Merge branch 'develop' into doc3

.dockerignore

@@ -0,0 +1 @@
+.gitignore

.gitignore

@@ -8,3 +8,4 @@ build/
 .cproject
 .pydevproject
 Makefile
+.test_env/

CMakeLists.txt

@@ -1,10 +1,6 @@
 cmake_minimum_required(VERSION 2.8)
 
 project(paddle CXX C)
-set(PADDLE_MAJOR_VERSION 0)
-set(PADDLE_MINOR_VERSION 9)
-set(PADDLE_PATCH_VERSION 0a0)
-set(PADDLE_VERSION ${PADDLE_MAJOR_VERSION}.${PADDLE_MINOR_VERSION}.${PADDLE_PATCH_VERSION})
 
 set(CMAKE_MODULE_PATH ${CMAKE_MODULE_PATH} "${CMAKE_SOURCE_DIR}/cmake")
 set(PROJ_ROOT ${CMAKE_SOURCE_DIR})
@@ -12,6 +8,17 @@ include(package)
 find_package(SWIG 2.0)
 find_package(CUDA QUIET)
 find_package(Protobuf REQUIRED)
+
+# Check protobuf library version.
+execute_process(COMMAND ${PROTOBUF_PROTOC_EXECUTABLE} --version
+	OUTPUT_VARIABLE PROTOBUF_VERSION)
+string(REPLACE "libprotoc " "" PROTOBUF_VERSION ${PROTOBUF_VERSION})
+
+set(PROTOBUF_3 OFF)
+if (${PROTOBUF_VERSION} VERSION_GREATER "3.0.0" OR ${PROTOBUF_VERSION} VERSION_EQUAL "3.0.0")
+    set(PROTOBUF_3 ON)
+endif()
+
 find_package(PythonLibs 2.7 REQUIRED)
 find_package(PythonInterp 2.7 REQUIRED)
 find_package(ZLIB REQUIRED)
@@ -45,7 +52,7 @@ option(ON_COVERALLS "Generating code coverage data on coveralls or not." OFF)
 option(COVERALLS_UPLOAD "Uploading the generated coveralls json." ON)
 
 if(NOT CMAKE_BUILD_TYPE)
-    set(CMAKE_BUILD_TYPE "RelWithDebInfo" CACHE STRING 
+    set(CMAKE_BUILD_TYPE "RelWithDebInfo" CACHE STRING
         "Choose the type of build, options are: Debug Release RelWithDebInfo MinSizeRel"
         FORCE)
 endif()
@@ -64,31 +71,11 @@ include(check_packages)
 include(swig)
 include(coveralls)
 
-# add PaddlePaddle version
+# Set PaddlePaddle version to Git tag name or Git commit ID.
-if(DEFINED ENV{PADDLE_VERSION})
+find_package(Git REQUIRED)
-    add_definitions(-DPADDLE_VERSION=\"$ENV{PADDLE_VERSION}\")
+# version.cmake will get the current PADDLE_VERSION
-else()
+include(version)
-    if(EXISTS ${PROJ_ROOT}/.svn/)
+add_definitions(-DPADDLE_VERSION=\"${PADDLE_VERSION}\")
-        find_package(Subversion REQUIRED)
-        if(SUBVERSION_FOUND)
-            Subversion_WC_INFO(${PROJ_ROOT} Project)
-            add_definitions(-DPADDLE_VERSION=${Project_WC_REVISION})
-        endif()
-    elseif(EXISTS ${PROJ_ROOT}/.git/)
-        find_package(Git REQUIRED)
-        execute_process(
-            COMMAND ${GIT_EXECUTABLE} log -1 --format=%H
-            WORKING_DIRECTORY ${PROJ_ROOT}
-            OUTPUT_VARIABLE GIT_SHA1
-            RESULT_VARIABLE GIT_RESULT
-            ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
-        if(NOT ${GIT_RESULT})
-            add_definitions(-DPADDLE_VERSION=\"${GIT_SHA1}\")
-        else()
-            message(WARNING "Cannot add paddle version from git tag")
-        endif()
-    endif()
-endif()
 
 
 if(NOT WITH_GPU)

RELEASE.md

@@ -0,0 +1,69 @@
+# Release v0.9.0
+
+## New Features:
+
+* New Layers
+  * bilinear interpolation layer.
+  * spatial pyramid-pool layer.
+  * de-convolution layer.
+  * maxout layer.
+* Support rectangle padding, stride, window and input for Pooling Operation.
+* Add —job=time in trainer, which can be used to print time info without compiler option -WITH_TIMER=ON.
+* Expose cost_weight/nce_layer in `trainer_config_helpers`
+* Add FAQ, concepts, h-rnn docs.
+* Add Bidi-LSTM and DB-LSTM to quick start demo @alvations
+* Add usage track scripts.
+
+## Improvements
+
+* Add Travis-CI for Mac OS X. Enable swig unittest in Travis-CI. Skip Travis-CI when only docs are changed.
+* Add code coverage tools.
+* Refine convolution layer to speedup and reduce GPU memory.
+* Speed up PyDataProvider2
+* Add ubuntu deb package build scripts.
+* Make Paddle use git-flow branching model.
+* PServer support no parameter blocks.
+
+## Bug Fixes
+
+* add zlib link to py_paddle
+* add input sparse data check for sparse layer at runtime
+* Bug fix for sparse matrix multiplication
+* Fix floating-point overflow problem of tanh
+* Fix some nvcc compile options
+* Fix a bug in yield dictionary in DataProvider
+* Fix SRL hang when exit.
+
+# Release v0.8.0beta.1
+New features:
+
+* Mac OSX is supported by source code. #138
+   * Both GPU and CPU versions of PaddlePaddle are supported.
+
+* Support CUDA 8.0
+
+* Enhance `PyDataProvider2`
+   * Add dictionary yield format. `PyDataProvider2` can yield a dictionary with key is data_layer's name, value is features.
+   * Add `min_pool_size` to control memory pool in provider.
+
+* Add `deb` install package & docker image for no_avx machines.
+   * Especially for cloud computing and virtual machines
+
+* Automatically disable `avx` instructions in cmake when machine's CPU don't support `avx` instructions.
+
+* Add Parallel NN api in trainer_config_helpers.
+
+* Add `travis ci` for Github
+
+Bug fixes:
+
+* Several bugs in trainer_config_helpers. Also complete the unittest for trainer_config_helpers
+* Check if PaddlePaddle is installed when unittest.
+* Fix bugs in GTX series GPU
+* Fix bug in MultinomialSampler
+
+Also more documentation was written since last release.
+
+# Release v0.8.0beta.0
+
+PaddlePaddle v0.8.0beta.0 release. The install package is not stable yet and it's a pre-release version.

cmake/version.cmake

@@ -0,0 +1,24 @@
+# Get the latest git tag.
+set(PADDLE_VERSION $ENV{PADDLE_VERSION})
+set(tmp_version "HEAD")
+while ("${PADDLE_VERSION}" STREQUAL "")
+  execute_process(
+    COMMAND ${GIT_EXECUTABLE} describe --tags --abbrev=0 ${tmp_version}
+    WORKING_DIRECTORY ${PROJ_ROOT}
+    OUTPUT_VARIABLE GIT_TAG_NAME
+    RESULT_VARIABLE GIT_RESULT
+    ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
+  if (NOT ${GIT_RESULT})
+    # Check the tag is a correct version
+    if (${GIT_TAG_NAME} MATCHES "v[0-9]+\\.[0-9]+\\.[0-9]+(\\.(a|b|rc)\\.[0-9]+)?")
+      string(REPLACE "v" "" PADDLE_VERSION ${GIT_TAG_NAME})
+    else()  # otherwise, get the previous git tag name.
+      set(tmp_version "${GIT_TAG_NAME}~1")
+    endif()
+  else()
+    set(PADDLE_VERSION "0.0.0")
+    message(WARNING "Cannot add paddle version from git tag")
+  endif()
+endwhile()
+
+message(STATUS "Paddle version is ${PADDLE_VERSION}")

demo/gan/.gitignore

@@ -0,0 +1,11 @@
+output/
+uniform_params/
+cifar_params/
+mnist_params/
+*.png
+.pydevproject
+.project
+*.log
+*.pyc
+data/mnist_data/
+data/cifar-10-batches-py/

demo/gan/README.md

@@ -0,0 +1,13 @@
+# Generative Adversarial Networks (GAN) 
+
+This demo implements GAN training described in the original GAN paper (https://arxiv.org/abs/1406.2661) and DCGAN (https://arxiv.org/abs/1511.06434).
+
+The general training procedures are implemented in gan_trainer.py. The neural network configurations are specified in gan_conf.py (for synthetic data) and gan_conf_image.py (for image data).
+
+In order to run the model, first download the corresponding data by running the shell script in ./data.
+Then you can run the command below. The flag -d specifies the training data (cifar, mnist or uniform) and flag --useGpu specifies whether to use gpu for training (0 is cpu, 1 is gpu).  
+
+$python gan_trainer.py -d cifar --use_gpu 1
+
+The generated images will be stored in ./cifar_samples/
+The corresponding models will be stored in ./cifar_params/

demo/gan/data/download_cifar.sh

@@ -0,0 +1,18 @@
+# Copyright (c) 2016 Baidu, Inc. All Rights Reserved
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+set -e
+wget https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz
+tar zxf cifar-10-python.tar.gz
+rm cifar-10-python.tar.gz
+

demo/gan/data/get_mnist_data.sh

@@ -0,0 +1,19 @@
+#!/usr/bin/env sh
+# This script downloads the mnist data and unzips it.
+set -e
+DIR="$( cd "$(dirname "$0")" ; pwd -P )"
+rm -rf "$DIR/mnist_data"
+mkdir "$DIR/mnist_data"
+cd "$DIR/mnist_data"
+
+echo "Downloading..."
+
+for fname in train-images-idx3-ubyte train-labels-idx1-ubyte t10k-images-idx3-ubyte t10k-labels-idx1-ubyte
+do
+    if [ ! -e $fname ]; then
+        wget --no-check-certificate http://yann.lecun.com/exdb/mnist/${fname}.gz
+        gunzip ${fname}.gz
+    fi
+done
+
+

demo/gan/gan_conf.py

@@ -0,0 +1,134 @@
+# Copyright (c) 2016 Baidu, Inc. All Rights Reserved
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from paddle.trainer_config_helpers import *
+
+mode = get_config_arg("mode", str, "generator")
+assert mode in set(["generator",
+                    "discriminator",
+                    "generator_training",
+                    "discriminator_training"])
+
+is_generator_training = mode == "generator_training"
+is_discriminator_training = mode == "discriminator_training"
+is_generator = mode == "generator"
+is_discriminator = mode == "discriminator"
+
+# The network structure below follows the ref https://arxiv.org/abs/1406.2661
+# Here we used two hidden layers and batch_norm
+
+print('mode=%s' % mode)
+# the dim of the noise (z) as the input of the generator network
+noise_dim = 10
+# the dim of the hidden layer
+hidden_dim = 10
+# the dim of the generated sample
+sample_dim = 2
+
+settings(
+    batch_size=128,
+    learning_rate=1e-4,
+    learning_method=AdamOptimizer(beta1=0.5)
+)
+
+def discriminator(sample):
+    """
+    discriminator ouputs the probablity of a sample is from generator
+    or real data.
+    The output has two dimenstional: dimension 0 is the probablity
+    of the sample is from generator and dimension 1 is the probabblity
+    of the sample is from real data.
+    """
+    param_attr = ParamAttr(is_static=is_generator_training)
+    bias_attr = ParamAttr(is_static=is_generator_training,
+                          initial_mean=1.0,
+                          initial_std=0)
+
+    hidden = fc_layer(input=sample, name="dis_hidden", size=hidden_dim,
+                    bias_attr=bias_attr,
+                    param_attr=param_attr,
+                    act=ReluActivation())
+
+    hidden2 = fc_layer(input=hidden, name="dis_hidden2", size=hidden_dim,
+                    bias_attr=bias_attr,
+                    param_attr=param_attr,
+                    act=LinearActivation())
+    
+    hidden_bn = batch_norm_layer(hidden2, 
+                     act=ReluActivation(), 
+                     name="dis_hidden_bn", 
+                     bias_attr=bias_attr, 
+                     param_attr=ParamAttr(is_static=is_generator_training,
+                           initial_mean=1.0,
+                           initial_std=0.02),
+                     use_global_stats=False)
+    
+    return fc_layer(input=hidden_bn, name="dis_prob", size=2,
+                    bias_attr=bias_attr,
+                    param_attr=param_attr,
+                    act=SoftmaxActivation())
+
+def generator(noise):
+    """
+    generator generates a sample given noise
+    """
+    param_attr = ParamAttr(is_static=is_discriminator_training)
+    bias_attr = ParamAttr(is_static=is_discriminator_training,
+                           initial_mean=1.0,
+                           initial_std=0)
+    
+    hidden = fc_layer(input=noise,
+                    name="gen_layer_hidden",
+                    size=hidden_dim,
+                    bias_attr=bias_attr,
+                    param_attr=param_attr,
+                    act=ReluActivation())
+
+    hidden2 = fc_layer(input=hidden, name="gen_hidden2", size=hidden_dim,
+                    bias_attr=bias_attr,
+                    param_attr=param_attr,
+                    act=LinearActivation())
+    
+    hidden_bn = batch_norm_layer(hidden2, 
+                     act=ReluActivation(), 
+                     name="gen_layer_hidden_bn", 
+                     bias_attr=bias_attr, 
+                     param_attr=ParamAttr(is_static=is_discriminator_training,
+                           initial_mean=1.0,
+                           initial_std=0.02),
+                     use_global_stats=False)
+    
+    return fc_layer(input=hidden_bn,
+                    name="gen_layer1",
+                    size=sample_dim,
+                    bias_attr=bias_attr,
+                    param_attr=param_attr,
+                    act=LinearActivation())
+
+if is_generator_training:
+    noise = data_layer(name="noise", size=noise_dim)
+    sample = generator(noise)
+
+if is_discriminator_training:
+    sample = data_layer(name="sample", size=sample_dim)
+
+if is_generator_training or is_discriminator_training:
+    label = data_layer(name="label", size=1)
+    prob = discriminator(sample)
+    cost = cross_entropy(input=prob, label=label)
+    classification_error_evaluator(input=prob, label=label, name=mode+'_error')
+    outputs(cost)
+
+if is_generator:
+    noise = data_layer(name="noise", size=noise_dim)
+    outputs(generator(noise))

doc/CMakeLists.txt

@@ -15,25 +15,11 @@ set(SPHINX_CACHE_DIR "${CMAKE_CURRENT_BINARY_DIR}/_doctrees")
 # HTML output directory
 set(SPHINX_HTML_DIR "${CMAKE_CURRENT_BINARY_DIR}/html")
 
-
-set(PADDLE_DOXYGEN_OUTPUT "${CMAKE_CURRENT_BINARY_DIR}/doxygen_xml")
-
 configure_file(
     "${CMAKE_CURRENT_SOURCE_DIR}/conf.py.in"
     "${BINARY_BUILD_DIR}/conf.py"
     @ONLY)
 
-configure_file(
-    "${CMAKE_CURRENT_SOURCE_DIR}/Doxyfile.in"
-    "${CMAKE_CURRENT_BINARY_DIR}/Doxyfile"
-    @ONLY
-  )
-
-add_custom_target(paddle_doxygen_docs ALL
-    ${DOXYGEN_EXECUTABLE} ${CMAKE_CURRENT_BINARY_DIR}/Doxyfile
-    WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
-)
-
 sphinx_add_target(paddle_docs
                   html
                   ${BINARY_BUILD_DIR}
@@ -41,6 +27,5 @@ sphinx_add_target(paddle_docs
                   ${CMAKE_CURRENT_SOURCE_DIR}
                   ${SPHINX_HTML_DIR})
 
-add_dependencies(paddle_docs 
+add_dependencies(paddle_docs
-  gen_proto_py
+  gen_proto_py)
-  paddle_doxygen_docs)

doc/about/index.rst

@@ -0,0 +1,14 @@
+ABOUT
+=======
+
+PaddlPaddle is an easy-to-use, efficient, flexible and scalable deep learning platform,
+which is originally developed by Baidu scientists and engineers for the purpose of applying deep learning to many products at Baidu.
+
+PaddlePaddle is now open source but far from complete, which is intended to be built upon, improved, scaled, and extended.
+We hope to build an active open source community both by providing feedback and by actively contributing to the source code.
+
+
+Credits
+--------
+
+We owe many thanks to `all contributors and developers <https://github.com/PaddlePaddle/Paddle/blob/develop/authors>`_ of PaddlePaddle!

doc/algorithm/index.rst

@@ -1,7 +0,0 @@
-Algorithm Tutorial
-==================
-
-..  toctree::
-  :maxdepth: 1
-
-  rnn/rnn.rst

doc/algorithm/rnn/bi_lstm.jpg

@@ -1 +0,0 @@
-../../demo/sentiment_analysis/bi_lstm.jpg

doc/algorithm/rnn/encoder-decoder-attention-model.png

@@ -1 +0,0 @@
-../../demo/text_generation/encoder-decoder-attention-model.png

doc/api/data_provider/index.rst

@@ -1,5 +1,5 @@
-DataProvider Introduction
+Introduction
-=========================
+==============
 DataProvider is a module that loads training or testing data into cpu or gpu
 memory for the following triaining or testing process.
 

doc/api/data_provider/pydataprovider2.rst

@@ -1,5 +1,5 @@
-How to use PyDataProvider2
+PyDataProvider2
-==========================
+=================
 
 We highly recommand users to use PyDataProvider2 to provide training or testing
 data to PaddlePaddle. The user only needs to focus on how to read a single

doc/api/index.rst

@@ -0,0 +1,36 @@
+API
+====
+
+DataProvider API
+----------------
+
+..  toctree::
+  :maxdepth: 1
+
+  data_provider/index.rst
+  data_provider/pydataprovider2.rst
+
+Model Config API
+----------------
+
+..  toctree::
+  :maxdepth: 1
+
+  trainer_config_helpers/index.rst
+  trainer_config_helpers/optimizers.rst
+  trainer_config_helpers/data_sources.rst
+  trainer_config_helpers/layers.rst
+  trainer_config_helpers/activations.rst 
+  trainer_config_helpers/poolings.rst
+  trainer_config_helpers/networks.rst
+  trainer_config_helpers/evaluators.rst
+  trainer_config_helpers/attrs.rst
+
+
+Applications API
+----------------
+
+..  toctree::
+  :maxdepth: 1
+
+  predict/swig_py_paddle_en.rst

doc/api/predict/swig_py_paddle_en.rst

@@ -1,5 +1,5 @@
-Python Prediction API
+Python Prediction
-=====================
+==================
 
 PaddlePaddle offers a set of clean prediction interfaces for python with the help of
 SWIG. The main steps of predict values in python are:

doc/api/trainer_config_helpers/attrs.rst

@@ -0,0 +1,5 @@
+Parameter Attributes
+=======================
+
+..  automodule:: paddle.trainer_config_helpers.attrs
+    :members: