Skip to content

P
Paddle

PaddlePaddle / Paddle
接近 2 年 前同步成功

通知 2323
Star 20933
Fork 5424
P
Paddle
提交 bc34d325 编写于 作者: B barrierye
浏览文件
操作

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

上级 e837a6de e9de2c1a
展开全部 隐藏空白更改
内联 并排
Showing with 43343 addition and 13583 deletion
.github/ISSUE_TEMPLATE/---feature-request-.md 0 → 100644
浏览文件 @ bc34d325
---
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 @ bc34d325
---
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 @ bc34d325
---
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 @ bc34d325
---
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 @ bc34d325
---
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 @ bc34d325
---
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 @ bc34d325
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
AUTHORS.md
浏览文件 @ bc34d325
......@@ -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 +43,9 @@
| QiJune | Jun Qi |
| qingqing01 | Qing-Qing Dang |
| reyoung | Yang Yu |
| Sand3r- | Michal Gallus |
| Superjom | Chun-Wei Yan |
| tensor-tang | Jian Tang |
| tianbingsz | Tian-Bing Xu |
| tpatejko | Tomasz Patejko |
| typhoonzero | Yi Wu |
......
CMakeLists.txt
浏览文件 @ bc34d325
......@@ -26,6 +26,11 @@ message(STATUS "C compiler: ${CMAKE_C_COMPILER}, version: "
"${CMAKE_C_COMPILER_ID} ${CMAKE_C_COMPILER_VERSION}")
if(WIN32)
set(CMAKE_STATIC_LIBRARY_PREFIX lib)
add_definitions("/DGOOGLE_GLOG_DLL_DECL=")
set(CMAKE_C_FLAGS_DEBUG "${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")
endif(WIN32)
if(NOT CMAKE_CROSSCOMPILING)
......@@ -41,6 +46,7 @@ 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)
......@@ -62,13 +68,14 @@ 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_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" ${PY_VERSION})
......@@ -104,6 +111,8 @@ if(ANDROID OR IOS)
"Disable RDMA when cross-compiling for Android and iOS" FORCE)
set(WITH_MKL OFF CACHE STRING
"Disable MKL when cross-compiling for Android and iOS" FORCE)
set(WITH_NGRAPH OFF CACHE STRING
"Disable nGraph when cross-compiling for Android and iOS" FORCE)
set(WITH_GOLANG OFF CACHE STRING
"Disable golang when cross-compiling for Android and iOS" FORCE)
......@@ -121,6 +130,21 @@ if (APPLE OR WIN32)
"Disable MKL for building on mac and windows" FORCE)
endif()
if (WIN32)
set(WITH_AVX OFF CACHE STRING
"Disable AVX when compiling for Windows" FORCE)
set(WITH_DSO OFF CACHE STRING
"Disable DSO when compiling for Windows" FORCE)
set(WITH_MKL OFF CACHE STRING
"Disable MKL when compiling for Windows" FORCE)
set(WITH_DISTRIBUTE OFF CACHE STRING
"Disable DISTRIBUTE when compiling for Windows" FORCE)
set(WITH_C_API OFF CACHE STRING
"Disable C_API when compiling for Windows" FORCE)
set(WITH_FLUID_ONLY ON CACHE STRING
"Enable FLUID_ONLY when compiling for Windows" FORCE)
endif()
set(THIRD_PARTY_PATH "${CMAKE_BINARY_DIR}/third_party" CACHE STRING
"A path setting third party libraries download & build directories.")
......@@ -172,6 +196,7 @@ 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/ngraph) # download, build, install nGraph
include(external/swig) # download, build, install swig
include(external/boost) # download boost
include(external/any) # download libn::any
......@@ -179,13 +204,17 @@ 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
if (NOT WIN32)
# there is no official support of warpctc, nccl, cupti in windows
include(external/warpctc) # download, build, install warpctc
include(cupti)
include(external/gzstream)
endif (NOT WIN32)
if(WITH_DISTRIBUTE)
......@@ -291,6 +320,14 @@ 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)
......
Dockerfile
浏览文件 @ bc34d325
......@@ -22,6 +22,29 @@ 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 \
......@@ -74,30 +97,48 @@ RUN localedef -i en_US -f UTF-8 en_US.UTF-8
RUN pip3 install -U wheel && \
pip3 install -U docopt PyYAML sphinx==1.5.6 && \
pip3 install sphinx-rtd-theme==0.1.9 recommonmark && \
pip3.6 install -U wheel && \
pip3.6 install -U docopt PyYAML sphinx==1.5.6 && \
pip3.6 install sphinx-rtd-theme==0.1.9 recommonmark && \
pip3.7 install -U wheel && \
pip3.7 install -U docopt PyYAML sphinx==1.5.6 && \
pip3.7 install sphinx-rtd-theme==0.1.9 recommonmark && \
easy_install -U pip && \
pip install -U wheel && \
pip install -U pip setuptools 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' && \
RUN pip3 install 'pre-commit==1.10.4' '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' && \
pip3.6 install 'pre-commit==1.10.4' 'ipython==5.3.0' && \
pip3.6 install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip3.6 install opencv-python && \
pip3.7 install 'pre-commit==1.10.4' 'ipython==5.3.0' && \
pip3.7 install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip3.7 install opencv-python && \
pip install 'pre-commit==1.10.4' 'ipython==5.3.0' && \
pip install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
pip install opencv-python
#For docstring checker
RUN pip3 install pylint pytest astroid isort
RUN pip3.6 install pylint pytest astroid isort
RUN pip3.7 install pylint pytest astroid isort
RUN pip install pylint pytest astroid isort LinkChecker
COPY ./python/requirements.txt /root/
RUN pip3 install -r /root/requirements.txt
RUN pip3.6 install -r /root/requirements.txt
RUN pip3.7 install -r /root/requirements.txt
RUN pip 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 pip3.6 install certifi urllib3[secure]
RUN pip3.7 install certifi urllib3[secure]
RUN pip install certifi urllib3[secure]
......
README.md
浏览文件 @ bc34d325
......@@ -2,8 +2,8 @@
[![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/1.0/getstarted/index_en.html)
[![Documentation Status](https://img.shields.io/badge/中文文档-最新-brightgreen.svg)](http://paddlepaddle.org/documentation/docs/zh/1.0/beginners_guide/index.html)
[![Documentation Status](https://img.shields.io/badge/docs-latest-brightgreen.svg?style=flat)](http://paddlepaddle.org/documentation/docs/en/1.1/getstarted/index_en.html)
[![Documentation Status](https://img.shields.io/badge/中文文档-最新-brightgreen.svg)](http://paddlepaddle.org/documentation/docs/zh/1.1/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)
......@@ -19,7 +19,7 @@ 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 1.0.1](https://github.com/PaddlePaddle/Paddle/tree/release/1.0.0)
### Latest PaddlePaddle Release: [Fluid 1.1.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.1)
### Install Latest Stable Release:
```
# Linux CPU
......@@ -27,9 +27,9 @@ pip install paddlepaddle
# Linux GPU cuda9cudnn7
pip install paddlepaddle-gpu
# Linux GPU cuda8cudnn7
pip install paddlepaddle-gpu==1.0.1.post87
pip install paddlepaddle-gpu==1.1.0.post87
# Linux GPU cuda8cudnn5
pip install paddlepaddle-gpu==1.0.1.post85
pip install paddlepaddle-gpu==1.1.0.post85
# For installation on other platform, refer to http://paddlepaddle.org/
```
......@@ -76,26 +76,26 @@ pip install paddlepaddle-gpu==1.0.1.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/zh/1.1/beginners_guide/index.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/1.1/getstarted/index_en.html) and
[Chinese](http://paddlepaddle.org/documentation/docs/zh/1.1/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/zh/1.1/user_guides/howto/training/cluster_howto.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/api/zh/1.1/fluid.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/zh/1.1/advanced_usage/development/contribute_to_paddle.html)
We appreciate your contributions!
......
benchmark/fluid/args.py
浏览文件 @ bc34d325
......@@ -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 @ bc34d325
......@@ -177,6 +177,7 @@ def train_parallel(train_args, test_args, args, train_prog, test_prog,
else:
build_strategy.reduce_strategy = fluid.BuildStrategy(
).ReduceStrategy.AllReduce
build_strategy.fuse_broadcast_op = args.fuse_broadcast_op
avg_loss = train_args[0]
......@@ -240,7 +241,6 @@ 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
......
cmake/configure.cmake
浏览文件 @ bc34d325
......@@ -50,11 +50,7 @@ if(NOT WITH_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)
if(WITH_AVX AND AVX_FOUND)
set(SIMD_FLAG ${AVX_FLAG})
elseif(SSE3_FOUND)
set(SIMD_FLAG ${SSE3_FLAG})
......@@ -222,3 +218,7 @@ 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/cuda.cmake
浏览文件 @ bc34d325
......@@ -157,6 +157,9 @@ 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
......@@ -196,10 +199,15 @@ elseif(CMAKE_BUILD_TYPE STREQUAL "MinSizeRel")
list(APPEND CUDA_NVCC_FLAGS ${CMAKE_CXX_FLAGS_RELEASE})
endif()
else(NOT WIN32)
if(CMAKE_BUILD_TYPE STREQUAL "Release")
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 build now. Please set visual studio build type to Release, x64 build.")
message(FATAL "Windows only support Release or Debug build now. Please set visual studio build type to Release/Debug, x64 build.")
endif()
endif(NOT WIN32)
......
cmake/cudnn.cmake
浏览文件 @ bc34d325
......@@ -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 ${CUDA_TOOLKIT_INCLUDE}
......
cmake/external/anakin.cmake
浏览文件 @ bc34d325
......@@ -58,19 +58,21 @@ ExternalProject_Add(
-DPROTOBUF_ROOT=${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 ${ANAKIN_SABER_LIB})
add_dependencies(anakin_saber extern_anakin protobuf mklml)
add_dependencies(anakin_saber extern_anakin)
list(APPEND external_project_dependencies anakin_shared anakin_saber)
cmake/external/boost.cmake
浏览文件 @ bc34d325
......@@ -28,34 +28,28 @@ if((NOT DEFINED BOOST_TAR) OR (NOT DEFINED BOOST_URL))
set(BOOST_TAR "boost_1_41_0" CACHE STRING "" FORCE)
set(BOOST_URL "http://paddlepaddledeps.cdn.bcebos.com/${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)
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)
......
cmake/external/dlpack.cmake 0 → 100644
浏览文件 @ bc34d325
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)
LIST(APPEND externl_project_dependencies dlpack)
cmake/external/eigen.cmake
浏览文件 @ bc34d325
......@@ -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 ${EIGEN_SOURCE_DIR}
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
......
cmake/external/gflags.cmake
浏览文件 @ bc34d325
......@@ -35,7 +35,12 @@ ExternalProject_Add(
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${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
......@@ -48,8 +53,8 @@ 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 ${GFLAGS_INSTALL_DIR}/lib/libgflags.lib
)
ENDIF()
ENDIF(WIN32)
ADD_LIBRARY(gflags STATIC IMPORTED GLOBAL)
......
cmake/external/glog.cmake
浏览文件 @ bc34d325
......@@ -46,7 +46,11 @@ ExternalProject_Add(
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${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=${CMAKE_C_FLAGS_RELEASE}
-DCMAKE_INSTALL_PREFIX=${GLOG_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR=${GLOG_INSTALL_DIR}/lib
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
......@@ -63,7 +67,7 @@ ExternalProject_Add(
IF(WIN32)
IF(NOT EXISTS "${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 ${GLOG_INSTALL_DIR}/lib/libglog.lib
)
ENDIF()
ENDIF(WIN32)
......
cmake/external/gtest.cmake
浏览文件 @ bc34d325
......@@ -50,7 +50,11 @@ IF(WITH_TESTING)
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${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
......
cmake/external/gzstream.cmake 0 → 100644
浏览文件 @ bc34d325
# 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.
#
IF(MOBILE_INFERENCE)
return()
ENDIF()
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(${GZSTREAM_INCLUDE_DIR})
ADD_DEPENDENCIES(gzstream extern_gzstream zlib)
cmake/external/mkldnn.cmake
浏览文件 @ bc34d325
......@@ -37,7 +37,6 @@ SET(CMAKE_INSTALL_RPATH_USE_LINK_PATH TRUE)
SET(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}" "${MKLDNN_INSTALL_DIR}/lib")
INCLUDE_DIRECTORIES(${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 ${MKLML_PROJECT})
......@@ -45,7 +44,7 @@ IF(${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 "-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}")
......@@ -54,7 +53,7 @@ ExternalProject_Add(
${EXTERNAL_PROJECT_LOG_ARGS}
DEPENDS ${MKLDNN_DEPENDS}
GIT_REPOSITORY "https://github.com/01org/mkl-dnn.git"
GIT_TAG "64e03a1939e0d526aa8e9f2e3f7dc0ad8d372944"
GIT_TAG "830a10059a018cd2634d94195140cf2d8790a75a"
PREFIX ${MKLDNN_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
......
cmake/external/ngraph.cmake 0 → 100644
浏览文件 @ bc34d325
# 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(ExternalProject)
SET(NGRAPH_PROJECT "extern_ngraph")
SET(NGRAPH_VERSION "0.9")
SET(NGRAPH_GIT_TAG "f9fd9d4cc318dc59dd4b68448e7fbb5f67a28bd0")
SET(NGRAPH_SOURCES_DIR ${THIRD_PARTY_PATH}/ngraph)
SET(NGRAPH_INSTALL_DIR ${THIRD_PARTY_PATH}/install/ngraph)
SET(NGRAPH_INC_DIR ${NGRAPH_INSTALL_DIR}/include)
SET(NGRAPH_SHARED_LIB_NAME libngraph.so.${NGRAPH_VERSION})
SET(NGRAPH_CPU_LIB_NAME libcpu_backend.so)
SET(NGRAPH_TBB_LIB_NAME libtbb.so.2)
SET(NGRAPH_GIT_REPO "https://github.com/NervanaSystems/ngraph.git")
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}/lib
)
if(UNIX AND NOT APPLE)
include(GNUInstallDirs)
SET(NGRAPH_LIB_DIR ${NGRAPH_INSTALL_DIR}/${CMAKE_INSTALL_LIBDIR})
else()
SET(NGRAPH_LIB_DIR ${NGRAPH_INSTALL_DIR}/lib)
endif()
MESSAGE(STATUS "nGraph lib will be installed at: ${NGRAPH_LIB_DIR}")
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})
# Workaround for nGraph expecting mklml to be in mkldnn install directory.
ExternalProject_Add_Step(
${NGRAPH_PROJECT}
PrepareMKL
COMMAND ${CMAKE_COMMAND} -E create_symlink ${MKLML_LIB} ${MKLDNN_INSTALL_DIR}/lib/libmklml_intel.so
COMMAND ${CMAKE_COMMAND} -E create_symlink ${MKLML_IOMP_LIB} ${MKLDNN_INSTALL_DIR}/lib/libiomp5.so
DEPENDEES download
DEPENDERS configure
)
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})
LIST(APPEND external_project_dependencies ngraph)
cmake/external/openblas.cmake
浏览文件 @ bc34d325
......@@ -17,12 +17,8 @@ IF(USE_EIGEN_FOR_BLAS)
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 +30,7 @@ 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})
......
cmake/external/protobuf.cmake
浏览文件 @ bc34d325
......@@ -30,66 +30,61 @@ UNSET_VAR(PROTOBUF_LITE_LIBRARY)
UNSET_VAR(PROTOBUF_LIBRARY)
UNSET_VAR(PROTOBUF_INCLUDE_DIR)
UNSET_VAR(Protobuf_PROTOC_EXECUTABLE)
function(protobuf_generate_python SRCS)
# shameless copy from https://github.com/Kitware/CMake/blob/master/Modules/FindProtobuf.cmake
if(NOT ARGN)
message(SEND_ERROR "Error: PROTOBUF_GENERATE_PYTHON() called without any proto files")
return()
endif()
if(NOT COMMAND protobuf_generate_python) # before cmake 3.4, protobuf_genrerate_python is not defined.
function(protobuf_generate_python SRCS)
# shameless copy from https://github.com/Kitware/CMake/blob/master/Modules/FindProtobuf.cmake
if(NOT ARGN)
message(SEND_ERROR "Error: PROTOBUF_GENERATE_PYTHON() called without any proto files")
return()
endif()
if(PROTOBUF_GENERATE_CPP_APPEND_PATH)
# Create an include path for each file specified
foreach(FIL ${ARGN})
get_filename_component(ABS_FIL ${FIL} ABSOLUTE)
get_filename_component(ABS_PATH ${ABS_FIL} PATH)
list(FIND _protobuf_include_path ${ABS_PATH} _contains_already)
if(${_contains_already} EQUAL -1)
list(APPEND _protobuf_include_path -I ${ABS_PATH})
endif()
endforeach()
else()
set(_protobuf_include_path -I ${CMAKE_CURRENT_SOURCE_DIR})
endif()
if(DEFINED PROTOBUF_IMPORT_DIRS AND NOT DEFINED Protobuf_IMPORT_DIRS)
set(Protobuf_IMPORT_DIRS "${PROTOBUF_IMPORT_DIRS}")
endif()
if(DEFINED Protobuf_IMPORT_DIRS)
foreach(DIR ${Protobuf_IMPORT_DIRS})
get_filename_component(ABS_PATH ${DIR} ABSOLUTE)
list(FIND _protobuf_include_path ${ABS_PATH} _contains_already)
if(${_contains_already} EQUAL -1)
list(APPEND _protobuf_include_path -I ${ABS_PATH})
endif()
endforeach()
endif()
set(${SRCS})
if(PROTOBUF_GENERATE_CPP_APPEND_PATH)
# Create an include path for each file specified
foreach(FIL ${ARGN})
get_filename_component(ABS_FIL ${FIL} ABSOLUTE)
get_filename_component(FIL_WE ${FIL} NAME_WE)
if(NOT PROTOBUF_GENERATE_CPP_APPEND_PATH)
get_filename_component(FIL_DIR ${FIL} DIRECTORY)
if(FIL_DIR)
set(FIL_WE "${FIL_DIR}/${FIL_WE}")
endif()
get_filename_component(ABS_PATH ${ABS_FIL} PATH)
list(FIND _protobuf_include_path ${ABS_PATH} _contains_already)
if(${_contains_already} EQUAL -1)
list(APPEND _protobuf_include_path -I ${ABS_PATH})
endif()
endforeach()
else()
set(_protobuf_include_path -I ${CMAKE_CURRENT_SOURCE_DIR})
endif()
if(DEFINED PROTOBUF_IMPORT_DIRS AND NOT DEFINED Protobuf_IMPORT_DIRS)
set(Protobuf_IMPORT_DIRS "${PROTOBUF_IMPORT_DIRS}")
endif()
list(APPEND ${SRCS} "${CMAKE_CURRENT_BINARY_DIR}/${FIL_WE}_pb2.py")
add_custom_command(
OUTPUT "${CMAKE_CURRENT_BINARY_DIR}/${FIL_WE}_pb2.py"
COMMAND ${Protobuf_PROTOC_EXECUTABLE} --python_out ${CMAKE_CURRENT_BINARY_DIR} ${_protobuf_include_path} ${ABS_FIL}
DEPENDS ${ABS_FIL} ${Protobuf_PROTOC_EXECUTABLE}
COMMENT "Running Python protocol buffer compiler on ${FIL}"
VERBATIM )
if(DEFINED Protobuf_IMPORT_DIRS)
foreach(DIR ${Protobuf_IMPORT_DIRS})
get_filename_component(ABS_PATH ${DIR} ABSOLUTE)
list(FIND _protobuf_include_path ${ABS_PATH} _contains_already)
if(${_contains_already} EQUAL -1)
list(APPEND _protobuf_include_path -I ${ABS_PATH})
endif()
endforeach()
endif()
set(${SRCS} ${${SRCS}} PARENT_SCOPE)
endfunction()
endif()
set(${SRCS})
foreach(FIL ${ARGN})
get_filename_component(ABS_FIL ${FIL} ABSOLUTE)
get_filename_component(FIL_WE ${FIL} NAME_WE)
if(NOT PROTOBUF_GENERATE_CPP_APPEND_PATH)
get_filename_component(FIL_DIR ${FIL} DIRECTORY)
if(FIL_DIR)
set(FIL_WE "${FIL_DIR}/${FIL_WE}")
endif()
endif()
list(APPEND ${SRCS} "${CMAKE_CURRENT_BINARY_DIR}/${FIL_WE}_pb2.py")
add_custom_command(
OUTPUT "${CMAKE_CURRENT_BINARY_DIR}/${FIL_WE}_pb2.py"
COMMAND ${PROTOBUF_PROTOC_EXECUTABLE} --python_out ${CMAKE_CURRENT_BINARY_DIR} ${_protobuf_include_path} ${ABS_FIL}
DEPENDS ${ABS_FIL} ${PROTOBUF_PROTOC_EXECUTABLE}
COMMENT "Running Python protocol buffer compiler on ${FIL}"
VERBATIM )
endforeach()
set(${SRCS} ${${SRCS}} PARENT_SCOPE)
endfunction()
# Print and set the protobuf library information,
# finish this cmake process and exit from this file.
......@@ -126,6 +121,7 @@ macro(PROMPT_PROTOBUF_LIB)
# FIND_Protobuf.cmake uses `Protobuf_PROTOC_EXECUTABLE`.
# make `protobuf_generate_cpp` happy.
SET(Protobuf_PROTOC_EXECUTABLE ${PROTOBUF_PROTOC_EXECUTABLE})
FOREACH(dep ${protobuf_DEPS})
ADD_DEPENDENCIES(protobuf ${dep})
ADD_DEPENDENCIES(protobuf_lite ${dep})
......@@ -144,7 +140,6 @@ endmacro()
set(PROTOBUF_ROOT "" CACHE PATH "Folder contains protobuf")
IF (WIN32)
SET(PROTOBUF_ROOT ${THIRD_PARTY_PATH}/install/protobuf)
MESSAGE(WARNING, "In windows, protobuf only support msvc build, please build it manually and put it at " ${PROTOBUF_ROOT})
ENDIF(WIN32)
if (NOT "${PROTOBUF_ROOT}" STREQUAL "")
......@@ -192,13 +187,20 @@ FUNCTION(build_protobuf TARGET_NAME BUILD_FOR_HOST)
SET(OPTIONAL_ARGS
"-DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}"
"-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}"
"-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}"
"-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}"
"-DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}"
"-DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}"
"-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}"
"-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}"
"-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}"
"-Dprotobuf_WITH_ZLIB=ON"
"-DZLIB_ROOT:FILEPATH=${ZLIB_ROOT}"
${EXTERNAL_OPTIONAL_ARGS})
SET(OPTIONAL_CACHE_ARGS "-DZLIB_ROOT:STRING=${ZLIB_ROOT}")
ENDIF()
IF(WIN32)
SET(OPTIONAL_ARGS ${OPTIONAL_ARGS} "-DCMAKE_GENERATOR_PLATFORM=x64")
ENDIF()
SET(PROTOBUF_REPO "https://github.com/google/protobuf.git")
SET(PROTOBUF_TAG "9f75c5aa851cd877fb0d93ccc31b8567a6706546")
......
cmake/external/pybind11.cmake
浏览文件 @ bc34d325
......@@ -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 @ bc34d325
......@@ -21,6 +21,48 @@ INCLUDE(python_module)
FIND_PACKAGE(PythonInterp ${PY_VERSION})
FIND_PACKAGE(PythonLibs ${PY_VERSION})
if(WIN32)
execute_process(COMMAND "${PYTHON_EXECUTABLE}" "-c"
"from distutils import sysconfig as s;import sys;import struct;
print(sys.prefix);
print(s.get_config_var('LDVERSION') or s.get_config_var('VERSION'));
"
RESULT_VARIABLE _PYTHON_SUCCESS
OUTPUT_VARIABLE _PYTHON_VALUES
ERROR_VARIABLE _PYTHON_ERROR_VALUE)
if(NOT _PYTHON_SUCCESS MATCHES 0)
set(PYTHONLIBS_FOUND FALSE)
return()
endif()
# Convert the process output into a list
string(REGEX REPLACE ";" "\\\\;" _PYTHON_VALUES ${_PYTHON_VALUES})
string(REGEX REPLACE "\n" ";" _PYTHON_VALUES ${_PYTHON_VALUES})
list(GET _PYTHON_VALUES 0 PYTHON_PREFIX)
list(GET _PYTHON_VALUES 1 PYTHON_LIBRARY_SUFFIX)
# Make sure all directory separators are '/'
string(REGEX REPLACE "\\\\" "/" PYTHON_PREFIX ${PYTHON_PREFIX})
set(PYTHON_LIBRARY
"${PYTHON_PREFIX}/libs/Python${PYTHON_LIBRARY_SUFFIX}.lib")
# when run in a venv, PYTHON_PREFIX points to it. But the libraries remain in the
# original python installation. They may be found relative to PYTHON_INCLUDE_DIR.
if(NOT EXISTS "${PYTHON_LIBRARY}")
get_filename_component(_PYTHON_ROOT ${PYTHON_INCLUDE_DIR} DIRECTORY)
set(PYTHON_LIBRARY
"${_PYTHON_ROOT}/libs/Python${PYTHON_LIBRARY_SUFFIX}.lib")
endif()
# raise an error if the python libs are still not found.
if(NOT EXISTS "${PYTHON_LIBRARY}")
message(FATAL_ERROR "Python libraries not found")
endif()
SET(PYTHON_LIBRARIES "${PYTHON_LIBRARY}")
endif(WIN32)
# Fixme: Maybe find a static library. Get SHARED/STATIC by FIND_PACKAGE.
ADD_LIBRARY(python SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET python PROPERTY IMPORTED_LOCATION ${PYTHON_LIBRARIES})
......
cmake/external/rocprim.cmake 0 → 100644
浏览文件 @ bc34d325
if (NOT WITH_AMD_GPU)
return()
endif()
# rocprim is "ROCm Parallel Primitives" for short.
# It is a header-only library providing HIP and HC parallel primitives
# for developing performant GPU-accelerated code on AMD ROCm platform.
if("x${HCC_HOME}" STREQUAL "x")
set(HCC_HOME "/opt/rocm/hcc")
endif()
INCLUDE(ExternalProject)
SET(ROCPRIM_SOURCE_DIR ${THIRD_PARTY_PATH}/rocprim)
SET(ROCPRIM_INSTALL_DIR ${THIRD_PARTY_PATH}/install/rocprim)
SET(ROCPRIM_INCLUDE_DIR ${ROCPRIM_INSTALL_DIR}/include)
ExternalProject_Add(
extern_rocprim
GIT_REPOSITORY "https://github.com/ROCmSoftwarePlatform/rocPRIM.git"
GIT_TAG 5bd41b96ab8d8343330fb2c3e1b96775bde3b3fc
PREFIX ${ROCPRIM_SOURCE_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${HCC_HOME}/bin/hcc
CMAKE_ARGS -DONLY_INSTALL=ON
CMAKE_ARGS -DBUILD_TEST=OFF
CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=${ROCPRIM_INSTALL_DIR}
INSTALL_DIR ${ROCPRIM_INSTALL_DIR}
${EXTERNAL_PROJECT_LOG_ARGS}
)
INCLUDE_DIRECTORIES(${ROCPRIM_INCLUDE_DIR})
if (${CMAKE_VERSION} VERSION_LESS "3.3.0")
set(dummyfile ${CMAKE_CURRENT_BINARY_DIR}/rocprim_dummy.c)
file(WRITE ${dummyfile} "const char *dummy_rocprim = \"${dummyfile}\";")
add_library(rocprim STATIC ${dummyfile})
else()
add_library(rocprim INTERFACE)
endif()
add_dependencies(rocprim extern_rocprim)
cmake/external/snappy.cmake
浏览文件 @ bc34d325
......@@ -24,7 +24,11 @@ set(SNAPPY_SOURCES_DIR ${THIRD_PARTY_PATH}/snappy)
set(SNAPPY_INSTALL_DIR ${THIRD_PARTY_PATH}/install/snappy)
set(SNAPPY_INCLUDE_DIR "${SNAPPY_INSTALL_DIR}/include" CACHE PATH "snappy include directory." FORCE)
set(SNAPPY_LIBRARIES "${SNAPPY_INSTALL_DIR}/lib/libsnappy.a")
if (WIN32)
set(SNAPPY_LIBRARIES "${SNAPPY_INSTALL_DIR}/lib/snappy.lib")
else(WIN32)
set(SNAPPY_LIBRARIES "${SNAPPY_INSTALL_DIR}/lib/libsnappy.a")
endif (WIN32)
ExternalProject_Add(
extern_snappy
......@@ -34,8 +38,12 @@ ExternalProject_Add(
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
-DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}
-DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
-DCMAKE_INSTALL_PREFIX=${SNAPPY_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR=${SNAPPY_INSTALL_DIR}/lib
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
......
cmake/external/snappystream.cmake
浏览文件 @ bc34d325
......@@ -18,36 +18,45 @@ ENDIF()
include (ExternalProject)
# NOTE: snappy is needed when linking with recordio
set(SNAPPYSTREAM_SOURCES_DIR ${THIRD_PARTY_PATH}/snappy_stream)
set(SNAPPYSTREAM_INSTALL_DIR ${THIRD_PARTY_PATH}/install/snappy_stream)
set(SNAPPYSTREAM_INCLUDE_DIR "${SNAPPYSTREAM_INSTALL_DIR}/include" CACHE PATH "snappy stream include directory." FORCE)
set(SNAPPYSTREAM_LIBRARIES "${SNAPPYSTREAM_INSTALL_DIR}/lib/libsnappystream.a")
ExternalProject_Add(
extern_snappystream
GIT_REPOSITORY "https://github.com/hoxnox/snappystream.git"
GIT_TAG "0.2.8"
PREFIX ${SNAPPYSTREAM_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
-DCMAKE_INSTALL_PREFIX=${SNAPPY_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR=${SNAPPY_INSTALL_DIR}/lib
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
-DSNAPPY_ROOT=${SNAPPY_INSTALL_DIR}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS
-DCMAKE_INSTALL_PREFIX:PATH=${SNAPPYSTREAM_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR:PATH=${SNAPPYSTREAM_INSTALL_DIR}/lib
-DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
DEPENDS snappy
)
if(WIN32)
# Fix me, VS2015 come without VLA support
set(SNAPPYSTREAM_LIBRARIES "${SNAPPYSTREAM_INSTALL_DIR}/lib/snappystream.lib")
MESSAGE(WARNING, "In windows, snappystream has no compile support for windows,
please build it manually and put it at " ${SNAPPYSTREAM_INSTALL_DIR})
else(WIN32)
set(SNAPPYSTREAM_LIBRARIES "${SNAPPYSTREAM_INSTALL_DIR}/lib/libsnappystream.a")
ExternalProject_Add(
extern_snappystream
GIT_REPOSITORY "https://github.com/hoxnox/snappystream.git"
GIT_TAG "0.2.8"
PREFIX ${SNAPPYSTREAM_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
-DCMAKE_C_FLAGS=${CMAKE_C_FLAGS}
-DCMAKE_C_FLAGS_DEBUG=${CMAKE_C_FLAGS_DEBUG}
-DCMAKE_C_FLAGS_RELEASE=${CMAKE_C_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
-DCMAKE_INSTALL_PREFIX=${SNAPPY_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR=${SNAPPY_INSTALL_DIR}/lib
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
-DSNAPPY_ROOT=${SNAPPY_INSTALL_DIR}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS
-DCMAKE_INSTALL_PREFIX:PATH=${SNAPPYSTREAM_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR:PATH=${SNAPPYSTREAM_INSTALL_DIR}/lib
-DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
DEPENDS snappy
)
endif(WIN32)
add_library(snappystream STATIC IMPORTED GLOBAL)
set_property(TARGET snappystream PROPERTY IMPORTED_LOCATION ${SNAPPYSTREAM_LIBRARIES})
......
cmake/external/xxhash.cmake 0 → 100644
浏览文件 @ bc34d325
INCLUDE(ExternalProject)
set(XXHASH_SOURCE_DIR ${THIRD_PARTY_PATH}/xxhash)
set(XXHASH_INSTALL_DIR ${THIRD_PARTY_PATH}/install/xxhash)
set(XXHASH_INCLUDE_DIR "${XXHASH_INSTALL_DIR}/include")
IF(WITH_STATIC_LIB)
SET(BUILD_CMD make lib)
ELSE()
IF(APPLE)
SET(BUILD_CMD sed -i \"\" "s/-Wstrict-prototypes -Wundef/-Wstrict-prototypes -Wundef -fPIC/g" ${XXHASH_SOURCE_DIR}/src/extern_xxhash/Makefile && make lib)
ELSE(APPLE)
SET(BUILD_CMD sed -i "s/-Wstrict-prototypes -Wundef/-Wstrict-prototypes -Wundef -fPIC/g" ${XXHASH_SOURCE_DIR}/src/extern_xxhash/Makefile && make lib)
ENDIF(APPLE)
ENDIF()
if(WIN32)
ExternalProject_Add(
extern_xxhash
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/Cyan4973/xxHash"
GIT_TAG "v0.6.5"
PREFIX ${XXHASH_SOURCE_DIR}
DOWNLOAD_NAME "xxhash"
UPDATE_COMMAND ""
BUILD_IN_SOURCE 1
PATCH_COMMAND
CONFIGURE_COMMAND
${CMAKE_COMMAND} ${XXHASH_SOURCE_DIR}/src/extern_xxhash/cmake_unofficial
-DCMAKE_INSTALL_PREFIX:PATH=${XXHASH_INSTALL_DIR}
-DCMAKE_BUILD_TYPE:STRING=${CMAKE_BUILD_TYPE}
-DCMAKE_VERBOSE_MAKEFILE:BOOL=OFF
-DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
-DBUILD_XXHSUM=OFF
-DCMAKE_GENERATOR_PLATFORM=x64
-DBUILD_SHARED_LIBS=OFF
${OPTIONAL_CACHE_ARGS}
TEST_COMMAND ""
)
else()
ExternalProject_Add(
extern_xxhash
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/Cyan4973/xxHash"
GIT_TAG "v0.6.5"
PREFIX ${XXHASH_SOURCE_DIR}
DOWNLOAD_NAME "xxhash"
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
BUILD_IN_SOURCE 1
PATCH_COMMAND
BUILD_COMMAND ${BUILD_CMD}
INSTALL_COMMAND export PREFIX=${XXHASH_INSTALL_DIR}/ && make install
TEST_COMMAND ""
)
endif()
if (WIN32)
set(XXHASH_LIBRARIES "${XXHASH_INSTALL_DIR}/lib/xxhash.lib")
else()
set(XXHASH_LIBRARIES "${XXHASH_INSTALL_DIR}/lib/libxxhash.a")
endif ()
INCLUDE_DIRECTORIES(${XXHASH_INCLUDE_DIR})
add_library(xxhash STATIC IMPORTED GLOBAL)
set_property(TARGET xxhash PROPERTY IMPORTED_LOCATION ${XXHASH_LIBRARIES})
include_directories(${XXHASH_INCLUDE_DIR})
add_dependencies(xxhash extern_xxhash)
LIST(APPEND external_project_dependencies xxhash)
IF(WITH_C_API)
INSTALL(DIRECTORY ${XXHASH_INCLUDE_DIR} DESTINATION third_party/xxhash)
IF(ANDROID)
INSTALL(FILES ${XXHASH_LIBRARIES} DESTINATION third_party/xxhash/lib/${ANDROID_ABI})
ELSE()
INSTALL(FILES ${XXHASH_LIBRARIES} DESTINATION third_party/xxhash/lib)
ENDIF()
ENDIF()
cmake/flags.cmake
浏览文件 @ bc34d325
......@@ -129,6 +129,9 @@ set(COMMON_FLAGS
-Wno-error=parentheses-equality # Warnings in pybind11
-Wno-error=ignored-attributes # Warnings in Eigen, gcc 6.3
-Wno-error=terminate # Warning in PADDLE_ENFORCE
-Wno-error=int-in-bool-context # Warning in Eigen gcc 7.2
-Wimplicit-fallthrough=0 # Warning in tinyformat.h
-Wno-error=maybe-uninitialized # Warning in boost gcc 7.2
)
set(GPU_COMMON_FLAGS
......
cmake/generic.cmake
浏览文件 @ bc34d325
......@@ -266,7 +266,11 @@ function(cc_library TARGET_NAME)
if("${cc_library_DEPS};" MATCHES "python;")
list(REMOVE_ITEM cc_library_DEPS python)
add_dependencies(${TARGET_NAME} python)
target_link_libraries(${TARGET_NAME} "-Wl,-undefined,dynamic_lookup")
if(WIN32)
target_link_libraries(${TARGET_NAME} ${PYTHON_LIBRARIES})
else()
target_link_libraries(${TARGET_NAME} "-Wl,-undefined,dynamic_lookup")
endif(WIN32)
endif()
target_link_libraries(${TARGET_NAME} ${cc_library_DEPS})
add_dependencies(${TARGET_NAME} ${cc_library_DEPS})
......@@ -288,6 +292,45 @@ function(cc_library TARGET_NAME)
endif(cc_library_SRCS)
endfunction(cc_library)
# The link operation under windows may exceeds the maximum characters limit, simply break the link command
# into multiple link opeartion can fix that, say
# original:
# lib /out:target.lib a.lib b.lib c.lib d.lib
# after:
# 1. lib /out:dummy_lib_1.lib a.lib b.lib
# 2. lib /out:dummy_lib_2.lib c.lib d.lib
# 1. lib /out:target.lib dummy_lib_1.lib dummy_lib_2.lib
function(sep_library TARGET_NAME)
set(options STATIC static SHARED shared)
set(oneValueArgs "")
set(multiValueArgs SRCS DEPS)
cmake_parse_arguments(sep_library "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
set(dummy_index 1)
set(dummy_offset 1)
# the dummy target would be consisted of limit size libraries
set(dummy_limit 50)
list(LENGTH sep_library_DEPS sep_all_len)
foreach(v ${sep_library_DEPS})
list(APPEND dummy_list ${v})
list(LENGTH dummy_list listlen )
if ((${listlen} GREATER ${dummy_limit}) OR (${dummy_offset} EQUAL ${sep_all_len}))
message("create dummy library ${TARGET_NAME}_dummy_lib_${dummy_index} for ${TARGET_NAME}")
cc_library(${TARGET_NAME}_dummy_lib_${dummy_index} STATIC DEPS ${dummy_list})
foreach(i ${dummy_list})
list(REMOVE_AT dummy_list 0)
endforeach()
list(APPEND ${TARGET_NAME}_dummy_list ${TARGET_NAME}_dummy_lib_${dummy_index})
MATH(EXPR dummy_index "${dummy_index}+1")
endif()
MATH(EXPR dummy_offset "${dummy_offset}+1")
endforeach()
if(${sep_library_SHARED})
cc_library(${TARGET_NAME} SHARED SRCS ${sep_library_SRCS} DEPS ${${TARGET_NAME}_dummy_list})
else(${sep_library_SHARED})
cc_library(${TARGET_NAME} STATIC SRCS ${sep_library_SRCS} DEPS ${${TARGET_NAME}_dummy_list})
endif(${sep_library_SHARED})
endfunction(sep_library)
function(cc_binary TARGET_NAME)
set(options "")
set(oneValueArgs "")
......@@ -306,8 +349,18 @@ function(cc_test TARGET_NAME)
set(oneValueArgs "")
set(multiValueArgs SRCS DEPS ARGS)
cmake_parse_arguments(cc_test "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
if(WIN32)
list(APPEND win32_deps shlwapi)
if("${cc_test_DEPS};" MATCHES "python;")
list(REMOVE_ITEM cc_test_DEPS python)
list(APPEND win32_deps ${PYTHON_LIBRARIES})
endif()
endif(WIN32)
add_executable(${TARGET_NAME} ${cc_test_SRCS})
target_link_libraries(${TARGET_NAME} ${cc_test_DEPS} paddle_gtest_main lod_tensor memory gtest gflags glog)
if(WIN32)
target_link_libraries(${TARGET_NAME} ${win32_deps})
endif(WIN32)
add_dependencies(${TARGET_NAME} ${cc_test_DEPS} paddle_gtest_main lod_tensor memory gtest gflags glog)
add_test(NAME ${TARGET_NAME}
COMMAND ${TARGET_NAME} ${cc_test_ARGS}
......@@ -408,25 +461,29 @@ function(hip_library TARGET_NAME)
else()
add_library(${TARGET_NAME} STATIC ${_cmake_options} ${_generated_files} ${_sources})
set_target_properties(${TARGET_NAME} PROPERTIES LINKER_LANGUAGE CXX)
target_link_libraries(${TARGET_NAME} /opt/rocm/hip/lib/libhip_hcc.so /opt/rocm/hip/lib/libhip_device.a)
find_fluid_modules(${TARGET_NAME})
target_link_libraries(${TARGET_NAME} /opt/rocm/hip/lib/libhip_hcc.so /opt/rocm/hip/lib/libhip_device.a /opt/rocm/rccl/lib/librccl.so /opt/rocm/hiprand/lib/libhiprand.so)
find_fluid_modules(${TARGET_NAME})
endif()
if (hip_library_DEPS)
add_dependencies(${TARGET_NAME} ${hip_library_DEPS})
target_link_libraries(${TARGET_NAME} ${hip_library_DEPS})
if("${hip_library_DEPS}" MATCHES "ARCHIVE_START")
# Support linking flags: --whole-archive (Linux) / -force_load (MacOS).
# WARNING: Please don't use ARCHIVE_START&ARCHIVE_END if TARGET_NAME will be linked by other libraries.
target_circle_link_libraries(${TARGET_NAME} ${hip_library_DEPS})
list(REMOVE_ITEM hip_library_DEPS ARCHIVE_START ARCHIVE_END)
else()
target_link_libraries(${TARGET_NAME} ${hip_library_DEPS})
endif()
# cpplint code style
foreach(source_file ${hip_library_SRCS})
string(REGEX REPLACE "\\.[^.]*$" "" source ${source_file})
if(EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${source}.h)
list(APPEND hip_library_HEADERS ${CMAKE_CURRENT_SOURCE_DIR}/${source}.h)
endif()
string(REGEX REPLACE "\\.[^.]*$" "" source ${source_file})
if(EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${source}.h)
list(APPEND hip_library_HEADERS ${CMAKE_CURRENT_SOURCE_DIR}/${source}.h)
endif()
endforeach()
else(hip_library_SRCS)
if (hip_library_DEPS)
merge_static_libs(${TARGET_NAME} ${hip_library_DEPS})
merge_static_libs(${TARGET_NAME} ${hip_library_DEPS})
else()
message(FATAL "Please specify source file or library in nv_library.")
message(FATAL "Please specify source file or library in nv_library.")
endif()
endif(hip_library_SRCS)
endif()
......@@ -633,7 +690,7 @@ function(py_test TARGET_NAME)
set(multiValueArgs SRCS DEPS ARGS ENVS)
cmake_parse_arguments(py_test "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
add_test(NAME ${TARGET_NAME}
COMMAND env FLAGS_init_allocated_mem=true FLAGS_cudnn_deterministic=true
COMMAND ${CMAKE_COMMAND} -E env FLAGS_init_allocated_mem=true FLAGS_cudnn_deterministic=true
FLAGS_cpu_deterministic=true
PYTHONPATH=${PADDLE_BINARY_DIR}/python ${py_test_ENVS}
${PYTHON_EXECUTABLE} -u ${py_test_SRCS} ${py_test_ARGS}
......
cmake/hip.cmake
浏览文件 @ bc34d325
......@@ -3,6 +3,8 @@ if(NOT WITH_AMD_GPU)
endif()
include_directories("/opt/rocm/include")
include_directories("/opt/rocm/hip/include")
include_directories("/opt/rocm/miopen/include")
include_directories("/opt/rocm/hipblas/include")
include_directories("/opt/rocm/hiprand/include")
include_directories("/opt/rocm/rocrand/include")
......@@ -11,20 +13,40 @@ include_directories("/opt/rocm/thrust")
list(APPEND EXTERNAL_LIBS "-L/opt/rocm/lib/ -lhip_hcc")
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -fPIC -DPADDLE_WITH_HIP -std=c++14" )
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -fPIC -DPADDLE_WITH_HIP -std=c++11" )
if(WITH_DSO)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_USE_DSO")
endif(WITH_DSO)
if(WITH_DOUBLE)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_TYPE_DOUBLE")
endif(WITH_DOUBLE)
if(WITH_TESTING)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_TESTING")
endif(WITH_TESTING)
if(WITH_DISTRIBUTE)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_DISTRIBUTE")
endif(WITH_DISTRIBUTE)
if(WITH_GRPC)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_GRPC")
endif(WITH_GRPC)
if(NOT WITH_GOLANG)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITHOUT_GOLANG")
endif(NOT WITH_GOLANG)
if(WITH_MKLDNN)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_WITH_MKLDNN")
endif(WITH_MKLDNN)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DANY_IMPL_ANY_CAST_MOVEABLE")
if(NOT WITH_RDMA)
set(HIP_HCC_FLAGS "${HIP_HCC_FLAGS} -DPADDLE_DISABLE_RDMA")
endif(NOT WITH_RDMA)
if(CMAKE_BUILD_TYPE STREQUAL "Debug")
list(APPEND HIP_HCC_FLAGS ${CMAKE_CXX_FLAGS_DEBUG})
elseif(CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo")
......
cmake/inference_lib.cmake
浏览文件 @ bc34d325
......@@ -22,183 +22,210 @@ function(copy TARGET)
list(LENGTH copy_lib_SRCS copy_lib_SRCS_len)
list(LENGTH copy_lib_DSTS copy_lib_DSTS_len)
if(NOT ${copy_lib_SRCS_len} EQUAL ${copy_lib_DSTS_len})
if (NOT ${copy_lib_SRCS_len} EQUAL ${copy_lib_DSTS_len})
message(FATAL_ERROR "${TARGET} source numbers are not equal to destination numbers")
endif()
endif ()
math(EXPR len "${copy_lib_SRCS_len} - 1")
add_custom_target(${TARGET} DEPENDS ${copy_lib_DEPS})
foreach(index RANGE ${len})
foreach (index RANGE ${len})
list(GET copy_lib_SRCS ${index} src)
list(GET copy_lib_DSTS ${index} dst)
add_custom_command(TARGET ${TARGET} PRE_BUILD
COMMAND mkdir -p "${dst}"
COMMAND cp -r "${src}" "${dst}"
COMMENT "copying ${src} -> ${dst}")
endforeach()
if (WIN32)
# windows cmd shell will not expand wildcard automatically.
# below expand the files,libs and copy them by rules.
file(GLOB header_files ${src} "*.h")
file(GLOB static_lib_files ${src} "*.lib")
file(GLOB dll_lib_files ${src} "*.dll")
set(src_files ${header_files} ${static_lib_files} ${dll_lib_files})
if (NOT "${src_files}" STREQUAL "")
list(REMOVE_DUPLICATES src_files)
endif ()
add_custom_command(TARGET ${TARGET} PRE_BUILD
COMMAND ${CMAKE_COMMAND} -E make_directory "${dst}"
)
foreach (src_file ${src_files})
add_custom_command(TARGET ${TARGET} PRE_BUILD
COMMAND ${CMAKE_COMMAND} -E copy "${src_file}" "${dst}"
COMMENT "copying ${src_file} -> ${dst}")
endforeach ()
else (WIN32) # not windows
add_custom_command(TARGET ${TARGET} PRE_BUILD
COMMAND mkdir -p "${dst}"
COMMAND cp -r "${src}" "${dst}"
COMMENT "copying ${src} -> ${dst}")
endif (WIN32) # not windows
endforeach ()
endfunction()
# third party
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/eigen3")
copy(eigen3_lib
SRCS ${EIGEN_INCLUDE_DIR}/Eigen/Core ${EIGEN_INCLUDE_DIR}/Eigen/src ${EIGEN_INCLUDE_DIR}/unsupported/Eigen
DSTS ${dst_dir}/Eigen ${dst_dir}/Eigen ${dst_dir}/unsupported
DEPS eigen3
)
SRCS ${EIGEN_INCLUDE_DIR}/Eigen/Core ${EIGEN_INCLUDE_DIR}/Eigen/src ${EIGEN_INCLUDE_DIR}/unsupported/Eigen
DSTS ${dst_dir}/Eigen ${dst_dir}/Eigen ${dst_dir}/unsupported
DEPS eigen3
)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/gflags")
copy(gflags_lib
SRCS ${GFLAGS_INCLUDE_DIR} ${GFLAGS_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS gflags
)
SRCS ${GFLAGS_INCLUDE_DIR} ${GFLAGS_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS gflags
)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/glog")
copy(glog_lib
SRCS ${GLOG_INCLUDE_DIR} ${GLOG_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS glog
)
SRCS ${GLOG_INCLUDE_DIR} ${GLOG_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS glog
)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/boost/")
copy(boost_lib
SRCS ${BOOST_INCLUDE_DIR}/boost
DSTS ${dst_dir}
DEPS boost
)
if(NOT PROTOBUF_FOUND)
SRCS ${BOOST_INCLUDE_DIR}/boost
DSTS ${dst_dir}
DEPS boost
)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/xxhash")
copy(xxhash_lib
SRCS ${XXHASH_INCLUDE_DIR} ${XXHASH_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS xxhash
)
if (NOT PROTOBUF_FOUND)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/protobuf")
copy(protobuf_lib
SRCS ${PROTOBUF_INCLUDE_DIR} ${PROTOBUF_LIBRARY}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS extern_protobuf
)
endif()
SRCS ${PROTOBUF_INCLUDE_DIR} ${PROTOBUF_LIBRARY}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS extern_protobuf
)
endif ()
if(NOT CBLAS_FOUND)
if (NOT CBLAS_FOUND)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/openblas")
copy(openblas_lib
SRCS ${CBLAS_INSTALL_DIR}/lib ${CBLAS_INSTALL_DIR}/include
DSTS ${dst_dir} ${dst_dir}
DEPS extern_openblas
)
SRCS ${CBLAS_INSTALL_DIR}/lib ${CBLAS_INSTALL_DIR}/include
DSTS ${dst_dir} ${dst_dir}
DEPS extern_openblas
)
elseif (WITH_MKLML)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/mklml")
copy(mklml_lib
SRCS ${MKLML_LIB} ${MKLML_IOMP_LIB} ${MKLML_INC_DIR}
DSTS ${dst_dir}/lib ${dst_dir}/lib ${dst_dir}
DEPS mklml
)
endif()
if(WITH_MKLDNN)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/mkldnn")
copy(mkldnn_lib
SRCS ${MKLDNN_INC_DIR} ${MKLDNN_SHARED_LIB}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS mkldnn
)
endif()
SRCS ${MKLML_LIB} ${MKLML_IOMP_LIB} ${MKLML_INC_DIR}
DSTS ${dst_dir}/lib ${dst_dir}/lib ${dst_dir}
DEPS mklml
)
endif ()
if (WITH_MKLDNN)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/mkldnn")
copy(mkldnn_lib
SRCS ${MKLDNN_INC_DIR} ${MKLDNN_SHARED_LIB}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS mkldnn
)
endif ()
if (NOT WIN32)
if(NOT MOBILE_INFERENCE AND NOT RPI)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappy")
copy(snappy_lib
SRCS ${SNAPPY_INCLUDE_DIR} ${SNAPPY_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS snappy)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappystream")
copy(snappystream_lib
SRCS ${SNAPPYSTREAM_INCLUDE_DIR} ${SNAPPYSTREAM_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS snappystream)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/zlib")
copy(zlib_lib
SRCS ${ZLIB_INCLUDE_DIR} ${ZLIB_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS zlib)
endif()
endif(NOT WIN32)
if (NOT MOBILE_INFERENCE AND NOT RPI)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappy")
copy(snappy_lib
SRCS ${SNAPPY_INCLUDE_DIR} ${SNAPPY_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS snappy)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/snappystream")
copy(snappystream_lib
SRCS ${SNAPPYSTREAM_INCLUDE_DIR} ${SNAPPYSTREAM_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS snappystream)
set(dst_dir "${FLUID_INSTALL_DIR}/third_party/install/zlib")
copy(zlib_lib
SRCS ${ZLIB_INCLUDE_DIR} ${ZLIB_LIBRARIES}
DSTS ${dst_dir} ${dst_dir}/lib
DEPS zlib)
endif ()
endif (NOT WIN32)
# paddle fluid module
set(src_dir "${PADDLE_SOURCE_DIR}/paddle/fluid")
set(dst_dir "${FLUID_INSTALL_DIR}/paddle/fluid")
set(module "framework")
if (NOT WIN32)
set(framework_lib_deps framework_py_proto)
endif(NOT WIN32)
set(framework_lib_deps framework_py_proto)
endif (NOT WIN32)
copy(framework_lib DEPS ${framework_lib_deps}
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/details/*.h ${PADDLE_BINARY_DIR}/paddle/fluid/framework/framework.pb.h
${src_dir}/${module}/ir/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/details ${dst_dir}/${module} ${dst_dir}/${module}/ir
)
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/details/*.h ${PADDLE_BINARY_DIR}/paddle/fluid/framework/framework.pb.h
${src_dir}/${module}/ir/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/details ${dst_dir}/${module} ${dst_dir}/${module}/ir
)
set(module "memory")
copy(memory_lib
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/detail/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/detail
)
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/detail/*.h ${src_dir}/${module}/allocation/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/detail ${dst_dir}/${module}/allocation
)
set(inference_deps paddle_fluid_shared paddle_fluid)
set(module "inference/api")
if (WITH_ANAKIN AND WITH_MKL)
copy(anakin_inference_lib DEPS paddle_inference_api inference_anakin_api
SRCS
${PADDLE_BINARY_DIR}/paddle/fluid/inference/api/libinference_anakin_api* # compiled anakin api
${ANAKIN_INSTALL_DIR} # anakin release
DSTS ${FLUID_INSTALL_DIR}/third_party/install/anakin ${FLUID_INSTALL_DIR}/third_party/install/anakin)
list(APPEND inference_deps anakin_inference_lib)
endif()
SRCS
${PADDLE_BINARY_DIR}/paddle/fluid/inference/api/libinference_anakin_api* # compiled anakin api
${ANAKIN_INSTALL_DIR} # anakin release
DSTS ${FLUID_INSTALL_DIR}/third_party/install/anakin ${FLUID_INSTALL_DIR}/third_party/install/anakin)
list(APPEND inference_deps anakin_inference_lib)
endif ()
set(module "inference")
copy(inference_lib DEPS ${inference_deps}
SRCS ${src_dir}/${module}/*.h ${PADDLE_BINARY_DIR}/paddle/fluid/inference/libpaddle_fluid.*
${src_dir}/${module}/api/paddle_inference_api.h
${PADDLE_BINARY_DIR}/paddle/fluid/inference/api/paddle_inference_pass.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module} ${dst_dir}/${module} ${dst_dir}/${module}
)
${src_dir}/${module}/api/paddle_*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module} ${dst_dir}/${module}
)
set(module "platform")
copy(platform_lib DEPS profiler_py_proto
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/dynload/*.h ${src_dir}/${module}/details/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/dynload ${dst_dir}/${module}/details
)
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/dynload/*.h ${src_dir}/${module}/details/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/dynload ${dst_dir}/${module}/details
)
set(module "string")
copy(string_lib
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/tinyformat/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/tinyformat
)
SRCS ${src_dir}/${module}/*.h ${src_dir}/${module}/tinyformat/*.h
DSTS ${dst_dir}/${module} ${dst_dir}/${module}/tinyformat
)
set(module "pybind")
copy(pybind_lib
SRCS ${CMAKE_CURRENT_BINARY_DIR}/paddle/fluid/${module}/pybind.h
DSTS ${dst_dir}/${module}
)
SRCS ${CMAKE_CURRENT_BINARY_DIR}/paddle/fluid/${module}/pybind.h
DSTS ${dst_dir}/${module}
)
# CMakeCache Info
copy(cmake_cache
SRCS ${CMAKE_CURRENT_BINARY_DIR}/CMakeCache.txt
DSTS ${FLUID_INSTALL_DIR})
SRCS ${CMAKE_CURRENT_BINARY_DIR}/CMakeCache.txt
DSTS ${FLUID_INSTALL_DIR})
# This command generates a complete fluid library for both train and inference
add_custom_target(fluid_lib_dist DEPENDS ${fluid_lib_dist_dep})
add_custom_target(fluid_lib_dist DEPENDS ${fluid_lib_dist_dep})
# Following commands generate a inference-only fluid library
# third_party, version.txt and CMakeCache.txt are the same position with ${FLUID_INSTALL_DIR}
copy(third_party DEPS fluid_lib_dist
SRCS ${FLUID_INSTALL_DIR}/third_party ${FLUID_INSTALL_DIR}/CMakeCache.txt
DSTS ${FLUID_INFERENCE_INSTALL_DIR} ${FLUID_INFERENCE_INSTALL_DIR}
)
SRCS ${FLUID_INSTALL_DIR}/third_party ${FLUID_INSTALL_DIR}/CMakeCache.txt
DSTS ${FLUID_INFERENCE_INSTALL_DIR} ${FLUID_INFERENCE_INSTALL_DIR}
)
# only need libpaddle_fluid.so/a and paddle_inference_api.h for inference-only library
# only need libpaddle_fluid.so/a and paddle_*.h for inference-only library
copy(inference_api_lib DEPS fluid_lib_dist
SRCS ${FLUID_INSTALL_DIR}/paddle/fluid/inference/libpaddle_fluid.*
${FLUID_INSTALL_DIR}/paddle/fluid/inference/paddle_inference_api.h
${FLUID_INSTALL_DIR}/paddle/fluid/inference/paddle_*.h
DSTS ${FLUID_INFERENCE_INSTALL_DIR}/paddle/lib ${FLUID_INFERENCE_INSTALL_DIR}/paddle/include
)
......@@ -206,20 +233,20 @@ add_custom_target(inference_lib_dist DEPENDS third_party inference_api_lib)
# paddle fluid version
function(version version_file)
execute_process(
COMMAND ${GIT_EXECUTABLE} log --pretty=format:%H -1
WORKING_DIRECTORY ${PADDLE_SOURCE_DIR}
OUTPUT_VARIABLE PADDLE_GIT_COMMIT)
file(WRITE ${version_file}
"GIT COMMIT ID: ${PADDLE_GIT_COMMIT}\n"
"WITH_MKL: ${WITH_MKL}\n"
"WITH_MKLDNN: ${WITH_MKLDNN}\n"
"WITH_GPU: ${WITH_GPU}\n")
if(WITH_GPU)
file(APPEND ${version_file}
"CUDA version: ${CUDA_VERSION}\n"
"CUDNN version: v${CUDNN_MAJOR_VERSION}\n")
endif()
execute_process(
COMMAND ${GIT_EXECUTABLE} log --pretty=format:%H -1
WORKING_DIRECTORY ${PADDLE_SOURCE_DIR}
OUTPUT_VARIABLE PADDLE_GIT_COMMIT)
file(WRITE ${version_file}
"GIT COMMIT ID: ${PADDLE_GIT_COMMIT}\n"
"WITH_MKL: ${WITH_MKL}\n"
"WITH_MKLDNN: ${WITH_MKLDNN}\n"
"WITH_GPU: ${WITH_GPU}\n")
if (WITH_GPU)
file(APPEND ${version_file}
"CUDA version: ${CUDA_VERSION}\n"
"CUDNN version: v${CUDNN_MAJOR_VERSION}\n")
endif ()
endfunction()
version(${FLUID_INSTALL_DIR}/version.txt)
version(${FLUID_INFERENCE_INSTALL_DIR}/version.txt)
cmake/operators.cmake 0 → 100644
浏览文件 @ bc34d325
set(PART_CUDA_KERNEL_FILES)
function(op_library TARGET)
# op_library is a function to create op library. The interface is same as
# cc_library. But it handle split GPU/CPU code and link some common library
# for ops.
set(cc_srcs)
set(cu_srcs)
set(hip_cu_srcs)
set(miopen_hip_cc_srcs)
set(cu_cc_srcs)
set(cudnn_cu_cc_srcs)
set(CUDNN_FILE)
set(mkldnn_cc_srcs)
set(MKLDNN_FILE)
set(op_common_deps operator op_registry math_function)
set(options "")
set(oneValueArgs "")
set(multiValueArgs SRCS DEPS)
set(pybind_flag 0)
cmake_parse_arguments(op_library "${options}" "${oneValueArgs}"
"${multiValueArgs}" ${ARGN})
list(LENGTH op_library_SRCS op_library_SRCS_len)
if (${op_library_SRCS_len} EQUAL 0)
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${TARGET}.cc)
list(APPEND cc_srcs ${TARGET}.cc)
endif()
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${TARGET}.cu.cc)
list(APPEND cu_cc_srcs ${TARGET}.cu.cc)
endif()
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${TARGET}.cu)
list(APPEND cu_srcs ${TARGET}.cu)
endif()
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${TARGET}.part.cu)
set(PART_CUDA_KERNEL_FILES ${CMAKE_CURRENT_SOURCE_DIR}/${TARGET}.part.cu
${PART_CUDA_KERNEL_FILES} PARENT_SCOPE)
list(APPEND cu_srcs ${CMAKE_CURRENT_SOURCE_DIR}/${TARGET}.part.cu)
endif()
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${TARGET}.hip.cu)
list(APPEND hip_cu_srcs ${TARGET}.hip.cu)
endif()
string(REPLACE "_op" "_cudnn_op" CUDNN_FILE "${TARGET}")
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${CUDNN_FILE}.cu.cc)
list(APPEND cudnn_cu_cc_srcs ${CUDNN_FILE}.cu.cc)
endif()
if(WITH_AMD_GPU)
string(REPLACE "_op" "_miopen_op" MIOPEN_FILE "${TARGET}")
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${MIOPEN_FILE}.hip.cc)
list(APPEND miopen_hip_cc_srcs ${MIOPEN_FILE}.hip.cc)
endif()
endif()
if(WITH_MKLDNN)
string(REPLACE "_op" "_mkldnn_op" MKLDNN_FILE "${TARGET}")
if (EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/${MKLDNN_FILE}.cc)
list(APPEND mkldnn_cc_srcs ${MKLDNN_FILE}.cc)
endif()
endif()
else()
foreach(src ${op_library_SRCS})
if (${src} MATCHES ".*\\.hip.cu$")
list(APPEND hip_cu_srcs ${src})
elseif (${src} MATCHES ".*\\.cu$")
list(APPEND cu_srcs ${src})
elseif(${src} MATCHES ".*_cudnn_op.cu.cc$")
list(APPEND cudnn_cu_cc_srcs ${src})
elseif(WITH_AMD_GPU AND ${src} MATCHES ".*_miopen_op.hip.cc$")
list(APPEND miopen_hip_cc_srcs ${src})
elseif(WITH_MKLDNN AND ${src} MATCHES ".*_mkldnn_op.cc$")
list(APPEND mkldnn_cc_srcs ${src})
elseif(${src} MATCHES ".*\\.cu.cc$")
list(APPEND cu_cc_srcs ${src})
elseif(${src} MATCHES ".*\\.cc$")
list(APPEND cc_srcs ${src})
else()
message(FATAL_ERROR "${TARGET} Source file ${src} should only be .cc or .cu")
endif()
endforeach()
endif()
list(LENGTH cc_srcs cc_srcs_len)
if (${cc_srcs_len} EQUAL 0)
message(FATAL_ERROR "The op library ${TARGET} should contains at least one .cc file")
endif()
if (WIN32)
# remove windows unsupported op, because windows has no nccl, no warpctc such ops.
foreach(windows_unsupport_op "nccl_op" "gen_nccl_id_op" "warpctc_op")
if ("${TARGET}" STREQUAL "${windows_unsupport_op}")
return()
endif()
endforeach()
endif(WIN32)
set(OP_LIBRARY ${TARGET} ${OP_LIBRARY} CACHE INTERNAL "op libs")
list(LENGTH op_library_DEPS op_library_DEPS_len)
if (${op_library_DEPS_len} GREATER 0)
set(DEPS_OPS ${TARGET} ${DEPS_OPS} PARENT_SCOPE)
endif()
if (WITH_GPU)
nv_library(${TARGET} SRCS ${cc_srcs} ${cu_cc_srcs} ${cudnn_cu_cc_srcs} ${mkldnn_cc_srcs} ${cu_srcs} DEPS ${op_library_DEPS}
${op_common_deps})
elseif (WITH_AMD_GPU)
hip_library(${TARGET} SRCS ${cc_srcs} ${hip_cu_srcs} ${miopen_hip_cc_srcs} ${mkldnn_cc_srcs} DEPS ${op_library_DEPS}
${op_common_deps})
else()
cc_library(${TARGET} SRCS ${cc_srcs} ${mkldnn_cc_srcs} DEPS ${op_library_DEPS}
${op_common_deps})
endif()
# Define operators that don't need pybind here.
foreach(manual_pybind_op "compare_op" "logical_op" "nccl_op"
"tensor_array_read_write_op" "tensorrt_engine_op" "conv_fusion_op"
"fusion_transpose_flatten_concat_op")
if ("${TARGET}" STREQUAL "${manual_pybind_op}")
set(pybind_flag 1)
endif()
endforeach()
# The registration of USE_OP, please refer to paddle/fluid/framework/op_registry.h.
# Note that it's enough to just adding one operator to pybind in a *_op.cc file.
# And for detail pybind information, please see generated paddle/pybind/pybind.h.
file(READ ${TARGET}.cc TARGET_CONTENT)
string(REGEX MATCH "REGISTER_OPERATOR\\(.*REGISTER_OPERATOR\\(" multi_register "${TARGET_CONTENT}")
string(REGEX MATCH "REGISTER_OPERATOR\\([a-z0-9_]*," one_register "${multi_register}")
if (one_register STREQUAL "")
string(REPLACE "_op" "" TARGET "${TARGET}")
else ()
string(REPLACE "REGISTER_OPERATOR(" "" TARGET "${one_register}")
string(REPLACE "," "" TARGET "${TARGET}")
endif()
# pybind USE_NO_KERNEL_OP
# HACK: if REGISTER_OP_CPU_KERNEL presents the operator must have kernel
string(REGEX MATCH "REGISTER_OP_CPU_KERNEL" regex_result "${TARGET_CONTENT}")
string(REPLACE "_op" "" TARGET "${TARGET}")
if (${pybind_flag} EQUAL 0 AND regex_result STREQUAL "")
file(APPEND ${pybind_file} "USE_NO_KERNEL_OP(${TARGET});\n")
set(pybind_flag 1)
endif()
# pybind USE_CPU_ONLY_OP
list(LENGTH cu_srcs cu_srcs_len)
list(LENGTH cu_cc_srcs cu_cc_srcs_len)
list(LENGTH mkldnn_cc_srcs mkldnn_cc_srcs_len)
list(LENGTH hip_cu_srcs hip_cu_srcs_len)
list(LENGTH miopen_hip_cc_srcs miopen_hip_cc_srcs_len)
if (${pybind_flag} EQUAL 0 AND ${mkldnn_cc_srcs_len} EQUAL 0 AND ${cu_srcs_len} EQUAL 0 AND ${cu_cc_srcs_len} EQUAL 0 AND
${hip_cu_srcs_len} EQUAL 0 AND ${miopen_hip_cc_srcs_len} EQUAL 0)
file(APPEND ${pybind_file} "USE_CPU_ONLY_OP(${TARGET});\n")
set(pybind_flag 1)
endif()
# pybind USE_OP_DEVICE_KERNEL for CUDNN
list(LENGTH cudnn_cu_cc_srcs cudnn_cu_cc_srcs_len)
if (WITH_GPU AND ${cudnn_cu_cc_srcs_len} GREATER 0)
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(${TARGET}, CUDNN);\n")
endif()
# pybind USE_OP_DEVICE_KERNEL for MIOPEN
if (WITH_AMD_GPU AND ${miopen_hip_cc_srcs_len} GREATER 0)
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(${TARGET}, MIOPEN);\n")
endif()
# pybind USE_OP_DEVICE_KERNEL for MKLDNN
if (WITH_MKLDNN AND ${mkldnn_cc_srcs_len} GREATER 0)
# Append first implemented MKLDNN activation operator
if (${MKLDNN_FILE} STREQUAL "activation_mkldnn_op")
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(relu, MKLDNN);\n")
else()
file(APPEND ${pybind_file} "USE_OP_DEVICE_KERNEL(${TARGET}, MKLDNN);\n")
endif()
endif()
# pybind USE_OP
if (${pybind_flag} EQUAL 0)
# NOTE(*): activation use macro to regist the kernels, set use_op manually.
if(${TARGET} STREQUAL "activation")
file(APPEND ${pybind_file} "USE_OP(relu);\n")
elseif(${TARGET} STREQUAL "fake_dequantize")
file(APPEND ${pybind_file} "USE_OP(fake_dequantize_max_abs);\n")
elseif(${TARGET} STREQUAL "fake_quantize")
file(APPEND ${pybind_file} "USE_OP(fake_quantize_abs_max);\n")
elseif(${TARGET} STREQUAL "tensorrt_engine_op")
message(STATUS "Pybind skips [tensorrt_engine_op], for this OP is only used in inference")
elseif(${TARGET} STREQUAL "fc")
# HACK: fc only have mkldnn and cpu, which would mismatch the cpu only condition
file(APPEND ${pybind_file} "USE_CPU_ONLY_OP(${TARGET});\n")
else()
file(APPEND ${pybind_file} "USE_OP(${TARGET});\n")
endif()
endif()
endfunction()
function(register_operators)
set(options "")
set(oneValueArgs "")
set(multiValueArgs EXCLUDES DEPS)
cmake_parse_arguments(register_operators "${options}" "${oneValueArgs}"
"${multiValueArgs}" ${ARGN})
file(GLOB OPS RELATIVE "${CMAKE_CURRENT_SOURCE_DIR}" "*_op.cc")
string(REPLACE "_mkldnn" "" OPS "${OPS}")
string(REPLACE ".cc" "" OPS "${OPS}")
list(REMOVE_DUPLICATES OPS)
list(LENGTH register_operators_DEPS register_operators_DEPS_len)
foreach(src ${OPS})
list(FIND register_operators_EXCLUDES ${src} _index)
if (${_index} EQUAL -1)
if (${register_operators_DEPS_len} GREATER 0)
op_library(${src} DEPS ${register_operators_DEPS})
else()
op_library(${src})
endif()
endif()
endforeach()
endfunction()
cmake/simd.cmake
浏览文件 @ bc34d325
......@@ -57,41 +57,46 @@ int main()
return 0;
}" SSE3_FOUND)
# Check AVX
set(CMAKE_REQUIRED_FLAGS ${AVX_FLAG})
set(AVX_FOUND_EXITCODE 1 CACHE STRING "Result from TRY_RUN" FORCE)
CHECK_CXX_SOURCE_RUNS("
#include <immintrin.h>
int main()
{
__m256 a = _mm256_set_ps (-1.0f, 2.0f, -3.0f, 4.0f, -1.0f, 2.0f, -3.0f, 4.0f);
__m256 b = _mm256_set_ps (1.0f, 2.0f, 3.0f, 4.0f, 1.0f, 2.0f, 3.0f, 4.0f);
__m256 result = _mm256_add_ps (a, b);
return 0;
}" AVX_FOUND)
# disable AVX by default on windows
if(NOT WIN32)
# Check AVX
set(CMAKE_REQUIRED_FLAGS ${AVX_FLAG})
set(AVX_FOUND_EXITCODE 1 CACHE STRING "Result from TRY_RUN" FORCE)
CHECK_CXX_SOURCE_RUNS("
#include <immintrin.h>
int main()
{
__m256 a = _mm256_set_ps (-1.0f, 2.0f, -3.0f, 4.0f, -1.0f, 2.0f, -3.0f, 4.0f);
__m256 b = _mm256_set_ps (1.0f, 2.0f, 3.0f, 4.0f, 1.0f, 2.0f, 3.0f, 4.0f);
__m256 result = _mm256_add_ps (a, b);
return 0;
}" AVX_FOUND)
# Check AVX 2
set(CMAKE_REQUIRED_FLAGS ${AVX2_FLAG})
set(AVX2_FOUND_EXITCODE 1 CACHE STRING "Result from TRY_RUN" FORCE)
CHECK_CXX_SOURCE_RUNS("
#include <immintrin.h>
int main()
{
__m256i a = _mm256_set_epi32 (-1, 2, -3, 4, -1, 2, -3, 4);
__m256i result = _mm256_abs_epi32 (a);
return 0;
}" AVX2_FOUND)
# Check AVX 2
set(CMAKE_REQUIRED_FLAGS ${AVX2_FLAG})
set(AVX2_FOUND_EXITCODE 1 CACHE STRING "Result from TRY_RUN" FORCE)
CHECK_CXX_SOURCE_RUNS("
#include <immintrin.h>
int main()
{
__m256i a = _mm256_set_epi32 (-1, 2, -3, 4, -1, 2, -3, 4);
__m256i result = _mm256_abs_epi32 (a);
return 0;
}" AVX2_FOUND)
# Check AVX512F
set(CMAKE_REQUIRED_FLAGS ${AVX512F_FLAG})
set(AVX512F_FOUND_EXITCODE 1 CACHE STRING "Result from TRY_RUN" FORCE)
CHECK_CXX_SOURCE_RUNS("
#include <immintrin.h>
int main()
{
__m512i a = _mm512_undefined_epi32();
return 0;
}" AVX512F_FOUND)
# Check AVX512F
set(CMAKE_REQUIRED_FLAGS ${AVX512F_FLAG})
set(AVX512F_FOUND_EXITCODE 1 CACHE STRING "Result from TRY_RUN" FORCE)
CHECK_CXX_SOURCE_RUNS("
#include <immintrin.h>
int main()
{
__m512i a = _mm512_set_epi32 (-1, 2, -3, 4, -1, 2, -3, 4,
13, -5, 6, -7, 9, 2, -6, 3);
__m512i result = _mm512_abs_epi32 (a);
return 0;
}" AVX512F_FOUND)
endif(NOT WIN32)
set(CMAKE_REQUIRED_FLAGS ${CMAKE_REQUIRED_FLAGS_RETAINED})
mark_as_advanced(MMX_FOUND SSE2_FOUND SSE3_FOUND AVX_FOUND AVX2_FOUND AVX512F_FOUND)
cmake/tensorrt.cmake
浏览文件 @ bc34d325
......@@ -34,4 +34,5 @@ if(TENSORRT_FOUND)
"Current TensorRT version is v${TENSORRT_MAJOR_VERSION}. ")
include_directories(${TENSORRT_INCLUDE_DIR})
list(APPEND EXTERNAL_LIBS ${TENSORRT_LIBRARY})
add_definitions(-DPADDLE_WITH_TENSORRT)
endif()
doc/v2/dev/contribute_to_paddle_en.md
浏览文件 @ bc34d325
../../../CONTRIBUTING.md
\ No newline at end of file
../../../CONTRIBUTING.md
paddle/CMakeLists.txt
浏览文件 @ bc34d325
......@@ -24,6 +24,7 @@ if(NOT WITH_FLUID_ONLY)
endif()
add_subdirectory(testing)
set(PYTHON_TESTS_DIR ${PADDLE_BINARY_DIR}/python/paddle/fluid/tests CACHE INTERNAL "python tests directory")
if(NOT MOBILE_INFERENCE AND NOT RPI AND NOT WITH_C_API)
add_subdirectory(fluid)
endif()
paddle/fluid/API.spec
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/CMakeLists.txt
浏览文件 @ bc34d325
......@@ -3,14 +3,9 @@ add_subdirectory(platform)
add_subdirectory(framework)
add_subdirectory(operators)
add_subdirectory(string)
if (NOT WIN32)
add_subdirectory(pybind)
add_subdirectory(recordio)
endif(NOT WIN32)
add_subdirectory(pybind)
if(WITH_INFERENCE)
# NOTE: please add subdirectory inference at last.
add_subdirectory(inference)
add_subdirectory(train)
endif()
# NOTE: please add subdirectory inference at last.
add_subdirectory(inference)
add_subdirectory(train)
paddle/fluid/framework/CMakeLists.txt
浏览文件 @ bc34d325
......@@ -31,11 +31,10 @@ function(windows_symbolic TARGET)
endfunction()
add_subdirectory(ir)
if (NOT WIN32)
add_subdirectory(details)
endif (NOT WIN32)
# ddim lib
proto_library(framework_proto SRCS framework.proto)
proto_library(async_executor_proto SRCS data_feed.proto)
cc_library(ddim SRCS ddim.cc DEPS eigen3 boost)
cc_test(ddim_test SRCS ddim_test.cc DEPS ddim)
......@@ -68,11 +67,7 @@ if(WITH_GPU)
else()
cc_test(mixed_vector_test SRCS mixed_vector_test.cc DEPS place memory device_context tensor)
endif()
if (NOT WIN32)
cc_library(lod_tensor SRCS lod_tensor.cc DEPS ddim place tensor framework_proto recordio version)
else()
cc_library(lod_tensor SRCS lod_tensor.cc DEPS ddim place tensor framework_proto version)
endif (NOT WIN32)
cc_library(lod_tensor SRCS lod_tensor.cc DEPS ddim place tensor framework_proto recordio version)
cc_test(lod_tensor_test SRCS lod_tensor_test.cc DEPS lod_tensor memory)
nv_test(lod_tensor_gpu_test SRCS lod_tensor_test.cu DEPS lod_tensor)
......@@ -122,13 +117,9 @@ cc_test(op_proto_maker_test SRCS op_proto_maker_test.cc DEPS op_proto_maker)
cc_library(op_info SRCS op_info.cc DEPS attribute framework_proto)
cc_library(shape_inference SRCS shape_inference.cc DEPS ddim attribute device_context)
if (NOT WIN32)
cc_library(operator SRCS operator.cc DEPS op_info device_context tensor scope glog
shape_inference data_transform lod_tensor profiler)
else()
cc_library(transfer_scope_cache SRCS transfer_scope_cache.cc DEPS scope framework_proto device_context)
cc_library(operator SRCS operator.cc DEPS op_info device_context tensor scope glog
shape_inference data_transform lod_tensor)
endif(NOT WIN32)
shape_inference data_transform lod_tensor profiler transfer_scope_cache)
cc_test(operator_test SRCS operator_test.cc DEPS operator op_registry device_context)
......@@ -136,44 +127,62 @@ cc_library(version SRCS version.cc)
cc_test(version_test SRCS version_test.cc DEPS version)
cc_library(proto_desc SRCS var_desc.cc op_desc.cc block_desc.cc program_desc.cc DEPS shape_inference op_info operator glog version)
cc_library(ngraph_bridge SRCS ngraph_bridge.cc DEPS operator framework_proto)
if(NOT WIN32)
cc_library(ngraph_operator SRCS ngraph_operator.cc DEPS ngraph_bridge operator op_info device_context tensor scope glog
shape_inference data_transform lod_tensor profiler)
endif(NOT WIN32)
cc_library(op_registry SRCS op_registry.cc DEPS op_proto_maker op_info operator glog proto_desc)
nv_test(op_registry_test SRCS op_registry_test.cc DEPS op_registry)
if (NOT WIN32)
py_proto_compile(framework_py_proto SRCS framework.proto)
py_proto_compile(framework_py_proto SRCS framework.proto data_feed.proto)
# Generate an empty __init__.py to make framework_py_proto as a valid python module.
add_custom_target(framework_py_proto_init ALL COMMAND ${CMAKE_COMMAND} -E touch __init__.py)
add_dependencies(framework_py_proto framework_py_proto_init)
add_custom_command(TARGET framework_py_proto POST_BUILD
COMMAND ${CMAKE_COMMAND} -E make_directory ${PADDLE_BINARY_DIR}/python/paddle/fluid/proto
COMMAND cp *.py ${PADDLE_BINARY_DIR}/python/paddle/fluid/proto/
COMMENT "Copy generated python proto into directory paddle/fluid/proto."
WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR})
if (NOT WIN32)
add_custom_command(TARGET framework_py_proto POST_BUILD
COMMAND ${CMAKE_COMMAND} -E make_directory ${PADDLE_BINARY_DIR}/python/paddle/fluid/proto
COMMAND cp *.py ${PADDLE_BINARY_DIR}/python/paddle/fluid/proto/
COMMENT "Copy generated python proto into directory paddle/fluid/proto."
WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR})
else(NOT WIN32)
string(REPLACE "/" "\\" proto_dstpath "${PADDLE_BINARY_DIR}/python/paddle/fluid/proto/")
add_custom_command(TARGET framework_py_proto POST_BUILD
COMMAND ${CMAKE_COMMAND} -E make_directory ${PADDLE_BINARY_DIR}/python/paddle/fluid/proto
COMMAND copy /Y *.py ${proto_dstpath}
COMMENT "Copy generated python proto into directory paddle/fluid/proto."
WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR})
endif(NOT WIN32)
cc_library(lod_rank_table SRCS lod_rank_table.cc DEPS lod_tensor)
cc_library(feed_fetch_method SRCS feed_fetch_method.cc DEPS lod_tensor scope glog)
cc_library(variable_helper SRCS variable_helper.cc DEPS lod_tensor)
cc_library(naive_executor SRCS naive_executor.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method graph_to_program_pass)
cc_library(naive_executor SRCS naive_executor.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method graph_to_program_pass variable_helper)
if(WITH_DISTRIBUTE)
cc_library(executor SRCS executor.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method sendrecvop_grpc cares grpc++_unsecure grpc_unsecure gpr graph_to_program_pass)
cc_library(executor SRCS executor.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method sendrecvop_grpc cares grpc++_unsecure grpc_unsecure gpr graph_to_program_pass variable_helper)
set(DISTRIBUTE_COMPILE_FLAGS "-Wno-non-virtual-dtor -Wno-error=non-virtual-dtor -Wno-error=delete-non-virtual-dtor")
set_source_files_properties(executor.cc PROPERTIES COMPILE_FLAGS ${DISTRIBUTE_COMPILE_FLAGS})
else()
cc_library(executor SRCS executor.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method graph_to_program_pass)
if(NOT WIN32)
cc_library(executor SRCS executor.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method graph_to_program_pass ngraph_operator variable_helper)
else(NOT WIN32)
cc_library(executor SRCS executor.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method graph_to_program_pass variable_helper)
endif(NOT WIN32)
cc_test(test_naive_executor SRCS naive_executor_test.cc DEPS naive_executor elementwise_add_op)
endif()
if (NOT WIN32)
cc_library(parallel_executor SRCS parallel_executor.cc DEPS
threaded_ssa_graph_executor scope_buffered_ssa_graph_executor
graph build_strategy
fast_threaded_ssa_graph_executor)
endif() # NOT WIN32
fast_threaded_ssa_graph_executor variable_helper)
cc_library(async_executor SRCS async_executor.cc data_feed.cc data_feed_factory.cc executor_thread_worker.cc DEPS op_registry device_context scope framework_proto glog lod_rank_table feed_fetch_method graph_to_program_pass async_executor_proto variable_helper)
cc_test(data_feed_test SRCS data_feed_test.cc DEPS async_executor)
cc_library(prune SRCS prune.cc DEPS framework_proto)
cc_test(prune_test SRCS prune_test.cc DEPS op_info prune recurrent_op device_context)
cc_test(var_type_inference_test SRCS var_type_inference_test.cc DEPS op_registry
......@@ -189,3 +198,6 @@ cc_test(tuple_test SRCS tuple_test.cc )
if (NOT WIN32)
cc_test(rw_lock_test SRCS rw_lock_test.cc)
endif (NOT WIN32)
cc_library(dlpack_tensor SRCS dlpack_tensor.cc DEPS tensor dlpack)
cc_test(dlpack_tensor_test SRCS dlpack_tensor_test.cc DEPS dlpack_tensor glog)
paddle/fluid/framework/async_executor.cc 0 → 100644
浏览文件 @ bc34d325
/* 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. */
#include "paddle/fluid/framework/async_executor.h"
#include "google/protobuf/io/zero_copy_stream_impl.h"
#include "google/protobuf/message.h"
#include "google/protobuf/text_format.h"
#include "gflags/gflags.h"
#include "paddle/fluid/framework/data_feed_factory.h"
#include "paddle/fluid/framework/executor_thread_worker.h"
#include "paddle/fluid/framework/feed_fetch_method.h"
#include "paddle/fluid/framework/feed_fetch_type.h"
#include "paddle/fluid/framework/lod_rank_table.h"
#include "paddle/fluid/framework/lod_tensor_array.h"
#include "paddle/fluid/framework/op_registry.h"
#include "paddle/fluid/framework/reader.h"
#include "paddle/fluid/inference/io.h"
#include "paddle/fluid/platform/place.h"
#include "paddle/fluid/pybind/pybind.h"
namespace paddle {
namespace framework {
AsyncExecutor::AsyncExecutor(Scope* scope, const platform::Place& place)
: root_scope_(scope), place_(place) {}
void AsyncExecutor::CreateThreads(
ExecutorThreadWorker* worker, const ProgramDesc& main_program,
const std::shared_ptr<DataFeed>& reader,
const std::vector<std::string>& fetch_var_names, Scope* root_scope,
const int thread_index, const bool debug) {
worker->SetThreadId(thread_index);
worker->SetDebug(debug);
worker->SetRootScope(root_scope);
worker->CreateThreadResource(main_program, place_);
worker->SetDataFeed(reader);
worker->SetFetchVarNames(fetch_var_names);
worker->BindingDataFeedMemory();
}
void PrepareReaders(std::vector<std::shared_ptr<DataFeed>>& readers, // NOLINT
const int thread_num, const DataFeedDesc& data_feed_desc,
const std::vector<std::string>& filelist) {
readers.resize(thread_num);
for (size_t i = 0; i < readers.size(); ++i) {
readers[i] = DataFeedFactory::CreateDataFeed(data_feed_desc.name());
readers[i]->Init(data_feed_desc); // set batch_size and queue_size here
}
readers[0]->SetFileList(filelist);
}
void AsyncExecutor::RunFromFile(const ProgramDesc& main_program,
const std::string& data_feed_desc_str,
const std::vector<std::string>& filelist,
const int thread_num,
const std::vector<std::string>& fetch_var_names,
const bool debug) {
std::vector<std::thread> threads;
auto& block = main_program.Block(0);
for (auto var_name : fetch_var_names) {
auto var_desc = block.FindVar(var_name);
auto shapes = var_desc->GetShape();
PADDLE_ENFORCE(shapes[shapes.size() - 1] == 1,
"var %s: Fetched var has wrong shape, "
"only variables with the last dimension size 1 supported",
var_name);
}
DataFeedDesc data_feed_desc;
google::protobuf::TextFormat::ParseFromString(data_feed_desc_str,
&data_feed_desc);
int actual_thread_num = thread_num;
int file_cnt = filelist.size();
PADDLE_ENFORCE(file_cnt > 0, "File list cannot be empty");
if (actual_thread_num > file_cnt) {
VLOG(1) << "Thread num = " << thread_num << ", file num = " << file_cnt
<< ". Changing thread_num = " << file_cnt;
actual_thread_num = file_cnt;
}
/*
readerDesc: protobuf description for reader initlization
argument: class_name, batch_size, use_slot, queue_size, buffer_size,
padding_index
reader:
1) each thread has a reader, reader will read input data and
put it into input queue
2) each reader has a Next() iterface, that can fetch an instance
from the input queue
*/
// todo: should be factory method for creating datafeed
std::vector<std::shared_ptr<DataFeed>> readers;
PrepareReaders(readers, actual_thread_num, data_feed_desc, filelist);
std::vector<std::shared_ptr<ExecutorThreadWorker>> workers;
workers.resize(actual_thread_num);
for (auto& worker : workers) {
worker.reset(new ExecutorThreadWorker);
}
// prepare thread resource here
for (int thidx = 0; thidx < actual_thread_num; ++thidx) {
CreateThreads(workers[thidx].get(), main_program, readers[thidx],
fetch_var_names, root_scope_, thidx, debug);
}
// start executing ops in multiple threads
for (int thidx = 0; thidx < actual_thread_num; ++thidx) {
threads.push_back(
std::thread(&ExecutorThreadWorker::TrainFiles, workers[thidx].get()));
}
for (auto& th : threads) {
th.join();
}
root_scope_->DropKids();
return;
}
} // einit_modelnd namespace framework
} // end namespace paddle
paddle/fluid/inference/analysis/tensorrt_subgraph_pass.h → paddle/fluid/framework/async_executor.h
浏览文件 @ bc34d325
/* 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. */
#pragma once
#include <string>
#include "paddle/fluid/inference/analysis/analysis_pass.h"
#include "paddle/fluid/inference/analysis/node.h"
#include "paddle/fluid/inference/analysis/subgraph_splitter.h"
namespace paddle {
namespace inference {
namespace analysis {
/*
* Parse the graph and replace TensorRT supported nodes with SubGraphNode
*/
class TensorRTSubGraphPass : public DataFlowGraphPass {
public:
// Tell whether to transform a sub-graph into TensorRT.
using NodeInsideSubgraphTeller = SubGraphFuse::NodeInsideSubgraphTeller;
explicit TensorRTSubGraphPass(const NodeInsideSubgraphTeller& teller);
bool Initialize(Argument* argument) override {
argument_ = argument;
return true;
}
// This class get a sub-graph as input and determine whether to transform this
// sub-graph into TensorRT.
void Run(DataFlowGraph* graph) override;
bool Finalize() override { return true; }
std::string repr() const override { return "tensorrt-sub-graph"; }
std::string description() const override { return "tensorrt sub graph pass"; }
private:
NodeInsideSubgraphTeller node_inside_subgraph_teller_;
Argument* argument_;
};
} // namespace analysis
} // namespace inference
} // namespace paddle
/* 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. */
#pragma once
#include <map>
#include <memory>
#include <mutex> // NOLINT
#include <set>
#include <string>
#include <thread> // NOLINT
#include <typeinfo>
#include <vector>
#include "paddle/fluid/framework/data_feed.pb.h"
#include "paddle/fluid/framework/executor.h"
#include "paddle/fluid/framework/executor_thread_worker.h"
#include "paddle/fluid/framework/program_desc.h"
#include "paddle/fluid/framework/scope.h"
namespace paddle {
namespace framework {
class AsyncExecutor {
public:
AsyncExecutor(Scope* scope, const platform::Place& place);
virtual ~AsyncExecutor() {}
void RunFromFile(const ProgramDesc& main_program,
const std::string& data_feed_desc_str,
const std::vector<std::string>& filelist,
const int thread_num,
const std::vector<std::string>& fetch_names,
const bool debug = false);
private:
void CreateThreads(ExecutorThreadWorker* worker,
const ProgramDesc& main_program,
const std::shared_ptr<DataFeed>& reader,
const std::vector<std::string>& fetch_var_names,
Scope* root_scope, const int thread_index,
const bool debug);
public:
Scope* root_scope_;
platform::Place place_;
};
} // namespace framework
} // namespace paddle
paddle/fluid/framework/attribute.cc
浏览文件 @ bc34d325
......@@ -64,6 +64,13 @@ Attribute GetAttrValue(const proto::OpDesc::Attr& attr_desc) {
case proto::AttrType::LONG: {
return attr_desc.l();
}
case proto::AttrType::LONGS: {
std::vector<int64_t> val(attr_desc.longs_size());
for (int i = 0; i < attr_desc.longs_size(); ++i) {
val[i] = attr_desc.longs(i);
}
return val;
}
default:
PADDLE_THROW("Unsupport attr type %d", attr_desc.type());
}
......
paddle/fluid/framework/attribute.h
浏览文件 @ bc34d325
......@@ -26,6 +26,113 @@ limitations under the License. */
namespace paddle {
namespace framework {
template <typename T>
struct ExtractAttribute {
explicit ExtractAttribute(const std::string& attr_name)
: attr_name_(attr_name) {}
T* operator()(Attribute& attr) const {
T* attr_value = nullptr;
try {
attr_value = &boost::get<T>(attr);
} catch (boost::bad_get& bad_get) {
PADDLE_THROW("Cannot get attribute %s by type %s, its type is %s",
attr_name_, paddle::platform::demangle(typeid(T).name()),
paddle::platform::demangle(attr.type().name()));
}
return attr_value;
}
const std::string& attr_name_;
};
// special handle bool
// FIXME(yuyang18): Currently we cast bool into int in python binding. It is
// hard to change the logic there. In another way, we should correct handle
// if the user set `some_flag=1`.
//
// FIX ME anytime if there is a better solution.
template <>
struct ExtractAttribute<bool> {
explicit ExtractAttribute(const std::string& attr_name)
: attr_name_(attr_name) {}
bool* operator()(Attribute& attr) const {
if (attr.type() == typeid(int)) { // NOLINT
int val = boost::get<int>(attr);
attr = static_cast<bool>(val);
} else if (attr.type() == typeid(float)) { // NOLINT
float val = boost::get<float>(attr);
attr = static_cast<bool>(val);
}
bool* attr_value = nullptr;
try {
attr_value = &boost::get<bool>(attr);
} catch (boost::bad_get& bad_get) {
PADDLE_THROW("Cannot get attribute %s by type bool, its type is %s",
attr_name_, paddle::platform::demangle(attr.type().name()));
}
return attr_value;
}
const std::string& attr_name_;
};
template <>
struct ExtractAttribute<int64_t> {
explicit ExtractAttribute(const std::string& attr_name)
: attr_name_(attr_name) {}
int64_t* operator()(Attribute& attr) const {
if (attr.type() == typeid(int)) { // NOLINT
int val = boost::get<int>(attr);
attr = static_cast<int64_t>(val);
} else if (attr.type() == typeid(float)) { // NOLINT
int val = boost::get<float>(attr);
attr = static_cast<int64_t>(val);
}
int64_t* attr_value = nullptr;
try {
attr_value = &boost::get<int64_t>(attr);
} catch (boost::bad_get& bad_get) {
PADDLE_THROW("Cannot get attribute %s by type int64_t, its type is %s",
attr_name_, paddle::platform::demangle(attr.type().name()));
}
return attr_value;
}
const std::string& attr_name_;
};
template <>
struct ExtractAttribute<std::vector<int64_t>> {
explicit ExtractAttribute(const std::string& attr_name)
: attr_name_(attr_name) {}
std::vector<int64_t>* operator()(Attribute& attr) const {
if (attr.type() == typeid(std::vector<int>)) { // NOLINT
std::vector<int> val = boost::get<std::vector<int>>(attr);
std::vector<int64_t> vec(val.begin(), val.end());
attr = vec;
} else if (attr.type() == typeid(std::vector<float>)) { // NOLINT
std::vector<float> val = boost::get<std::vector<float>>(attr);
std::vector<int64_t> vec(val.begin(), val.end());
attr = vec;
}
std::vector<int64_t>* attr_value = nullptr;
try {
attr_value = &boost::get<std::vector<int64_t>>(attr);
} catch (boost::bad_get& bad_get) {
PADDLE_THROW("Cannot get attribute %s by type int64_t, its type is %s",
attr_name_, paddle::platform::demangle(attr.type().name()));
}
return attr_value;
}
const std::string& attr_name_;
};
template <typename T>
inline proto::AttrType AttrTypeID() {
Attribute tmp = T();
......@@ -42,7 +149,11 @@ class AttrReader {
inline const T& Get(const std::string& name) const {
PADDLE_ENFORCE(attrs_.count(name) != 0, "%s should be in AttributeMap",
name);
return boost::get<T>(attrs_.at(name));
Attribute& attr = const_cast<Attribute&>(attrs_.at(name));
ExtractAttribute<T> extract_attr(name);
T* attr_value = extract_attr(attr);
return *attr_value;
}
private:
......@@ -82,7 +193,7 @@ class DefaultValueSetter {
public:
explicit DefaultValueSetter(T default_value)
: default_value_(default_value) {}
void operator()(T& value) const { value = default_value_; }
void operator()(T& value) const { value = default_value_; } // NOLINT
private:
T default_value_;
......@@ -117,84 +228,6 @@ class EnumInContainer {
std::unordered_set<T> container_;
};
template <typename T>
struct ExtractAttribute {
explicit ExtractAttribute(const std::string& attr_name)
: attr_name_(attr_name) {}
T* operator()(Attribute& attr) const {
T* attr_value = nullptr;
try {
attr_value = &boost::get<T>(attr);
} catch (boost::bad_get& bad_get) {
PADDLE_THROW("Cannot get attribute %s by type %s, its type is %s",
attr_name_, paddle::platform::demangle(typeid(T).name()),
paddle::platform::demangle(attr.type().name()));
}
return attr_value;
}
const std::string& attr_name_;
};
// special handle bool
// FIXME(yuyang18): Currently we cast bool into int in python binding. It is
// hard to change the logic there. In another way, we should correct handle
// if the user set `some_flag=1`.
//
// FIX ME anytime if there is a better solution.
template <>
struct ExtractAttribute<bool> {
explicit ExtractAttribute(const std::string& attr_name)
: attr_name_(attr_name) {}
bool* operator()(Attribute& attr) const {
if (attr.type() == typeid(int)) { // NOLINT
int val = boost::get<int>(attr);
attr = static_cast<bool>(val);
} else if (attr.type() == typeid(float)) { // NOLINT
float val = boost::get<float>(attr);
attr = static_cast<bool>(val);
}
bool* attr_value = nullptr;
try {
attr_value = &boost::get<bool>(attr);
} catch (boost::bad_get& bad_get) {
PADDLE_THROW("Cannot get attribute %s by type bool, its type is %s",
attr_name_, paddle::platform::demangle(attr.type().name()));
}
return attr_value;
}
const std::string& attr_name_;
};
template <>
struct ExtractAttribute<int64_t> {
explicit ExtractAttribute(const std::string& attr_name)
: attr_name_(attr_name) {}
int64_t* operator()(Attribute& attr) const {
if (attr.type() == typeid(int)) { // NOLINT
int val = boost::get<int>(attr);
attr = static_cast<int64_t>(val);
} else if (attr.type() == typeid(float)) { // NOLINT
int val = boost::get<float>(attr);
attr = static_cast<int64_t>(val);
}
int64_t* attr_value = nullptr;
try {
attr_value = &boost::get<int64_t>(attr);
} catch (boost::bad_get& bad_get) {
PADDLE_THROW("Cannot get attribute %s by type int64_t, its type is %s",
attr_name_, paddle::platform::demangle(attr.type().name()));
}
return attr_value;
}
const std::string& attr_name_;
};
// check whether a certain attribute fit its limits
// an attribute can have more than one limits
template <typename T>
......@@ -235,7 +268,7 @@ class TypedAttrChecker {
return *this;
}
void operator()(AttributeMap& attr_map) const {
void operator()(AttributeMap& attr_map) const { // NOLINT
if (!attr_map.count(attr_name_)) {
// user do not set this attr
PADDLE_ENFORCE(!default_value_setter_.empty(),
......@@ -271,7 +304,7 @@ class OpAttrChecker {
return *(checker.target<TypedAttrChecker<T>>());
}
void Check(AttributeMap& attr_map) const {
void Check(AttributeMap& attr_map) const { // NOLINT
for (const auto& checker : attr_checkers_) {
checker(attr_map);
}
......
paddle/fluid/framework/data_device_transform_test.cu
浏览文件 @ bc34d325
......@@ -17,7 +17,7 @@ limitations under the License. */
#include "paddle/fluid/framework/lod_tensor.h"
#include "paddle/fluid/framework/op_info.h"
#include "paddle/fluid/framework/op_registry.h"
#include "paddle/fluid/operators/elementwise_op_function.h"
#include "paddle/fluid/operators/elementwise/elementwise_op_function.h"
#include "paddle/fluid/operators/math/math_function.h"
#include "paddle/fluid/platform/device_context.h"
#include "paddle/fluid/platform/init.h"
......
paddle/fluid/framework/data_feed.cc 0 → 100644
浏览文件 @ bc34d325
/* 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. */
#include "google/protobuf/io/zero_copy_stream_impl.h"
#include "google/protobuf/message.h"
#include "google/protobuf/text_format.h"
#include "gflags/gflags.h"
#include "paddle/fluid/framework/data_feed.h"
#include "paddle/fluid/framework/feed_fetch_method.h"
#include "paddle/fluid/framework/feed_fetch_type.h"
namespace paddle {
namespace framework {
std::vector<std::string> DataFeed::filelist_;
size_t DataFeed::file_idx_;
std::mutex DataFeed::mutex_for_pick_file_;
bool DataFeed::finish_set_filelist_;
void DataFeed::AddFeedVar(Variable* var, const std::string& name) {
CheckInit();
for (size_t i = 0; i < use_slots_.size(); ++i) {
if (name == use_slots_[i]) {
if (use_slots_is_dense_[i]) {
feed_vec_[i] = MixTensor(var->GetMutable<Tensor>());
} else {
feed_vec_[i] = MixTensor(var->GetMutable<LoDTensor>());
}
}
}
}
bool DataFeed::SetFileList(const std::vector<std::string>& files) {
std::unique_lock<std::mutex> lock(mutex_for_pick_file_);
CheckInit();
if (finish_set_filelist_) {
VLOG(3) << "info: you have set the filelist.";
return false;
}
PADDLE_ENFORCE(files.size(), "You have set an empty filelist.");
filelist_.assign(files.begin(), files.end());
file_idx_ = 0;
finish_set_filelist_ = true;
return true;
}
void DataFeed::SetBatchSize(int batch_size) {
PADDLE_ENFORCE(batch_size > 0, "Illegal batch size: %d.", batch_size);
default_batch_size_ = batch_size;
}
bool DataFeed::PickOneFile(std::string* filename) {
std::unique_lock<std::mutex> lock(mutex_for_pick_file_);
if (file_idx_ == filelist_.size()) {
return false;
}
*filename = filelist_[file_idx_++];
return true;
}
void DataFeed::CheckInit() {
PADDLE_ENFORCE(finish_init_, "Initialization did not succeed.");
}
void DataFeed::CheckSetFileList() {
PADDLE_ENFORCE(finish_set_filelist_, "Set filelist did not succeed.");
}
void DataFeed::CheckStart() {
PADDLE_ENFORCE(finish_start_, "Datafeed has not started running yet.");
}
template <typename T>
void PrivateQueueDataFeed<T>::SetQueueSize(int queue_size) {
PADDLE_ENFORCE(queue_size > 0, "Illegal queue size: %d.", queue_size);
queue_size_ = queue_size;
queue_ = std::unique_ptr<paddle::operators::reader::BlockingQueue<T>>(
new paddle::operators::reader::BlockingQueue<T>(queue_size_));
}
template <typename T>
bool PrivateQueueDataFeed<T>::Start() {
CheckSetFileList();
read_thread_ = std::thread(&PrivateQueueDataFeed::ReadThread, this);
read_thread_.detach();
finish_start_ = true;
return true;
}
template <typename T>
void PrivateQueueDataFeed<T>::ReadThread() {
std::string filename;
while (PickOneFile(&filename)) {
file_.open(filename.c_str()); // is_text_feed
PADDLE_ENFORCE(file_.good(), "Open file<%s> fail.", filename.c_str());
T instance;
while (ParseOneInstance(&instance)) {
queue_->Send(instance);
}
file_.close();
}
queue_->Close();
}
template <typename T>
int PrivateQueueDataFeed<T>::Next() {
CheckStart();
int index = 0;
T instance;
T ins_vec;
while (index < default_batch_size_) {
if (!queue_->Receive(&instance)) {
break;
}
AddInstanceToInsVec(&ins_vec, instance, index++);
}
batch_size_ = index;
if (batch_size_ != 0) {
PutToFeedVec(ins_vec);
}
return batch_size_;
}
#ifdef _WIN32
template class PrivateQueueDataFeed<std::vector<MultiSlotType>>;
#endif
void MultiSlotDataFeed::Init(
const paddle::framework::DataFeedDesc& data_feed_desc) {
finish_init_ = false;
finish_set_filelist_ = false;
finish_start_ = false;
PADDLE_ENFORCE(data_feed_desc.has_multi_slot_desc(),
"Multi_slot_desc has not been set.");
paddle::framework::MultiSlotDesc multi_slot_desc =
data_feed_desc.multi_slot_desc();
SetBatchSize(data_feed_desc.batch_size());
SetQueueSize(data_feed_desc.batch_size());
size_t all_slot_num = multi_slot_desc.slots_size();
all_slots_.resize(all_slot_num);
all_slots_type_.resize(all_slot_num);
use_slots_index_.resize(all_slot_num);
use_slots_.clear();
use_slots_is_dense_.clear();
for (size_t i = 0; i < all_slot_num; ++i) {
const auto& slot = multi_slot_desc.slots(i);
all_slots_[i] = slot.name();
all_slots_type_[i] = slot.type();
use_slots_index_[i] = slot.is_used() ? use_slots_.size() : -1;
if (slot.is_used()) {
use_slots_.push_back(all_slots_[i]);
use_slots_is_dense_.push_back(slot.is_dense());
}
}
feed_vec_.resize(use_slots_.size());
finish_init_ = true;
}
bool MultiSlotDataFeed::CheckFile(const char* filename) {
CheckInit(); // get info of slots
std::ifstream fin(filename);
if (!fin.good()) {
VLOG(1) << "error: open file<" << filename << "> fail";
return false;
}
std::string line;
int instance_cout = 0;
std::string all_slots_alias = "";
for (const auto& alias : all_slots_) {
all_slots_alias += alias + " ";
}
std::string use_slots_alias = "";
for (const auto& alias : use_slots_) {
use_slots_alias += alias + " ";
}
VLOG(3) << "total slots num: " << all_slots_.size();
VLOG(3) << "total slots alias: " << all_slots_alias;
VLOG(3) << "used slots num: " << use_slots_.size();
VLOG(3) << "used slots alias: " << use_slots_alias;
while (getline(fin, line)) {
++instance_cout;
const char* str = line.c_str();
char* endptr = const_cast<char*>(str);
int len = line.length();
for (size_t i = 0; i < all_slots_.size(); ++i) {
int num = strtol(endptr, &endptr, 10);
if (num < 0) {
VLOG(1) << "error: the number of ids is a negative number: " << num;
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
} else if (num == 0) {
VLOG(1)
<< "error: the number of ids can not be zero, you need "
"padding it in data generator; or if there is something wrong"
" with the data, please check if the data contains unresolvable "
"characters.";
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
} else if (errno == ERANGE || num > INT_MAX) {
VLOG(1) << "error: the number of ids greater than INT_MAX";
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
}
if (all_slots_type_[i] == "float") {
for (int i = 0; i < num; ++i) {
strtof(endptr, &endptr);
if (errno == ERANGE) {
VLOG(1) << "error: the value is out of the range of "
"representable values for float";
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
}
if (i + 1 != num && endptr - str == len) {
VLOG(1) << "error: there is a wrong with the number of ids.";
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
}
}
} else if (all_slots_type_[i] == "uint64") {
for (int i = 0; i < num; ++i) {
strtoull(endptr, &endptr, 10);
if (errno == ERANGE) {
VLOG(1) << "error: the value is out of the range of "
"representable values for uint64_t";
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
}
if (i + 1 != num && endptr - str == len) {
VLOG(1) << "error: there is a wrong with the number of ids.";
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
}
}
} else {
VLOG(1) << "error: this type<" << all_slots_type_[i]
<< "> is not supported";
return false;
}
}
if (endptr - str != len) {
VLOG(1) << "error: there is some data at the end of the line.";
VLOG(1) << "please check line<" << instance_cout << "> in file<"
<< filename << ">";
return false;
}
}
VLOG(3) << "instances cout: " << instance_cout;
VLOG(3) << "The file format is correct";
return true;
}
bool MultiSlotDataFeed::ParseOneInstance(std::vector<MultiSlotType>* instance) {
std::string line;
if (getline(file_, line)) {
int use_slots_num = use_slots_.size();
instance->resize(use_slots_num);
// parse line
const char* str = line.c_str();
char* endptr = const_cast<char*>(str);
int pos = 0;
for (size_t i = 0; i < use_slots_index_.size(); ++i) {
int idx = use_slots_index_[i];
int num = strtol(&str[pos], &endptr, 10);
PADDLE_ENFORCE(
num,
"The number of ids can not be zero, you need padding "
"it in data generator; or if there is something wrong with "
"the data, please check if the data contains unresolvable "
"characters.\nplease check this error line: %s",
str);
if (idx != -1) {
(*instance)[idx].Init(all_slots_type_[i]);
if ((*instance)[idx].GetType()[0] == 'f') { // float
for (int j = 0; j < num; ++j) {
float feasign = strtof(endptr, &endptr);
(*instance)[idx].AddValue(feasign);
}
} else if ((*instance)[idx].GetType()[0] == 'u') { // uint64
for (int j = 0; j < num; ++j) {
uint64_t feasign = (uint64_t)strtoull(endptr, &endptr, 10);
(*instance)[idx].AddValue(feasign);
}
}
pos = endptr - str;
} else {
for (int j = 0; j <= num; ++j) {
pos = line.find_first_of(' ', pos + 1);
}
}
}
} else {
return false;
}
return true;
}
void MultiSlotDataFeed::AddInstanceToInsVec(
std::vector<MultiSlotType>* ins_vec,
const std::vector<MultiSlotType>& instance, int index) {
if (index == 0) {
ins_vec->resize(instance.size());
for (size_t i = 0; i < instance.size(); ++i) {
(*ins_vec)[i].Init(instance[i].GetType());
(*ins_vec)[i].InitOffset();
}
}
for (size_t i = 0; i < instance.size(); ++i) {
(*ins_vec)[i].AddIns(instance[i]);
}
}
void MultiSlotDataFeed::PutToFeedVec(
const std::vector<MultiSlotType>& ins_vec) {
for (size_t i = 0; i < use_slots_.size(); ++i) {
const auto& type = ins_vec[i].GetType();
const auto& offset = ins_vec[i].GetOffset();
int total_instance = static_cast<int>(offset.back());
if (type[0] == 'f') { // float
const auto& feasign = ins_vec[i].GetFloatData();
if (feed_vec_[i].IsDense()) {
int size_in_each_batch = total_instance / batch_size_;
float* tensor_ptr = feed_vec_[i].GetTensor()->mutable_data<float>(
{batch_size_, size_in_each_batch}, platform::CPUPlace());
memcpy(tensor_ptr, &feasign[0], total_instance * sizeof(float));
} else {
float* tensor_ptr = feed_vec_[i].GetLoDTensor()->mutable_data<float>(
{total_instance, 1}, platform::CPUPlace());
memcpy(tensor_ptr, &feasign[0], total_instance * sizeof(float));
LoD data_lod{offset};
feed_vec_[i].GetLoDTensor()->set_lod(data_lod);
}
} else if (type[0] == 'u') { // uint64
// no uint64_t type in paddlepaddle
const auto& feasign = ins_vec[i].GetUint64Data();
if (feed_vec_[i].IsDense()) {
int size_in_each_batch = total_instance / batch_size_;
int64_t* tensor_ptr = feed_vec_[i].GetTensor()->mutable_data<int64_t>(
{batch_size_, size_in_each_batch}, platform::CPUPlace());
memcpy(tensor_ptr, &feasign[0], total_instance * sizeof(int64_t));
} else {
int64_t* tensor_ptr =
feed_vec_[i].GetLoDTensor()->mutable_data<int64_t>(
{total_instance, 1}, platform::CPUPlace());
memcpy(tensor_ptr, &feasign[0], total_instance * sizeof(int64_t));
LoD data_lod{offset};
feed_vec_[i].GetLoDTensor()->set_lod(data_lod);
}
}
}
}
} // namespace framework
} // namespace paddle
paddle/fluid/framework/data_feed.h 0 → 100644
浏览文件 @ bc34d325
/* 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. */
#pragma once
#include <fstream>
#include <memory>
#include <mutex> // NOLINT
#include <string>
#include <thread> // NOLINT
#include <vector>
#include "paddle/fluid/framework/data_feed.pb.h"
#include "paddle/fluid/framework/lod_tensor.h"
#include "paddle/fluid/framework/reader.h"
#include "paddle/fluid/framework/variable.h"
#include "paddle/fluid/operators/reader/blocking_queue.h"
namespace paddle {
namespace framework {
// Pack Tensor type and LoDTensor type into MixTensor type, in order
// to record either Tensor or LoDTensor information at the same time.
class MixTensor {
public:
MixTensor() {}
explicit MixTensor(LoDTensor* lodtensor) {
is_dense_ = false;
lodtensor_ = lodtensor;
}
explicit MixTensor(Tensor* tensor) {
is_dense_ = true;
tensor_ = tensor;
}
bool IsDense() { return is_dense_; }
LoDTensor* GetLoDTensor() {
PADDLE_ENFORCE(!is_dense_, "Let a dense var return a LoDTensor ptr.");
return lodtensor_;
}
Tensor* GetTensor() {
PADDLE_ENFORCE(is_dense_, "Let a sparse var return a Tensor ptr.");
return tensor_;
}
private:
bool is_dense_;
LoDTensor* lodtensor_;
Tensor* tensor_;
};
// DataFeed is the base virtual class for all ohther DataFeeds.
// It is used to read files and parse the data for subsequent trainer.
// Example:
// DataFeed* reader =
// paddle::framework::DataFeedFactory::CreateDataFeed(data_feed_name);
// reader->Init(data_feed_desc); // data_feed_desc is a protobuf object
// reader->SetFileList(filelist);
// const std::vector<std::string> & use_slot_alias =
// reader->GetUseSlotAlias();
// for (auto name: use_slot_alias){ // for binding memory
// reader->AddFeedVar(scope->Var(name), name);
// }
// reader->Start();
// while (reader->Next()) {
// // trainer do something
// }
class DataFeed {
public:
DataFeed() {}
virtual ~DataFeed() {}
virtual void Init(const paddle::framework::DataFeedDesc& data_feed_desc) = 0;
virtual bool CheckFile(const char* filename) {
PADDLE_THROW("This function(CheckFile) is not implemented.");
}
// Set filelist for DataFeed.
// Pay attention that it must init all readers before call this function.
// Otherwise, Init() function will init finish_set_filelist_ flag.
virtual bool SetFileList(const std::vector<std::string>& files);
virtual bool Start() = 0;
// The trainer calls the Next() function, and the DataFeed will load a new
// batch to the feed_vec. The return value of this function is the batch
// size of the current batch.
virtual int Next() = 0;
// Get all slots' alias which defined in protofile
virtual const std::vector<std::string>& GetAllSlotAlias() {
return all_slots_;
}
// Get used slots' alias which defined in protofile
virtual const std::vector<std::string>& GetUseSlotAlias() {
return use_slots_;
}
// This function is used for binding feed_vec memory
virtual void AddFeedVar(Variable* var, const std::string& name);
protected:
// The following three functions are used to check if it is executed in this
// order:
// Init() -> SetFileList() -> Start() -> Next()
virtual void CheckInit();
virtual void CheckSetFileList();
virtual void CheckStart();
virtual void SetBatchSize(
int batch); // batch size will be set in Init() function
// This function is used to pick one file from the global filelist(thread
// safe).
virtual bool PickOneFile(std::string* filename);
static std::vector<std::string> filelist_;
static size_t file_idx_;
static std::mutex mutex_for_pick_file_;
// the alias of used slots, and its order is determined by
// data_feed_desc(proto object)
std::vector<std::string> use_slots_;
std::vector<bool> use_slots_is_dense_;
// the alias of all slots, and its order is determined by data_feed_desc(proto
// object)
std::vector<std::string> all_slots_;
std::vector<std::string> all_slots_type_;
std::vector<int>
use_slots_index_; // -1: not used; >=0: the index of use_slots_
// The data read by DataFeed will be stored here
std::vector<MixTensor> feed_vec_;
// the batch size defined by user
int default_batch_size_;
// current batch size
int batch_size_;
bool finish_init_;
static bool finish_set_filelist_;
bool finish_start_;
};
// PrivateQueueDataFeed is the base virtual class for ohther DataFeeds.
// It use a read-thread to read file and parse data to a private-queue
// (thread level), and get data from this queue when trainer call Next().
template <typename T>
class PrivateQueueDataFeed : public DataFeed {
public:
PrivateQueueDataFeed() {}
virtual ~PrivateQueueDataFeed() {}
virtual void Init(const paddle::framework::DataFeedDesc& data_feed_desc) = 0;
virtual bool Start();
virtual int Next();
protected:
// The thread implementation function for reading file and parse.
virtual void ReadThread();
// This function is used to set private-queue size, and the most
// efficient when the queue size is close to the batch size.
virtual void SetQueueSize(int queue_size);
// The reading and parsing method called in the ReadThread.
virtual bool ParseOneInstance(T* instance) = 0;
// This function is used to put instance to vec_ins
virtual void AddInstanceToInsVec(T* vec_ins, const T& instance,
int index) = 0;
// This function is used to put ins_vec to feed_vec
virtual void PutToFeedVec(const T& ins_vec) = 0;
// The thread for read files
std::thread read_thread_;
// using ifstream one line and one line parse is faster
// than using fread one buffer and one buffer parse.
// for a 601M real data:
// ifstream one line and one line parse: 6034 ms
// fread one buffer and one buffer parse: 7097 ms
std::ifstream file_;
size_t queue_size_;
// The queue for store parsed data
std::unique_ptr<paddle::operators::reader::BlockingQueue<T>> queue_;
};
// This class define the data type of instance(ins_vec) in MultiSlotDataFeed
class MultiSlotType {
public:
MultiSlotType() {}
~MultiSlotType() {}
void Init(const std::string& type) {
CheckType(type);
if (type_[0] == 'f') {
float_feasign_.clear();
} else if (type_[0] == 'u') {
uint64_feasign_.clear();
}
type_ = type;
}
void InitOffset() {
offset_.resize(1);
// LoDTensor' lod is counted from 0, the size of lod
// is one size larger than the size of data.
offset_[0] = 0;
}
const std::vector<size_t>& GetOffset() const { return offset_; }
void AddValue(const float v) {
CheckFloat();
float_feasign_.push_back(v);
}
void AddValue(const uint64_t v) {
CheckUint64();
uint64_feasign_.push_back(v);
}
void AddIns(const MultiSlotType& ins) {
if (ins.GetType()[0] == 'f') { // float
CheckFloat();
auto& vec = ins.GetFloatData();
offset_.push_back(offset_.back() + vec.size());
float_feasign_.insert(float_feasign_.end(), vec.begin(), vec.end());
} else if (ins.GetType()[0] == 'u') { // uint64
CheckUint64();
auto& vec = ins.GetUint64Data();
offset_.push_back(offset_.back() + vec.size());
uint64_feasign_.insert(uint64_feasign_.end(), vec.begin(), vec.end());
}
}
const std::vector<float>& GetFloatData() const { return float_feasign_; }
const std::vector<uint64_t>& GetUint64Data() const { return uint64_feasign_; }
const std::string& GetType() const { return type_; }
private:
void CheckType(const std::string& type) const {
PADDLE_ENFORCE((type == "uint64") || (type == "float"),
"There is no this type<%s>.", type);
}
void CheckFloat() const {
PADDLE_ENFORCE(type_[0] == 'f', "Add %s value to float slot.", type_);
}
void CheckUint64() const {
PADDLE_ENFORCE(type_[0] == 'u', "Add %s value to uint64 slot.", type_);
}
std::vector<float> float_feasign_;
std::vector<uint64_t> uint64_feasign_;
std::string type_;
std::vector<size_t> offset_;
};
// This DataFeed is used to feed multi-slot type data.
// The format of multi-slot type data:
// [n feasign_0 feasign_1 ... feasign_n]*
class MultiSlotDataFeed
: public PrivateQueueDataFeed<std::vector<MultiSlotType>> {
public:
MultiSlotDataFeed() {}
virtual ~MultiSlotDataFeed() {}
virtual void Init(const paddle::framework::DataFeedDesc& data_feed_desc);
virtual bool CheckFile(const char* filename);
protected:
virtual void AddInstanceToInsVec(std::vector<MultiSlotType>* vec_ins,
const std::vector<MultiSlotType>& instance,
int index);
virtual bool ParseOneInstance(std::vector<MultiSlotType>* instance);
virtual void PutToFeedVec(const std::vector<MultiSlotType>& ins_vec);
};
} // namespace framework
} // namespace paddle
paddle/fluid/framework/data_feed.proto 0 → 100644
浏览文件 @ bc34d325
/* 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. */
syntax = "proto2";
package paddle.framework;
message Slot {
required string name = 1;
required string type = 2;
optional bool is_dense = 3 [ default = false ];
optional bool is_used = 4 [ default = false ];
}
message MultiSlotDesc { repeated Slot slots = 1; }
message DataFeedDesc {
optional string name = 1;
optional int32 batch_size = 2 [ default = 32 ];
optional MultiSlotDesc multi_slot_desc = 3;
}
paddle/fluid/framework/data_feed_factory.cc 0 → 100644
浏览文件 @ bc34d325
/* 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 "paddle/fluid/framework/data_feed_factory.h"
#include <memory>
#include <string>
#include <unordered_map>
#include "paddle/fluid/framework/data_feed.h"
namespace paddle {
namespace framework {
typedef std::shared_ptr<DataFeed> (*Createdata_feedFunction)();
typedef std::unordered_map<std::string, Createdata_feedFunction> data_feedMap;
data_feedMap g_data_feed_map;
#define REGISTER_DATAFEED_CLASS(data_feed_class) \
namespace { \
std::shared_ptr<DataFeed> Creator_##data_feed_class() { \
return std::shared_ptr<DataFeed>(new data_feed_class); \
} \
class __Registerer_##data_feed_class { \
public: \
__Registerer_##data_feed_class() { \
g_data_feed_map[#data_feed_class] = &Creator_##data_feed_class; \
} \
}; \
__Registerer_##data_feed_class g_registerer_##data_feed_class; \
} // namespace
std::string DataFeedFactory::DataFeedTypeList() {
std::string data_feed_types;
for (auto iter = g_data_feed_map.begin(); iter != g_data_feed_map.end();
++iter) {
if (iter != g_data_feed_map.begin()) {
data_feed_types += ", ";
}
data_feed_types += iter->first;
}
return data_feed_types;
}
std::shared_ptr<DataFeed> DataFeedFactory::CreateDataFeed(
std::string data_feed_class) {
if (g_data_feed_map.count(data_feed_class) < 1) {
exit(-1);
}
return g_data_feed_map[data_feed_class]();
}
REGISTER_DATAFEED_CLASS(MultiSlotDataFeed);
} // namespace framework
} // namespace paddle
paddle/fluid/framework/data_feed_factory.h 0 → 100644
浏览文件 @ bc34d325
/* 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. */
#pragma once
#include <memory>
#include <string>
#include "paddle/fluid/framework/data_feed.h"
namespace paddle {
namespace framework {
class DataFeedFactory {
public:
static std::string DataFeedTypeList();
static std::shared_ptr<DataFeed> CreateDataFeed(std::string data_feed_class);
};
} // namespace framework
} // namespace paddle
paddle/fluid/framework/data_feed_test.cc 0 → 100644
浏览文件 @ bc34d325
// 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 "paddle/fluid/framework/data_feed.h"
#include <fcntl.h>
#include <chrono> // NOLINT
#include <fstream>
#include <iostream>
#include <map>
#include <mutex> // NOLINT
#include <set>
#include <thread> // NOLINT
#include <utility>
#include <vector>
#include "google/protobuf/io/zero_copy_stream_impl.h"
#include "google/protobuf/text_format.h"
#include "gtest/gtest.h"
#include "paddle/fluid/framework/data_feed_factory.h"
#include "paddle/fluid/framework/lod_tensor.h"
#include "paddle/fluid/framework/scope.h"
paddle::framework::DataFeedDesc load_datafeed_param_from_file(
const char* filename) {
paddle::framework::DataFeedDesc data_feed_desc;
int file_descriptor = open(filename, O_RDONLY);
PADDLE_ENFORCE(file_descriptor != -1, "Can not open %s.", filename);
google::protobuf::io::FileInputStream fileInput(file_descriptor);
google::protobuf::TextFormat::Parse(&fileInput, &data_feed_desc);
close(file_descriptor);
return data_feed_desc;
}
const std::vector<std::string> load_filelist_from_file(const char* filename) {
std::vector<std::string> filelist;
std::ifstream fin(filename);
PADDLE_ENFORCE(fin.good(), "Can not open %s.", filename);
std::string line;
while (getline(fin, line)) {
filelist.push_back(line);
}
fin.close();
return filelist;
}
void GenerateFileForTest(const char* protofile, const char* filelist) {
std::ofstream w_protofile(protofile);
w_protofile << "name: \"MultiSlotDataFeed\"\n"
"batch_size: 2\n"
"multi_slot_desc {\n"
" slots {\n"
" name: \"uint64_sparse_slot\"\n"
" type: \"uint64\"\n"
" is_dense: false\n"
" is_used: true\n"
" }\n"
" slots {\n"
" name: \"float_sparse_slot\"\n"
" type: \"float\"\n"
" is_dense: false\n"
" is_used: true\n"
" }\n"
" slots {\n"
" name: \"uint64_dense_slot\"\n"
" type: \"uint64\"\n"
" is_dense: true\n"
" is_used: true\n"
" }\n"
" slots {\n"
" name: \"float_dense_slot\"\n"
" type: \"float\"\n"
" is_dense: true\n"
" is_used: true\n"
" }\n"
" slots {\n"
" name: \"not_used_slot\"\n"
" type: \"uint64\"\n"
" is_dense: false\n"
" is_used: false\n"
" }\n"
"}";
w_protofile.close();
std::ofstream w_filelist(filelist);
int total_file = 4;
for (int i = 0; i < total_file; ++i) {
std::string filename = "TestMultiSlotDataFeed.data." + std::to_string(i);
w_filelist << filename;
if (i + 1 != total_file) {
w_filelist << std::endl;
}
std::ofstream w_datafile(filename.c_str());
w_datafile << "3 3978 620 82 1 1926.08 1 1926 1 6.02 1 1996\n"
"2 1300 2983353 1 985.211 1 8 1 0.618 1 12\n"
"1 19260827 2 3.14 2.718 1 27 1 2.236 1 28\n";
w_datafile.close();
}
w_filelist.close();
}
class MultiTypeSet {
public:
MultiTypeSet() {
uint64_set_.clear();
float_set_.clear();
}
~MultiTypeSet() {}
void AddValue(uint64_t v) { uint64_set_.insert(v); }
void AddValue(float v) { float_set_.insert(v); }
const std::set<uint64_t>& GetUint64Set() const { return uint64_set_; }
const std::set<float>& GetFloatSet() const { return float_set_; }
private:
std::set<uint64_t> uint64_set_;
std::set<float> float_set_;
};
void GetElemSetFromReader(std::vector<MultiTypeSet>* reader_elem_set,
const paddle::framework::DataFeedDesc& data_feed_desc,
const std::vector<std::string>& filelist,
const int thread_num) {
int used_slot_num = 0;
for (auto i = 0; i < data_feed_desc.multi_slot_desc().slots_size(); ++i) {
if (data_feed_desc.multi_slot_desc().slots(i).is_used()) {
++used_slot_num;
}
}
reader_elem_set->resize(used_slot_num);
std::vector<std::thread> threads;
std::vector<std::shared_ptr<paddle::framework::DataFeed>> readers;
readers.resize(thread_num);
for (int i = 0; i < thread_num; ++i) {
readers[i] = paddle::framework::DataFeedFactory::CreateDataFeed(
data_feed_desc.name());
readers[i]->Init(data_feed_desc);
}
readers[0]->SetFileList(filelist);
std::mutex mu;
for (int idx = 0; idx < thread_num; ++idx) {
threads.emplace_back(std::thread([&, idx] {
std::unique_ptr<paddle::framework::Scope> scope(
new paddle::framework::Scope());
const auto& multi_slot_desc = data_feed_desc.multi_slot_desc();
std::map<std::string, const paddle::framework::LoDTensor*>
lodtensor_targets;
std::map<std::string, const paddle::framework::Tensor*> tensor_targets;
for (int i = 0; i < multi_slot_desc.slots_size(); ++i) {
const auto& slot = multi_slot_desc.slots(i);
if (slot.is_used()) {
const auto& name = slot.name();
readers[idx]->AddFeedVar(scope->Var(name), name);
if (slot.is_dense()) {
tensor_targets[name] =
&scope->FindVar(name)->Get<paddle::framework::Tensor>();
} else {
lodtensor_targets[name] =
&scope->FindVar(name)->Get<paddle::framework::LoDTensor>();
}
}
}
readers[idx]->Start();
while (readers[idx]->Next()) {
int index = 0;
for (int k = 0; k < multi_slot_desc.slots_size(); ++k) {
const auto& slot = multi_slot_desc.slots(k);
if (!slot.is_used()) {
continue;
}
if (slot.is_dense()) { // dense branch
const paddle::framework::Tensor* tens = tensor_targets[slot.name()];
if (slot.type() == "uint64") {
const int64_t* data = tens->data<int64_t>();
int batch_size = tens->dims()[0];
int dim = tens->dims()[1];
for (int i = 0; i < batch_size; ++i) {
for (int j = 0; j < dim; ++j) {
std::lock_guard<std::mutex> lock(mu);
(*reader_elem_set)[index].AddValue(
(uint64_t)data[i * dim + j]);
}
}
} else if (slot.type() == "float") {
const float* data = tens->data<float>();
int batch_size = tens->dims()[0];
int dim = tens->dims()[1];
for (int i = 0; i < batch_size; ++i) {
for (int j = 0; j < dim; ++j) {
std::lock_guard<std::mutex> lock(mu);
(*reader_elem_set)[index].AddValue(data[i * dim + j]);
}
}
} else {
PADDLE_THROW("Error type in proto file.");
}
} else { // sparse branch
const paddle::framework::LoDTensor* tens =
lodtensor_targets[slot.name()];
if (slot.type() == "uint64") {
const int64_t* data = tens->data<int64_t>();
for (size_t i = 0; i < tens->NumElements(); ++i) {
std::pair<size_t, size_t> element = tens->lod_element(0, i);
for (size_t j = element.first; j < element.second; ++j) {
std::lock_guard<std::mutex> lock(mu);
(*reader_elem_set)[index].AddValue((uint64_t)data[j]);
}
}
} else if (slot.type() == "float") {
const float* data = tens->data<float>();
for (size_t i = 0; i < tens->NumElements(); ++i) {
std::pair<size_t, size_t> element = tens->lod_element(0, i);
for (size_t j = element.first; j < element.second; ++j) {
std::lock_guard<std::mutex> lock(mu);
(*reader_elem_set)[index].AddValue(data[j]);
}
}
} else {
PADDLE_THROW("Error type in proto file.");
}
} // end sparse branch
++index;
} // end slots loop
} // end while Next()
})); // end anonymous function
}
for (auto& th : threads) {
th.join();
}
}
void CheckIsUnorderedSame(const std::vector<MultiTypeSet>& s1,
const std::vector<MultiTypeSet>& s2) {
EXPECT_EQ(s1.size(), s2.size());
for (size_t i = 0; i < s1.size(); ++i) {
// check for uint64
const std::set<uint64_t>& uint64_s1 = s1[i].GetUint64Set();
const std::set<uint64_t>& uint64_s2 = s2[i].GetUint64Set();
EXPECT_EQ(uint64_s1.size(), uint64_s2.size());
auto uint64_it1 = uint64_s1.begin();
auto uint64_it2 = uint64_s2.begin();
while (uint64_it1 != uint64_s1.end()) {
EXPECT_EQ(*uint64_it1, *uint64_it2);
++uint64_it1;
++uint64_it2;
}
// check for float
const std::set<float>& float_s1 = s1[i].GetFloatSet();
const std::set<float>& float_s2 = s2[i].GetFloatSet();
EXPECT_EQ(float_s1.size(), float_s2.size());
auto float_it1 = float_s1.begin();
auto float_it2 = float_s2.begin();
while (float_it1 != float_s1.end()) {
EXPECT_EQ(*float_it1, *float_it2);
++float_it1;
++float_it2;
}
}
}
void GetElemSetFromFile(std::vector<MultiTypeSet>* file_elem_set,
const paddle::framework::DataFeedDesc& data_feed_desc,
const std::vector<std::string>& filelist) {
int used_slot_num = 0;
for (auto i = 0; i < data_feed_desc.multi_slot_desc().slots_size(); ++i) {
if (data_feed_desc.multi_slot_desc().slots(i).is_used()) {
++used_slot_num;
}
}
file_elem_set->resize(used_slot_num);
for (const auto& file : filelist) {
std::ifstream fin(file.c_str());
PADDLE_ENFORCE(fin.good(), "Can not open %s.", file.c_str());
while (1) {
bool end_flag = false;
int index = 0;
for (auto i = 0; i < data_feed_desc.multi_slot_desc().slots_size(); ++i) {
int num;
if (fin >> num) {
auto slot = data_feed_desc.multi_slot_desc().slots(i);
auto type = slot.type();
if (type == "uint64") {
while (num--) {
uint64_t feasign;
fin >> feasign;
if (slot.is_used()) {
(*file_elem_set)[index].AddValue(feasign);
}
}
} else if (type == "float") {
while (num--) {
float feasign;
fin >> feasign;
if (slot.is_used()) {
(*file_elem_set)[index].AddValue(feasign);
}
}
} else {
PADDLE_THROW("Error type in proto file.");
}
if (slot.is_used()) {
++index;
}
} else {
end_flag = true;
break;
}
}
if (end_flag) {
break;
}
}
fin.close();
}
}
TEST(DataFeed, MultiSlotUnitTest) {
const char* protofile = "data_feed_desc.prototxt";
const char* filelist_name = "filelist.txt";
GenerateFileForTest(protofile, filelist_name);
const std::vector<std::string> filelist =
load_filelist_from_file(filelist_name);
paddle::framework::DataFeedDesc data_feed_desc =
load_datafeed_param_from_file(protofile);
std::vector<MultiTypeSet> reader_elem_set;
std::vector<MultiTypeSet> file_elem_set;
GetElemSetFromReader(&reader_elem_set, data_feed_desc, filelist, 4);
GetElemSetFromFile(&file_elem_set, data_feed_desc, filelist);
CheckIsUnorderedSame(reader_elem_set, file_elem_set);
}
paddle/fluid/framework/details/CMakeLists.txt
浏览文件 @ bc34d325
cc_library(var_handle SRCS var_handle.cc DEPS place framework_proto node)
cc_library(op_handle_base SRCS op_handle_base.cc DEPS var_handle device_context lod_tensor)
cc_library(op_graph_view SRCS op_graph_view.cc DEPS op_handle_base)
cc_library(scale_loss_grad_op_handle SRCS scale_loss_grad_op_handle.cc DEPS op_handle_base scope lod_tensor ddim memory)
cc_library(fetch_op_handle SRCS fetch_op_handle.cc DEPS op_handle_base scope lod_tensor ddim memory)
cc_library(computation_op_handle SRCS computation_op_handle.cc DEPS framework_proto scope place operator op_registry)
......@@ -16,32 +17,40 @@ if(WITH_GPU)
dynload_cuda variable_visitor)
nv_library(reduce_op_handle SRCS reduce_op_handle.cc DEPS op_handle_base variable_visitor scope ddim dynload_cuda)
nv_library(broadcast_op_handle SRCS broadcast_op_handle.cc DEPS op_handle_base scope ddim memory variable_visitor dynload_cuda)
nv_library(fused_broadcast_op_handle SRCS fused_broadcast_op_handle.cc DEPS broadcast_op_handle)
else()
cc_library(all_reduce_op_handle SRCS all_reduce_op_handle.cc DEPS op_handle_base scope lod_tensor ddim memory
variable_visitor)
cc_library(reduce_op_handle SRCS reduce_op_handle.cc DEPS op_handle_base variable_visitor scope ddim)
cc_library(broadcast_op_handle SRCS broadcast_op_handle.cc DEPS op_handle_base scope ddim memory variable_visitor)
cc_library(fused_broadcast_op_handle SRCS fused_broadcast_op_handle.cc DEPS broadcast_op_handle)
endif()
cc_library(data_balance_op_handle SRCS data_balance_op_handle.cc DEPS op_handle_base scope lod_tensor)
cc_library(gather_op_handle SRCS gather_op_handle.cc DEPS op_handle_base scope ddim memory variable_visitor)
cc_library(fuse_vars_op_handle SRCS fuse_vars_op_handle.cc DEPS op_handle_base scope)
if(WITH_GPU)
cc_library(modify_op_lock_and_record_event_pass SRCS modify_op_lock_and_record_event_pass.cc DEPS computation_op_handle op_graph_view multi_devices_helper)
if (WITH_GPU)
cc_library(reference_count_pass SRCS reference_count_pass.cc DEPS computation_op_handle scale_loss_grad_op_handle rpc_op_handle
all_reduce_op_handle reduce_op_handle broadcast_op_handle data_balance_op_handle graph graph_helper pass)
endif()
cc_library(sequential_execution_pass SRCS sequential_execution_pass.cc DEPS graph graph_helper pass)
cc_library(all_reduce_deps_pass SRCS all_reduce_deps_pass.cc DEPS graph graph_helper pass)
cc_library(multi_devices_graph_pass SRCS multi_devices_graph_pass.cc DEPS multi_devices_helper computation_op_handle
scale_loss_grad_op_handle rpc_op_handle all_reduce_op_handle reduce_op_handle broadcast_op_handle data_balance_op_handle)
scale_loss_grad_op_handle rpc_op_handle all_reduce_op_handle reduce_op_handle broadcast_op_handle data_balance_op_handle fused_broadcast_op_handle)
if(WITH_GPU)
cc_library(ssa_graph_executor SRCS ssa_graph_executor.cc DEPS graph framework_proto reference_count_pass)
else()
cc_library(ssa_graph_executor SRCS ssa_graph_executor.cc DEPS graph framework_proto)
set(SSA_GRAPH_EXECUTOR_DEPS graph framework_proto sequential_execution_pass modify_op_lock_and_record_event_pass all_reduce_deps_pass)
if (WITH_GPU)
list(APPEND SSA_GRAPH_EXECUTOR_DEPS reference_count_pass)
endif()
cc_library(ssa_graph_executor SRCS ssa_graph_executor.cc DEPS ${SSA_GRAPH_EXECUTOR_DEPS})
cc_library(threaded_ssa_graph_executor SRCS threaded_ssa_graph_executor.cc DEPS fetch_op_handle ssa_graph_executor scope
simple_threadpool device_context)
......@@ -54,8 +63,9 @@ cc_library(scope_buffered_ssa_graph_executor SRCS scope_buffered_ssa_graph_execu
# device_context reduce_op_handle )
cc_library(fast_threaded_ssa_graph_executor SRCS fast_threaded_ssa_graph_executor.cc
DEPS fetch_op_handle ssa_graph_executor scope simple_threadpool device_context)
cc_test(fused_broadcast_op_test SRCS fused_broadcast_op_handle_test.cc DEPS fused_broadcast_op_handle)
cc_library(build_strategy SRCS build_strategy.cc DEPS
graph_viz_pass multi_devices_graph_pass
multi_devices_graph_print_pass multi_devices_graph_check_pass
fuse_elewise_add_act_pass)
fuse_elewise_add_act_pass multi_batch_merge_pass)
paddle/fluid/framework/details/all_reduce_deps_pass.cc 0 → 100644
浏览文件 @ bc34d325
// 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 <algorithm>
#include <string>
#include <unordered_map>
#include <unordered_set>
#include <vector>
#include "paddle/fluid/framework/details/all_reduce_deps_pass.h"
#include "paddle/fluid/framework/details/all_reduce_op_handle.h"
#include "paddle/fluid/framework/details/multi_devices_helper.h"
#include "paddle/fluid/framework/details/op_graph_view.h"
#include "paddle/fluid/framework/details/var_handle.h"
#include "paddle/fluid/framework/ir/graph_helper.h"
#include "paddle/fluid/framework/op_proto_maker.h"
namespace paddle {
namespace framework {
namespace details {
static constexpr char kAllOpDescs[] = "all_op_descs";
VarHandle* GetValidInput(const OpHandleBase* a) {
for (auto p : a->Inputs()) {
VarHandle* b = dynamic_cast<VarHandle*>(p);
if (b) {
return b;
}
}
return nullptr;
}
std::unique_ptr<ir::Graph> AllReduceDepsPass::ApplyImpl(
std::unique_ptr<ir::Graph> graph) const {
auto graph_ops = ir::FilterByNodeWrapper<OpHandleBase>(*graph);
// get vars order
int order = 0;
std::unordered_map<std::string, int> vars;
// TODO(gongwb): use graph topology sort to find the order of operators.
// Note that must assert topology sort is stable
auto& ops = Get<const std::vector<OpDesc*>>(kAllOpDescs);
for (auto* op_desc : ops) {
auto outputs = op_desc->Outputs();
for (auto& o_it : outputs) {
for (auto& v : o_it.second) { // values
vars[v] = order;
}
}
order++;
}
std::vector<OpHandleBase*> dist_ops;
// get allreduce ops.
for (auto& op : graph_ops) {
// FIXME(gongwb):add broad cast.
if (op->Name() == "all_reduce" || op->Name() == "reduce") {
dist_ops.push_back(op);
}
}
VLOG(10) << "dist_ops size:" << dist_ops.size() << std::endl;
std::sort(dist_ops.begin(), dist_ops.end(), [&](OpHandleBase* op1,
OpHandleBase* op2) {
VarHandle* i0 = dynamic_cast<VarHandle*>(GetValidInput(op1));
VarHandle* i1 = dynamic_cast<VarHandle*>(GetValidInput(op2));
PADDLE_ENFORCE(i0 != nullptr && i1 != nullptr, "%s convert to %s error",
op1->DebugString(), op2->DebugString());
auto l_it = vars.find(i0->name_);
auto r_it = vars.find(i1->name_);
if (l_it->second < r_it->second) return true;
if (l_it->second == r_it->second) {
return i0->name_ < i1->name_;
}
return false;
});
// add dependency.
auto& sorted_ops = dist_ops;
for (size_t i = 1; i < sorted_ops.size(); ++i) {
auto* dep_var = new DummyVarHandle(graph->CreateControlDepVar());
auto* pre_op = sorted_ops[i - 1];
auto* op = sorted_ops[i];
pre_op->AddOutput(dep_var);
op->AddInput(dep_var);
graph->Get<GraphDepVars>(kGraphDepVars).emplace(dep_var);
VLOG(10) << "add all_reduce sequential dependencies between " << pre_op
<< " and " << op;
VLOG(10) << "pre_op:" << pre_op->DebugString()
<< ", op:" << op->DebugString();
}
return graph;
}
} // namespace details
} // namespace framework
} // namespace paddle
REGISTER_PASS(all_reduce_deps_pass,
paddle::framework::details::AllReduceDepsPass)
.RequirePassAttr(paddle::framework::details::kAllOpDescs);
paddle/fluid/framework/details/all_reduce_deps_pass.h 0 → 100644
浏览文件 @ bc34d325
// 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.
#pragma once
#include "paddle/fluid/framework/ir/graph.h"
#include "paddle/fluid/framework/ir/pass.h"
namespace paddle {
namespace framework {
namespace details {
// TODO(gongwb): overlap allreduce with backward computation.
class AllReduceDepsPass : public ir::Pass {
protected:
std::unique_ptr<ir::Graph> ApplyImpl(
std::unique_ptr<ir::Graph> graph) const override;
};
} // namespace details
} // namespace framework
} // namespace paddle
paddle/fluid/framework/details/all_reduce_op_handle.cc
浏览文件 @ bc34d325
......@@ -23,7 +23,7 @@ namespace paddle {
namespace framework {
namespace details {
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
AllReduceOpHandle::AllReduceOpHandle(ir::Node *node,
const std::vector<Scope *> &local_scopes,
const std::vector<platform::Place> &places,
......@@ -34,7 +34,7 @@ AllReduceOpHandle::AllReduceOpHandle(ir::Node *node,
nccl_ctxs_(ctxs) {
if (nccl_ctxs_) {
for (auto &p : places_) {
this->dev_ctxes_[p] = nccl_ctxs_->DevCtx(p);
this->SetDeviceContext(p, nccl_ctxs_->DevCtx(p));
}
}
}
......@@ -46,7 +46,7 @@ AllReduceOpHandle::AllReduceOpHandle(ir::Node *node,
#endif
void AllReduceOpHandle::RunImpl() {
platform::RecordEvent record_event(Name(), dev_ctxes_.begin()->second);
platform::RecordEvent record_event(Name(), dev_ctxes_.cbegin()->second);
if (NoDummyInputSize() == 1) {
return; // No need to all reduce when GPU count = 1;
......@@ -74,7 +74,7 @@ void AllReduceOpHandle::RunImpl() {
}
if (platform::is_gpu_place(lod_tensors[0]->place())) {
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
PADDLE_ENFORCE(nccl_ctxs_, "nccl_ctxs should not be nullptr.");
int dtype = -1;
size_t numel = 0;
......@@ -127,7 +127,7 @@ void AllReduceOpHandle::RunImpl() {
*local_scopes_[i]->FindVar(kLocalExecScopeName)->Get<Scope *>();
auto &p = places_[i];
auto *var = scope.FindVar(out_var_handles[i]->name_);
auto *dev_ctx = dev_ctxes_[p];
auto *dev_ctx = dev_ctxes_.at(p);
RunAndRecordEvent(p, [&trg, var, dev_ctx, p] {
auto &tensor_gpu = *var->GetMutable<framework::LoDTensor>();
......
paddle/fluid/framework/details/all_reduce_op_handle.h
浏览文件 @ bc34d325
......@@ -20,7 +20,7 @@
#include "paddle/fluid/framework/details/op_handle_base.h"
#include "paddle/fluid/framework/lod_tensor.h"
#include "paddle/fluid/framework/scope.h"
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
#include "paddle/fluid/platform/nccl_helper.h"
#endif
......@@ -29,7 +29,7 @@ namespace framework {
namespace details {
struct AllReduceOpHandle : public OpHandleBase {
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
AllReduceOpHandle(ir::Node *node, const std::vector<Scope *> &local_scopes,
const std::vector<platform::Place> &places,
const platform::NCCLContextMap *ctxs);
......@@ -49,7 +49,7 @@ struct AllReduceOpHandle : public OpHandleBase {
private:
std::vector<Scope *> local_scopes_;
std::vector<platform::Place> places_;
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
const platform::NCCLContextMap *nccl_ctxs_;
#endif
};
......
paddle/fluid/framework/details/broadcast_op_handle.cc
浏览文件 @ bc34d325
......@@ -48,16 +48,27 @@ void BroadcastOpHandle::RunImpl() {
var_scopes.emplace_back(s->FindVar(kLocalExecScopeName)->Get<Scope *>());
}
BroadcastOneVar(*in_var_handle, out_var_handles, var_scopes);
}
void BroadcastOpHandle::BroadcastOneVar(
const VarHandle &in_var_handle,
const std::vector<VarHandle *> &out_var_handles,
const std::vector<const Scope *> &var_scopes) {
auto *in_var =
var_scopes.at(in_var_handle->scope_idx_)->FindVar(in_var_handle->name_);
var_scopes.at(in_var_handle.scope_idx_)->FindVar(in_var_handle.name_);
PADDLE_ENFORCE_NOT_NULL(in_var);
Tensor &in_tensor = VariableVisitor::GetMutableTensor(in_var);
if (UNLIKELY(!in_tensor.IsInitialized())) {
VLOG(3) << "in var " << in_var_handle.name_ << "not inited, return!";
return;
}
InitOutputValue(*in_var_handle, out_var_handles);
InitOutputValue(in_var_handle, out_var_handles);
if (platform::is_cpu_place(in_tensor.place())) {
for (auto *out_var_handle : out_var_handles) {
if (out_var_handle->IsTheSameVar(*in_var_handle)) {
if (out_var_handle->IsTheSameVar(in_var_handle)) {
continue;
}
auto &out_p = out_var_handle->place_;
......@@ -71,7 +82,7 @@ void BroadcastOpHandle::RunImpl() {
});
}
} else {
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
VarHandle *out_handle = nullptr;
int root_id = boost::get<platform::CUDAPlace>(in_tensor.place()).device;
std::vector<std::function<void()>> broadcast_calls;
......@@ -114,12 +125,12 @@ void BroadcastOpHandle::RunImpl() {
}
}
if (!out_handle->IsTheSameVar(*in_var_handle)) {
auto out_var = var_scopes.at(in_var_handle->scope_idx_)
if (!out_handle->IsTheSameVar(in_var_handle)) {
auto out_var = var_scopes.at(in_var_handle.scope_idx_)
->FindVar(out_var_handles[0]->name_);
paddle::framework::TensorCopy(
in_tensor, in_var_handle->place_,
*(dev_ctxes_.at(in_var_handle->place_)),
in_tensor, in_var_handle.place_,
*(dev_ctxes_.at(in_var_handle.place_)),
&VariableVisitor::GetMutableTensor(out_var));
}
});
......
paddle/fluid/framework/details/broadcast_op_handle.h
浏览文件 @ bc34d325
......@@ -24,7 +24,7 @@
#include "paddle/fluid/framework/selected_rows.h"
#include "paddle/fluid/platform/device_context.h"
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
#include "paddle/fluid/platform/nccl_helper.h"
#endif
......@@ -34,7 +34,7 @@ namespace details {
struct BroadcastOpHandle : public OpHandleBase {
public:
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
BroadcastOpHandle(ir::Node *node, const std::vector<Scope *> &local_scopes,
const std::vector<platform::Place> &places,
const platform::NCCLContextMap *nccl_ctxs)
......@@ -44,7 +44,8 @@ struct BroadcastOpHandle : public OpHandleBase {
nccl_ctxs_(nccl_ctxs) {
if (nccl_ctxs_) {
for (auto &p_ctx : nccl_ctxs_->contexts_) {
dev_ctxes_[platform::CUDAPlace(p_ctx.first)] = p_ctx.second.ctx_.get();
this->SetDeviceContext(platform::CUDAPlace(p_ctx.first),
p_ctx.second.ctx_.get());
}
}
}
......@@ -61,10 +62,13 @@ struct BroadcastOpHandle : public OpHandleBase {
protected:
void RunImpl() override;
private:
void BroadcastOneVar(const VarHandle &in_var_handle,
const std::vector<VarHandle *> &out_var_handles,
const std::vector<const Scope *> &var_scopes);
std::vector<Scope *> local_scopes_;
std::vector<platform::Place> places_;
#ifdef PADDLE_WITH_CUDA
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
const platform::NCCLContextMap *nccl_ctxs_;
#endif
......
paddle/fluid/framework/details/broadcast_op_handle_test.cc
浏览文件 @ bc34d325
......@@ -12,232 +12,12 @@
// See the License for the specific language governing permissions and
// limitations under the License.
#include "paddle/fluid/framework/details/broadcast_op_handle.h"
#include "gtest/gtest.h"
#include "paddle/fluid/platform/device_context.h"
#include "paddle/fluid/framework/details/broadcast_op_handle_test.h"
namespace paddle {
namespace framework {
namespace details {
namespace f = paddle::framework;
namespace p = paddle::platform;
// test data amount
const f::DDim kDims = {20, 20};
struct TestBroadcastOpHandle {
std::vector<std::unique_ptr<p::DeviceContext>> ctxs_;
std::vector<Scope*> local_scopes_;
std::vector<Scope*> param_scopes_;
Scope g_scope_;
std::unique_ptr<OpHandleBase> op_handle_;
std::vector<std::unique_ptr<VarHandleBase>> vars_;
std::vector<p::Place> gpu_list_;
bool use_gpu_;
#ifdef PADDLE_WITH_CUDA
std::unique_ptr<platform::NCCLContextMap> nccl_ctxs_;
#endif
void WaitAll() {
for (size_t j = 0; j < ctxs_.size(); ++j) {
ctxs_[j]->Wait();
}
#ifdef PADDLE_WITH_CUDA
if (nccl_ctxs_) {
nccl_ctxs_->WaitAll();
}
#endif
}
void InitCtxOnGpu(bool use_gpu) {
use_gpu_ = use_gpu;
if (use_gpu_) {
#ifdef PADDLE_WITH_CUDA
int count = p::GetCUDADeviceCount();
if (count <= 1) {
LOG(WARNING) << "Cannot test multi-gpu Broadcast, because the CUDA "
"device count is "
<< count;
exit(0);
}
for (int i = 0; i < count; ++i) {
auto p = p::CUDAPlace(i);
gpu_list_.push_back(p);
ctxs_.emplace_back(new p::CUDADeviceContext(p));
}
nccl_ctxs_.reset(new platform::NCCLContextMap(gpu_list_));
#else
PADDLE_THROW("CUDA is not support.");
#endif
} else {
int count = 8;
for (int i = 0; i < count; ++i) {
auto p = p::CPUPlace();
gpu_list_.push_back(p);
ctxs_.emplace_back(new p::CPUDeviceContext(p));
}
#ifdef PADDLE_WITH_CUDA
nccl_ctxs_.reset(nullptr);
#endif
}
}
void InitBroadcastOp(size_t input_scope_idx) {
for (size_t j = 0; j < gpu_list_.size(); ++j) {
local_scopes_.push_back(&(g_scope_.NewScope()));
Scope& local_scope = local_scopes_.back()->NewScope();
*local_scopes_.back()
->Var(details::kLocalExecScopeName)
->GetMutable<Scope*>() = &local_scope;
local_scope.Var("out");
param_scopes_.emplace_back(&local_scope);
}
param_scopes_[input_scope_idx]->Var("input");
std::unique_ptr<ir::Node> n =
ir::CreateNodeForTest("node0", ir::Node::Type::kOperation);
if (use_gpu_) {
#ifdef PADDLE_WITH_CUDA
op_handle_.reset(new BroadcastOpHandle(n.get(), local_scopes_, gpu_list_,
nccl_ctxs_.get()));
#else
PADDLE_THROW("CUDA is not support.");
#endif
} else {
#ifdef PADDLE_WITH_CUDA
op_handle_.reset(new BroadcastOpHandle(n.get(), local_scopes_, gpu_list_,
nccl_ctxs_.get()));
#else
op_handle_.reset(
new BroadcastOpHandle(n.get(), local_scopes_, gpu_list_));
#endif
}
std::unique_ptr<ir::Node> v =
ir::CreateNodeForTest("node1", ir::Node::Type::kVariable);
auto* in_var_handle = new VarHandle(v.get(), 1, input_scope_idx, "input",
gpu_list_[input_scope_idx]);
vars_.emplace_back(in_var_handle);
op_handle_->AddInput(in_var_handle);
// add dummy var
std::unique_ptr<ir::Node> v2 =
ir::CreateNodeForTest("node2", ir::Node::Type::kVariable);
vars_.emplace_back(new DummyVarHandle(v2.get()));
DummyVarHandle* dummy_var_handle =
static_cast<DummyVarHandle*>(vars_.back().get());
dummy_var_handle->ClearGeneratedOp();
op_handle_->AddInput(dummy_var_handle);
for (size_t j = 0; j < gpu_list_.size(); ++j) {
if (!use_gpu_) {
op_handle_->SetDeviceContext(gpu_list_[j], ctxs_[j].get());
}
std::unique_ptr<ir::Node> v3 =
ir::CreateNodeForTest("node3", ir::Node::Type::kVariable);
VarHandle* out_var_handle =
new VarHandle(v3.get(), 2, j, "out", gpu_list_[j]);
vars_.emplace_back(out_var_handle);
op_handle_->AddOutput(out_var_handle);
}
// add dummy var
std::unique_ptr<ir::Node> v4 =
ir::CreateNodeForTest("node4", ir::Node::Type::kVariable);
vars_.emplace_back(new DummyVarHandle(v4.get()));
DummyVarHandle* out_dummy_var_handle =
static_cast<DummyVarHandle*>(vars_.back().get());
out_dummy_var_handle->ClearGeneratedOp();
op_handle_->AddOutput(out_dummy_var_handle);
}
void TestBroadcastLodTensor(size_t input_scope_idx) {
auto in_var = param_scopes_[input_scope_idx]->FindVar("input");
PADDLE_ENFORCE_NOT_NULL(in_var);
auto in_lod_tensor = in_var->GetMutable<f::LoDTensor>();
in_lod_tensor->mutable_data<float>(kDims, gpu_list_[input_scope_idx]);
std::vector<float> send_vector(static_cast<size_t>(f::product(kDims)));
for (size_t k = 0; k < send_vector.size(); ++k) {
send_vector[k] = k;
}
f::LoD lod{{0, 10, 20}};
paddle::framework::TensorFromVector<float>(
send_vector, *(ctxs_[input_scope_idx]), in_lod_tensor);
in_lod_tensor->set_lod(lod);
in_lod_tensor->Resize(kDims);
op_handle_->Run(false);
WaitAll();
p::CPUPlace cpu_place;
for (size_t j = 0; j < gpu_list_.size(); ++j) {
auto out_var = param_scopes_[j]->FindVar("out");
PADDLE_ENFORCE_NOT_NULL(out_var);
auto out_tensor = out_var->Get<f::LoDTensor>();
PADDLE_ENFORCE_EQ(out_tensor.lod(), lod, "lod is not equal.");
f::Tensor result_tensor;
f::TensorCopySync(out_tensor, cpu_place, &result_tensor);
float* ct = result_tensor.mutable_data<float>(cpu_place);
for (int64_t i = 0; i < f::product(kDims); ++i) {
ASSERT_NEAR(ct[i], send_vector[i], 1e-5);
}
}
}
void TestBroadcastSelectedRows(size_t input_scope_idx) {
auto in_var = param_scopes_[input_scope_idx]->FindVar("input");
PADDLE_ENFORCE_NOT_NULL(in_var);
auto in_selected_rows = in_var->GetMutable<f::SelectedRows>();
auto value = in_selected_rows->mutable_value();
value->mutable_data<float>(kDims, gpu_list_[input_scope_idx]);
int height = static_cast<int>(kDims[0]) * 2;
std::vector<int64_t> rows{0, 1, 2, 3, 3, 0, 14, 7, 3, 1,
2, 4, 6, 3, 1, 1, 1, 1, 3, 7};
in_selected_rows->set_height(height);
in_selected_rows->set_rows(rows);
std::vector<float> send_vector(static_cast<size_t>(f::product(kDims)));
for (size_t k = 0; k < send_vector.size(); ++k) {
send_vector[k] = k;
}
paddle::framework::TensorFromVector<float>(
send_vector, *(ctxs_[input_scope_idx]), value);
op_handle_->Run(false);
WaitAll();
p::CPUPlace cpu_place;
for (size_t j = 0; j < gpu_list_.size(); ++j) {
auto out_var = param_scopes_[j]->FindVar("out");
PADDLE_ENFORCE_NOT_NULL(out_var);
auto& out_select_rows = out_var->Get<f::SelectedRows>();
auto rt = out_select_rows.value();
PADDLE_ENFORCE_EQ(out_select_rows.height(), height,
"height is not equal.");
for (size_t k = 0; k < out_select_rows.rows().size(); ++k) {
PADDLE_ENFORCE_EQ(out_select_rows.rows()[k], rows[k]);
}
f::Tensor result_tensor;
f::TensorCopySync(rt, cpu_place, &result_tensor);
float* ct = result_tensor.data<float>();
for (int64_t i = 0; i < f::product(kDims); ++i) {
ASSERT_NEAR(ct[i], send_vector[i], 1e-5);
}
}
}
};
TEST(BroadcastTester, TestCPUBroadcastTestLodTensor) {
TestBroadcastOpHandle test_op;
size_t input_scope_idx = 0;
......
paddle/fluid/framework/details/broadcast_op_handle_test.h 0 → 100644
浏览文件 @ bc34d325
// 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.
#pragma once
#include <string>
#include <vector>
#include "gtest/gtest.h"
#include "paddle/fluid/framework/details/broadcast_op_handle.h"
#include "paddle/fluid/platform/device_context.h"
namespace paddle {
namespace framework {
namespace details {
namespace f = paddle::framework;
namespace p = paddle::platform;
// test data amount
const f::DDim kDims = {20, 20};
struct TestBroadcastOpHandle {
std::vector<std::unique_ptr<p::DeviceContext>> ctxs_;
std::vector<Scope*> local_scopes_;
std::vector<Scope*> param_scopes_;
Scope g_scope_;
OpHandleBase* op_handle_;
std::vector<VarHandleBase*> vars_;
std::vector<std::unique_ptr<ir::Node>> nodes_;
std::vector<p::Place> place_list_;
bool use_gpu_;
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
std::unique_ptr<platform::NCCLContextMap> nccl_ctxs_;
#endif
void WaitAll() {
for (size_t j = 0; j < ctxs_.size(); ++j) {
ctxs_[j]->Wait();
}
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
if (nccl_ctxs_) {
nccl_ctxs_->WaitAll();
}
#endif
}
void InitCtxOnGpu(bool use_gpu) {
use_gpu_ = use_gpu;
if (use_gpu_) {
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
int count = p::GetCUDADeviceCount();
if (count <= 1) {
LOG(WARNING) << "Cannot test multi-gpu Broadcast, because the CUDA "
"device count is "
<< count;
exit(0);
}
for (int i = 0; i < count; ++i) {
auto p = p::CUDAPlace(i);
place_list_.push_back(p);
ctxs_.emplace_back(new p::CUDADeviceContext(p));
}
nccl_ctxs_.reset(new platform::NCCLContextMap(place_list_));
#else
PADDLE_THROW("CUDA is not support.");
#endif
} else {
int count = 8;
for (int i = 0; i < count; ++i) {
auto p = p::CPUPlace();
place_list_.push_back(p);
ctxs_.emplace_back(new p::CPUDeviceContext(p));
}
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
nccl_ctxs_.reset(nullptr);
#endif
}
}
void InitBroadcastOp(size_t input_scope_idx) {
nodes_.clear();
for (size_t j = 0; j < place_list_.size(); ++j) {
local_scopes_.push_back(&(g_scope_.NewScope()));
Scope& local_scope = local_scopes_.back()->NewScope();
*local_scopes_.back()
->Var(details::kLocalExecScopeName)
->GetMutable<Scope*>() = &local_scope;
local_scope.Var("out");
param_scopes_.emplace_back(&local_scope);
}
param_scopes_[input_scope_idx]->Var("input");
nodes_.emplace_back(
ir::CreateNodeForTest("node0", ir::Node::Type::kOperation));
if (use_gpu_) {
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
op_handle_ = new BroadcastOpHandle(nodes_.back().get(), local_scopes_,
place_list_, nccl_ctxs_.get());
#else
PADDLE_THROW("CUDA is not support.");
#endif
} else {
#if defined(PADDLE_WITH_CUDA) && !defined(_WIN32)
op_handle_ = new BroadcastOpHandle(nodes_.back().get(), local_scopes_,
place_list_, nccl_ctxs_.get());
#else
op_handle_ = new BroadcastOpHandle(nodes_.back().get(), local_scopes_,
place_list_);
#endif
}
nodes_.emplace_back(
ir::CreateNodeForTest("node1", ir::Node::Type::kVariable));
auto* in_var_handle = new VarHandle(nodes_.back().get(), 1, input_scope_idx,
"input", place_list_[input_scope_idx]);
vars_.emplace_back(in_var_handle);
op_handle_->AddInput(in_var_handle);
// add dummy var
nodes_.emplace_back(
ir::CreateNodeForTest("node2", ir::Node::Type::kVariable));
vars_.emplace_back(new DummyVarHandle(nodes_.back().get()));
DummyVarHandle* dummy_var_handle =
static_cast<DummyVarHandle*>(vars_.back());
dummy_var_handle->ClearGeneratedOp();
op_handle_->AddInput(dummy_var_handle);
for (size_t j = 0; j < place_list_.size(); ++j) {
if (!use_gpu_) {
op_handle_->SetDeviceContext(place_list_[j], ctxs_[j].get());
}
nodes_.emplace_back(
ir::CreateNodeForTest("node3", ir::Node::Type::kVariable));
VarHandle* out_var_handle =
new VarHandle(nodes_.back().get(), 2, j, "out", place_list_[j]);
vars_.emplace_back(out_var_handle);
op_handle_->AddOutput(out_var_handle);
}
// add dummy var
nodes_.emplace_back(
ir::CreateNodeForTest("node4", ir::Node::Type::kVariable));
vars_.emplace_back(new DummyVarHandle(nodes_.back().get()));
DummyVarHandle* out_dummy_var_handle =
static_cast<DummyVarHandle*>(vars_.back());
out_dummy_var_handle->ClearGeneratedOp();
op_handle_->AddOutput(out_dummy_var_handle);
}
std::vector<float> InitLoDTensor(const std::string& varname,
size_t input_scope_idx, const f::LoD& lod,
float val_scalar = 0.0) {
auto var = param_scopes_[input_scope_idx]->FindVar(varname);
PADDLE_ENFORCE_NOT_NULL(var);
auto lod_tensor = var->GetMutable<f::LoDTensor>();
std::vector<float> send_vector(static_cast<size_t>(f::product(kDims)));
for (size_t k = 0; k < send_vector.size(); ++k) {
send_vector[k] = k + val_scalar;
}
paddle::framework::TensorFromVector<float>(
send_vector, *(ctxs_[input_scope_idx]), lod_tensor);
lod_tensor->set_lod(lod);
lod_tensor->Resize(kDims);
return send_vector;
}
std::vector<float> InitSelectedRows(const std::string& varname,
size_t input_scope_idx,
const std::vector<int64_t>& rows,
int height, float value_scalar = 0.0) {
std::vector<float> send_vector(static_cast<size_t>(f::product(kDims)));
for (size_t k = 0; k < send_vector.size(); ++k) {
send_vector[k] = k + value_scalar;
}
auto var = param_scopes_[input_scope_idx]->FindVar(varname);
PADDLE_ENFORCE_NOT_NULL(var);
auto selected_rows = var->GetMutable<f::SelectedRows>();
auto value = selected_rows->mutable_value();
value->mutable_data<float>(kDims, place_list_[input_scope_idx]);
selected_rows->set_height(height);
selected_rows->set_rows(rows);
paddle::framework::TensorFromVector<float>(
send_vector, *(ctxs_[input_scope_idx]), value);
return send_vector;
}
void SelectedRowsEqual(const std::string& varname, int input_scope_idx,
const std::vector<float>& send_vector,
const std::vector<int64_t>& rows, int height) {
auto var = param_scopes_[input_scope_idx]->FindVar(varname);
PADDLE_ENFORCE_NOT_NULL(var);
auto& selected_rows = var->Get<f::SelectedRows>();
auto rt = selected_rows.value();
PADDLE_ENFORCE_EQ(selected_rows.height(), height, "height is not equal.");
for (size_t k = 0; k < selected_rows.rows().size(); ++k) {
PADDLE_ENFORCE_EQ(selected_rows.rows()[k], rows[k]);
}
p::CPUPlace cpu_place;
f::Tensor result_tensor;
f::TensorCopySync(rt, cpu_place, &result_tensor);
float* ct = result_tensor.data<float>();
for (int64_t i = 0; i < f::product(kDims); ++i) {
ASSERT_NEAR(ct[i], send_vector[i], 1e-5);
}
}
void LoDTensorEqual(const std::string& varname,
const std::vector<float>& send_vec, const f::LoD& lod,
framework::Scope* scope) {
p::CPUPlace cpu_place;
auto var = scope->FindVar(varname);
PADDLE_ENFORCE_NOT_NULL(var);
auto tensor = var->Get<f::LoDTensor>();
PADDLE_ENFORCE_EQ(tensor.lod(), lod, "lod is not equal.");
f::Tensor result_tensor;
f::TensorCopySync(tensor, cpu_place, &result_tensor);
float* ct = result_tensor.mutable_data<float>(cpu_place);
for (int64_t k = 0; k < f::product(kDims); ++k) {
ASSERT_NEAR(ct[k], send_vec[k], 1e-5);
}
}
void TestBroadcastLodTensor(size_t input_scope_idx) {
f::LoD lod{{0, 10, 20}};
auto send_vector = InitLoDTensor("input", input_scope_idx, lod);
op_handle_->Run(false);
WaitAll();
for (size_t j = 0; j < place_list_.size(); ++j) {
LoDTensorEqual("out", send_vector, lod, param_scopes_[j]);
}
}
void TestBroadcastSelectedRows(size_t input_scope_idx) {
std::vector<int64_t> rows{0, 1, 2, 3, 3, 0, 14, 7, 3, 1,
2, 4, 6, 3, 1, 1, 1, 1, 3, 7};
int height = static_cast<int>(kDims[0] * 2);
auto send_vector = InitSelectedRows("input", input_scope_idx, rows, height);
op_handle_->Run(false);
WaitAll();
for (size_t j = 0; j < place_list_.size(); ++j) {
SelectedRowsEqual("out", input_scope_idx, send_vector, rows, height);
}
}
};
} // namespace details
} // namespace framework
} // namespace paddle
paddle/fluid/framework/details/computation_op_handle.cc
浏览文件 @ bc34d325
......@@ -29,15 +29,21 @@ ComputationOpHandle::ComputationOpHandle(ir::Node *node, Scope *scope,
void ComputationOpHandle::RunImpl() {
WaitInputVarGenerated(place_);
this->RunAndRecordEvent([this] {
auto run_func = [this]() {
op_->Run(*scope_->FindVar(kLocalExecScopeName)->Get<Scope *>(), place_);
});
};
if (is_lock_and_record_event_free_) {
run_func();
} else {
this->RunAndRecordEvent(run_func);
}
}
bool ComputationOpHandle::NeedWait(VarHandleBase *in_var) {
bool need_wait =
in_var && in_var->GeneratedOp() &&
in_var->GeneratedOp()->DeviceContext(place_) != dev_ctxes_[place_];
in_var->GeneratedOp()->DeviceContext(place_) != dev_ctxes_.at(place_);
return need_wait;
}
......
paddle/fluid/framework/details/computation_op_handle.h
浏览文件 @ bc34d325
......@@ -36,6 +36,8 @@ struct ComputationOpHandle : public OpHandleBase {
const platform::Place &GetPlace() const { return place_; }
void SetLockAndRecordEventFree(bool b) { is_lock_and_record_event_free_ = b; }
protected:
void RunImpl() override;
......@@ -45,6 +47,7 @@ struct ComputationOpHandle : public OpHandleBase {
std::unique_ptr<OperatorBase> op_;
Scope *scope_;
platform::Place place_;
bool is_lock_and_record_event_free_{false};
};
} // namespace details
} // namespace framework
......
paddle/fluid/framework/details/op_graph_view.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/details/op_graph_view.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/dlpack_tensor.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/dlpack_tensor.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/dlpack_tensor_test.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/eigen.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/executor.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/executor.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/executor_thread_worker.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/executor_thread_worker.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/depthwise_conv_mkldnn_pass.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/depthwise_conv_mkldnn_pass.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/graph.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/graph.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/is_test_pass.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/is_test_pass_tester.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/multi_batch_merge_pass.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/multi_batch_merge_pass.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/node.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/node.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/node_test.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ir/pass.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/lod_tensor.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ngraph_bridge.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ngraph_bridge.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ngraph_operator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/ngraph_operator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/op_desc.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/op_desc.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/operator.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/operator.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/scope.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/scope.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/tensor.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/tensor.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/threadpool.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/transfer_scope_cache.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/transfer_scope_cache.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/type_defs.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/variable_helper.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/framework/variable_helper.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/analysis/node.cc 已删除 100644 → 0
浏览文件 @ e837a6de
此差异已折叠。
paddle/fluid/inference/analysis/node.h 已删除 100644 → 0
浏览文件 @ e837a6de
此差异已折叠。
paddle/fluid/inference/analysis/passes/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/analysis/passes/passes.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/analysis_config.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/api.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/details/reset_tensor_array.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/helper.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/paddle_anakin_config.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/paddle_analysis_config.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/paddle_api.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/paddle_pass_builder.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/api/paddle_pass_builder.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/io.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tensorrt/convert/prelu_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tensorrt/convert/split_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tensorrt/plugin/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tensorrt/plugin/serialize.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tensorrt/plugin/trt_plugin.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tensorrt/plugin/trt_plugin.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tests/api/analyzer_dam_tester.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tests/api/config_printer.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/tests/test.cmake 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/utils/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/utils/benchmark.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/utils/benchmark.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/inference/utils/benchmark_tester.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/CMakeLists.txt
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/aligned_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/allocator_facade.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/allocator_facade_test.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/allocator_strategy.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/allocator_strategy.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/best_fit_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/best_fit_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/buffered_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/buffered_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/conditional_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/cpu_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/cpu_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/cuda_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/cuda_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/legacy_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/legacy_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/locked_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/locked_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/pinned_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/pinned_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/retry_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/retry_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/retry_allocator_test.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/zero_size_allocator.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/allocation/zero_size_allocator.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/malloc.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/malloc.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/memory/malloc_test.cc 已删除 100644 → 0
浏览文件 @ e837a6de
此差异已折叠。
paddle/fluid/memory/memcpy.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/add_position_encoding_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/add_position_encoding_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/affine_grid_cudnn_op.cu.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/affine_grid_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/affine_grid_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/bilinear_interp_op.h 已删除 100644 → 0
浏览文件 @ e837a6de
此差异已折叠。
paddle/fluid/operators/controlflow/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/conv_cudnn_op_cache.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/conv_fusion_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/conv_fusion_op.cu.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/conv_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/conv_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/conv_transpose_mkldnn_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/csp/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/detection/density_prior_box_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/distributed/parameter_prefetch.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/distributed_ops/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/distributed_ops/fake_init_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/dropout_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/elementwise/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/expand_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/extract_rows_op.cc 已删除 100644 → 0
浏览文件 @ e837a6de
此差异已折叠。
paddle/fluid/operators/fc_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/fused/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/gather.cu.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/gather.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/gather_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/gather_op.cu
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/grid_sampler_cudnn_op.cu.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/grid_sampler_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/grid_sampler_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/group_norm_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/group_norm_op.cu 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/group_norm_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/hash_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/hash_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/interpolate_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/load_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/lrn_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/lrn_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/blas.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_code.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_code.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_gen.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_gen.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_kernel_crf_decode.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_kernel_impl.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_kernel_layer_norm.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/math/jit_kernel_refer.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/mean_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/mean_op.cu
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/mean_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/metrics/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/mul_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/mul_op.cu.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/nce_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/nce_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/optimizers/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/optimizers/lars_momentum_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/optimizers/lars_momentum_op.cu 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/optimizers/lars_momentum_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/pad2d_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/pad2d_op.cu
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/pool_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/pool_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/prelu_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reader/create_ctr_reader_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reader/ctr_reader.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reader/ctr_reader.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reader/ctr_reader_test.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reduce_ops/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reduce_ops/reduce_max_op.part.cu 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reduce_ops/reduce_min_op.part.cu 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/reduce_ops/reduce_sum_op.part.cu 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/scale_op.cu
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/scale_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/scatter.cu.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/scatter.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/selu_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/selu_op.cu 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/selu_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/sequence_ops/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/sign_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/sign_op.cu
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/softmax_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/space_to_depth_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/space_to_depth_op.cu 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/space_to_depth_op.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/spp_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/stack_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/stack_op.cu
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/stack_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/sum_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/sum_op.cu
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/sum_op.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/tensor_array_to_tensor_op.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/tensorrt/CMakeLists.txt 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/unpool_op.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/operators/warpctc_cudnn_op.cu.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/cpu_helper.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/cpu_info.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/cpu_info.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/cuda_device_guard.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/cuda_device_guard.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/enforce.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/float16.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/gpu_info.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/gpu_info.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/init.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/init.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/lock_guard_ptr.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/mkldnn_reuse.h 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/nccl_helper.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/place.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/port.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/profiler.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/profiler.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/platform/variant.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/pybind/CMakeLists.txt
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/pybind/async_executor_py.cc 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/pybind/const_value.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/pybind/protobuf.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/pybind/pybind.cc
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/pybind/tensor_py.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/fluid/string/printf.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/legacy/utils/ThreadLocal.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/legacy/utils/Util.h
浏览文件 @ bc34d325
此差异已折叠。
paddle/scripts/paddle_build.sh
浏览文件 @ bc34d325
此差异已折叠。
paddle/testing/CMakeLists.txt
浏览文件 @ bc34d325
此差异已折叠。
paddle/testing/paddle_gtest_main.cc
浏览文件 @ bc34d325
此差异已折叠。
python/CMakeLists.txt
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/dataset/wmt16.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/__init__.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/async_executor.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/clip.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/contrib/reader/ctr_reader.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/contrib/utils/__init__.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/contrib/utils/hdfs_utils.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/contrib/utils/lookup_table_utils.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/data_feed_desc.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/distribute_lookup_table.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/evaluator.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/executor.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/framework.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/initializer.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/io.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/layers/io.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/layers/nn.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/layers/ops.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/metrics.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/nets.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/op.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/optimizer.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/regularizer.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/tests/demo/async_executor.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/tests/unittests/dist_mnist_lars.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/tests/unittests/dist_save_load.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/tests/unittests/test_hash_op.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/tests/unittests/test_metrics.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/fluid/tests/unittests/test_selu_op.py 0 → 100644
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/utils/__init__.py
浏览文件 @ bc34d325
此差异已折叠。
python/paddle/v2/dataset/wmt16.py
浏览文件 @ bc34d325
此差异已折叠。
python/requirements.txt
浏览文件 @ bc34d325
此差异已折叠。
python/setup.py.in
浏览文件 @ bc34d325
此差异已折叠。
tools/manylinux1/Dockerfile.x64
浏览文件 @ bc34d325
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