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

Author: sneaxiy

.travis.yml

@@ -27,15 +27,6 @@ script:
     # 43min timeout
     paddle/scripts/paddle_docker_build.sh ${JOB}
     if [ $? -eq 0 ] || [ $? -eq 142 ]; then true; else exit 1; fi;
-  - |
-    if [[ "$JOB" != "doc" ]]; then exit 0; fi;
-    # For document only
-    if [[ "$TRAVIS_PULL_REQUEST" != "false" ]]; then exit 0; fi;
-    if [[ "$TRAVIS_BRANCH" != "develop"  && ! "$TRAVIS_BRANCH" =~ ^v[[:digit:]]+\.[[:digit:]]+(\.[[:digit:]]+)?(-\S*)?$ ]]; then exit 0; fi;
-    export DEPLOY_DOCS_SH=https://raw.githubusercontent.com/PaddlePaddle/PaddlePaddle.org/master/scripts/deploy/deploy_docs.sh
-    export DOCS_DIR=`pwd`
-    cd ..
-    curl $DEPLOY_DOCS_SH | bash -s $CONTENT_DEC_PASSWD $TRAVIS_BRANCH $DOCS_DIR $DOCS_DIR/build/doc/
 notifications:
   email:
     on_success: change

AUTHORS.md

@@ -46,6 +46,7 @@
 | tianbingsz | Tian-Bing Xu |
 | tpatejko | Tomasz Patejko |
 | typhoonzero | Yi Wu |
+| velconia | Qi-Yang Min |
 | wanghaoshuang | Hao-Shuang Wang |
 | wangyang59 | Yang Wang |
 | wangzhen-nlp | Zhen Wang |

CMakeLists.txt

@@ -65,7 +65,14 @@ option(REPLACE_ENFORCE_GLOG "Replace PADDLE_ENFORCE with glog/CHECK for better d
 option(WITH_ANAKIN      "Compile with Anakin library"                   OFF)
 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(WITH_SYSTEM_BLAS   "Use system blas library"           OFF)
+option(PY_VERSION       "Compile PaddlePaddle with python3 support"     ${PY_VERSION})
+
+# PY_VERSION
+if(NOT PY_VERSION)
+  set(PY_VERSION 2.7)
+endif()
 
 # CMAKE_BUILD_TYPE
 if(NOT CMAKE_BUILD_TYPE)
@@ -103,6 +110,11 @@ if(ANDROID OR IOS)
     add_definitions(-DPADDLE_MOBILE_INFERENCE)
 endif()
 
+if (APPLE OR WIN32)
+    set(WITH_MKL OFF CACHE STRING
+        "Disable MKL for building on mac and windows" FORCE)
+endif()
+
 set(THIRD_PARTY_PATH "${CMAKE_BINARY_DIR}/third_party" CACHE STRING
   "A path setting third party libraries download & build directories.")
 
@@ -125,6 +137,12 @@ else()
     set(THIRD_PARTY_BUILD_TYPE Release)
 endif()
 
+if(WITH_MKL)
+  option(MKL_SPLIT_GEMM "PaddlePaddle MKL gemm would split to small ones" OFF)
+  if (MKL_SPLIT_GEMM)
+    add_definitions(-DPADDLE_MKL_SPLIT_GEMM)
+  endif()
+endif()
 set(WITH_MKLML ${WITH_MKL})
 if (NOT DEFINED WITH_MKLDNN)
     if (WITH_MKL AND AVX2_FOUND)
@@ -141,6 +159,7 @@ endif()
 ########################################################################################
 
 include(external/mklml)     # download mklml package
+include(external/libxsmm)   # download, build, install libxsmm
 include(external/zlib)      # download, build, install zlib
 include(external/gflags)    # download, build, install gflags
 include(external/glog)      # download, build, install glog
@@ -156,6 +175,7 @@ include(external/any)       # download libn::any
 include(external/eigen)     # download eigen3
 include(external/pybind11)  # download pybind11
 include(external/cares)
+include(external/cub)
 
 if(WITH_DISTRIBUTE)
     if(WITH_GRPC)
@@ -182,6 +202,14 @@ include(external/snappy)    # download snappy
 include(external/snappystream)
 include(external/threadpool)
 
+if(WITH_GPU)
+    include(cuda)
+    include(tensorrt)
+    include(external/anakin)
+else()
+  set(WITH_ANAKIN OFF CACHE STRING "Anakin is valid only when GPU is set." FORCE)
+endif()
+
 include(cudnn)              # set cudnn libraries, must before configure
 include(cupti)
 include(configure)          # add paddle env configuration
@@ -210,14 +238,6 @@ set(EXTERNAL_LIBS
     ${PYTHON_LIBRARIES}
 )
 
-if(WITH_GPU)
-    include(cuda)
-    include(tensorrt)
-    include(external/anakin)
-else()
-  set(WITH_ANAKIN OFF CACHE STRING "Anakin is valid only when GPU is set." FORCE)
-endif()
-
 if(WITH_AMD_GPU)
     find_package(HIP)
     include(hip)
@@ -227,6 +247,10 @@ if(WITH_MKLML)
     list(APPEND EXTERNAL_LIBS ${MKLML_IOMP_LIB})
 endif()
 
+if(WITH_LIBXSMM)
+    list(APPEND EXTERNAL_LIBS ${LIBXSMM_LIBS})
+endif()
+
 if(WITH_MKLDNN)
     list(APPEND EXTERNAL_LIBS ${MKLDNN_LIB})
 endif()
@@ -266,7 +290,3 @@ if(WITH_DOC)
     find_python_module(recommonmark REQUIRED)
     add_subdirectory(doc)
 endif()
-
-if (WITH_CONTRIB)
-    add_subdirectory(paddle/contrib)
-endif()

Dockerfile

@@ -80,7 +80,7 @@ RUN pip install pre-commit 'ipython==5.3.0' && \
     pip install opencv-python
 
 #For docstring checker
-RUN pip install pylint pytest astroid isort
+RUN pip install pylint pytest astroid isort LinkChecker
 
 COPY ./python/requirements.txt /root/
 RUN pip install -r /root/requirements.txt

README.md

@@ -18,7 +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.
 
-### Lastest PaddlePaddle Version: [Fluid](https://github.com/PaddlePaddle/Paddle/tree/develop/paddle/fluid)
+
+### Latest PaddlePaddle Release: [Fluid 0.14.0](https://github.com/PaddlePaddle/Paddle/tree/v0.14.0)
+### Install Latest Stable Release:
+```
+# Linux CPU
+pip install paddlepaddle
+# Linux GPU cuda9cudnn7
+pip install paddlepaddle-gpu
+# Linux GPU cuda8cudnn7
+pip install paddlepaddle-gpu==0.14.0.post87
+# Linux GPU cuda8cudnn5
+pip install paddlepaddle-gpu==0.14.0.post85
+
+# For installation on other platform, refer to http://paddlepaddle.org/
+```
 
 ## Features
 

benchmark/fluid/fluid_benchmark.py

@@ -85,8 +85,7 @@ def dist_transpile(trainer_id, args):
         trainer_id,
         pservers=pserver_endpoints,
         trainers=trainers,
-        sync_mode=not args.async_mode,
-        slice_var_up=not args.no_split_var)
+        sync_mode=not args.async_mode)
     if training_role == "PSERVER":
         pserver_program = t.get_pserver_program(current_endpoint)
         pserver_startup_program = t.get_startup_program(current_endpoint,
@@ -210,7 +209,7 @@ def train_parallel(avg_loss, infer_prog, optimizer, train_reader, test_reader,
     # generate fake:
     if args.use_fake_data:
         for var in feed_var_list:
-            v = startup_prog.global_block().clone_variable(var)
+            v = startup_prog.global_block()._clone_variable(var)
             var.persistable = True
             v.persistable = True
 

benchmark/paddle/image/run.sh

@@ -1,3 +1,5 @@
+#!/bin/bash
+
 set -e
 
 function train() {

benchmark/paddle/image/run_mkl_infer.sh

@@ -1,3 +1,5 @@
+#!/bin/bash
+
 set -e
 
 function clock_to_seconds() {

benchmark/paddle/image/run_mkl_train.sh

@@ -1,3 +1,5 @@
+#!/bin/bash
+
 set -e
 
 function train() {

benchmark/paddle/image/run_openblas_infer.sh

@@ -1,3 +1,5 @@
+#!/bin/bash
+
 set -e
 
 function clock_to_seconds() {