fix conflict

Author: Shixiaowei02

CMakeLists.txt

@@ -59,19 +59,14 @@ option(WITH_DISTRIBUTE  "Compile with distributed support"              OFF)
 option(WITH_BRPC_RDMA     "Use brpc rdma as the rpc protocal"           OFF)
 option(ON_INFER         "Turn on inference optimization."               OFF)
 ################################ Internal Configurations #######################################
-option(WITH_ANAKIN      "Compile with Anakin library"                   OFF)
 option(WITH_AMD_GPU     "Compile PaddlePaddle with AMD GPU"             OFF)
 option(WITH_NGRAPH      "Compile PaddlePaddle with nGraph support."     OFF)
 option(WITH_PROFILER    "Compile PaddlePaddle with GPU profiler and gperftools"        OFF)
-option(WITH_JEMALLOC    "Compile PaddlePaddle with jemalloc"            OFF)
 option(WITH_COVERAGE    "Compile PaddlePaddle with code coverage"       OFF)
 option(COVERALLS_UPLOAD "Package code coverage data to coveralls"       OFF)
 option(WITH_PSLIB       "Compile with pslib support"                    OFF)
 option(WITH_CONTRIB     "Compile the third-party contributation"        OFF)
 option(REPLACE_ENFORCE_GLOG "Replace PADDLE_ENFORCE with glog/CHECK for better debug." OFF)
-# TODO(Superjomn) Remove WITH_ANAKIN option if not needed latter.
-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_INFERENCE_API_TEST   "Test fluid inference C++ high-level api interface"  OFF)
 option(WITH_HIGH_LEVEL_API_TEST   "Test fluid python high-level api interface"  OFF)
@@ -185,6 +180,7 @@ if(WITH_BRPC_RDMA)
     endif()
 endif()
 
+include(anakin_subgraph)
 
 include(external/threadpool)
 include(flags)              # set paddle compile flags
@@ -207,24 +203,12 @@ if(WITH_DGC)
     add_definitions(-DPADDLE_WITH_DGC)
 endif()
 
-if(WITH_MKL OR WITH_MKLML)
-    include(external/anakin)
-elseif()
-    set(WITH_ANAKIN OFF CACHE STRING "Anakin is used in MKL only now." FORCE)
-endif()
-
 if (WITH_PROFILER)
     find_package(Gperftools REQUIRED)
     include_directories(${GPERFTOOLS_INCLUDE_DIR})
     add_definitions(-DWITH_GPERFTOOLS)
 endif()
 
-if (WITH_JEMALLOC)
-    find_package(JeMalloc REQUIRED)
-    include_directories(${JEMALLOC_INCLUDE_DIR})
-    add_definitions(-DPADDLE_WITH_JEMALLOC)
-endif()
-
 include(generic)            # simplify cmake module
 include(package)            # set paddle packages
 include(ccache)             # set ccache for compilation

Dockerfile

@@ -1,7 +1,6 @@
 # A image for building paddle binaries
 # Use cuda devel base image for both cpu and gpu environment
 # When you modify it, please be aware of cudnn-runtime version
-# and libcudnn.so.x in paddle/scripts/docker/build.sh
 FROM nvidia/cuda:8.0-cudnn7-devel-ubuntu16.04
 MAINTAINER PaddlePaddle Authors <paddle-dev@baidu.com>
 
@@ -76,7 +75,7 @@ RUN curl -s -q https://glide.sh/get | sh
 # 2. Manually add ~IPluginFactory() in IPluginFactory class of NvInfer.h, otherwise, it couldn't work in paddle.
 #    See https://github.com/PaddlePaddle/Paddle/issues/10129 for details.
 
-RUN wget -q https://paddlepaddledeps.cdn.bcebos.com/TensorRT-4.0.1.6-ubuntu14.04.x86_64-gnu.cuda.8.0.cudnn7.0.tar.gz --no-check-certificate && \
+RUN wget -q https://paddlepaddledeps.bj.bcebos.com/TensorRT-4.0.1.6-ubuntu14.04.x86_64-gnu.cuda.8.0.cudnn7.0.tar.gz --no-check-certificate && \
     tar -zxf TensorRT-4.0.1.6-ubuntu14.04.x86_64-gnu.cuda.8.0.cudnn7.0.tar.gz -C /usr/local && \
     cp -rf /usr/local/TensorRT/include /usr && \
     cp -rf /usr/local/TensorRT/lib /usr

README.md

@@ -18,20 +18,21 @@ learning to many products at Baidu.
 Our vision is to enable deep learning for everyone via PaddlePaddle.
 Please refer to our [release announcement](https://github.com/PaddlePaddle/Paddle/releases) to track the latest feature of PaddlePaddle.
 
-### Latest PaddlePaddle Release: [Fluid 1.4.1](https://github.com/PaddlePaddle/Paddle/tree/release/1.4)
+### Latest PaddlePaddle Release: [Fluid 1.5.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.5)
 ### Install Latest Stable Release:
 ```
 # Linux CPU
 pip install paddlepaddle
 # Linux GPU cuda9cudnn7
 pip install paddlepaddle-gpu
+# Linux GPU cuda10cudnn7
+pip install paddlepaddle-gpu==1.5.0.post107
 # Linux GPU cuda8cudnn7
-pip install paddlepaddle-gpu==1.4.1.post87
-# Linux GPU cuda8cudnn5
-pip install paddlepaddle-gpu==1.4.1.post85
+pip install paddlepaddle-gpu==1.5.0.post87
 
 # For installation on other platform, refer to http://paddlepaddle.org/
 ```
+Now our developers could acquire Tesla V100 online computing resources for free. If you create a program by AI Studio, you would obtain 12 hours to train models online per day. If you could insist on that for five consecutive days, then you would own extra 48 hours. [Click here to start](http://ai.baidu.com/support/news?action=detail&id=981).
 
 ## Features
 
@@ -98,9 +99,11 @@ We provide [English](http://www.paddlepaddle.org/documentation/docs/en/1.4/begin
 
    We appreciate your contributions!
 
-## Ask Questions
+## Communication
 
-You are welcome to submit questions and bug reports as [Github Issues](https://github.com/PaddlePaddle/Paddle/issues).
+- [Github Issues](https://github.com/PaddlePaddle/Paddle/issues): bug reports, feature requests, install issues, usage issues, etc.
+- QQ discussion group: 432676488 (PaddlePaddle).
+- [Forums](http://ai.baidu.com/forum/topic/list/168?pageNo=1): discuss implementations, research, etc.
 
 ## Copyright and License
 PaddlePaddle is provided under the [Apache-2.0 license](LICENSE).

README_cn.md

@@ -16,20 +16,21 @@ PaddlePaddle (PArallel Distributed Deep LEarning) 是一个简单易用、高效
 
 跟进PaddlePaddle最新特性请参考我们的[版本说明](https://github.com/PaddlePaddle/Paddle/releases)
 
-### PaddlePaddle最新版本: [Fluid 1.4.1](https://github.com/PaddlePaddle/Paddle/tree/release/1.4)
+### PaddlePaddle最新版本: [Fluid 1.5.0](https://github.com/PaddlePaddle/Paddle/tree/release/1.5)
 ### 安装最新稳定版本:
 ```
 # Linux CPU
 pip install paddlepaddle
 # Linux GPU cuda9cudnn7
 pip install paddlepaddle-gpu
+# Linux GPU cuda10cudnn7
+pip install paddlepaddle-gpu==1.5.0.post107
 # Linux GPU cuda8cudnn7
-pip install paddlepaddle-gpu==1.4.1.post87
-# Linux GPU cuda8cudnn5
-pip install paddlepaddle-gpu==1.4.1.post85
+pip install paddlepaddle-gpu==1.5.0.post87
 
 # 其他平台上的安装指引请参考 http://paddlepaddle.org/
 ```
+PaddlePaddle用户可领取**免费Tesla V100在线算力资源**，训练模型更高效。**每日登陆即送12小时**，**连续五天运行再加送48小时**，[前往使用免费算力](https://ai.baidu.com/support/news?action=detail&id=981)。
 
 ## 特性
 
@@ -80,9 +81,11 @@ pip install paddlepaddle-gpu==1.4.1.post85
 
    欢迎您的贡献!
 
-## 答疑
+## 交流与反馈
 
-欢迎您将问题和bug报告以[Github Issues](https://github.com/PaddlePaddle/Paddle/issues)的形式提交
+- 欢迎您通过[Github Issues](https://github.com/PaddlePaddle/Paddle/issues)来提交问题、报告与建议
+- QQ群: 432676488 (PaddlePaddle)
+- [论坛](http://ai.baidu.com/forum/topic/list/168): 欢迎大家在PaddlePaddle论坛分享在使用PaddlePaddle中遇到的问题和经验, 营造良好的论坛氛围
 
 ## 版权和许可证
 PaddlePaddle由[Apache-2.0 license](LICENSE)提供

benchmark/fluid/README.md

@@ -59,7 +59,7 @@ python -c 'from recordio_converter import *; prepare_mnist("data", 1)'
 ## Run Distributed Benchmark on Kubernetes Cluster
 
 You may need to build a Docker image before submitting a cluster job onto Kubernetes, or you will
-have to start all those processes mannually on each node, which is not recommended.
+have to start all those processes manually on each node, which is not recommended.
 
 To build the Docker image, you need to choose a paddle "whl" package to run with, you may either
 download it from

cmake/FindJeMalloc.cmake

@@ -30,6 +30,7 @@ endif(NOT WITH_PROFILER)
 
 if(WITH_AVX AND AVX_FOUND)
     set(SIMD_FLAG ${AVX_FLAG})
+    add_definitions(-DPADDLE_WITH_AVX)
 elseif(SSE3_FOUND)
     set(SIMD_FLAG ${SSE3_FLAG})
 endif()
@@ -98,23 +99,15 @@ if(WITH_GPU)
         endif()
         include_directories(${TENSORRT_INCLUDE_DIR})
     endif()
-    if(WITH_ANAKIN)
+    if(ANAKIN_FOUND)
         if(${CUDA_VERSION_MAJOR} VERSION_LESS 8)
-            message(WARNING "Anakin needs CUDA >= 8.0 to compile. Force WITH_ANAKIN=OFF")
-            set(WITH_ANAKIN OFF CACHE STRING "Anakin is valid only when CUDA >= 8.0." FORCE)
+            message(WARNING "Anakin needs CUDA >= 8.0 to compile. Force ANAKIN_FOUND = OFF")
+            set(ANAKIN_FOUND OFF CACHE STRING "Anakin is valid only when CUDA >= 8.0." FORCE)
         endif()
         if(${CUDNN_MAJOR_VERSION} VERSION_LESS 7)
-            message(WARNING "Anakin needs CUDNN >= 7.0 to compile. Force WITH_ANAKIN=OFF")
-            set(WITH_ANAKIN OFF CACHE STRING "Anakin is valid only when CUDNN >= 7.0." FORCE)
+            message(WARNING "Anakin needs CUDNN >= 7.0 to compile. Force ANAKIN_FOUND = OFF")
+            set(ANAKIN_FOUND OFF CACHE STRING "Anakin is valid only when CUDNN >= 7.0." FORCE)
         endif()
-        add_definitions(-DWITH_ANAKIN)
-    endif()
-    if(WITH_ANAKIN)
-        # NOTICE(minqiyang): the end slash is important because $CUDNN_INCLUDE_DIR
-        # is a softlink to real cudnn.h directory
-        set(ENV{CUDNN_INCLUDE_DIR} "${CUDNN_INCLUDE_DIR}/")
-        get_filename_component(CUDNN_LIBRARY_DIR ${CUDNN_LIBRARY} DIRECTORY)
-        set(ENV{CUDNN_LIBRARY} ${CUDNN_LIBRARY_DIR})
     endif()
 elseif(WITH_AMD_GPU)
     add_definitions(-DPADDLE_WITH_HIP)

cmake/configure.cmake

@@ -141,31 +141,27 @@ endfunction()
 message(STATUS "CUDA detected: " ${CUDA_VERSION})
 if (${CUDA_VERSION} LESS 7.0)
   set(paddle_known_gpu_archs ${paddle_known_gpu_archs})
-  add_definitions("-DPADDLE_CUDA_BINVER=\"60\"")
 elseif (${CUDA_VERSION} LESS 8.0) # CUDA 7.x
   set(paddle_known_gpu_archs ${paddle_known_gpu_archs7})
   list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
   list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"70\"")
 elseif (${CUDA_VERSION} LESS 9.0) # CUDA 8.x
   set(paddle_known_gpu_archs ${paddle_known_gpu_archs8})
   list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
   list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
   # CUDA 8 may complain that sm_20 is no longer supported. Suppress the
   # warning for now.
   list(APPEND CUDA_NVCC_FLAGS "-Wno-deprecated-gpu-targets")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"80\"")
 elseif (${CUDA_VERSION} LESS 10.0) # CUDA 9.x
   set(paddle_known_gpu_archs ${paddle_known_gpu_archs9})
   list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
   list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"90\"")
 elseif (${CUDA_VERSION} LESS 11.0) # CUDA 10.x
   set(paddle_known_gpu_archs ${paddle_known_gpu_archs10})
   list(APPEND CUDA_NVCC_FLAGS "-D_MWAITXINTRIN_H_INCLUDED")
   list(APPEND CUDA_NVCC_FLAGS "-D__STRICT_ANSI__")
-  add_definitions("-DPADDLE_CUDA_BINVER=\"100\"")
 endif()
+add_definitions("-DPADDLE_CUDA_BINVER=\"${CUDA_VERSION_MAJOR}${CUDA_VERSION_MINOR}\"")
 
 include_directories(${CUDA_INCLUDE_DIRS})
 if(NOT WITH_DSO)

cmake/cuda.cmake

@@ -96,7 +96,7 @@ if(CUDNN_FOUND)
         endif()
 
         message(STATUS "Current cuDNN header is ${CUDNN_INCLUDE_DIR}/cudnn.h. "
-            "Current cuDNN version is v${CUDNN_MAJOR_VERSION}. ")
+            "Current cuDNN version is v${CUDNN_MAJOR_VERSION}.${CUDNN_MINOR_VERSION}. ")
 
     endif()
 endif()

cmake/cudnn.cmake

@@ -38,5 +38,3 @@ ADD_LIBRARY(dgc STATIC IMPORTED GLOBAL)
 SET_PROPERTY(TARGET dgc PROPERTY IMPORTED_LOCATION ${DGC_LIBRARIES})
 ADD_DEPENDENCIES(dgc extern_dgc)
 
-LIST(APPEND external_project_dependencies dgc)
-

cmake/external/anakin.cmake

@@ -12,6 +12,13 @@ if(NOT WITH_FAST_MATH)
   add_definitions(-DEIGEN_FAST_MATH=0)
 endif()
 
+if(WIN32)
+    set(EIGEN_GIT_REPOSITORY https://github.com/wopeizl/eigen-git-mirror)
+    set(EIGEN_GIT_TAG support_cuda9_win)
+else()
+    set(EIGEN_GIT_REPOSITORY https://github.com/eigenteam/eigen-git-mirror)
+    set(EIGEN_GIT_TAG 917060c364181f33a735dc023818d5a54f60e54c)
+endif()
 if(WITH_AMD_GPU)
     ExternalProject_Add(
         extern_eigen3
@@ -29,10 +36,10 @@ else()
     ExternalProject_Add(
         extern_eigen3
         ${EXTERNAL_PROJECT_LOG_ARGS}
-        GIT_REPOSITORY  "https://github.com/eigenteam/eigen-git-mirror"
+        GIT_REPOSITORY  "${EIGEN_GIT_REPOSITORY}"
         # eigen on cuda9.1 missing header of math_funtions.hpp
         # https://stackoverflow.com/questions/43113508/math-functions-hpp-not-found-when-using-cuda-with-eigen
-        GIT_TAG         917060c364181f33a735dc023818d5a54f60e54c
+        GIT_TAG         ${EIGEN_GIT_TAG}
         PREFIX          ${EIGEN_SOURCE_DIR}
         DOWNLOAD_NAME   "eigen"
         UPDATE_COMMAND  ""

cmake/external/dgc.cmake

@@ -18,7 +18,7 @@ SET(GFLAGS_SOURCES_DIR ${THIRD_PARTY_PATH}/gflags)
 SET(GFLAGS_INSTALL_DIR ${THIRD_PARTY_PATH}/install/gflags)
 SET(GFLAGS_INCLUDE_DIR "${GFLAGS_INSTALL_DIR}/include" CACHE PATH "gflags include directory." FORCE)
 IF(WIN32)
-  set(GFLAGS_LIBRARIES "${GFLAGS_INSTALL_DIR}/lib/libgflags.lib" CACHE FILEPATH "GFLAGS_LIBRARIES" FORCE)
+  set(GFLAGS_LIBRARIES "${GFLAGS_INSTALL_DIR}/lib/gflags_static.lib" CACHE FILEPATH "GFLAGS_LIBRARIES" FORCE)
 ELSE(WIN32)
   set(GFLAGS_LIBRARIES "${GFLAGS_INSTALL_DIR}/lib/libgflags.a" CACHE FILEPATH "GFLAGS_LIBRARIES" FORCE)
 ENDIF(WIN32)
@@ -50,13 +50,6 @@ ExternalProject_Add(
                      -DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
                      -DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
 )
-IF(WIN32)
-  IF(NOT EXISTS "${GFLAGS_INSTALL_DIR}/lib/libgflags.lib")
-    add_custom_command(TARGET extern_gflags POST_BUILD
-            COMMAND cmake -E copy ${GFLAGS_INSTALL_DIR}/lib/gflags_static.lib ${GFLAGS_INSTALL_DIR}/lib/libgflags.lib
-            )
-  ENDIF()
-ENDIF(WIN32)
 ADD_LIBRARY(gflags STATIC IMPORTED GLOBAL)
 SET_PROPERTY(TARGET gflags PROPERTY IMPORTED_LOCATION ${GFLAGS_LIBRARIES})
 ADD_DEPENDENCIES(gflags extern_gflags)