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

Author: chengduoZH

CMakeLists.txt

@@ -39,7 +39,6 @@ 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_TENSORRT    "Compile PaddlePaddle with TensorRT 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)
@@ -180,13 +179,9 @@ set(EXTERNAL_LIBS
 
 if(WITH_GPU)
     include(cuda)
+    include(tensorrt)
 endif(WITH_GPU)
 
-# TensorRT depends on GPU.
-if (NOT WITH_GPU)
-  set(WITH_TENSORRT OFF)
-endif()
-
 if(WITH_AMD_GPU)
     find_package(HIP)
     include(hip)

Dockerfile

@@ -46,7 +46,7 @@ ENV PATH=${PATH}:${GOROOT}/bin:${GOPATH}/bin
 RUN curl -s -q https://glide.sh/get | sh
 
 # Install TensorRT
-# The unnecessary files has been removed to make the library small.
+# The unnecessary files has been removed to make the library small. It only contains include and lib now.
 RUN wget -qO- http://paddlepaddledeps.bj.bcebos.com/TensorRT-4.0.0.3.Ubuntu-16.04.4.x86_64-gnu.cuda-8.0.cudnn7.0.tar.gz | \
     tar -xz -C /usr/local && \
     cp -rf /usr/local/TensorRT/include /usr && \

Dockerfile.android

@@ -27,7 +27,7 @@ RUN git config --global credential.helper store
 # Fix locales to en_US.UTF-8
 RUN localedef -i en_US -f UTF-8 en_US.UTF-8
 
-RUN pip install --upgrade pip && \
+RUN pip install --upgrade pip==9.0.3 && \
     pip install -U 'protobuf==3.1.0' && \
     pip install -U wheel sphinx && \
     pip install pre-commit

paddle/scripts/check_env.sh renamed to benchmark/paddle/image/check_env.sh

@@ -80,6 +80,16 @@ if(WITH_GPU)
     # Include cuda and cudnn
     include_directories(${CUDNN_INCLUDE_DIR})
     include_directories(${CUDA_TOOLKIT_INCLUDE})
+
+    if(TENSORRT_FOUND)
+        if(${CUDA_VERSION_MAJOR} VERSION_LESS 8)
+            message(FATAL_ERROR "TensorRT needs CUDA >= 8.0 to compile")
+        endif()
+        if(${CUDNN_MAJOR_VERSION} VERSION_LESS 7)
+            message(FATAL_ERROR "TensorRT needs CUDNN >= 7.0 to compile")
+        endif()
+        include_directories(${TENSORRT_INCLUDE_DIR})
+    endif()
 elseif(WITH_AMD_GPU)
     add_definitions(-DPADDLE_WITH_HIP)
     set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -D__HIP_PLATFORM_HCC__")

cmake/configure.cmake

@@ -0,0 +1,33 @@
+if(NOT WITH_GPU)
+    return()
+endif()
+
+set(TENSORRT_ROOT "/usr" CACHE PATH "TENSORRT ROOT")
+find_path(TENSORRT_INCLUDE_DIR NvInfer.h
+    PATHS ${TENSORRT_ROOT} ${TENSORRT_ROOT}/include
+    $ENV{TENSORRT_ROOT} $ENV{TENSORRT_ROOT}/include
+    NO_DEFAULT_PATH
+)
+
+find_library(TENSORRT_LIBRARY NAMES libnvinfer.so libnvinfer.a
+    PATHS ${TENSORRT_ROOT} ${TENSORRT_ROOT}/lib
+    $ENV{TENSORRT_ROOT} $ENV{TENSORRT_ROOT}/lib
+    NO_DEFAULT_PATH
+    DOC "Path to TensorRT library.")
+
+if(TENSORRT_INCLUDE_DIR AND TENSORRT_LIBRARY)
+    set(TENSORRT_FOUND ON)
+else()
+    set(TENSORRT_FOUND OFF)
+endif()
+
+if(TENSORRT_FOUND)
+    file(READ ${TENSORRT_INCLUDE_DIR}/NvInfer.h TENSORRT_VERSION_FILE_CONTENTS)
+    string(REGEX MATCH "define NV_TENSORRT_MAJOR +([0-9]+)" TENSORRT_MAJOR_VERSION
+        "${TENSORRT_VERSION_FILE_CONTENTS}")
+    string(REGEX REPLACE "define NV_TENSORRT_MAJOR +([0-9]+)" "\\1"
+        TENSORRT_MAJOR_VERSION "${TENSORRT_MAJOR_VERSION}")
+
+    message(STATUS "Current TensorRT header is ${TENSORRT_INCLUDE_DIR}/NvInfer.h. "
+        "Current TensorRT version is v${TENSORRT_MAJOR_VERSION}. ")
+endif()

cmake/tensorrt.cmake

@@ -3,7 +3,9 @@ add_custom_target(paddle_apis ALL
 
 add_custom_target(paddle_docs ALL
                   DEPENDS paddle_v2_docs paddle_v2_docs_cn
-                  paddle_fluid_docs paddle_fluid_docs_cn)
+                  paddle_fluid_docs paddle_fluid_docs_cn
+                  paddle_mobile_docs paddle_mobile_docs_cn)
 
 add_subdirectory(v2)
 add_subdirectory(fluid)
+add_subdirectory(mobile)

doc/CMakeLists.txt

@@ -33,3 +33,45 @@ Xavier
     :members:
     :noindex:
 
+MSRA
+------
+
+..  autoclass:: paddle.fluid.initializer.MSRA
+    :members:
+    :noindex:
+
+ConstantInitializer
+-------------------
+
+..  autoclass:: paddle.fluid.initializer.ConstantInitializer
+    :members:
+    :noindex:
+
+UniformInitializer
+------------------
+
+..  autoclass:: paddle.fluid.initializer.UniformInitializer
+    :members:
+    :noindex:
+
+NormalInitializer
+-----------------
+
+..  autoclass:: paddle.fluid.initializer.NormalInitializer
+    :members:
+    :noindex:
+
+XavierInitializer
+-----------------
+
+..  autoclass:: paddle.fluid.initializer.XavierInitializer
+    :members:
+    :noindex:
+    MSRA
+    ------
+
+MSRAInitializer
+-----------------
+..  autoclass:: paddle.fluid.initializer.MSRAInitializer
+    :members:
+    :noindex:

doc/fluid/api/initializer.rst

@@ -84,7 +84,7 @@ Running an operator can be asynchronized. There is a thread pool to execute an `
 
 ## Synchronize GPU Kernels
 
-The GPU is a non-blocking device. The different streams need be synchronized when switing streams. In current implementation, the synchronization based on the following algorithm:
+The GPU is a non-blocking device. The different streams need be synchronized when switching streams. In current implementation, the synchronization based on the following algorithm:
 
 1. `OpHandle` will record `DeviceContext` that it is used.
 2. In `OpHandle::Run`, if the `DeviceContext` of current operator is different from `DeviceContext` of any input variable, just wait the generate operator of this input variable.

doc/fluid/design/concepts/parallel_executor.md

@@ -0,0 +1,57 @@
+## Distributed training overview doc
+
+Currently Paddle Fluid use parameter server architecture to support distributed training.
+
+For synchronous and asynchronous training, the differences are mostly in the logic of parameter server. Now we have already support synchronous training.
+
+### Synchronous training
+
+The training process of synchronous training is:
+
+![synchronous distributed training](./src/sync_distributed_training.png)
+
+1. Pserver
+	1. set `barrier_condition_` to 0 and waits for trainers to send gradient.
+1. Trainer
+	1. Trainer read minibatch of data, run forward-backward with local parameter copy and get the gradients for parameters.
+	1. Trainer use split op to split all the gradient into blocks. The split method is determined at compile time.
+	1. Trainer use send_op to send all the split gradients to corresponding parameter server.
+	1. After trainer send all the gradients, it will send a `BATCH_BARRIER_MESSAGE` to all pservers.
+	1. Trainer call GetVariable to pserver and wait for `barrier_condition_` on pserver to be 1.
+1. Pserver
+   1. Pserver will count the number of `BATCH_BARRIER_MESSAGE`.
+	1. When the count of `BATCH_BARRIER_MESSAGE` is equal to the number of Trainer. Pserver thinks it received all gradient from all trainers.
+	1. Pserver will run the optimization block to optimize the parameters.
+	1. After optimization, pserver set `barrier_condition_` to 1.
+	1. Pserver wait for `FETCH_BARRIER_MESSAGE`.
+1. Trainer.
+	1. The trainer uses GetVariable to get all the parameters from pserver.
+	1. Trainer sends a `FETCH_BARRIER_MESSAGE` to each pserver.
+1. Pserver.
+	1. when the number of `FETCH_BARRIER_MESSAGE` reach the number of all trainers. Pserver think all the parameters have been got. it will go back to 1. to set `barrier_condition_` to 0.
+
+### Asynchronous training
+In the above process. There are two barriers for all trainers to synchronize with each other. In asynchronous training, these two barriers are not needed. The trainer can just send gradients to pserver and then get parameters back.
+
+The training process of asynchronous training can be:
+
+![asynchronous distributed training](./src/async_distributed_training.png)
+
+1. Pserver:
+	1. Each parameter has a queue to receive its gradient from trainers.
+	1. Each parameter has a thread to read data from the queue and run optimize block, using the gradient to optimize the parameter.
+	1. Using an independent thread to handle RPC call `GetVariable` for trainers to get parameters back.(Maybe here we should use a thread pool to speed up fetching the parameters.)
+
+1. Trainer:
+	1. Trainer read a batch of data. Run forward and backward with local parameter copy and get the gradients for parameters.
+	1. Trainer split all gradients to blocks and then send these gradient blocks to pservers(pserver will put them into the queue).
+	2. Trainer gets all parameters back from pserver.
+
+### Note:
+There are also some conditions that need to consider. For exmaple:
+
+1. If trainer needs to wait for the pserver to apply it's gradient and then get back the parameters back.
+1. If we need a lock between parameter update and parameter fetch.
+1. If one parameter must be on one server, or it can also be split and send to multiple parameter servers.
+
+The above architecture of asynchronous training can support different mode, we can have a detailed test in the future for these problems.