Skip to content

P
Paddle

PaddlePaddle / Paddle
大约 1 年 前同步成功

通知 2298
Star 20931
Fork 5422
P
Paddle
提交 0dc4b298 编写于 作者: X xzl
浏览文件
操作

deal with conflit

上级 931a23f1 2711bd92
展开全部 隐藏空白更改
内联 并排
Showing with 33151 addition and 8486 deletion
.pre-commit-config.yaml
浏览文件 @ 0dc4b298
......@@ -31,6 +31,3 @@
- id: go-fmt
types:
- go
- id: gometalinter
types:
- go
CMakeLists.txt
浏览文件 @ 0dc4b298
......@@ -86,6 +86,14 @@ if(ANDROID OR IOS)
"Disable MKLDNN when cross-compiling for Android and iOS" FORCE)
set(WITH_MKLML OFF CACHE STRING
"Disable MKLML package when cross-compiling for Android and iOS" FORCE)
# Compile PaddlePaddle mobile inference library
if (NOT WITH_C_API)
set(WITH_C_API ON CACHE STRING
"Always compile the C_API when cross-compiling for Android and iOS" FORCE)
endif()
set(MOBILE_INFERENCE ON)
add_definitions(-DPADDLE_MOBILE_INFERENCE)
endif()
set(THIRD_PARTY_PATH "${CMAKE_BINARY_DIR}/third_party" CACHE STRING
......@@ -97,6 +105,12 @@ if (WITH_C_API AND WITH_PYTHON)
"different Python interpreter from compiling.")
endif()
if(MOBILE_INFERENCE)
set(THIRD_PARTY_BUILD_TYPE MinSizeRel)
else()
set(THIRD_PARTY_BUILD_TYPE Release)
endif()
########################################################################################
include(external/mklml) # download mklml package
......@@ -113,6 +127,7 @@ include(external/warpctc) # download, build, install warpctc
include(external/any) # download libn::any
include(external/eigen) # download eigen3
include(external/pybind11) # download pybind11
include(external/nccl)
include(cudnn) # set cudnn libraries, must before configure
include(configure) # add paddle env configuration
......@@ -145,7 +160,7 @@ set(EXTERNAL_LIBS
if(WITH_GPU)
list(APPEND EXTERNAL_LIBS ${CUDA_LIBRARIES} ${CUDA_rt_LIBRARY})
if(NOT WITH_DSO)
list(APPEND EXTERNAL_LIBS ${CUDNN_LIBRARY} ${CUDA_CUBLAS_LIBRARIES} ${CUDA_curand_LIBRARY})
list(APPEND EXTERNAL_LIBS ${CUDNN_LIBRARY} ${CUDA_CUBLAS_LIBRARIES} ${CUDA_curand_LIBRARY} ${NCCL_LIBRARY})
endif(NOT WITH_DSO)
endif(WITH_GPU)
......@@ -160,9 +175,11 @@ endif(USE_NNPACK)
add_subdirectory(proto)
# "add_subdirectory(go)" should be placed after the following loine,
# because it depends on paddle/optimizer.
add_subdirectory(paddle/optimizer)
if(NOT MOBILE_INFERENCE)
# "add_subdirectory(go)" should be placed after the following loine,
# because it depends on paddle/optimizer.
add_subdirectory(paddle/optimizer)
endif()
# "add_subdirectory(paddle)" and "add_subdirectory(python)" should be
# placed after this block, because they depends on it.
......
Dockerfile
浏览文件 @ 0dc4b298
......@@ -22,7 +22,7 @@ COPY ./paddle/scripts/docker/root/ /root/
RUN apt-get update && \
apt-get install -y \
git python-pip python-dev openssh-server bison \
git python-pip python-dev openssh-server bison libnccl-dev \
wget unzip unrar tar xz-utils bzip2 gzip coreutils ntp \
curl sed grep graphviz libjpeg-dev zlib1g-dev \
python-matplotlib gcc-4.8 g++-4.8 \
......
cmake/configure.cmake
浏览文件 @ 0dc4b298
......@@ -24,6 +24,10 @@ if(WITH_DOUBLE)
add_definitions(-DPADDLE_TYPE_DOUBLE)
endif(WITH_DOUBLE)
if(WITH_TESTING)
add_definitions(-DPADDLE_WITH_TESTING)
endif(WITH_TESTING)
if(NOT WITH_TIMER)
add_definitions(-DPADDLE_DISABLE_TIMER)
endif(NOT WITH_TIMER)
......@@ -49,19 +53,20 @@ if(NOT WITH_GOLANG)
endif(NOT WITH_GOLANG)
if(NOT WITH_GPU)
add_definitions(-DPADDLE_ONLY_CPU)
add_definitions(-DHPPL_STUB_FUNC)
list(APPEND CMAKE_CXX_SOURCE_FILE_EXTENSIONS cu)
else()
add_definitions(-DPADDLE_WITH_CUDA)
FIND_PACKAGE(CUDA REQUIRED)
if(${CUDA_VERSION_MAJOR} VERSION_LESS 7)
message(FATAL_ERROR "Paddle need CUDA >= 7.0 to compile")
message(FATAL_ERROR "Paddle needs CUDA >= 7.0 to compile")
endif()
if(NOT CUDNN_FOUND)
message(FATAL_ERROR "Paddle need cudnn to compile")
message(FATAL_ERROR "Paddle needs cudnn to compile")
endif()
set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS} "-Xcompiler ${SIMD_FLAG}")
......
cmake/external/eigen.cmake
浏览文件 @ 0dc4b298
......@@ -8,7 +8,7 @@ ExternalProject_Add(
extern_eigen3
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/RLovelett/eigen.git"
GIT_TAG "master"
GIT_TAG 70661066beef694cadf6c304d0d07e0758825c10
PREFIX ${EIGEN_SOURCE_DIR}
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
......
cmake/external/gflags.cmake
浏览文件 @ 0dc4b298
......@@ -36,6 +36,7 @@ ExternalProject_Add(
# change this back to the official Github repo once my PR is
# merged.
GIT_REPOSITORY "https://github.com/wangkuiyi/gflags.git"
GIT_TAG 986964c07427ecb9cdb5bd73f73ebbd40e54dadb
PREFIX ${GFLAGS_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
......@@ -45,11 +46,11 @@ ExternalProject_Add(
-DCMAKE_INSTALL_PREFIX=${GFLAGS_INSTALL_DIR}
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DBUILD_TESTING=OFF
-DCMAKE_BUILD_TYPE=Release
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${GFLAGS_INSTALL_DIR}
-DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
-DCMAKE_BUILD_TYPE:STRING=Release
-DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
)
ADD_LIBRARY(gflags STATIC IMPORTED GLOBAL)
......
cmake/external/glog.cmake
浏览文件 @ 0dc4b298
......@@ -31,6 +31,7 @@ ExternalProject_Add(
${EXTERNAL_PROJECT_LOG_ARGS}
DEPENDS gflags
GIT_REPOSITORY "https://github.com/google/glog.git"
GIT_TAG v0.3.5
PREFIX ${GLOG_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
......@@ -43,12 +44,12 @@ ExternalProject_Add(
-DWITH_GFLAGS=ON
-Dgflags_DIR=${GFLAGS_INSTALL_DIR}/lib/cmake/gflags
-DBUILD_TESTING=OFF
-DCMAKE_BUILD_TYPE=Release
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${GLOG_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR:PATH=${GLOG_INSTALL_DIR}/lib
-DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
-DCMAKE_BUILD_TYPE:STRING=Release
-DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
)
ADD_LIBRARY(glog STATIC IMPORTED GLOBAL)
......
cmake/external/gtest.cmake
浏览文件 @ 0dc4b298
......@@ -56,11 +56,11 @@ IF(WITH_TESTING)
-DBUILD_GMOCK=ON
-Dgtest_disable_pthreads=ON
-Dgtest_force_shared_crt=ON
-DCMAKE_BUILD_TYPE=Release
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${GTEST_INSTALL_DIR}
-DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
-DCMAKE_BUILD_TYPE:STRING=Release
-DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
)
ADD_LIBRARY(gtest STATIC IMPORTED GLOBAL)
......
cmake/external/nccl.cmake 0 → 100644
浏览文件 @ 0dc4b298
include(ExternalProject)
set(NCCL_SOURCE_DIR ${THIRD_PARTY_PATH}/nccl)
include_directories(${NCCL_SOURCE_DIR}/src/extern_nccl/src)
if(WITH_DSO)
# If we use DSO, we do not build nccl, just download the dependencies
set(NCCL_BUILD_COMMAND "")
set(NCCL_INSTALL_COMMAND "")
set(NCCL_INSTALL_DIR "")
else()
# otherwise, we build nccl and link it.
set(NCCL_INSTALL_DIR ${THIRD_PARTY_PATH}/install/nccl)
# Note: cuda 8.0 is needed to make nccl
# When cuda is not installed on the system directory, need to set CUDA_HOME to your cuda root
set(NCCL_BUILD_COMMAND "make -j 8")
set(NCCL_INSTALL_COMMAND "make install PREFIX=${NCCL_INSTALL_DIR}")
endif()
ExternalProject_Add(
extern_nccl
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/NVIDIA/nccl.git"
GIT_TAG "v1.3.4-1"
PREFIX "${NCCL_SOURCE_DIR}"
UPDATE_COMMAND ""
CONFIGURE_COMMAND ""
BUILD_COMMAND "${NCCL_BUILD_COMMAND}"
INSTALL_COMMAND "${NCCL_INSTALL_COMMAND}"
INSTALL_DIR "${NCCL_INSTALL_DIR}"
TEST_COMMAND ""
)
if(WITH_DSO)
if(${CMAKE_VERSION} VERSION_LESS "3.3.0")
set(dummyfile ${CMAKE_CURRENT_BINARY_DIR}/lib_nccl_dummy.c)
file(WRITE ${dummyfile} "const char * dummy_nccl = \"${dummyfile}\";")
add_library(nccl STATIC ${dummyfile})
else()
add_library(nccl INTERFACE)
endif()
else()
add_library(nccl STATIC IMPORTED GLOBAL)
set_property(TARGET nccl PROPERTY IMPORTED_LOCATION
${NCCL_INSTALL_DIR}/lib/libnccl_static.a)
endif()
add_dependencies(nccl extern_nccl)
cmake/external/protobuf.cmake
浏览文件 @ 0dc4b298
......@@ -191,12 +191,12 @@ FUNCTION(build_protobuf TARGET_NAME BUILD_FOR_HOST)
${OPTIONAL_ARGS}
-Dprotobuf_BUILD_TESTS=OFF
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DCMAKE_BUILD_TYPE=Release
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
-DCMAKE_INSTALL_PREFIX=${PROTOBUF_INSTALL_DIR}
-DCMAKE_INSTALL_LIBDIR=lib
CMAKE_CACHE_ARGS
-DCMAKE_INSTALL_PREFIX:PATH=${PROTOBUF_INSTALL_DIR}
-DCMAKE_BUILD_TYPE:STRING=Release
-DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
-DCMAKE_VERBOSE_MAKEFILE:BOOL=OFF
-DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
${OPTIONAL_CACHE_ARGS}
......
cmake/external/warpctc.cmake
浏览文件 @ 0dc4b298
......@@ -35,6 +35,7 @@ ExternalProject_Add(
extern_warpctc
${EXTERNAL_PROJECT_LOG_ARGS}
GIT_REPOSITORY "https://github.com/gangliao/warp-ctc.git"
GIT_TAG b63a0644654a3e0ed624c85a1767bc8193aead09
PREFIX ${WARPCTC_SOURCES_DIR}
UPDATE_COMMAND ""
CMAKE_ARGS -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
......@@ -48,9 +49,9 @@ ExternalProject_Add(
-DCMAKE_DISABLE_FIND_PACKAGE_Torch=ON
-DBUILD_SHARED=ON
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DCMAKE_BUILD_TYPE=Release
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS -DCMAKE_BUILD_TYPE:STRING=Release
CMAKE_CACHE_ARGS -DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
-DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
-DCMAKE_INSTALL_PREFIX:PATH=${WARPCTC_INSTALL_DIR}
)
......
cmake/external/zlib.cmake
浏览文件 @ 0dc4b298
......@@ -42,11 +42,11 @@ ExternalProject_Add(
-DBUILD_SHARED_LIBS=OFF
-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-DCMAKE_MACOSX_RPATH=ON
-DCMAKE_BUILD_TYPE=Release
-DCMAKE_BUILD_TYPE=${THIRD_PARTY_BUILD_TYPE}
${EXTERNAL_OPTIONAL_ARGS}
CMAKE_CACHE_ARGS -DCMAKE_INSTALL_PREFIX:PATH=${ZLIB_INSTALL_DIR}
-DCMAKE_POSITION_INDEPENDENT_CODE:BOOL=ON
-DCMAKE_BUILD_TYPE:STRING=Release
-DCMAKE_BUILD_TYPE:STRING=${THIRD_PARTY_BUILD_TYPE}
)
LIST(APPEND external_project_dependencies zlib)
......
cmake/generic.cmake
浏览文件 @ 0dc4b298
......@@ -389,13 +389,60 @@ function(go_test TARGET_NAME)
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
endfunction(go_test)
# Modification of standard 'protobuf_generate_cpp()' with protobuf-lite support
# Usage:
# paddle_protobuf_generate_cpp(<proto_srcs> <proto_hdrs> <proto_files>)
function(paddle_protobuf_generate_cpp SRCS HDRS)
if(NOT ARGN)
message(SEND_ERROR "Error: paddle_protobuf_generate_cpp() called without any proto files")
return()
endif()
set(${SRCS})
set(${HDRS})
if (MOBILE_INFERENCE)
set(EXTRA_FLAG "lite:")
else()
set(EXTRA_FLAG "")
endif()
foreach(FIL ${ARGN})
get_filename_component(ABS_FIL ${FIL} ABSOLUTE)
get_filename_component(FIL_WE ${FIL} NAME_WE)
set(_protobuf_protoc_src "${CMAKE_CURRENT_BINARY_DIR}/${FIL_WE}.pb.cc")
set(_protobuf_protoc_hdr "${CMAKE_CURRENT_BINARY_DIR}/${FIL_WE}.pb.h")
list(APPEND ${SRCS} "${_protobuf_protoc_src}")
list(APPEND ${HDRS} "${_protobuf_protoc_hdr}")
add_custom_command(
OUTPUT "${_protobuf_protoc_src}"
"${_protobuf_protoc_hdr}"
COMMAND ${CMAKE_COMMAND} -E make_directory "${CMAKE_CURRENT_BINARY_DIR}"
COMMAND ${PROTOBUF_PROTOC_EXECUTABLE}
-I${CMAKE_CURRENT_SOURCE_DIR}
--cpp_out "${EXTRA_FLAG}${CMAKE_CURRENT_BINARY_DIR}" ${ABS_FIL}
DEPENDS ${ABS_FIL} protoc
COMMENT "Running C++ protocol buffer compiler on ${FIL}"
VERBATIM )
endforeach()
set_source_files_properties(${${SRCS}} ${${HDRS}} PROPERTIES GENERATED TRUE)
set(${SRCS} ${${SRCS}} PARENT_SCOPE)
set(${HDRS} ${${HDRS}} PARENT_SCOPE)
endfunction()
function(proto_library TARGET_NAME)
set(oneValueArgs "")
set(multiValueArgs SRCS DEPS)
cmake_parse_arguments(proto_library "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
set(proto_srcs)
set(proto_hdrs)
protobuf_generate_cpp(proto_srcs proto_hdrs ${proto_library_SRCS})
paddle_protobuf_generate_cpp(proto_srcs proto_hdrs ${proto_library_SRCS})
cc_library(${TARGET_NAME} SRCS ${proto_srcs} DEPS ${proto_library_DEPS} protobuf)
endfunction()
......
cmake/util.cmake
浏览文件 @ 0dc4b298
......@@ -73,25 +73,43 @@ function(link_paddle_exe TARGET_NAME)
generate_rdma_links()
endif()
target_circle_link_libraries(${TARGET_NAME}
ARCHIVE_START
paddle_gserver
paddle_function
ARCHIVE_END
paddle_pserver
paddle_trainer_lib
paddle_network
paddle_math
paddle_utils
paddle_parameter
paddle_proto
paddle_cuda
paddle_optimizer
${EXTERNAL_LIBS}
${CMAKE_THREAD_LIBS_INIT}
${CMAKE_DL_LIBS}
${RDMA_LD_FLAGS}
${RDMA_LIBS})
if(MOBILE_INFERENCE)
target_circle_link_libraries(${TARGET_NAME}
ARCHIVE_START
paddle_gserver
paddle_function
ARCHIVE_END
paddle_math
paddle_utils
paddle_parameter
paddle_proto
paddle_cuda
${EXTERNAL_LIBS}
${CMAKE_THREAD_LIBS_INIT}
${CMAKE_DL_LIBS}
${RDMA_LD_FLAGS}
${RDMA_LIBS})
else()
target_circle_link_libraries(${TARGET_NAME}
ARCHIVE_START
paddle_gserver
paddle_function
ARCHIVE_END
paddle_pserver
paddle_trainer_lib
paddle_network
paddle_math
paddle_utils
paddle_parameter
paddle_proto
paddle_cuda
paddle_optimizer
${EXTERNAL_LIBS}
${CMAKE_THREAD_LIBS_INIT}
${CMAKE_DL_LIBS}
${RDMA_LD_FLAGS}
${RDMA_LIBS})
endif()
if(ANDROID)
target_link_libraries(${TARGET_NAME} log)
......
doc/api/v1/index_cn.rst
浏览文件 @ 0dc4b298
......@@ -21,7 +21,7 @@ Model Config API
trainer_config_helpers/optimizers.rst
trainer_config_helpers/data_sources.rst
trainer_config_helpers/layers.rst
trainer_config_helpers/activations.rst
trainer_config_helpers/activations.rst
trainer_config_helpers/poolings.rst
trainer_config_helpers/networks.rst
trainer_config_helpers/evaluators.rst
......
doc/api/v2/config/layer.rst
浏览文件 @ 0dc4b298
......@@ -345,6 +345,11 @@ clip
.. autoclass:: paddle.v2.layer.clip
:noindex:
resize
------
.. autoclass:: paddle.v2.layer.resize
:noindex:
slope_intercept
---------------
.. autoclass:: paddle.v2.layer.slope_intercept
......
doc/api/v2/config/networks.rst
浏览文件 @ 0dc4b298
......@@ -125,3 +125,8 @@ simple_attention
:members: simple_attention
:noindex:
dot_product_attention
---------------------
.. automodule:: paddle.v2.networks
:members: dot_product_attention
:noindex:
doc/design/block.md
浏览文件 @ 0dc4b298
......@@ -5,12 +5,12 @@
Both deep learning systems and programming languages help users describe computation procedures. These systems use various representations of computation:
- Caffe, Torch, and Paddle: sequences of layers.
- TensorFlow, Caffe2, Mxnet: graphs of operators.
- TensorFlow, Caffe2, Mxnet: graph of operators.
- PaddlePaddle: nested blocks, like C++ and Java programs.
## Block in Programming Languages and Deep Learning
In programming languages, a block is a pair of curly braces that includes local variables definitions and a sequence of instructions, or operators.
In programming languages, a block is a pair of curly braces that includes local variables definitions and a sequence of instructions or operators.
Blocks work with control flow structures like `if`, `else`, and `for`, which have equivalents in deep learning:
......@@ -24,14 +24,14 @@ A key difference is that a C++ program describes a one pass computation, whereas
## Stack Frames and the Scope Hierarchy
The existence of the backward makes the execution of a block of traditional programs and PaddlePaddle different to each other:
The existence of the backward pass makes the execution of a block of PaddlePaddle different from traditional programs:
| programming languages | PaddlePaddle |
|-----------------------|-------------------------------|
| stack | scope hierarchy |
| stack frame | scope |
| push at entering block| push at entering block |
| pop at leaving block | destroy at minibatch completes|
| programming languages | PaddlePaddle |
|-----------------------|---------------------------------|
| stack | scope hierarchy |
| stack frame | scope |
| push at entering block| push at entering block |
| pop at leaving block | destroy when minibatch completes|
1. In traditional programs:
......@@ -42,9 +42,9 @@ The existence of the backward makes the execution of a block of traditional prog
1. In PaddlePaddle
- When the execution enters a block, PaddlePaddle adds a new scope, where it realizes variables.
- PaddlePaddle doesn't pop a scope after the execution of the block because variables therein are to be used by the backward pass. So it has a stack forest known as a *scope hierarchy*.
- PaddlePaddle doesn't pop a scope after the execution of the block because variables therein are used by the backward pass. So it has a stack forest known as a *scope hierarchy*.
- The height of the highest tree is the maximum depth of nested blocks.
- After the process of a minibatch, PaddlePaddle destroys the scope hierarchy.
- After the processing of a minibatch, PaddlePaddle destroys the scope hierarchy.
## Use Blocks in C++ and PaddlePaddle Programs
......@@ -55,17 +55,23 @@ Let us consolidate the discussion by presenting some examples.
The following C++ programs shows how blocks are used with the `if-else` structure:
```c++
namespace pd = paddle;
int x = 10;
int y = 20;
int out;
int y = 1;
int z = 10;
bool cond = false;
int o1, o2;
if (cond) {
int z = x + y;
out = softmax(z);
o1 = z;
o2 = pd::layer::softmax(z);
} else {
int z = fc(x);
out = z;
int d = pd::layer::fc(z);
o1 = d;
o2 = d+1;
}
```
An equivalent PaddlePaddle program from the design doc of the [IfElseOp operator](./if_else_op.md) is as follows:
......@@ -73,57 +79,55 @@ An equivalent PaddlePaddle program from the design doc of the [IfElseOp operator
```python
import paddle as pd
x = var(10)
y = var(20)
cond = var(false)
ie = pd.create_ifelseop(inputs=[x], output_num=1)
x = minibatch([10, 20, 30]) # shape=[None, 1]
y = var(1) # shape=[1], value=1
z = minibatch([10, 20, 30]) # shape=[None, 1]
cond = larger_than(x, 15) # [false, true, true]
ie = pd.ifelse()
with ie.true_block():
x = ie.inputs(true, 0)
z = operator.add(x, y)
ie.set_output(true, 0, operator.softmax(z))
d = pd.layer.add_scalar(x, y)
ie.output(d, pd.layer.softmax(d))
with ie.false_block():
x = ie.inputs(false, 0)
z = layer.fc(x)
ie.set_output(true, 0, operator.softmax(z))
out = b(cond)
d = pd.layer.fc(z)
ie.output(d, d+1)
o1, o2 = ie(cond)
```
In both examples, the left branch computes `softmax(x+y)` and the right branch computes `fc(x)`.
In both examples, the left branch computes `x+y` and `softmax(x+y)`, the right branch computes `fc(x)` and `x+1` .
The difference is that variables in the C++ program contain scalar values, whereas those in the PaddlePaddle programs are mini-batches of instances.
A difference is that variables in the C++ program contain scalar values, whereas those in the PaddlePaddle programs are mini-batches of instances. The `ie.input(true, 0)` invocation returns instances in the 0-th input, `x`, that corresponds to true values in `cond` as the local variable `x`, where `ie.input(false, 0)` returns instances corresponding to false values.
### Blocks with `for` and `RNNOp`
The following RNN model from the [RNN design doc](./rnn.md)
The following RNN model in PaddlePaddle from the [RNN design doc](./rnn.md) :
```python
x = sequence([10, 20, 30])
m = var(0)
W = tensor()
U = tensor()
rnn = create_rnn(inputs=[input])
with rnn.stepnet() as net:
x = net.set_inputs(0)
h = net.add_memory(init=m)
fc_out = pd.matmul(W, x)
hidden_out = pd.matmul(U, h.pre(n=1))
sum = pd.add_two(fc_out, hidden_out)
act = pd.sigmoid(sum)
h.update(act) # update memory with act
net.set_outputs(0, act, hidden_out) # two outputs
x = sequence([10, 20, 30]) # shape=[None, 1]
m = var(0) # shape=[1]
W = var(0.314, param=true) # shape=[1]
U = var(0.375, param=true) # shape=[1]
rnn = pd.rnn()
with rnn.step():
h = rnn.memory(init = m)
h_prev = rnn.previous_memory(h)
a = layer.fc(W, x)
b = layer.fc(U, h_prev)
s = pd.add(a, b)
act = pd.sigmoid(s)
rnn.update_memory(h, act)
rnn.output(a, b)
o1, o2 = rnn()
print o1, o2
```
has its equivalent C++ program as follows
```c++
int* x = {10, 20, 30};
int m = 0;
int W = some_value();
int U = some_other_value();
int* m = {0};
int* W = {0.314};
int* U = {0.375};
int mem[sizeof(x) / sizeof(x[0]) + 1];
int o1[sizeof(x) / sizeof(x[0]) + 1];
......@@ -131,25 +135,21 @@ int o2[sizeof(x) / sizeof(x[0]) + 1];
for (int i = 1; i <= sizeof(x)/sizeof(x[0]); ++i) {
int x = x[i-1];
if (i == 1) mem[0] = m;
int fc_out = W * x;
int hidden_out = Y * mem[i-1];
int sum = fc_out + hidden_out;
int a = W * x;
int b = Y * mem[i-1];
int s = fc_out + hidden_out;
int act = sigmoid(sum);
mem[i] = act;
o1[i] = act;
o2[i] = hidden_out;
}
print_array(o1);
print_array(o2);
```
## Compilation and Execution
Like TensorFlow programs, a PaddlePaddle program is written in Python. The first part describes a neural network as a protobuf message, and the rest part executes the message for training or inference.
Like TensorFlow, a PaddlePaddle program is written in Python. The first part describes a neural network as a protobuf message, and the rest executes the message for training or inference.
The generation of this protobuf message is like what a compiler generates a binary executable file. The execution of the message that the OS executes the binary file.
The generation of this protobuf message is similar to how a compiler generates a binary executable file. The execution of the message is similar to how the OS executes the binary file.
## The "Binary Executable File Format"
......@@ -186,10 +186,10 @@ Also, the RNN operator in above example is serialized into a protobuf message of
```
OpDesc {
inputs = {0} // the index of x
outputs = {5, 3} // indices of act and hidden_out
inputs = {0} // the index of x in vars of BlockDesc above
outputs = {5, 3} // indices of act and hidden_out in vars of BlockDesc above
attrs {
"memories" : {1} // the index of h
"states" : {1} // the index of h
"step_net" : <above step net>
}
};
......@@ -203,32 +203,32 @@ This `OpDesc` value is in the `ops` field of the `BlockDesc` value representing
During the generation of the Protobuf message, the Block should store VarDesc (the Protobuf message which describes Variable) and OpDesc (the Protobuf message which describes Operator).
VarDesc in a block should have its name scope to avoid local variables affect parent block's name scope.
Child block's name scopes should inherit the parent's so that OpDesc in child block can reference a VarDesc that stored in parent block. For example
Child block's name scopes should inherit the parent's so that OpDesc in child block can reference a VarDesc that stored in parent block. For example:
```python
a = pd.Varaible(shape=[20, 20])
a = pd.Variable(shape=[20, 20])
b = pd.fc(a, params=["fc.w", "fc.b"])
rnn = pd.create_rnn()
with rnn.stepnet() as net:
x = net.set_inputs(a)
with rnn.stepnet():
x = a.as_step_input()
# reuse fc's parameter
fc_without_b = pd.get_variable("fc.w")
net.set_outputs(fc_without_b)
rnn.output(fc_without_b)
out = rnn()
```
the method `pd.get_variable` can help retrieve a Variable by a name, a Variable may store in a parent block, but might be retrieved in a child block, so block should have a variable scope that supports inheritance.
The method `pd.get_variable` can help retrieve a Variable by the name. The Variable may be stored in a parent block, but might be retrieved in a child block, so block should have a variable scope that supports inheritance.
In compiler design, the symbol table is a data structure created and maintained by compilers to store information about the occurrence of various entities such as variable names, function names, classes, etc.
To store the definition of variables and operators, we define a C++ class `SymbolTable`, like the one used in compilers.
`SymbolTable` can do the following stuff:
`SymbolTable` can do the following:
- store the definitions (some names and attributes) of variables and operators,
- to verify if a variable was declared,
- to make it possible to implement type checking (offer Protobuf message pointers to `InferShape` handlers).
- verify if a variable was declared,
- make it possible to implement type checking (offer Protobuf message pointers to `InferShape` handlers).
```c++
......@@ -240,19 +240,18 @@ class SymbolTable {
OpDesc* NewOp(const string& name="");
// TODO determine whether name is generated by python or C++
// currently assume that a unique name will be generated by C++ if the
// argument name left default.
VarDesc* NewVar(const string& name="");
// TODO determine whether name is generated by python or C++.
// Currently assume that a unique name will be generated by C++ if the
// argument name is left default.
VarDesc* Var(const string& name="");
// find a VarDesc by name, if recursive true, find parent's SymbolTable
// find a VarDesc by name, if recursive is true, find parent's SymbolTable
// recursively.
// this interface is introduced to support InferShape, find protobuf messages
// of variables and operators, pass pointers into InferShape.
// operator
//
// NOTE maybe some C++ classes such as VarDescBuilder and OpDescBuilder should
// be proposed and embedded into pybind to enable python operate on C++ pointers.
// be proposed and embedded into pybind to enable python operation on C++ pointers.
VarDesc* FindVar(const string& name, bool recursive=true);
OpDesc* FindOp(const string& name);
......@@ -270,7 +269,7 @@ class SymbolTable {
After all the description of variables and operators is added into SymbolTable,
the block has enough information to run.
The `Block` class takes a `BlockDesc` as input, and provide `Run` and `InferShape` functions.
The `Block` class takes a `BlockDesc` as input, and provides `Run` and `InferShape` functions.
```c++
......@@ -302,7 +301,7 @@ public:
void CreateVariables(const framework::Scope& scope);
void CreateOperators();
// some other necessary interfaces of NetOp are list below
// some other necessary interfaces of NetOp are listed below
// ...
private:
......@@ -316,15 +315,14 @@ private:
Block inherits from OperatorBase, which has a Run method.
Block's Run method will run its operators sequentially.
There is another important interface called `Eval`, which take some arguments called targets, and generate a minimal graph which takes targets as the end points and creates a new Block,
after `Run`, `Eval` will get the latest value and return the targets.
There is another important interface called `Eval`, which takes some arguments called targets and generates a minimal graph which treats targets as the end points and creates a new Block. After `Run`, `Eval` will get the latest value and return the targets.
The definition of Eval is as follows:
```c++
// clean a block description by targets using the corresponding dependency graph.
// return a new BlockDesc with minimal number of operators.
// NOTE not return a Block but the block's description so that this can be distributed
// NOTE: The return type is not a Block but the block's description so that this can be distributed
// to a cluster.
BlockDesc Prune(const BlockDesc& desc, vector<string> targets);
......
doc/design/executor.md 0 → 100644
浏览文件 @ 0dc4b298
# Executor Design Doc
## Motivation
We use executor to do the runtime evaluation of a `ProgramDesc`.
## Overview
An executor takes a `ProgramDesc`, a `block_id` and a `Scope`. The `ProgramDesc` is a list of blocks and each block contains the protobuf definition of all the parameters and operators. The `block_id` specifies the entrance block. And the `Scope` is the container of all the variable instance, which is persistent throughout different runs.
### What does executor do?
It evaluates all the operators in the `block_id`th block of a `ProgramDesc`.
### What does executor NOT do?
It does not do runtime optimization, meaning intelligently parse the dependency of each op a choose which one to be run and in which order they should be run.
It does not do graph partitioning, meaning dividing the `ProgramDesc` into several small pieces and executing them on different devices.
## Implementation
`Executor` evaluates a `ProgramDesc`. Essentially, it instantiates Variables and Operators, then run all the operators in sequence. [[code]](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/executor.cc)
doc/design/gan_api.md 0 → 100644
浏览文件 @ 0dc4b298
# Design for GAN
GAN (General Adversarial Net [https://arxiv.org/abs/1406.2661]) is an important model for unsupervised learning and widely used in many areas.
It applies several important concepts in machine learning system design, including building and running subgraphs, dependency tracing, different optimizers in one executor and so forth.
In our GAN design, we wrap it as a user-friendly easily customized python API to design different models. We take the conditional DC-GAN (Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks [https://arxiv.org/abs/1511.06434]) as an example due to its good performance on image generation.
<p align="center">
<img src="./test.dot.png" width = "35%" align="center"/><br/>
Figure 1. The overall running logic of GAN. The black solid arrows indicate the forward pass; the green dashed arrows indicate the backward pass of generator training; the red dashed arrows indicate the backward pass of the discriminator training. The BP pass of the green (red) arrow should only update the parameters in the green (red) boxes. The diamonds indicate the data providers. d\_loss and g\_loss marked in red and green are the two targets we would like to run.
</p>
The operators, layers and functions required/optional to build a GAN demo is summarized in https://github.com/PaddlePaddle/Paddle/issues/4563.
<p align="center">
<img src="./dcgan.png" width = "90%" align="center"/><br/>
Figure 2. Photo borrowed from the original DC-GAN paper.
</p>
## The Conditional-GAN might be a class.
This design we adopt the popular open source design in https://github.com/carpedm20/DCGAN-tensorflow and https://github.com/rajathkmp/DCGAN. It contains following data structure:
- DCGAN(object): which contains everything required to build a GAN model. It provides following member functions methods as API:
- __init__(...): Initialize hyper-parameters (like conv dimension and so forth), and declare model parameters of discriminator and generator as well.
- generator(z, y=None): Generate a fake image from input noise z. If the label y is provided, the conditional GAN model will be chosen.
Returns a generated image.
- discriminator(image):
Given an image, decide if it is from a real source or a fake one.
Returns a 0/1 binary label.
- build_model(self):
build the whole GAN model, define training loss for both generator and discrimator.
## Discussion on Engine Functions required to build GAN
- Trace the tensor and variable dependency in the engine executor. (Very critical, otherwise GAN can'be be trained correctly)
- Different optimizers responsible for optimizing different loss.
To be more detailed, we introduce our design of DCGAN as following:
### Class member Function: Initializer
- Set up hyper-parameters, including condtional dimension, noise dimension, batch size and so forth.
- Declare and define all the model variables. All the discriminator parameters are included in the list self.theta_D and all the generator parameters are included in the list self.theta_G.
```python
class DCGAN(object):
def __init__(self, y_dim=None):
# hyper parameters
self.y_dim = y_dim # conditional gan or not
self.batch_size = 100
self.z_dim = z_dim # input noise dimension
# define parameters of discriminators
self.D_W0 = pd.Variable(shape=[3,3, 1, 128], data=pd.gaussian_normal_randomizer())
self.D_b0 = pd.Variable(np.zeros(128)) # variable also support initialization using a numpy data
self.D_W1 = pd.Variable(shape=[784, 128], data=pd.gaussian_normal_randomizer())
self.D_b1 = pd.Variable(np.zeros(128)) # variable also support initialization using a numpy data
self.D_W2 = pd.Varialble(np.random.rand(128, 1))
self.D_b2 = pd.Variable(np.zeros(128))
self.theta_D = [self.D_W0, self.D_b0, self.D_W1, self.D_b1, self.D_W2, self.D_b2]
# define parameters of generators
self.G_W0 = pd.Variable(shape=[784, 128], data=pd.gaussian_normal_randomizer())
self.G_b0 = pd.Variable(np.zeros(128)) # variable also support initialization using a numpy data
self.G_W1 = pd.Variable(shape=[784, 128], data=pd.gaussian_normal_randomizer())
self.G_b1 = pd.Variable(np.zeros(128)) # variable also support initialization using a numpy data
self.G_W2 = pd.Varialble(np.random.rand(128, 1))
self.G_b2 = pd.Variable(np.zeros(128))
self.theta_G = [self.G_W0, self.G_b0, self.G_W1, self.G_b1, self.G_W2, self.G_b2]
```
### Class member Function: Generator
- Given a noisy input z, returns a fake image.
- Concatenation, batch-norm, FC operations required;
- Deconv layer required, which is missing now...
```python
class DCGAN(object):
def generator(self, z, y = None):
# input z: the random noise
# input y: input data label (optional)
# output G_im: generated fake images
if not self.y_dim:
z = pd.layer.concat(1, [z, y])
G_h0 = pd.layer.fc(z, self.G_w0, self.G_b0)
G_h0_bn = pd.layer.batch_norm(G_h0)
G_h0_relu = pd.layer.relu(G_h0_bn)
G_h1 = pd.layer.deconv(G_h0_relu, self.G_w1, self.G_b1)
G_h1_bn = pd.layer.batch_norm(G_h1)
G_h1_relu = pd.layer.relu(G_h1_bn)
G_h2 = pd.layer.deconv(G_h1_relu, self.G_W2, self.G_b2))
G_im = pd.layer.tanh(G_im)
return G_im
```
### Class member function: Discriminator
- Given a noisy input z, returns a fake image.
- Concatenation, Convolution, batch-norm, FC, Leaky-ReLU operations required;
```python
class DCGAN(object):
def discriminator(self, image):
# input image: either generated images or real ones
# output D_h2: binary logit of the label
D_h0 = pd.layer.conv2d(image, w=self.D_w0, b=self.D_b0)
D_h0_bn = pd.layer.batchnorm(h0)
D_h0_relu = pd.layer.lrelu(h0_bn)
D_h1 = pd.layer.conv2d(D_h0_relu, w=self.D_w1, b=self.D_b1)
D_h1_bn = pd.layer.batchnorm(D_h1)
D_h1_relu = pd.layer.lrelu(D_h1_bn)
D_h2 = pd.layer.fc(D_h1_relu, w=self.D_w2, b=self.D_b2)
return D_h2
```
### Class member function: Build the model
- Define data readers as placeholders to hold the data;
- Build generator and discriminators;
- Define two training losses for discriminator and generator, respectively.
If we have execution dependency engine to back-trace all tensors, the module building our GAN model will be like this:
```python
class DCGAN(object):
def build_model(self):
if self.y_dim:
self.y = pd.data(pd.float32, [self.batch_size, self.y_dim])
self.images = pd.data(pd.float32, [self.batch_size, self.im_size, self.im_size])
self.faked_images = pd.data(pd.float32, [self.batch_size, self.im_size, self.im_size])
self.z = pd.data(tf.float32, [None, self.z_size])
# step 1: generate images by generator, classify real/fake images with discriminator
if self.y_dim: # if conditional GAN, includes label
self.G = self.generator(self.z, self.y)
self.D_t = self.discriminator(self.images)
# generated fake images
self.sampled = self.sampler(self.z, self.y)
self.D_f = self.discriminator(self.G)
else: # original version of GAN
self.G = self.generator(self.z)
self.D_t = self.discriminator(self.images)
# generate fake images
self.sampled = self.sampler(self.z)
self.D_f = self.discriminator(self.images)
# step 2: define the two losses
self.d_loss_real = pd.reduce_mean(pd.cross_entropy(self.D_t, np.ones(self.batch_size))
self.d_loss_fake = pd.reduce_mean(pd.cross_entropy(self.D_f, np.zeros(self.batch_size))
self.d_loss = self.d_loss_real + self.d_loss_fake
self.g_loss = pd.reduce_mean(pd.cross_entropy(self.D_f, np.ones(self.batch_szie))
```
If we do not have dependency engine but blocks, the module building our GAN model will be like this:
```python
class DCGAN(object):
def build_model(self, default_block):
# input data in the default block
if self.y_dim:
self.y = pd.data(pd.float32, [self.batch_size, self.y_dim])
self.images = pd.data(pd.float32, [self.batch_size, self.im_size, self.im_size])
# self.faked_images = pd.data(pd.float32, [self.batch_size, self.im_size, self.im_size])
self.z = pd.data(tf.float32, [None, self.z_size])
# step 1: generate images by generator, classify real/fake images with discriminator
with pd.default_block().g_block():
if self.y_dim: # if conditional GAN, includes label
self.G = self.generator(self.z, self.y)
self.D_g = self.discriminator(self.G, self.y)
else: # original version of GAN
self.G = self.generator(self.z)
self.D_g = self.discriminator(self.G, self.y)
self.g_loss = pd.reduce_mean(pd.cross_entropy(self.D_g, np.ones(self.batch_szie))
with pd.default_block().d_block():
if self.y_dim: # if conditional GAN, includes label
self.D_t = self.discriminator(self.images, self.y)
self.D_f = self.discriminator(self.G, self.y)
else: # original version of GAN
self.D_t = self.discriminator(self.images)
self.D_f = self.discriminator(self.G)
# step 2: define the two losses
self.d_loss_real = pd.reduce_mean(pd.cross_entropy(self.D_t, np.ones(self.batch_size))
self.d_loss_fake = pd.reduce_mean(pd.cross_entropy(self.D_f, np.zeros(self.batch_size))
self.d_loss = self.d_loss_real + self.d_loss_fake
```
Some small confusion and problems with this design:
- D\_g and D\_f are actually the same thing, but has to be written twice; i.e., if we want to run two sub-graphs conceptually, the same codes have to be written twice if they are shared by the graph.
- Requires ability to create a block anytime, rather than in if-else or rnn only;
## Main function for the demo:
Generally, the user of GAN just need to the following things:
- Define an object as DCGAN class;
- Build the DCGAN model;
- Specify two optimizers for two different losses with respect to different parameters.
```python
# pd for short, should be more concise.
from paddle.v2 as pd
import numpy as np
import logging
if __name__ == "__main__":
# dcgan class in the default graph/block
# if we use dependency engine as tensorflow
# the codes, will be slightly different like:
# dcgan = DCGAN()
# dcgan.build_model()
with pd.block() as def_block:
dcgan = DCGAN()
dcgan.build_model(def_block)
# load mnist data
data_X, data_y = self.load_mnist()
# Two subgraphs required!!!
with pd.block().d_block():
d_optim = pd.train.Adam(lr = .001, beta= .1)
d_step = d_optim.minimize(dcgan.d_loss, dcgan.theta_D)
with pd.block.g_block():
g_optim = pd.train.Adam(lr = .001, beta= .1)
g_step = pd.minimize(dcgan.g_loss, dcgan.theta_G)
# executor
sess = pd.executor()
# training
for epoch in xrange(10000):
for batch_id in range(N / batch_size):
idx = ...
# sample a batch
batch_im, batch_label = data_X[idx:idx+batch_size], data_y[idx:idx+batch_size]
# sample z
batch_z = np.random.uniform(-1., 1., [batch_size, z_dim])
if batch_id % 2 == 0:
sess.run(d_step,
feed_dict = {dcgan.images: batch_im,
dcgan.y: batch_label,
dcgan.z: batch_z})
else:
sess.run(g_step,
feed_dict = {dcgan.z: batch_z})
```
# More thinking about dependency engine v.s. block design:
- What if we just want to run an intermediate result? Do we need to run the whole block/graph?
- Should we call eval() to get the fake images in the first stage? And then train the discriminator in the second stage?
doc/design/graph_survey.md 0 → 100644
浏览文件 @ 0dc4b298
## Survey on Graph
Neural network framework often provides symbolic API for users to write network topology conveniently. This doc manily focus on symbolic API in most popular neural network frameworks, and try to find out how to parse symbolic configuration to a portable file, such as protobuf or json.
### Mxnet
The core concept of symbolic API is `Symbol`. Mxnet implements `Symbol` class in C++, and export to Python using C-API. Please refer to the comments in Mxnet:
`Symbol` is help class used to represent the operator node in Graph.
`Symbol` acts as an interface for building graphs from different components like Variable, Functor and Group. `Symbol` is also exported to python front-end (while Graph is not) to enable quick test and deployment. Conceptually, symbol is the final operation of a graph and thus including all the information required (the graph) to evaluate its output value.
A simple network topology wrote by Symbol is as follows:
```python
def get_symbol(num_classes=10, **kwargs):
data = mx.symbol.Variable('data')
data = mx.symbol.Flatten(data=data)
fc1 = mx.symbol.FullyConnected(data = data, name='fc1', num_hidden=128)
act1 = mx.symbol.Activation(data = fc1, name='relu1', act_type="relu")
fc2 = mx.symbol.FullyConnected(data = act1, name = 'fc2', num_hidden = 64)
act2 = mx.symbol.Activation(data = fc2, name='relu2', act_type="relu")
fc3 = mx.symbol.FullyConnected(data = act2, name='fc3', num_hidden=num_classes)
mlp = mx.symbol.SoftmaxOutput(data = fc3, name = 'softmax')
return mlp
```
Varible here is actually a Symbol. Every basic Symbol will correspond to one Node, and every Node has its own NodeAttr. There is a op field in NodeAttr class, when a Symbol represents Variable(often input data), the op field is null.
Symbol contains a data member, std::vector<NodeEntry> outputs, and NodeEntry cantains a poniter to Node. We can follow the Node pointer to get all the Graph.
And Symbol can be saved to a Json file.
Here is a detailed example:
```
>>> import mxnet as mx
>>> data = mx.symbol.Variable('data')
>>> print data.debug_str()
Variable:data
>>> data = mx.symbol.Flatten(data=data)
>>> print data.debug_str()
Symbol Outputs:
output[0]=flatten0(0)
Variable:data
--------------------
Op:Flatten, Name=flatten0
Inputs:
arg[0]=data(0) version=0
>>> fc1 = mx.symbol.FullyConnected(data = data, name='fc1', num_hidden=128)
>>> print fc1.debug_str()
Symbol Outputs:
output[0]=fc1(0)
Variable:data
--------------------
Op:Flatten, Name=flatten0
Inputs:
arg[0]=data(0) version=0
Variable:fc1_weight
Variable:fc1_bias
--------------------
Op:FullyConnected, Name=fc1
Inputs:
arg[0]=flatten0(0)
arg[1]=fc1_weight(0) version=0
arg[2]=fc1_bias(0) version=0
Attrs:
num_hidden=128
```
### TensorFlow
The core concept of symbolic API is `Tensor`. Tensorflow defines `Tensor` in Python. Please refer to the comments in TensorFlow:
A `Tensor` is a symbolic handle to one of the outputs of an `Operation`. It does not hold the values of that operation's output, but instead provides a means of computing those values in a TensorFlow [Session](https://www.tensorflow.org/api_docs/python/tf/Session).
A simple example is as follows:
```python
# Build a dataflow graph.
c = tf.constant([[1.0, 2.0], [3.0, 4.0]])
d = tf.constant([[1.0, 1.0], [0.0, 1.0]])
e = tf.matmul(c, d)
# Construct a `Session` to execute the graph.
sess = tf.Session()
# Execute the graph and store the value that `e` represents in `result`.
result = sess.run(e)
```
The main method of `Tensor` is as follows:
```python
@property
def op(self):
"""The `Operation` that produces this tensor as an output."""
return self._op
@property
def dtype(self):
"""The `DType` of elements in this tensor."""
return self._dtype
@property
def graph(self):
"""The `Graph` that contains this tensor."""
return self._op.graph
@property
def name(self):
"""The string name of this tensor."""
if not self._op.name:
raise ValueError("Operation was not named: %s" % self._op)
return "%s:%d" % (self._op.name, self._value_index)
@property
def device(self):
"""The name of the device on which this tensor will be produced, or None."""
return self._op.device
```
Tensor can be taken as target to run by session. Tensor contains all the information of Graph, and tracks data dependency.
Here is a detailed example:
```
>>> import tensorflow as tf
>>> c = tf.constant([[1.0, 2.0], [3.0, 4.0]])
>>> print c.graph
<tensorflow.python.framework.ops.Graph object at 0x10f256d50>
>>> d = tf.constant([[1.0, 1.0], [0.0, 1.0]])
>>> print d.graph
<tensorflow.python.framework.ops.Graph object at 0x10f256d50>
>>> e = tf.matmul(c, d)
>>> print e.graph
<tensorflow.python.framework.ops.Graph object at 0x10f256d50>
```
### Dynet
The core concept of symbolic API is `Expression`, and Dynet defines `Expression` class in C++.
A simple example is as follows:
```cpp
ComputationGraph cg;
Expression W = parameter(cg, pW);
Expression in = input(cg, xs[i]);
Expression label = input(cg, ys[i]);
Expression pred = W * in;
Expression loss = square(pred - label);
```
The input data and parameter are also represented by Expression. Every basci Expression corresponds to a Node. And input data is also a Node.
Expression has a data member ComputationGraph, and ComputationGraph will be modified in users' configuring process. Expression can be a running target, beacuse Expression contains all dependency.
Here is a detailed example:
write topology in C++
```
ComputationGraph cg;
Expression W = parameter(cg, pW);
cg.print_graphviz();
Expression pred = W * xs[i];
cg.print_graphviz();
Expression loss = square(pred - ys[i]);
cg.print_graphviz();
```
compile and print
```
# first print
digraph G {
rankdir=LR;
nodesep=.05;
N0 [label="v0 = parameters({1}) @ 0x7ffe4de00110"];
}
# second print
digraph G {
rankdir=LR;
nodesep=.05;
N0 [label="v0 = parameters({1}) @ 0x7ffe4de00110"];
N1 [label="v1 = v0 * -0.98"];
N0 -> N1;
}
# third print
digraph G {
rankdir=LR;
nodesep=.05;
N0 [label="v0 = parameters({1}) @ 0x7ffe4de00110"];
N1 [label="v1 = v0 * -0.98"];
N0 -> N1;
N2 [label="v2 = -1.88387 - v1"];
N1 -> N2;
N3 [label="v3 = -v2"];
N2 -> N3;
N4 [label="v4 = square(v3)"];
N3 -> N4;
}
```
### Conclusion
Actually, Symbol/Tensor/Expression in Mxnet/TensorFlow/Dynet are the same level concepts. We use a unified name Expression here, this level concept has following features:
- Users wirte topoloy with symbolic API, and all return value is Expression, including input data and parameter.
- Expression corresponds with a global Graph, and Expression can also be composed.
- Expression tracks all dependency and can be taken as a run target
doc/design/if_else_op.md
浏览文件 @ 0dc4b298
IfOp should have only one branch. An IfOp operator takes a `cond` variable whose value must be a vector of N boolean elements. Its return value has N instances. If cond[i] == True, input instance input[i] will go through true_block() and generate output[i]; otherwise it will produce output from false_bloack().
# The `IfElse` Operator
```python
import paddle as pd
PaddlePaddle's `IfElse` operator differs from TensorFlow's:
x = var()
y = var()
cond = var()
default_value = var()
b = pd.create_ifelseop(inputs=[x], output_num=1)
with b.true_block():
x = b.inputs(0)
z = operator.add(x, y)
b.set_output(0, operator.softmax(z))
with b.false_block():
x = b.inputs(0)
z = layer.fc(x)
b.set_output(0, operator.softmax(z))
out = b(cond)
```
- the TensorFlow version takes a scalar boolean value as the condition so that the whole mini-batch goes to either the true or the false branch, whereas
- the PaddlePaddle version takes a vector of boolean value as the condition, and instances corresponding to true values go to the true branch, those corresponding to false values go to the false branch.
## Example
The following PaddlePaddle program shows the usage of the IfElse operator:
If only true_block is set in an IfElseOp, a special case is that we can have a default value for false as:
```python
import paddle as pd
x = var()
y = var()
cond = var()
default_value = var()
b = pd.create_ifelseop(inputs=[x], output_num=1, default_value)
with b.true_block():
x = b.inputs(0)
z = operator.add(x, y)
b.set_output(0, operator.softmax(z))
x = minibatch([10, 20, 30]) # shape=[None, 1]
y = var(1) # shape=[1], value=1
z = minibatch([10, 20, 30]) # shape=[None, 1]
cond = larger_than(x, 15) # [false, true, true]
ie = pd.ifelse()
with ie.true_block():
d = pd.layer.add(x, y)
ie.output(d, pd.layer.softmax(d))
with ie.false_block():
d = pd.layer.fc(z)
ie.output(d, d+1)
o1, o2 = ie(cond)
```
out = b(cond)
A challenge to implement the `IfElse` operator is to infer those variables to be split, or, say, to identify the variable of the mini-batch or those derived from the mini-batch.
An equivalent C++ program is as follows:
```c++
namespace pd = paddle;
int x = 10;
int y = 1;
int z = 10;
bool cond = false;
int o1, o2;
if (cond) {
int d = x + y;
o1 = z;
o2 = pd::layer::softmax(z);
} else {
int d = pd::layer::fc(z);
o1 = d;
o2 = d+1;
}
```
where default_value is a list of vars for `cond` == False.
doc/design/images/graph_construction_example.dot
浏览文件 @ 0dc4b298
......@@ -33,7 +33,6 @@ digraph ImageClassificationGraph {
cost -> MSE_Grad [color=red];
d_cost -> MSE_Grad [color=red];
x -> MSE_Grad [color=red];
l -> MSE_Grad [color=red];
y -> MSE_Grad -> d_y [color=red];
......
doc/design/infer_var_type.md 0 → 100644
浏览文件 @ 0dc4b298
# Design Doc: InferVarType
## The Problem Posed
The variable in our design can hold variant types. Such as `LoDTensor` and `SelectedRows`. An operator should be able to inference the variable types of its output.
For example, a `lookup table` operator takes two `LoDTensor`; one is a float tensor as the embedding table, the other is an int tensor as word ID. The gradient operator of `lookup table` will generate a `SelectedRows` as its output. A `sum` operator can take both `LoDTensor` and `SelectedRows` as its inputs and will generate a `LoDTensor` if any of its inputs is `LoDTensor`, otherwise, the `sum` operator will generate `SelectedRows` as its output.
The variable type will be constant at runtime. Every variable's type can either be set by the user (input data and parameter) or be inferred by the operator in compile time.
## Proposed Solution
The `InferVarType` is a compile-time function which is registered to each operator. The inferface of that function is:
```c++
using InferVarTypeFN = std::function<
void (const OpDescBind& /*op_desc*/, BlockDescBind* /*block*/)>;
```
It takes an operator description as its input and will write the output variable type and store them in block description.
The `InferVarTypeFN` will be registered in `OpInfo`, to replace `infer_var_type_` field. The `OpInfo` should be
```cpp
struct OpInfo {
InferVarTypeFN infer_var_type_;
...
};
```
The default `InferVarType` will set output type as `LoDTensor`. It can be done by `GetInferVarType()`.
```cpp
void DefaultInferVarType(const OpDescBind& op_desc, BlockDescBind* block) {
// set the output type of variable as `LoDTensor`.
// ...
}
struct OpInfo {
InferVarTypeFN infer_var_type_;
InferVarTypeFN GetInferVarType() const {
if (infer_var_type_) {
return infer_var_type_;
} else {
return DefaultInferVarType;
}
}
};
```
## Register InferVarType
We provide a thin base class for registering an `InferVarTypeFN`. To use a base class will ease the implementation of registry since we can detect the registry entry is an `InferVarTypeFN` or not.
```cpp
class VarTypeInferer {
public:
virtual void operator()(const OpDescBind& op_desc, BlockDescBind* block) const = 0;
}
```
Operator developers can write the specialize `VarTypeInferer` as follow.
```cpp
class SpecialVarTypeInferer : public VarTypeInferer {
public:
virtual void operator()(const OpDescBind& op_desc, BlockDescBind* block) const {
// .. own logic
}
}
```
Then user can register the `InferVarType` just like `GradOpDescMaker` and `OpInfoMaker`.
```
REGISTER_OPERATOR(some_op, OpType, SpecialVarTypeInferer, ...);
```
doc/design/model_format.md 0 → 100644
浏览文件 @ 0dc4b298
# Design Doc: Model Format
## Motivation
The model is the output of training process. One complete model consists of two parts, namely, the **topology** and the **parameters**. To support industrial deployment, we need to make the model format must be self-completed and do not expose any training source code.
As a result, In PaddlePaddle, the **topology** represents as a [ProgramDesc](https://github.com/PaddlePaddle/Paddle/blob/1c0a4c901c9fc881d120249c703b15d1c50dae7d/doc/design/program.md), which describes the model structure. The **parameters** contain all the trainable weights in the model, we must support large size parameter, and efficient serialization/deserialization.
## Implementation
The topology is saved as a plain text, in detail, a self-contain protobuf file.
The parameters are saved as a binary file. As we all know, the protobuf message has the limits of [64M size](https://developers.google.com/protocol-buffers/docs/reference/cpp/google.protobuf.io.coded_stream#CodedInputStream.SetTotalBytesLimit.details). We do a (benchmark experiment)[https://github.com/PaddlePaddle/Paddle/pull/4610], its result shows protobuf is not fit in this scene.
As a result, we design a particular format for tensor serialization. By default, arbitrary tensor in Paddle is a [LoDTensor](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/lod_tensor.md), and has a description information proto of (LoDTensorDesc)[https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/framework.proto#L99]. We save the DescProto as the byte string header, it contains the necessary information, such as the `dims`, the `name` of the tensor, and the `LoD` information in [LoDTensor](https://github.com/PaddlePaddle/Paddle/blob/1c0a4c901c9fc881d120249c703b15d1c50dae7d/paddle/framework/lod_tensor.md). Tensor stores value in a continuous memory buffer, for speed we dump the raw memory to disk and save it as the byte string content. So, the binary format of one tensor is,
|HeaderLength|ContentLength|**LoDTensorDesc**|**TensorValue**|
In detail, tensor's byte view as the table shows. Note that all the signed value written in little-endian.
```text
[offset] [type] [description]
0004 4 bytes integer HeaderLength, the length of LoDTensorDesc
0008 4 bytes integer ContentLength, the length of LodTensor Buffer
0009 1 bytes char TensorDesc
00010 1 bytes char TensorDesc
...
00100 1 bytes char TensorValue
00101 1 bytes char TensorValue
00102 1 bytes char TensorValue ..
...
```
## Summary
We introduce the model format, the `ProgramDesc` describe the **topology**, and a bunch of particular format binary tensors describes the **parameters**.
doc/design/optimizer.md 0 → 100644
浏览文件 @ 0dc4b298
## Optimizer Design
### The Problem
A PaddlePaddle program, or a block, is a sequence of operators operating variables. A training program needs to do three kinds of works:
1. the forward pass, which computes intermediate results and the cost(s),
1. the backward pass, which derives gradients from intermediate results and costs, and
1. the optimization pass, which update model parameters to optimize the cost(s).
These works rely on three kinds of operators:
1. forward operators,
1. gradient operators, and
1. optimization operators.
It's true that users should be able to create all these operators manually by calling some low-level API, but it would be much more convenient if they could only describe the forward pass and let PaddlePaddle create the backward and optimization operators automatically.
In this design, we propose a high-level API that automatically derives the optimisation pass and operators from the forward pass.
### High-level Python API to describe the training process
1. User write code to describe the network:
```python
images = layer.data("images")
labels = layer.data("labels")
w1 = pd.var("w1")
b1 = pd.var("b1")
hidden = layer.fc(images, w=w1, b=b1)
cost = layer.mse(hidden, labels)
```
The above code snippet will create forward operators in [Block](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/block.md).
2. Users create a certain kind of Optimizer with some argument.
```python
optimizer = AdagradOptimizer(learing_rate=0.001)
```
3. Users use the optimizer to `minimize` a certain `cost` through updating parameters in parameter_list.
```python
opt_op_list = optimizer.minimize(cost, parameter_list=[w1, b1])
```
The above code snippet will create gradient and optimization operators in Block. The return value of `minimize()` is list of optimization operators that will be run by session.
4. Users use Session/Executor to run this opt_op_list as target to do training.
```python
sess.run(target= opt_op_list, ...)
```
#### Optimizer Python interface:
```python
class Optimizer(object):
"""Optimizer Base class.
"""
def __init__(self):
pass
def create_optimization_pass(self, parameters_and_grads):
"""Add optimization operators to update gradients to variables.
Args:
parameters_and_grads: a list of (variable, gradient) pair to update.
Returns:
optmization_op_list: a list of optimization operator that will update parameter using gradient.
"""
return None
def minimize(self, loss, parameter_list):
"""Add operations to minimize `loss` by updating `parameter_list`.
This method combines interface `append_backward_ops()` and
`create_optimization_pass()` into one.
"""
params_grads = self.create_backward_pass(loss, parameter_list)
update_ops = self.create_optimization_pass(params_grads)
return update_ops
```
Users can inherit the Optimizer above to create their own Optimizer with some special logic, such as AdagradOptimizer.
doc/design/program.md
浏览文件 @ 0dc4b298
# Design Doc: ProgramDesc
# Design Doc: PaddlePaddle Programs
The basic structure of a PaddlePaddle program is some nested blocks, as a C++ or Java program.
## Compile and Execution
A PaddlePaddle program consists of two parts -- the first generates a `ProgramDesc` protobuf message that describes the program, and the second runs this message using a C++ class `Executor`.
As described in [graph.md](./graph.md), the first five lines of the following PaddlePaddle program
A simple example PaddlePaddle program can be found in [graph.md](./graph.md):
```python
x = layer.data("images")
......@@ -13,36 +15,112 @@ optimize(cost)
train(cost, reader=mnist.train())
```
generates, or compiles, a PaddelPaddle program, which is represented by the following protobuf message:
The first five lines of the following PaddlePaddle program generates, or, compiles, the `ProgramDesc` message. The last line runs it.
```protobuf
message ProgramDesc {
repeated BlockDesc blocks = 1;
## Programs and Blocks
The basic structure of a PaddlePaddle program is some nested blocks, as a C++ or Java program.
- program: some nested blocks
- [block](./block.md):
- some local variable definitions, and
- a sequence of operators
The concept of block comes from usual programs. For example, the following C++ program has three blocks:
```c++
int main() { // block 0
int i = 0;
if (i < 10) { // block 1
for (int j = 0; j < 10; j++) { // block 2
}
}
return 0;
}
```
The following PaddlePaddle program has three blocks:
```python
import paddle as pd // block 0
x = minibatch([10, 20, 30]) # shape=[None, 1]
y = var(1) # shape=[1], value=1
z = minibatch([10, 20, 30]) # shape=[None, 1]
cond = larger_than(x, 15) # [false, true, true]
ie = pd.ifelse()
with ie.true_block(): // block 1
d = pd.layer.add_scalar(x, y)
ie.output(d, pd.layer.softmax(d))
with ie.false_block(): // block 2
d = pd.layer.fc(z)
ie.output(d, d+1)
o1, o2 = ie(cond)
```
## `BlockDesc` and `ProgramDesc`
All protobuf messages are defined in `framework.proto`.
`BlockDesc` is straight-forward -- it includes local variable definitions, `vars`, and a sequence of operators, `ops`.
```protobuf
message BlockDesc {
required int32 parent = 1;
repeated VarDesc vars = 2;
repeated OpDesc ops = 3;
}
```
The parent ID indicates the parent block so that operators in a block can refer to variables defined locally and also those defined in their ancestor blocks.
All hierarchical blocks in a program are flattened and stored in an array. The block ID is the index of the block in this array.
```protobuf
message ProgramDesc {
repeated BlockDesc blocks = 1;
}
```
### Global Block
The global block is the first one in the above array.
## Operators that Use Blocks
In the above example, the operator `IfElseOp` has two blocks -- the true branch and the false branch.
The definition of `OpDesc` shows that an operator could have some attributes:
```protobuf
message OpDesc {
AttrDesc attrs = 1;
...
}
```
and an attribute could be of type block, which is, in fact, a block ID as described above:
```
message AttrDesc {
required AttrType type = 1;
required string name = 1;
// index into ProgramDesc::blocks when type==BLOCK
optional int32 block = 2;
enum AttrType {
INT = 1,
STRING = 2,
...
BLOCK = ...
}
required AttrType type = 2;
optional int32 block = 10; // when type == BLOCK
...
}
```
When each of the first five lines runs, related Python function, e.g., `layer.fc`, calls C++ InferShape functions. This InferShape function needs to access the properties of VarDesc's accessed by the current OpDesc. These VarDesc's might not be defined in the current block, but in some ancestor blocks. This requires that we can trace the parent of a block.
A nested block is often an attribute of an operator, most likely, an IfElseOp or a WhileOp. In above solution, all blocks are in `ProgramDesc::blocks`, this implicitly assigns a zero-based ID to each block -- the index of the block in `ProgramDesc::blocks`. So that `AttrDesc::block` could be an integer block ID.
## InferShape
With this design, the InferShape function should take the following parameters:
......
doc/design/prune.md 0 → 100644
浏览文件 @ 0dc4b298
# Prune
## Motivation
We want to support running inference, training and checkpointing in one `ProgramDesc`. We implement
`void Prune(const ProgramDesc* input, ProgramDesc* output)` function, which takes a `ProgramDesc`
and generate a pruned `ProgramDesc`.
## Challenge
Pruning need to support both variables and operators being evaluation targets. Consider the following
different situations.
```python
# Case 1: run foward pass.
cost_np = session.run(target=cost)
# Case 2: run backward passing.
opts_np, _ = session.run(target=[cost, opt])
# Case 3: run checkpointing
_ = session.run(target=checkpoint)
```
## Solution
To support evaluation of operators, we add `is_target` field in the `OpDesc`.
```c++
message OpDesc {
required string type = 3;
repeated Var inputs = 1;
repeated Var outputs = 2;
repeated Attr attrs = 4;
optional bool is_target = 5 [ default = false ];
};
```
To support evaluation of variables, we add [fetch_op](https://github.com/PaddlePaddle/Paddle/pull/4599).
For each variable in the `target`, we insert a `fetch_op` into the `ProgramDesc` with `variable` being
`fetch_op`'s input. Then we also set `fetch_op` is a target.
### Algorithm
If an operator needs to be run, it must fall into one of the following cases:
1. It is the target.
2. It is depended by some other ops, meaning its output is some other op's input.
The first case can be checked by `op_desc.is_traget()` . The second case can be implement as
```c++
bool HasDependentVar(const OpDesc& op_desc, const std::set<string>& dependent_vars) {
for (auto& var : op_desc.outputs()) {
for (auto& argu : var.arguments()) {
if (dependent_vars.count(argu) != 0) {
return true;
}
}
}
return false;
}
```
Then the whole algorithm can be implemented as the following [code](https://github.com/tonyyang-svail/Paddle/blob/prune_impl/paddle/framework/prune.cc).
doc/design/python_api.md 0 → 100644
浏览文件 @ 0dc4b298
# Design Doc: Python API
Due to the refactorization of the PaddlePaddle core, we need Python classes to construct corresponding protobuf messages that describe a DL program.
| Python classes | Protobuf messages |
| --- | --- |
| Program | ProgramDesc |
| Block | BlockDesc |
| Operator | OpDesc |
| Variable | VarDesc |
Please be aware that these Python classes need to maintain some construction-time information, which are not part of the protobuf messages.
## Core Concepts
### Program
A `ProgramDesc` describes a [DL program](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/program.md), which is composed of an array of `BlockDesc`s. The `BlockDesc`s in a `ProgramDesc` can have a tree-like hierarchical structure. However, the `ProgramDesc` onlys stores a flattened array of `BlockDesc`s. A `BlockDesc` refers to its parent block by its index in the array. For example, operators in the step block of an RNN operator need to be able to access variables in its ancestor blocks.
Whenever we create a block, we need to set its parent block to the current block, hence the Python class `Program` needs to maintain a data member `current_block`.
```python
class Program(objects):
def __init__(self):
self.desc = core.NewProgram() # a C++ ProgramDesc pointer.
self.blocks = vector<Block>()
self.blocks.append(Block(self, -1)) # the global block
self.current_block = 0 # initialized to the global block
def global_block():
return self.blocks[0]
def current_block():
return self.get_block(self.current_block)
def rollback():
self.current_block = self.current_block().parent_idx
def create_block():
new_block_idx = len(self.block)
self.blocks.append(Block(self, self.current_block))
self.current_block = new_block_idx
return current_block()
```
`Program` is an accessor to the protobuf message `ProgramDesc`, which is created in C++ space, because the InferShape function is in C++, which manipulates `VarDesc` messages, which are in turn members of `BlockDesc`, which is a member of `ProgramDesc`.
`Program` creates the first block as the global block in its constructor. All parameters and their initializer operators are in the global block.
### Block
A [Block](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/block.md) includes
1. a map from variable names to an instance of the Python `Variable` class, and
1. a list of `Operator` instances.
```python
class Block(objects):
def __init__(self, program, parent_idx):
self.desc = core.NewBlock(program.desc)
self.program = program
self.vars = map<string, Variable>()
self.ops = vector<Operator>()
self.parent_idx = parent_idx
def create_var(self, ...):
return Variable(self, ...)
def _create_global_var(self, ...):
program.global_block().create_var(...)
def create_parameter(self, name, ...):
# Parameter is a subclass of variable. See Parameter section for details.
self.vars[name] = Parameter(self._create_global_var(...), ...)
return self.vars[name]
def append_operator(self, ...):
self.ops.append(Operator(self, ...))
def prepend_operator(self, ...): # Parameter's ctor prepands initialize operators.
self.ops.prepend(Operator(self, ...))
```
`create_parameter` is necessary because parameters are global variables, defined in the global block, but can be created in some sub-blocks. For example, an FC layer in the step block of an RNN operator.
`prepend_operator` is necessary because the constructor of `Parameter` needs to create the initialize (or load) operator of the parameter, and would like to put it in the *preamble* of the global block.
### Operator
The `Operator` class fills in the `OpDesc` message and calls the C++ function `InferShape` to infer the output shapes from the input shapes.
```python
class Operator(object):
def __init__(self,
block, # Block
type, # string
inputs, # dict<string, Variable>
outputs,# dict<stirng, Variable>
attrs # dict<string, Any>
):
self.desc = core.NewOpDesc(block.desc, type, inputs, outputs, attrs)
core.infer_shape(self.desc, inputs, outputs)
def type(self):
return self.desc.type()
```
`Operator` creates the `OpDesc` message in C++ space, so that it can call the `InferShape` function, which is in C++.
### Variable
Operators take Variables as its inputs and outputs.
```python
class Variable(object):
def __init__(self,
block=None, # Block
name=None, # string
shape, # tuple
dtype="float32", # string
lod_level=None # int
):
if name is None:
name = unique_name_generator()
self.name = name
self.block = block
self.desc = core.NewVarDesc(block.desc, name, shape, lod_level)
self.writer = None
```
Please be aware of `self.writer`, that tracks operator who creates the variable. It possible that there are more than one operators who write a variable, but in Python space, each write to a variable is represented by a Variable class. This is guaranteed by the fact that **`core.NewVarDesc` must NOT create a new `VarDesc` message if its name already exists in the specified block**.
### Parameter
A parameter is a global variable with an initializer (or load) operator.
```python
class Parameter(Variable):
def __init__(self,
block=None, # Block
name=None, # string
shape, # tuple
dtype="float32", # string
lod_level=None # int
trainable, # bool
initialize_op_attrs,
optimize_op_attrs):
super(Parameter, self).__init__(block, name, shape, dtype, lod_level)
self.trainable = trainable
self.optimize_op_attrs = optimize_op_attrs
block.prepend(Operator(block, # Block
initialize_op_attrs['type'], # string
None, # no inputs
self, # output is the parameter
initialize_op_attrs)
```
When users create a parameter, they can call
```python
program.create_parameter(
...,
init_attr={
type: "uniform_random",
min: -1.0,
max: 1.0,
})
)
```
In above example, `init_attr.type` names an initialize operator. It can also name the load operator
```python
init_attr={
type: "load",
filename: "something.numpy",
}
```
`optimize_op_attrs` is not in the `VarDesc` message, but kept in the Python instance, as it will be used in the Python space when creating the optimize operator's `OpDesc`, and will be in the `OpDesc` message.
## Layer Function
A layer is a Python function that creates some operators and variables. Layers simplify the work of application programmers.
Layer functions take `Variable` and configuration parameters as its input and return the output variable(s).
For example, `FullyConnected` take one or more variable as its input. The input could be input data or another layer's output. There are many configuration options for a `FullyConnected` layer, such as layer size, activation, parameter names, initialization strategies of parameters, and so on. The `FullyConnected` layer will return an output variable.
### Necessity for reusing code between layer functions
There are a lot of code that can be reused. Such as
* Give the default value of configuration. e.g., default initialize strategy for parameters is uniform random with `min = -1.0`, `max = 1.0`. and default initialize strategy for bias is to fill zero.
* Append the activation operator.
* Create a temporary variable.
* Create parameter.
* Generate a unique name.
* Add a bias.
* ...
A mechanism to reuse code between layer functions is necessary. It will be around [150 lines of code](https://github.com/PaddlePaddle/Paddle/pull/4724/files#diff-823b27e07e93914ada859232ae23f846R12) if we write a `FullyConnected` layer without any helper functions.
### Comparision between global functions and helper class
The `FullyConnected` layer will be as follow when we provide global functions:
```python
def fc_layer(input, size, param_attr=None, bias_attr=None, act=None, name=None):
if name is None:
name = unique_name("fc")
input = multiple_input(input)
param_attr = default_param_attr(param_attr)
param_attr = multiple_param_attr(param_attr, len(input))
# mul
mul_results = []
for ipt, attr in zip(input, param_attr):
shape = ipt.shape[1:] + [size]
w = g_program.global_block().create_parameter(shape, ipt.dtype, name, attr)
tmp = create_tmp_var(name)
g_program.current_block().append_op("mul", {ipt, w}, {tmp})
mul_results.append(tmp)
# add sum
...
# add bias
...
# add activation
...
return out
```
We can provide many helpers functions for layer developers. However, there are several disadvantages for global helper functions:
1. We need a namespace for these methods, then layer developers can quickly figure out what method they can use.
2. Global functions will force layer developers to pass its parameter time by time.
So we provide a helper class, `LayerHelper`, to share code between layer functions. The `FullyConnected` Layer will be as follow.
```python
def fc_layer(input, size, param_attr=None, bias_attr=None, act=None, name=None):
helper = LayerHelper(locals()) # pass all parameter to LayerHelper
mul_results = []
for ipt, param in helper.iter_multiple_input_and_param():
w = helper.create_parameter(shape=ipt.shape[1:] + [size], dtype = ipt.dtype)
tmp = helper.create_tmp_variable()
helper.append_op('mul', {ipt, w}, {tmp})
mul_results.append(tmp)
pre_bias = helper.add_sum(mul_results)
pre_activation = helper.add_bias(pre_bias)
return helper.add_activation(pre_activation)
```
We not only use the fewer lines of code to write `fc_layer` but also make the code clearer to understand. At the same time, layer developers can figure out what function they can invoke by typing `helper.` in a python editor.
### Implementation of layer helper
We just keep all parameters of a layer function as a dictionary in layer helper as a private data member. Every method of layer helper will look up the dictionary after it is invoked. In that way, we can implement a layer helper for all layer functions even some layer does not contain some operator. For example, The `activation` is used by the FullyConnected layer or convolution layers, but a cross-entropy layer does not use it. The example code of `add_activation` are:
```python
class LayerHelper(object):
def __init__(self, **kwargs): # kwargs is short for `keyword arguments`
self.kwargs = kwargs
def add_activation(self, input_var):
act = self.kwargs.get("act", None) # default value is None
if act is None: # do nothing if no act
return input_var
tmp = self.create_tmp_var(self)
self.append_op(type=act, input=input_var, output=tmp)
return tmp
```
## Optimizer
[Optimizer Design Doc](./optimizer.md)
doc/design/refactor/session.md 0 → 100644
浏览文件 @ 0dc4b298
# Design Doc: Session
## Abstract
The *session* object encapsulates the environment in which the
computation graph is executed.
We will have the *local* session and *remote* session, they offer the
same [interface](#interface). The local session encapsulates the local
runtime environment and the remote session encapsulates the cluster
runtime environment.
The local runtime environment contains:
1. computation devices (i.e., CPU, GPU) handles, and
1. the [scope](../scope.md) which holds all variables.
The remote runtime environment contains:
1. computation devices (i.e., CPU and GPU on node 0, 1) in a cluster,
and
1. the distributed [scope](../scope.md) in a cluster which holds all
variables.
The user can create a remote session on Paddle Cloud and evaluate the
computation graph with it. In this way, the user can control the
remote computation resource in a cluster from his local computer.
## Background
The current design has an implicit global session in which
`paddle.eval()` is executed. The pain point is:
Since the user is not able to explicitly switch between runtime
environments, the user cannot run a topology in two independent
environments.
For example, in reinforcement learning, the user may want to have a
stale model for inference and a fresh model for training, and only
replace the stale model with the fresh model periodically.
Furthermore, we have no concept that encapsulates a remote environment
that executes a computation graph.
We need the session object to address above issues.
## Session
A session is an object that owns the runtime environment. All
computations are executed through `session.eval()`.
### Interface
```python
eval(
targets,
feed_dict=None,
)
```
Evaluates the target Operations or Variables in `targets`.
- *targets*: the evaluation targets. Can be a single Operation or
Variable, or a list with the Operations or Variables as
elements. The value returned by `eval()` has the same shape as the
`target` argument.
The PaddlePaddle program is represented by
the [ProgramDesc](../design/program.md), `eval()` will infer the
ProgramDesc from the given targets and run the PaddlePaddle
program. Please
see
[this graph](./distributed_architecture.md#local-training-architecture) for
the detailed illustration for the local session
and
[this graph](./distributed_architecture.md#distributed-training-architecture) for
the detailed illustration for the remote session.
- *feed_dict*: a dictionary that contains the tensors which override
the edges of the computation graph.
feed_dict not only can provide the input data, it can override any
OP's input as well:
```python
a = pd.constant(2.0, name="a")
b = pd.variable(name="b")
c = pd.mul(a,b)
sess.eval(targets=c, feed_dict={"b":3.0}) # returns 6.0
```
```python
close()
```
Closes the session and releases the scope that the session owns.
### Create a Local Session
```python
session(
devices=None
)
```
Creates a new session. One session owns one global scope, so creating
multiple sessions will create different scopes.
- *devices*: a single `string` or a list of `string` of device names,
the corresponding devices will be the computation devices for
`eval()`. If not specified, all available devices (e.g., all GPUs)
will be used. The user doesn't need to specify the CPU device since
it will be always used. Multiple sessions can use the same device.
#### Example
```Python
a = paddle.constant(1.0)
b = paddle.constant(2.0)
c = a + b
sess = paddle.session(devices=["gpu:0", "gpu:1", "fpga:0"])
sess.eval(c)
sess.close()
```
### Create a Remote Session
```python
create_cloud_job(
name,
num_trainer,
mem_per_trainer,
gpu_per_trainer,
cpu_per_trainer,
num_ps,
mem_per_ps,
cpu_per_ps,
)
```
Creates a Paddle Cloud job. Fails if the job name exists.
```python
get_cloud_job(
name
)
```
Gets a Paddle Cloud job.
```python
remote_session(
job
)
```
- *job*: the Paddle Cloud job.
#### Example
```Python
reader = paddle.reader.recordio("/pfs/home/peter/mnist-train-*") # data stored on Paddle Cloud
image = reader.column(0)
label = reader.column(1)
fc1 = paddle.op.fc(image, size=256, act="sigmoid")
fc2 = paddle.op.fc(fc1, size=10, act="softmax")
cost = paddle.op.cross_entropy(fc2, label)
opt = paddle.optimizer.sgd(cost)
job = paddle.create_cloud_job("test", 3, "1G", 1, 1, 2, "1G", 1)
sess = paddle.remote_ession(job)
for i in range(1000):
sess.eval(opt)
sess.close()
```
doc/design/refactorization.md
浏览文件 @ 0dc4b298
# Design Doc: Refactorization Overview
The goal of refactorizaiton include:
The goals of refactoring include:
1. Make it easy for external contributors to write new elementory computaiton operations.
1. Make the codebase clean and readable.
1. Introduce a new design of computation representation -- a computation graph of operators and variables.
1. The graph representation helps implementing auto-scalable and auto fault recoverable distributed computing.
1. Making it easy for external contributors to write new elementary computation operations.
1. Making the codebase clean and readable.
1. Designing a new computation representation -- a computation graph of operators and variables.
1. Implementing auto-scalability and auto fault recoverable distributed computing with the help of computation graphs.
## Computation Graphs
1. PaddlePaddle represent the computation, training and inference of DL models, by computation graphs.
1. PaddlePaddle represents the computation, training and inference of Deep Learning models, by computation graphs.
1. Please dig into [computation graphs](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/graph.md) for a solid example.
1. Please refer to [computation graphs](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/graph.md) for a concrete example.
1. Users write Python programs to describe the graphs and run it (locally or remotely).
1. Users write Python programs to describe the graphs and run them (locally or remotely).
1. A graph is composed of *variables* and *operators*.
1. The description of graphs must be able to be serialized/deserialized, so it
1. The description of graphs must be serializable/deserializable, so that:
1. could to be sent to the cloud for distributed execution, and
1. be sent to clients for mobile or enterprise deployment.
1. It can be sent to the cloud for distributed execution, and
1. It can be sent to clients for mobile or enterprise deployment.
1. The Python program do
1. The Python program does two things
1. *compilation*: runs a Python program to generate a protobuf message representation of the graph and send it to
1. *Compilation* runs a Python program to generate a protobuf message representation of the graph and send it to
1. the C++ library `libpaddle.so` for local execution,
1. the master process of a distributed training job for training, or
1. the server process of a Kubernetes serving job for distributed serving.
1. *execution*: according to the protobuf message, constructs instances of class `Variable` and `OperatorBase`, and run them.
1. *Execution* executes the graph by constructing instances of class [`Variable`](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/variable.h#L24) and [`OperatorBase`](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/operator.h#L70), according to the protobuf message.
## Description and Realization
## Description and Realization of Computation Graph
At compile time, the Python program generates protobuf message representation of the graph, or the description of the graph.
At compile time, the Python program generates a protobuf message representation of the graph, or a description of the graph.
At runtime, the C++ program realizes the graph and run it.
At runtime, the C++ program realizes the graph and runs it.
| | Representation (protobuf messages) | Realization (C++ class objects) |
|---|---|---|
......@@ -42,30 +42,31 @@ At runtime, the C++ program realizes the graph and run it.
|Operation|[OpDesc](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/framework.proto#L35)|[Operator](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/operator.h#L64)|
|Block|BlockDesc|Block|
The word *graph* is exchangable with *block* in this document. A graph represent computation steps and local variables as a C++/Java program block, or a pair of { and }.
The word *graph* is interchangeable with *block* in this document. A graph consists of computation steps and local variables similar to a C++/Java program block, or a pair of parentheses(`{` and `}`).
## Compilation and Execution
1. Run an applicaton Python program to describe the graph. In particular,
1. Run a Python program to describe the graph. In particular, the Python application program does the following:
1. create VarDesc to represent local/intermediate variables,
1. create operators and set attributes,
1. validate attribute values,
1. inference the type and the shape of variables,
1. plan for memory-reuse for variables,
1. generate backward and optimization part of the Graph.
1. possiblly split the graph for distributed training.
1. Create `VarDesc` to represent local/intermediate variables,
1. Create operators and set attributes,
1. Validate attribute values,
1. Infer the type and the shape of variables,
1. Plan memory-reuse for variables,
1. Generate the backward graph
1. Add optimization operators to the computation graph.
1. Optionally, split the graph for distributed training.
1. The invocation of `train` or `infer` in the application Python program:
1. The invocation of `train` or [`infer`](https://github.com/PaddlePaddle/Paddle/blob/develop/python/paddle/v2/inference.py#L108) methods in the Python program does the following:
1. create a new Scope instance in the [scope hierarchy](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/scope.md) for each run of a block,
1. Create a new Scope instance in the [scope hierarchy](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/scope.md) for each run of a block,
1. realize local variables defined in the BlockDesc message in the new scope,
1. a scope is similar to the stack frame in programming languages,
1. create an instance of class `Block`, in which,
1. Create an instance of class `Block`, in which,
1. realize operators in the BlockDesc message,
1. run the Block by calling
1. Run the Block by calling
1. `Block::Eval(vector<Variable>* targets)` for forward and backward computations, or
1. `Block::Eval(vector<Operator>* targets)` for optimization.
......@@ -76,14 +77,14 @@ The word *graph* is exchangable with *block* in this document. A graph represen
Compile Time -> IR -> Runtime
```
### Benefit
### Benefits of IR
- Optimization
```text
Compile Time -> IR -> Optimized IR -> Runtime
```
- Send automatically partitioned IR to different nodes.
- Automatic data parallel
- Automatically send partitioned IR to different nodes.
- Automatic Data Parallelism
```text
Compile Time
|-> Single GPU IR
......@@ -92,7 +93,7 @@ Compile Time -> IR -> Runtime
|-> Node-1 (runs trainer-IR-1)
|-> Node-2 (runs pserver-IR)
```
- Automatic model parallel (planned for future)
- Automatic Model Parallelism (planned for future)
---
......@@ -105,10 +106,10 @@ Compile Time -> IR -> Runtime
# Operator
![class_diagram](http://api.paddlepaddle.org/graphviz?dot=https://gist.githubusercontent.com/reyoung/53df507f6749762675dff3e7ce53372f/raw/dd598e8f1976f5759f58af5e5ef94738a6b2e661/op.dot)
* `Operator` is the fundamental building block as the user interface.
* Operator stores input/output variable name, and attributes.
* The `InferShape` interface is used to infer output variable shapes by its input shapes.
* Use `Run` to compute `input variables` to `output variables`.
* `Operator` is the fundamental building block of the user interface.
* Operator stores input/output variable names and attributes.
* The `InferShape` interface is used to infer the shape of the output variables based on the shapes of the input variables.
* Use `Run` to compute the `output` variables from the `input` variables.
---
......@@ -126,30 +127,29 @@ Compile Time -> IR -> Runtime
# Why separate Kernel and Operator
* Separate GPU and CPU code.
* Make Paddle can run without GPU.
* Make one operator (which is user interface) can contain many implementations.
* Same mul op, different FP16, FP32 Kernel. different MKL, eigen kernel.
* Make Paddle capable of running without GPU.
* Make one operator (which is a user interface) and create many implementations.
* For example, same multiplication op can have different implementations kernels such as FP16 kernel, FP32 kernel, MKL, eigen kernel.
---
# Libraries for Kernel development
* `Eigen::Tensor` contains basic math and element-wise functions.
* Note that `Eigen::Tensor` has broadcast implementation.
* Limit number of `tensor.device(dev) = ` in your code.
* `thrust::tranform` and `std::transform`.
* `thrust` has the same API as C++ standard library. Using `transform` can quickly implement a customized elementwise kernel.
* `thrust` has more complex API, like `scan`, `reduce`, `reduce_by_key`.
* Limit the number of `tensor.device(dev) = ` in your code.
* `thrust::transform` and `std::transform`.
* `thrust` has the same API as C++ standard library. Using `transform`, one can quickly implement customized element-wise kernels.
* `thrust`, in addition, supports more complex APIs, like `scan`, `reduce`, `reduce_by_key`.
* Hand-writing `GPUKernel` and `CPU` code
* Do not write `.h`. CPU Kernel should be in `.cc`. GPU kernel should be in `.cu`. (`GCC` cannot compile GPU code.)
* Do not write in header (`.h`) files. CPU Kernel should be in cpp source (`.cc`) and GPU kernels should be in cuda (`.cu`) files. (GCC cannot compile GPU code.)
---
# Operator Register
# Operator Registration
## Why register is necessary?
## Why is registration necessary?
We need a method to build mappings between Op type names and Op classes.
## How to do the register?
Maintain a map, whose key is the type name and value is corresponding Op constructor.
## How is registration implemented?
Maintaining a map, whose key is the type name and the value is the corresponding Op constructor.
---
# The Registry Map
......@@ -169,7 +169,7 @@ Maintain a map, whose key is the type name and value is corresponding Op constru
# Related Concepts
### Op_Maker
It's constructor takes `proto` and `checker`. They are compeleted during Op_Maker's construction. ([ScaleOpMaker](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/operators/scale_op.cc#L37))
It's constructor takes `proto` and `checker`. They are completed during Op_Maker's construction. ([ScaleOpMaker](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/operators/scale_op.cc#L37))
### Register Macros
```cpp
......@@ -177,34 +177,30 @@ REGISTER_OP(op_type, op_class, op_maker_class, grad_op_type, grad_op_class)
REGISTER_OP_WITHOUT_GRADIENT(op_type, op_class, op_maker_class)
```
### `USE` Macros
make sure the registration process is executed and linked.
---
# Register Process
1. Write Op class, as well as its gradient Op class if there is.
2. Write Op maker class. In the constructor, describe its inputs, outputs, and attributes.
3. Invoke macro `REGISTER_OP`. The macro will
1. call maker class to complete `proto` and `checker`
2. with the completed `proto` and `checker`, build a new key-value pair in the `OpInfoMap`
4. Invoke `USE` macro in where the Op is used to make sure it is linked.
# Registration Process
1. Write an Op class and its gradient Op class, if required.
2. Write an Op maker class. In the constructor of this class, describe the inputs, outputs and attributes of the operator.
3. Invoke the macro `REGISTER_OP`. This macro will
1. Call maker class to complete `proto` and `checker`
2. Using the completed `proto` and `checker`, it will add a new key-value pair to the `OpInfoMap`
---
# Backward Module (1/2)
### Create Backward Operator
- Mapping from forwarding Op to backward Op
- Mapping from forward Op to backward Op
![backward](https://gist.githubusercontent.com/dzhwinter/a6fbd4623ee76c459f7f94591fd1abf0/raw/61026ab6e518e66bde66a889bc42557a1fccff33/backward.png)
---
# Backward Module (2/2)
### Build Backward Network
- **Input** graph of forwarding operators
- **Output** graph of backward operators
- **corner case in construction**
- shared variable => insert `Add` operator
- no gradient => insert `fill_zero_grad` operator
- recursive netOp => call `Backward` recursively
- **Input**: a graph of forward operators
- **Output**: a graph of backward operators
- **Corner cases in construction**
- Shared Variables => insert an `Add` operator to combine gradients
- No Gradient => insert a `fill_zero_grad` operator
- Recursive NetOp => call `Backward` recursively
- RNN Op => recursively call `Backward` on stepnet
- RNN Op => recursively call `Backward` on stepnet
......@@ -213,41 +209,41 @@ make sure the registration process is executed and linked.
* `Tensor` is an n-dimension array with type.
* Only dims and data pointers are stored in `Tensor`.
* All operators on `Tensor` is written in `Operator` or global functions.
* variable length Tensor design [LoDTensor](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/lod_tensor.md)
* `Variable` is the inputs and outputs of an operator. Not just `Tensor`.
* step_scopes in RNN is a variable and not a tensor.
* `Scope` is where variables store at.
* map<string/*var name */, Variable>
* `Scope` has a hierarchical structure. The local scope can get variable from its parent scope.
* All operations on `Tensor` are written in `Operator` or global functions.
* Variable length Tensor design [LoDTensor](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/lod_tensor.md)
* `Variable` instances are the inputs and the outputs of an operator, not just `Tensor`.
* `step_scopes` in RNN is a variable and not a tensor.
* `Scope` is where variables are stored.
* map<string `var name`, Variable>
* `Scope` has a hierarchical structure. The local scope can get variables from its parent scope.
---
# Block (in design)
## the difference with original RNNOp
- as an operator is more intuitive than `RNNOp`,
- offers new interface `Eval(targets)` to deduce the minimal block to `Run`,
- fits the compile-time/ runtime separation design.
- during the compilation, `SymbolTable` stores `VarDesc`s and `OpDesc`s and serialize to a `BlockDesc`
- when graph executes, a Block with `BlockDesc` passed in creates `Op` and `Var` then `Run`
## the difference between original RNNOp and Block
- As an operator is more intuitive than `RNNOp`,
- Offers a new interface `Eval(targets)` to deduce the minimal block to `Run`,
- Fits the compile-time/ runtime separation design paradigm.
- During the compilation, `SymbolTable` stores `VarDesc`s and `OpDesc`s and serialize to a `BlockDesc`
- When graph executes, a Block with `BlockDesc` is passed. It then creates `Op` and `Var` instances and then invokes `Run`.
---
# Milestone
- take Paddle/books as the main line, the requirement of the models motivates framework refactoring,
- model migration
- framework development gives **priority support** to model migration, for example,
- Take Paddle/books as the main line, the requirement of the models motivates framework refactoring,
- Model migration
- Framework development gives **priority support** to model migration, for example,
- the MNIST demo needs a Python interface,
- the RNN models require the framework to support `LoDTensor`.
- determine some timelines,
- heavily-relied Ops need to be migrated first,
- different models can be migrated parallelly.
- improve the framework at the same time
- accept imperfection, concentrated on solving the specific problem at the right price.
- Determine some timelines,
- Frequently used Ops need to be migrated first,
- Different models can be migrated in parallel.
- Improve the framework at the same time
- Accept imperfection, concentrate on solving the specific problem at the right price.
---
# Control the migration quality
- compare the performance of migrated models with old ones.
- follow google C style
- build the automatic workflow of generating Python/C++ documentations
- the documentation of layers and ops should be written inside the code
- take the documentation quality into account when doing PR
- preview the documentations, read and improve them from users' perspective
- Compare the performance of migrated models with old ones.
- Follow the google C++ style guide.
- Build the automatic workflow of generating Python/C++ documentations.
- The documentation of layers and ops should be written inside the code.
- Take the documentation quality into account when submitting pull requests.
- Preview the documentations, read and improve them from a user's perspective.
doc/design/register_grad_op.md 0 → 100644
浏览文件 @ 0dc4b298
# Design Doc: Gradient Operators Registration
## The Problem Posed
Currently, for each C++ operator class definition, a *gradient operator creator* function is registered, which takes as input a C++ operator instance and returns the corresponding gradient operator instance.
However, we noticed two problems with the current design:
1. As we decided to separate the *compilation* and the *execution* phases, we need to change the creator to take an `OpDesc` protobuf message in a `ProgramDesc` and inserts corresponding `OpDesc` messages into the `ProgramDesc` message.
1. For some operators, the gradient computation can be written in terms of existing operators. For example, the gradient of *minus* operator consists of two operators -- an *identity* operator followed by a *scale* operator. Hence the registration mechanism needs to support mapping from an operator to a set of operators for the gradient computation.
## The Current Implementation
Instances of the C++ class `OpInfo` are stored an associative map whose key is the operator type. The `grad_op_type` indicates the associated gradient operator type. An operator can create the gradient operator by invoking `OpInfo::creator_` of the gradient operator. The pseudo code is as follows
```cpp
struct OpInfo {
std::function<OperatorBase*(...)> creator_;
std::string grad_op_type_;
...
};
map<string, OpInfo> OpInfoMap;
OperatorBase* CreateGradientOperator(const OperatorBase& op) {
return OpInfoMap.at(op.Type()).creator_(...);
}
```
## Proposed Solution
The mapping relationship between an operator and its gradient operators is a function. The interface of this function is:
```cpp
// (OpDesc) --> vector<OpDesc>
std::function<std::vector<OpDescBind>(const OpDescBind&)>;
```
The function takes an `OpDescBind` of the forward operator and returns one or many gradient operator descriptions. `OpDescBind` is a C++ wrapper for the protobuf message `OpDesc` for rapid manipulation of `OpDesc`.
The `GradOpDescMaker` will be registered in `OpInfo` and will replace the `grad_op_type_` field. The `OpInfo` should look like
```cpp
struct OpInfo {
std::function<std::vector<std::unique_ptr<OpDescBind>>(const OpDescBind&)> grad_op_maker_;
...
};
```
The `grad_op_maker_ ` is a `nullptr` if the operator does not have any associated gradient operators.
We propose a base class called `GradOpDescMakerBase` to let operator developers generate `Gradient Operators` easily. The public interface of that class is
```cpp
class GradOpDescMakerBase {
public:
GradOpDescMakerBase(const OpDescBind& );
virtual std::vector<std::unique_ptr<OpDescBind>> operator()()const = 0;
};
```
We can convert `GradOpDescMakerBase` to `std::function<std::vector<std::unique_ptr<OpDescBind>>(const OpDescBind&)>` by
```cpp
using GradOpMaker = ...;
std::function<std::vector<OpDescBind>(const OpDescBind&)> func;
func = [] (const OpDescBind& fwd_op) {
GradOpMaker maker(fwd_op);
return maker();
};
```
We can write many helper functions since the `GradOpDescMakerBase` is a class now. The basic helper functions get the variables of `Input`, `Output`, `InputGradient` and `OutputGradient` in the forwarding operator.
We should change register macros at the same time. In the current solution, there is no difference between forwarding operators and backward operators. So `REGISTER_OP` just register one operator. If the `REGISTER_OPERATOR ` contains `OpProtoAndCheckerMaker` and `GradOpDescMaker`, we just list them in the same macro. It can be done by a macro contains `__VA_ARGS__`.
The user interface should be
```cpp
vector<OpDesc> MinusOpGradMaker(OpDesc) {...}
REGISTER_OPERATOR(minus, MinusOp, MinusOpProtoAndCheckerMaker, SumOpGradMaker);
// Developers can still manually implement gradient operator.
REGISTER_OPERATOR(minus_grad, MinusGradOp);
```
The interface of current `REGISTER_OP` macro could not be changed. In `REGISTER_OP`, it will invoke `REGISTER_OPERATOR` two times and generate GradOpDescMaker inside.
```cpp
REGISTER_OP(minus, MinusOp, MinusOpProtoAndCheckerMaker, minus_grad, MinusGradOp);
```
doc/design/regularization.md 0 → 100644
浏览文件 @ 0dc4b298
# Regularization in PaddlePaddle
## Introduction to Regularization
A central problem in machine learning is how to design an algorithm that will perform well not just on the training data, but also on new data. Many strategies are used by machine learning practitioners to reduce the test error, possibly at the expense of increased training error. These strategies are collectively known as **regularization**.
### Parameter Norm Penalties
Most common regularization approaches in deep learning are based on limiting the capacity of the models by adding a parameter norm penalty to the objective function `J`. This is given as follows:
<img src="./images/loss_equation.png" align="center"/><br/>
The parameter `alpha` is a hyperparameter that weights the relative contribution of the norm penalty term, `omega`, relative to the standard objective function `J`.
The most commonly used norm penalties are the L2 norm penalty and the L1 norm penalty. These are given as follows:
##### L2 Regularization:
<img src="./images/l2_regularization.png" align="center"/><br/>
##### L1 Regularization
<img src="./images/l1_regularization.png" align="center"/><br/>
A much more detailed mathematical background of reguilarization can be found [here](http://www.deeplearningbook.org/contents/regularization.html).
## How to do Regularization in PaddlePaddle
On surveying existing frameworks like Tensorflow, PyTorch, Caffe, etc, it can be seen that there are 2 common approaches of doing regularization:
1. Making regularization a part of the optimizer using an attribute like `weight_decay` that is used to control the scale of the L2 Penalty. This approach is used in PyTorch as follows:
```python
opt = torch.optim.SGD(params, lr=0.2, weight_decay=0.2)
```
At every optimization step, this code will add the gradient of the L2 Norm of the params to the gradient of the params with respect to the loss function. This can seen in the following code snippet:
```python
if weight_decay != 0:
d_p.add_(weight_decay, p.data)
```
This is a very restyrictive way of doing regularization and does not give the users enough flexibility.
**Advantages**:
- It is easy to implement for us.
- Faster execution of backward. However, it can be done manually by advanced users too.
**Disadvantages**:
- Not flexible for other regularizations such as L1/L0 regularization.
- Does not allow for different regularization coefficient for different parameters. For example, in most models, ony the weight matrices are regularized and the bias vectors are unregularized.
- Tightly coupled optimizer and regularization implementation.
2. Adding regularization ops to the graph through Python API. This approach is used by Tensorflow and Caffe. Using this approach, we manually add regularization ops to the graph and then add the regularization loss to the final loss function before sending them to the optimizer.
**Advantages**:
- Allows for greater flexibility to the users of Paddle. Using this approach, the users can put different regularization to different parameters and also choose parameters that are not a part of regularization.
- Makes it easy for the users to customize and extend the framework.
**Disadvantages**:
- Implementation requires comprehensive design and time.
## Proposal for Regularization in PaddlePaddle
### Low-Level implementation
In the new design, we propose to create new operations for regularization. For now, we can add 2 ops thgat correspond to the most frequently used regularizations:
- L2_regularization_op
- L1_regularization_op
These ops can be like any other ops with their own CPU/GPU implementations either using Eigen or separate Cpu and GPU kernels. As the initial implementation, we can implement their kernels using Eigen following the abstraction pattern implemented for [Activation Ops](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/operators/accuracy_op.h). This abstraction pattern can make it very easy to implement new regularization schemes. other than L1 and L2 norm penalties.
The idea of building ops for regularization is in sync with the refactored Paddle philosophy of using operators to represent any computation unit. The way these ops will be added to the computation graph, will be decided by the [layer functions](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md#layer-function) in Python API.
### Computation Graph
Below is an example of a really simple feed forward neural network.
<img src="./images/feed_forward.png" align="center"/><br/>
The Python API will modify this computation graph to add regularization operators. The modified computation graph will look as follows:
<img src="./images/feed_forward_regularized.png" align="center"/><br/>
   
### Python API implementation for Regularization
Using the low level ops, `L2_regularization_op` and `L1_regularization_op`, any user can add regularization to their computation graphs. However, this will require a lot of lines of code and we should design Python APIs that support regularization. An example of such an API can be seen in [Keras](https://keras.io/regularizers/). As per the PaddlePaddle [Python API design](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md), the layer functions are responsible for creating operators, operator parameters and variables. Since regularization is a property of parameters, it makes sense to create these in the layer functions.
#### Creation of Regularization ops
There are two possibilities for creating the regularization ops:
1. We create these ops immediately while building the computation graph.
2. We add these ops in a lazy manner, just before the backward, similar to the way the optimization ops are added.
The proposal is to add these ops in a lazy manner just before the backward pass.
#### Storage of Regularization attributes
Since we want to create the regularization ops in a lazy manner, the regularization attributes (type of regularization and weight of regularization penalty) can be stored as attributes of the [`Parameter`](https://github.com/PaddlePaddle/Paddle/blob/develop/python/paddle/v2/framework/framework.py#L421) class. This is because regularization is a property of the parameters and storing regularization properties with Parameters also allows for shared parameters.
#### High-level API
In PaddlePaddle Python API, users will primarily rely on [layer functions](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md#layer-function) to create neural network layers. Hence, we lso need to provide regularization functionality in layer functions. The design of these APIs can be postponed for later right now. A good reference for these APIs can be found in [Keras](https://keras.io/regularizers/) and also by looking at Tensorflow in [`tf.contrib.layers`](https://www.tensorflow.org/api_guides/python/contrib.layers).
doc/design/scope.md
浏览文件 @ 0dc4b298
......@@ -37,7 +37,7 @@ Scope is an association of a name to variable. All variables belong to `Scope`.
```cpp
class Scope {
public:
Variable* NewVar(const std::string& name);
Variable* Var(const std::string& name);
const Variable* FindVar(const std::string& name) const;
private:
......@@ -98,7 +98,7 @@ class Scope {
Variable* FindVar(const std::string& name) const;
// return if already contains same name variable.
Variable* NewVar(const std::string& name);
Variable* Var(const std::string& name);
private:
std::shared_ptr<Scope> parent_;
......@@ -107,7 +107,7 @@ class Scope {
```
## Only scope can create a variable
To ensure `only scope can create a variable`, we should mark `Variable`'s constructor as a private member function, and Scope is a friend class of Variable. And then only `NewVar` can construct `Variable`.
To ensure `only scope can create a variable`, we should mark `Variable`'s constructor as a private member function, and Scope is a friend class of Variable. And then only `Var` can construct `Variable`.
## When scope destroyed, all variables inside this scope should be destroyed together
......@@ -121,4 +121,4 @@ Also, as the parent scope is a `shared_ptr`, we can only `Create()` a scope shar
## Orthogonal interface
`FindVar` will return `nullptr` when `name` is not found. It can be used as `Contains` method. `NewVar` will return an `Error` when there is a name conflict locally. Combine `FindVar` and `NewVar`, we can implement `NewVar` easily.
`FindVar` will return `nullptr` when `name` is not found. It can be used as `Contains` method. `Var` will return an `Error` when there is a name conflict locally. Combine `FindVar` and `Var`, we can implement `Var` easily.
doc/design/selected_rows.md 0 → 100644
浏览文件 @ 0dc4b298
# Design Doc: Selected Rows
`SelectedRows` is a type of sparse tensor data type, which is designed to support `embedding` operators. The gradient of embedding table is a sparse tensor. Only a few rows are non-zero values in this tensor. It is straight-forward to represent a sparse tensor by the following sparse tensor data structure:
```cpp
class SelectedRows {
private:
vector<int> rows_;
Tensor value_;
int height_;
};
```
The field `height_` is the first dimension of `SelectedRows`. The `rows` are the indices of the non-zero rows of `SelectedRows`. The `value_` field is an N-dim tensor of shape `[rows.size() /* NUM_ROWS */, ...]`, which supplies values for each row. The dimension of `SelectedRows` satisfies `[height_] + value_.shape[1:]`.
Suppose that a SelectedRows-typed variable `x` has many rows, but only two of them have values -- row 73 is `[1, 2]` and row 84 is `[3, 4]`, the `SelectedRows` representation would be:
```
x = SelectedRow {
rows = [73, 84],
value = [[1, 2], [3,4]]
}
```
## SelectedRows in Protobuf
`SelectedRows` is a type of `Variable`. `VarDesc` in protobuf should describe the `SelectedRows` information. Only the tensor dimension of a `SelectedRows` will be described in compile-time because the `rows_` and `value_` are dependent on the training data.
So we use `TensorDesc` to unify `data_type` and `dims`. A LodTensorDesc contains a `TensorDesc` and `lod_level`. The description of `SelectedRows` is a Tensor description.
```proto
message TensorDesc {
required DataType data_type = 1;
repeated int64 dims = 2; // [UNK, 640, 480] is saved as [-1, 640, 480]
}
message LodTensorDesc {
required TensorDesc tensor = 1;
optional int lod_level = 2;
}
message VarDesc {
required string name = 1;
enum VarType {
LOD_TENSOR = 0;
SELECTED_ROWS = 1;
}
required VarType type = 2;
optional LodTensorDesc lod_desc = 3;
optional TensorDesc selected_rows_desc = 4;
optional bool persistable = 5 [ default = false ];
}
```
## InferShape for Selected Rows
Just like `LoD` information, `InferShape` method will infer the output tensor type as well. The operator should decide whether its output is a `SelectedRows` or `Dense` tensor.
For example, the gradient operator of `TableLookup` will always generate `SelectedRows`. Its `InferShape` method should be like following
```cpp
void TableLookupGrad::InferShape(context) {
...
context.SetDataType("Embedding.Grad", kSelectedRows);
}
```
## Sparse Operators
There are several operators that need to be written to support `SelectedRows`. These are:
1. Operators which generate `SelectedRows` gradient. e.g. Gradient of `TableLookupOp`.
2. Optimize operators which support `SelectedRows` gradient. e.g. `SGD` or `AdaGrad` for `SelectedRows`. However, there should be only one `SGD` operator. `OpWithKernel::Run` should select a suitable kernel for both `dense` tensor or `SelectedRows`.
doc/design/tensor_array.md
浏览文件 @ 0dc4b298
# Design for TensorArray
This design doc presents the necessity of a new C++ class `TensorArray`.
In addition to the very simple C++ implementation
```c++
class TensorArray {
public:
explicit TensorArray(const LoDTensor&);
explicit TensorArray(size_t size);
private:
vector<LoDTensor> values_;
};
```
We also need to expose it to PaddlePaddle's Python API,
because users would want to use it with our very flexible operators `WhileLoop`.
An example for a RNN based on dynamic operators is
```python
input = pd.data(...)
num_steps = Var(12)
TensorArray states(size=num_steps)
TensorArray step_inputs(unstack_from=input)
TensorArray step_outputs(size=num_steps)
W = Tensor(...)
U = Tensor(...)
default_state = some_op()
step = Var(1)
wloop = paddle.create_whileloop(loop_vars=[step])
with wloop.frame():
wloop.break_if(pd.equal(step, num_steps)
pre_state = states.read(step-1, default_state)
step_input = step_inputs.read(step)
state = pd.sigmoid(pd.matmul(U, pre_state) + pd.matmul(W, step_input))
states.write(step, state)
step_outputs.write(step, state) # output state
step.update(state+1)
output = step_outputs.stack()
```
## Background
Steps are one of the core concepts of RNN. In each time step of RNN, there should be several input segments, states, and output segments; all these components act like arrays, for example, call `states[step_id]` will get the state in `step_id`th time step.
An RNN can be implemented with the following pseudocode
```c++
Array states;
Array input_segments;
Array output_segments;
Parameter W, U;
step = 1
seq_len = 12
while_loop {
if (step == seq_len) break;
states[step] = sigmoid(W * states[step-1] + U * input_segments[step]);
output_segments[step] = states[step] // take state as output
step++;
}
```
According to the [RNN roadmap](https://github.com/PaddlePaddle/Paddle/issues/4561), there are several different RNNs that PaddlePaddle will eventually support.
Currently, the basic RNN implementation supported by PaddlePaddle is the `recurrent_op` which takes tensors as input and splits them into `input_segments`.
Since a tensor cannot store variable-length sequences directly, PaddlePaddle implements the tensor with level of details (`LoDTensor` for short).
Segmenting the `LoDTensor` is much more complicated than splitting a tensor, that makes it necessary to refactor the `recurrent_op` with `LoDTensor` segmenting support.
As the next step in RNN support, `dynamic_recurrent_op` should be introduced to handle inputs with variable-length sequences.
The implementation is similar to `recurrent_op`.
The key difference is the way **the original input `LoDTensors` and outupts are split to get the `input_segments` and the `output_segments`.**
Though it can't be built over `recurrent_op` or `dynamic_recurrent_op` directly,
the logic behind splitting a tensor or a LoD tensor into `input_segments` remains the same.
## Why `TensorArray`
The logic behind splitting the inputs to segments, states and outputs is similar and can be shared in a seperate module.
The array of `states`, `input_segments` and `output_segments` would be exposed to users when writing a dynamic RNN model similar to the above pseudo codes.
So there should be an array-like container, which can store the segments of a tensor or LoD tensor.
**This container can store an array of tensors and provides several methods to split a tensor or a LoD tensor** .
This is where the notion of `TensorArray` comes from.
## Introduce TensorArray to uniform all the three RNNs
TensorArray as a new concept is borrowed from TensorFlow,
it is meant to be used with dynamic iteration primitives such as `while_loop` and `map_fn`.
This concept can be used to support our new design of dynamic operations, and help to refactor some existing variant-sentence-related layers,
such as `RecurrentGradientMachine`.
such as `recurrent_op`, `RecurrentGradientMachine`.
In [our design for dynamic RNN](https://github.com/PaddlePaddle/Paddle/pull/4401),
`TensorArray` is used to segment inputs and store states in all time steps.
By providing some methods similar to a C++ array,
the definition of some state-based dynamic models such as RNN could be more natural and highly flexible.
## Dynamic-Related Methods
Some basic methods should be proposed as follows:
### stack()
Pack the values in a `TensorArray` into a tensor with rank one higher than each tensor in `values`.
### unstack(axis=0)
Unpacks the given dimension of a rank-`R` tensor into rank-`(R-1)` tensors.
### concat()
Return the values in the `TensorArray` as a concatenated Tensor.
### write(index, value, data_shared=true)
Write value into index of the TensorArray.
### read(index)
Read the value at location `index` in the `TensorArray`.
### size()
Return the number of values.
the definition of some state-based dynamic models such as RNN can be more natural and highly flexible.
## Dynamic-operations on TensorArray
`TensorArray` will be used directly when defining dynamic models, so some operators listed below should be implemented
```python
# several helper operators for TensorArray
def tensor_array_stack(ta, tensor):
'''
get a tensor array `ta`, return a packed `tensor`.
'''
pass
def tensor_array_unstack(tensor, ta):
'''
get a `tensor`, unstack it and get a tensor array `ta`.
'''
pass
def tensor_array_write(ta, index, tensor, data_shared):
'''
get a `tensor` and a scalar tensor `index`, write `tensor` into index-th
value of the tensor array `ta`.
`data_shared` is an attribute that specifies whether to copy or reference the tensors.
'''
pass
def tensor_array_read(ta, index, tensor):
'''
get a tensor array `ta`, a scalar tensor `index`, read the index-th value of
`ta` and return as the `tensor`.
'''
pass
def tensor_array_size(ta, tensor):
'''
get a tensor array `ta`, return the size of `ta` and return as the scalar `tensor`.
'''
pass
```
It is trivial for users to use so many low-level operators, so some helper methods should be proposed in python wrapper to make `TensorArray` easier to use,
for example
```python
class TensorArray:
def __init__(self, name):
self.name = name
self.desc = TensorArrayDesc()
def stack(self, name=None):
'''
Pack the values in a `TensorArray` into a tensor with rank one higher
than each tensor in `values`.
`stack` can be used to split tensor into time steps for RNN or whileloop.
@name: str
the name of the variable to output.
'''
tensor = Var(name)
tensor_array_stack(self.name, tensor)
return tensor
def unstack(self, input):
'''
Unpacks the given dimension of a rank-`R` tensor into rank-`(R-1)` tensors.
`unstack` can be used to concatenate all the time steps for RNN or whileloop.
@input: str
the name of input tensor
'''
tensor_array_unstack(tensor, self.name)
def write(self, index, value, data_shared=True):
'''
Write value into index of the TensorArray.
If `data_shared` is set to True, than the index-th value in TensorArray will
be shared with the tensor passed in.
@index: str
name of a scalar tensor
@value: str
name of a tensor
@data_shared: bool
'''
tensor_array_write(self.name, index, value, data_shared)
def read(self, index, output):
'''
Read the value at location `index` in the `TensorArray`.
@index: str
name of a scalar tensor
@output:
name of a output variable
'''
tensor_array_read(self.name, index, output)
def size(self, output):
'''
Return the number of values.
@output: str
name of a scalar tensor
'''
tensor_array_size(self.name, output)
```
## LoDTensor-related Supports
The `RecurrentGradientMachine` in Paddle serves as a flexible RNN layer; it takes variant length sequences as input,
because each step of RNN could only take a tensor-represented batch of data as input,
The `RecurrentGradientMachine` in Paddle serves as a flexible RNN layer; it takes varience-length sequences as input, and output sequences too.
Since each step of RNN can only take a tensor-represented batch of data as input,
some preprocess should be taken on the inputs such as sorting the sentences by their length in descending order and cut each word and pack to new batches.
Such cut-like operations can be embedded into `TensorArray` as general methods called `unpack` and `pack`.
Such cut-like operations can be embedded into `TensorArray` as general methods called `unpack` and `pack`,
these two operations are similar to `stack` and `unstack` except that they operate on variable-length sequences formated as a LoD tensor rather than a tensor.
Some definitions are like
```python
def unpack(level):
'''
Split LodTensor in some `level` and generate batches, if set `sort_by_length`,
will sort by length.
With these two methods, a variant-sentence-RNN can be implemented like
Returns:
- a new `TensorArray`, whose values are LodTensors and represents batches
of data.
- an int32 Tensor, which stores the map from the new batch's indices to
original LoDTensor
'''
pass
def pack(level, indices_map):
'''
Recover the original LoD-arranged LoDTensor with the values in a `TensorArray`
and `level` and `indices_map`.
'''
pass
```
With these two methods, a varience-length sentence supported RNN can be implemented like
```c++
// input is the varient-length data
......@@ -58,16 +269,3 @@ LoDTensor rnn_output = ta.pack(ta, indice_map);
```
the code above shows that by embedding the LoDTensor-related preprocess operations into `TensorArray`,
the implementation of a RNN that supports varient-length sentences is far more concise than `RecurrentGradientMachine` because the latter mixes all the codes together, hard to read and extend.
some details are as follows.
### unpack(level, sort_by_length)
Split LodTensor in some `level` and generate batches, if set `sort_by_length`, will sort by length.
Returns:
- a new `TensorArray`, whose values are LodTensors and represents batches of data.
- an int32 Tensor, which stores the map from the new batch's indices to original LoDTensor
### pack(level, indices_map)
Recover the original LoD-arranged LoDTensor with the values in a `TensorArray` and `level` and `indices_map`.
doc/design/test.dot 0 → 100644
浏览文件 @ 0dc4b298
digraph Test {
z -> generator -> G_img;
G_img -> discriminator -> D_f -> d_loss_f;
label0 -> d_loss_f -> d_loss;
img -> discriminator -> D_t -> d_loss_t;
label1 -> d_loss_t -> d_loss;
d_loss -> d_loss_t[color=red, style=dashed];
d_loss -> d_loss_f[color=red, style=dashed];
d_loss_t -> D_t[color=red, style=dashed];
d_loss_f -> D_f[color=red, style=dashed];
D_t -> discriminator[color=red, style=dashed];
D_f -> discriminator[color=red, style=dashed];
D_f -> g_loss;
label2 -> g_loss;
g_loss -> D_f[color=green, style=dashed];
D_f -> discriminator[color=green, style=dashed];
discriminator -> G_img[color=green, style=dashed];
G_img -> generator[color=green, style=dashed];
discriminator [color=red, shape=box];
generator [color=green, shape=box];
z [shape=diamond];
img [shape=diamond];
label0 [shape=diamond];
label1 [shape=diamond];
label2 [shape=diamond];
d_loss [color=red];
g_loss [color=green];
}
doc/design/var_desc.md
浏览文件 @ 0dc4b298
......@@ -16,16 +16,23 @@ The computation graph is constructed by Data Node and Operation Node. The concep
## Definition of VarDesc
A VarDesc should have a name and value, in PaddlePaddle, the value will always be a tensor. Since we use LoDTensor most of the time. We add a LoDTesnorDesc to represent it.
A VarDesc should have a name, and value. The are two kinds of variable type in compile time, they are `LoDTensor` and `SelectedRows`.
```proto
message VarDesc {
required string name = 1;
optional LoDTensorDesc lod_tensor = 2;
enum VarType {
LOD_TENSOR = 0;
SELECTED_ROWS = 1;
}
required VarType type = 2;
optional LoDTensorDesc lod_desc = 3;
optional TensorDesc selected_rows_desc = 4;
optional bool persistable = 5 [ default = false ];
}
```
## Definition of LodTensorDesc
## Definition of TensorDesc
```proto
enum DataType {
......@@ -38,87 +45,25 @@ enum DataType {
FP64 = 6;
}
message LoDTensorDesc {
message TensorDesc {
required DataType data_type = 1;
repeated int32 dims = 2; // [UNK, 640, 480] is saved as [-1, 640, 480]
optional int32 lod_level = 3 [default=0];
repeated int64 dims = 2; // [UNK, 640, 480] is saved as [-1, 640, 480]
}
```
## Definition of Variable in Python
In Python API, layer will take Variable as Input, and return Variable as Output. There should be a class `Variable` in python to help create and manage Variable.
```python
image = Variable(dims=[-1, 640, 480])
# fc1 and fc2 are both Variable
fc1 = layer.fc(input=image, output_size=10)
fc2 = layer.fc(input=fc1, output_size=20)
```
### what should class `Variable` Have
1. `name`.a name of string type is used to mark the value of the Variable.
1. `initializer`. Since our Tensor does not have value. we will always use some Operator to fullfill it when run. So we should have a initialize method to help add the init operator.
1. `operator`. Variable should record which operator produce itself. The reaon is:
- we use pd.eval(targets=[var1, var2]) to run the related ops to get the value of var1 and var2. var.op is used to trace the dependency of the current variable.
In PaddlePaddle, we use Block to describe Computation Graph, so in the code we will use Block but not Graph.
```python
import VarDesc
import LoDTensorDesc
import framework
def AddInitialOperator(variable, initializer):
# add an initialize Operator to block to init this Variable
class Variable(object):
def __init__(self, name, dims, type, initializer):
self._block = get_default_block()
self._name = name
self.op = None
tensor_desc = LoDTensorDesc(data_type=type, dims=dims)
_var_desc = VarDesc(name=name, lod_tensor=tensor_desc)
self._var = framework.CreateVar(_var_desc)
self._block.add_var(self)
A TensorDesc describes `SelectedRows` and `LoDTensor`. For details of `SelectedRows`, please reference [`SelectedRows`](./selected_rows.md).
# add initial op according to initializer
if initializer is not None:
AddInitialOperator(self, initializer)
def dims(self):
return self._var.dims()
def data_type(self):
return self._var.data_type()
## Definition of LodTensorDesc
def to_proto(self):
pass
```proto
message LoDTensorDesc {
required TensorDesc tensor = 1;
optional int lod_level = 2;
}
```
Then we can use this Variable to create a fc layer in Python.
A LoDTensorDesc contains a tensor and a lod_level.
```python
import paddle as pd
def flatten_size(X, num_flatten_dims):
prod = 1 # of last num_flatten_dims
for i in xrange(num_flatten_dims):
prod = prod * X.dims[-i-1]
return prod
def layer.fc(X, output_size, num_flatten_dims):
W = Variable(pd.random_uniform(), type=FP32, dims=[flatten_size(X, num_flatten_dims), output_size])
b = Variable(pd.random_uniform(), type=FP32, dims=[output_size])
out = Variable(type=FP32)
y = operator.fc(X, W, b, output=out) # fc will put fc op input into out
pd.InferShape(y)
return out
x = Variable(dims=[-1, 640, 480])
y = layer.fc(x, output_size=100)
z = layer.fc(y, output_size=200)
## Definition of Variable in Python
paddle.eval(targets=[z], ...)
print(z)
```
For Variable in Python, please reference [`Python API`](./python_api.md).
doc/faq/local/index_cn.rst
浏览文件 @ 0dc4b298
......@@ -174,7 +174,7 @@ decoder_inputs = paddle.layer.fc(
1. 两者都是对梯度的截断，但截断时机不同，前者在 :code:`optimzier` 更新网络参数时应用；后者在激活函数反向计算时被调用；
2. 截断对象不同：前者截断可学习参数的梯度，后者截断回传给前层的梯度;
除此之外，还可以通过减小学习律或者对数据进行归一化处理来解决这类问题。
除此之外，还可以通过减小学习率或者对数据进行归一化处理来解决这类问题。
5. 如何调用 infer 接口输出多个layer的预测结果
-----------------------------------------------
......
doc/howto/cross_compiling/cross_compiling_for_android_cn.md
浏览文件 @ 0dc4b298
# 构建Android平台上的PaddlePaddle库
用户可通过交叉编译的方式，在用户熟悉的开发平台（Linux，Mac OS X和Windows）上编译Android平台上适用的PaddlePaddle库。
用户可通过如下两种方式，交叉编译Android平台上适用的PaddlePaddle库：
- 基于Docker容器的编译方式
- 基于Linux交叉编译环境的编译方式
## 基于Docker容器的编译方式
Docker能在所有主要操作系统（包括Linux，Mac OS X和Windows）上运行，因此，使用基于Docker容器的编译方式，用户可在自己熟悉的开发平台上编译Android平台上适用的PaddlePaddle库。
### 构建PaddlePaddle的Android开发镜像
我们把PaddlePaddle的交叉编译环境打包成一个镜像，称为开发镜像，里面涵盖了交叉编译Android版PaddlePaddle库需要的所有编译工具。
```bash
$ git clone https://github.com/PaddlePaddle/Paddle.git
$ cd Paddle
$ docker build -t username/paddle-android:dev . -f Dockerfile.android
```
### 编译PaddlePaddle C-API库
构建好开发镜像后，即可使用开发镜像来编译Android版PaddlePaddle C-API库。
Android的Docker开发镜像向用户提供两个可配置的参数：
| Argument | Optional Values | Default |
|-----------------|-------------------------|---------|
|`ANDROID_ABI` |`armeabi-v7a, arm64-v8a` | `armeabi-v7a` |
|`ANDROID_API` |`>= 21` | `21` |
- 编译`armeabi-v7a`，`Android API 21`的PaddlePaddle库
```bash
$ docker run -it --rm -v $PWD:/paddle -e "ANDROID_ABI=armeabi-v7a" -e "ANDROID_API=21" username/paddle-android:dev
```
- 编译`arm64-v8a`，`Android API 21`的PaddlePaddle库
```bash
$ docker run -it --rm -v $PWD:/paddle -e "ANDROID_ABI=arm64-v8a" -e "ANDROID_API=21" username/paddle-android:dev
```
执行上述`docker run`命令时，容器默认执行[paddle/scripts/docker/build_android.sh](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/scripts/docker/build_android.sh)脚本。该脚本中记录了交叉编译Android版PaddlePaddle库常用的CMake配置，并且会根据`ANDROID_ABI`和`ANDROID_API`自动构建独立工具链、进行编译和安装。由于arm64架构要求Android API不小于21。因此当`ANDROID_ABI=arm64-v8a`，`ANDROID_API<21`时，Docker容器中将默认使用`Android API 21`的编译工具链。用户可以参考下文**配置交叉编译参数**章节，根据个人的需求修改定制Docker容器所执行的脚本。编译安装结束之后，PaddlePaddle的C-API库将被安装到`$PWD/install_android`目录，所依赖的第三方库同时也被安装到`$PWD/install_android/third_party`目录。
## 基于Linux交叉编译环境的编译方式
本文档将以Linux x86-64平台为例，介绍交叉编译Android平台上适用的PaddlePaddle库的方法和步骤。
## 准备交叉编译环境
### 准备交叉编译环境
从源码交叉编译PaddlePaddle，用户需要提前准备好交叉编译环境。Android平台上使用的C/C++交叉编译工具链为[Android NDK](https://developer.android.com/ndk/downloads/index.html?hl=zh-cn)，用户可自行前往下载预编译好的版本，也可通过以下命令获取：
......@@ -13,18 +50,27 @@ unzip -q android-ndk-r14b-linux-x86_64.zip
```
Android NDK中包含了所有Android API级别、所有架构（arm/arm64/x86/mips）需要用到的编译工具和系统库。用户可根据自己的编译目标架构、所需支持的最低Android API级别，构建[独立工具链](https://developer.android.google.cn/ndk/guides/standalone_toolchain.html?hl=zh-cn)。
比如：
- 构建`armeabi-v7a`、 `Android API 21`的独立工具链：
```bash
your/path/to/android-ndk-r14b-linux-x86_64/build/tools/make-standalone-toolchain.sh \
--arch=arm --platform=android-21 --install-dir=your/path/to/my_standalone_toolchain
--arch=arm --platform=android-21 --install-dir=your/path/to/arm_standalone_toolchain
```
此命令将在your/path/to/my_standalone_toolchain目录生成一套编译工具链，面向架构为32位ARM架构，支持的最小的Android API级别为21，使用的编译器为arm-linux-androideabi-gcc (GCC) 4.9。
此命令将在`your/path/to/arm_standalone_toolchain`目录生成一套独立编译工具链，面向架构为32位ARM架构，支持的最小的Android API级别为21，支持编译器`arm-linux-androideabi-gcc (GCC) 4.9`和`clang 3.8`。
注意：**PaddlePaddle要求使用的编译工具链所支持的Andoid API级别不小于21**。
- 构建`arm64-v8a`、 `Android API 21`的独立工具链：
```bash
your/path/to/android-ndk-r14b-linux-x86_64/build/tools/make-standalone-toolchain.sh \
--arch=arm64 --platform=android-21 --install-dir=your/path/to/arm64_standalone_toolchain
```
## 配置交叉编译参数
此命令将在`your/path/to/arm64_standalone_toolchain`目录生成一套独立编译工具链，面向架构为64位ARM64架构，支持的最小Android API级别为21，支持编译器`arm-linux-androideabi-gcc (GCC) 4.9`和`clang 3.8`。
注意：**PaddlePaddle要求使用的编译工具链所支持的Android API级别不小于21**。
### 配置交叉编译参数
CMake系统对交叉编译提供了支持[cmake-toolchains](https://cmake.org/cmake/help/v3.0/manual/cmake-toolchains.7.html#cross-compiling)。为了简化cmake配置，PaddlePaddle为交叉编译提供了工具链配置文档[cmake/cross_compiling/android.cmake](https://github.com/PaddlePaddle/Paddle/blob/develop/cmake/cross_compiling/android.cmake)，以提供一些默认的编译器和编译参数相关配置。注意，从CMake 3.7版本开始，CMake官方对Android平台的交叉编译提供了通用的支持。PaddlePaddle若检测到用户使用的CMake版本不低于3.7时，将会将用户传进来的配置参数传递CMake系统，交由CMake系统本身来处理。有关参数配置的详细说明见[cmake-toolchains](https://cmake.org/cmake/help/v3.7/manual/cmake-toolchains.7.html#cross-compiling)。
......@@ -36,32 +82,57 @@ CMake系统对交叉编译提供了支持[cmake-toolchains](https://cmake.org/cm
Android平台可选配置参数：
- `ANDROID_STANDALONE_TOOLCHAIN`，独立工具链所在的绝对路径，或者相对于构建目录的相对路径。PaddlePaddle的CMake系统将根据该值自动推导和设置需要使用的交叉编译器、sysroot、以及Android API级别；否则，用户需要在cmake时手动设置这些值。无默认值。
- `ANDROID_ABI`，目标架构ABI。目前只支持`armeabi-v7a`，默认值为`armeabi-v7a`。
- `ANDROID_TOOLCHAIN`，目标工具链。可设置`gcc/clang`，默认值为`clang`。
- CMake 3.7以上，将会始终使用`clang`工具链；CMake 3.7以下，可设置`ANDROID_TOOLCHAIN=gcc`以使用`gcc`工具链。
- Android官方提供的`clang`编译器要求系统支持`GLIBC 2.15`以上。
- `ANDROID_ABI`，目标架构ABI。目前支持`armeabi-v7a`和`arm64-v8a`，默认值为`armeabi-v7a`。
- `ANDROID_NATIVE_API_LEVEL`，工具链的Android API级别。若没有显式设置，PaddlePaddle将根据`ANDROID_STANDALONE_TOOLCHAIN`的值自动推导得到。
- `ANROID_ARM_MODE`，是否使用ARM模式。可设置`ON/OFF`，默认值为`ON`。
- `ANDROID_ARM_NEON`，是否使用NEON指令。目前必须设置成`ON`，默认值为`ON`。
- `ANROID_ARM_MODE`，是否使用ARM模式。
- `ANDROID_ABI=armeabi-v7a`时，可设置`ON/OFF`，默认值为`ON`；
- `ANDROID_ABI=arm64-v8a`时，不需要设置。
- `ANDROID_ARM_NEON`，是否使用NEON指令。
- `ANDROID_ABI=armeabi-v7a`时，可设置`ON/OFF`，默认值为`ON`；
- `ANDROID_ABI=arm64-v8a`时，不需要设置。
其他配置参数：
- `USE_EIGEN_FOR_BLAS`，是否使用Eigen库进行矩阵计算。可设置`ON/OFF`，默认值为`OFF`。
- `HOST_C/CXX_COMPILER`，宿主机的C/C++编译器。在编译宿主机版protoc可执行文件和目标机版OpenBLAS库时需要用到。默认设置成环境变量`CC`的值；若环境变量`CC`没有设置，则设置成`cc`编译器。
一种常用的cmake配置如下：
常用的cmake配置如下：
```bash
cmake -DCMAKE_SYSTEM_NAME=Android \
-DANDROID_STANDALONE_TOOLCHAIN=your/path/to/my_standalone_toolchain \
-DANDROID_STANDALONE_TOOLCHAIN=your/path/to/arm_standalone_toolchain \
-DANDROID_ABI=armeabi-v7a \
-DANDROID_ARM_NEON=ON \
-DANDROID_ARM_MODE=ON \
-DUSE_EIGEN_FOR_BLAS=ON \
-DCMAKE_INSTALL_PREFIX=your/path/to/install \
-DWITH_C_API=ON \
-DWITH_SWIG_PY=OFF \
..
```
```
cmake -DCMAKE_SYSTEM_NAME=Android \
-DANDROID_STANDALONE_TOOLCHAIN=your/path/to/arm64_standalone_toolchain \
-DANDROID_ABI=arm64-v8a \
-DUSE_EIGEN_FOR_BLAS=OFF \
-DCMAKE_INSTALL_PREFIX=your/path/to/install \
-DWITH_C_API=ON \
-DWITH_SWIG_PY=OFF \
..
```
用户还可根据自己的需求设置其他编译参数。比如希望最小化生成的库的大小，可以设置`CMAKE_BUILD_TYPE`为`MinSizeRel`；若希望最快的执行速度，则可设置`CMAKE_BUILD_TYPE`为`Release`。亦可以通过手动设置`CMAKE_C/CXX_FLAGS_MINSIZEREL/RELEASE`来影响PaddlePaddle的编译过程。
## 编译和安装
**性能TIPS**，为了达到最快的计算速度，在CMake参数配置上，有以下建议：
- 设置`CMAKE_BUILD_TYPE`为`Release`
- 使用`clang`编译工具链
- `armeabi-v7a`时，设置`USE_EIGEN_BLAS=ON`，使用Eigen进行矩阵计算；`arm64-v8a`时，设置`USE_EIGEN_FOR_BLAS=OFF`，使用OpenBLAS进行矩阵计算
### 编译和安装
CMake配置完成后，执行以下命令，PaddlePaddle将自动下载和编译所有第三方依赖库、编译和安装PaddlePaddle预测库。
......@@ -72,4 +143,4 @@ make install
注意：如果你曾经在源码目录下编译过其他平台的PaddlePaddle库，请先使用`rm -rf`命令删除`third_party`目录和`build`目录，以确保所有的第三方依赖库和PaddlePaddle代码都是针对新的CMake配置重新编译的。
执行完安装命令后，`your/path/to/install`目录中会包含`include`和`lib`目录，其中`include`中包含C-API的头文件，`lib`中包含一个Android版本的库。自此，PaddlePaddle的已经安装完成，用户可将`your/path/to/install`目录下的生成文件用于深度学习相关Android App中，调用方法见C-API文档。
执行完安装命令后，`your/path/to/install`目录中会包含`include`、`lib`和`third_party`目录，其中`include`中包含C-API的头文件，`lib`中包含若干个不同Android ABI的PaddlePaddle库，`third_party`中包含所依赖的所有第三方库。自此，PaddlePaddle的已经安装完成，用户可将`your/path/to/install`目录下的生成文件用于深度学习相关Android App中，调用方法见C-API文档。
doc/howto/deep_model/rnn/rnn_config_cn.rst
浏览文件 @ 0dc4b298
......@@ -21,7 +21,7 @@ wmt14数据的提供文件在 `python/paddle/v2/dataset/wmt14.py <https://github
循环神经网络在每个时间步骤顺序地处理序列。下面列出了 LSTM 的架构的示例。
.. image:: ../../../tutorials/sentiment_analysis/bi_lstm.jpg
.. image:: src/bi_lstm.jpg
:align: center
一般来说，循环网络从 :math:`t=1` 到 :math:`t=T` 或者反向地从 :math:`t=T` 到 :math:`t=1` 执行以下操作。
......@@ -96,7 +96,7 @@ Sequence to Sequence Model with Attention
我们将使用 sequence to sequence model with attention
作为例子演示如何配置复杂的循环神经网络模型。该模型的说明如下图所示。
.. image:: ../../../tutorials/text_generation/encoder-decoder-attention-model.png
.. image:: src/encoder-decoder-attention-model.png
:align: center
在这个模型中，源序列 :math:`S = \{s_1, \dots, s_T\}`
......
doc/howto/deep_model/rnn/rnn_config_en.rst
浏览文件 @ 0dc4b298
......@@ -19,7 +19,7 @@ Simple Gated Recurrent Neural Network
Recurrent neural network process a sequence at each time step sequentially. An example of the architecture of LSTM is listed below.
.. image:: ../../../tutorials/sentiment_analysis/src/bi_lstm.jpg
.. image:: src/bi_lstm.jpg
:align: center
Generally speaking, a recurrent network perform the following operations from :math:`t=1` to :math:`t=T`, or reversely from :math:`t=T` to :math:`t=1`.
......@@ -78,7 +78,7 @@ Sequence to Sequence Model with Attention
-----------------------------------------
We will use the sequence to sequence model with attention as an example to demonstrate how you can configure complex recurrent neural network models. An illustration of the sequence to sequence model with attention is shown in the following figure.
.. image:: ../../../tutorials/text_generation/encoder-decoder-attention-model.png
.. image:: src/encoder-decoder-attention-model.png
:align: center
In this model, the source sequence :math:`S = \{s_1, \dots, s_T\}` is encoded with a bidirectional gated recurrent neural networks. The hidden states of the bidirectional gated recurrent neural network :math:`H_S = \{H_1, \dots, H_T\}` is called *encoder vector* The decoder is a gated recurrent neural network. When decoding each token :math:`y_t`, the gated recurrent neural network generates a set of weights :math:`W_S^t = \{W_1^t, \dots, W_T^t\}`, which are used to compute a weighted sum of the encoder vector. The weighted sum of the encoder vector is utilized to condition the generation of the token :math:`y_t`.
......
doc/howto/dev/new_op_cn.md
浏览文件 @ 0dc4b298
......@@ -206,7 +206,7 @@ MulOp(const std::string &type, const framework::VariableNameMap &inputs,
- `REGISTER_OP` ： 注册`ops::MulOp`类，类型名为`mul`，该类的`ProtoMaker`为`ops::MulOpMaker`，注册`ops::MulOpGrad`，类型名为`mul_grad`。
- `REGISTER_OP_WITHOUT_GRADIENT` ： 用于注册没有反向的Op。
- `REGISTER_OP_CPU_KERNEL` ：注册`ops::MulKernel`类，并特化模板参数为`paddle::platform::CPUPlace`和`float`类型，同理，注册`ops::MulKernel`类。
- `REGISTER_OP_CPU_KERNEL` ：注册`ops::MulKernel`类，并特化模板参数为`paddle::platform::CPUPlace`和`float`类型，同理，注册`ops::MulGradKernel`类。
- 在 `.cu`文件中注册GPU Kernel。
......@@ -285,41 +285,27 @@ class TestMulGradOp(GradientChecker):
'Y': np.random.random((84, 100)).astype("float32")
}
def test_cpu_gpu_compare(self):
self.compare_grad(self.op, self.inputs)
def test_normal(self):
def test_check_grad_normal(self):
# mul op will enlarge the relative error
self.check_grad(
self.op, self.inputs, ["X", "Y"], "Out", max_relative_error=0.5)
self.check_grad(['X', 'Y'], 'Out', max_relative_error=0.5)
def test_ignore_x(self):
def test_check_grad_ingore_x(self):
self.check_grad(
self.op,
self.inputs, ["Y"],
"Out",
max_relative_error=0.5,
no_grad_set={"X"})
['Y'], 'Out', max_relative_error=0.5, no_grad_set=set("X"))
def test_ignore_y(self):
def test_check_grad_ingore_y(self):
self.check_grad(
self.op,
self.inputs, ["X"],
"Out",
max_relative_error=0.5,
no_grad_set={"Y"})
['X'], 'Out', max_relative_error=0.5, no_grad_set=set('Y'))
```
下面解释代码中一些关键的地方:
- 调用`create_op("mul")`创建反向Op对应的前向Op。
- 调用`compare_grad`函数对比CPU、GPU计算结果。
- `test_normal`中调用`check_grad`使用数值法检测梯度正确性和稳定性。
- 第一个参数`self.op` : 前向Op。
- 第二个参数`self.inputs` : 输入词典，词典的Key和`ProtoMaker`定义保持一致。
- 第三个参数`["X", "Y"]` : 指定对输入变量`X`、`Y`做梯度检测。
- 第四个参数`"Out"` : 指定前向网络最终的输出目标变量`Out`
- `test_ignore_x`和`test_ignore_y`分支用来测试只需要计算一个输入梯度的情况。
- `test_check_grad_normal`中调用`check_grad`使用数值法检测梯度正确性和稳定性。
- 第一个参数`["X", "Y"]` : 指定对输入变量`X`、`Y`做梯度检测。
- 第二个参数`"Out"` : 指定前向网络最终的输出目标变量`Out`。
- 第三个参数`max_relative_error`：指定检测梯度时能容忍的最大错误值。
- `test_check_grad_ingore_x`和`test_check_grad_ingore_y`分支用来测试只需要计算一个输入梯度的情况。
### 编译和执行单元测试
......
doc/howto/dev/new_op_en.md
浏览文件 @ 0dc4b298
......@@ -205,7 +205,7 @@ The definition of its corresponding backward operator, if applicable, is similar
- `REGISTER_OP` registers the `ops::MulOp` class, type named `mul`, its type `ProtoMaker` is `ops::MulOpMaker`, registering `ops::MulOpGrad` as `mul_grad`.
- `REGISTER_OP_WITHOUT_GRADIENT` registers an operator without gradient.
- `REGISTER_OP_CPU_KERNEL` registers `ops::MulKernel` class and specialized template types `paddle::platform::CPUPlace` and `float`, which also registers `ops::MulKernel`.
- `REGISTER_OP_CPU_KERNEL` registers `ops::MulKernel` class and specialized template types `paddle::platform::CPUPlace` and `float`, which also registers `ops::MulGradKernel`.
- Registering GPU Kernel in `.cu` files
......@@ -293,41 +293,27 @@ class TestMulGradOp(GradientChecker):
'Y': np.random.random((84, 100)).astype("float32")
}
def test_cpu_gpu_compare(self):
self.compare_grad(self.op, self.inputs)
def test_normal(self):
def test_check_grad_normal(self):
# mul op will enlarge the relative error
self.check_grad(
self.op, self.inputs, ["X", "Y"], "Out", max_relative_error=0.5)
self.check_grad(['X', 'Y'], 'Out', max_relative_error=0.5)
def test_ignore_x(self):
def test_check_grad_ingore_x(self):
self.check_grad(
self.op,
self.inputs, ["Y"],
"Out",
max_relative_error=0.5,
no_grad_set={"X"})
['Y'], 'Out', max_relative_error=0.5, no_grad_set=set("X"))
def test_ignore_y(self):
def test_check_grad_ingore_y(self):
self.check_grad(
self.op,
self.inputs, ["X"],
"Out",
max_relative_error=0.5,
no_grad_set={"Y"})
['X'], 'Out', max_relative_error=0.5, no_grad_set=set('Y'))
```
Some key points in the code above include:
- `create_op("mul")` creates the backward operator's corresponding forward operator.
- `compare_grad` compares results between utilizing the CPU and the GPU.
- `test_normal` calls `check_grad` to validate scaling tests' correctness and stability through numeric methods.
- The first variable `self.op` denotes the forward operator.
- The second variable `self.inputs` denotes the input dictionary, which has its key value identical to its `ProtoMaker` definitions.
- The third variable `["X", "Y"]` appoints `X` and `Y` to be scale tested.
- The fourth variable `"Out"` points to the network's final output target `Out`.
- `test_ignore_x` and `test_ignore_y`branches test the cases where there is only one scaling input.
- The first variable `["X", "Y"]` appoints `X` and `Y` to be scale tested.
- The second variable `"Out"` points to the network's final output target `Out`.
- The third variable `max_relative_error` points to the maximum relative tolerance error during scaling tests.
- `test_check_grad_ingore_x` and `test_check_grad_ingore_y`branches test the cases where there is only one scaling input.
### Compiling and Running
......
doc/howto/usage/cluster/src/word2vec/api_train_v2.py 0 → 100644
浏览文件 @ 0dc4b298
import gzip
import math
import paddle.v2 as paddle
embsize = 32
hiddensize = 256
N = 5
def wordemb(inlayer):
wordemb = paddle.layer.embedding(
input=inlayer,
size=embsize,
param_attr=paddle.attr.Param(
name="_proj",
initial_std=0.001,
learning_rate=1,
l2_rate=0,
sparse_update=True))
return wordemb
def main():
# for local training
cluster_train = False
if not cluster_train:
paddle.init(use_gpu=False, trainer_count=1)
else:
paddle.init(
use_gpu=False,
trainer_count=2,
port=7164,
ports_num=1,
ports_num_for_sparse=1,
num_gradient_servers=1)
word_dict = paddle.dataset.imikolov.build_dict()
dict_size = len(word_dict)
firstword = paddle.layer.data(
name="firstw", type=paddle.data_type.integer_value(dict_size))
secondword = paddle.layer.data(
name="secondw", type=paddle.data_type.integer_value(dict_size))
thirdword = paddle.layer.data(
name="thirdw", type=paddle.data_type.integer_value(dict_size))
fourthword = paddle.layer.data(
name="fourthw", type=paddle.data_type.integer_value(dict_size))
nextword = paddle.layer.data(
name="fifthw", type=paddle.data_type.integer_value(dict_size))
Efirst = wordemb(firstword)
Esecond = wordemb(secondword)
Ethird = wordemb(thirdword)
Efourth = wordemb(fourthword)
contextemb = paddle.layer.concat(input=[Efirst, Esecond, Ethird, Efourth])
hidden1 = paddle.layer.fc(input=contextemb,
size=hiddensize,
act=paddle.activation.Sigmoid(),
layer_attr=paddle.attr.Extra(drop_rate=0.5),
bias_attr=paddle.attr.Param(learning_rate=2),
param_attr=paddle.attr.Param(
initial_std=1. / math.sqrt(embsize * 8),
learning_rate=1))
predictword = paddle.layer.fc(input=hidden1,
size=dict_size,
bias_attr=paddle.attr.Param(learning_rate=2),
act=paddle.activation.Softmax())
def event_handler(event):
if isinstance(event, paddle.event.EndIteration):
if event.batch_id % 100 == 0:
with gzip.open("batch-" + str(event.batch_id) + ".tar.gz",
'w') as f:
trainer.save_parameter_to_tar(f)
result = trainer.test(
paddle.batch(
paddle.dataset.imikolov.test(word_dict, N), 32))
print "Pass %d, Batch %d, Cost %f, %s, Testing metrics %s" % (
event.pass_id, event.batch_id, event.cost, event.metrics,
result.metrics)
cost = paddle.layer.classification_cost(input=predictword, label=nextword)
parameters = paddle.parameters.create(cost)
adagrad = paddle.optimizer.AdaGrad(
learning_rate=3e-3,
regularization=paddle.optimizer.L2Regularization(8e-4))
trainer = paddle.trainer.SGD(cost,
parameters,
adagrad,
is_local=not cluster_train)
trainer.train(
paddle.batch(paddle.dataset.imikolov.train(word_dict, N), 32),
num_passes=30,
event_handler=event_handler)
if __name__ == '__main__':
main()
doc/howto/usage/cluster/src/word2vec/api_train_v2_cluster.py 0 → 100644
浏览文件 @ 0dc4b298
import math
import os
import paddle.v2 as paddle
import pickle
embsize = 32
hiddensize = 256
N = 5
cluster_train_file = "./train_data_dir/train/train.txt"
cluster_test_file = "./test_data_dir/test/test.txt"
node_id = os.getenv("OMPI_COMM_WORLD_RANK")
if not node_id:
raise EnvironmentError("must provied OMPI_COMM_WORLD_RANK")
def wordemb(inlayer):
wordemb = paddle.layer.embedding(
input=inlayer,
size=embsize,
param_attr=paddle.attr.Param(
name="_proj",
initial_std=0.001,
learning_rate=1,
l2_rate=0,
sparse_update=True))
return wordemb
def cluster_reader_cluster(filename, node_id):
def cluster_reader():
with open("-".join([filename, "%05d" % int(node_id)]), "r") as f:
for l in f:
csv_data = [int(cell) for cell in l.split(",")]
yield tuple(csv_data)
return cluster_reader
def main():
# get arguments from env
# for local training
TRUTH = ["true", "True", "TRUE", "1", "yes", "Yes", "YES"]
cluster_train = os.getenv('PADDLE_CLUSTER_TRAIN', "False") in TRUTH
use_gpu = os.getenv('PADDLE_INIT_USE_GPU', "False")
if not cluster_train:
paddle.init(
use_gpu=use_gpu,
trainer_count=int(os.getenv("PADDLE_INIT_TRAINER_COUNT", "1")))
else:
paddle.init(
use_gpu=use_gpu,
trainer_count=int(os.getenv("PADDLE_INIT_TRAINER_COUNT", "1")),
port=int(os.getenv("PADDLE_INIT_PORT", "7164")),
ports_num=int(os.getenv("PADDLE_INIT_PORTS_NUM", "1")),
ports_num_for_sparse=int(
os.getenv("PADDLE_INIT_PORTS_NUM_FOR_SPARSE", "1")),
num_gradient_servers=int(
os.getenv("PADDLE_INIT_NUM_GRADIENT_SERVERS", "1")),
trainer_id=int(os.getenv("PADDLE_INIT_TRAINER_ID", "0")),
pservers=os.getenv("PADDLE_INIT_PSERVERS", "127.0.0.1"))
fn = open("thirdparty/wuyi_train_thdpty/word_dict.pickle", "r")
word_dict = pickle.load(fn)
fn.close()
dict_size = len(word_dict)
firstword = paddle.layer.data(
name="firstw", type=paddle.data_type.integer_value(dict_size))
secondword = paddle.layer.data(
name="secondw", type=paddle.data_type.integer_value(dict_size))
thirdword = paddle.layer.data(
name="thirdw", type=paddle.data_type.integer_value(dict_size))
fourthword = paddle.layer.data(
name="fourthw", type=paddle.data_type.integer_value(dict_size))
nextword = paddle.layer.data(
name="fifthw", type=paddle.data_type.integer_value(dict_size))
Efirst = wordemb(firstword)
Esecond = wordemb(secondword)
Ethird = wordemb(thirdword)
Efourth = wordemb(fourthword)
contextemb = paddle.layer.concat(input=[Efirst, Esecond, Ethird, Efourth])
hidden1 = paddle.layer.fc(input=contextemb,
size=hiddensize,
act=paddle.activation.Sigmoid(),
layer_attr=paddle.attr.Extra(drop_rate=0.5),
bias_attr=paddle.attr.Param(learning_rate=2),
param_attr=paddle.attr.Param(
initial_std=1. / math.sqrt(embsize * 8),
learning_rate=1))
predictword = paddle.layer.fc(input=hidden1,
size=dict_size,
bias_attr=paddle.attr.Param(learning_rate=2),
act=paddle.activation.Softmax())
def event_handler(event):
if isinstance(event, paddle.event.EndIteration):
if event.batch_id % 100 == 0:
result = trainer.test(
paddle.batch(
cluster_reader_cluster(cluster_test_file, node_id), 32))
print "Pass %d, Batch %d, Cost %f, %s, Testing metrics %s" % (
event.pass_id, event.batch_id, event.cost, event.metrics,
result.metrics)
cost = paddle.layer.classification_cost(input=predictword, label=nextword)
parameters = paddle.parameters.create(cost)
adagrad = paddle.optimizer.AdaGrad(
learning_rate=3e-3,
regularization=paddle.optimizer.L2Regularization(8e-4))
trainer = paddle.trainer.SGD(cost,
parameters,
adagrad,
is_local=not cluster_train)
trainer.train(
paddle.batch(cluster_reader_cluster(cluster_train_file, node_id), 32),
num_passes=30,
event_handler=event_handler)
if __name__ == '__main__':
main()
doc/howto/usage/cluster/src/word2vec/prepare.py 0 → 100644
浏览文件 @ 0dc4b298
import paddle.v2 as paddle
import tarfile
import os
import pickle
SPLIT_COUNT = 3
N = 5
def file_len(fd):
for i, l in enumerate(fd):
pass
return i + 1
def split_from_reader_by_line(filename, reader, split_count):
fn = open(filename, "w")
for batch_id, batch_data in enumerate(reader()):
batch_data_str = [str(d) for d in batch_data]
fn.write(",".join(batch_data_str))
fn.write("\n")
fn.close()
fn = open(filename, "r")
total_line_count = file_len(fn)
fn.close()
per_file_lines = total_line_count / split_count + 1
cmd = "split -d -a 5 -l %d %s %s-" % (per_file_lines, filename, filename)
os.system(cmd)
word_dict = paddle.dataset.imikolov.build_dict()
with open("word_dict.pickle", "w") as dict_f:
pickle.dump(word_dict, dict_f)
split_from_reader_by_line("train.txt",
paddle.dataset.imikolov.train(word_dict, N),
SPLIT_COUNT)
split_from_reader_by_line("test.txt",
paddle.dataset.imikolov.test(word_dict, N),
SPLIT_COUNT)
doc/tutorials/image_classification/index_cn.md 已删除 100644 → 0
浏览文件 @ 931a23f1
图像分类教程
==========
在本教程中，我们将使用CIFAR-10数据集训练一个卷积神经网络，并使用这个神经网络来对图片进行分类。如下图所示，卷积神经网络可以辨识图片中的主体，并给出分类结果。
<center>![Image Classification](./image_classification.png)</center>
## 数据准备
首先下载CIFAR-10数据集。下面是CIFAR-10数据集的官方网址：
<https://www.cs.toronto.edu/~kriz/cifar.html>
我们准备了一个脚本，可以用于从官方网站上下载CIFAR-10数据集，转为jpeg文件并存入特定的目录。使用这个脚本前请确认已经安装了pillow及相关依赖模块。可以参照下面的命令进行安装：
1. 安装pillow
```bash
sudo apt-get install libjpeg-dev
pip install pillow
```
2. 下载数据集
```bash
cd demo/image_classification/data/
sh download_cifar.sh
```
CIFAR-10数据集包含60000张32x32的彩色图片。图片分为10类，每个类包含6000张。其中50000张图片作为训练集，10000张作为测试集。
下图展示了所有的图片类别，每个类别中随机抽取了10张图片。
<center>![Image Classification](./cifar.png)</center>
脚本运行完成后，我们应当会得到一个名为cifar-out的文件夹，其下子文件夹的结构如下
```
train
---airplane
---automobile
---bird
---cat
---deer
---dog
---frog
---horse
---ship
---truck
test
---airplane
---automobile
---bird
---cat
---deer
---dog
---frog
---horse
---ship
---truck
```
cifar-out下包含`train`和`test`两个文件夹，其中分别包含了CIFAR-10中的训练集和测试集。这两个文件夹下各自有10个子文件夹，每个子文件夹下存储相应分类的图片。将图片按照上述结构存储好之后，我们就可以着手对分类模型进行训练了。
## 预处理
数据下载之后，还需要进行预处理，将数据转换为Paddle的格式。我们可以通过如下命令进行预处理工作：
```
cd demo/image_classification/
sh preprocess.sh
```
其中`preprocess.sh` 调用 `./demo/image_classification/preprocess.py` 对图片进行预处理
```sh
export PYTHONPATH=$PYTHONPATH:../../
data_dir=./data/cifar-out
python preprocess.py -i $data_dir -s 32 -c 1
```
`./demo/image_classification/preprocess.py` 使用如下参数：
- `-i` 或 `--input` 给出输入数据所在路径；
- `-s` 或 `--size` 给出图片尺寸；
- `-c` 或 `--color` 标示图片是彩色图或灰度图
## 模型训练
在开始训练之前，我们需要先创建一个模型配置文件。下面我们给出了一个配置示例。**注意**，这里的列出的和`vgg_16_cifar.py`文件稍有差别，因为该文件可适用于预测。
```python
from paddle.trainer_config_helpers import *
data_dir='data/cifar-out/batches/'
meta_path=data_dir+'batches.meta'
args = {'meta':meta_path, 'mean_img_size': 32,
'img_size': 32, 'num_classes': 10,
'use_jpeg': 1, 'color': "color"}
define_py_data_sources2(train_list=data_dir+"train.list",
test_list=data_dir+'test.list',
module='image_provider',
obj='processData',
args=args)
settings(
batch_size = 128,
learning_rate = 0.1 / 128.0,
learning_method = MomentumOptimizer(0.9),
regularization = L2Regularization(0.0005 * 128))
img = data_layer(name='image', size=3*32*32)
lbl = data_layer(name="label", size=10)
# small_vgg is predined in trainer_config_helpers.network
predict = small_vgg(input_image=img, num_channels=3)
outputs(classification_cost(input=predict, label=lbl))
```
在第一行中我们载入用于定义网络的函数。
```python
from paddle.trainer_config_helpers import *
```
之后定义的`define_py_data_sources2`使用Python数据提供器，其中 `args`将在`image_provider.py`进行使用，该文件负责产生图片数据并传递给Paddle系统
- `meta`: 训练集平均值。
- `mean_img_size`: 平均特征图的高度及宽度。
- `img_size`：输入图片的高度及宽度。
- `num_classes`：类别个数。
- `use_jpeg`：处理过程中数据存储格式。
- `color`：标示是否为彩色图片。
`settings`用于设置训练算法。在下面的例子中，learning rate被设置为0.1除以batch size，而weight decay则为0.0005乘以batch size。
```python
settings(
batch_size = 128,
learning_rate = 0.1 / 128.0,
learning_method = MomentumOptimizer(0.9),
regularization = L2Regularization(0.0005 * 128)
)
```
`small_vgg`定义了网络结构。这里我们使用的是一个小的VGG网络。关于VGG卷积神经网络的描述可以参考：[http://www.robots.ox.ac.uk/~vgg/research/very_deep/](http://www.robots.ox.ac.uk/~vgg/research/very_deep/)。
```python
# small_vgg is predined in trainer_config_helpers.network
predict = small_vgg(input_image=img, num_channels=3)
```
配置创建完毕后，可以运行脚本train.sh来训练模型。
```bash
config=vgg_16_cifar.py
output=./cifar_vgg_model
log=train.log
paddle train \
--config=$config \
--dot_period=10 \
--log_period=100 \
--test_all_data_in_one_period=1 \
--use_gpu=1 \
--save_dir=$output \
2>&1 | tee $log
python -m paddle.utils.plotcurve -i $log > plot.png
```
- 这里我们使用的是GPU模式进行训练。如果你没有GPU环境，可以设置`use_gpu=0`。
- `./demo/image_classification/vgg_16_cifar.py`是网络和数据配置文件。各项参数的详细说明可以在命令行参数相关文档中找到。
- 脚本`plotcurve.py`依赖于python的`matplotlib`模块。因此如果这个脚本运行失败，也许是因为需要安装`matplotlib`。
在训练完成后，训练及测试误差曲线图会被`plotcurve.py`脚本保存在 `plot.png`中。下面是一个误差曲线图的示例：
<center>![Training and testing curves.](./plot.png)</center>
## 预测
在训练完成后，模型及参数会被保存在路径`./cifar_vgg_model/pass-%05d`下。例如第300个pass的模型会被保存在`./cifar_vgg_model/pass-00299`。
要对一个图片的进行分类预测，我们可以使用`predict.sh`，该脚本将输出预测分类的标签：
```
sh predict.sh
```
predict.sh:
```
model=cifar_vgg_model/pass-00299/
image=data/cifar-out/test/airplane/seaplane_s_000978.png
use_gpu=1
python prediction.py $model $image $use_gpu
```
## 练习
在CUB-200数据集上使用VGG模型训练一个鸟类图片分类模型。相关的鸟类数据集可以从如下地址下载，其中包含了200种鸟类的照片（主要来自北美洲）。
<http://www.vision.caltech.edu/visipedia/CUB-200.html>
## 细节探究
### 卷积神经网络
卷积神经网络是一种使用卷积层的前向神经网络，很适合构建用于理解图片内容的模型。一个典型的神经网络如下图所示：
![Convolutional Neural Network](./lenet.png)
一个卷积神经网络包含如下层：
- 卷积层：通过卷积操作从图片或特征图中提取特征
- 池化层：使用max-pooling对特征图下采样
- 全连接层：使输入层到隐藏层的神经元是全部连接的。
卷积神经网络在图片分类上有着惊人的性能，这是因为它发掘出了图片的两类重要信息：局部关联性质和空间不变性质。通过交替使用卷积和池化处理， 卷积神经网络能够很好的表示这两类信息。
关于如何定义网络中的层，以及如何在层之间进行连接，请参考Layer文档。
doc/tutorials/image_classification/index_en.md 已删除 100644 → 0
浏览文件 @ 931a23f1
Image Classification Tutorial
==============================
This tutorial will guide you through training a convolutional neural network to classify objects using the CIFAR-10 image classification dataset.
As shown in the following figure, the convolutional neural network can recognize the main object in images, and output the classification result.
<center>![Image Classification](./image_classification.png)</center>
## Data Preparation
First, download CIFAR-10 dataset. CIFAR-10 dataset can be downloaded from its official website.
<https://www.cs.toronto.edu/~kriz/cifar.html>
We have prepared a script to download and process CIFAR-10 dataset. The script will download CIFAR-10 dataset from the official dataset.
It will convert it to jpeg images and organize them into a directory with the required structure for the tutorial. Make sure that you have installed pillow and its dependents.
Consider the following commands:
1. install pillow dependents
```bash
sudo apt-get install libjpeg-dev
pip install pillow
```
2. download data and preparation
```bash
cd demo/image_classification/data/
sh download_cifar.sh
```
The CIFAR-10 dataset consists of 60000 32x32 color images in 10 classes, with 6000 images per class. There are 50000 training images and 10000 test images.
Here are the classes in the dataset, as well as 10 random images from each:
<center>![Image Classification](./cifar.png)</center>
After downloading and converting, we should find a directory (cifar-out) containing the dataset in the following format:
```
train
---airplane
---automobile
---bird
---cat
---deer
---dog
---frog
---horse
---ship
---truck
test
---airplane
---automobile
---bird
---cat
---deer
---dog
---frog
---horse
---ship
---truck
```
It has two directories:`train` and `test`. These two directories contain training data and testing data of CIFAR-10, respectively. Each of these two folders contains 10 sub-folders, ranging from `airplane` to `truck`. Each sub-folder contains images with the corresponding label. After the images are organized into this structure, we are ready to train an image classification model.
## Preprocess
After the data has been downloaded, it needs to be pre-processed into the Paddle format. We can run the following command for preprocessing.
```
cd demo/image_classification/
sh preprocess.sh
```
`preprocess.sh` calls `./demo/image_classification/preprocess.py` to preprocess image data.
```sh
export PYTHONPATH=$PYTHONPATH:../../
data_dir=./data/cifar-out
python preprocess.py -i $data_dir -s 32 -c 1
```
`./demo/image_classification/preprocess.py` has the following arguments
- `-i` or `--input` specifes the input data directory.
- `-s` or `--size` specifies the processed size of images.
- `-c` or `--color` specifes whether images are color images or gray images.
## Model Training
We need to create a model config file before training the model. An example of the config file (vgg_16_cifar.py) is listed below. **Note**, it is slightly different from the `vgg_16_cifar.py` which also applies to the prediction.
```python
from paddle.trainer_config_helpers import *
data_dir='data/cifar-out/batches/'
meta_path=data_dir+'batches.meta'
args = {'meta':meta_path, 'mean_img_size': 32,
'img_size': 32, 'num_classes': 10,
'use_jpeg': 1, 'color': "color"}
define_py_data_sources2(train_list=data_dir+"train.list",
test_list=data_dir+'test.list',
module='image_provider',
obj='processData',
args=args)
settings(
batch_size = 128,
learning_rate = 0.1 / 128.0,
learning_method = MomentumOptimizer(0.9),
regularization = L2Regularization(0.0005 * 128))
img = data_layer(name='image', size=3*32*32)
lbl = data_layer(name="label", size=10)
# small_vgg is predined in trainer_config_helpers.network
predict = small_vgg(input_image=img, num_channels=3)
outputs(classification_cost(input=predict, label=lbl))
```
The first line imports python functions for defining networks.
```python
from paddle.trainer_config_helpers import *
```
Then define an `define_py_data_sources2` which use python data provider
interface. The arguments in `args` are used in `image_provider.py` which
yeilds image data and transform them to Paddle.
- `meta`: the mean value of training set.
- `mean_img_size`: the size of mean feature map.
- `img_size`: the height and width of input image.
- `num_classes`: the number of classes.
- `use_jpeg`: the data storage type when preprocessing.
- `color`: specify color image.
`settings` specifies the training algorithm. In the following example,
it specifies learning rate as 0.1, but divided by batch size, and the weight decay
is 0.0005 and multiplied by batch size.
```python
settings(
batch_size = 128,
learning_rate = 0.1 / 128.0,
learning_method = MomentumOptimizer(0.9),
regularization = L2Regularization(0.0005 * 128)
)
```
The `small_vgg` specifies the network. We use a small version of VGG convolutional network as our network
for classification. A description of VGG network can be found here [http://www.robots.ox.ac.uk/~vgg/research/very_deep/](http://www.robots.ox.ac.uk/~vgg/research/very_deep/).
```python
# small_vgg is predined in trainer_config_helpers.network
predict = small_vgg(input_image=img, num_channels=3)
```
After writing the config, we can train the model by running the script train.sh.
```bash
config=vgg_16_cifar.py
output=./cifar_vgg_model
log=train.log
paddle train \
--config=$config \
--dot_period=10 \
--log_period=100 \
--test_all_data_in_one_period=1 \
--use_gpu=1 \
--save_dir=$output \
2>&1 | tee $log
python -m paddle.utils.plotcurve -i $log > plot.png
```
- Here we use GPU mode to train. If you have no gpu environment, just set `use_gpu=0`.
- `./demo/image_classification/vgg_16_cifar.py` is the network and data configuration file. The meaning of the other flags can be found in the documentation of the command line flags.
- The script `plotcurve.py` requires the python module of `matplotlib`, so if it fails, maybe you need to install `matplotlib`.
After training finishes, the training and testing error curves will be saved to `plot.png` using `plotcurve.py` script. An example of the plot is shown below:
<center>![Training and testing curves.](./plot.png)</center>
## Prediction
After we train the model, the model file as well as the model parameters are stored in path `./cifar_vgg_model/pass-%05d`. For example, the model of the 300-th pass is stored at `./cifar_vgg_model/pass-00299`.
To make a prediction for an image, one can run `predict.sh` as follows. The script will output the label of the classfiication.
```
sh predict.sh
```
predict.sh:
```
model=cifar_vgg_model/pass-00299/
image=data/cifar-out/test/airplane/seaplane_s_000978.png
use_gpu=1
python prediction.py $model $image $use_gpu
```
## Exercise
Train a image classification of birds using VGG model and CUB-200 dataset. The birds dataset can be downloaded here. It contains an image dataset with photos of 200 bird species (mostly North American).
<http://www.vision.caltech.edu/visipedia/CUB-200.html>
## Delve into Details
### Convolutional Neural Network
A Convolutional Neural Network is a feedforward neural network that uses convolution layers. It is very suitable for building neural networks that process and understand images. A standard convolutional neural network is shown below:
![Convolutional Neural Network](./lenet.png)
Convolutional Neural Network contains the following layers:
- Convolutional layer: It uses convolution operation to extract features from an image or a feature map.
- Pooling layer: It uses max-pooling to downsample feature maps.
- Fully Connected layer: It uses fully connected connections to transform features.
Convolutional Neural Network achieves amazing performance for image classification because it exploits two important characteristics of images: *local correlation* and *spatial invariance*. By iteratively applying convolution and max-pooing operations, convolutional neural network can well represent these two characteristics of images.
For more details of how to define layers and their connections, please refer to the documentation of layers.
doc/tutorials/index_cn.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/index_en.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/rec/ml_dataset_cn.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/rec/ml_dataset_en.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/rec/ml_regression_cn.rst 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/rec/ml_regression_en.rst 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/rec/rec_regression_network.png 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/sentiment_analysis/index_cn.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/sentiment_analysis/index_en.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/sentiment_analysis/lstm.png 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/sentiment_analysis/src/lstm.png 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/text_generation/index_cn.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/tutorials/text_generation/index_en.md 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
doc/v1_api_tutorials/README.md 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
go/cmd/master/master.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/cmd/pserver/pserver.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/glide.lock
浏览文件 @ 0dc4b298
此差异已折叠。
go/glide.yaml
浏览文件 @ 0dc4b298
此差异已折叠。
go/master/c/client.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/master/client.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/master/client_internal_test.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/master/etcd_client.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/master/service.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/client/c/cclient.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/client/client.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/client/client_test.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/client/etcd_client.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/etcd_client.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/optimizer.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/optimizer_test.go
浏览文件 @ 0dc4b298
此差异已折叠。
go/pserver/service.go
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/api/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/api/Util.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/capi/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/capi/Matrix.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/capi/tests/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/attribute.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/attribute.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/backward.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/backward.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/backward_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/block_desc.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/block_desc.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/data_type.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/details/op_registry.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/executor.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/executor.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/feed_fetch_method.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/feed_fetch_type.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/framework.proto
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/grad_op_builder.cc 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
paddle/framework/grad_op_builder_test.cc 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
paddle/framework/grad_op_desc_maker.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/lod_tensor.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/lod_tensor.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/lod_tensor.md
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/lod_tensor_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/lod_tensor_test.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/op_desc.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/op_desc.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/op_info.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/op_proto_maker.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/op_registry.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/op_registry.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/operator.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/operator.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/operator_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/program_desc.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/program_desc.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/program_desc_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/proto_desc.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/prune.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/prune_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/saver.proto 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/scope.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/scope.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/scope_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/selected_rows.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/selected_rows.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/selected_rows_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/shape_inference.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/tensor.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/tensor_array.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/tensor_array.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/tensor_array_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/tensor_impl.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/tensor_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/type_defs.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/var_desc.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/var_desc.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/var_type_inference.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/var_type_inference_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/framework/variable.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/BlockExpandOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/CosSimOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/CropOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/DepthwiseConvOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/GemmConvOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/GemmConvOpTest.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/Im2ColTest.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/MulOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/PadOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/RowConvOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/function/SwitchOp.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/layers/CostLayer.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/layers/Layer.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/layers/Layer.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/layers/MKLDNNBase.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/layers/MKLDNNLayer.cpp 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/layers/MKLDNNLayer.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/layers/PoolLayer.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/tests/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/tests/MKLDNNTester.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/tests/mkldnn_branches_conv.conf 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/tests/mkldnn_branches_fc.conf 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/gserver/tests/mkldnn_branches_pool.conf 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/MKLDNNMatrix.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/MKLDNNMatrix.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/Matrix.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/RowBuffer.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/SparseMatrix.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/Vector.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/tests/test_Matrix.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/math/tests/test_Tensor.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/memory/memcpy.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/memory/memcpy.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/memory/memory.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/memory/memory_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/accuracy_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/accuracy_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/accuracy_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/activation_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/activation_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/activation_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adadelta_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adadelta_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adadelta_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adagrad_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adagrad_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adagrad_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adam_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adam_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adam_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adamax_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/adamax_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/batch_norm_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/batch_norm_op.md 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/clip_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/clip_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/concat_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/concat_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/cond_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/cond_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv2d_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv2d_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv2dtranspose_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv2dtranspose_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv2dtranspose_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv_cudnn_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv_cudnn_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv_shift_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv_shift_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/conv_shift_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/cos_sim_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/cos_sim_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/crop_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/crop_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/cross_entropy_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/decayed_adagrad_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/decayed_adagrad_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/decayed_adagrad_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/dropout_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/dropout_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/dropout_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/dynamic_recurrent_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/dynamic_recurrent_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/dynamic_recurrent_op_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/elementwise_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/fc_op.cc 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
paddle/operators/feed_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/fetch_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/fill_constant_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gather.cu.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gather.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gather_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gather_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gather_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gru_unit_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gru_unit_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/gru_unit_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/images/batch_norm_fork.dot 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/images/batch_norm_fork.png 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/images/batch_norm_op_kernel.png 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/increment_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/increment_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/increment_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lookup_table_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lookup_table_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lookup_table_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lstm_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lstm_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lstm_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lstm_unit_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lstm_unit_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/lstm_unit_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/margin_rank_loss_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/margin_rank_loss_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/margin_rank_loss_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/CMakeLists.txt 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/hl_activation_functions.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/hl_avx_functions.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/hl_avx_functions.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/hl_cpu_functions.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/hl_functions.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/hl_gpu_functions.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/lstm_cpu_kernel.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/lstm_gpu_kernel.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/detail/lstm_kernel.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/im2col.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/im2col.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/im2col.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/lstm_compute.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/lstm_compute.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/lstm_compute.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/math_function_test.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/matmul.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/pooling.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/pooling.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/pooling.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/selected_rows_functor.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/selected_rows_functor.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/selected_rows_functor.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/selected_rows_functor_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/selected_rows_functor_test.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/sequence2batch.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/sequence2batch.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/sequence2batch.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/softmax.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/softmax.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/softmax.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/vol2col.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/vol2col.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/vol2col.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/math/vol2col_test.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/matmul_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/matmul_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/matmul_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/mean_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/mean_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/minus_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/minus_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/momentum_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/momentum_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/momentum_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/mul_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/mul_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/multiplex_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/multiplex_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/multiplex_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/name_convention.md
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/net_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pad_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pad_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pool_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pool_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pool_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pool_with_index_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pool_with_index_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/pool_with_index_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/prelu_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/prelu_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/proximal_gd_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/proximal_gd_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/rank_loss_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/rank_loss_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/recurrent_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/recurrent_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/reduce_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/reduce_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/reduce_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/reshape_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/reshape_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/rmsprop_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/rmsprop_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/rmsprop_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/rowwise_add_op.cc 已删除 100644 → 0
浏览文件 @ 931a23f1
此差异已折叠。
paddle/operators/save_restore_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scale_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scale_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scatter.cu.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scatter.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scatter_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scatter_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scatter_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/scatter_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sequence_concat_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sequence_concat_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sequence_concat_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sequence_pool_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sequence_softmax_op.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sequence_softmax_op.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sequence_softmax_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sgd_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sgd_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sgd_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sigmoid_cross_entropy_with_logits_op.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/softmax_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/softmax_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/split_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/split_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sum_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sum_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/sum_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/top_k_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/top_k_op.cu
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/top_k_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/transpose_op.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/operators/transpose_op.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/optimizer/adam_optimizer.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/optimizer/adam_optimizer.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/optimizer/lr_policy.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/optimizer/optimizer.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/optimizer/sgd_optimizer.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/optimizer/sgd_optimizer.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/cudnn_helper.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/device_context.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/device_context.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/dynload/cublas.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/dynload/nccl.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/dynload/nccl.h 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/enforce.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/enforce_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/gpu_info.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/gpu_info.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/hostdevice.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/macros.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/nccl_test.cu 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/place.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/place.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/platform/variant.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/pserver/LightNetwork.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/pserver/test/SocketTest.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/pybind/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/pybind/print_operators_doc.cc 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/pybind/protobuf.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/pybind/pybind.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/pybind/tensor_py.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/scripts/cluster_train_v2/fabric/conf.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/scripts/cluster_train_v2/fabric/run.sh 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/scripts/docker/build.sh
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/scripts/submit_local.sh.in
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/string/to_string_test.cc
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/trainer/MergeModel.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/trainer/tests/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/utils/Flags.cpp
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/utils/Util.h
浏览文件 @ 0dc4b298
此差异已折叠。
paddle/utils/Version.h
浏览文件 @ 0dc4b298
此差异已折叠。
proto/CMakeLists.txt
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/event.py
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/backward.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/executor.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/framework.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/layer_helper.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/layers.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/nets.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/op.py
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/optimizer.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/.gitignore 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_adadelta_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_adagrad_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_adam_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_adamax_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_batch_norm_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_conv_shift_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_fit_a_line.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_gru_unit_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_increment_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_infer_shape.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_layers.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_lstm_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_matmul_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_momentum_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_op_support_gpu.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_operator_desc.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_optimizer.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_parameter.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_pool2d_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_pool3d_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_pool_max_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_program.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_proximal_gd_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_rmsprop_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_rnn_helpers.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_selected_rows.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_seq_concat_op.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_tensor_array.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_variable.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/framework/tests/test_word2vec.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/inference.py
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/parameters.py
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/tests/test_paramconf_order.py 0 → 100644
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/topology.py
浏览文件 @ 0dc4b298
此差异已折叠。
python/paddle/v2/trainer.py
浏览文件 @ 0dc4b298
此差异已折叠。
v1_api_demo/README.md
浏览文件 @ 0dc4b298
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