NNPACK_LIBS)
list(APPEND NNPACK_LIBS ${NNPACK_LIB} ${PTHREADPOOL_LIB})
if (NNPACK_UKERNELS_LIB)
list(APPEND NNPACK_LIBS ${NNPACK_UKERNELS_LIB})
endif()
if (NNPACK_CPUFEATURES_LIB)
list(APPEND NNPACK_LIBS ${NNPACK_CPUFEATURES_LIB})
endif()
if(NOT ANDROID)
list(APPEND NNPACK_LIBS "rt")
endif()
else()
message(FATAL_ERROR "Cannot find NNPACK in (${NNPACK_ROOT})")
endif()
cmake/external/openblas.cmake
浏览文件 @ a9ea99d7
...@@ -11,11 +11,6 @@ ...@@ -11,11 +11,6 @@
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and # See the License for the specific language governing permissions and
# limitations under the License. # limitations under the License.
IF(USE_EIGEN_FOR_BLAS)
return()
ENDIF(USE_EIGEN_FOR_BLAS)
INCLUDE(cblas) INCLUDE(cblas)
IF(NOT ${CBLAS_FOUND}) IF(NOT ${CBLAS_FOUND})
...@@ -40,38 +35,12 @@ IF(NOT ${CBLAS_FOUND}) ...@@ -40,38 +35,12 @@ IF(NOT ${CBLAS_FOUND})
SET(OPENBLAS_CC "${CMAKE_C_COMPILER} -Wno-unused-but-set-variable -Wno-unused-variable") SET(OPENBLAS_CC "${CMAKE_C_COMPILER} -Wno-unused-but-set-variable -Wno-unused-variable")
SET(OPENBLAS_COMMIT "v0.2.20") SET(OPENBLAS_COMMIT "v0.2.20")
IF(CMAKE_CROSSCOMPILING) IF(APPLE)
SET(OPTIONAL_ARGS HOSTCC=${HOST_C_COMPILER}) SET(OPENBLAS_CC "${CMAKE_C_COMPILER} -isysroot ${CMAKE_OSX_SYSROOT}")
GET_FILENAME_COMPONENT(CROSS_SUFFIX ${CMAKE_C_COMPILER} DIRECTORY) ENDIF()
SET(CROSS_SUFFIX ${CROSS_SUFFIX}/) SET(OPTIONAL_ARGS "")
IF(ANDROID) IF(CMAKE_SYSTEM_PROCESSOR MATCHES "^x86(_64)?$")
IF(ANDROID_ABI MATCHES "^armeabi(-v7a)?$") SET(OPTIONAL_ARGS DYNAMIC_ARCH=1 NUM_THREADS=64)
# use softfp
SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV7 ARM_SOFTFP_ABI=1 USE_THREAD=0)
ELSEIF(ANDROID_ABI STREQUAL "arm64-v8a")
SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV8 BINARY=64 USE_THREAD=0)
ENDIF()
ELSEIF(IOS)
IF(CMAKE_OSX_ARCHITECTURES MATCHES "arm64")
SET(OPENBLAS_CC "${OPENBLAS_CC} ${CMAKE_C_FLAGS} -isysroot ${CMAKE_OSX_SYSROOT}")
SET(OPENBLAS_CC "${OPENBLAS_CC} -arch arm64")
SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV8 BINARY=64 USE_THREAD=0 CROSS_SUFFIX=${CROSS_SUFFIX})
ELSE()
MESSAGE(FATAL_ERROR "OpenBLAS only support arm64 architectures on iOS. "
"You can set IOS_USE_VECLIB_FOR_BLAS=ON or USE_EIGEN_FOR_BLAS=ON to use other blas library instead.")
ENDIF()
ELSEIF(RPI)
# use hardfp
SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} TARGET=ARMV7 USE_THREAD=0)
ENDIF()
ELSE()
IF(APPLE)
SET(OPENBLAS_CC "${CMAKE_C_COMPILER} -isysroot ${CMAKE_OSX_SYSROOT}")
ENDIF()
SET(OPTIONAL_ARGS "")
IF(CMAKE_SYSTEM_PROCESSOR MATCHES "^x86(_64)?$")
SET(OPTIONAL_ARGS DYNAMIC_ARCH=1 NUM_THREADS=64)
ENDIF()
ENDIF() ENDIF()
SET(COMMON_ARGS CC=${OPENBLAS_CC} NO_SHARED=1 NO_LAPACK=1 libs) SET(COMMON_ARGS CC=${OPENBLAS_CC} NO_SHARED=1 NO_LAPACK=1 libs)
...@@ -92,25 +61,6 @@ IF(NOT ${CBLAS_FOUND}) ...@@ -92,25 +61,6 @@ IF(NOT ${CBLAS_FOUND})
ELSE() ELSE()
ENDIF(NOT WIN32) ENDIF(NOT WIN32)
SET(CBLAS_PROVIDER openblas) SET(CBLAS_PROVIDER openblas)
IF(WITH_C_API)
INSTALL(DIRECTORY ${CBLAS_INC_DIR} DESTINATION third_party/openblas)
# Because libopenblas.a is a symbolic link of another library, thus need to
# install the whole directory.
IF(ANDROID)
SET(TMP_INSTALL_DIR third_party/openblas/lib/${ANDROID_ABI})
ELSE()
SET(TMP_INSTALL_DIR third_party/openblas/lib)
ENDIF()
INSTALL(CODE "execute_process(
COMMAND ${CMAKE_COMMAND} -E copy_directory ${CBLAS_INSTALL_DIR}/lib
${CMAKE_INSTALL_PREFIX}/${TMP_INSTALL_DIR}
)"
)
INSTALL(CODE "MESSAGE(STATUS \"Installing: \"
\"${CBLAS_INSTALL_DIR}/lib -> ${CMAKE_INSTALL_PREFIX}/${TMP_INSTALL_DIR}\"
)"
)
ENDIF()
ENDIF(NOT ${CBLAS_FOUND}) ENDIF(NOT ${CBLAS_FOUND})
MESSAGE(STATUS "BLAS library: ${CBLAS_LIBRARIES}") MESSAGE(STATUS "BLAS library: ${CBLAS_LIBRARIES}")
...@@ -136,7 +86,6 @@ ENDIF() ...@@ -136,7 +86,6 @@ ENDIF()
IF(NOT ${CBLAS_FOUND}) IF(NOT ${CBLAS_FOUND})
ADD_DEPENDENCIES(cblas extern_openblas) ADD_DEPENDENCIES(cblas extern_openblas)
LIST(APPEND external_project_dependencies cblas)
ELSE() ELSE()
IF("${CBLAS_PROVIDER}" STREQUAL "MKLML") IF("${CBLAS_PROVIDER}" STREQUAL "MKLML")
ADD_DEPENDENCIES(cblas mklml) ADD_DEPENDENCIES(cblas mklml)
......
cmake/external/protobuf.cmake
浏览文件 @ a9ea99d7
...@@ -129,7 +129,6 @@ macro(PROMPT_PROTOBUF_LIB) ...@@ -129,7 +129,6 @@ macro(PROMPT_PROTOBUF_LIB)
ADD_DEPENDENCIES(protoc ${dep}) ADD_DEPENDENCIES(protoc ${dep})
ENDFOREACH() ENDFOREACH()
LIST(APPEND external_project_dependencies protobuf)
RETURN() RETURN()
endmacro() endmacro()
macro(SET_PROTOBUF_VERSION) macro(SET_PROTOBUF_VERSION)
...@@ -204,15 +203,6 @@ FUNCTION(build_protobuf TARGET_NAME BUILD_FOR_HOST) ...@@ -204,15 +203,6 @@ FUNCTION(build_protobuf TARGET_NAME BUILD_FOR_HOST)
SET(PROTOBUF_REPO "https://github.com/google/protobuf.git") SET(PROTOBUF_REPO "https://github.com/google/protobuf.git")
SET(PROTOBUF_TAG "9f75c5aa851cd877fb0d93ccc31b8567a6706546") SET(PROTOBUF_TAG "9f75c5aa851cd877fb0d93ccc31b8567a6706546")
IF(MOBILE_INFERENCE)
# The reason why the official version is not used is described in
# https://github.com/PaddlePaddle/Paddle/issues/6114
SET(PROTOBUF_REPO "https://github.com/qingqing01/protobuf.git")
SET(PROTOBUF_TAG "v3.2.0")
IF(NOT BUILD_FOR_HOST)
SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} "-Dprotobuf_BUILD_PROTOC_BINARIES=OFF")
ENDIF()
ENDIF()
ExternalProject_Add( ExternalProject_Add(
${TARGET_NAME} ${TARGET_NAME}
...@@ -240,19 +230,7 @@ FUNCTION(build_protobuf TARGET_NAME BUILD_FOR_HOST) ...@@ -240,19 +230,7 @@ FUNCTION(build_protobuf TARGET_NAME BUILD_FOR_HOST)
) )
ENDFUNCTION() ENDFUNCTION()
IF(NOT MOBILE_INFERENCE) SET(PROTOBUF_VERSION 3.1.0)
SET(PROTOBUF_VERSION 3.1)
ELSE()
SET(PROTOBUF_VERSION 3.2)
ENDIF()
IF(CMAKE_CROSSCOMPILING)
build_protobuf(protobuf_host TRUE)
LIST(APPEND external_project_dependencies protobuf_host)
SET(PROTOBUF_PROTOC_EXECUTABLE ${protobuf_host_PROTOC_EXECUTABLE}
CACHE FILEPATH "protobuf executable." FORCE)
ENDIF()
IF(NOT PROTOBUF_FOUND) IF(NOT PROTOBUF_FOUND)
build_protobuf(extern_protobuf FALSE) build_protobuf(extern_protobuf FALSE)
...@@ -266,20 +244,7 @@ IF(NOT PROTOBUF_FOUND) ...@@ -266,20 +244,7 @@ IF(NOT PROTOBUF_FOUND)
SET(PROTOBUF_PROTOC_LIBRARY ${extern_protobuf_PROTOC_LIBRARY} SET(PROTOBUF_PROTOC_LIBRARY ${extern_protobuf_PROTOC_LIBRARY}
CACHE FILEPATH "protoc library." FORCE) CACHE FILEPATH "protoc library." FORCE)
IF(WITH_C_API) SET(PROTOBUF_PROTOC_EXECUTABLE ${extern_protobuf_PROTOC_EXECUTABLE}
INSTALL(DIRECTORY ${PROTOBUF_INCLUDE_DIR} DESTINATION third_party/protobuf) CACHE FILEPATH "protobuf executable." FORCE)
IF(ANDROID) PROMPT_PROTOBUF_LIB(extern_protobuf)
INSTALL(FILES ${PROTOBUF_LITE_LIBRARY} DESTINATION third_party/protobuf/lib/${ANDROID_ABI})
ELSE()
INSTALL(FILES ${PROTOBUF_LITE_LIBRARY} DESTINATION third_party/protobuf/lib)
ENDIF()
ENDIF()
IF(CMAKE_CROSSCOMPILING)
PROMPT_PROTOBUF_LIB(protobuf_host extern_protobuf)
ELSE()
SET(PROTOBUF_PROTOC_EXECUTABLE ${extern_protobuf_PROTOC_EXECUTABLE}
CACHE FILEPATH "protobuf executable." FORCE)
PROMPT_PROTOBUF_LIB(extern_protobuf)
ENDIF()
ENDIF(NOT PROTOBUF_FOUND) ENDIF(NOT PROTOBUF_FOUND)
cmake/external/pslib.cmake
浏览文件 @ a9ea99d7
...@@ -70,8 +70,3 @@ ExternalProject_Add( ...@@ -70,8 +70,3 @@ ExternalProject_Add(
ADD_LIBRARY(pslib SHARED IMPORTED GLOBAL) ADD_LIBRARY(pslib SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET pslib PROPERTY IMPORTED_LOCATION ${PSLIB_LIB}) SET_PROPERTY(TARGET pslib PROPERTY IMPORTED_LOCATION ${PSLIB_LIB})
ADD_DEPENDENCIES(pslib ${PSLIB_PROJECT}) ADD_DEPENDENCIES(pslib ${PSLIB_PROJECT})
LIST(APPEND external_project_dependencies pslib)
IF(WITH_C_API)
INSTALL(FILES ${PSLIB_LIB} ${PSLIB_IOMP_LIB} DESTINATION lib)
ENDIF()
cmake/external/pslib_brpc.cmake
浏览文件 @ a9ea99d7
...@@ -70,8 +70,3 @@ ExternalProject_Add( ...@@ -70,8 +70,3 @@ ExternalProject_Add(
ADD_LIBRARY(pslib_brpc SHARED IMPORTED GLOBAL) ADD_LIBRARY(pslib_brpc SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET pslib_brpc PROPERTY IMPORTED_LOCATION ${PSLIB_BRPC_LIB}) SET_PROPERTY(TARGET pslib_brpc PROPERTY IMPORTED_LOCATION ${PSLIB_BRPC_LIB})
ADD_DEPENDENCIES(pslib_brpc ${PSLIB_BRPC_PROJECT}) ADD_DEPENDENCIES(pslib_brpc ${PSLIB_BRPC_PROJECT})
LIST(APPEND external_project_dependencies pslib_brpc)
IF(WITH_C_API)
INSTALL(FILES ${PSLIB_BRPC_LIB} ${PSLIB_BRPC_IOMP_LIB} DESTINATION lib)
ENDIF()
cmake/external/snappy.cmake
浏览文件 @ a9ea99d7
...@@ -12,10 +12,6 @@ ...@@ -12,10 +12,6 @@
# See the License for the specific language governing permissions and # See the License for the specific language governing permissions and
# limitations under the License. # limitations under the License.
if(MOBILE_INFERENCE OR RPI)
return()
endif()
include (ExternalProject) include (ExternalProject)
# NOTE: snappy is needed when linking with recordio # NOTE: snappy is needed when linking with recordio
...@@ -24,6 +20,12 @@ set(SNAPPY_SOURCES_DIR ${THIRD_PARTY_PATH}/snappy) ...@@ -24,6 +20,12 @@ set(SNAPPY_SOURCES_DIR ${THIRD_PARTY_PATH}/snappy)
set(SNAPPY_INSTALL_DIR ${THIRD_PARTY_PATH}/install/snappy) set(SNAPPY_INSTALL_DIR ${THIRD_PARTY_PATH}/install/snappy)
set(SNAPPY_INCLUDE_DIR "${SNAPPY_INSTALL_DIR}/include" CACHE PATH "snappy include directory." FORCE) set(SNAPPY_INCLUDE_DIR "${SNAPPY_INSTALL_DIR}/include" CACHE PATH "snappy include directory." FORCE)
if(WIN32)
SET(SNAPPY_CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /wd4244 /wd4267")
else()
SET(SNAPPY_CMAKE_CXX_FLAGS ${CMAKE_CXX_FLAGS})
endif()
ExternalProject_Add( ExternalProject_Add(
extern_snappy extern_snappy
GIT_REPOSITORY "https://github.com/google/snappy" GIT_REPOSITORY "https://github.com/google/snappy"
...@@ -35,7 +37,7 @@ ExternalProject_Add( ...@@ -35,7 +37,7 @@ ExternalProject_Add(
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS} -DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
-DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG} -DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}
-DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE} -DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS} -DCMAKE_CXX_FLAGS=${SNAPPY_CMAKE_CXX_FLAGS}
-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE} -DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG} -DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
-DCMAKE_INSTALL_PREFIX=${SNAPPY_INSTALL_DIR} -DCMAKE_INSTALL_PREFIX=${SNAPPY_INSTALL_DIR}
......
cmake/external/snappystream.cmake
浏览文件 @ a9ea99d7
...@@ -12,10 +12,6 @@ ...@@ -12,10 +12,6 @@
# See the License for the specific language governing permissions and # See the License for the specific language governing permissions and
# limitations under the License. # limitations under the License.
IF(MOBILE_INFERENCE OR RPI)
return()
ENDIF()
include (ExternalProject) include (ExternalProject)
set(SNAPPYSTREAM_SOURCES_DIR ${THIRD_PARTY_PATH}/snappy_stream) set(SNAPPYSTREAM_SOURCES_DIR ${THIRD_PARTY_PATH}/snappy_stream)
......
cmake/external/swig.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
IF(NOT WITH_SWIG_PY)
return()
ENDIF()
FIND_PACKAGE(SWIG)
IF(NOT SWIG_FOUND)
# build swig as an external project
INCLUDE(ExternalProject)
SET(SWIG_SOURCES_DIR ${THIRD_PARTY_PATH}/swig)
SET(SWIG_INSTALL_DIR ${THIRD_PARTY_PATH}/install/swig)
SET(SWIG_TARGET_VERSION "3.0.2")
SET(SWIG_DOWNLOAD_SRC_MD5 "62f9b0d010cef36a13a010dc530d0d41")
SET(SWIG_DOWNLOAD_WIN_MD5 "3f18de4fc09ab9abb0d3be37c11fbc8f")
IF(WIN32)
# swig.exe available as pre-built binary on Windows:
ExternalProject_Add(swig
URL http://prdownloads.sourceforge.net/swig/swigwin-${SWIG_TARGET_VERSION}.zip
URL_MD5 ${SWIG_DOWNLOAD_WIN_MD5}
SOURCE_DIR ${SWIG_SOURCES_DIR}
CONFIGURE_COMMAND ""
BUILD_COMMAND ""
INSTALL_COMMAND ""
UPDATE_COMMAND ""
)
SET(SWIG_DIR ${SWIG_SOURCES_DIR} CACHE FILEPATH "SWIG Directory" FORCE)
SET(SWIG_EXECUTABLE ${SWIG_SOURCES_DIR}/swig.exe CACHE FILEPATH "SWIG Executable" FORCE)
ELSE(WIN32)
# swig uses bison find it by cmake and pass it down
FIND_PACKAGE(BISON)
# From SWIG configure
ExternalProject_Add(swig
GIT_REPOSITORY https://github.com/swig/swig.git
GIT_TAG rel-3.0.10
PREFIX ${SWIG_SOURCES_DIR}
CONFIGURE_COMMAND cd <SOURCE_DIR> && ./autogen.sh && ./configure
--prefix=${SWIG_INSTALL_DIR} --without-pcre
BUILD_COMMAND cd <SOURCE_DIR> && make
INSTALL_COMMAND cd <SOURCE_DIR> && make install
UPDATE_COMMAND ""
)
SET(SWIG_DIR ${SWIG_INSTALL_DIR}/share/swig/${SWIG_TARGET_VERSION})
SET(SWIG_EXECUTABLE ${SWIG_INSTALL_DIR}/bin/swig)
ENDIF(WIN32)
LIST(APPEND external_project_dependencies swig)
ENDIF(NOT SWIG_FOUND)
cmake/external/threadpool.cmake
浏览文件 @ a9ea99d7
...@@ -26,5 +26,3 @@ else() ...@@ -26,5 +26,3 @@ else()
endif() endif()
add_dependencies(simple_threadpool extern_threadpool) add_dependencies(simple_threadpool extern_threadpool)
LIST(APPEND external_project_dependencies simple_threadpool)
cmake/external/warpctc.cmake
浏览文件 @ a9ea99d7
...@@ -12,10 +12,6 @@ ...@@ -12,10 +12,6 @@
# See the License for the specific language governing permissions and # See the License for the specific language governing permissions and
# limitations under the License. # limitations under the License.
IF(MOBILE_INFERENCE)
return()
ENDIF()
INCLUDE(ExternalProject) INCLUDE(ExternalProject)
SET(WARPCTC_SOURCES_DIR ${THIRD_PARTY_PATH}/warpctc) SET(WARPCTC_SOURCES_DIR ${THIRD_PARTY_PATH}/warpctc)
...@@ -87,5 +83,3 @@ INCLUDE_DIRECTORIES(${THIRD_PARTY_PATH}/install) # For Paddle code to include wa ...@@ -87,5 +83,3 @@ INCLUDE_DIRECTORIES(${THIRD_PARTY_PATH}/install) # For Paddle code to include wa
ADD_LIBRARY(warpctc SHARED IMPORTED GLOBAL) ADD_LIBRARY(warpctc SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET warpctc PROPERTY IMPORTED_LOCATION ${WARPCTC_LIBRARIES}) SET_PROPERTY(TARGET warpctc PROPERTY IMPORTED_LOCATION ${WARPCTC_LIBRARIES})
ADD_DEPENDENCIES(warpctc extern_warpctc) ADD_DEPENDENCIES(warpctc extern_warpctc)
LIST(APPEND external_project_dependencies warpctc)
cmake/external/xbyak.cmake
浏览文件 @ a9ea99d7
...@@ -55,4 +55,3 @@ else() ...@@ -55,4 +55,3 @@ else()
endif() endif()
add_dependencies(xbyak ${XBYAK_PROJECT}) add_dependencies(xbyak ${XBYAK_PROJECT})
list(APPEND external_project_dependencies xbyak)
cmake/external/xxhash.cmake
浏览文件 @ a9ea99d7
...@@ -71,14 +71,3 @@ add_library(xxhash STATIC IMPORTED GLOBAL) ...@@ -71,14 +71,3 @@ add_library(xxhash STATIC IMPORTED GLOBAL)
set_property(TARGET xxhash PROPERTY IMPORTED_LOCATION ${XXHASH_LIBRARIES}) set_property(TARGET xxhash PROPERTY IMPORTED_LOCATION ${XXHASH_LIBRARIES})
include_directories(${XXHASH_INCLUDE_DIR}) include_directories(${XXHASH_INCLUDE_DIR})
add_dependencies(xxhash extern_xxhash) add_dependencies(xxhash extern_xxhash)
LIST(APPEND external_project_dependencies xxhash)
IF(WITH_C_API)
INSTALL(DIRECTORY ${XXHASH_INCLUDE_DIR} DESTINATION third_party/xxhash)
IF(ANDROID)
INSTALL(FILES ${XXHASH_LIBRARIES} DESTINATION third_party/xxhash/lib/${ANDROID_ABI})
ELSE()
INSTALL(FILES ${XXHASH_LIBRARIES} DESTINATION third_party/xxhash/lib)
ENDIF()
ENDIF()
cmake/external/zlib.cmake
浏览文件 @ a9ea99d7
...@@ -57,14 +57,3 @@ ENDIF(WIN32) ...@@ -57,14 +57,3 @@ ENDIF(WIN32)
ADD_LIBRARY(zlib STATIC IMPORTED GLOBAL) ADD_LIBRARY(zlib STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET zlib PROPERTY IMPORTED_LOCATION ${ZLIB_LIBRARIES}) SET_PROPERTY(TARGET zlib PROPERTY IMPORTED_LOCATION ${ZLIB_LIBRARIES})
ADD_DEPENDENCIES(zlib extern_zlib) ADD_DEPENDENCIES(zlib extern_zlib)
LIST(APPEND external_project_dependencies zlib)
IF(WITH_C_API)
INSTALL(DIRECTORY ${ZLIB_INCLUDE_DIR} DESTINATION third_party/zlib)
IF(ANDROID)
INSTALL(FILES ${ZLIB_LIBRARIES} DESTINATION third_party/zlib/lib/${ANDROID_ABI})
ELSE()
INSTALL(FILES ${ZLIB_LIBRARIES} DESTINATION third_party/zlib/lib)
ENDIF()
ENDIF()
cmake/flags.cmake
浏览文件 @ a9ea99d7
...@@ -21,7 +21,7 @@ function(CheckCompilerCXX11Flag) ...@@ -21,7 +21,7 @@ function(CheckCompilerCXX11Flag)
if (${CMAKE_CXX_COMPILER_VERSION} VERSION_LESS 3.3) if (${CMAKE_CXX_COMPILER_VERSION} VERSION_LESS 3.3)
message(FATAL_ERROR "Unsupported Clang version. Clang >= 3.3 required.") message(FATAL_ERROR "Unsupported Clang version. Clang >= 3.3 required.")
endif() endif()
endif() endif()
endif() endif()
endfunction() endfunction()
...@@ -147,19 +147,12 @@ set(GPU_COMMON_FLAGS ...@@ -147,19 +147,12 @@ set(GPU_COMMON_FLAGS
-Wno-error=unused-function # Warnings in Numpy Header. -Wno-error=unused-function # Warnings in Numpy Header.
-Wno-error=array-bounds # Warnings in Eigen::array -Wno-error=array-bounds # Warnings in Eigen::array
) )
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -m64")
else(NOT WIN32)
set(COMMON_FLAGS
"/w") #disable all warnings.
set(GPU_COMMON_FLAGS
"/w") #disable all warnings
endif(NOT WIN32) endif(NOT WIN32)
if (APPLE) if (APPLE)
if(NOT CMAKE_CROSSCOMPILING) # On Mac OS X build fat binaries with x86_64 architectures by default.
# On Mac OS X build fat binaries with x86_64 architectures by default. set (CMAKE_OSX_ARCHITECTURES "x86_64" CACHE STRING "Build architectures for OSX" FORCE)
set (CMAKE_OSX_ARCHITECTURES "x86_64" CACHE STRING "Build architectures for OSX" FORCE)
endif()
# On Mac OS X register class specifier is deprecated and will cause warning error on latest clang 10.0 # On Mac OS X register class specifier is deprecated and will cause warning error on latest clang 10.0
set (COMMON_FLAGS -Wno-deprecated-register) set (COMMON_FLAGS -Wno-deprecated-register)
endif(APPLE) endif(APPLE)
...@@ -195,8 +188,7 @@ safe_set_static_flag() ...@@ -195,8 +188,7 @@ safe_set_static_flag()
CMAKE_CXX_FLAGS_MINSIZEREL CMAKE_CXX_FLAGS_RELWITHDEBINFO CMAKE_CXX_FLAGS_MINSIZEREL CMAKE_CXX_FLAGS_RELWITHDEBINFO
CMAKE_C_FLAGS CMAKE_C_FLAGS_DEBUG CMAKE_C_FLAGS_RELEASE CMAKE_C_FLAGS CMAKE_C_FLAGS_DEBUG CMAKE_C_FLAGS_RELEASE
CMAKE_C_FLAGS_MINSIZEREL CMAKE_C_FLAGS_RELWITHDEBINFO) CMAKE_C_FLAGS_MINSIZEREL CMAKE_C_FLAGS_RELWITHDEBINFO)
if(${flag_var} MATCHES "/W3") string(REGEX REPLACE "(^| )/W[0-9]( |$)" " " ${flag_var} "${${flag_var}}")
string(REGEX REPLACE "/W3" "/w" ${flag_var} "${${flag_var}}") set(flag_var "${flag_var} /w")
endif(${flag_var} MATCHES "/W3")
endforeach(flag_var) endforeach(flag_var)
endif(WIN32) endif(WIN32)
cmake/generic.cmake
浏览文件 @ a9ea99d7
...@@ -90,11 +90,11 @@ ...@@ -90,11 +90,11 @@
# including binary directory for generated headers. # including binary directory for generated headers.
include_directories(${CMAKE_CURRENT_BINARY_DIR}) include_directories(${CMAKE_CURRENT_BINARY_DIR})
if(NOT APPLE AND NOT ANDROID) if(NOT APPLE)
find_package(Threads REQUIRED) find_package(Threads REQUIRED)
link_libraries(${CMAKE_THREAD_LIBS_INIT}) link_libraries(${CMAKE_THREAD_LIBS_INIT})
set(CMAKE_CXX_LINK_EXECUTABLE "${CMAKE_CXX_LINK_EXECUTABLE} -pthread -ldl -lrt") set(CMAKE_CXX_LINK_EXECUTABLE "${CMAKE_CXX_LINK_EXECUTABLE} -pthread -ldl -lrt")
endif(NOT APPLE AND NOT ANDROID) endif(NOT APPLE)
set_property(GLOBAL PROPERTY FLUID_MODULES "") set_property(GLOBAL PROPERTY FLUID_MODULES "")
# find all fluid modules is used for paddle fluid static library # find all fluid modules is used for paddle fluid static library
...@@ -304,7 +304,7 @@ function(cc_library TARGET_NAME) ...@@ -304,7 +304,7 @@ function(cc_library TARGET_NAME)
if(cc_library_DEPS) if(cc_library_DEPS)
merge_static_libs(${TARGET_NAME} ${cc_library_DEPS}) merge_static_libs(${TARGET_NAME} ${cc_library_DEPS})
else() else()
message(FATAL "Please specify source file or library in cc_library.") message(FATAL_ERROR "Please specify source files or libraries in cc_library(${TARGET_NAME} ...).")
endif() endif()
endif(cc_library_SRCS) endif(cc_library_SRCS)
endfunction(cc_library) endfunction(cc_library)
...@@ -388,6 +388,7 @@ function(cc_test TARGET_NAME) ...@@ -388,6 +388,7 @@ function(cc_test TARGET_NAME)
endif() endif()
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cpu_deterministic=true) set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cpu_deterministic=true)
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_init_allocated_mem=true) set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_init_allocated_mem=true)
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_limit_of_tmp_allocation=4294967296) # 4G
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cudnn_deterministic=true) set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cudnn_deterministic=true)
# No unit test should exceed 10 minutes. # No unit test should exceed 10 minutes.
set_tests_properties(${TARGET_NAME} PROPERTIES TIMEOUT 600) set_tests_properties(${TARGET_NAME} PROPERTIES TIMEOUT 600)
...@@ -460,6 +461,7 @@ function(nv_test TARGET_NAME) ...@@ -460,6 +461,7 @@ function(nv_test TARGET_NAME)
endif() endif()
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cpu_deterministic=true) set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cpu_deterministic=true)
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_init_allocated_mem=true) set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_init_allocated_mem=true)
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_limit_of_tmp_allocation=4294967296) # 4G
set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cudnn_deterministic=true) set_property(TEST ${TARGET_NAME} PROPERTY ENVIRONMENT FLAGS_cudnn_deterministic=true)
endif() endif()
endfunction(nv_test) endfunction(nv_test)
...@@ -655,12 +657,6 @@ function(paddle_protobuf_generate_cpp SRCS HDRS) ...@@ -655,12 +657,6 @@ function(paddle_protobuf_generate_cpp SRCS HDRS)
set(${SRCS}) set(${SRCS})
set(${HDRS}) set(${HDRS})
if (MOBILE_INFERENCE)
set(EXTRA_FLAG "lite:")
else()
set(EXTRA_FLAG "")
endif()
foreach(FIL ${ARGN}) foreach(FIL ${ARGN})
get_filename_component(ABS_FIL ${FIL} ABSOLUTE) get_filename_component(ABS_FIL ${FIL} ABSOLUTE)
get_filename_component(FIL_WE ${FIL} NAME_WE) get_filename_component(FIL_WE ${FIL} NAME_WE)
...@@ -677,7 +673,7 @@ function(paddle_protobuf_generate_cpp SRCS HDRS) ...@@ -677,7 +673,7 @@ function(paddle_protobuf_generate_cpp SRCS HDRS)
COMMAND ${CMAKE_COMMAND} -E make_directory "${CMAKE_CURRENT_BINARY_DIR}" COMMAND ${CMAKE_COMMAND} -E make_directory "${CMAKE_CURRENT_BINARY_DIR}"
COMMAND ${PROTOBUF_PROTOC_EXECUTABLE} COMMAND ${PROTOBUF_PROTOC_EXECUTABLE}
-I${CMAKE_CURRENT_SOURCE_DIR} -I${CMAKE_CURRENT_SOURCE_DIR}
--cpp_out "${EXTRA_FLAG}${CMAKE_CURRENT_BINARY_DIR}" ${ABS_FIL} --cpp_out "${CMAKE_CURRENT_BINARY_DIR}" ${ABS_FIL}
DEPENDS ${ABS_FIL} protoc DEPENDS ${ABS_FIL} protoc
COMMENT "Running C++ protocol buffer compiler on ${FIL}" COMMENT "Running C++ protocol buffer compiler on ${FIL}"
VERBATIM ) VERBATIM )
...@@ -714,9 +710,10 @@ function(py_test TARGET_NAME) ...@@ -714,9 +710,10 @@ function(py_test TARGET_NAME)
set(oneValueArgs "") set(oneValueArgs "")
set(multiValueArgs SRCS DEPS ARGS ENVS) set(multiValueArgs SRCS DEPS ARGS ENVS)
cmake_parse_arguments(py_test "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN}) cmake_parse_arguments(py_test "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
add_test(NAME ${TARGET_NAME} add_test(NAME ${TARGET_NAME}
COMMAND ${CMAKE_COMMAND} -E env FLAGS_init_allocated_mem=true FLAGS_cudnn_deterministic=true COMMAND ${CMAKE_COMMAND} -E env FLAGS_init_allocated_mem=true FLAGS_cudnn_deterministic=true
FLAGS_cpu_deterministic=true FLAGS_cpu_deterministic=true FLAGS_limit_of_tmp_allocation=4294967296 # 4G
PYTHONPATH=${PADDLE_BINARY_DIR}/python ${py_test_ENVS} PYTHONPATH=${PADDLE_BINARY_DIR}/python ${py_test_ENVS}
${PYTHON_EXECUTABLE} -u ${py_test_SRCS} ${py_test_ARGS} ${PYTHON_EXECUTABLE} -u ${py_test_SRCS} ${py_test_ARGS}
WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}) WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR})
...@@ -748,7 +745,7 @@ function(grpc_library TARGET_NAME) ...@@ -748,7 +745,7 @@ function(grpc_library TARGET_NAME)
#FIXME(putcn): the follwoing line is supposed to generate *.pb.h and cc, but #FIXME(putcn): the follwoing line is supposed to generate *.pb.h and cc, but
# somehow it didn't. line 602 to 604 is to patching this. Leaving this here # somehow it didn't. line 602 to 604 is to patching this. Leaving this here
# for now to enable dist CI. # for now to enable dist CI.
protobuf_generate_cpp(grpc_proto_srcs grpc_proto_hdrs "${ABS_PROTO}") paddle_protobuf_generate_cpp(grpc_proto_srcs grpc_proto_hdrs "${ABS_PROTO}")
set(grpc_grpc_srcs "${CMAKE_CURRENT_BINARY_DIR}/${PROTO_WE}.grpc.pb.cc") set(grpc_grpc_srcs "${CMAKE_CURRENT_BINARY_DIR}/${PROTO_WE}.grpc.pb.cc")
set(grpc_grpc_hdrs "${CMAKE_CURRENT_BINARY_DIR}/${PROTO_WE}.grpc.pb.h") set(grpc_grpc_hdrs "${CMAKE_CURRENT_BINARY_DIR}/${PROTO_WE}.grpc.pb.h")
cc_library("${TARGET_NAME}_proto" SRCS "${grpc_proto_srcs}") cc_library("${TARGET_NAME}_proto" SRCS "${grpc_proto_srcs}")
...@@ -791,7 +788,7 @@ function(brpc_library TARGET_NAME) ...@@ -791,7 +788,7 @@ function(brpc_library TARGET_NAME)
get_filename_component(PROTO_WE ${brpc_library_PROTO} NAME_WE) get_filename_component(PROTO_WE ${brpc_library_PROTO} NAME_WE)
get_filename_component(PROTO_PATH ${ABS_PROTO} PATH) get_filename_component(PROTO_PATH ${ABS_PROTO} PATH)
protobuf_generate_cpp(brpc_proto_srcs brpc_proto_hdrs "${ABS_PROTO}") paddle_protobuf_generate_cpp(brpc_proto_srcs brpc_proto_hdrs "${ABS_PROTO}")
cc_library("${TARGET_NAME}_proto" SRCS "${brpc_proto_srcs}") cc_library("${TARGET_NAME}_proto" SRCS "${brpc_proto_srcs}")
cc_library("${TARGET_NAME}" SRCS "${brpc_library_SRCS}" DEPS "${TARGET_NAME}_proto" "${brpc_library_DEPS}") cc_library("${TARGET_NAME}" SRCS "${brpc_library_SRCS}" DEPS "${TARGET_NAME}_proto" "${brpc_library_DEPS}")
endfunction() endfunction()
cmake/hip.cmake
浏览文件 @ a9ea99d7
...@@ -11,8 +11,6 @@ include_directories("/opt/rocm/rocrand/include") ...@@ -11,8 +11,6 @@ include_directories("/opt/rocm/rocrand/include")
include_directories("/opt/rocm/rccl/include") include_directories("/opt/rocm/rccl/include")
include_directories("/opt/rocm/thrust") include_directories("/opt/rocm/thrust")
list(APPEND EXTERNAL_LIBS "-L/opt/rocm/lib/ -lhip_hcc")
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -fPIC -DPADDLE_WITH_HIP -std=c++11" ) set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -fPIC -DPADDLE_WITH_HIP -std=c++11" )
if(WITH_DSO) if(WITH_DSO)
...@@ -31,22 +29,12 @@ if(WITH_GRPC) ...@@ -31,22 +29,12 @@ if(WITH_GRPC)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_GRPC") set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_GRPC")
endif(WITH_GRPC) endif(WITH_GRPC)
if(NOT WITH_GOLANG)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITHOUT_GOLANG")
endif(NOT WITH_GOLANG)
if(WITH_MKLDNN) if(WITH_MKLDNN)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_MKLDNN") set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_MKLDNN")
endif(WITH_MKLDNN) endif(WITH_MKLDNN)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DANY_IMPL_ANY_CAST_MOVEABLE") set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DANY_IMPL_ANY_CAST_MOVEABLE")
if(NOT WITH_RDMA)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_DISABLE_RDMA")
endif(NOT WITH_RDMA)
if(CMAKE_BUILD_TYPE STREQUAL "Debug") if(CMAKE_BUILD_TYPE STREQUAL "Debug")
list(APPEND HIP_HCC_FLAGS ${CMAKE_CXX_FLAGS_DEBUG}) list(APPEND HIP_HCC_FLAGS ${CMAKE_CXX_FLAGS_DEBUG})
elseif(CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo") elseif(CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo")
......
cmake/inference_lib.cmake
浏览文件 @ a9ea99d7
...@@ -149,25 +149,23 @@ if (WITH_NGRAPH) ...@@ -149,25 +149,23 @@ if (WITH_NGRAPH)
) )
endif () endif ()
if (NOT MOBILE_INFERENCE AND NOT RPI) set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappy")
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappy") copy(snappy_lib
copy(snappy_lib SRCS ${SNAPPY_INCLUDE_DIR} ${SNAPPY_LIBRARIES}
SRCS ${SNAPPY_INCLUDE_DIR} ${SNAPPY_LIBRARIES} DSTS ${dst_dir} ${dst_dir}/lib
DSTS ${dst_dir} ${dst_dir}/lib DEPS snappy)
DEPS snappy)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappystream") set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappystream")
copy(snappystream_lib copy(snappystream_lib
SRCS ${SNAPPYSTREAM_INCLUDE_DIR} ${SNAPPYSTREAM_LIBRARIES} SRCS ${SNAPPYSTREAM_INCLUDE_DIR} ${SNAPPYSTREAM_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib DSTS ${dst_dir} ${dst_dir}/lib
DEPS snappystream) DEPS snappystream)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/zlib") set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/zlib")
copy(zlib_lib copy(zlib_lib
SRCS ${ZLIB_INCLUDE_DIR} ${ZLIB_LIBRARIES} SRCS ${ZLIB_INCLUDE_DIR} ${ZLIB_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib DSTS ${dst_dir} ${dst_dir}/lib
DEPS zlib) DEPS zlib)
endif ()
# paddle fluid module # paddle fluid module
set(src_dir "${PADDLE_SOURCE_DIR}/paddle/fluid") set(src_dir "${PADDLE_SOURCE_DIR}/paddle/fluid")
......
cmake/operators.cmake
浏览文件 @ a9ea99d7
...@@ -52,8 +52,8 @@ function(op_library TARGET) ...@@ -52,8 +52,8 @@ function(op_library TARGET)
endif() endif()
if(WITH_MKLDNN) if(WITH_MKLDNN)
string(REPLACE "_op" "_mkldnn_op" MKLDNN_FILE "${TARGET}") string(REPLACE "_op" "_mkldnn_op" MKLDNN_FILE "${TARGET}")
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${MKLDNN_FILE}.cc) if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/mkldnn/${MKLDNN_FILE}.cc)
list(APPEND mkldnn_cc_srcs ${MKLDNN_FILE}.cc) list(APPEND mkldnn_cc_srcs mkldnn/${MKLDNN_FILE}.cc)
endif() endif()
endif() endif()
else() else()
...@@ -153,7 +153,11 @@ function(op_library TARGET) ...@@ -153,7 +153,11 @@ function(op_library TARGET)
# pybind USE_OP_DEVICE_KERNEL for CUDNN # pybind USE_OP_DEVICE_KERNEL for CUDNN
list(LENGTH cudnn_cu_cc_srcs cudnn_cu_cc_srcs_len) list(LENGTH cudnn_cu_cc_srcs cudnn_cu_cc_srcs_len)
if (WITH_GPU AND ${cudnn_cu_cc_srcs_len} GREATER 0) if (WITH_GPU AND ${cudnn_cu_cc_srcs_len} GREATER 0)
if(${TARGET} STREQUAL "activation")
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(relu, CUDNN);\n")
else()
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(${TARGET}, CUDNN);\n") file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(${TARGET}, CUDNN);\n")
endif()
endif() endif()
# pybind USE_OP_DEVICE_KERNEL for MIOPEN # pybind USE_OP_DEVICE_KERNEL for MIOPEN
...@@ -168,6 +172,9 @@ function(op_library TARGET) ...@@ -168,6 +172,9 @@ function(op_library TARGET)
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(relu, MKLDNN);\n") file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(relu, MKLDNN);\n")
elseif(${MKLDNN_FILE} STREQUAL "conv_mkldnn_op") elseif(${MKLDNN_FILE} STREQUAL "conv_mkldnn_op")
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL_WITH_CUSTOM_TYPE(conv2d, MKLDNN, FP32);\n") file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL_WITH_CUSTOM_TYPE(conv2d, MKLDNN, FP32);\n")
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL_WITH_CUSTOM_TYPE(conv2d, MKLDNN, S8);\n")
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL_WITH_CUSTOM_TYPE(conv2d, MKLDNN, U8);\n")
else() else()
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(${TARGET}, MKLDNN);\n") file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(${TARGET}, MKLDNN);\n")
endif() endif()
......
cmake/rdma.cmake 已删除 100644 → 0
浏览文件 @ eac5a0aa
# user should download rdma first from subversion repository
# execute following instruction to download svn mannally
# svn co https://svn.baidu.com/sys/ip/trunk/rdma/sockrdmav1 rdma/
# svn co https://svn.baidu.com/sys/ip/trunk/rdma/thirdparty rdma/
# we use static output in svn repositories to avoid implict bugs from not standard runtime env.
if(WITH_RDMA)
set(RDMA_ROOT $ENV{RDMA_ROOT} CACHE PATH "Folder contains RDMA sock library and thirdparty library")
function(generate_rdma_links)
#redirect to current DIR to isolate the pollution from system runtime environment
#it can benifits unified control for different gcc environment.
#e.g, by default gcc48 did not refer /usr/lib64 which could contain low version
#runtime libraries that will crash process while loading it. That redirect trick
#can fix it.
execute_process(
COMMAND mkdir -p librdma
COMMAND ln -s -f /usr/lib64/libibverbs.so.1.0.0 librdma/libibverbs.so.1
COMMAND ln -s -f /usr/lib64/libibverbs.so.1.0.0 librdma/libibverbs.so
COMMAND ln -s -f /usr/lib64/librdmacm.so.1.0.0 librdma/librdmacm.so.1
COMMAND ln -s -f /usr/lib64/librdmacm.so.1.0.0 librdma/librdmacm.so
COMMAND ln -s -f /lib64/libnl.so.1.1.4 librdma/libnl.so.1
COMMAND ln -s -f /lib64/libnl.so.1.1.4 librdma/libnl.so
WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
)
endfunction(generate_rdma_links)
#check and set headers
find_path(RDMA_INC_SXISOCK sxi_sock.h PATHS ${RDMA_ROOT}/sockrdmav1/output/include)
find_path(RDMA_INC_XIO libxio.h PATHS ${RDMA_ROOT}/thirdparty/output/accelio)
find_path(RDMA_INC_EVENT event2 PATHS ${RDMA_ROOT}/thirdparty/output/libevent)
find_path(RDMA_INC_NUMA numa.h PATHS ${RDMA_ROOT}/thirdparty/output/libnuma)
#check and set libs
find_library(RDMA_LIB_SXISOCK NAMES sxisock PATHS ${RDMA_ROOT}/sockrdmav1/output)
find_library(RDMA_LIB_XIO NAMES xio PATHS ${RDMA_ROOT}/thirdparty/output/accelio)
find_library(RDMA_LIB_EVENT NAMES event PATHS ${RDMA_ROOT}/thirdparty/output/libevent)
find_library(RDMA_LIB_EVENT_CORE NAMES event_core PATHS ${RDMA_ROOT}/thirdparty/output/libevent)
find_library(RDMA_LIB_EVENT_EXTRA NAMES event_extra PATHS ${RDMA_ROOT}/thirdparty/output/libevent)
find_library(RDMA_LIB_EVENT_PTHREADS NAMES event_pthreads PATHS ${RDMA_ROOT}/thirdparty/output/libevent)
find_library(RDMA_LIB_NUMA NAMES numa PATHS ${RDMA_ROOT}/thirdparty/output/libnuma)
if(
RDMA_INC_SXISOCK AND
RDMA_INC_XIO AND
RDMA_INC_EVENT AND
RDMA_INC_NUMA AND
RDMA_LIB_SXISOCK AND
RDMA_LIB_XIO AND
RDMA_LIB_EVENT AND
RDMA_LIB_EVENT_CORE AND
RDMA_LIB_EVENT_EXTRA AND
RDMA_LIB_EVENT_PTHREADS AND
RDMA_LIB_NUMA
)
set(RDMA_INC_DIR
${RDMA_INC_SXISOCK}
${RDMA_INC_XIO}
${RDMA_INC_EVENT}
${RDMA_INC_NUMA})
set(RDMA_LIBS
${RDMA_LIB_SXISOCK}
${RDMA_LIB_XIO}
${RDMA_LIB_EVENT}
${RDMA_LIB_EVENT_CORE}
${RDMA_LIB_EVENT_EXTRA}
${RDMA_LIB_EVENT_PTHREADS}
${RDMA_LIB_NUMA}
)
set(RDMA_LD_FLAGS "-L./librdma -libverbs -lrdmacm -Xlinker -rpath ./librdma")
include_directories("${RDMA_INC_DIR}")
else()
#if this module is not called, RDMA_INC_DIR RDMA_LIBS will be null, so top module always refer this variable
message(FATAL_ERROR, "RDMA libraries are not found, try to set RDMA_ROOT or check all related libraries.")
endif()
else(WITH_RDMA)
set(RDMA_LIBS "")
set(RDMA_LD_FLAGS "")
add_definitions(-DPADDLE_DISABLE_RDMA)
endif(WITH_RDMA)
cmake/system.cmake
浏览文件 @ a9ea99d7
...@@ -74,21 +74,6 @@ MARK_AS_ADVANCED(HOST_SYSTEM CPU_CORES) ...@@ -74,21 +74,6 @@ MARK_AS_ADVANCED(HOST_SYSTEM CPU_CORES)
MESSAGE(STATUS "Found Paddle host system: ${HOST_SYSTEM}, version: ${HOST_SYSTEM_VERSION}") MESSAGE(STATUS "Found Paddle host system: ${HOST_SYSTEM}, version: ${HOST_SYSTEM_VERSION}")
MESSAGE(STATUS "Found Paddle host system's CPU: ${CPU_CORES} cores") MESSAGE(STATUS "Found Paddle host system's CPU: ${CPU_CORES} cores")
# configuration for cross-compiling
IF(DEFINED CMAKE_SYSTEM_NAME)
INCLUDE(cross_compiling/host)
IF(${CMAKE_SYSTEM_NAME} STREQUAL "Android")
SET(ANDROID TRUE)
INCLUDE(cross_compiling/android)
ELSEIF(${CMAKE_SYSTEM_NAME} STREQUAL "RPi")
SET(RPI TRUE)
INCLUDE(cross_compiling/raspberry_pi)
ELSEIF(${CMAKE_SYSTEM_NAME} STREQUAL "iOS")
SET(IOS TRUE)
INCLUDE(cross_compiling/ios)
ENDIF()
ENDIF()
# external dependencies log output # external dependencies log output
SET(EXTERNAL_PROJECT_LOG_ARGS SET(EXTERNAL_PROJECT_LOG_ARGS
LOG_DOWNLOAD 0 # Wrap download in script to log output LOG_DOWNLOAD 0 # Wrap download in script to log output
......
cmake/tensorrt.cmake
浏览文件 @ a9ea99d7
...@@ -33,6 +33,5 @@ if(TENSORRT_FOUND) ...@@ -33,6 +33,5 @@ if(TENSORRT_FOUND)
message(STATUS "Current TensorRT header is ${TENSORRT_INCLUDE_DIR}/NvInfer.h. " message(STATUS "Current TensorRT header is ${TENSORRT_INCLUDE_DIR}/NvInfer.h. "
"Current TensorRT version is v${TENSORRT_MAJOR_VERSION}. ") "Current TensorRT version is v${TENSORRT_MAJOR_VERSION}. ")
include_directories(${TENSORRT_INCLUDE_DIR}) include_directories(${TENSORRT_INCLUDE_DIR})
list(APPEND EXTERNAL_LIBS ${TENSORRT_LIBRARY})
add_definitions(-DPADDLE_WITH_TENSORRT) add_definitions(-DPADDLE_WITH_TENSORRT)
endif() endif()
cmake/util.cmake
浏览文件 @ a9ea99d7
...@@ -53,118 +53,3 @@ function(target_circle_link_libraries TARGET_NAME) ...@@ -53,118 +53,3 @@ function(target_circle_link_libraries TARGET_NAME)
"-Wl,--end-group") "-Wl,--end-group")
endif() endif()
endfunction() endfunction()
# compile_cu_as_cpp
# Make a cu file compiled as C++
# Arguments: Source files
macro(compile_cu_as_cpp)
foreach(s ${ARGN})
set_source_files_properties(${s} PROPERTIES LANGUAGE CXX)
set_source_files_properties(${s} PROPERTIES COMPILE_FLAGS "-x c++")
endforeach()
endmacro()
# link_paddle_exe
# add paddle library for a paddle executable, such as trainer, pserver.
#
# It will handle WITH_PYTHON etc.
function(link_paddle_exe TARGET_NAME)
if(WITH_RDMA)
generate_rdma_links()
endif()
if(MOBILE_INFERENCE)
target_circle_link_libraries(${TARGET_NAME}
ARCHIVE_START
paddle_gserver
paddle_function
ARCHIVE_END
paddle_math
paddle_utils
paddle_parameter
paddle_proto
paddle_cuda
${EXTERNAL_LIBS}
${CMAKE_THREAD_LIBS_INIT}
${CMAKE_DL_LIBS}
${RDMA_LD_FLAGS}
${RDMA_LIBS})
else()
target_circle_link_libraries(${TARGET_NAME}
ARCHIVE_START
paddle_gserver
paddle_function
ARCHIVE_END
paddle_pserver
paddle_trainer_lib
paddle_network
paddle_math
paddle_utils
paddle_parameter
paddle_proto
paddle_cuda
paddle_optimizer
${EXTERNAL_LIBS}
${CMAKE_THREAD_LIBS_INIT}
${CMAKE_DL_LIBS}
${RDMA_LD_FLAGS}
${RDMA_LIBS})
endif()
if(ANDROID)
target_link_libraries(${TARGET_NAME} log)
endif(ANDROID)
if(WITH_MKLML AND MKLML_LIB_DIR AND MKLML_IOMP_LIB)
target_link_libraries(${TARGET_NAME} "-L${MKLML_LIB_DIR} -liomp5 -Wl,--as-needed")
endif()
add_dependencies(${TARGET_NAME} ${external_project_dependencies})
endfunction()
# link_paddle_test
# Link a paddle unittest for target
# TARGET_NAME: the unittest target name
# Rest Arguemnts: not used.
function(link_paddle_test TARGET_NAME)
link_paddle_exe(${TARGET_NAME})
target_link_libraries(${TARGET_NAME}
paddle_test_main
paddle_test_util
${GTEST_LIBRARIES})
endfunction()
# add_unittest_without_exec
#
# create a paddle unittest. not specifically define how to run this unittest.
# TARGET_NAME: the unittest target name, same as executable file name
# Rest Arguments: the source files to compile this unittest.
macro(add_unittest_without_exec TARGET_NAME)
add_executable(${TARGET_NAME} ${ARGN})
link_paddle_test(${TARGET_NAME})
endmacro()
# add_unittest
# create a paddle unittest and just to execute this binary to make unittest.
#
# TARGET_NAME: the unittest target name, same as executable file name
# Rest Arguments: the source files to compile this unittest.
macro(add_unittest TARGET_NAME)
add_unittest_without_exec(${TARGET_NAME} ${ARGN})
add_test(${TARGET_NAME} ${TARGET_NAME})
endmacro()
# add_simple_unittest
# create a paddle unittest with file name. It just compile ${TARGET_NAME}.cpp to
# ${TARGET_NAME} and then execute it.
macro(add_simple_unittest TARGET_NAME)
add_unittest(${TARGET_NAME} ${TARGET_NAME}.cpp)
endmacro()
# Creates C resources file from files in given resource file
function(create_resources res_file output_file)
add_custom_command(
OUTPUT ${output_file}
COMMAND python ARGS ${PADDLE_SOURCE_DIR}/cmake/make_resource.py ${res_file} ${output_file}
DEPENDS ${res_file} ${PADDLE_SOURCE_DIR}/cmake/make_resource.py)
endfunction()
cmake/version.cmake
浏览文件 @ a9ea99d7
...@@ -31,8 +31,23 @@ while ("${PADDLE_VERSION}" STREQUAL "") ...@@ -31,8 +31,23 @@ while ("${PADDLE_VERSION}" STREQUAL "")
set(tmp_version "${GIT_TAG_NAME}~1") set(tmp_version "${GIT_TAG_NAME}~1")
endif() endif()
else() else()
# otherwise, we always set PADDLE_VERSION to 0.0.0 to represent latest execute_process(
set(PADDLE_VERSION "0.0.0") COMMAND ${GIT_EXECUTABLE} describe --exact-match --tags ${tmp_version}
WORKING_DIRECTORY ${PADDLE_SOURCE_DIR}
OUTPUT_VARIABLE GIT_EXACT_TAG_NAME
RESULT_VARIABLE GIT_EXACT_TAG_RESULT
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
if (NOT ${GIT_EXACT_TAG_NAME})
# Check if current branch is tag branch
if (${GIT_EXACT_TAG_NAME} MATCHES "v${TAG_VERSION_REGEX}")
string(REPLACE "v" "" PADDLE_VERSION ${GIT_EXACT_TAG_NAME})
else()
set(PADDLE_VERSION "0.0.0")
endif()
else()
# otherwise, we always set PADDLE_VERSION to 0.0.0 to represent latest
set(PADDLE_VERSION "0.0.0")
endif()
endif() endif()
else() else()
set(PADDLE_VERSION "0.0.0") set(PADDLE_VERSION "0.0.0")
......
doc/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
add_custom_target(paddle_apis ALL
DEPENDS paddle_v2_apis)
add_custom_target(paddle_docs ALL
DEPENDS paddle_v2_docs paddle_v2_docs_cn
paddle_mobile_docs paddle_mobile_docs_cn)
add_subdirectory(v2)
add_subdirectory(mobile)
doc/about/about_us.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
=========
关于我们
=========
什么是PaddlePaddle
--------------------
- PaddlePaddle是百度自主研发并开源的深度学习框架，它能够让开发者和企业安全、快速地实现自己的AI想法
- 项目团队汇聚了全球顶级的深度学习科学家，致力于为开发者和企业提供最好的深度学习研发体验
- 框架具有易学、易用、安全、高效四大特性，是最适合中国开发者和企业的深度学习工具
PaddlePaddle的技术特色
-------------------------
- 新一代深度学习框架： PaddlePaddle是基于“深度学习编程语言”的新一代深度学习框架，在保证性能的同时，极大的提升了框架对模型的表达能力，能够描述任意潜在可能出现的模型
- 对大规模计算更加友好：经过百度内多种大规模计算业务的打磨，PaddlePaddle在分布式计算上表现优异，基于EDL技术能够节约大量计算资源，同时也能支持大规模稀疏模型的训练
- 提供可视化的深度学习：通过Visual DL可以帮助开发者方便的观测训练整体趋势、数据样本质量和中间结果、参数分布和变化趋势、以及模型的结构，帮助开发者更便捷的完成编程过程
提供基于PaddlePaddle的教育体系
--------------------------------
- 深度学习课程：百度与中国市场顶级的教育、培训机构共同开发了深度学习精品课程以及学习教材，帮助开发者从零掌握深度学习
- 深度学习实训：对于目的是科研和学习的用户，PaddlePaddle提供了无需安装、线上运行的开发环境，并提供算法、算力、数据支持
- 线下培训：提供丰富、高质量的线下教育活动，如青年教师培训、线下实战营、沙龙等多种形式的培训和交流
提供基于PaddlePaddle的AI服务
------------------------------
- EadyDL：可以帮助零算法基础的企业快速完成一个深度学习任务，只需少量的数据即可得到优质的模型
- AI市场：提供标准化的AI 能力、产品的交易机制，帮助企业快速找到所需，有效开展AI业务
- 深度学习竞赛： PaddlePaddle汇聚顶尖深度学习开发者，企业可以发布自己的商业问题，通过竞赛方式快速找到最优的解决方案
你对PaddlePaddle有任何的问题都可以通过以下方式联系到我们
-----------------------------------------------------------
- 学习/使用问题：可以在 `PaddlePaddle开源社区 <https://github.com/PaddlePaddle/Paddle/issues>`_，以及 `PaddlePaddle中文社区 <http://ai.baidu.com/forum/topic/list/168>`_ 向我们反馈
- 对PaddlePaddle框架发展的建议：可发送邮件至Paddle-better@baidu.com
我们期待与你一起打造世界顶级深度学习框架，共同推动AI技术的进步
PaddlePaddle团队
doc/mobile/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
if(NOT DEFINED SPHINX_THEME)
set(SPHINX_THEME default)
endif()
if(NOT DEFINED SPHINX_THEME_DIR)
set(SPHINX_THEME_DIR)
endif()
# configured documentation tools and intermediate build results
set(BINARY_BUILD_DIR_EN "${CMAKE_CURRENT_BINARY_DIR}/en/_build")
# Sphinx cache with pickled ReST documents
set(SPHINX_CACHE_DIR_EN "${CMAKE_CURRENT_BINARY_DIR}/en/_doctrees")
# HTML output director
set(SPHINX_HTML_DIR_EN "${CMAKE_CURRENT_BINARY_DIR}/en/html")
set(IMPORT_PADDLE_STRING "")
set(IMPORT_PADDLEV2_STRING "")
configure_file(
"${CMAKE_CURRENT_SOURCE_DIR}/../templates/conf.py.en.in"
"${BINARY_BUILD_DIR_EN}/conf.py"
@ONLY)
sphinx_add_target(paddle_mobile_docs
html
${BINARY_BUILD_DIR_EN}
${SPHINX_CACHE_DIR_EN}
${CMAKE_CURRENT_SOURCE_DIR}
${SPHINX_HTML_DIR_EN})
# configured documentation tools and intermediate build results
set(BINARY_BUILD_DIR_CN "${CMAKE_CURRENT_BINARY_DIR}/cn/_build")
# Sphinx cache with pickled ReST documents
set(SPHINX_CACHE_DIR_CN "${CMAKE_CURRENT_BINARY_DIR}/cn/_doctrees")
# HTML output director
set(SPHINX_HTML_DIR_CN "${CMAKE_CURRENT_BINARY_DIR}/cn/html")
configure_file(
"${CMAKE_CURRENT_SOURCE_DIR}/../templates/conf.py.cn.in"
"${BINARY_BUILD_DIR_CN}/conf.py"
@ONLY)
sphinx_add_target(paddle_mobile_docs_cn
html
${BINARY_BUILD_DIR_CN}
${SPHINX_CACHE_DIR_CN}
${CMAKE_CURRENT_SOURCE_DIR}
${SPHINX_HTML_DIR_CN})
doc/mobile/cross_compiling_for_android_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Android平台编译指南
用户可通过如下两种方式，交叉编译Android平台上适用的PaddlePaddle库：
- [基于Docker容器的编译方式](#基于docker容器的编译方式)
- [基于Linux交叉编译环境的编译方式](#基于linux交叉编译环境的编译方式)
## 基于Docker容器的编译方式
Docker能在所有主要操作系统（包括Linux，Mac OS X和Windows）上运行，因此，使用基于Docker容器的编译方式，用户可在自己熟悉的开发平台上编译Android平台上适用的PaddlePaddle库。
### 构建PaddlePaddle的Android开发镜像
我们把PaddlePaddle的交叉编译环境打包成一个镜像，称为开发镜像，里面涵盖了交叉编译Android版PaddlePaddle库需要的所有编译工具。
```bash
$ git clone https://github.com/PaddlePaddle/Paddle.git
$ cd Paddle
$ docker build -t username/paddle-android:dev . -f Dockerfile.android
```
用户也可以使用PaddlePaddle提供的官方开发镜像：
```bash
$ docker pull paddlepaddle/paddle:latest-dev-android
```
对于国内用户，我们提供了加速访问的镜像源：
```bash
$ docker pull docker.paddlepaddlehub.com/paddle:latest-dev-android
```
### 编译PaddlePaddle C-API库
构建好开发镜像后，即可使用开发镜像来编译Android版PaddlePaddle C-API库。
Android的Docker开发镜像向用户提供两个可配置的参数：
<table class="docutils">
<colgroup>
<col width="25%" />
<col width="50%" />
<col width="25%" />
</colgroup>
<thead valign="bottom">
<tr class="row-odd">
<th class="head">Argument</th>
<th class="head">Optional Values</th>
<th class="head">Default</th>
</tr>
</thead>
<tbody valign="top">
<tr class="row-even">
<td>ANDROID_ABI</td>
<td>armeabi-v7a, arm64-v8a</td>
<td>armeabi-v7a</td>
</tr>
<tr class="row-odd">
<td>ANDROID_API</td>
<td>>= 16</td>
<td>21</td>
</tr>
</tbody>
</table>
- 编译`armeabi-v7a`，`Android API 21`的PaddlePaddle库
```bash
$ docker run -it --rm -v $PWD:/paddle -w /paddle -e "ANDROID_ABI=armeabi-v7a" -e "ANDROID_API=21" username/paddle-android:dev ./paddle/scripts/paddle_build.sh build_android
```
- 编译`arm64-v8a`，`Android API 21`的PaddlePaddle库
```bash
$ docker run -it --rm -v $PWD:/paddle -w /paddle -e "ANDROID_ABI=arm64-v8a" -e "ANDROID_API=21" username/paddle-android:dev ./paddle/scripts/paddle_build.sh build_android
```
执行上述`docker run`命令时，容器执行[paddle/scripts/paddle_build.sh build_android](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/scripts/paddle_build.sh)脚本。该脚本中记录了交叉编译Android版PaddlePaddle库常用的CMake配置，并且会根据`ANDROID_ABI`和`ANDROID_API`自动构建独立工具链、进行编译和安装。由于arm64架构要求Android API不小于21。因此当`ANDROID_ABI=arm64-v8a`，`ANDROID_API<21`时，Docker容器中将默认使用`Android API 21`的编译工具链。用户可以参考下文[配置交叉编译参数](#配置交叉编译参数)章节，根据个人的需求修改定制Docker容器所执行的脚本。编译安装结束之后，PaddlePaddle的C-API库将被安装到`$PWD/install_android`目录，所依赖的第三方库同时也被安装到`$PWD/install_android/third_party`目录。
## 基于Linux交叉编译环境的编译方式
本文档将以Linux x86-64平台为例，介绍交叉编译Android平台上适用的PaddlePaddle库的方法和步骤。
### 准备交叉编译环境
从源码交叉编译PaddlePaddle，用户需要提前准备好交叉编译环境。Android平台上使用的C/C++交叉编译工具链为[Android NDK](https://developer.android.com/ndk/downloads/index.html?hl=zh-cn)，用户可自行前往下载预编译好的版本，也可通过以下命令获取：
```bash
wget -q https://dl.google.com/android/repository/android-ndk-r14b-linux-x86_64.zip
unzip -q android-ndk-r14b-linux-x86_64.zip
```
Android NDK中包含了所有Android API级别、所有架构（arm/arm64/x86/mips）需要用到的编译工具和系统库。用户可根据自己的编译目标架构、所需支持的最低Android API级别，构建[独立工具链](https://developer.android.google.cn/ndk/guides/standalone_toolchain.html?hl=zh-cn)。
- 构建`armeabi-v7a`、 `Android API 21`的独立工具链：
```bash
your/path/to/android-ndk-r14b-linux-x86_64/build/tools/make-standalone-toolchain.sh \
--arch=arm --platform=android-21 --install-dir=your/path/to/arm_standalone_toolchain
```
此命令将在`your/path/to/arm_standalone_toolchain`目录生成一套独立编译工具链，面向架构为32位ARM架构，支持的最小的Android API级别为21，支持编译器`arm-linux-androideabi-gcc (GCC) 4.9`和`clang 3.8`。
- 构建`arm64-v8a`、 `Android API 21`的独立工具链：
```bash
your/path/to/android-ndk-r14b-linux-x86_64/build/tools/make-standalone-toolchain.sh \
--arch=arm64 --platform=android-21 --install-dir=your/path/to/arm64_standalone_toolchain
```
此命令将在`your/path/to/arm64_standalone_toolchain`目录生成一套独立编译工具链，面向架构为64位ARM64架构，支持的最小Android API级别为21，支持编译器`arm-linux-androideabi-gcc (GCC) 4.9`和`clang 3.8`。
### 配置交叉编译参数
CMake系统对交叉编译提供了支持[cmake-toolchains](https://cmake.org/cmake/help/v3.0/manual/cmake-toolchains.7.html#cross-compiling)。为了简化cmake配置，PaddlePaddle为交叉编译提供了工具链配置文档[cmake/cross_compiling/android.cmake](https://github.com/PaddlePaddle/Paddle/blob/develop/cmake/cross_compiling/android.cmake)，以提供一些默认的编译器和编译参数相关配置。注意，从CMake 3.7版本开始，CMake官方对Android平台的交叉编译提供了通用的支持。PaddlePaddle若检测到用户使用的CMake版本不低于3.7时，将会将用户传进来的配置参数传递CMake系统，交由CMake系统本身来处理。有关参数配置的详细说明见[cmake-toolchains](https://cmake.org/cmake/help/v3.7/manual/cmake-toolchains.7.html#cross-compiling)。
交叉编译Android版本的PaddlePaddle库时，有一些必须配置的参数：
- `CMAKE_SYSTEM_NAME`，CMake编译的目标平台，必须设置为`Android`。在设置`CMAKE_SYSTEM_NAME=Android`后，PaddlePaddle的CMake系统才认为是在交叉编译Android系统的版本，并自动编译PaddlePaddle所需的所有第三方库。此外，还会强制设置一些PaddlePaddle参数的值（`WITH_GPU=OFF`、`WITH_AVX=OFF`、`WITH_PYTHON=OFF`、`WITH_RDMA=OFF`、`WITH_MKL=OFF`、`WITH_GOLANG=OFF`）。
- `WITH_C_API`，必须设置为`ON`。在Android平台上只支持使用C-API来预测。
- `WITH_SWIG_PY`，必须设置为`OFF`。在Android平台上不支持通过swig调用来训练或者预测。
Android平台可选配置参数：
- `ANDROID_STANDALONE_TOOLCHAIN`，独立工具链所在的绝对路径，或者相对于构建目录的相对路径。PaddlePaddle的CMake系统将根据该值自动推导和设置需要使用的交叉编译器、sysroot、以及Android API级别；否则，用户需要在cmake时手动设置这些值。无默认值。
- `ANDROID_TOOLCHAIN`，目标工具链。可设置`gcc/clang`，默认值为`clang`。
- CMake 3.7以上，将会始终使用`clang`工具链；CMake 3.7以下，可设置`ANDROID_TOOLCHAIN=gcc`以使用`gcc`工具链。
- Android官方提供的`clang`编译器要求系统支持`GLIBC 2.15`以上。
- `ANDROID_ABI`，目标架构ABI。目前支持`armeabi-v7a`和`arm64-v8a`，默认值为`armeabi-v7a`。
- `ANDROID_NATIVE_API_LEVEL`，工具链的Android API级别。若没有显式设置，PaddlePaddle将根据`ANDROID_STANDALONE_TOOLCHAIN`的值自动推导得到。
- `ANROID_ARM_MODE`，是否使用ARM模式。
- `ANDROID_ABI=armeabi-v7a`时，可设置`ON/OFF`，默认值为`ON`；
- `ANDROID_ABI=arm64-v8a`时，不需要设置。
- `ANDROID_ARM_NEON`，是否使用NEON指令。
- `ANDROID_ABI=armeabi-v7a`时，可设置`ON/OFF`，默认值为`ON`；
- `ANDROID_ABI=arm64-v8a`时，不需要设置。
其他配置参数：
- `USE_EIGEN_FOR_BLAS`，是否使用Eigen库进行矩阵计算。可设置`ON/OFF`，默认值为`OFF`。
- `HOST_C/CXX_COMPILER`，宿主机的C/C++编译器。在编译宿主机版protoc可执行文件和目标机版OpenBLAS库时需要用到。默认设置成环境变量`CC/CXX`的值；若环境变量`CC/CXX`没有设置，则设置成`cc/c++`编译器。
常用的cmake配置如下：
```bash
cmake -DCMAKE_SYSTEM_NAME=Android \
-DANDROID_STANDALONE_TOOLCHAIN=your/path/to/arm_standalone_toolchain \
-DANDROID_ABI=armeabi-v7a \
-DANDROID_ARM_NEON=ON \
-DANDROID_ARM_MODE=ON \
-DUSE_EIGEN_FOR_BLAS=ON \
-DCMAKE_INSTALL_PREFIX=your/path/to/install \
-DWITH_C_API=ON \
-DWITH_SWIG_PY=OFF \
..
```
```
cmake -DCMAKE_SYSTEM_NAME=Android \
-DANDROID_STANDALONE_TOOLCHAIN=your/path/to/arm64_standalone_toolchain \
-DANDROID_ABI=arm64-v8a \
-DUSE_EIGEN_FOR_BLAS=OFF \
-DCMAKE_INSTALL_PREFIX=your/path/to/install \
-DWITH_C_API=ON \
-DWITH_SWIG_PY=OFF \
..
```
用户还可根据自己的需求设置其他编译参数。
- 设置`CMAKE_BUILD_TYPE`为`MinSizeRel`，最小化生成的库的大小。
- 设置`CMAKE_BUILD_TYPE`为`Release`，获得最快的执行速度，
- 用户亦可以通过手动设置`CMAKE_C/CXX_FLAGS`来影响PaddlePaddle的编译过程。
**性能TIPS**，为了达到最快的计算速度，在CMake参数配置上，有以下建议：
- 设置`CMAKE_BUILD_TYPE`为`Release`
- 使用`clang`编译工具链
- `armeabi-v7a`时，设置`USE_EIGEN_BLAS=ON`，使用Eigen进行矩阵计算；`arm64-v8a`时，设置`USE_EIGEN_FOR_BLAS=OFF`，使用OpenBLAS进行矩阵计算
### 编译和安装
CMake配置完成后，执行以下命令，PaddlePaddle将自动下载和编译所有第三方依赖库、编译和安装PaddlePaddle预测库。
```bash
make
make install
```
注意：如果你曾经在源码目录下编译过其他平台的PaddlePaddle库，请先使用`rm -rf`命令删除`third_party`目录和`build`目录，以确保所有的第三方依赖库和PaddlePaddle代码都是针对新的CMake配置重新编译的。
执行完安装命令后，`your/path/to/install`目录中会包含`include`、`lib`和`third_party`目录，其中`include`中包含C-API的头文件，`lib`中包含若干个不同Android ABI的PaddlePaddle库，`third_party`中包含所依赖的所有第三方库。自此，PaddlePaddle的已经安装完成，用户可将`your/path/to/install`目录下的生成文件用于深度学习相关Android App中，调用方法见C-API文档。
doc/mobile/cross_compiling_for_android_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/mobile/cross_compiling_for_ios_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
# iOS平台编译指南
交叉编译iOS平台上适用的PaddlePaddle库，需要在MacOS系统上进行。本文的将介绍在MacOS上，从源码交叉编译iOS平台上适用的PaddlePaddle库。
## 准备交叉编译环境
Apple官方为iOS开发提供了完整的交叉编译工具和集成开发环境，用户从App Store下载安装Xcode即可。也可自行前往官网下载，[Xcode](https://developer.apple.com/cn/xcode/)。安装完成之后，可在命令行执行`xcodebuild -version`，判断是否安装成功。
```bash
$ xcodebuild -version
Xcode 9.0
Build version 9A235
```
## 配置交叉编译参数
PaddlePaddle为交叉编译提供了工具链配置文档[cmake/cross_compiling/ios.cmake](https://github.com/PaddlePaddle/Paddle/blob/develop/cmake/cross_compiling/ios.cmake)，以提供一些默认的编译器和编译参数配置。
交叉编译iOS版本的PaddlePaddle库时，有一些必须配置的参数：
- `CMAKE_SYSTEM_NAME`，CMake编译的目标平台，必须设置为`iOS`。在设置`CMAKE_SYSTEM_NAME=iOS`后，PaddlePaddle的CMake系统会自动编译所有的第三方依赖库，并且强制设置一些PaddlePaddle参数的值（`WITH_C_API=ON`、`WITH_GPU=OFF`、`WITH_AVX=OFF`、`WITH_PYTHON=OFF`、`WITH_RDMA=OFF`）。
- `WITH_C_API`，是否编译C-API预测库，必须设置为ON。在iOS平台上只支持使用C-API来预测。
- `WITH_SWIG_PY`，必须设置为`OFF`。在iOS平台上不支持通过swig调用来训练或者预测。
iOS平台可选配置参数：
- `IOS_PLATFORM`，可设置为`OS`（默认值）或`SIMULATOR`。
- `OS`，构建目标为`arm`架构的iPhone或者iPad等物理设备。
- `SIMULATOR`，构建目标为`x86`架构的模拟器平台。
- `IOS_ARCH`，目标架构。针对不同的`IOS_PLATFORM`，可设置的目标架构如下表所示，默认编译所有架构：
<table class="docutils">
<colgroup>
<col width="35%" />
<col width="65%" />
</colgroup>
<thead valign="bottom">
<tr class="row-odd">
<th class="head">IOS_PLATFORM</th>
<th class="head">IOS_ARCH</th>
</tr>
</thead>
<tbody valign="top">
<tr class="row-even">
<td>OS</td>
<td>armv7, armv7s, arm64 </td>
</tr>
<tr class="row-odd">
<td>SIMULATOR</td>
<td>i386, x86_64 </td>
</tr>
</tbody>
</table>
- `IOS_DEPLOYMENT_TARGET`，最小的iOS部署版本，默认值为`7.0`。
- `IOS_ENABLE_BITCODE`，是否使能[Bitcode](https://developer.apple.com/library/content/documentation/IDEs/Conceptual/AppDistributionGuide/AppThinning/AppThinning.html#//apple_ref/doc/uid/TP40012582-CH35-SW3)，可设置`ON/OFF`，默认值为`ON`。
- `IOS_USE_VECLIB_FOR_BLAS`，是否使用[vecLib](https://developer.apple.com/documentation/accelerate/veclib)框架进行BLAS矩阵计算，可设置`ON/OFF`，默认值为`OFF`。
- `IOS_DEVELOPMENT_ROOT`，`Developer`目录，可显式指定为`/path/to/platform/Developer`。若未显式指定，PaddlePaddle将会根据`IOS_PLATFORM`自动选择`Xcode`对应`platform`的`Developer`目录。
- `IOS_SDK_ROOT`，所使用`SDK`的根目录，可显式指定为`/path/to/platform/Developer/SDKs/SDK`。若未显式指定，PaddlePaddle将会自动选择`IOS_DEVELOPMENT_ROOT`目录下最新的`SDK`版本。
其他配置参数：
- `USE_EIGEN_FOR_BLAS`，是否使用Eigen库进行矩阵计算，在`IOS_USE_VECLIB_FOR_BLAS=OFF`时有效。可设置`ON/OFF`，默认值为`OFF`。
- `HOST_C/CXX_COMPILER`，宿主机的C/C++编译器。默认值为环境变量`CC/CXX`的值；若环境变量`CC/CXX`未设置，则使用`cc/c++`编译器。
常用的cmake配置如下：
```bash
cmake -DCMAKE_SYSTEM_NAME=iOS \
-DIOS_PLATFORM=OS \
-DIOS_ARCH="armv7;arm64" \
-DIOS_ENABLE_BITCODE=ON \
-DIOS_USE_VECLIB_FOR_BLAS=ON \
-DCMAKE_INSTALL_PREFIX=your/path/to/install \
-DWITH_C_API=ON \
-DWITH_TESTING=OFF \
-DWITH_SWIG_PY=OFF \
..
```
```bash
cmake -DCMAKE_SYSTEM_NAME=iOS \
-DIOS_PLATFORM=SIMULATOR \
-DIOS_ARCH="x86_64" \
-DIOS_USE_VECLIB_FOR_BLAS=ON \
-DCMAKE_INSTALL_PREFIX=your/path/to/install \
-DWITH_C_API=ON \
-DWITH_TESTING=OFF \
-DWITH_SWIG_PY=OFF \
..
```
用户还可根据自己的需求设置其他编译参数。比如希望最小化生成库的大小，可以设置`CMAKE_BUILD_TYPE`为`MinSizeRel`；若希望得到最快的执行速度，则可设置`CMAKE_BUILD_TYPE`为`Release`。亦可以通过手动设置`CMAKE_C/CXX_FLAGS`来影响PaddlePaddle的编译过程。
**性能TIPS**，为了达到最快的计算速度，在CMake参数配置上，有以下建议：
- 设置`CMAKE_BUILD_TYPE`为`Release`
- 设置`IOS_USE_VECLIB_FOR_BLAS=ON`，调用`vecLib`框架提供的BLAS函数进行矩阵计算。
## 编译和安装
CMake配置完成后，执行以下命令，PaddlePaddle将自动下载和编译所有第三方依赖库、编译和安装PaddlePaddle预测库。
```
$ make
$ make install
```
注意：如果你曾在源码目录下编译过其他平台的PaddlePaddle库，请先使用`rm -rf`命令删除`third_party`目录和`build`目录，以确保所有的第三方依赖库和PaddlePaddle代码都是针对新的CMake配置重新编译的。
执行完安装命令后，`your/path/to/install`目录中会包含以下内容：
- `include`目录，其中包含所有C-API的头文件
- `lib`目录，其中包含PaddlePaddle的C-API静态库
- `third_party`目录，其中包含所依赖的所有第三方库
注意，如果PaddlePaddle库需要同时支持真机和模拟器，则需要分别编译真机和模拟器版本，然后使用`lipo`工具合并fat库。
自此，PaddlePaddle库已经安装完成，用户可将合成的fat库用于深度学习相关的iOS App中，调用方法见C-API文档。
doc/mobile/cross_compiling_for_ios_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/mobile/cross_compiling_for_raspberry_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
# Raspberry Pi平台编译指南
通常有两个方法来构建基于 Rasspberry Pi 的版本：
1. 通过ssh等方式登录到Raspberry Pi系统上来构建。所需的开发工具和第三方库可以参考 [`/Dockerfile`](https://github.com/PaddlePaddle/Paddle/blob/develop/Dockerfile)。
1. 另一个方法是交叉编译。这篇文档介绍在 Linux/x64 上交叉编译Raspberry Pi平台上适用的PaddlePaddle的方法和步骤。
## 安装交叉编译器
克隆下面 Github repo
```bash
git clone https://github.com/raspberrypi/tools.git
```
即可在 `./tools/tree/master/arm-bcm2708/gcc-linaro-arm-linux-gnueabihf-raspbian-x64` 目录里找到交叉编译器 arm-linux-gnueabihf-gcc 4.8.3。运行该编译工具链需要一台 Linux x64 机器上以及 2.14版本以上的 glibc。
## 配置交叉编译参数
CMake[支持交叉编译](https://cmake.org/cmake/help/v3.0/manual/cmake-toolchains.7.html#cross-compiling)。PaddlePaddle for Raspberry Pi的配置信息在[cmake/cross_compiling/raspberry_pi.cmake](https://github.com/PaddlePaddle/Paddle/blob/develop/cmake/cross_compiling/raspberry_pi.cmake)。
交叉编译Raspberry Pi版本PaddlePaddle库时，有一些必须配置的参数：
- `CMAKE_SYSTEM_NAME`：CMake编译的目标平台，必须配置为`RPi`。在设置`CMAKE_SYSTEM_NAME=RPi`后，PaddlePaddle的CMake系统才认为在是在交叉编译Raspberry Pi系统的版本，并自动编译宿主机版protoc可执行文件、目标机版protobuf库、以及目标机版OpenBLAS库。
- `RPI_TOOLCHAIN`：编译工具链所在的绝对路径，或者相对于构建目录的相对路径。PaddlePaddle的CMake系统将根据该值自动设置需要使用的交叉编译器；否则，用户需要在cmake时手动设置这些值。无默认值。
- `RPI_ARM_NEON`：是否使用NEON指令。目前必须设置成`ON`，默认值为`ON`。
- `HOST_C/CXX_COMPILER`，宿主机的C/C++编译器。在编译宿主机版protoc可执行文件和目标机版OpenBLAS库时需要用到。默认设置成环境变量`CC`的值；若环境变量`CC`没有设置，则设置成`cc`编译器。
一个常用的CMake配置如下：
```
cmake -DCMAKE_SYSTEM_NAME=RPi \
-DRPI_TOOLCHAIN=your/path/to/arm-bcm2708/gcc-linaro-arm-linux-gnueabihf-raspbian-x64 \
-DRPI_ARM_NEON=ON \
-DCMAKE_INSTALL_PREFIX=your/path/to/install \
-DWITH_GPU=OFF \
-DWITH_C_API=ON \
-DWITH_PYTHON=OFF \
-DWITH_SWIG_PY=OFF \
..
```
其中`WITH_C_API=ON`表示需要构建推理库。
用户还可根据自己的需求设置其他编译参数。比如希望最小化生成的库的大小，可以设置`CMAKE_BUILD_TYPE`为`MinSizeRel`；若希望最快的执行速度，则可设置`CMAKE_BUILD_TYPE`为`Release`。
## 编译和安装
CMake配置完成后，执行以下命令，PaddlePaddle将自动下载和编译所有第三方依赖库、编译和安装PaddlePaddle。
```bash
make
make install
```
注意：如果你曾经在源码目录下编译过其他平台的PaddlePaddle库，请先使用`rm -rf`命令删除`third_party`目录和`build`目录，以确保所有的第三方依赖库和PaddlePaddle代码都是针对新的CMake配置重新编译的。
执行完安装命令后，`your/path/to/install`目录中会包含`include`和`lib`目录，其中`include`中包含C-API的头文件，`lib`中包含一个Raspberry Pi版本的库。
doc/mobile/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/mobile/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/survey/cluster_bootstrapping_tools.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/survey/dynamic_graph.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/survey/op_fusion_design.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/templates/conf.py.cn.in 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/templates/conf.py.en.in 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/templates/layout.html 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/config/activation.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/config/attr.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/config/evaluators.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/config/layer.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/config/networks.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/config/optimizer.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/config/pooling.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/data.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/data/data_reader.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/data/dataset.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/data/image.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/model_configs.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/overview.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/api/run_logic.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/build_and_install/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/build_and_install/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/build_and_install/paddleci.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/build_and_install/pip_install_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/build_and_install/pip_install_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/README.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/checkpointing.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/data_dispatch.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/master_server.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/pserver_client.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/save_model.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/dataset.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/dataset.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/init_lock.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/paddle-etcd.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/paddle-ps-0.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/paddle-ps-1.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/submit-job.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/trainer.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/src/trainer.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/cluster_train/submit-job.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/interface/00.why_plain_c.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/interface/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/interface/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/mkl/image/engine.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/mkl/image/gradients.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/mkl/image/layers.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/mkl/image/matrix.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/mkl/image/overview.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/mkl/mkl_packed.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/design/mkl/mkldnn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/contribute_to_paddle_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/contribute_to_paddle_en.md 已删除 120000 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/new_layer_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/new_layer_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/src/FullyConnected.jpg 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/src/doc_en.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/write_docs_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/dev/write_docs_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/build_and_install/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/build_and_install/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/cluster/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/cluster/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/local/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/local/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/local/src/reduce_min_pool_size.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/local/src/word2vec_config.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/local/src/word2vec_dataprovider.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/model/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/model/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/parameter/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/faq/parameter/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/getstarted/concepts/src/infer.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/getstarted/concepts/src/train.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/getstarted/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/getstarted/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/getstarted/quickstart_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/getstarted/quickstart_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/compile_paddle_lib_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/compile_paddle_lib_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/images/csr.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/images/sequence_data.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/images/workflow_of_CAPI.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/workflow_of_capi_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/capi/workflow_of_capi_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/cmd_argument_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/cmd_argument_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/multi_cluster/k8s_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/multi_cluster/k8s_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/preparations_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/preparations_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/Dockerfile 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/efs_mount.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/managed_policy.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/ps_cn.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/ps_en.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/trainer.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/trainer_cn.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cluster/src/word2vec/prepare.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cmd_parameter/arguments_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cmd_parameter/arguments_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cmd_parameter/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cmd_parameter/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cmd_parameter/use_case_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/cmd_parameter/use_case_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/optimization/nvvp1.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/optimization/nvvp2.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/optimization/nvvp3.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/optimization/nvvp4.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/hierarchical_layer_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/hierarchical_layer_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/hrnn_rnn_api_compare_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/hrnn_rnn_api_compare_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/recurrent_group_cn.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/recurrent_group_en.md 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/rnn_config_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/rnn_config_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/src/bi_lstm.jpg 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/howto/rnn/src/glossary_rnn.dot 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/FullyConnected.jpg 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/add_security_group.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/bi_lstm.jpg 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/checkpointing.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/create_efs.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/csr.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/data_dispatch.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/dataset.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/dataset.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/doc_en.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/efs_mount.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/engine.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/file_storage.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/file_storage.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/glossary_rnn.dot 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/glossary_rnn_with_memory.dot 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/gradients.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/init_lock.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/init_lock.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/k8s-paddle-arch.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/layers.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/managed_policy.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/matrix.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/nvvp1.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/nvvp2.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/nvvp3.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/nvvp4.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/overview.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-cloud-in-data-center.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-etcd.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-etcd.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-model-sharding.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-model-sharding.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-ps-0.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-ps-1.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-ps.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-task-queues.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-task-queues.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-task-states.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/paddle-task-states.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/ps_cn.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/ps_en.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/pserver_and_trainer.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/pserver_init.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/pserver_init.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/route53_create_recordset.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/route53_create_zone.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/sequence_data.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/simple_full_recurrent.dot 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/submit-job.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/submit-job.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/trainer.graffle 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/trainer.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/trainer_cn.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/worker_security_group.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/images/workflow_of_CAPI.png 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/index_cn.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
doc/v2/index_en.rst 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/.gitignore 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/cmd/master/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/cmd/master/master.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/cmd/pserver/.gitignore 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/cmd/pserver/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/cmd/pserver/pserver.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/connection/conn.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/c/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/c/client.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/client.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/client_internal_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/client_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/etcd_client.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/inmem_store.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/service.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/service_internal_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/master/service_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/proto/.gitignore 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/.gitignore 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/cclient.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/test/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/test/test_cclient.c 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/test/test_mnist.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/test/test_train.py 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/c/test/testdata/optimizer.pb 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/client.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/client_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/etcd_client.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/client/etcd_client_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/etcd_client.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/optimizer.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/optimizer_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/service.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/service_internal_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/pserver/service_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/utils/networkhelper/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/utils/networkhelper/helper.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
go/utils/networkhelper/helper_test.go 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/CMakeLists.txt
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/API.spec
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/block_desc.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/commit.h.in 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/details/graph_test_base.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/details/inplace_op_pass.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/details/inplace_op_pass.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/details/memory_optimize_helper.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/details/memory_optimize_pass.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/details/memory_optimize_pass.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/executor.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/inplace_op_inference.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/inplace_op_inference_test.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/graph.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/graph.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/node.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/pass.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/repeated_fc_relu_fuse_pass.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/repeated_fc_relu_fuse_pass.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/squared_mat_sub_fuse_pass.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ir/squared_mat_sub_fuse_pass.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/ngraph_bridge.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/framework/ngraph_bridge.h 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/framework/ngraph_operator.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/framework/ngraph_operator.h 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/framework/op_info.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/operator.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/operator.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/operator_kernel_configs.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/python_headers.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/scope.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/tensor.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/framework/type_defs.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/imperative/README.md 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/imperative/layer.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/imperative/layer.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/imperative/tracer.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/imperative/tracer.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/imperative/type_defs.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/CMakeLists.txt 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/CMakeLists.txt 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/fc.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/fc.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/op_converter.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/registrar.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/registrar.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/test_fc_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/convert/ut_helper.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/engine.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/engine.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/anakin/test_anakin_engine.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/api/api.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/api/details/CMakeLists.txt 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/api/helper.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/tensorrt/trt_int8_calibrator.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/inference/tensorrt/trt_int8_calibrator.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/memory/CMakeLists.txt
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/memory/memcpy.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/activation_cudnn.cu.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/activation_cudnn_op.cu.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/alloc_continuous_space_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/beam_search_op.cu.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/beam_search_op_test.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/benchmark/CMakeLists.txt 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester_config.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester_config.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/concat_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/conv_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/conv_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/conv_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/data_norm_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/data_norm_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/box_clip_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/box_clip_op.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/box_clip_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/mask_util.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/mask_util.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/mask_util_test.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/yolov3_loss_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/detection/yolov3_loss_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/expand_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/expand_op.cu
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/fc_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/fc_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/gather_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/gather_op.cu
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/gather_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/gru_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/gru_op.cu.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/gru_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/hash_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/hash_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/is_empty_op.cu.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/embseqpool.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/embseqpool.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/hopv.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/hopv.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/matmul.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/matmul.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/sgd.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/gen/sgd.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/helper.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/jit/test.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/lrn_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/lstm_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/lstmp_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/lstmp_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/beam_search.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/beam_search.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/beam_search.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/beam_search_test.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/blas.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/sample_prob.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/sample_prob.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/sample_prob.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/tree2col.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/tree2col.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/math/tree2col.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/activation_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/batch_norm_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/concat_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/conv_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/dequantize_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/fc_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/lrn_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/pool_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/quantize_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/softmax_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/sum_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/mkldnn/transpose_mkldnn_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/nce_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/CMakeLists.txt 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ngraph_bridge.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ngraph_bridge.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ngraph_engine.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ngraph_engine.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ngraph_engine_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ngraph_engine_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ngraph_ops.h 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/ngraph/ops/CMakeLists.txt 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/accuracy_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/activation_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/adam_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/batch_norm_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/binary_unary_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/concat_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/conv2d_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/cross_entropy_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/elementwise_add_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/elementwise_node.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/momentum_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/op_bridge.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/pool2d_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/softmax_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/ngraph/ops/sum_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/norm_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/pool_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/pool_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/py_func_op.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/quantize_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/sample_logits_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/sample_logits_op.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/sample_logits_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/scale_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/shape_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/shuffle_channel_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/shuffle_channel_op.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/shuffle_channel_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/slice_op.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/softmax_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/stack_op.cu
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/sum_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/transpose_mkldnn_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/tree_conv_op.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/tree_conv_op.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/tree_conv_op.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/operators/yolov3_loss_op.cc 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/operators/yolov3_loss_op.h 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/fluid/platform/cpu_info.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/cudnn_desc.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/cudnn_desc_test.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/enforce.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/event.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/gpu_info.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/init.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/mkldnn_utils.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/nccl_helper.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/place.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/profiler.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/profiler.cu 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/platform/profiler.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/CMakeLists.txt
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/const_value.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/imperative.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/imperative.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/inference_api.cc 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/inference_api.h 0 → 100644
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/ir.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/protobuf.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/pybind/pybind.cc
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/string/pretty_log.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/string/printf.h
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/fluid/train/demo/README.md
浏览文件 @ a9ea99d7
此差异已折叠。
paddle/legacy/api/Arguments.cpp 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/legacy/api/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/legacy/api/ConfigParser.cpp 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/legacy/api/Evaluator.cpp 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
paddle/legacy/api/GradientMachine.cpp 已删除 100644 → 0
浏览文件 @ eac5a0aa
此差异已折叠。
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或