Skip to content

P
Paddle

Crayon鑫 / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle
提交 955fad7a 编写于 作者: P phlrain
浏览文件
操作

Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop

上级 a7fe3b50 8e4ad008
展开全部 隐藏空白更改
内联 并排
Showing with 45280 addition and 42429 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
.github/ISSUE_TEMPLATE/---feature-request-.md 0 → 100644
浏览文件 @ 955fad7a
---
name: 建议(Feature request)
about: 您可以提出您的建议。 You could use this template for reporting a suggestion  issue.
---
欢迎您对PaddlePaddle提出建议，非常感谢您对PaddlePaddle的贡献！
在留下您的建议时，辛苦您同步提供如下信息：
- 版本、环境信息
1）PaddlePaddle版本：请提供您的PaddlePaddle版本号，例如1.1
2）CPU/GPU：您是否使用GPU进行训练，如是，请提供您的CUDA和cuDNN版本号
3）系统环境：请您描述系统类型、版本，例如Mac OS 10.14
- 复现信息：如为报错，请给出复现环境、复现步骤
- 建议描述：请您详细描述，您认为需优化的功能
Thank you for contributing to PaddlePaddle.
Before submitting the issue, you could search issue in the github in case that there was a similar issue submitted or resolved before.
Please make sure that this is a feature request.
**System information**
-PaddlePaddle version （eg.1.1）or CommitID
-CPU: including CPUMKL/OpenBlas/MKLDNN version
-GPU: including CUDA/CUDNN version
-OS Platform (eg.Mac OS 10.14)
**To Reproduce**
Steps to reproduce the behavior
**Describe the feature and the current behavior/state.**
**Any Other info.**
.github/ISSUE_TEMPLATE/---inference-issue-.md 0 → 100644
浏览文件 @ 955fad7a
---
name: 预测（Inference Issue）
about: 您可以提问预测中报错、应用等问题。 You could use this template for reporting an inference issue.
---
为使您的问题得到快速解决，在建立Issue前，请您先通过如下方式搜索是否有相似问题:【搜索issue关键字】【使用labels筛选】【官方文档】
如果您没有查询到相似问题，为快速解决您的提问，建立issue时请提供如下细节信息：
- 标题：简洁、精准描述您的问题，例如“最新预测库的API文档在哪儿 ”
- 版本、环境信息：
   1）PaddlePaddle版本：请提供您的PaddlePaddle版本号（如1.1）或CommitID
   2）CPU：预测若用CPU，请提供CPU型号，MKL/OpenBlas/MKLDNN/等数学库使用情况
   3）GPU：预测若用GPU，请提供GPU型号、CUDA和CUDNN版本号
   4）系统环境：请您描述系统类型、版本（如Mac OS 10.14），Python版本
-预测信息
   1）C++预测：请您提供预测库安装包的版本信息，及其中的version.txt文件
   2）CMake包含路径的完整命令
   3）API信息（如调用请提供）
   4）预测库来源：官网下载/特殊环境（如BCLOUD编译）
- 复现信息：如为报错，请给出复现环境、复现步骤
- 问题描述：请详细描述您的问题，同步贴出报错信息、日志/代码关键片段
Thank you for contributing to PaddlePaddle.
Before submitting the issue, you could search issue in the github in case that th
If there is no solution,please make sure that this is an inference issue including the following details :
**System information**
-PaddlePaddle version （eg.1.1）or CommitID
-CPU: including CPUMKL/OpenBlas/MKLDNN version
-GPU: including CUDA/CUDNN version
-OS Platform (eg.Mac OS 10.14)
-Python version
-Cmake orders
-C++version.txt
-API information
**To Reproduce**
Steps to reproduce the behavior
**Describe your current behavior**
**Code to reproduce the issue**
**Other info / logs**
.github/ISSUE_TEMPLATE/---installation-issue-.md 0 → 100644
浏览文件 @ 955fad7a
---
name: 安装（Installation Issue）
about: 您可以提问安装、编译出现报错等问题。 You could use this template for reporting an installation
 issue.
---
为使您的问题得到快速解决，在建立Issue前，请您先通过如下方式搜索是否有相似问题:【搜索issue关键字】【使用labels筛选】【官方文档】
建立issue时，为快速解决问题，请您根据使用情况给出如下信息：
- 标题：请包含关键词“安装错误”/“编译错误”，例如“Mac编译错误”
- 版本、环境信息：
   1）PaddlePaddle版本：请提供您的PaddlePaddle版本号（如1.1）或CommitID
   2）CPU：请提供CPU型号，MKL/OpenBlas/MKLDNN/等数学库的使用情况
   3）GPU：请提供GPU型号，CUDA和CUDNN版本号
   4）系统环境：请说明系统类型、版本（如Mac OS 10.14）、Python版本
- 安装方式信息：
1）pip安装/docker安装
2）本地编译：请提供cmake命令，编译命令
3）docker编译：请提供docker镜像，编译命令           
 特殊环境请注明：如离线安装等
- 复现信息：如为报错，请给出复现环境、复现步骤
- 问题描述：请详细描述您的问题，同步贴出报错信息、日志/代码关键片段
Thank you for contributing to PaddlePaddle.
Before submitting the issue, you could search issue in Github in case that there was a similar issue submitted or resolved before.
If there is no solution,please make sure that this is an installation issue including the following details:
**System information**
-PaddlePaddle version （eg.1.1）or CommitID
-CPU: including CPUMKL/OpenBlas/MKLDNN version
-GPU: including CUDA/CUDNN version
-OS Platform (eg. Mac OS 10.14)
-Python version
- Install method: pip install/install with docker/build from source(without docker)/build within docker
- Other special cases that you think may be related to this problem, eg. offline install, special internet condition  
**To Reproduce**
Steps to reproduce the behavior
**Describe your current behavior**
**Code to reproduce the issue**
**Other info / logs**
.github/ISSUE_TEMPLATE/---model-issue-.md 0 → 100644
浏览文件 @ 955fad7a
---
name: 模型（Model Issue）
about: 您可以提问模型、算法、数据集方向的使用报错等问题。You could use this template for reporting a model/
algorithm/dataset  issue.
---
为使您的问题得到快速解决，在建立Issue前，请您先通过如下方式搜索是否有相似问题:【搜索issue关键字】【使用labels筛选】【官方文档】
建立issue时，为快速解决问题，请您根据使用情况给出如下信息：
- 标题：简洁、精准描述您的问题，例如“ssd 模型前置lstm报错  ”
- 版本、环境信息：
   1）PaddlePaddle版本：请提供PaddlePaddle版本号，例如1.1或CommitID
   2）CPU：请提供CPU型号，MKL/OpenBlas/MKLDNN/等数学库的使用情况
   3）GPU：请提供GPU型号，CUDA和CUDNN版本号
   4）系统环境：请说明系统类型、版本（例如Mac OS 10.14），Python版本
- 模型信息
   1）模型名称 2）使用数据集名称 3）使用算法名称 4）模型链接
- 复现信息：如为报错，请给出复现环境、复现步骤
- 问题描述：请详细描述您的问题，同步贴出报错信息、日志/代码关键片段
Thank you for contributing to PaddlePaddle.
Before submitting the issue, you could search issue in the github.Probably there was a similar issue submitted or resolved before.
If there is no solution,please make sure that this is a issue of models including the following details:
**System information**
-PaddlePaddle version （eg.1.1）or CommitID
-CPU: including CPUMKL/OpenBlas/MKLDNN version
-GPU: including CUDA/CUDNN version
-OS Platform (eg.Mac OS 10.14)
-Python version
-Name of Models&Dataset/details of operator
**To Reproduce**
Steps to reproduce the behavior
**Describe your current behavior**
**Code to reproduce the issue**
**Other info / logs**
.github/ISSUE_TEMPLATE/---others-.md 0 → 100644
浏览文件 @ 955fad7a
---
name: 其他（Others）
about: 如上述分类未包含您的问题，可在此提出。 You could use this template for reporting other issues
---
为使您的问题得到快速解决，在建立Issues前，请您先通过如下方式搜索是否有相似问题:【搜索issue关键字】【使用labels筛选】【官方文档】
如果您没有查询到相似问题，为快速解决您的提问，建立issue时请提供如下细节信息：
- 标题：简洁、精准概括您的问题
- 版本、环境信息：
   1）PaddlePaddle版本：请提供您的PaddlePaddle版本号，例如1.1或CommitID
   2）CPU/GPU：如果您使用GPU训练，请提供GPU驱动版本、CUDA和cuDNN版本号
   3）系统环境：请您描述系统类型、版本，例如Mac OS 10.14
   4）Python版本号
   5）显存信息
- 复现信息：如为报错，请给出复现环境、复现步骤
- 问题描述：请详细描述您的问题，同步贴出报错信息、日志/代码关键片段
Thank you for contributing to PaddlePaddle.
Before submitting the issue, you could search issue in the github in case that there was a similar issue submitted or resolved before.
If there is no solution,please provide us with the following details :
**System information**
-PaddlePaddle version （eg.1.1）or CommitID
-CPU: including CPUMKL/OpenBlas/MKLDNN version
-GPU: including CUDA/cuDNN version
-OS Platform and Distribution(eg.Mac OS 10.14)
-Python version
**To Reproduce**
Steps to reproduce the behavior
**Describe your current behavior**
**Code to reproduce the issue**
**Other info / logs**
.github/ISSUE_TEMPLATE/---training-issue-.md 0 → 100644
浏览文件 @ 955fad7a
---
name: 训练（Training issue）
about: 您可以提问训练中报错、应用、出core等问题。 You could use this template for reporting an training
 issue.
---
为使您的问题得到快速解决，在建立Issues前，请您先通过如下方式搜索是否有相似问题:【搜索issue关键字】【使用labels筛选】【官方文档】
如果您没有查询到相似问题，为快速解决您的提问，建立issue时请提供如下细节信息：
- 标题：简洁、精准概括您的问题，例如“Insufficient Memory xxx" ”
- 版本、环境信息：
   1）PaddlePaddle版本：请提供您的PaddlePaddle版本号，例如1.1或CommitID
   2）CPU：预测若用CPU，请提供CPU型号，MKL/OpenBlas/MKLDNN/等数学库使用情况
   3）GPU：预测若用GPU，请提供GPU型号、CUDA和CUDNN版本号
   4）系统环境：请您描述系统类型、版本，例如Mac OS 10.14，Python版本
- 训练信息
   1）单机/多机，单卡/多卡
   2）显存信息
   3）Operator信息
- 复现信息：如为报错，请给出复现环境、复现步骤
- 问题描述：请详细描述您的问题，同步贴出报错信息、日志、可复现的代码片段
Thank you for contributing to PaddlePaddle.
Before submitting the issue, you could search issue in the github in case that there was a similar issue submitted or resolved before.
If there is no solution,please make sure that this is a training issue including the following details:
**System information**
-PaddlePaddle version （eg.1.1）or CommitID
-CPU: including CPUMKL/OpenBlas/MKLDNN version
-GPU: including CUDA/CUDNN version
-OS Platform (eg.Mac OS 10.14)
-Other imformation: Distriuted training/informantion of operator/
Graphics card storage
**To Reproduce**
Steps to reproduce the behavior
**Describe your current behavior**
**Code to reproduce the issue**
**Other info / logs**
.gitignore
浏览文件 @ 955fad7a
python/paddle/fluid/tests/unittests/reader_reset_test.recordio
paddle/operators/check_t.save
paddle/operators/check_tensor.ls
paddle/operators/tensor.save
python/paddle/v2/fluid/tests/book/image_classification_resnet.inference.model/
python/paddle/v2/fluid/tests/book/image_classification_vgg.inference.model/
python/paddle/v2/fluid/tests/book/label_semantic_roles.inference.model/
paddle/fluid/operators/distributed/send_recv.proto
*.DS_Store
*.vs
build/
......@@ -28,3 +30,5 @@ third_party/
build_*
# clion workspace.
cmake-build-*
paddle/fluid/operators/distributed/send_recv.proto
model_test
.pre-commit-config.yaml
浏览文件 @ 955fad7a
......@@ -42,12 +42,6 @@ repos:
entry: bash ./tools/codestyle/pylint_pre_commit.hook
language: system
files: \.(py)$
- repo: https://github.com/PaddlePaddle/pre-commit-golang
sha: 8337620115c25ff8333f1b1a493bd031049bd7c0
hooks:
- id: go-fmt
types:
- go
- repo: local
hooks:
- id: copyright_checker
......
.travis.yml
浏览文件 @ 955fad7a
......@@ -4,7 +4,6 @@ cache:
- $HOME/.ccache
- $HOME/.cache/pip
- $TRAVIS_BUILD_DIR/build/third_party
- $TRAVIS_BUILD_DIR/build_android/third_party
sudo: required
dist: trusty
services:
......@@ -13,7 +12,6 @@ os:
- linux
env:
- JOB=check_style
- JOB=build_android
addons:
ssh_known_hosts: 13.229.163.131
before_install:
......
AUTHORS.md
浏览文件 @ 955fad7a
......@@ -25,6 +25,7 @@
| kexinzhao | Ke-Xin Zhao |
| kuke | Yi-Bing Liu |
| lcy-seso | Ying Cao |
| cjld | Dun Liang |
| lipeng-unisound | Peng Li |
| liuyuan | Yuan Liu |
| livc | Zhao Li |
......@@ -42,7 +43,10 @@
| QiJune | Jun Qi |
| qingqing01 | Qing-Qing Dang |
| reyoung | Yang Yu |
| Sand3r- | Michal Gallus |
| sfraczek | Sylwester Fraczek |
| Superjom | Chun-Wei Yan |
| tensor-tang | Jian Tang |
| tianbingsz | Tian-Bing Xu |
| tpatejko | Tomasz Patejko |
| typhoonzero | Yi Wu |
......@@ -51,6 +55,7 @@
| wangyang59 | Yang Wang |
| wangzhen-nlp | Zhen Wang |
| wen-bo-yang | Wen-Bo Yang |
| wojtuss | Wojciech Uss |
| wwhu | Wei-Wei Hu |
| xinghai-sun | Xing-Hai Sun |
| Xreki | Yi-Qun Liu |
......
CMakeLists.txt
浏览文件 @ 955fad7a
......@@ -24,13 +24,24 @@ message(STATUS "CXX compiler: ${CMAKE_CXX_COMPILER}, version: "
"${CMAKE_CXX_COMPILER_ID} ${CMAKE_CXX_COMPILER_VERSION}")
message(STATUS "C compiler: ${CMAKE_C_COMPILER}, version: "
"${CMAKE_C_COMPILER_ID} ${CMAKE_C_COMPILER_VERSION}")
message(STATUS "AR tools: ${CMAKE_AR}")
if(WIN32)
set(CMAKE_SUPPRESS_REGENERATION ON)
set(CMAKE_STATIC_LIBRARY_PREFIX lib)
add_definitions("/DGOOGLE_GLOG_DLL_DECL=")
set(CMAKE_C_FLAGS_DEBUG "${CMAKE_C_FLAGS_DEBUG} /bigobj /MTd")
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_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)
if(NOT CMAKE_CROSSCOMPILING)
find_package(CUDA QUIET)
endif(NOT CMAKE_CROSSCOMPILING)
find_package(CUDA QUIET)
find_package(Git REQUIRED)
find_package(Threads REQUIRED)
......@@ -41,34 +52,24 @@ option(WITH_GPU "Compile PaddlePaddle with NVIDIA GPU" ${CUDA_F
option(WITH_AMD_GPU "Compile PaddlePaddle with AMD GPU" OFF)
option(WITH_AVX "Compile PaddlePaddle with AVX intrinsics" ${AVX_FOUND})
option(WITH_MKL "Compile PaddlePaddle with MKL support." ${AVX_FOUND})
option(WITH_NGRAPH "Compile PaddlePaddle with nGraph support." OFF)
option(WITH_DSO "Compile PaddlePaddle with dynamic linked CUDA" ON)
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_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" OFF)
option(WITH_DOC "Compile PaddlePaddle with documentation" OFF)
option(WITH_PROFILER "Compile PaddlePaddle with GPU profiler and gperftools" OFF)
option(WITH_JEMALLOC "Compile PaddlePaddle with jemalloc" OFF)
option(WITH_COVERAGE "Compile PaddlePaddle with code coverage" 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(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_FAST_BUNDLE_TEST "Bundle tests that can be run in a single process together to reduce launch overhead" OFF)
option(WITH_PSLIB "Compile with pslib support" 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(WITH_ANAKIN "Compile with Anakin library" OFF)
option(ANAKIN_BUILD_FAT_BIN "Build anakin cuda fat-bin lib for all device plantform, ignored when WITH_ANAKIN=OFF" OFF)
option(ANAKIN_BUILD_CROSS_PLANTFORM "Build anakin lib for any nvidia device plantform. ignored when WITH_ANAKIN=OFF" ON)
option(WITH_GRPC "Use grpc as the default rpc framework" ${WITH_DISTRIBUTE})
option(WITH_BRPC_RDMA "Use brpc rdma as the rpc protocal" OFF)
option(WITH_INFERENCE "Compile fluid inference library" ON)
option(ON_INFER "Turn on inference optimization." OFF)
option(WITH_INFERENCE_API_TEST "Test fluid inference high-level api interface" OFF)
option(WITH_SYSTEM_BLAS "Use system blas library" OFF)
option(PY_VERSION "Compile PaddlePaddle with python3 support" ${PY_VERSION})
......@@ -87,38 +88,14 @@ if(NOT CMAKE_BUILD_TYPE)
FORCE)
endif()
if(ANDROID OR IOS)
if(ANDROID)
if(${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)
if (APPLE)
set(WITH_MKL OFF CACHE STRING
"Disable MKL 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)
"Disable MKL for building on mac" FORCE)
endif()
if (APPLE OR WIN32)
set(WITH_MKL OFF CACHE STRING
"Disable MKL for building on mac and windows" FORCE)
if (WIN32)
set(WITH_DISTRIBUTE OFF CACHE STRING
"Disable DISTRIBUTE when compiling for Windows" FORCE)
endif()
set(THIRD_PARTY_PATH "${CMAKE_BINARY_DIR}/third_party" CACHE STRING
......@@ -130,21 +107,7 @@ set(FLUID_INSTALL_DIR "${CMAKE_BINARY_DIR}/fluid_install_dir" CACHE STRING
set(FLUID_INFERENCE_INSTALL_DIR "${CMAKE_BINARY_DIR}/fluid_inference_install_dir" CACHE STRING
"A path setting fluid inference shared and static libraries")
if (WITH_C_API AND WITH_PYTHON)
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(THIRD_PARTY_BUILD_TYPE Release)
set(WITH_MKLML ${WITH_MKL})
if (NOT DEFINED WITH_MKLDNN)
......@@ -172,22 +135,31 @@ include(external/protobuf) # download, build, install protobuf
include(external/python) # download, build, install python
include(external/openblas) # download, build, install openblas
include(external/mkldnn) # download, build, install mkldnn
include(external/swig) # download, build, install swig
include(external/ngraph) # download, build, install nGraph
include(external/boost) # download boost
include(external/any) # download libn::any
include(external/eigen) # download eigen3
include(external/pybind11) # download pybind11
include(external/cares)
include(external/cub)
if (NOT WIN32)
# there is no official support of snappystream, warpctc, nccl, cupti in windows
include(external/rocprim)
include(external/xxhash) # download xxhash
include(external/dlpack)
include(external/snappy) # download snappy
include(external/snappystream) # download snappystream
include(external/warpctc) # download, build, install warpctc
if (NOT WIN32)
# there is no official support of nccl, cupti in windows
include(cupti)
include(external/gzstream)
endif (NOT WIN32)
if(WITH_PSLIB)
include(external/libmct)
include(external/pslib_brpc)
include(external/pslib)
endif(WITH_PSLIB)
if(WITH_DISTRIBUTE)
if(WITH_GRPC)
include(external/grpc)
......@@ -225,79 +197,48 @@ elseif()
set(WITH_ANAKIN OFF CACHE STRING "Anakin is used in MKL only now." FORCE)
endif()
if (WITH_PROFILER)
find_package(Gperftools REQUIRED)
include_directories(${GPERFTOOLS_INCLUDE_DIR})
add_definitions(-DWITH_GPERFTOOLS)
endif()
if (WITH_JEMALLOC)
find_package(JeMalloc REQUIRED)
include_directories(${JEMALLOC_INCLUDE_DIR})
add_definitions(-DPADDLE_WITH_JEMALLOC)
endif()
include(generic) # simplify cmake module
include(package) # set paddle packages
include(ccache) # set ccache for compilation
include(util) # set unittest and link libs
include(rdma) # set rdma libraries
include(version) # set PADDLE_VERSION
include(coveralls) # set code coverage
include(inference_lib) # add paddle fluid inference libraries
include_directories("${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_AMD_GPU)
find_package(HIP)
include(hip)
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(CMAKE_CXX_FLAGS_RELWITHDEBINFO "-O3 -g -DNDEBUG")
set(CMAKE_C_FLAGS_RELWITHDEBINFO "-O3 -g -DNDEBUG")
if (ON_INFER)
message(STATUS "On inference mode, will take place some specific optimization.")
add_definitions(-DPADDLE_ON_INFERENCE)
else()
#TODO(luotao), combine this warning with `make inference_lib_dist` command.
message(WARNING "On inference mode, will take place some specific optimization. Turn on the ON_INFER flag when building inference_lib only.")
endif()
add_subdirectory(paddle)
if(WITH_PYTHON)
add_subdirectory(python)
endif()
if(WITH_DOC)
find_package(Sphinx REQUIRED)
find_python_module(recommonmark REQUIRED)
add_subdirectory(doc)
endif()
CONTRIBUTING.md
浏览文件 @ 955fad7a
......@@ -128,7 +128,7 @@ Please install pre-commit, which automatically reformat the changes to C/C++ and
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/).
......
Dockerfile
浏览文件 @ 955fad7a
......@@ -11,17 +11,38 @@ RUN /bin/bash -c 'if [[ -n ${UBUNTU_MIRROR} ]]; then sed -i 's#http://archive.ub
# ENV variables
ARG WITH_GPU
ARG WITH_AVX
ARG WITH_DOC
ENV WOBOQ OFF
ENV WITH_GPU=${WITH_GPU:-ON}
ENV WITH_AVX=${WITH_AVX:-ON}
ENV WITH_DOC=${WITH_DOC:-OFF}
ENV HOME /root
# Add bash enhancements
COPY ./paddle/scripts/docker/root/ /root/
# Prepare packages for Python
RUN apt-get update && \
apt-get install -y make build-essential libssl-dev zlib1g-dev libbz2-dev \
libreadline-dev libsqlite3-dev wget curl llvm libncurses5-dev libncursesw5-dev \
xz-utils tk-dev libffi-dev liblzma-dev
# Install Python3.6
RUN mkdir -p /root/python_build/ && wget -q https://www.sqlite.org/2018/sqlite-autoconf-3250300.tar.gz && \
tar -zxf sqlite-autoconf-3250300.tar.gz && cd sqlite-autoconf-3250300 && \
./configure -prefix=/usr/local && make -j8 && make install && cd ../ && rm sqlite-autoconf-3250300.tar.gz && \
wget -q https://www.python.org/ftp/python/3.6.0/Python-3.6.0.tgz && \
tar -xzf Python-3.6.0.tgz && cd Python-3.6.0 && \
CFLAGS="-Wformat" ./configure --prefix=/usr/local/ --enable-shared > /dev/null && \
make -j8 > /dev/null && make altinstall > /dev/null
# Install Python3.7
RUN wget -q https://www.python.org/ftp/python/3.7.0/Python-3.7.0.tgz && \
tar -xzf Python-3.7.0.tgz && cd Python-3.7.0 && \
CFLAGS="-Wformat" ./configure --prefix=/usr/local/ --enable-shared > /dev/null && \
make -j8 > /dev/null && make altinstall > /dev/null
RUN rm -r /root/python_build
RUN apt-get update && \
apt-get install -y --allow-downgrades patchelf \
python3 python3-dev python3-pip \
......@@ -54,8 +75,9 @@ RUN curl -s -q https://glide.sh/get | sh
# and its size is only one-third of the official one.
# 2. Manually add ~IPluginFactory() in IPluginFactory class of NvInfer.h, otherwise, it couldn't work in paddle.
# See https://github.com/PaddlePaddle/Paddle/issues/10129 for details.
RUN wget -qO- http://paddlepaddledeps.cdn.bcebos.com/TensorRT-4.0.0.3.Ubuntu-16.04.4.x86_64-gnu.cuda-8.0.cudnn7.0.tar.gz | \
tar -xz -C /usr/local && \
RUN wget -q https://paddlepaddledeps.cdn.bcebos.com/TensorRT-4.0.1.6-ubuntu14.04.x86_64-gnu.cuda.8.0.cudnn7.0.tar.gz --no-check-certificate && \
tar -zxf TensorRT-4.0.1.6-ubuntu14.04.x86_64-gnu.cuda.8.0.cudnn7.0.tar.gz -C /usr/local && \
cp -rf /usr/local/TensorRT/include /usr && \
cp -rf /usr/local/TensorRT/lib /usr
......@@ -71,34 +93,52 @@ RUN localedef -i en_US -f UTF-8 en_US.UTF-8
# specify sphinx version as 1.5.6 and remove -U option for [pip install -U
# sphinx-rtd-theme] since -U option will cause sphinx being updated to newest
# version(1.7.1 for now), which causes building documentation failed.
RUN pip3 install -U wheel && \
pip3 install -U docopt PyYAML sphinx==1.5.6 && \
pip3 install sphinx-rtd-theme==0.1.9 recommonmark && \
RUN pip3 --no-cache-dir install -U wheel && \
pip3 --no-cache-dir install -U docopt PyYAML sphinx==1.5.6 && \
pip3 --no-cache-dir install sphinx-rtd-theme==0.1.9 recommonmark && \
pip3.6 --no-cache-dir install -U wheel && \
pip3.6 --no-cache-dir install -U docopt PyYAML sphinx==1.5.6 && \
pip3.6 --no-cache-dir install sphinx-rtd-theme==0.1.9 recommonmark && \
pip3.7 --no-cache-dir install -U wheel && \
pip3.7 --no-cache-dir install -U docopt PyYAML sphinx==1.5.6 && \
pip3.7 --no-cache-dir install sphinx-rtd-theme==0.1.9 recommonmark && \
easy_install -U pip && \
pip install -U wheel && \
pip install -U docopt PyYAML sphinx==1.5.6 && \
pip install sphinx-rtd-theme==0.1.9 recommonmark
RUN pip3 install pre-commit 'ipython==5.3.0' && \
pip3 install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip3 install opencv-python && \
pip install pre-commit 'ipython==5.3.0' && \
pip install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip install opencv-python
pip --no-cache-dir install -U pip setuptools wheel && \
pip --no-cache-dir install -U docopt PyYAML sphinx==1.5.6 && \
pip --no-cache-dir install sphinx-rtd-theme==0.1.9 recommonmark
RUN pip3 --no-cache-dir install 'pre-commit==1.10.4' 'ipython==5.3.0' && \
pip3 --no-cache-dir install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip3 --no-cache-dir install opencv-python && \
pip3.6 --no-cache-dir install 'pre-commit==1.10.4' 'ipython==5.3.0' && \
pip3.6 --no-cache-dir install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip3.6 --no-cache-dir install opencv-python && \
pip3.7 --no-cache-dir install 'pre-commit==1.10.4' 'ipython==5.3.0' && \
pip3.7 --no-cache-dir install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip3.7 --no-cache-dir install opencv-python && \
pip --no-cache-dir install 'pre-commit==1.10.4' 'ipython==5.3.0' && \
pip --no-cache-dir install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip --no-cache-dir install opencv-python
#For docstring checker
RUN pip3 install pylint pytest astroid isort
RUN pip install pylint pytest astroid isort LinkChecker
RUN pip3 --no-cache-dir install pylint pytest astroid isort
RUN pip3.6 --no-cache-dir install pylint pytest astroid isort
RUN pip3.7 --no-cache-dir install pylint pytest astroid isort
RUN pip --no-cache-dir install pylint pytest astroid isort LinkChecker
COPY ./python/requirements.txt /root/
RUN pip3 install -r /root/requirements.txt
RUN pip install -r /root/requirements.txt
RUN pip3 --no-cache-dir install -r /root/requirements.txt
RUN pip3.6 --no-cache-dir install -r /root/requirements.txt
RUN pip3.7 --no-cache-dir install -r /root/requirements.txt
RUN pip --no-cache-dir install -r /root/requirements.txt
# To fix https://github.com/PaddlePaddle/Paddle/issues/1954, we use
# the solution in https://urllib3.readthedocs.io/en/latest/user-guide.html#ssl-py2
RUN apt-get install -y libssl-dev libffi-dev
RUN pip3 install certifi urllib3[secure]
RUN pip install certifi urllib3[secure]
RUN apt-get install -y libssl-dev libffi-dev && apt-get clean -y
RUN pip3 --no-cache-dir install certifi urllib3[secure]
RUN pip3.6 --no-cache-dir install certifi urllib3[secure]
RUN pip3.7 --no-cache-dir install certifi urllib3[secure]
RUN pip --no-cache-dir install certifi urllib3[secure]
# Install woboq_codebrowser to /woboq
......@@ -108,6 +148,14 @@ RUN git clone https://github.com/woboq/woboq_codebrowser /woboq && \
-DCMAKE_BUILD_TYPE=Release . \
make)
# ar mishandles 4GB files
# https://sourceware.org/bugzilla/show_bug.cgi?id=14625
# remove them when apt-get support 2.27 and higher version
RUN wget -q https://launchpad.net/ubuntu/+archive/primary/+sourcefiles/binutils/2.27-9ubuntu1/binutils_2.27.orig.tar.gz && \
tar -xzf binutils_2.27.orig.tar.gz && \
cd binutils-2.27 && \
./configure && make -j && make install && cd .. && rm -rf binutils-2.27 binutils_2.27.orig.tar.gz
# Configure OpenSSH server. c.f. https://docs.docker.com/engine/examples/running_ssh_service
RUN mkdir /var/run/sshd
RUN echo 'root:root' | chpasswd
......
Dockerfile.android 已删除 100644 → 0
浏览文件 @ a7fe3b50
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 @ 955fad7a
# 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)
[![Documentation Status](https://img.shields.io/badge/docs-latest-brightgreen.svg?style=flat)](http://www.paddlepaddle.org/docs/develop/documentation/en/getstarted/index_en.html)
[![Documentation Status](https://img.shields.io/badge/中文文档-最新-brightgreen.svg)](http://www.paddlepaddle.org/docs/develop/documentation/zh/getstarted/index_cn.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/1.3/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)
Welcome to the PaddlePaddle GitHub.
PaddlePaddle (PArallel Distributed Deep LEarning) is an easy-to-use,
......@@ -18,8 +18,7 @@ learning to many products at Baidu.
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.
### Latest PaddlePaddle Release: [Fluid 1.0.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.0.0)
### Latest PaddlePaddle Release: [Fluid 1.3.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.3)
### Install Latest Stable Release:
```
# Linux CPU
......@@ -27,9 +26,9 @@ pip install paddlepaddle
# Linux GPU cuda9cudnn7
pip install paddlepaddle-gpu
# Linux GPU cuda8cudnn7
pip install paddlepaddle-gpu==0.15.0.post87
pip install paddlepaddle-gpu==1.3.0.post87
# Linux GPU cuda8cudnn5
pip install paddlepaddle-gpu==0.15.0.post85
pip install paddlepaddle-gpu==1.3.0.post85
# For installation on other platform, refer to http://paddlepaddle.org/
```
......@@ -76,30 +75,29 @@ pip install paddlepaddle-gpu==0.15.0.post85
## Installation
It is recommended to read [this doc](http://paddlepaddle.org/documentation/docs/zh/1.0/beginners_guide/index.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.
## Documentation
We provide [English](http://paddlepaddle.org/documentation/docs/en/1.0.0/getstarted/index_en.html) and
[Chinese](http://paddlepaddle.org/documentation/docs/zh/1.0/beginners_guide/index.html) documentation.
We provide [English](http://paddlepaddle.org/documentation/docs/en/1.3/beginners_guide/index_en.html) and
[Chinese](http://paddlepaddle.org/documentation/docs/zh/1.3/beginners_guide/index.html) documentation.
- [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.
- [Distributed Training](http://paddlepaddle.org/documentation/docs/zh/1.0/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.
- [Python API](http://paddlepaddle.org/documentation/api/zh/1.0/fluid.html)
- [Python API](http://paddlepaddle.org/documentation/docs/en/1.3/api/index_en.html)
Our new API enables much shorter programs.
- [How to Contribute](http://paddlepaddle.org/documentation/docs/zh/1.0/advanced_usage/development/contribute_to_paddle.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!
## Ask Questions
You are welcome to submit questions and bug reports as [Github Issues](https://github.com/PaddlePaddle/Paddle/issues).
......
README_cn.md 0 → 100644
浏览文件 @ 955fad7a
# 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 @ a7fe3b50
# 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 @ a7fe3b50
# 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 @ 955fad7a
......@@ -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 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
ADD https://raw.githubusercontent.com/PaddlePaddle/cloud/develop/docker/paddle_k8s /usr/bin
......
benchmark/fluid/args.py
浏览文件 @ 955fad7a
......@@ -142,5 +142,10 @@ def parse_args():
choices=['reduce', 'all_reduce'],
default='all_reduce',
help='Specify the reduce strategy, can be reduce, all_reduce')
parser.add_argument(
'--fuse_broadcast_op',
action='store_true',
help='If set, would fuse multiple broadcast operators into one fused_broadcast operator.'
)
args = parser.parse_args()
return args
benchmark/fluid/fluid_benchmark.py
浏览文件 @ 955fad7a
......@@ -81,9 +81,11 @@ def dist_transpile(trainer_id, args, train_prog, startup_prog):
# the role, should be either PSERVER or TRAINER
training_role = os.getenv("PADDLE_TRAINING_ROLE")
config = distribute_transpiler.DistributeTranspilerConfig()
config = fluid.DistributeTranspilerConfig()
config.slice_var_up = not args.no_split_var
config.min_block_size = 1048576
t = distribute_transpiler.DistributeTranspiler(config=config)
t.transpile(
trainer_id,
# NOTE: *MUST* use train_prog, for we are using with guard to
......@@ -240,7 +242,6 @@ def train_parallel(train_args, test_args, args, train_prog, test_prog,
if args.use_fake_data or args.use_reader_op:
try:
fetch_ret = exe.run(fetch_list)
except fluid.core.EOFException as eof:
break
......
benchmark/paddle/image/alexnet.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
# 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 @ a7fe3b50
#!/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 @ a7fe3b50
# 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 @ a7fe3b50
# 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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ a7fe3b50
# 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 @ a7fe3b50
# 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 @ a7fe3b50
#!/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 @ a7fe3b50
#!/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 @ 955fad7a
......@@ -35,8 +35,6 @@ import os
import argparse
import time
import paddle.v2 as paddle
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument(
"--embedding_dim",
......
benchmark/tensorflow/mnist.py
浏览文件 @ 955fad7a
......@@ -21,7 +21,6 @@ import time
import numpy as np
import tensorflow as tf
import paddle.v2 as paddle
DTYPE = tf.float32
......
benchmark/tensorflow/resnet.py
浏览文件 @ 955fad7a
......@@ -27,7 +27,6 @@ import argparse
import time
import numpy as np
import paddle.v2 as paddle
import tensorflow as tf
DTYPE = tf.float32
......
benchmark/tensorflow/stacked_dynamic_lstm.py
浏览文件 @ 955fad7a
......@@ -21,8 +21,6 @@ import argparse
import time
import tensorflow as tf
import paddle.v2 as paddle
def parse_args():
parser = argparse.ArgumentParser("LSTM model benchmark.")
......
benchmark/tensorflow/vgg.py
浏览文件 @ 955fad7a
......@@ -13,7 +13,6 @@
# limitations under the License.
"""VGG16 benchmark in TensorFlow"""
import tensorflow as tf
import paddle.v2 as paddle
import numpy as np
import argparse
import time
......
cmake/FindGperftools.cmake 0 → 100644
浏览文件 @ 955fad7a
# Tries to find Gperftools.
#
# Usage of this module as follows:
#
# find_package(Gperftools)
#
# Variables used by this module, they can change the default behaviour and need
# to be set before calling find_package:
#
# Gperftools_ROOT_DIR Set this variable to the root installation of
# Gperftools if the module has problems finding
# the proper installation path.
#
# Variables defined by this module:
#
# GPERFTOOLS_FOUND System has Gperftools libs/headers
# GPERFTOOLS_LIBRARIES The Gperftools libraries (tcmalloc & profiler)
# GPERFTOOLS_INCLUDE_DIR The location of Gperftools headers
find_library(GPERFTOOLS_TCMALLOC
NAMES tcmalloc
HINTS ${Gperftools_ROOT_DIR}/lib)
find_library(GPERFTOOLS_PROFILER
NAMES profiler
HINTS ${Gperftools_ROOT_DIR}/lib)
find_library(GPERFTOOLS_TCMALLOC_AND_PROFILER
NAMES tcmalloc_and_profiler
HINTS ${Gperftools_ROOT_DIR}/lib)
find_path(GPERFTOOLS_INCLUDE_DIR
NAMES gperftools/heap-profiler.h
HINTS ${Gperftools_ROOT_DIR}/include)
set(GPERFTOOLS_LIBRARIES ${GPERFTOOLS_TCMALLOC_AND_PROFILER})
include(FindPackageHandleStandardArgs)
find_package_handle_standard_args(
Gperftools
DEFAULT_MSG
GPERFTOOLS_LIBRARIES
GPERFTOOLS_INCLUDE_DIR)
mark_as_advanced(
Gperftools_ROOT_DIR
GPERFTOOLS_TCMALLOC
GPERFTOOLS_PROFILER
GPERFTOOLS_TCMALLOC_AND_PROFILER
GPERFTOOLS_LIBRARIES
GPERFTOOLS_INCLUDE_DIR)
# create IMPORTED targets
if (Gperftools_FOUND AND NOT TARGET gperftools::tcmalloc)
add_library(gperftools::tcmalloc UNKNOWN IMPORTED)
set_target_properties(gperftools::tcmalloc PROPERTIES
IMPORTED_LOCATION ${GPERFTOOLS_TCMALLOC}
INTERFACE_INCLUDE_DIRECTORIES "${GPERFTOOLS_INCLUDE_DIR}")
add_library(gperftools::profiler UNKNOWN IMPORTED)
set_target_properties(gperftools::profiler PROPERTIES
IMPORTED_LOCATION ${GPERFTOOLS_PROFILER}
INTERFACE_INCLUDE_DIRECTORIES "${GPERFTOOLS_INCLUDE_DIR}")
endif()
cmake/FindJeMalloc.cmake 0 → 100644
浏览文件 @ 955fad7a
# - Find JeMalloc library
# Find the native JeMalloc includes and library
#
# JEMALLOC_INCLUDE_DIR - where to find jemalloc.h, etc.
# JEMALLOC_LIBRARIES - List of libraries when using jemalloc.
# JEMALLOC_FOUND - True if jemalloc found.
find_path(JEMALLOC_INCLUDE_DIR
NAMES jemalloc/jemalloc.h
HINTS ${JEMALLOC_ROOT_DIR}/include)
find_library(JEMALLOC_LIBRARIES
NAMES jemalloc
HINTS ${JEMALLOC_ROOT_DIR}/lib)
include(FindPackageHandleStandardArgs)
find_package_handle_standard_args(jemalloc DEFAULT_MSG JEMALLOC_LIBRARIES JEMALLOC_INCLUDE_DIR)
mark_as_advanced(
JEMALLOC_LIBRARIES
JEMALLOC_INCLUDE_DIR)
if (JEMALLOC_FOUND)
add_library(jemalloc::jemalloc UNKNOWN IMPORTED)
set_target_properties(jemalloc::jemalloc PROPERTIES
IMPORTED_LOCATION ${JEMALLOC_LIBRARIES}
INTERFACE_INCLUDE_DIRECTORIES "${JEMALLOC_INCLUDE_DIR}")
endif()
cmake/FindSphinx.cmake 已删除 100644 → 0
浏览文件 @ a7fe3b50
# - 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 @ 955fad7a
......@@ -64,24 +64,18 @@ endif()
## Then find the reference-cblas. www.netlib.org/blas/
set(REFERENCE_CBLAS_ROOT $ENV{REFERENCE_CBLAS_ROOT} CACHE PATH
"Folder contains reference-cblas")
if(NOT CMAKE_CROSSCOMPILING)
set(REFERENCE_CBLAS_INCLUDE_SEARCH_PATHS
${REFERENCE_CBLAS_ROOT}/include
/usr/include
/usr/include/cblas
)
set(REFERENCE_CBLAS_INCLUDE_SEARCH_PATHS
${REFERENCE_CBLAS_ROOT}/include
/usr/include
/usr/include/cblas
)
set(REFERENCE_CBLAS_LIB_SEARCH_PATHS
${REFERENCE_CBLAS_ROOT}/lib
/usr/lib
/usr/lib/blas/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()
set(REFERENCE_CBLAS_LIB_SEARCH_PATHS
${REFERENCE_CBLAS_ROOT}/lib
/usr/lib
/usr/lib/blas/reference/
/usr/lib/reference/
)
if(WITH_SYSTEM_BLAS)
find_path(REFERENCE_CBLAS_INCLUDE_DIR NAMES cblas.h PATHS
......@@ -98,10 +92,3 @@ if(WITH_SYSTEM_BLAS)
message(STATUS "Found reference-cblas (include: ${CBLAS_INC_DIR}, library: ${CBLAS_LIBRARIES})")
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 @ 955fad7a
......@@ -20,45 +20,18 @@ if(WITH_DSO)
add_definitions(-DPADDLE_USE_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)
add_definitions(-DPADDLE_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)
add_definitions(-DPADDLE_DISABLE_PROFILER)
endif(NOT WITH_PROFILER)
if(NOT CMAKE_CROSSCOMPILING)
if(WITH_AVX AND AVX512F_FOUND)
set(SIMD_FLAG ${AVX512F_FLAG})
elseif(WITH_AVX AND AVX2_FOUND)
set(SIMD_FLAG ${AVX2_FLAG})
elseif(WITH_AVX AND AVX_FOUND)
set(SIMD_FLAG ${AVX_FLAG})
elseif(SSE3_FOUND)
set(SIMD_FLAG ${SSE3_FLAG})
endif()
if(WITH_AVX AND AVX_FOUND)
set(SIMD_FLAG ${AVX_FLAG})
elseif(SSE3_FOUND)
set(SIMD_FLAG ${SSE3_FLAG})
endif()
if(WIN32)
......@@ -84,12 +57,13 @@ if(WIN32)
endif(NOT MSVC)
endif(WIN32)
if(NOT WITH_GOLANG)
add_definitions(-DPADDLE_WITHOUT_GOLANG)
endif(NOT WITH_GOLANG)
if(WITH_PSLIB)
add_definitions(-DPADDLE_WITH_PSLIB)
endif()
if(WITH_GPU)
add_definitions(-DPADDLE_WITH_CUDA)
add_definitions(-DEIGEN_USE_GPU)
FIND_PACKAGE(CUDA REQUIRED)
......@@ -133,6 +107,7 @@ if(WITH_GPU)
message(WARNING "Anakin needs CUDNN >= 7.0 to compile. Force WITH_ANAKIN=OFF")
set(WITH_ANAKIN OFF CACHE STRING "Anakin is valid only when CUDNN >= 7.0." FORCE)
endif()
add_definitions(-DWITH_ANAKIN)
endif()
if(WITH_ANAKIN)
# NOTICE(minqiyang): the end slash is important because $CUDNN_INCLUDE_DIR
......@@ -152,7 +127,12 @@ endif()
if (WITH_MKLML AND 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_CXX_CREATE_SHARED_LIBRARY_FORBIDDEN_FLAGS ${OPENMP_FLAGS})
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${OPENMP_FLAGS}")
......@@ -166,55 +146,6 @@ if(WITH_DISTRIBUTE)
add_definitions(-DPADDLE_WITH_DISTRIBUTE)
endif()
if(WITH_GOLANG)
# we need to symlink Paddle directory into GOPATH. If we
# don't do it and we have code that depends on Paddle, go
# get ./... will download a new Paddle repo from Github,
# without the changes in our current Paddle repo that we
# want to build.
set(GOPATH "${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)
add_definitions(-DPADDLE_WITH_GRPC)
endif(WITH_GRPC)
......@@ -222,3 +153,7 @@ endif(WITH_GRPC)
if(WITH_BRPC_RDMA)
add_definitions(-DPADDLE_WITH_BRPC_RDMA)
endif(WITH_BRPC_RDMA)
if(ON_INFER)
add_definitions(-DPADDLE_ON_INFERENCE)
endif(ON_INFER)
cmake/cross_compiling/android.cmake 已删除 100644 → 0
浏览文件 @ a7fe3b50
# 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 @ a7fe3b50
# 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 @ a7fe3b50
# 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 @ a7fe3b50
# 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 @ 955fad7a
......@@ -5,6 +5,8 @@ endif()
set(paddle_known_gpu_archs "30 35 50 52 60 61 70")
set(paddle_known_gpu_archs7 "30 35 50 52")
set(paddle_known_gpu_archs8 "30 35 50 52 60 61")
set(paddle_known_gpu_archs9 "30 35 50 52 60 61 70")
set(paddle_known_gpu_archs10 "30 35 50 52 60 61 70 75")
######################################################################################
# A function for automatic detection of GPUs installed (if autodetection is enabled)
......@@ -59,11 +61,9 @@ endfunction()
# select_nvcc_arch_flags(out_variable)
function(select_nvcc_arch_flags out_variable)
# List of arch names
set(archs_names "Kepler" "Maxwell" "Pascal" "All" "Manual")
set(archs_names "Kepler" "Maxwell" "Pascal" "Volta" "Turing" "All" "Manual")
set(archs_name_default "All")
if(NOT CMAKE_CROSSCOMPILING)
list(APPEND archs_names "Auto")
endif()
list(APPEND archs_names "Auto")
# set CUDA_ARCH_NAME strings (so it will be seen as dropbox in CMake-Gui)
set(CUDA_ARCH_NAME ${archs_name_default} CACHE STRING "Select target NVIDIA GPU achitecture.")
......@@ -93,6 +93,8 @@ function(select_nvcc_arch_flags out_variable)
set(cuda_arch_bin "60 61")
elseif(${CUDA_ARCH_NAME} STREQUAL "Volta")
set(cuda_arch_bin "70")
elseif(${CUDA_ARCH_NAME} STREQUAL "Turing")
set(cuda_arch_bin "75")
elseif(${CUDA_ARCH_NAME} STREQUAL "All")
set(cuda_arch_bin ${paddle_known_gpu_archs})
elseif(${CUDA_ARCH_NAME} STREQUAL "Auto")
......@@ -139,10 +141,12 @@ endfunction()
message(STATUS "CUDA detected: " ${CUDA_VERSION})
if (${CUDA_VERSION} LESS 7.0)
set(paddle_known_gpu_archs ${paddle_known_gpu_archs})
add_definitions("-DPADDLE_CUDA_BINVER=\"60\"")
elseif (${CUDA_VERSION} LESS 8.0) # CUDA 7.x
set(paddle_known_gpu_archs ${paddle_known_gpu_archs7})
list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
add_definitions("-DPADDLE_CUDA_BINVER=\"70\"")
elseif (${CUDA_VERSION} LESS 9.0) # CUDA 8.x
set(paddle_known_gpu_archs ${paddle_known_gpu_archs8})
list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
......@@ -150,13 +154,24 @@ elseif (${CUDA_VERSION} LESS 9.0) # CUDA 8.x
# CUDA 8 may complain that sm_20 is no longer supported. Suppress the
# warning for now.
list(APPEND CUDA_NVCC_FLAGS "-Wno-deprecated-gpu-targets")
add_definitions("-DPADDLE_CUDA_BINVER=\"80\"")
elseif (${CUDA_VERSION} LESS 10.0) # CUDA 9.x
set(paddle_known_gpu_archs ${paddle_known_gpu_archs9})
list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
add_definitions("-DPADDLE_CUDA_BINVER=\"90\"")
elseif (${CUDA_VERSION} LESS 11.0) # CUDA 10.x
set(paddle_known_gpu_archs ${paddle_known_gpu_archs10})
list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
add_definitions("-DPADDLE_CUDA_BINVER=\"100\"")
endif()
include_directories(${CUDA_INCLUDE_DIRS})
list(APPEND EXTERNAL_LIBS ${CUDA_LIBRARIES} ${CUDA_rt_LIBRARY})
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)
set_property(GLOBAL PROPERTY CUDA_MODULES ${CUDNN_LIBRARY} ${CUDA_CUBLAS_LIBRARIES} ${CUDA_curand_LIBRARY})
endif(WIN32)
endif(NOT WITH_DSO)
# setting nvcc arch flags
......@@ -185,21 +200,27 @@ list(APPEND CUDA_NVCC_FLAGS "-w")
list(APPEND CUDA_NVCC_FLAGS "--expt-relaxed-constexpr")
if (NOT WIN32)
if(CMAKE_BUILD_TYPE STREQUAL "Debug")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_DEBUG})
elseif(CMAKE_BUILD_TYPE STREQUAL "Release")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE})
elseif(CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELWITHDEBINFO})
elseif(CMAKE_BUILD_TYPE STREQUAL "MinSizeRel")
# nvcc 9 does not support -Os. Use Release flags instead
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE})
endif()
if(CMAKE_BUILD_TYPE STREQUAL "Debug")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_DEBUG})
elseif(CMAKE_BUILD_TYPE STREQUAL "Release")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE})
elseif(CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELWITHDEBINFO})
elseif(CMAKE_BUILD_TYPE STREQUAL "MinSizeRel")
# nvcc 9 does not support -Os. Use Release flags instead
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE})
endif()
else(NOT WIN32)
if(CMAKE_BUILD_TYPE STREQUAL "Release")
list(APPEND CUDA_NVCC_FLAGS "-O3 -DNDEBUG")
else()
message(FATAL "Windows only support Release build now. Please set visual studio build type to Release, x64 build.")
list(APPEND CUDA_NVCC_FLAGS "-Xcompiler \"/wd 4244 /wd 4267 /wd 4819\"")
list(APPEND CUDA_NVCC_FLAGS "--compiler-options;/bigobj")
if(CMAKE_BUILD_TYPE STREQUAL "Debug")
list(APPEND CUDA_NVCC_FLAGS "-g -G")
# match the cl's _ITERATOR_DEBUG_LEVEL
list(APPEND CUDA_NVCC_FLAGS "-D_DEBUG")
elseif(CMAKE_BUILD_TYPE STREQUAL "Release")
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.")
endif()
endif(NOT WIN32)
......
cmake/cudnn.cmake
浏览文件 @ 955fad7a
......@@ -2,7 +2,12 @@ if(NOT WITH_GPU)
return()
endif()
set(CUDNN_ROOT "/usr" CACHE PATH "CUDNN ROOT")
if(WIN32)
set(CUDNN_ROOT ${CUDA_TOOLKIT_ROOT_DIR})
else(WIN32)
set(CUDNN_ROOT "/usr" CACHE PATH "CUDNN ROOT")
endif(WIN32)
find_path(CUDNN_INCLUDE_DIR cudnn.h
PATHS ${CUDNN_ROOT} ${CUDNN_ROOT}/include
$ENV{CUDNN_ROOT} $ENV{CUDNN_ROOT}/include ${CUDA_TOOLKIT_INCLUDE}
......@@ -39,9 +44,9 @@ if(WIN32)
set(CUDNN_LIB_NAME "cudnn.lib" "cudnn64_7.dll")
endif(WIN32)
if(Apple)
if(APPLE)
set(CUDNN_LIB_NAME "libcudnn.dylib" "libcudnn.so")
endif(Apple)
endif(APPLE)
find_library(CUDNN_LIBRARY NAMES ${CUDNN_LIB_NAME} # libcudnn_static.a
PATHS ${CUDNN_CHECK_LIBRARY_DIRS} ${CUDNN_INCLUDE_DIR} ${__libpath_hist}
......@@ -84,6 +89,7 @@ if(CUDNN_FOUND)
if(NOT CUDNN_MAJOR_VERSION)
set(CUDNN_VERSION "???")
else()
add_definitions("-DPADDLE_CUDNN_BINVER=\"${CUDNN_MAJOR_VERSION}\"")
math(EXPR CUDNN_VERSION
"${CUDNN_MAJOR_VERSION} * 1000 +
${CUDNN_MINOR_VERSION} * 100 + ${CUDNN_PATCHLEVEL_VERSION}")
......
cmake/external/anakin.cmake
浏览文件 @ 955fad7a
......@@ -58,19 +58,19 @@ ExternalProject_Add(
-DPROTOBUF_ROOT=${THIRD_PARTY_PATH}/install/protobuf
-DMKLML_ROOT=${THIRD_PARTY_PATH}/install/mklml
-DENABLE_OP_TIMER=${ANAKIN_ENABLE_OP_TIMER}
-DBUILD_FAT_BIN=${ANAKIN_BUILD_FAT_BIN}
-DBUILD_CROSS_PLANTFORM=${ANAKIN_BUILD_CROSS_PLANTFORM}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${ANAKIN_INSTALL_DIR}
)
message(STATUS "Anakin for inference is enabled")
message(STATUS "Anakin is set INCLUDE:${ANAKIN_INCLUDE} LIBRARY:${ANAKIN_LIBRARY}")
add_dependencies(extern_anakin protobuf mklml)
add_library(anakin_shared SHARED IMPORTED GLOBAL)
set_property(TARGET anakin_shared PROPERTY IMPORTED_LOCATION ${ANAKIN_SHARED_LIB})
add_dependencies(anakin_shared extern_anakin protobuf mklml)
add_dependencies(anakin_shared extern_anakin)
add_library(anakin_saber SHARED IMPORTED GLOBAL)
set_property(TARGET anakin_saber PROPERTY IMPORTED_LOCATION ${ANAKIN_SABER_LIB})
add_dependencies(anakin_saber extern_anakin protobuf mklml)
list(APPEND external_project_dependencies anakin_shared anakin_saber)
add_dependencies(anakin_saber extern_anakin)
cmake/external/any.cmake 已删除 100644 → 0
浏览文件 @ a7fe3b50
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 ${dummyfile})
else()
add_library(lib_any INTERFACE)
endif()
add_dependencies(lib_any extern_lib_any)
add_definitions(-DANY_IMPL_ANY_CAST_MOVEABLE)
LIST(APPEND external_project_dependencies lib_any)
cmake/external/boost.cmake
浏览文件 @ 955fad7a
......@@ -23,39 +23,30 @@ set(BOOST_PROJECT "extern_boost")
# checked that the devtools package of CentOS 6 installs boost 1.41.0.
# So we use 1.41.0 here.
set(BOOST_VER "1.41.0")
if((NOT DEFINED BOOST_TAR) OR (NOT DEFINED BOOST_URL))
message(STATUS "use pre defined download url")
set(BOOST_TAR "boost_1_41_0" CACHE STRING "" FORCE)
set(BOOST_URL "http://paddlepaddledeps.cdn.bcebos.com/${BOOST_TAR}.tar.gz" CACHE STRING "" FORCE)
endif()
IF (WIN32)
MESSAGE(WARNING, "In windows, boost can not be downloaded automaticlly, please build it manually and put it at " ${THIRD_PARTY_PATH}install/boost)
else()
MESSAGE(STATUS "BOOST_TAR: ${BOOST_TAR}, BOOST_URL: ${BOOST_URL}")
ENDIF(WIN32)
set(BOOST_TAR "boost_1_41_0" CACHE STRING "" FORCE)
set(BOOST_URL "http://paddlepaddledeps.cdn.bcebos.com/${BOOST_TAR}.tar.gz" CACHE STRING "" FORCE)
MESSAGE(STATUS "BOOST_TAR: ${BOOST_TAR}, BOOST_URL: ${BOOST_URL}")
set(BOOST_SOURCES_DIR ${THIRD_PARTY_PATH}/boost)
set(BOOST_DOWNLOAD_DIR "${BOOST_SOURCES_DIR}/src/${BOOST_PROJECT}")
set(BOOST_INCLUDE_DIR "${BOOST_DOWNLOAD_DIR}/${BOOST_TAR}" CACHE PATH "boost include directory." FORCE)
set_directory_properties(PROPERTIES CLEAN_NO_CUSTOM 1)
set(BOOST_INCLUDE_DIR "${BOOST_DOWNLOAD_DIR}" CACHE PATH "boost include directory." FORCE)
set_directory_properties(PROPERTIES CLEAN_NO_CUSTOM 1)
include_directories(${BOOST_INCLUDE_DIR})
if (NOT WIN32)
ExternalProject_Add(
${BOOST_PROJECT}
${EXTERNAL_PROJECT_LOG_ARGS}
DOWNLOAD_DIR ${BOOST_DOWNLOAD_DIR}
DOWNLOAD_COMMAND wget --no-check-certificate ${BOOST_URL} -c -q -O ${BOOST_TAR}.tar.gz
&& tar zxf ${BOOST_TAR}.tar.gz
URL ${BOOST_URL}
DOWNLOAD_NO_PROGRESS 1
PREFIX ${BOOST_SOURCES_DIR}
CONFIGURE_COMMAND ""
BUILD_COMMAND ""
INSTALL_COMMAND ""
UPDATE_COMMAND ""
)
endif(NOT WIN32)
)
if (${CMAKE_VERSION} VERSION_LESS "3.3.0" OR NOT WIN32)
set(dummyfile ${CMAKE_CURRENT_BINARY_DIR}/boost_dummy.c)
......@@ -66,5 +57,4 @@ else()
endif()
add_dependencies(boost ${BOOST_PROJECT})
list(APPEND external_project_dependencies boost)
set(Boost_INCLUDE_DIR ${BOOST_INCLUDE_DIR})
cmake/external/brpc.cmake
浏览文件 @ 955fad7a
......@@ -14,14 +14,16 @@
INCLUDE(ExternalProject)
find_library(SSL_LIBRARY NAMES ssl)
find_package(OpenSSL REQUIRED)
message(STATUS "ssl:" ${OPENSSL_SSL_LIBRARY})
message(STATUS "crypto:" ${OPENSSL_CRYPTO_LIBRARY})
ADD_LIBRARY(ssl SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET ssl PROPERTY IMPORTED_LOCATION ${SSL_LIBRARY})
SET_PROPERTY(TARGET ssl PROPERTY IMPORTED_LOCATION ${OPENSSL_SSL_LIBRARY})
find_library(CRYPTO_LIBRARY NAMES crypto)
ADD_LIBRARY(crypto SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET crypto PROPERTY IMPORTED_LOCATION ${CRYPTO_LIBRARY})
SET_PROPERTY(TARGET crypto PROPERTY IMPORTED_LOCATION ${OPENSSL_CRYPTO_LIBRARY})
SET(BRPC_SOURCES_DIR ${THIRD_PARTY_PATH}/brpc)
SET(BRPC_INSTALL_DIR ${THIRD_PARTY_PATH}/install/brpc)
......@@ -31,14 +33,15 @@ SET(BRPC_LIBRARIES "${BRPC_INSTALL_DIR}/lib/libbrpc.a" CACHE FILEPATH "brpc libr
INCLUDE_DIRECTORIES(${BRPC_INCLUDE_DIR})
# Reference https://stackoverflow.com/questions/45414507/pass-a-list-of-prefix-paths-to-externalproject-add-in-cmake-args
set(prefix_path "${THIRD_PARTY_PATH}/install/gflags|${THIRD_PARTY_PATH}/install/leveldb|${THIRD_PARTY_PATH}/install/snappy|${THIRD_PARTY_PATH}/install/gtest|${THIRD_PARTY_PATH}/install/protobuf|${THIRD_PARTY_PATH}/install/zlib")
set(prefix_path "${THIRD_PARTY_PATH}/install/gflags|${THIRD_PARTY_PATH}/install/leveldb|${THIRD_PARTY_PATH}/install/snappy|${THIRD_PARTY_PATH}/install/gtest|${THIRD_PARTY_PATH}/install/protobuf|${THIRD_PARTY_PATH}/install/zlib|${THIRD_PARTY_PATH}/install/glog")
# If minimal .a is need, you can set WITH_DEBUG_SYMBOLS=OFF
ExternalProject_Add(
extern_brpc
${EXTERNAL_PROJECT_LOG_ARGS}
# TODO(gongwb): change to de newst repo when they changed.
GIT_REPOSITORY "https://github.com/gongweibao/brpc"
GIT_TAG "7dc04defad1fd4173aae170c3fcbde131b65155a"
GIT_TAG "e9b67ec1b7458f2af5fae76451afe1e27e01b4b4"
PREFIX ${BRPC_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
......@@ -50,7 +53,7 @@ ExternalProject_Add(
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
-DCMAKE_PREFIX_PATH=${prefix_path}
-DBRPC_WITH_GLOG=ON
-DWITH_GLOG=ON
-DIOBUF_WITH_HUGE_BLOCK=ON
-DBRPC_WITH_RDMA=${WITH_BRPC_RDMA}
${EXTERNAL_OPTIONAL_ARGS}
......@@ -65,5 +68,4 @@ ADD_LIBRARY(brpc STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET brpc PROPERTY IMPORTED_LOCATION ${BRPC_LIBRARIES})
ADD_DEPENDENCIES(brpc extern_brpc)
LIST(APPEND external_project_dependencies brpc)
add_definitions(-DBRPC_WITH_GLOG)
cmake/external/cares.cmake
浏览文件 @ 955fad7a
......@@ -13,7 +13,7 @@
# limitations under the License.
#
IF(MOBILE_INFERENCE OR NOT WITH_DISTRIBUTE)
IF(NOT WITH_DISTRIBUTE)
return()
ENDIF()
......
cmake/external/cub.cmake
浏览文件 @ 955fad7a
......@@ -31,5 +31,3 @@ else()
endif()
add_dependencies(cub extern_cub)
LIST(APPEND externl_project_dependencies cub)
cmake/external/dlpack.cmake 0 → 100644
浏览文件 @ 955fad7a
include(ExternalProject)
set(DLPACK_SOURCE_DIR ${THIRD_PARTY_PATH}/dlpack)
set(DLPACK_INCLUDE_DIR ${DLPACK_SOURCE_DIR}/src/extern_dlpack/include)
include_directories(${DLPACK_INCLUDE_DIR})
ExternalProject_Add(
extern_dlpack
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/dmlc/dlpack.git"
GIT_TAG "v0.2"
PREFIX ${DLPACK_SOURCE_DIR}
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
BUILD_COMMAND ""
INSTALL_COMMAND ""
TEST_COMMAND ""
)
if(${CMAKE_VERSION} VERSION_LESS "3.3.0")
set(dummyfile ${CMAKE_CURRENT_BINARY_DIR}/dlpack_dummy.c)
file(WRITE ${dummyfile} "const char *dummy = \"${dummyfile}\";")
add_library(dlpack STATIC ${dummyfile})
else()
add_library(dlpack INTERFACE)
endif()
add_dependencies(dlpack extern_dlpack)
cmake/external/eigen.cmake
浏览文件 @ 955fad7a
......@@ -17,7 +17,7 @@ if(WITH_AMD_GPU)
extern_eigen3
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/sabreshao/hipeigen.git"
GIT_TAG 0cba03ff9f8f9f70bbd92ac5857b031aa8fed6f9
GIT_TAG 7cb2b6e5a4b4a1efe658abb215cd866c6fb2275e
PREFIX ${EIGEN_SOURCE_DIR}
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
......@@ -52,5 +52,3 @@ else()
endif()
add_dependencies(eigen3 extern_eigen3)
LIST(APPEND external_project_dependencies eigen3)
cmake/external/gflags.cmake
浏览文件 @ 955fad7a
......@@ -35,7 +35,12 @@ ExternalProject_Add(
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
-DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}
-DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}
-DBUILD_STATIC_LIBS=ON
-DCMAKE_INSTALL_PREFIX=${GFLAGS_INSTALL_DIR}
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DBUILD_TESTING=OFF
......@@ -48,22 +53,19 @@ ExternalProject_Add(
IF(WIN32)
IF(NOT EXISTS "${GFLAGS_INSTALL_DIR}/lib/libgflags.lib")
add_custom_command(TARGET extern_gflags POST_BUILD
COMMAND cmake -E rename ${GFLAGS_INSTALL_DIR}/lib/gflags_static.lib ${GFLAGS_INSTALL_DIR}/lib/libgflags.lib
)
COMMAND cmake -E copy ${GFLAGS_INSTALL_DIR}/lib/gflags_static.lib ${GFLAGS_INSTALL_DIR}/lib/libgflags.lib
)
ENDIF()
ENDIF(WIN32)
ADD_LIBRARY(gflags STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET gflags PROPERTY IMPORTED_LOCATION ${GFLAGS_LIBRARIES})
ADD_DEPENDENCIES(gflags extern_gflags)
LIST(APPEND external_project_dependencies gflags)
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 ${GFLAGS_LIBRARIES} DESTINATION third_party/gflags/lib)
ENDIF()
ENDIF()
# On Windows (including MinGW), the Shlwapi library is used by gflags if available.
if (WIN32)
include(CheckIncludeFileCXX)
check_include_file_cxx("shlwapi.h" HAVE_SHLWAPI)
if (HAVE_SHLWAPI)
set_property(GLOBAL PROPERTY OS_DEPENDENCY_MODULES shlwapi.lib)
endif(HAVE_SHLWAPI)
endif (WIN32)
cmake/external/glog.cmake
浏览文件 @ 955fad7a
......@@ -20,20 +20,16 @@ SET(GLOG_INCLUDE_DIR "${GLOG_INSTALL_DIR}/include" CACHE PATH "glog include dire
IF(WIN32)
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)
SET(GLOG_LIBRARIES "${GLOG_INSTALL_DIR}/lib/libglog.a" CACHE FILEPATH "glog library." FORCE)
SET(GLOG_CMAKE_CXX_FLAGS ${CMAKE_CXX_FLAGS})
ENDIF(WIN32)
INCLUDE_DIRECTORIES(${GLOG_INCLUDE_DIR})
IF(ANDROID AND ${CMAKE_SYSTEM_VERSION} VERSION_LESS "21")
# Using the unofficial glog for Android API < 21
SET(GLOG_REPOSITORY "https://github.com/Xreki/glog.git")
SET(GLOG_TAG "8a547150548b284382ccb6582408e9140ff2bea8")
ELSE()
SET(GLOG_REPOSITORY "https://github.com/google/glog.git")
SET(GLOG_TAG "v0.3.5")
ENDIF()
SET(GLOG_REPOSITORY "https://github.com/google/glog.git")
SET(GLOG_TAG "v0.3.5")
ExternalProject_Add(
extern_glog
......@@ -45,8 +41,12 @@ ExternalProject_Add(
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_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_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
-DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}
-DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}
-DCMAKE_INSTALL_PREFIX=${GLOG_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR=${GLOG_INSTALL_DIR}/lib
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
......@@ -63,7 +63,7 @@ ExternalProject_Add(
IF(WIN32)
IF(NOT EXISTS "${GLOG_INSTALL_DIR}/lib/libglog.lib")
add_custom_command(TARGET extern_glog POST_BUILD
COMMAND cmake -E rename ${GLOG_INSTALL_DIR}/lib/glog.lib ${GLOG_INSTALL_DIR}/lib/libglog.lib
COMMAND cmake -E copy ${GLOG_INSTALL_DIR}/lib/glog.lib ${GLOG_INSTALL_DIR}/lib/libglog.lib
)
ENDIF()
ENDIF(WIN32)
......@@ -72,14 +72,3 @@ ADD_LIBRARY(glog STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET glog PROPERTY IMPORTED_LOCATION ${GLOG_LIBRARIES})
ADD_DEPENDENCIES(glog extern_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 @ 955fad7a
......@@ -13,7 +13,7 @@
# limitations under the License.
#
IF(MOBILE_INFERENCE OR NOT WITH_DISTRIBUTE)
IF(NOT WITH_DISTRIBUTE)
return()
ENDIF()
......
cmake/external/gtest.cmake
浏览文件 @ 955fad7a
......@@ -12,8 +12,12 @@
# See the License for the specific language governing permissions and
# limitations under the License.
IF(WITH_TESTING)
ENABLE_TESTING()
#FIXME:(gongwb) Move brpc's gtest dependency.
IF(WITH_TESTING OR (WITH_DISTRIBUTE AND NOT WITH_GRPC))
IF(WITH_TESTING)
ENABLE_TESTING()
ENDIF(WITH_TESTING)
INCLUDE(ExternalProject)
SET(GTEST_SOURCES_DIR ${THIRD_PARTY_PATH}/gtest)
......@@ -50,7 +54,11 @@ IF(WITH_TESTING)
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
-DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}
-DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}
-DCMAKE_INSTALL_PREFIX=${GTEST_INSTALL_DIR}
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DBUILD_GMOCK=ON
......@@ -71,5 +79,4 @@ IF(WITH_TESTING)
SET_PROPERTY(TARGET gtest_main PROPERTY IMPORTED_LOCATION ${GTEST_MAIN_LIBRARIES})
ADD_DEPENDENCIES(gtest_main extern_gtest)
LIST(APPEND external_project_dependencies gtest gtest_main)
ENDIF(WITH_TESTING)
ENDIF(WITH_TESTING OR (WITH_DISTRIBUTE AND NOT WITH_GRPC))
cmake/external/gzstream.cmake 0 → 100644
浏览文件 @ 955fad7a
# 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.
#
include (ExternalProject)
# NOTE: gzstream is needed when linking with ctr reader.
SET(GZSTREAM_SOURCES_DIR ${THIRD_PARTY_PATH}/gzstream)
SET(GZSTREAM_INSTALL_DIR ${THIRD_PARTY_PATH}/install/gzstream)
SET(GZSTREAM_INCLUDE_DIR "${GZSTREAM_INSTALL_DIR}/include/" CACHE PATH "gzstream include directory." FORCE)
ExternalProject_Add(
extern_gzstream
DEPENDS zlib
GIT_REPOSITORY "https://github.com/jacquesqiao/gzstream.git"
GIT_TAG ""
PREFIX ${GZSTREAM_SOURCES_DIR}
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
BUILD_IN_SOURCE 1
BUILD_COMMAND make EXTERN_CPPFLAGS="-I${THIRD_PARTY_PATH}/install/zlib/include" EXTERM_LDFLAGS="-L${THIRD_PARTY_PATH}/install/zlib/lib" -j8
INSTALL_COMMAND mkdir -p ${GZSTREAM_INSTALL_DIR}/lib/ && mkdir -p ${GZSTREAM_INSTALL_DIR}/include/
&& cp ${GZSTREAM_SOURCES_DIR}/src/extern_gzstream/libgzstream.a ${GZSTREAM_INSTALL_DIR}/lib
&& cp -r ${GZSTREAM_SOURCES_DIR}/src/extern_gzstream/gzstream.h ${GZSTREAM_INSTALL_DIR}/include
)
ADD_LIBRARY(gzstream STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET gzstream PROPERTY IMPORTED_LOCATION
"${GZSTREAM_INSTALL_DIR}/lib/libgzstream.a")
include_directories(${GZSTREAM_INCLUDE_DIR})
ADD_DEPENDENCIES(gzstream extern_gzstream zlib)
cmake/external/leveldb.cmake
浏览文件 @ 955fad7a
......@@ -24,8 +24,8 @@ ExternalProject_Add(
extern_leveldb
${EXTERNAL_PROJECT_LOG_ARGS}
PREFIX ${LEVELDB_SOURCES_DIR}
URL "https://github.com/google/leveldb/archive/v1.18.tar.gz"
URL_MD5 "73770de34a2a5ab34498d2e05b2b7fa0"
GIT_REPOSITORY "https://github.com/google/leveldb"
GIT_TAG v1.18
CONFIGURE_COMMAND ""
BUILD_COMMAND CXXFLAGS=-fPIC make -j ${NUM_OF_PROCESSOR} libleveldb.a
INSTALL_COMMAND mkdir -p ${LEVELDB_INSTALL_DIR}/lib/
......@@ -39,6 +39,3 @@ ADD_DEPENDENCIES(extern_leveldb snappy)
ADD_LIBRARY(leveldb STATIC IMPORTED GLOBAL)
SET_PROPERTY(TARGET leveldb PROPERTY IMPORTED_LOCATION ${LEVELDB_LIBRARIES})
ADD_DEPENDENCIES(leveldb extern_leveldb)
LIST(APPEND external_project_dependencies leveldb)
cmake/external/libmct.cmake 0 → 100644
浏览文件 @ 955fad7a
# Copyright (c) 2017 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_LIBMCT})
return()
ENDIF(NOT ${WITH_LIBMCT})
IF(WIN32 OR APPLE)
MESSAGE(WARNING
"Windows or Mac is not supported with LIBMCT in Paddle yet."
"Force WITH_LIBMCT=OFF")
SET(WITH_LIBMCT OFF CACHE STRING "Disable LIBMCT package in Windows and MacOS" FORCE)
return()
ENDIF()
INCLUDE(ExternalProject)
SET(LIBMCT_PROJECT "extern_libmct")
IF((NOT DEFINED LIBMCT_VER) OR (NOT DEFINED LIBMCT_URL))
MESSAGE(STATUS "use pre defined download url")
SET(LIBMCT_VER "0.1.0" CACHE STRING "" FORCE)
SET(LIBMCT_NAME "libmct" CACHE STRING "" FORCE)
SET(LIBMCT_URL "https://raw.githubusercontent.com/PaddlePaddle/Fleet/release/${LIBMCT_VER}/${LIBMCT_NAME}.tar.gz" CACHE STRING "" FORCE)
ENDIF()
MESSAGE(STATUS "LIBMCT_NAME: ${LIBMCT_NAME}, LIBMCT_URL: ${LIBMCT_URL}")
SET(LIBMCT_SOURCE_DIR "${THIRD_PARTY_PATH}/libmct")
SET(LIBMCT_DOWNLOAD_DIR "${LIBMCT_SOURCE_DIR}/src/${LIBMCT_PROJECT}")
SET(LIBMCT_DST_DIR "libmct")
SET(LIBMCT_INSTALL_ROOT "${THIRD_PARTY_PATH}/install")
SET(LIBMCT_INSTALL_DIR ${LIBMCT_INSTALL_ROOT}/${LIBMCT_DST_DIR})
SET(LIBMCT_ROOT ${LIBMCT_INSTALL_DIR})
SET(LIBMCT_INC_DIR ${LIBMCT_ROOT}/include)
SET(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}" "${LIBMCT_ROOT}/lib")
INCLUDE_DIRECTORIES(${LIBMCT_INC_DIR})
FILE(WRITE ${LIBMCT_DOWNLOAD_DIR}/CMakeLists.txt
"PROJECT(LIBMCT)\n"
"cmake_minimum_required(VERSION 3.0)\n"
"install(DIRECTORY ${LIBMCT_NAME}/include ${LIBMCT_NAME}/lib \n"
" DESTINATION ${LIBMCT_DST_DIR})\n")
ExternalProject_Add(
${LIBMCT_PROJECT}
${EXTERNAL_PROJECT_LOG_ARGS}
PREFIX ${LIBMCT_SOURCE_DIR}
DOWNLOAD_DIR ${LIBMCT_DOWNLOAD_DIR}
DOWNLOAD_COMMAND wget --no-check-certificate ${LIBMCT_URL} -c -q -O ${LIBMCT_NAME}.tar.gz
&& tar zxvf ${LIBMCT_NAME}.tar.gz
DOWNLOAD_NO_PROGRESS 1
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=${LIBMCT_INSTALL_ROOT}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${LIBMCT_INSTALL_ROOT}
)
if (${CMAKE_VERSION} VERSION_LESS "3.3.0" OR NOT WIN32)
set(dummyfile ${CMAKE_CURRENT_BINARY_DIR}/boost_dummy.c)
file(WRITE ${dummyfile} "const char *dummy = \"${dummyfile}\";")
add_library(libmct STATIC ${dummyfile})
else()
add_library(libmct INTERFACE)
endif()
ADD_DEPENDENCIES(libmct ${LIBMCT_PROJECT})
cmake/external/libxsmm.cmake
浏览文件 @ 955fad7a
......@@ -19,8 +19,8 @@ IF(NOT WITH_LIBXSMM)
return()
ENDIF()
IF(WIN32 OR APPLE OR ANDROID OR IOS)
MESSAGE(WARNING "Windows, Mac or Mobile are not supported with libxsmm in Paddle yet.")
IF(WIN32 OR APPLE)
MESSAGE(WARNING "Windows, Mac are not supported with libxsmm in Paddle yet.")
SET(WITH_LIBXSMM OFF CACHE STRING "Disable LIBXSMM" FORCE)
return()
ENDIF()
......@@ -53,5 +53,3 @@ MESSAGE(STATUS "Libxsmm library: ${LIBXSMM_LIBS}")
include_directories(${LIBXSMM_INCLUDE_DIR})
ADD_DEFINITIONS(-DPADDLE_WITH_LIBXSMM)
ADD_DEPENDENCIES(libxsmm extern_libxsmm)
LIST(APPEND external_project_dependencies libxsmm)
cmake/external/mkldnn.cmake
浏览文件 @ 955fad7a
......@@ -23,21 +23,27 @@ SET(MKLDNN_SOURCES_DIR ${THIRD_PARTY_PATH}/mkldnn)
SET(MKLDNN_INSTALL_DIR ${THIRD_PARTY_PATH}/install/mkldnn)
SET(MKLDNN_INC_DIR "${MKLDNN_INSTALL_DIR}/include" CACHE PATH "mkldnn include directory." FORCE)
IF(WIN32 OR APPLE)
IF(APPLE)
MESSAGE(WARNING
"Windows or Mac is not supported with MKLDNN in Paddle yet."
"Mac is not supported with MKLDNN in Paddle yet."
"Force WITH_MKLDNN=OFF")
SET(WITH_MKLDNN OFF CACHE STRING "Disable MKLDNN in Windows and MacOS" FORCE)
SET(WITH_MKLDNN OFF CACHE STRING "Disable MKLDNN in MacOS" FORCE)
return()
ENDIF()
SET(MKLDNN_LIB "${MKLDNN_INSTALL_DIR}/lib/libmkldnn.so" CACHE FILEPATH "mkldnn library." FORCE)
MESSAGE(STATUS "Set ${MKLDNN_INSTALL_DIR}/lib to runtime path")
# Introduce variables:
# * CMAKE_INSTALL_LIBDIR
INCLUDE(GNUInstallDirs)
SET(LIBDIR "lib")
if(CMAKE_INSTALL_LIBDIR MATCHES ".*lib64$")
SET(LIBDIR "lib64")
endif()
MESSAGE(STATUS "Set ${MKLDNN_INSTALL_DIR}/l${LIBDIR} to runtime path")
SET(CMAKE_INSTALL_RPATH_USE_LINK_PATH TRUE)
SET(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}" "${MKLDNN_INSTALL_DIR}/lib")
SET(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}" "${MKLDNN_INSTALL_DIR}/${LIBDIR}")
INCLUDE_DIRECTORIES(${MKLDNN_INC_DIR}) # For MKLDNN code to include internal headers.
INCLUDE_DIRECTORIES(${THIRD_PARTY_PATH}/install) # For Paddle code to include mkldnn.h
IF(${CBLAS_PROVIDER} STREQUAL "MKLML")
SET(MKLDNN_DEPENDS ${MKLML_PROJECT})
......@@ -45,22 +51,34 @@ IF(${CBLAS_PROVIDER} STREQUAL "MKLML")
ELSE()
MESSAGE(FATAL_ERROR "Should enable MKLML when build MKLDNN")
ENDIF()
SET(MKLDNN_FLAG "-Wno-error=strict-overflow -Wno-error=unused-result")
SET(MKLDNN_FLAG "${MKLDNN_FLAG} -Wno-unused-result -Wno-unused-value")
SET(MKLDNN_CFLAG "${CMAKE_C_FLAGS} ${MKLDNN_FLAG}")
SET(MKLDNN_CXXFLAG "${CMAKE_CXX_FLAGS} ${MKLDNN_FLAG}")
IF(NOT WIN32)
SET(MKLDNN_FLAG "-Wno-error=strict-overflow -Wno-error=unused-result -Wno-error=array-bounds")
SET(MKLDNN_FLAG "${MKLDNN_FLAG} -Wno-unused-result -Wno-unused-value")
SET(MKLDNN_CFLAG "${CMAKE_C_FLAGS} ${MKLDNN_FLAG}")
SET(MKLDNN_CXXFLAG "${CMAKE_CXX_FLAGS} ${MKLDNN_FLAG}")
ELSE()
SET(MKLDNN_CXXFLAG "${CMAKE_CXX_FLAGS} /EHsc")
ENDIF(NOT WIN32)
ExternalProject_Add(
${MKLDNN_PROJECT}
${EXTERNAL_PROJECT_LOG_ARGS}
DEPENDS ${MKLDNN_DEPENDS}
GIT_REPOSITORY "https://github.com/01org/mkl-dnn.git"
GIT_TAG "64e03a1939e0d526aa8e9f2e3f7dc0ad8d372944"
GIT_REPOSITORY "https://github.com/intel/mkl-dnn.git"
GIT_TAG "863ff6e7042cec7d2e29897fe9f0872e0888b0fc"
PREFIX ${MKLDNN_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
CMAKE_ARGS -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
CMAKE_ARGS -DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
CMAKE_ARGS -DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
CMAKE_ARGS -DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
CMAKE_ARGS -DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}
CMAKE_ARGS -DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}
CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=${MKLDNN_INSTALL_DIR}
CMAKE_ARGS -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE}
CMAKE_ARGS -DCMAKE_POSITION_INDEPENDENT_CODE=ON
CMAKE_ARGS -DMKLROOT=${MKLML_ROOT}
CMAKE_ARGS -DCMAKE_C_FLAGS=${MKLDNN_CFLAG}
CMAKE_ARGS -DCMAKE_CXX_FLAGS=${MKLDNN_CXXFLAG}
......@@ -68,13 +86,17 @@ ExternalProject_Add(
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${MKLDNN_INSTALL_DIR}
-DMKLROOT:PATH=${MKLML_ROOT}
)
if(WIN32)
SET(MKLDNN_LIB "${MKLDNN_INSTALL_DIR}/${LIBDIR}/mkldnn.lib" CACHE FILEPATH "mkldnn library." FORCE)
else(WIN32)
SET(MKLDNN_LIB "${MKLDNN_INSTALL_DIR}/${LIBDIR}/libmkldnn.so" CACHE FILEPATH "mkldnn library." FORCE)
endif(WIN32)
ADD_LIBRARY(shared_mkldnn SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET shared_mkldnn PROPERTY IMPORTED_LOCATION ${MKLDNN_LIB})
ADD_DEPENDENCIES(shared_mkldnn ${MKLDNN_PROJECT})
MESSAGE(STATUS "MKLDNN library: ${MKLDNN_LIB}")
add_definitions(-DPADDLE_WITH_MKLDNN)
LIST(APPEND external_project_dependencies shared_mkldnn)
# generate a static dummy target to track mkldnn dependencies
# for cc_library(xxx SRCS xxx.c DEPS mkldnn)
......@@ -86,13 +108,13 @@ ADD_DEPENDENCIES(mkldnn ${MKLDNN_PROJECT})
# copy the real so.0 lib to install dir
# it can be directly contained in wheel or capi
SET(MKLDNN_SHARED_LIB ${MKLDNN_INSTALL_DIR}/libmkldnn.so.0)
ADD_CUSTOM_COMMAND(OUTPUT ${MKLDNN_SHARED_LIB}
COMMAND cp ${MKLDNN_LIB} ${MKLDNN_SHARED_LIB}
DEPENDS mkldnn)
if(WIN32)
SET(MKLDNN_SHARED_LIB ${MKLDNN_INSTALL_DIR}/bin/mkldnn.dll)
else(WIN32)
SET(MKLDNN_SHARED_LIB ${MKLDNN_INSTALL_DIR}/libmkldnn.so.0)
ADD_CUSTOM_COMMAND(OUTPUT ${MKLDNN_SHARED_LIB}
COMMAND ${CMAKE_COMMAND} -E copy ${MKLDNN_LIB} ${MKLDNN_SHARED_LIB}
DEPENDS mkldnn shared_mkldnn)
endif(WIN32)
ADD_CUSTOM_TARGET(mkldnn_shared_lib ALL DEPENDS ${MKLDNN_SHARED_LIB})
IF(WITH_C_API)
INSTALL(FILES ${MKLDNN_SHARED_LIB} DESTINATION lib)
ENDIF()
ADD_DEPENDENCIES(mkldnn_shared_lib ${MKLDNN_PROJECT} mkldnn)
cmake/external/mklml.cmake
浏览文件 @ 955fad7a
......@@ -16,61 +16,62 @@ IF(NOT ${WITH_MKLML})
return()
ENDIF(NOT ${WITH_MKLML})
IF(WIN32 OR APPLE)
MESSAGE(WARNING
"Windows or Mac is not supported with MKLML in Paddle yet."
"Force WITH_MKLML=OFF")
SET(WITH_MKLML OFF CACHE STRING "Disable MKLML package in Windows and MacOS" FORCE)
IF(APPLE)
MESSAGE(WARNING "Mac is not supported with MKLML in Paddle yet. Force WITH_MKLML=OFF.")
SET(WITH_MKLML OFF CACHE STRING "Disable MKLML package in MacOS" FORCE)
return()
ENDIF()
INCLUDE(ExternalProject)
SET(MKLML_PROJECT "extern_mklml")
IF((NOT DEFINED MKLML_VER) OR (NOT DEFINED MKLML_URL))
MESSAGE(STATUS "use pre defined download url")
SET(MKLML_VER "mklml_lnx_2019.0.20180710" CACHE STRING "" FORCE)
SET(MKLML_URL "http://paddlepaddledeps.cdn.bcebos.com/${MKLML_VER}.tgz" CACHE STRING "" FORCE)
ENDIF()
MESSAGE(STATUS "MKLML_VER: ${MKLML_VER}, MKLML_URL: ${MKLML_URL}")
SET(MKLML_SOURCE_DIR "${THIRD_PARTY_PATH}/mklml")
SET(MKLML_DOWNLOAD_DIR "${MKLML_SOURCE_DIR}/src/${MKLML_PROJECT}")
SET(MKLML_DST_DIR "mklml")
SET(MKLML_INSTALL_ROOT "${THIRD_PARTY_PATH}/install")
SET(MKLML_INSTALL_DIR ${MKLML_INSTALL_ROOT}/${MKLML_DST_DIR})
SET(MKLML_ROOT ${MKLML_INSTALL_DIR})
SET(MKLML_INC_DIR ${MKLML_ROOT}/include)
SET(MKLML_LIB_DIR ${MKLML_ROOT}/lib)
SET(MKLML_LIB ${MKLML_LIB_DIR}/libmklml_intel.so)
SET(MKLML_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5.so)
SET(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}" "${MKLML_ROOT}/lib")
INCLUDE_DIRECTORIES(${MKLML_INC_DIR})
SET(TIME_VERSION "2019.0.1.20181227")
IF(WIN32)
SET(MKLML_VER "mklml_win_${TIME_VERSION}" CACHE STRING "" FORCE)
SET(MKLML_URL "https://paddlepaddledeps.cdn.bcebos.com/${MKLML_VER}.zip" CACHE STRING "" FORCE)
SET(MKLML_LIB ${MKLML_LIB_DIR}/mklml.lib)
SET(MKLML_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5md.lib)
SET(MKLML_SHARED_LIB ${MKLML_LIB_DIR}/mklml.dll)
SET(MKLML_SHARED_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5md.dll)
ELSE()
#TODO(intel-huying):
# Now enable Erf function in mklml library temporarily, it will be updated as offical version later.
SET(MKLML_VER "Glibc225_vsErf_mklml_lnx_${TIME_VERSION}" CACHE STRING "" FORCE)
SET(MKLML_URL "http://paddlepaddledeps.cdn.bcebos.com/${MKLML_VER}.tgz" CACHE STRING "" FORCE)
SET(MKLML_LIB ${MKLML_LIB_DIR}/libmklml_intel.so)
SET(MKLML_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5.so)
SET(MKLML_SHARED_LIB ${MKLML_LIB_DIR}/libmklml_intel.so)
SET(MKLML_SHARED_IOMP_LIB ${MKLML_LIB_DIR}/libiomp5.so)
ENDIF()
FILE(WRITE ${MKLML_DOWNLOAD_DIR}/CMakeLists.txt
"PROJECT(MKLML)\n"
"cmake_minimum_required(VERSION 3.0)\n"
"install(DIRECTORY ${MKLML_VER}/include ${MKLML_VER}/lib \n"
" DESTINATION ${MKLML_DST_DIR})\n")
SET(MKLML_PROJECT "extern_mklml")
MESSAGE(STATUS "MKLML_VER: ${MKLML_VER}, MKLML_URL: ${MKLML_URL}")
SET(MKLML_SOURCE_DIR "${THIRD_PARTY_PATH}/mklml")
SET(MKLML_DOWNLOAD_DIR "${MKLML_SOURCE_DIR}/src/${MKLML_PROJECT}")
ExternalProject_Add(
${MKLML_PROJECT}
${EXTERNAL_PROJECT_LOG_ARGS}
PREFIX ${MKLML_SOURCE_DIR}
PREFIX ${MKLML_SOURCE_DIR}
URL ${MKLML_URL}
DOWNLOAD_DIR ${MKLML_DOWNLOAD_DIR}
DOWNLOAD_COMMAND wget --no-check-certificate ${MKLML_URL} -c -q -O ${MKLML_VER}.tgz
&& tar zxf ${MKLML_VER}.tgz
DOWNLOAD_NO_PROGRESS 1
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=${MKLML_INSTALL_ROOT}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${MKLML_INSTALL_ROOT}
CONFIGURE_COMMAND ""
BUILD_COMMAND ""
UPDATE_COMMAND ""
INSTALL_COMMAND
${CMAKE_COMMAND} -E copy_directory ${MKLML_DOWNLOAD_DIR}/include ${MKLML_INC_DIR} &&
${CMAKE_COMMAND} -E copy_directory ${MKLML_DOWNLOAD_DIR}/lib ${MKLML_LIB_DIR}
)
INCLUDE_DIRECTORIES(${MKLML_INC_DIR})
ADD_LIBRARY(mklml SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET mklml PROPERTY IMPORTED_LOCATION ${MKLML_LIB})
ADD_DEPENDENCIES(mklml ${MKLML_PROJECT})
LIST(APPEND external_project_dependencies mklml)
IF(WITH_C_API)
INSTALL(FILES ${MKLML_LIB} ${MKLML_IOMP_LIB} DESTINATION lib)
ENDIF()
cmake/external/ngraph.cmake 0 → 100644
浏览文件 @ 955fad7a
# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
add_library(ngraph INTERFACE)
IF(WIN32 OR APPLE)
MESSAGE(WARNING
"Windows or Mac is not supported with nGraph in Paddle yet."
"Force WITH_NGRAPH=OFF")
SET(WITH_NGRAPH OFF CACHE STRING "Disable nGraph in Windows and MacOS" FORCE)
ENDIF()
IF(${WITH_NGRAPH} AND NOT ${WITH_MKLDNN})
MESSAGE(WARNING
"nGraph needs mkl-dnn to be enabled."
"Force WITH_NGRAPH=OFF")
SET(WITH_NGRAPH OFF CACHE STRING "Disable nGraph if mkl-dnn is disabled" FORCE)
ENDIF()
IF(NOT ${WITH_NGRAPH})
return()
ENDIF()
INCLUDE(GNUInstallDirs)
INCLUDE(ExternalProject)
SET(NGRAPH_PROJECT "extern_ngraph")
SET(NGRAPH_GIT_TAG "a444f7a959b7d87f2c117c9b57a4c387759e481e")
SET(NGRAPH_SOURCES_DIR ${THIRD_PARTY_PATH}/ngraph)
SET(NGRAPH_INSTALL_DIR ${THIRD_PARTY_PATH}/install/ngraph)
SET(NGRAPH_INC_DIR ${NGRAPH_INSTALL_DIR}/include)
SET(NGRAPH_LIB_DIR ${NGRAPH_INSTALL_DIR}/${CMAKE_INSTALL_LIBDIR})
SET(NGRAPH_SHARED_LIB_NAME libngraph.so)
SET(NGRAPH_CPU_LIB_NAME libcpu_backend.so)
if(CMAKE_BUILD_TYPE STREQUAL "Debug")
SET(NGRAPH_TBB_LIB_NAME libtbb_debug.so.2)
else()
SET(NGRAPH_TBB_LIB_NAME libtbb.so.2)
endif()
SET(NGRAPH_GIT_REPO "https://github.com/NervanaSystems/ngraph.git")
SET(NGRAPH_SHARED_LIB ${NGRAPH_LIB_DIR}/${NGRAPH_SHARED_LIB_NAME})
SET(NGRAPH_CPU_LIB ${NGRAPH_LIB_DIR}/${NGRAPH_CPU_LIB_NAME})
SET(NGRAPH_TBB_LIB ${NGRAPH_LIB_DIR}/${NGRAPH_TBB_LIB_NAME})
ExternalProject_Add(
${NGRAPH_PROJECT}
${EXTERNAL_PROJECT_LOG_ARGS}
DEPENDS ${MKLDNN_PROJECT} ${MKLML_PROJECT}
GIT_REPOSITORY ${NGRAPH_GIT_REPO}
GIT_TAG ${NGRAPH_GIT_TAG}
PREFIX ${NGRAPH_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=${NGRAPH_INSTALL_DIR}
CMAKE_ARGS -DNGRAPH_UNIT_TEST_ENABLE=FALSE
CMAKE_ARGS -DNGRAPH_TOOLS_ENABLE=FALSE
CMAKE_ARGS -DNGRAPH_INTERPRETER_ENABLE=FALSE
CMAKE_ARGS -DNGRAPH_DEX_ONLY=TRUE
CMAKE_ARGS -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE}
CMAKE_ARGS -DMKLDNN_INCLUDE_DIR=${MKLDNN_INC_DIR}
CMAKE_ARGS -DMKLDNN_LIB_DIR=${MKLDNN_INSTALL_DIR}/${CMAKE_INSTALL_LIBDIR}
CMAKE_ARGS -DMKLML_LIB_DIR=${MKLML_INSTALL_DIR}/lib
)
add_dependencies(ngraph ${NGRAPH_PROJECT})
target_compile_definitions(ngraph INTERFACE -DPADDLE_WITH_NGRAPH)
target_include_directories(ngraph INTERFACE ${NGRAPH_INC_DIR})
target_link_libraries(ngraph INTERFACE ${NGRAPH_SHARED_LIB})
cmake/external/nnpack.cmake 已删除 100644 → 0
浏览文件 @ a7fe3b50
# 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 @ 955fad7a
......@@ -11,18 +11,9 @@
# 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(USE_EIGEN_FOR_BLAS)
return()
ENDIF(USE_EIGEN_FOR_BLAS)
INCLUDE(cblas)
# IF(WIN32 AND NOT ${CBLAS_FOUND})
IF(NOT ${CBLAS_FOUND})
INCLUDE(ExternalProject)
SET(CBLAS_SOURCES_DIR ${THIRD_PARTY_PATH}/openblas)
......@@ -34,6 +25,7 @@ IF(NOT ${CBLAS_FOUND})
CACHE FILEPATH "openblas library." FORCE)
ADD_DEFINITIONS(-DPADDLE_USE_OPENBLAS)
IF (WIN32)
SET(CBLAS_FOUND true)
MESSAGE(WARNING, "In windows, openblas only support msvc build, please build it manually and put it at " ${CBLAS_INSTALL_DIR})
......@@ -43,38 +35,12 @@ IF(NOT ${CBLAS_FOUND})
SET(OPENBLAS_CC "${CMAKE_C_COMPILER} -Wno-unused-but-set-variable -Wno-unused-variable")
SET(OPENBLAS_COMMIT "v0.2.20")
IF(CMAKE_CROSSCOMPILING)
SET(OPTIONAL_ARGS HOSTCC=${HOST_C_COMPILER})
GET_FILENAME_COMPONENT(CROSS_SUFFIX ${CMAKE_C_COMPILER} DIRECTORY)
SET(CROSS_SUFFIX ${CROSS_SUFFIX}/)
IF(ANDROID)
IF(ANDROID_ABI MATCHES "^armeabi(-v7a)?$")
# 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()
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()
SET(COMMON_ARGS CC=${OPENBLAS_CC} NO_SHARED=1 NO_LAPACK=1 libs)
......@@ -95,25 +61,6 @@ IF(NOT ${CBLAS_FOUND})
ELSE()
ENDIF(NOT WIN32)
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})
MESSAGE(STATUS "BLAS library: ${CBLAS_LIBRARIES}")
......@@ -139,7 +86,6 @@ ENDIF()
IF(NOT ${CBLAS_FOUND})
ADD_DEPENDENCIES(cblas extern_openblas)
LIST(APPEND external_project_dependencies cblas)
ELSE()
IF("${CBLAS_PROVIDER}" STREQUAL "MKLML")
ADD_DEPENDENCIES(cblas mklml)
......
cmake/external/protobuf.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/pslib.cmake 0 → 100644
浏览文件 @ 955fad7a
# Copyright (c) 2017 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_PSLIB})
return()
ENDIF(NOT ${WITH_PSLIB})
IF(WIN32 OR APPLE)
MESSAGE(WARNING
"Windows or Mac is not supported with PSLIB in Paddle yet."
"Force WITH_PSLIB=OFF")
SET(WITH_PSLIB OFF CACHE STRING "Disable PSLIB package in Windows and MacOS" FORCE)
return()
ENDIF()
INCLUDE(ExternalProject)
SET(PSLIB_PROJECT "extern_pslib")
IF((NOT DEFINED PSLIB_VER) OR (NOT DEFINED PSLIB_URL))
MESSAGE(STATUS "use pre defined download url")
SET(PSLIB_VER "0.1.0" CACHE STRING "" FORCE)
SET(PSLIB_NAME "pslib" CACHE STRING "" FORCE)
SET(PSLIB_URL "https://raw.githubusercontent.com/PaddlePaddle/Fleet/release/${PSLIB_VER}/${PSLIB_NAME}.tar.gz" CACHE STRING "" FORCE)
ENDIF()
MESSAGE(STATUS "PSLIB_NAME: ${PSLIB_NAME}, PSLIB_URL: ${PSLIB_URL}")
SET(PSLIB_SOURCE_DIR "${THIRD_PARTY_PATH}/pslib")
SET(PSLIB_DOWNLOAD_DIR "${PSLIB_SOURCE_DIR}/src/${PSLIB_PROJECT}")
SET(PSLIB_DST_DIR "pslib")
SET(PSLIB_INSTALL_ROOT "${THIRD_PARTY_PATH}/install")
SET(PSLIB_INSTALL_DIR ${PSLIB_INSTALL_ROOT}/${PSLIB_DST_DIR})
SET(PSLIB_ROOT ${PSLIB_INSTALL_DIR})
SET(PSLIB_INC_DIR ${PSLIB_ROOT}/include)
SET(PSLIB_LIB_DIR ${PSLIB_ROOT}/lib)
SET(PSLIB_LIB ${PSLIB_LIB_DIR}/libps.so)
SET(PSLIB_IOMP_LIB ${PSLIB_LIB_DIR}/libiomp5.so) #todo what is this
SET(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}" "${PSLIB_ROOT}/lib")
INCLUDE_DIRECTORIES(${PSLIB_INC_DIR})
FILE(WRITE ${PSLIB_DOWNLOAD_DIR}/CMakeLists.txt
"PROJECT(PSLIB)\n"
"cmake_minimum_required(VERSION 3.0)\n"
"install(DIRECTORY ${PSLIB_NAME}/include ${PSLIB_NAME}/lib \n"
" DESTINATION ${PSLIB_DST_DIR})\n")
ExternalProject_Add(
${PSLIB_PROJECT}
${EXTERNAL_PROJECT_LOG_ARGS}
PREFIX ${PSLIB_SOURCE_DIR}
DOWNLOAD_DIR ${PSLIB_DOWNLOAD_DIR}
DOWNLOAD_COMMAND wget --no-check-certificate ${PSLIB_URL} -c -q -O ${PSLIB_NAME}.tar.gz
&& tar zxvf ${PSLIB_NAME}.tar.gz
DOWNLOAD_NO_PROGRESS 1
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=${PSLIB_INSTALL_ROOT}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${PSLIB_INSTALL_ROOT}
)
ADD_LIBRARY(pslib SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET pslib PROPERTY IMPORTED_LOCATION ${PSLIB_LIB})
ADD_DEPENDENCIES(pslib ${PSLIB_PROJECT})
cmake/external/pslib_brpc.cmake 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/pybind11.cmake
浏览文件 @ 955fad7a
......@@ -26,7 +26,7 @@ ExternalProject_Add(
extern_pybind
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/pybind/pybind11.git"
GIT_TAG "v2.1.1"
GIT_TAG "v2.2.4"
PREFIX ${PYBIND_SOURCE_DIR}
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
......
cmake/external/python.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/rocprim.cmake 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/snappy.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/snappystream.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/swig.cmake 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
cmake/external/threadpool.cmake
浏览文件 @ 955fad7a
......@@ -26,5 +26,3 @@ else()
endif()
add_dependencies(simple_threadpool extern_threadpool)
LIST(APPEND external_project_dependencies simple_threadpool)
cmake/external/warpctc.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/xbyak.cmake
浏览文件 @ 955fad7a
......@@ -55,4 +55,3 @@ else()
endif()
add_dependencies(xbyak ${XBYAK_PROJECT})
list(APPEND external_project_dependencies xbyak)
cmake/external/xxhash.cmake 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
cmake/external/zlib.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/flags.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/generic.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/hip.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/inference_lib.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/operators.cmake 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
cmake/rdma.cmake 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
cmake/simd.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/system.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/tensorrt.cmake
浏览文件 @ 955fad7a
......@@ -33,5 +33,5 @@ if(TENSORRT_FOUND)
message(STATUS "Current TensorRT header is ${TENSORRT_INCLUDE_DIR}/NvInfer.h. "
"Current TensorRT version is v${TENSORRT_MAJOR_VERSION}. ")
include_directories(${TENSORRT_INCLUDE_DIR})
list(APPEND EXTERNAL_LIBS ${TENSORRT_LIBRARY})
add_definitions(-DPADDLE_WITH_TENSORRT)
endif()
cmake/util.cmake
浏览文件 @ 955fad7a
此差异已折叠。
cmake/version.cmake
浏览文件 @ 955fad7a
此差异已折叠。
doc/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/about/about_us.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/mobile/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/mobile/cross_compiling_for_android_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/mobile/cross_compiling_for_android_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/mobile/cross_compiling_for_ios_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/mobile/cross_compiling_for_ios_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/mobile/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/mobile/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/survey/cluster_bootstrapping_tools.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/survey/dynamic_graph.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/survey/op_fusion_design.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/templates/conf.py.cn.in 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/templates/conf.py.en.in 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/templates/layout.html 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/config/activation.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/config/attr.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/config/evaluators.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/config/layer.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/config/networks.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/config/optimizer.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/config/pooling.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/data.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/data/data_reader.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/data/dataset.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/data/image.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/model_configs.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/overview.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/api/run_logic.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/build_and_install/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/build_and_install/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/build_and_install/paddleci.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/build_and_install/pip_install_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/build_and_install/pip_install_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/README.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/checkpointing.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/data_dispatch.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/master_server.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/pserver_client.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/save_model.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/src/dataset.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/src/dataset.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/src/init_lock.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/src/paddle-etcd.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
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 @ a7fe3b50
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 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/src/submit-job.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/src/trainer.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/src/trainer.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/cluster_train/submit-job.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/interface/00.why_plain_c.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/interface/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/interface/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/mkl/image/engine.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/mkl/image/gradients.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/mkl/image/layers.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/mkl/image/matrix.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/mkl/image/overview.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/mkl/mkl_packed.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/design/mkl/mkldnn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/contribute_to_paddle_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/contribute_to_paddle_en.md 已删除 120000 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/new_layer_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/new_layer_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/src/FullyConnected.jpg 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/src/doc_en.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/write_docs_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/dev/write_docs_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/build_and_install/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/build_and_install/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/cluster/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/cluster/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/local/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/local/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
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 @ a7fe3b50
此差异已折叠。
doc/v2/faq/local/src/word2vec_config.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/local/src/word2vec_dataprovider.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/model/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/model/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/parameter/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/faq/parameter/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/getstarted/concepts/src/infer.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/getstarted/concepts/src/train.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/getstarted/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/getstarted/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/getstarted/quickstart_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/getstarted/quickstart_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/compile_paddle_lib_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/compile_paddle_lib_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/images/csr.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/images/sequence_data.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/images/workflow_of_CAPI.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/workflow_of_capi_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/capi/workflow_of_capi_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/cmd_argument_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/cmd_argument_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/multi_cluster/k8s_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/multi_cluster/k8s_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/preparations_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/preparations_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/Dockerfile 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/efs_mount.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/managed_policy.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/ps_cn.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/ps_en.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/trainer.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/trainer_cn.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cluster/src/word2vec/prepare.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cmd_parameter/arguments_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cmd_parameter/arguments_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cmd_parameter/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cmd_parameter/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cmd_parameter/use_case_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/cmd_parameter/use_case_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/optimization/nvvp1.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/optimization/nvvp2.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/optimization/nvvp3.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/optimization/nvvp4.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/hierarchical_layer_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/hierarchical_layer_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
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 @ a7fe3b50
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 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/recurrent_group_cn.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/recurrent_group_en.md 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/rnn_config_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/rnn_config_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/src/bi_lstm.jpg 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/howto/rnn/src/glossary_rnn.dot 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/FullyConnected.jpg 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/add_security_group.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/bi_lstm.jpg 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/checkpointing.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/create_efs.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/csr.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/data_dispatch.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/dataset.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/dataset.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/doc_en.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/efs_mount.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/engine.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/file_storage.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/file_storage.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/glossary_rnn.dot 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/glossary_rnn_with_memory.dot 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/gradients.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/init_lock.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/init_lock.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/k8s-paddle-arch.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/layers.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/managed_policy.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/matrix.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/nvvp1.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/nvvp2.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/nvvp3.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/nvvp4.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/overview.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-cloud-in-data-center.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-etcd.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-etcd.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-model-sharding.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-model-sharding.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-ps-0.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-ps-1.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-ps.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-task-queues.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-task-queues.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-task-states.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/paddle-task-states.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/ps_cn.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/ps_en.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/pserver_and_trainer.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/pserver_init.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/pserver_init.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/route53_create_recordset.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/route53_create_zone.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/sequence_data.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/simple_full_recurrent.dot 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/submit-job.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/submit-job.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/trainer.graffle 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/trainer.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/trainer_cn.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/worker_security_group.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/images/workflow_of_CAPI.png 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/index_cn.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
doc/v2/index_en.rst 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/.gitignore 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/cmd/master/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/cmd/master/master.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/cmd/pserver/.gitignore 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/cmd/pserver/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/cmd/pserver/pserver.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/connection/conn.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/c/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/c/client.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/client.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/client_internal_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/client_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/etcd_client.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/inmem_store.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/service.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/service_internal_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/master/service_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/proto/.gitignore 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/.gitignore 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/cclient.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/test/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/test/test_cclient.c 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/test/test_mnist.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/test/test_train.py 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/c/test/testdata/optimizer.pb 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/client.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/client_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/etcd_client.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/client/etcd_client_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/etcd_client.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/optimizer.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/optimizer_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/service.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/service_internal_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/pserver/service_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/utils/networkhelper/CMakeLists.txt 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/utils/networkhelper/helper.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
go/utils/networkhelper/helper_test.go 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/CMakeLists.txt
浏览文件 @ 955fad7a
此差异已折叠。
paddle/contrib/float16/README.md
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/API.spec
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/CMakeLists.txt
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/array.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/async_executor.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/async_executor.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/attribute.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/attribute.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/block_desc.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/commit.h.in 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_feed.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_feed.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_feed.proto 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_feed_factory.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_feed_factory.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_feed_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_type.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/data_type.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ddim.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ddim.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/all_reduce_deps_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/all_reduce_deps_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/eager_deletion_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/early_delete_op_handle.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/graph_test_base.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/inplace_op_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/inplace_op_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/memory_optimize_helper.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/memory_optimize_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/memory_optimize_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/op_graph_view.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/details/op_graph_view.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/dim.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/dlpack_tensor.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/dlpack_tensor.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/dlpack_tensor_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/eigen.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/executor.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/executor.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/executor_thread_worker.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/executor_thread_worker.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/garbage_collector.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/inplace_op_inference.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/inplace_op_inference_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/cpu_quantize_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/cpu_quantize_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/cpu_quantize_pass_tester.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/cpu_quantize_squash_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/cpu_quantize_squash_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/fuse_pass_base.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/graph.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/graph.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/is_test_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/is_test_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/is_test_pass_tester.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/lock_free_optimize_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/lock_free_optimize_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/multi_batch_merge_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/multi_batch_merge_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/node.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/node.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/node_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/pass.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/pass.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/repeated_fc_relu_fuse_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/repeated_fc_relu_fuse_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/seqpool_concat_fuse_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/seqpool_concat_fuse_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/squared_mat_sub_fuse_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/squared_mat_sub_fuse_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/sync_batch_norm_pass.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/ir/sync_batch_norm_pass.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/lod_tensor.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/op_desc.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/op_desc.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/op_info.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/op_kernel_type.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/operator.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/operator.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/operator_kernel_configs.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/python_headers.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/rw_lock.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/scope.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/scope.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/scope_pool.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/scope_pool.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/tensor.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/tensor.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/threadpool.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/transfer_scope_cache.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/transfer_scope_cache.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/type_defs.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/unroll_array_ops.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/unroll_array_ops_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/var_type.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/var_type_traits.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/var_type_traits.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/var_type_traits_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/variable.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/variable_helper.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/framework/variable_helper.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/README.md 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/engine.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/engine.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/layer.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/layer.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/profiler.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/profiler.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/tracer.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/tracer.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/imperative/type_defs.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/fc.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/fc.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/op_converter.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/registrar.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/registrar.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/test_fc_op.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/convert/ut_helper.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/engine.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/engine.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/anakin/test_anakin_engine.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/analysis/node.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/inference/analysis/node.h 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/inference/analysis/passes/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/analysis/passes/passes.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/analysis/passes/passes.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/analysis_config.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/api.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/details/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/details/reset_tensor_array.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/helper.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/paddle_anakin_config.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/paddle_analysis_config.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/paddle_api.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/paddle_pass_builder.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/api/paddle_pass_builder.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/engine.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/io.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/io.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/convert/prelu_op.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/convert/split_op.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/op_teller.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/op_teller.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/plugin/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/plugin/trt_plugin.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/plugin/trt_plugin.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/trt_int8_calibrator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tensorrt/trt_int8_calibrator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tests/api/analyzer_dam_tester.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tests/api/config_printer.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/tests/test.cmake 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/utils/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/utils/benchmark.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/utils/benchmark.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/utils/benchmark_tester.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/utils/visualizer.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/inference/utils/visualizer.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/CMakeLists.txt
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/aligned_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/aligned_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/allocator_facade.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/allocator_facade.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/allocator_facade_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/allocator_strategy.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/allocator_strategy.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/best_fit_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/best_fit_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/buffered_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/buffered_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/conditional_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/conditional_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/cpu_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/cpu_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/cuda_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/cuda_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/legacy_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/legacy_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/locked_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/locked_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/pinned_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/pinned_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/retry_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/retry_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/retry_allocator_test.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/zero_size_allocator.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/allocation/zero_size_allocator.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/malloc.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/malloc.h
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/memory/malloc_test.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/memory/memcpy.cc
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/accuracy_op.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/accuracy_op.cu 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/activation_cudnn.cu.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/activation_cudnn_op.cu.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/adadelta_op.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adadelta_op.cu 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adadelta_op.h 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adagrad_op.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adagrad_op.cu 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adagrad_op.h 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adam_op.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adam_op.cu 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adam_op.h 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adamax_op.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adamax_op.cu 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/adamax_op.h 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/add_position_encoding_op.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/add_position_encoding_op.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/affine_grid_cudnn_op.cu.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/affine_grid_op.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/affine_grid_op.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/alloc_continuous_space_op.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/auc_op.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/auc_op.h 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/batch_norm_op.cu 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/batch_norm_op.cu.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/beam_search_op.cu.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/beam_search_op_test.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/benchmark/CMakeLists.txt 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester_config.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/benchmark/op_tester_config.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/bilinear_interp_op.cc 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/bilinear_interp_op.cu 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/bilinear_interp_op.h 已删除 100644 → 0
浏览文件 @ a7fe3b50
此差异已折叠。
paddle/fluid/operators/bpr_loss_op.cc 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/bpr_loss_op.h 0 → 100644
浏览文件 @ 955fad7a
此差异已折叠。
paddle/fluid/operators/cast_op.cc
浏览文件 @ 955fad7a
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