Skip to content

P
PaddleDetection

PaddlePaddle / PaddleDetection
大约 1 年 前同步成功

通知 694
Star 11112
Fork 2696
P
PaddleDetection

体验新版 GitCode，发现更多精彩内容 >>

提交 ba8660a1 编写于 作者: D dangqingqing
浏览文件
操作

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

上级 544ff784 fc5c432a
展开全部 隐藏空白更改
内联 并排
Showing with 4709 addition and 2649 deletion
README.md
浏览文件 @ ba8660a1
......@@ -61,32 +61,32 @@ Please refer to our [release announcement](https://github.com/PaddlePaddle/Paddl
## Installation
It is recommended to check out the
[Docker installation guide](http://doc.paddlepaddle.org/develop/doc/getstarted/build_and_install/docker_install_en.html)
[Docker installation guide](http://www.paddlepaddle.org/docs/develop/documentation/en/getstarted/build_and_install/docker_install_en.html)
before looking into the
[build from source guide](http://doc.paddlepaddle.org/develop/doc/getstarted/build_and_install/build_from_source_en.html).
[build from source guide](http://www.paddlepaddle.org/docs/develop/documentation/en/getstarted/build_and_install/build_from_source_en.html).
## Documentation
We provide [English](http://doc.paddlepaddle.org/develop/doc/) and
[Chinese](http://doc.paddlepaddle.org/doc_cn/) documentation.
We provide [English](http://www.paddlepaddle.org/docs/develop/documentation/en/getstarted/index_en.html) and
[Chinese](http://www.paddlepaddle.org/docs/develop/documentation/zh/getstarted/index_cn.html) documentation.
- [Deep Learning 101](http://book.paddlepaddle.org/index.html)
- [Deep Learning 101](http://www.paddlepaddle.org/docs/develop/book/01.fit_a_line/index.html)
You might want to start from this online interactive book that can run in a Jupyter Notebook.
- [Distributed Training](http://doc.paddlepaddle.org/develop/doc/howto/usage/cluster/cluster_train_en.html)
- [Distributed Training](http://www.paddlepaddle.org/docs/develop/documentation/en/howto/usage/cluster/cluster_train_en.html)
You can run distributed training jobs on MPI clusters.
- [Distributed Training on Kubernetes](http://doc.paddlepaddle.org/develop/doc/howto/usage/k8s/k8s_en.html)
- [Distributed Training on Kubernetes](http://www.paddlepaddle.org/docs/develop/documentation/en/howto/usage/cluster/k8s_en.html)
You can also run distributed training jobs on Kubernetes clusters.
- [Python API](http://doc.paddlepaddle.org/develop/doc/api/index_en.html)
- [Python API](http://www.paddlepaddle.org/docs/develop/documentation/en/api/index_en.html)
Our new API enables much shorter programs.
- [How to Contribute](http://doc.paddlepaddle.org/develop/doc/howto/dev/contribute_to_paddle_en.html)
- [How to Contribute](http://www.paddlepaddle.org/docs/develop/documentation/en/howto/dev/contribute_to_paddle_en.html)
We appreciate your contributions!
......
benchmark/paddle/image/alexnet.py
浏览文件 @ ba8660a1
......@@ -6,8 +6,18 @@ height = 227
width = 227
num_class = 1000
batch_size = get_config_arg('batch_size', int, 128)
gp = get_config_arg('layer_num', int, 1)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {'height': height, 'width': width, 'color': True, 'num_class': num_class}
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list", None, module="provider", obj="process", args=args)
......@@ -31,7 +41,7 @@ net = img_pool_layer(input=net, pool_size=3, stride=2)
# conv2
net = img_conv_layer(
input=net, filter_size=5, num_filters=256, stride=1, padding=2, groups=1)
input=net, filter_size=5, num_filters=256, stride=1, padding=2, groups=gp)
net = img_cmrnorm_layer(input=net, size=5, scale=0.0001, power=0.75)
net = img_pool_layer(input=net, pool_size=3, stride=2)
......@@ -40,11 +50,11 @@ net = img_conv_layer(
input=net, filter_size=3, num_filters=384, stride=1, padding=1)
# conv4
net = img_conv_layer(
input=net, filter_size=3, num_filters=384, stride=1, padding=1, groups=1)
input=net, filter_size=3, num_filters=384, stride=1, padding=1, groups=gp)
# conv5
net = img_conv_layer(
input=net, filter_size=3, num_filters=256, stride=1, padding=1, groups=1)
input=net, filter_size=3, num_filters=256, stride=1, padding=1, groups=gp)
net = img_pool_layer(input=net, pool_size=3, stride=2)
net = fc_layer(
......@@ -59,6 +69,9 @@ net = fc_layer(
layer_attr=ExtraAttr(drop_rate=0.5))
net = fc_layer(input=net, size=1000, act=SoftmaxActivation())
lab = data_layer('label', num_class)
loss = cross_entropy(input=net, label=lab)
outputs(loss)
if is_infer:
outputs(net)
else:
lab = data_layer('label', num_class)
loss = cross_entropy(input=net, label=lab)
outputs(loss)
benchmark/paddle/image/googlenet.py
浏览文件 @ ba8660a1
......@@ -7,13 +7,15 @@ num_class = 1000
batch_size = get_config_arg('batch_size', int, 128)
use_gpu = get_config_arg('use_gpu', bool, True)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list" if not is_infer else None,
......
benchmark/paddle/image/provider.py
浏览文件 @ ba8660a1
......@@ -14,6 +14,7 @@ def initHook(settings, height, width, color, num_class, **kwargs):
else:
settings.data_size = settings.height * settings.width
settings.is_infer = kwargs.get('is_infer', False)
settings.num_samples = kwargs.get('num_samples', 2560)
if settings.is_infer:
settings.slots = [dense_vector(settings.data_size)]
else:
......@@ -23,7 +24,7 @@ def initHook(settings, height, width, color, num_class, **kwargs):
@provider(
init_hook=initHook, min_pool_size=-1, cache=CacheType.CACHE_PASS_IN_MEM)
def process(settings, file_list):
for i in xrange(2560 if settings.is_infer else 1024):
for i in xrange(settings.num_samples):
img = np.random.rand(1, settings.data_size).reshape(-1, 1).flatten()
if settings.is_infer:
yield img.astype('float32')
......
benchmark/paddle/image/resnet.py
浏览文件 @ ba8660a1
......@@ -7,13 +7,15 @@ num_class = 1000
batch_size = get_config_arg('batch_size', int, 64)
layer_num = get_config_arg("layer_num", int, 50)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list" if not is_infer else None,
......
benchmark/paddle/image/run_mkldnn_infer.sh → benchmark/paddle/image/run_mkl_infer.sh
浏览文件 @ ba8660a1
......@@ -37,7 +37,7 @@ function infer() {
--trainer_count=1 \
--num_passes=1 \
--save_dir="models/${topology}-${layer_num}" \
--config_args="batch_size=128,layer_num=${layer_num}" \
--config_args="batch_size=128,layer_num=${layer_num},num_samples=256" \
> /dev/null 2>&1
echo "Done"
fi
......@@ -79,8 +79,9 @@ fi
# inference benchmark
for use_mkldnn in True False; do
for batchsize in 1 2 4 8 16; do
infer googlenet v1 $batchsize $use_mkldnn
infer resnet 50 $batchsize $use_mkldnn
infer vgg 19 $batchsize $use_mkldnn
infer resnet 50 $batchsize $use_mkldnn
infer googlenet v1 $batchsize $use_mkldnn
infer alexnet 2 $batchsize $use_mkldnn
done
done
benchmark/paddle/image/run_mkldnn_train.sh → benchmark/paddle/image/run_mkl_train.sh
浏览文件 @ ba8660a1
......@@ -28,6 +28,10 @@ function train() {
--test_period=100 \
--config_args=$args \
2>&1 | tee ${log}
avg_time=`tail ${log} -n 1 | awk -F ' ' '{print $8}' | sed 's/avg=//'`
fps=`awk 'BEGIN{printf "%.2f",('$bs' / '$avg_time' * 1000)}'`
echo "FPS: $fps images/sec" 2>&1 | tee -a ${log}
}
if [ ! -f "train.list" ]; then
......@@ -43,5 +47,6 @@ for use_mkldnn in True False; do
train vgg 19 $batchsize $use_mkldnn
train resnet 50 $batchsize $use_mkldnn
train googlenet v1 $batchsize $use_mkldnn
train alexnet 2 $batchsize $use_mkldnn
done
done
benchmark/paddle/image/run_openblas_infer.sh 0 → 100755
浏览文件 @ ba8660a1
set -e
function clock_to_seconds() {
hours=`echo $1 | awk -F ':' '{print $1}'`
mins=`echo $1 | awk -F ':' '{print $2}'`
secs=`echo $1 | awk -F ':' '{print $3}'`
echo `awk 'BEGIN{printf "%.2f",('$secs' + '$mins' * 60 + '$hours' * 3600)}'`
}
function infer() {
unset OMP_NUM_THREADS MKL_NUM_THREADS OMP_DYNAMIC KMP_AFFINITY
topology=$1
layer_num=$2
bs=$3
thread=`nproc`
if [ $thread -gt $bs ]; then
thread=$bs
fi
log="logs/infer-${topology}-${layer_num}-${thread}openblas-${bs}.log"
models_in="models/${topology}-${layer_num}/pass-00000/"
if [ ! -d $models_in ]; then
echo "./run_mkl_infer.sh to save the model first"
exit 0
fi
log_period=$((32 / bs))
paddle train --job=test \
--config="${topology}.py" \
--use_mkldnn=False \
--use_gpu=False \
--trainer_count=$thread \
--log_period=$log_period \
--config_args="batch_size=${bs},layer_num=${layer_num},is_infer=True,num_samples=256" \
--init_model_path=$models_in \
2>&1 | tee ${log}
# calculate the last 5 logs period time of 160(=32*5) samples,
# the time before are burning time.
start=`tail ${log} -n 7 | head -n 1 | awk -F ' ' '{print $2}' | xargs`
end=`tail ${log} -n 2 | head -n 1 | awk -F ' ' '{print $2}' | xargs`
start_sec=`clock_to_seconds $start`
end_sec=`clock_to_seconds $end`
fps=`awk 'BEGIN{printf "%.2f",(160 / ('$end_sec' - '$start_sec'))}'`
echo "Last 160 samples start: ${start}(${start_sec} sec), end: ${end}(${end_sec} sec;" >> ${log}
echo "FPS: $fps images/sec" 2>&1 | tee -a ${log}
}
if [ ! -f "train.list" ]; then
echo " " > train.list
fi
if [ ! -f "test.list" ]; then
echo " " > test.list
fi
if [ ! -d "logs" ]; then
mkdir logs
fi
# inference benchmark
for batchsize in 1 2 4 8 16; do
infer vgg 19 $batchsize
infer resnet 50 $batchsize
infer googlenet v1 $batchsize
infer alexnet 2 $batchsize
done
benchmark/paddle/image/run_openblas_train.sh 0 → 100755
浏览文件 @ ba8660a1
set -e
function train() {
unset OMP_NUM_THREADS MKL_NUM_THREADS OMP_DYNAMIC KMP_AFFINITY
topology=$1
layer_num=$2
bs=$3
thread=`nproc`
# each trainer_count use only 1 core to avoid conflict
log="logs/train-${topology}-${layer_num}-${thread}openblas-${bs}.log"
args="batch_size=${bs},layer_num=${layer_num}"
config="${topology}.py"
paddle train --job=time \
--config=$config \
--use_mkldnn=False \
--use_gpu=False \
--trainer_count=$thread \
--log_period=3 \
--test_period=30 \
--config_args=$args \
2>&1 | tee ${log}
avg_time=`tail ${log} -n 1 | awk -F ' ' '{print $8}' | sed 's/avg=//'`
fps=`awk 'BEGIN{printf "%.2f",('$bs' / '$avg_time' * 1000)}'`
echo "FPS: $fps images/sec" 2>&1 | tee -a ${log}
}
if [ ! -f "train.list" ]; then
echo " " > train.list
fi
if [ ! -d "logs" ]; then
mkdir logs
fi
# training benchmark
for batchsize in 64 128 256; do
train vgg 19 $batchsize
train resnet 50 $batchsize
train googlenet v1 $batchsize
train alexnet 2 $batchsize
done
benchmark/paddle/image/vgg.py
浏览文件 @ ba8660a1
......@@ -7,13 +7,15 @@ num_class = 1000
batch_size = get_config_arg('batch_size', int, 64)
layer_num = get_config_arg('layer_num', int, 19)
is_infer = get_config_arg("is_infer", bool, False)
num_samples = get_config_arg('num_samples', int, 2560)
args = {
'height': height,
'width': width,
'color': True,
'num_class': num_class,
'is_infer': is_infer
'is_infer': is_infer,
'num_samples': num_samples
}
define_py_data_sources2(
"train.list" if not is_infer else None,
......
cmake/external/protobuf.cmake
浏览文件 @ ba8660a1
......@@ -253,9 +253,9 @@ IF(NOT PROTOBUF_FOUND)
IF(WITH_C_API)
INSTALL(DIRECTORY ${PROTOBUF_INCLUDE_DIR} DESTINATION third_party/protobuf)
IF(ANDROID)
INSTALL(FILES ${PROTOBUF_LIBRARY} DESTINATION third_party/protobuf/lib/${ANDROID_ABI})
INSTALL(FILES ${PROTOBUF_LITE_LIBRARY} DESTINATION third_party/protobuf/lib/${ANDROID_ABI})
ELSE()
INSTALL(FILES ${PROTOBUF_LIBRARY} DESTINATION third_party/protobuf/lib)
INSTALL(FILES ${PROTOBUF_LITE_LIBRARY} DESTINATION third_party/protobuf/lib)
ENDIF()
ENDIF()
......
doc/api/v2/config/layer.rst
浏览文件 @ ba8660a1
......@@ -467,7 +467,7 @@ lambda_cost
:noindex:
square_error_cost
--------
-----------------
.. autoclass:: paddle.v2.layer.square_error_cost
:noindex:
......@@ -533,7 +533,7 @@ Miscs
=====
dropout
--------------
--------
.. autoclass:: paddle.v2.layer.dropout
:noindex:
......
doc/api/v2/fluid/layers.rst
浏览文件 @ ba8660a1
......@@ -19,17 +19,17 @@ dynamic_lstm
:noindex:
data
---------
----
.. autofunction:: paddle.v2.fluid.layers.data
:noindex:
mean
---------
----
.. autofunction:: paddle.v2.fluid.layers.mean
:noindex:
mul
---------
---
.. autofunction:: paddle.v2.fluid.layers.mul
:noindex:
......@@ -45,13 +45,13 @@ elementwise_div
dropout
---------
-------
.. autofunction:: paddle.v2.fluid.layers.dropout
:noindex:
reshape
---------
--------
.. autofunction:: paddle.v2.fluid.layers.reshape
:noindex:
......@@ -81,67 +81,67 @@ transpose
sigmoid_cross_entropy_with_logits
---------
---------------------------------
.. autofunction:: paddle.v2.fluid.layers.esigmoid_cross_entropy_with_logits
:noindex:
cast
---------
----
.. autofunction:: paddle.v2.fluid.layers.cast
:noindex:
concat
---------
-------
.. autofunction:: paddle.v2.fluid.layers.concat
:noindex:
sums
---------
----
.. autofunction:: paddle.v2.fluid.layers.sums
:noindex:
linear_chain_crf
---------
----------------
.. autofunction:: paddle.v2.fluid.layers.linear_chain_crf
:noindex:
assign
---------
-------
.. autofunction:: paddle.v2.fluid.layers.embedding
:noindex:
split_lod_tensor
---------
----------------
.. autofunction:: paddle.v2.fluid.layers.split_lod_tensor
:noindex:
merge_lod_tensor
---------
----------------
.. autofunction:: paddle.v2.fluid.layers.merge_lod_tensor
:noindex:
cos_sim
---------
--------
.. autofunction:: paddle.v2.fluid.layers.cos_sim
:noindex:
cross_entropy
---------
-------------
.. autofunction:: paddle.v2.fluid.layers.cross_entropy
:noindex:
square_error_cost
---------
-----------------
.. autofunction:: paddle.v2.fluid.layers.square_error_cost
:noindex:
......@@ -153,74 +153,80 @@ accuracy
sequence_conv
---------
-------------
.. autofunction:: paddle.v2.fluid.layers.sequence_conv
:noindex:
conv2d
---------
------
.. autofunction:: paddle.v2.fluid.layers.conv2d
:noindex:
sequence_pool
---------
-------------
.. autofunction:: paddle.v2.fluid.layers.sequence_pool
:noindex:
sequence_first_step
-------------------
.. autofunction:: paddle.v2.fluid.layers.sequence_first_step
:noindex:
sequence_last_step
------------------
.. autofunction:: paddle.v2.fluid.layers.sequence_last_step
:noindex:
pool2d
---------
------
.. autofunction:: paddle.v2.fluid.layers.pool2d
:noindex:
batch_norm
---------
----------
.. autofunction:: paddle.v2.fluid.layers.batch_norm
:noindex:
beam_search_decode
---------
------------------
.. autofunction:: paddle.v2.fluid.layers.beam_search_decode
:noindex:
lstm
---------
.. autofunction:: paddle.v2.fluid.layers.lstm
:noindex:
lod_rank_table
---------
--------------
.. autofunction:: paddle.v2.fluid.layers.lod_rank_table
:noindex:
max_sequence_len
---------
----------------
.. autofunction:: paddle.v2.fluid.layers.max_sequence_len
:noindex:
topk
---------
-----
.. autofunction:: paddle.v2.fluid.layers.topk
:noindex:
lod_tensor_to_array
---------
-------------------
.. autofunction:: paddle.v2.fluid.layers.lod_tensor_to_array
:noindex:
array_to_lod_tensor
---------
-------------------
.. autofunction:: paddle.v2.fluid.layers.array_to_lod_tensor
:noindex:
......@@ -228,26 +234,26 @@ array_to_lod_tensor
fill_constant
---------
-------------
.. autofunction:: paddle.v2.fluid.layers.fill_constant
:noindex:
fill_constant_batch_size_like
---------
-----------------------------
.. autofunction:: paddle.v2.fluid.layers.fill_constant_batch_size_like
:noindex:
ones
---------
----
.. autofunction:: paddle.v2.fluid.layers.ones
:noindex:
zeros
---------
-----
.. autofunction:: paddle.v2.fluid.layers.zeros
:noindex:
......@@ -259,14 +265,14 @@ increment
array_write
---------
-----------
.. autofunction:: paddle.v2.fluid.layers.array_write
:noindex:
create_array
---------
------------
.. autofunction:: paddle.v2.fluid.layers.create_array
:noindex:
......@@ -278,25 +284,55 @@ less_than
array_read
---------
----------
.. autofunction:: paddle.v2.fluid.layers.array_read
:noindex:
shrink_memory
---------
--------------
.. autofunction:: paddle.v2.fluid.layers.shrink_memory
:noindex:
array_length
---------
-------------
.. autofunction:: paddle.v2.fluid.layers.array_length
:noindex:
conv2d_transpose
---------
----------------
.. autofunction:: paddle.v2.fluid.layers.conv2d_transpose
:noindex:
sequence_expand
---------------
.. autofunction:: paddle.v2.fluid.layers.sequence_expand
:noindex:
lstm_unit
---------
.. autofunction:: paddle.v2.fluid.layers.lstm_unit
:noindex:
sequence_softmax
----------------
.. autofunction:: paddle.v2.fluid.layers.sequence_softmax
:noindex:
reduce_sum
----------
.. autofunction:: paddle.v2.fluid.layers.reduce_sum
:noindex:
reduce_mean
---------
.. autofunction:: paddle.v2.fluid.layers.reduce_mean
:noindex:
doc/api/v2/fluid/nets.rst
浏览文件 @ ba8660a1
......@@ -3,19 +3,19 @@ Nets
===========
simple_img_conv_pool
-----------
--------------------
.. autofunction:: paddle.v2.fluid.nets.simple_img_conv_pool
:noindex:
img_conv_group
-----------
---------------
.. autofunction:: paddle.v2.fluid.nets.img_conv_group
:noindex:
sequence_conv_pool
-----------
------------------
.. autofunction:: paddle.v2.fluid.nets.sequence_conv_pool
:noindex:
......
doc/api/v2/fluid/optimizer.rst
浏览文件 @ ba8660a1
......@@ -18,7 +18,7 @@ SGDOptimizer
MomentumOptimizer
-----------
-----------------
.. automodule:: paddle.v2.fluid.optimizer
:members: MomentumOptimizer
:noindex:
......@@ -26,14 +26,14 @@ MomentumOptimizer
AdagradOptimizer
-----------
----------------
.. automodule:: paddle.v2.fluid.optimizer
:members: AdagradOptimizer
:noindex:
AdamOptimizer
-----------
-------------
.. automodule:: paddle.v2.fluid.optimizer
:members: AdamOptimizer
:noindex:
......@@ -47,7 +47,7 @@ AdamaxOptimizer
DecayedAdagradOptimizer
-----------
-----------------------
.. automodule:: paddle.v2.fluid.optimizer
:members: DecayedAdagradOptimizer
:noindex:
......
doc/api/v2/fluid/regularizer.rst
浏览文件 @ ba8660a1
......@@ -3,14 +3,14 @@ Regularizer
===========
WeightDecayRegularizer
-----------
----------------------
.. automodule:: paddle.v2.fluid.regularizer
:members: WeightDecayRegularizer
:noindex:
L2DecayRegularizer
-----------
------------------
.. automodule:: paddle.v2.fluid.regularizer
:members: L2DecayRegularizer
:noindex:
......@@ -18,7 +18,7 @@ L2DecayRegularizer
L1DecayRegularizer
-----------
-------------------
.. automodule:: paddle.v2.fluid.regularizer
:members: L1DecayRegularizer
......
doc/design/executor.md
浏览文件 @ ba8660a1
# Executor Design Doc
## Motivation
In [fluid](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/fluid.md), we encourage the user to use deep learning programming paradigms to describe the training process. When the user-written Python program is executed, it will first create a protobuf message
[`ProgramDesc`](https://github.com/PaddlePaddle/Paddle/blob/a91efdde6910ce92a78e3aa7157412c4c88d9ee8/paddle/framework/framework.proto#L145) that describes the process and is conceptually like an [abstract syntax tree](https://en.wikipedia.org/wiki/Abstract_syntax_tree).
We use executor to do the runtime evaluation of a `ProgramDesc`.
The executor runs the `ProgramDesc` like an interpreter. `ProgramDesc` contains the intrinsics (operators in this case) and variables which will be used, executor explicitly executes the stored precompiled code.
## 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.
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 in the block. The `block_id` specifies the entrance block. And the `Scope` is the container of all the variable instances, which is persistent throughout different runs.
### What does executor do?
## Executor
It evaluates all the operators in the `block_id`th block of a `ProgramDesc`.
The `Executor` explicitly executes all the intrinsics (operators here) in the `block_id`th block of a `ProgramDesc`. Essentially, it instantiates Variables and Operators, then runs all the operators in sequence one-by-one.
It is very similar to how a push stack frame works when entering a block, following which it cleans up all the temporary variables when a mini-batch is finished. It does not however, have the stack frame pop process.
### What does executor NOT do?
### The interface
```c++
Executor(places);
```
A executor does not own any computing resources, a user can only construct an executor using the specified places.
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.
### Running an Executor
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)
```
void Run(ProgramDesc, Scope, block_id, create_local_scope);
```
An `Executor` only provides a unified way to execute `ProgramDesc`. `ProgramDesc` is the target that will be executed, the `Scope` specifies the variable container, the `block_id` indicates the entrance block and `create_local_scope` is a boolean that states whether it will destroy the temporary variables after the execution is finished.
doc/design/kernel_hint_design.md 0 → 100644
浏览文件 @ ba8660a1
## Problem
In PaddlePaddle's [Design](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/switch_kernel.md), one Operator may have multiple kernels. Users may have some personal preference to choose a certain type of kernel for an operator, such as `force_cpu` to choose a CPU kernel, `use_cudnn` to choose a CUDNN kernel, we need to provide a way for users to do this.
In the current design, we use KernelType to describe one kernel.
```cpp
struct KernelType {
Place place_;
DataType data_type_;
LayoutType layout_;
};
```
`place_` `data_type_` and `layout_` can be got from the input tensors of the operator, `GetActualKernelType(inputs)` use inputs to infer the proper kernel key that fit the incoming data, but users can not directly configure it.
The [design](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/switch_kernel.md) also provides a virtual method `GetExpectedKernelType` that user can overload and use to choose the KernelType they want to use.
So we should send the information user defined in proto to `GetExpectedKernelType` for choosing a kernel.
The problem is, how should we define and send the information for `GetExpectedKernelType` to use?
## Solution
### Potential choice
1. Do nothing, let the user add the information they want to operator‘s attribute and get them inside `GetExpectedKernelType`, this can work properly. But there is a little problem that users may define many kinds of hints for the same purpose, such as `force_cpu`, `use_cpu`, `cpu_kernel` to choose CPU kernel, and `use_cudnn`, `force_cudnn`, `cudnn_kernel` to choose CUDNN kernel.
2. Pre-define all the needed option and use a single attr key such as `kernel_hint` for the user, this is not so flexible if the user wants to define some more kind of hint.
### Final choice
To provide enough flexibility while avoiding confusion definition, we can define some global constants for these attribute names, such as `force_cpu`, `use_cudnn`, `use_mkldnn` for a user to choose.
In C++
```cpp
const std::string kForceCPU = "force_cpu";
const std::string kUseCUDNN = "use_cudnn";
const std::string kUseMKLDNN = "use_mkldnn";
KernelType GetExpectedKernelType() {
if (Attr<bool>(kForceCPU)) {
return KernelType(CPUPlace, ...)
} else {
...
}
}
```
In Python code
```python
FORCE_CPU = core.kForceCPU()
def xx_layer(..., force_cpu=false):
layer_helper = LayerHelper(...)
layer_helper.append_op(
type="xx",
attr={FORCE_CPU: force_cpu})
```
doc/design/mkl/mkl_packed.md
浏览文件 @ ba8660a1
......@@ -30,10 +30,10 @@
由于在现有的某些情况下（例如RNN），多次调用 cblas_?gemm 会使用相同的原数据，因此，每次调用时对原数据的重复Packing便成为了冗余。
为了最大程度减少多次调用 cblas_?gemm 在Packing上的耗时，Intel® MKL 引入了以下四个API:
* cblas_?gemm_alloc
* cblas_?gemm_pack
* cblas_?gemm_compute
* cblas_?gemm_free
* [cblas_?gemm_alloc](https://software.intel.com/en-us/mkl-developer-reference-c-cblas-gemm-alloc)
* [cblas_?gemm_pack](https://software.intel.com/en-us/mkl-developer-reference-c-cblas-gemm-pack)
* [cblas_?gemm_compute](https://software.intel.com/en-us/mkl-developer-reference-c-cblas-gemm-compute)
* [cblas_?gemm_free](https://software.intel.com/en-us/mkl-developer-reference-c-cblas-gemm-free)
通过使用这些API，我们可以先完成对原数据的Packing操作，再把已转换为Packed格式的数据传递给那些复用同一数据的gemm_compute函数，从而避免了Packing冗余。
......@@ -84,7 +84,20 @@ PaddlePaddle/Paddle
2. 对比优化后layer与相对应的PaddlePaddle原有layer, 在batch mode下的结果。
### Python API
TBD
计划在`paddle/utils.Flags`中添加`use_mkl_packed`的flag，用于选择是否使用相关功能，并且当编译时`WITH_MKL=ON`的情况下，默认设置为`true`。
同时，在`python/paddle/trainer/config_parser.py`中对应的layer处，添加`use_mkl_packed`这个选择，方便用户在Python端选择是否启用这个功能。
具体实现方式比如：
```python
use_mkl_packed = bool(int(g_command_config_args.get("use_mkl_packed", 0)))
if use_mkl_packed:
self.layer_type = mkl_packed_*
```
所有相关的`layer_type`会以*mkl_packed_*开头，这些会在`MKLPacked*Layer`注册layer的时候保证，以示区分。
### Benchmarking
会添加相应的脚本用于测试和对比在使用MKL Packed recurrent layers 前后的网络性能。
......
doc/design/operator_kernel_type.md 0 → 100644
浏览文件 @ ba8660a1
# Design Doc: The Keys of Operator Kernel Type
## Problem
An operator can have different kernel implementations, and each operator will have a map to store the related kernels. Fluid uses `OpKernelType` as a key to identify a unique Kernel. Before an operator runs, an certain kernel must be chosen by a key of `OpKernelType`. Currently, `OpKernelType` is defined as follows:
```cpp
struct OpKernelType {
platform::Place place_;
proto::DataType data_type_;
};
```
For more details, please refer to [codes](https://github.com/PaddlePaddle/Paddle/blob/2d5ec16bc8a09fb8e0f62c89b116b0cd1d333907/paddle/framework/operator.h#L348-L374) in github.
It contains two keys, `Place` and `DataType`. And these two keys will be hashed to a unique key to represent a certain type of kernel. However, these two keys are not enough. We need a more complete representation of `OpKernelType`.
We often implement a kernel of an operator with some computing library in certain device(place). Please remind that computing library and device are not one-to-one corresponding. A device can have a lot of computing libraries and a computing library can also support several devices.
For example, Eigen library can support Nvidia GPU/AMD GPU/CPU. And MKLDNN library can support Intel CPU/Intel FPGA. Both `Place` and `Library` should be a key of `OpKernelType`.
It's obvious that different DataTypes, like fp64/fp32/int8 will have different kernels. But the data layout of a Tensor will also lead to different implementation. Please refer to the batch norm operator [kernels](https://github.com/PaddlePaddle/Paddle/blob/a948fac4d0ad7e0412d373b8aabeb711c2899563/paddle/operators/batch_norm_op.cc#L180-L209). Data Layout should also be taken into consideration.
## Solution
There are four keys to determine a kernel type of an operator: `Place`/`Library`/`DataType`/`Layout`.
```cpp
struct OpKernelType {
platform::Place place_;
platform::Library library_;
proto::DataType data_type_;
framework::Layout layout_;
};
```
Following is the details:
### Place
`Place` is defined as follows:
```cpp
typedef boost::variant<CUDAPlace, ROCmPlace, FPGAPlace, CPUPlace> Place;
```
`Place` is to represent the device memory where data is locating.
### Library
One operator kernel is usually implemented based on one library. `Library` is defined as a enum variable:
```cpp
enum Library { Plain, MKLDNN, CUDNN };
```
We use `Plain` enumerator to represent default library. Since most operators in Fluid are implemented based on `Eigen` library, we take `Eigen` library as the `Plain` enumerator.
A library usually has a corresponding `DeviceContext` which contains some handles needed by computation. Fluid now have two default DeviceContexts in CPU and CUDA, `CPUDeviceContext` and `CUDADeviceContext`. `CPUDeviceContext` contains a Eigen library handle and `CDUADeviceContext` contains a Eigen library handle and cuBLAS handle.
If we want to support new Library, a new enumerator need to be added to `Library` and a new corresponding `LibraryDeviceContext` will be created.
### DataType
`DataType` is defined in [framework.proto](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/framework.proto). Currently, int32/int64/fp32/fp64 are supported.
### Layout
Actually, a Tensor is a view of a block of memory. Besides a pointer to the memory, we also have to get some other descriptions of this block of memory, such as shape(ddim), stride, and layout.
Different layout leads to different implementation of operator kernel. There are mainly 4 principles we have to follow to support layout in our fluid framework.
- We take layout as a data member of Tensor. Layout is actually a enum variable. If fluid is built with MKLDNN, then, the memory format in MKLDNN will be added into this enum variable too.
- Users have to set layout for input data. And some operators like fill_constant/random, also have to set layout of generating data. Of course, we can have some default layout, like NCHW.
- The inference of Layout is at run-time, not compile-time.
- Every operator have to implement different kernels for different layouts. Let's take MKLDNN as an example, if we want to implement a MKLDNN convolution operator, we have to realize all the kernels for different layout, list at [here](http://01org.github.io/mkl-dnn/structmkldnn_1_1memory.html). And we will have a special macro to do registering kernels for MKLDNN operators.
`Layout` is also defined as a enum variable:
```cpp
enum Layout {
kNCHW,
kNHWC,
#ifdef PADDLE_WITH_MKLDNN
knChw8c
...
#endif
};
```
doc/design/refactor/multi_cpu.md 0 → 100644
浏览文件 @ ba8660a1
# Design Doc: Execute the Program with Multi CPU
## Abstract
This Design Doc propose an approach to make the user-defined Op graph
running with multi-CPU, we will use an auto transpiler to convert the user-defined
Op graph to a multi-CPU Op graph, and run `ParallelDo` Op to run the graph.
## Transpiler
<img src="src/multi-threads/single-thread@3x.png" width="300">
After converted:
<img src="src/multi-threads/multi-threads@3x.png" width="1000">
## Implement
- `Multi-CPU Transpiler` will convert the graph to a multi-CPU graph
which would be executed with multi-threads.
- `BlockingCounter` will `Init/Decrement` an atomic counter, and Blocking `Wait`
for the atomic counter become `0`:
```cpp
BlockingCounter bc(thread_count);
for (int i = 0; i < thread_count; ++i) {
thread_pool->Start([&bc] {bc.DecrementCount(); })
}
bc.Wait();
```
- `ParallelDo` Operator
- Initialize a thread pool which is a Singleton.
- Use a block id as the input, and create run the specify Block on independent scope
with multi-threads.
- Initialize a `BlockingCounter` instance and wait until all threads are done.
- `Split` Operator will split the Input Tensor into a TensorArray.
- `Merge` merge all the gradients which calculated in different threads
with `mean/sum/max/min...` method, and then run the Optimizer Op to optimize `W`.
## TODO
- Improve the optimizer stage with multi-threads, since we could
assign the parameters to the different threads and execute
optimizer with multi-threads.
doc/design/switch_kernel.md 0 → 100644
浏览文件 @ ba8660a1
## Background
Every operator has many kernels because there are multiple data types, places, data layout that Fluid supports. We use the `KernelType` to describe kernel types that operators can hold.
The `KernelType` is as follows.
```
struct KernelType {
Place place_;
DataType data_type_;
LayoutType layout_;
};
```
The `place_` is a descriptor of the device and the computational library, e.g., `MKLDNNPlace`, `CUDAPlace`.
The `data_type_` is the data type that this kernel performs on, e.g., `FP32`, `INT64`. Note that one kernel may have inputs with different data types. However, it will be a major `data_type`. For example, the `cross_entropy` takes `int64` as it label, and `double`/`float` as its input logit and output cost. The major `data_type` of `cross_entropy` is `float`/`double`.
The `layout` is useful for some computational library. One example is that MKLDNN uses many kinds of layout, such as `nChw8c`. Each kind of layout will invoke the different kernel.
## Problem
We register a kernel for every operator and every kernel type ideally. However, it is impracticable for the following situations.
1. Some operators, like CRF, are complicated and inefficient to be implemented on GPU. The CRF operator will only have a CPU kernel.
2. Some operators will take too many memory. It is better to force them into CPU. However, the rest of operators in this neural network will be performed on GPU, i.e., model parallel problem.
3. Some layout and place are particular. One example is that MKLDNN uses `nChw8` and there is no other library uses `nChw8c`.
Problems under these situations are similar. We can formalise this problem as follow.
We register kernels with types $KT = \{kt_1, kt_2, kt_3, ...\}$ for one operator. The inputs of this operator should be run on kernel type $kt_{?}$, which the $kt_{?} \notin KT$. How to cast the input of this operator from $kt_{?}$ to any of kernel type in $KT$.
## Solution
It is clearly that transforming inputs of an operator toadapt another kernel type is not related to the particular operator. So we should register these transformation methods as global methods.
We can infer a kernel type from the inputs of an operators. We let this kernel type as `actual kernel type`, which means this kernel type is the actually kernel type that operator should be performed.
We can get a kernel type by 1) The configuration of operator description. (Users may want to force use `MKL` for `conv` operator). 2) The place of the current executor. (Executor is running on GPU). This kernel type is what we expect the operator will be performed on. We let this kernel type as `expect kernel type`.
We transform the input data from `actual` to `expect` if the expect kernel type is not as same as actual kernel type.
The algorithm is described as follow
```cpp
using DataTransformationFN = std::function<void(const Tensor& in, Tensor* out)>;
using KernelTypePair = std::pair<KernelType, KernelType>;
map<KernelTypePair, DataTransformationFN> g_data_transformation_;
void OpWithKernel::Run() {
vec<Tensor> inputs = ...
auto actual_kernel_type = GetActualKernelType(inputs);
// The expected kernel type is related to actual kernel type.
// For the most operators, the expected kernel type is as same as
// actual kernel type.
//
// So we pass `actual_kernel_type` as a parameter of
// GetExpectedKernelType
auto expect_kernel_type = GetExpectedKernelType(actual_kernel_type);
auto trans = g_data_transformation_[{actual_kernel_type, expect_kernel_type}];
kernel.run(trans(inputs));
}
```
doc/getstarted/build_and_install/build_from_source_cn.rst
浏览文件 @ ba8660a1
......@@ -70,13 +70,13 @@ PaddlePaddle编译需要使用到下面的依赖（包含但不限于），其
:header: "依赖", "版本", "说明"
:widths: 10, 15, 30
"CMake", ">=3.5", ""
"CMake", ">=3.2", ""
"GCC", "4.8.2", "推荐使用CentOS的devtools2"
"Python", "2.7.x", "依赖libpython2.7.so"
"pip", ">=9.0", ""
"numpy", "", ""
"Python", "2.7.x", "依赖libpython2.7.so"
"pip", ">=9.0", ""
"numpy", "", ""
"SWIG", ">=2.0", ""
"Go", ">=1.8", "可选"
"Go", ">=1.8", "可选"
.. _build_options:
......
doc/getstarted/build_and_install/build_from_source_en.rst
浏览文件 @ ba8660a1
......@@ -76,13 +76,13 @@ will be downloaded automatically.
:header: "Dependency", "Version", "Description"
:widths: 10, 15, 30
"CMake", ">=3.5", ""
"CMake", ">=3.2", ""
"GCC", "4.8.2", "Recommend devtools2 for CentOS"
"Python", "2.7.x", "Need libpython2.7.so"
"pip", ">=9.0", ""
"numpy", "", ""
"Python", "2.7.x", "Need libpython2.7.so"
"pip", ">=9.0", ""
"numpy", "", ""
"SWIG", ">=2.0", ""
"Go", ">=1.8", "Optional"
"Go", ">=1.8", "Optional"
.. _build_options:
......
doc/getstarted/build_and_install/docker_install_cn.rst
浏览文件 @ ba8660a1
......@@ -128,7 +128,7 @@ PaddlePaddle Book是为用户和开发者制作的一个交互式的Jupyter Note
AVX是一种CPU指令集，可以加速PaddlePaddle的计算。最新的PaddlePaddle Docker镜像默认
是开启AVX编译的，所以，如果您的电脑不支持AVX，需要单独
`编译 <./build_from_source_cn.rst>`_ PaddlePaddle为no-avx版本。
`编译 <./build_from_source_cn.html>`_ PaddlePaddle为no-avx版本。
以下指令能检查Linux电脑是否支持AVX：
......
doc/getstarted/build_and_install/docker_install_en.rst
浏览文件 @ ba8660a1
......@@ -137,7 +137,7 @@ GPU driver installed before move on.
AVX is a kind of CPU instruction can accelerate PaddlePaddle's calculations.
The latest PaddlePaddle Docker image turns AVX on by default, so, if your
computer doesn't support AVX, you'll probably need to
`build <./build_from_source_en.rst>`_ with :code:`WITH_AVX=OFF`.
`build <./build_from_source_en.html>`_ with :code:`WITH_AVX=OFF`.
The following command will tell you whether your computer supports AVX.
......
doc/getstarted/build_and_install/pip_install_cn.rst
浏览文件 @ ba8660a1
......@@ -37,11 +37,11 @@ PaddlePaddle可以使用常用的Python包管理工具
:header: "版本说明", "cp27-cp27mu", "cp27-cp27m", "C-API"
:widths: 1, 3, 3, 3
"cpu_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cpu_avx_openblas", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "暂无"
"cuda7.5_cudnn5_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn5_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn7_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cpu_avx_mkl", "`paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cpu_avx_openblas", "`paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "暂无"
"cuda7.5_cudnn5_avx_mkl", "`paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn5_avx_mkl", "`paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn7_avx_mkl", "`paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
.. _pip_dependency:
......
doc/getstarted/build_and_install/pip_install_en.rst
浏览文件 @ ba8660a1
......@@ -40,11 +40,11 @@ If the links below shows up the login form, just click "Log in as guest" to star
:header: "version", "cp27-cp27mu", "cp27-cp27m", "C-API"
:widths: 1, 3, 3, 3
"cpu_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cpu_avx_openblas", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "Not Available"
"cuda7.5_cudnn5_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn5_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn7_avx_mkl", "`paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle-0.10.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <http://guest@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cpu_avx_mkl", "`paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxCp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cpu_avx_openblas", "`paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_CpuAvxOpenblas/.lastSuccessful/paddlepaddle-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "Not Available"
"cuda7.5_cudnn5_avx_mkl", "`paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda75cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn5_avx_mkl", "`paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda80cudnn5cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
"cuda8.0_cudnn7_avx_mkl", "`paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27mu-linux_x86_64.whl>`_", "`paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddlepaddle_gpu-0.11.0-cp27-cp27m-linux_x86_64.whl>`_", "`paddle.tgz <https://guest:@paddleci.ngrok.io/repository/download/Manylinux1_Cuda8cudnn7cp27cp27mu/.lastSuccessful/paddle.tgz>`_"
.. _pip_dependency:
......
doc/howto/dev/new_op_cn.md
浏览文件 @ ba8660a1
......@@ -53,7 +53,7 @@ Kernel实现 | CPU、CUDA共享Kernel实现在`.h`文件中，否则，CPU
```cpp
class MulOpMaker : public framework::OpProtoAndCheckerMaker {
public:
MulOpMaker(framework::OpProto *proto, framework::OpAttrChecker *op_checker)
MulOpMaker(OpProto *proto, OpAttrChecker *op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("X", "(Tensor), 2D tensor of size (M x K)");
AddInput("Y", "(Tensor), 2D tensor of size (K x N)");
......@@ -82,7 +82,7 @@ The equation is: Out = X * Y
template <typename AttrType>
class ScaleOpMaker : public framework::OpProtoAndCheckerMaker {
public:
ScaleOpMaker(framework::OpProto *proto, framework::OpAttrChecker *op_checker)
ScaleOpMaker(OpProto *proto, OpAttrChecker *op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("X", "The input tensor of scale operator.").NotInGradient();
AddOutput("Out", "The output tensor of scale operator.").NotInGradient();
......
doc/howto/dev/new_op_en.md
浏览文件 @ ba8660a1
......@@ -50,7 +50,7 @@ First, define `ProtoMaker` to describe the Operator's input, output, and additio
```cpp
class MulOpMaker : public framework::OpProtoAndCheckerMaker {
public:
MulOpMaker(framework::OpProto *proto, framework::OpAttrChecker *op_checker)
MulOpMaker(OpProto *proto, OpAttrChecker *op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("X", "(Tensor), 2D tensor of size (M x K)");
AddInput("Y", "(Tensor), 2D tensor of size (K x N)");
......@@ -79,7 +79,7 @@ An additional example [`ScaleOp`](https://github.com/PaddlePaddle/Paddle/blob/de
template <typename AttrType>
class ScaleOpMaker : public framework::OpProtoAndCheckerMaker {
public:
ScaleOpMaker(framework::OpProto *proto, framework::OpAttrChecker *op_checker)
ScaleOpMaker(OpProto *proto, OpAttrChecker *op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("X", "The input tensor of scale operator.").NotInGradient();
AddOutput("Out", "The output tensor of scale operator.").NotInGradient();
......
doc/howto/index_cn.rst
浏览文件 @ ba8660a1
......@@ -9,9 +9,6 @@
usage/cmd_parameter/index_cn.rst
usage/cluster/cluster_train_cn.md
usage/k8s/k8s_basis_cn.md
usage/k8s/k8s_cn.md
usage/k8s/k8s_distributed_cn.md
开发标准
--------
......
doc/howto/index_en.rst
浏览文件 @ ba8660a1
......@@ -9,8 +9,6 @@ Usage
usage/cmd_parameter/index_en.rst
usage/cluster/cluster_train_en.md
usage/k8s/k8s_en.md
usage/k8s/k8s_aws_en.md
Development
------------
......
doc/howto/usage/cluster/cluster_train_cn.md
浏览文件 @ ba8660a1
# PaddlePaddle分布式训练
* [概述](#概述)
* [环境准备](#环境准备)
* [启动参数说明](#启动参数说明)
* [启动参数服务器](#启动参数服务器)
* [启动计算节点](#启动计算节点)
* [准备数据集](#准备数据集)
* [准备训练程序](#准备训练程序)
* [使用分布式计算平台或工具](#使用分布式计算平台或工具)
* [使用Fabric启动集群作业](#使用fabric启动集群作业)
* [准备一个Linux集群](#准备一个linux集群)
* [启动集群作业](#启动集群作业)
* [终止集群作业](#终止集群作业)
* [检查集群训练结果](#检查集群训练结果)
* [检查模型输出](#检查模型输出)
* [在OpenMPI集群中提交训练作业](#在openmpi集群中提交训练作业)
* [准备OpenMPI集群](#准备OpenMPI集群)
* [启动集群作业](#启动集群作业-1)
* [在Kubernetes集群中提交训练作业](#在kubernetes集群中提交训练作业)
# 分布式训练
## 概述
本文将介绍如何使用PaddlePaddle在不同的集群框架下完成分布式训练。分布式训练架构如下图所示：
<img src="https://user-images.githubusercontent.com/13348433/31772175-5f419eca-b511-11e7-9db7-5231fe3d9ccb.png" width="500">
......@@ -32,10 +15,11 @@
在使用同步SGD训练神经网络时，PaddlePaddle使用同步屏障（barrier），使梯度的提交和参数的更新按照顺序方式执行。在异步SGD中，则并不会等待所有trainer提交梯度才更新参数，这样极大地提高了计算的并行性：参数服务器之间不相互依赖，并行地接收梯度和更新参数，参数服务器也不会等待计算节点全部都提交梯度之后才开始下一步，计算节点之间也不会相互依赖，并行地执行模型的训练。可以看出，虽然异步SGD方式会提高参数更新并行度, 但是并不能保证参数同步更新，在任意时间某一台参数服务器上保存的参数可能比另一台要更新，与同步SGD相比，梯度会有噪声。
## 环境准备
1. 准备您的计算集群。计算集群通常由一组（几台到几千台规模）的Linux服务器组成。服务器之间可以通过局域网（LAN）联通，每台服务器具有集群中唯一的IP地址（或者可被DNS解析的主机名）。集群中的每台计算机通常被成为一个“节点”。
1. 我们需要在集群的所有节点上安装 PaddlePaddle。 如果要启用GPU，还需要在节点上安装对应的GPU驱动以及CUDA。PaddlePaddle的安装可以参考[build_and_install](https://github.com/PaddlePaddle/Paddle/tree/develop/doc/getstarted/build_and_install)的多种安装方式。我们推荐使用[Docker](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/getstarted/build_and_install/docker_install_cn.rst)安装方式来快速安装PaddlePaddle。
1. 我们需要在集群的所有节点上安装 PaddlePaddle。 如果要启用GPU，还需要在节点上安装对应的GPU驱动以及CUDA。PaddlePaddle的安装可以参考[build_and_install](http://www.paddlepaddle.org/docs/develop/documentation/zh/getstarted/build_and_install/index_cn.html)的多种安装方式。我们推荐使用[Docker](http://www.paddlepaddle.org/docs/develop/documentation/zh/getstarted/build_and_install/docker_install_cn.html)安装方式来快速安装PaddlePaddle。
安装完成之后，执行下面的命令可以查看已经安装的版本（docker安装方式可以进入docker容器执行：`docker run -it paddlepaddle/paddle:[tag] /bin/bash`）：
```bash
......@@ -63,12 +47,12 @@ $ paddle pserver --port=7164 --ports_num=1 --ports_num_for_sparse=1 --num_gradie
$ stdbuf -oL /usr/bin/nohup paddle pserver --port=7164 --ports_num=1 --ports_num_for_sparse=1 --num_gradient_servers=1 &> pserver.log
```
| 参数 | 是否必选 | 默认值 | 说明 |
| ------------- | ------------- | ------------- | ------------- |
| port | 必选 | 7164 | pserver监听的起始端口，根据ports_num决定<br>总端口个数，从起始端口监听多个端口用于通信 |
| ports_num | 必选 | 1 | 监听的端口个数 |
| ports_num_for_sparse | 必选 | 1 | 用于稀疏类型参数通信的端口个数 |
| num_gradient_servers | 必选 | 1 | 当前训练任务pserver总数 |
参数说明
- port：**必选，默认7164**，pserver监听的起始端口，根据ports_num决定总端口个数，从起始端口监听多个端口用于通信
- ports_num：**必选，默认1**，监听的端口个数
- ports_num_for_sparse：**必选，默认1**，用于稀疏类型参数通信的端口个数
- num_gradient_servers：**必选，默认1**，当前训练任务pserver总数
### 启动计算节点
执行以下命令启动使用python编写的trainer程序（文件名为任意文件名，如train.py）
......@@ -105,16 +89,16 @@ paddle.init(
pservers="127.0.0.1")
```
| 参数 | 是否必选 | 默认 | 说明 |
| ------------- | ------------- | ------------- | ------------- |
| use_gpu | 可选 | False | 是否启用GPU训练 |
| trainer_count | 必选 | 1 | 当前训练任务trainer总个数 |
| port | 必选 | 7164 | 连接到pserver的端口 |
| ports_num | 必选 | 1 | 连接到pserver的端口个数 |
| ports_num_for_sparse | 必选 | 1 | 和pserver之间用于稀疏类型参数通信的端口个数 |
| num_gradient_servers | 必选 | 1 | 当前训练任务pserver总数 |
| trainer_id | 必选 | 0 | 每个trainer的唯一ID，从0开始的整数 |
| pservers | 必选 | 127.0.0.1 | 当前训练任务启动的pserver的IP列表，多个IP使用“,”隔开 |
参数说明
- use_gpu： **可选，默认False**，是否启用GPU训练
- trainer_count：**必选，默认1**，当前训练任务trainer总个数
- port：**必选，默认7164**，连接到pserver的端口
- ports_num：**必选，默认1**，连接到pserver的端口个数
- ports_num_for_sparse：**必选，默认1**，和pserver之间用于稀疏类型参数通信的端口个数
- num_gradient_servers：**必选，默认1**，当前训练任务pserver总数
- trainer_id：**必选，默认0**，每个trainer的唯一ID，从0开始的整数
- pservers：**必选，默认127.0.0.1**，当前训练任务启动的pserver的IP列表，多个IP使用“,”隔开
### 准备数据集
......@@ -171,7 +155,7 @@ test.txt-00002
- `my_lib.py`：会被`train.py`调用的一些用户定义的库函数，比如PIL库等。
- `word_dict.pickle`：在`train.py`中会使用到的字典数据文件。
- `train.py`：训练程序，代码参考[api_train_v2_cluster.py](https://github.com/PaddlePaddle/Paddle/tree/develop/doc/howto/usage/cluster/src/word2vec/prepare.py)。***注意：*** 对于本样例代码，在使用不同的分布式计算平台时，您可能需要修改`train.py`开头的部分（如下），以便获得训练数据的位置和获取环境变量配置：
- `train.py`：训练程序，代码参考[api_train_v2_cluster.py](https://github.com/PaddlePaddle/Paddle/tree/develop/doc/howto/usage/cluster/src/word2vec/api_train_v2_cluster.py)。***注意：*** 对于本样例代码，在使用不同的分布式计算平台时，您可能需要修改`train.py`开头的部分（如下），以便获得训练数据的位置和获取环境变量配置：
```python
cluster_train_file = "./train_data_dir/train/train.txt"
......@@ -195,91 +179,10 @@ PaddlePaddle可以使用多种分布式计算平台构建分布式计算任务
在使用分布式计算平台进行训练时，任务被调度在集群中时，分布式计算平台通常会通过API或者环境变量提供任务运行需要的参数，比如节点的ID、IP和任务节点个数等。
### 使用Fabric启动集群作业
#### 准备一个Linux集群
可以在`paddle/scripts/cluster_train_v2/fabric/docker_cluster`目录下，执行`kubectl -f ssh_servers.yaml`启动一个测试集群，并使用`kubectl get po -o wide`获得这些节点的IP地址。
#### 启动集群作业
`paddle.py` 提供了自动化脚本来启动不同节点中的所有 PaddlePaddle 集群进程。默认情况下，所有命令行选项可以设置为 `paddle.py` 命令选项并且 `paddle.py` 将透明、自动地将这些选项应用到 PaddlePaddle 底层进程。
`paddle.py` 为方便作业启动提供了两个独特的命令选项。
- `job_dispatch_package` 设为本地 `workspace` 目录，它将被分发到 `conf.py` 中设置的所有节点。它有助于帮助频繁修改和访问工作区文件的用户减少负担，否则频繁的多节点工作空间部署可能会很麻烦。
- `job_workspace` 设为已部署的工作空间目录，`paddle.py` 将跳过分发阶段直接启动所有节点的集群作业。它可以帮助减少分发延迟。
`cluster_train/run.sh` 提供了命令样例来运行 `doc/howto/usage/cluster/src/word2vec` 集群任务，只需用您定义的目录修改 `job_dispatch_package` 和 `job_workspace`，然后：
```
sh run.sh
```
集群作业将会在几秒后启动。
#### 终止集群作业
`paddle.py`能获取`Ctrl + C` SIGINT 信号来自动终止它启动的所有进程。只需中断 `paddle.py` 任务来终止集群作业。如果程序崩溃你也可以手动终止。
#### 检查集群训练结果
详细信息请检查 $workspace/log 里的日志，每一个节点都有相同的日志结构。
`paddle_trainer.INFO`
提供几乎所有训练的内部输出日志，与本地训练相同。这里检验运行时间模型的收敛。
`paddle_pserver2.INFO`
提供 pserver 运行日志，有助于诊断分布式错误。
`server.log`
提供 parameter server 进程的 stderr 和 stdout。训练失败时可以检查错误日志。
`train.log`
提供训练过程的 stderr 和 stdout。训练失败时可以检查错误日志。
#### 检查模型输出
运行完成后，模型文件将被写入节点 0 的 `output` 目录中。
工作空间中的 `nodefile` 表示当前集群作业的节点 ID。
### 在OpenMPI集群中提交训练作业
#### 准备OpenMPI集群
执行下面的命令以启动3个节点的OpenMPI集群和一个"head"节点：
```bash
paddle/scripts/cluster_train_v2/openmpi/docker_cluster
kubectl create -f head.yaml
kubectl create -f mpi-nodes.yaml
```
然后可以从head节点ssh无密码登录到OpenMPI的每个节点上。
#### 启动集群作业
您可以按照下面的步骤在OpenMPI集群中提交paddle训练任务：
```bash
# 获得head和node节点的IP地址
kubectl get po -o wide
# 将node节点的IP地址保存到machines文件中
kubectl get po -o wide | grep nodes | awk '{print $6}' > machines
# 拷贝必要的文件到head节点
scp -i ssh/id_rsa.mpi.pub machines prepare.py train.py start_mpi_train.sh tutorial@[headIP]:~
# ssh 登录到head节点
ssh -i ssh/id_rsa.mpi.pub tutorial@[headIP]
# --------------- 以下操作均在head节点中执行 ---------------
# 准备训练数据
python prepare.py
# 拷贝训练程序和字典文件到每台MPI节点
cat machines | xargs -i scp word_dict.pickle train.py start_mpi_train.sh machines {}:/home/tutorial
# 创建日志目录
mpirun -hostfile machines -n 3 mkdir /home/tutorial/logs
# 拷贝训练数据到各自的节点
scp train.txt-00000 test.txt-00000 [node1IP]:/home/tutorial
scp train.txt-00001 test.txt-00001 [node2IP]:/home/tutorial
scp train.txt-00002 test.txt-00002 [node3IP]:/home/tutorial
# 启动训练任务
mpirun -hostfile machines -n 3 /home/tutorial/start_mpi_train.sh
```
### 在Kubernetes集群中提交训练作业
## 在不同集群中运行
此部分的使用方法可以参考[here](../k8s/k8s_distributed_cn.md)。
- [fabric集群](fabric_cn.md)
- [openmpi集群](openmpi_cn.md)
- [kubernetes单机](k8s_cn.md)
- [kubernetes distributed分布式](k8s_distributed_cn.md)
- [AWS上运行kubernetes集群训练](k8s_aws_cn.md)
doc/howto/usage/cluster/cluster_train_en.md
浏览文件 @ ba8660a1
# PaddlePaddle Distributed Training
* [Introduction](#introduction)
* [Preparations](#preparations)
* [Command-line arguments](#command-line-arguments)
* [Starting parameter server](#starting-parameter-server)
* [Starting trainer](#starting-trainer)
* [Prepare Training Dataset](#prepare-training-dataset)
* [Prepare Training program](#prepare-training-program)
* [Use cluster platforms or cluster management tools](#use-cluster-platforms-or-cluster-management-tools)
* [Cluster Training Using Fabric](#cluster-training-using-fabric)
* [Prepare a Linux cluster](#prepare-a-linux-cluster)
* [Launching Cluster Job](#launching-cluster-job)
* [Kill Cluster Job](#kill-cluster-job)
* [Check Cluster Training Result](#check-cluster-training-result)
* [Check Model Output](#check-model-output)
* [Cluster Training Using OpenMPI](#cluster-training-using-openmpi)
* [Prepare an OpenMPI cluster](#prepare-an-openmpi-cluster)
* [Launching Cluster Job](#launching-cluster-job-1)
* [Cluster Training Using Kubernetes](#cluster-training-using-kubernetes)
# Distributed Training
## Introduction
......@@ -35,7 +16,7 @@ When training with synchronize SGD, PaddlePaddle uses an internal "synchronize b
## Preparations
1. Prepare your computer cluster. It's normally a bunch of Linux servers connected by LAN. Each server will be assigned a unique IP address. The computers in the cluster can be called "nodes".
2. Install PaddlePaddle on every node. If you are going to take advantage of GPU cards, you'll also need to install proper driver and CUDA libraries. To install PaddlePaddle please read [this build and install](https://github.com/PaddlePaddle/Paddle/tree/develop/doc/getstarted/build_and_install) document. We strongly recommend using [Docker installation](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/getstarted/build_and_install/docker_install_en.rst).
2. Install PaddlePaddle on every node. If you are going to take advantage of GPU cards, you'll also need to install proper driver and CUDA libraries. To install PaddlePaddle please read [this build and install](http://www.paddlepaddle.org/docs/develop/documentation/en/getstarted/build_and_install/index_en.html) document. We strongly recommend using [Docker installation](http://www.paddlepaddle.org/docs/develop/documentation/en/getstarted/build_and_install/docker_install_en.html).
After installation, you can check the version by typing the below command (run a docker container if using docker: `docker run -it paddlepaddle/paddle:[tag] /bin/bash`):
......@@ -67,12 +48,12 @@ If you wish to run parameter servers in background, and save a log file, you can
$ stdbuf -oL /usr/bin/nohup paddle pserver --port=7164 --ports_num=1 --ports_num_for_sparse=1 --num_gradient_servers=1 &> pserver.log
```
| param | required | default | description |
| ------------- | ------------- | ------------- | ------------- |
| port | required | 7164 | port which parameter server will listen on. If ports_num greater than 1, parameter server will listen on multiple ports for more network throughput |
| ports_num | required | 1 | total number of ports will listen on |
| ports_num_for_sparse | required | 1 | number of ports which serves sparse parameter update |
| num_gradient_servers | required | 1 | total number of gradient servers |
Parameter Description
- port: **required, default 7164**, port which parameter server will listen on. If ports_num greater than 1, parameter server will listen on multiple ports for more network throughput.
- ports_num: **required, default 1**, total number of ports will listen on.
- ports_num_for_sparse: **required, default 1**, number of ports which serves sparse parameter update.
- num_gradient_servers: **required, default 1**, total number of gradient servers.
### Starting trainer
Type the command below to start the trainer(name the file whatever you want, like "train.py")
......@@ -111,16 +92,16 @@ paddle.init(
pservers="127.0.0.1")
```
| param | required | default | description |
| ------------- | ------------- | ------------- | ------------- |
| use_gpu | optional | False | set to "True" to enable GPU training |
| trainer_count | required | 1 | total count of trainers in the training job |
| port | required | 7164 | port to connect to parameter server |
| ports_num | required | 1 | number of ports for communication |
| ports_num_for_sparse | required | 1 | number of ports for sparse type caculation |
| num_gradient_servers | required | 1 | total number of gradient server |
| trainer_id | required | 0 | ID for every trainer, start from 0 |
| pservers | required | 127.0.0.1 | list of IPs of parameter servers, separated by "," |
Parameter Description
- use_gpu: **optional, default False**, set to "True" to enable GPU training.
- trainer_count: **required, default 1**, total count of trainers in the training job.
- port: **required, default 7164**, port to connect to parameter server.
- ports_num: **required, default 1**, number of ports for communication.
- ports_num_for_sparse: **required, default 1**, number of ports for sparse type caculation.
- num_gradient_servers: **required, default 1**, total number of gradient server.
- trainer_id: **required, default 0**, ID for every trainer, start from 0.
- pservers: **required, default 127.0.0.1**, list of IPs of parameter servers, separated by ",".
### Prepare Training Dataset
......@@ -178,7 +159,7 @@ Your workspace may looks like:
- `my_lib.py`: user defined libraries, like PIL libs. This is optional.
- `word_dict.pickle`: dict file for training word embeding.
- `train.py`: training program. Sample code: [api_train_v2_cluster.py](https://github.com/PaddlePaddle/Paddle/tree/develop/doc/howto/usage/cluster/src/word2vec/prepare.py). ***NOTE:*** You may need to modify the head part of `train.py` when using different cluster platform to retrive configuration environment variables:
- `train.py`: training program. Sample code: [api_train_v2_cluster.py](https://github.com/PaddlePaddle/Paddle/tree/develop/doc/howto/usage/cluster/src/word2vec/api_train_v2_cluster.py). ***NOTE:*** You may need to modify the head part of `train.py` when using different cluster platform to retrive configuration environment variables:
```python
cluster_train_file = "./train_data_dir/train/train.txt"
......@@ -202,92 +183,9 @@ We'll introduce cluster job management on these platforms. The examples can be f
These cluster platforms provide API or environment variables for training processes, when the job is dispatched to different nodes. Like node ID, IP or total number of nodes etc.
### Cluster Training Using Fabric
#### Prepare a Linux cluster
Run `kubectl -f ssh_servers.yaml` under the directory: `paddle/scripts/cluster_train_v2/fabric/docker_cluster` will launch a demo cluster. Run `kubectl get po -o wide` to get IP addresses of these nodes.
#### Launching Cluster Job
`paddle.py` provides automatical scripts to start all PaddlePaddle cluster processes in different nodes. By default, all command line options can be set as `paddle.py` command options and `paddle.py` will transparently and automatically set these options to PaddlePaddle lower level processes.
`paddle.py`provides two distinguished command option for easy job launching.
- `job_dispatch_package` set it with local `workspace` directory, it will be dispatched to all nodes which is set in `conf.py`. It could be helpful for frequently manipulating workspace files. otherwise, frequent multi-nodes workspace deployment is very annoying.
- `job_workspace` set it with already deployed workspace directory, `paddle.py` will skip dispatch stage to directly launch cluster job with all nodes. It could help to reduce heavy
dispatch latency.
`cluster_train/run.sh` provides command line sample to run `demo/recommendation` cluster job, just modify `job_dispatch_package` and `job_workspace` with your defined directory, then:
```
sh run.sh
```
The cluster Job will start in several seconds.
#### Kill Cluster Job
`paddle.py` can capture `Ctrl + C` SIGINT signal to automatically kill all processes launched by it. So just stop `paddle.py` to kill cluster job. You should manually kill the job if the program crashed.
#### Check Cluster Training Result
Check log in $workspace/log for details, each node owns same log structure.
`paddle_trainer.INFO`
It provides almost all internal output log for training, same as local training. Check runtime model convergence here.
`paddle_pserver2.INFO`
It provides parameter server running log, which could help to diagnose distributed error.
`server.log`
It provides stderr and stdout of parameter server process. Check error log if training crashes.
`train.log`
It provides stderr and stdout of trainer process. Check error log if training crashes.
#### Check Model Output
After one pass finished, model files will be written in `output` directory in node 0.
`nodefile` in workspace indicates the node id of current cluster job.
### Cluster Training Using OpenMPI
#### Prepare an OpenMPI cluster
Run the following command to start a 3-node MPI cluster and one "head" node.
```bash
cd paddle/scripts/cluster_train_v2/openmpi/docker_cluster
kubectl create -f head.yaml
kubectl create -f mpi-nodes.yaml
```
Then you can log in to every OpenMPI node using ssh without input any passwords.
#### Launching Cluster Job
Follow the steps to launch a PaddlePaddle training job in OpenMPI cluster:\
```bash
# find out node IP addresses
kubectl get po -o wide
# generate a "machines" file containing node IP addresses
kubectl get po -o wide | grep nodes | awk '{print $6}' > machines
# copy necessary files onto "head" node
scp -i ssh/id_rsa.mpi.pub machines prepare.py train.py start_mpi_train.sh tutorial@[headIP]:~
# login to head node using ssh
ssh -i ssh/id_rsa.mpi.pub tutorial@[headIP]
# --------------- in head node ---------------
# prepare training data
python prepare.py
# copy training data and dict file to MPI nodes
cat machines | xargs -i scp word_dict.pickle train.py start_mpi_train.sh machines {}:/home/tutorial
# creat a directory for storing log files
mpirun -hostfile machines -n 3 mkdir /home/tutorial/logs
# copy training data to every node
scp train.txt-00000 test.txt-00000 [node1IP]:/home/tutorial
scp train.txt-00001 test.txt-00001 [node2IP]:/home/tutorial
scp train.txt-00002 test.txt-00002 [node3IP]:/home/tutorial
# start the job
mpirun -hostfile machines -n 3 /home/tutorial/start_mpi_train.sh
```
### Cluster Training Using Kubernetes
## Use different clusters
The details can be found [here](../k8s/k8s_cn.md)
- [fabric](fabric_en.md)
- [openmpi](openmpi_en.md)
- [kubernetes](k8s_en.md)
- [kubernetes on AWS](k8s_aws_en.md)
doc/howto/usage/cluster/fabric_cn.md 0 → 100644
浏览文件 @ ba8660a1
# 使用fabric启动集群训练
## 准备一个Linux集群
可以在`paddle/scripts/cluster_train_v2/fabric/docker_cluster`目录下，执行`kubectl -f ssh_servers.yaml`启动一个测试集群，并使用`kubectl get po -o wide`获得这些节点的IP地址。
## 启动集群作业
`paddle.py` 提供了自动化脚本来启动不同节点中的所有 PaddlePaddle 集群进程。默认情况下，所有命令行选项可以设置为 `paddle.py` 命令选项并且 `paddle.py` 将透明、自动地将这些选项应用到 PaddlePaddle 底层进程。
`paddle.py` 为方便作业启动提供了两个独特的命令选项。
- `job_dispatch_package` 设为本地 `workspace` 目录，它将被分发到 `conf.py` 中设置的所有节点。它有助于帮助频繁修改和访问工作区文件的用户减少负担，否则频繁的多节点工作空间部署可能会很麻烦。
- `job_workspace` 设为已部署的工作空间目录，`paddle.py` 将跳过分发阶段直接启动所有节点的集群作业。它可以帮助减少分发延迟。
`cluster_train/run.sh` 提供了命令样例来运行 `doc/howto/usage/cluster/src/word2vec` 集群任务，只需用您定义的目录修改 `job_dispatch_package` 和 `job_workspace`，然后：
```
sh run.sh
```
集群作业将会在几秒后启动。
## 终止集群作业
`paddle.py`能获取`Ctrl + C` SIGINT 信号来自动终止它启动的所有进程。只需中断 `paddle.py` 任务来终止集群作业。如果程序崩溃你也可以手动终止。
## 检查集群训练结果
详细信息请检查 $workspace/log 里的日志，每一个节点都有相同的日志结构。
`paddle_trainer.INFO`
提供几乎所有训练的内部输出日志，与本地训练相同。这里检验运行时间模型的收敛。
`paddle_pserver2.INFO`
提供 pserver 运行日志，有助于诊断分布式错误。
`server.log`
提供 parameter server 进程的 stderr 和 stdout。训练失败时可以检查错误日志。
`train.log`
提供训练过程的 stderr 和 stdout。训练失败时可以检查错误日志。
## 检查模型输出
运行完成后，模型文件将被写入节点 0 的 `output` 目录中。
工作空间中的 `nodefile` 表示当前集群作业的节点 ID。
doc/howto/usage/cluster/fabric_en.md 0 → 100644
浏览文件 @ ba8660a1
# Cluster Training Using Fabric
## Prepare a Linux cluster
Run `kubectl -f ssh_servers.yaml` under the directory: `paddle/scripts/cluster_train_v2/fabric/docker_cluster` will launch a demo cluster. Run `kubectl get po -o wide` to get IP addresses of these nodes.
## Launching Cluster Job
`paddle.py` provides automatical scripts to start all PaddlePaddle cluster processes in different nodes. By default, all command line options can be set as `paddle.py` command options and `paddle.py` will transparently and automatically set these options to PaddlePaddle lower level processes.
`paddle.py`provides two distinguished command option for easy job launching.
- `job_dispatch_package` set it with local `workspace` directory, it will be dispatched to all nodes which is set in `conf.py`. It could be helpful for frequently manipulating workspace files. otherwise, frequent multi-nodes workspace deployment is very annoying.
- `job_workspace` set it with already deployed workspace directory, `paddle.py` will skip dispatch stage to directly launch cluster job with all nodes. It could help to reduce heavy
dispatch latency.
`cluster_train/run.sh` provides command line sample to run `demo/recommendation` cluster job, just modify `job_dispatch_package` and `job_workspace` with your defined directory, then:
```
sh run.sh
```
The cluster Job will start in several seconds.
## Kill Cluster Job
`paddle.py` can capture `Ctrl + C` SIGINT signal to automatically kill all processes launched by it. So just stop `paddle.py` to kill cluster job. You should manually kill the job if the program crashed.
## Check Cluster Training Result
Check log in $workspace/log for details, each node owns same log structure.
`paddle_trainer.INFO`
It provides almost all internal output log for training, same as local training. Check runtime model convergence here.
`paddle_pserver2.INFO`
It provides parameter server running log, which could help to diagnose distributed error.
`server.log`
It provides stderr and stdout of parameter server process. Check error log if training crashes.
`train.log`
It provides stderr and stdout of trainer process. Check error log if training crashes.
## Check Model Output
After one pass finished, model files will be written in `output` directory in node 0.
`nodefile` in workspace indicates the node id of current cluster job.
doc/howto/usage/cluster/k8s_aws_cn.md 0 → 120000
浏览文件 @ ba8660a1
k8s_aws_en.md
\ No newline at end of file
doc/howto/usage/k8s/k8s_aws_en.md → doc/howto/usage/cluster/k8s_aws_en.md
浏览文件 @ ba8660a1
......@@ -493,7 +493,7 @@ spec:
spec:
containers:
- name: paddle-data
image: paddledev/paddle-tutorial:k8s_data
image: paddlepaddle/paddle-tutorial:k8s_data
imagePullPolicy: Always
volumeMounts:
- mountPath: "/efs"
......@@ -522,7 +522,7 @@ NAME DESIRED SUCCESSFUL AGE
paddle-data 1 1 6m
```
Data preparation is done by docker image `paddledev/paddle-tutorial:k8s_data`, see [here](src/k8s_data/README.md) for how to build this docker image and source code.
Data preparation is done by docker image `paddlepaddle/paddle-tutorial:k8s_data`, see [here](src/k8s_data/README.md) for how to build this docker image and source code.
#### Start Training
......@@ -545,7 +545,7 @@ spec:
claimName: efsvol
containers:
- name: trainer
image: paddledev/paddle-tutorial:k8s_train
image: paddlepaddle/paddle-tutorial:k8s_train
command: ["bin/bash", "-c", "/root/start.sh"]
env:
- name: JOB_NAME
......@@ -617,7 +617,7 @@ kubectl --kubeconfig=kubeconfig log -f POD_NAME
Run `kubectl --kubeconfig=kubeconfig describe job paddle-cluster-job` to check training job status. It will complete in around 20 minutes.
The details for start `pserver` and `trainer` are hidden inside docker image `paddledev/paddle-tutorial:k8s_train`, see [here](src/k8s_train/README.md) for how to build the docker image and source code.
The details for start `pserver` and `trainer` are hidden inside docker image `paddlepaddle/paddle-tutorial:k8s_train`, see [here](src/k8s_train/README.md) for how to build the docker image and source code.
#### Inspect Training Output
......
doc/howto/usage/k8s/k8s_cn.md → doc/howto/usage/cluster/k8s_cn.md
浏览文件 @ ba8660a1
# Kubernetes单机训练
在这篇文档里，我们介绍如何在 Kubernetes 集群上启动一个单机使用CPU的Paddle训练作业。在下一篇中，我们将介绍如何启动分布式训练作业。
在这篇文档里，我们介绍如何在 Kubernetes 集群上启动一个单机使用CPU的PaddlePaddle训练作业。在下一篇中，我们将介绍如何启动分布式训练作业。
## 制作Docker镜像
在一个功能齐全的Kubernetes机群里，通常我们会安装Ceph等分布式文件系统来存储训练数据。这样的话，一个分布式Paddle训练任务中的每个进程都可以从Ceph读取数据。在这个例子里，我们只演示一个单机作业，所以可以简化对环境的要求，把训练数据直接放在
Paddle的Docker image里。为此，我们需要制作一个包含训练数据的Paddle镜像。
在一个功能齐全的Kubernetes机群里，通常我们会安装Ceph等分布式文件系统来存储训练数据。这样的话，一个分布式PaddlePaddle训练任务中
的每个进程都可以从Ceph读取数据。在这个例子里，我们只演示一个单机作业，所以可以简化对环境的要求，把训练数据直接放在
PaddlePaddle的Docker Image里。为此，我们需要制作一个包含训练数据的PaddlePaddle镜像。
PaddlePaddle的 `paddlepaddle/paddle:cpu-demo-latest` 镜像里有PaddlePaddle的源码与demo，
（请注意，默认的PaddlePaddle生产环境镜像 `paddlepaddle/paddle:latest` 是不包括源码的，PaddlePaddle的各版本镜像可以参考
[Docker Installation Guide](http://paddlepaddle.org/docs/develop/documentation/zh/getstarted/build_and_install/docker_install_cn.html)），
下面我们使用这个镜像来下载数据到Docker Container中，并把这个包含了训练数据的Container保存为一个新的镜像。
Paddle 的 [Quick Start Tutorial](http://www.paddlepaddle.org/doc/demo/quick_start/index_en.html)
里介绍了用Paddle源码中的脚本下载训练数据的过程。
而 `paddledev/paddle:cpu-demo-latest` 镜像里有 Paddle 源码与demo，（ 请注意，默认的
Paddle镜像 `paddledev/paddle:cpu-latest` 是不包括源码的, Paddle的各版本镜像可以参考 [Docker installation guide](http://www.paddlepaddle.org/doc/build/docker_install.html) ），所以我们使用这个镜像来下载训练数据到Docker container中，然后把这个包含了训练数据的container保存为一个新的镜像。
### 运行容器
```
$ docker run --name quick_start_data -it paddledev/paddle:cpu-demo-latest
$ docker run --name quick_start_data -it paddlepaddle/paddle:cpu-demo-latest
```
### 下载数据
......@@ -103,7 +104,7 @@ spec:
restartPolicy: Never
```
### 创建Paddle Job
### 创建PaddlePaddle Job
使用上文创建的yaml文件创建Kubernetes Job，命令为：
......
doc/howto/usage/k8s/k8s_distributed_cn.md → doc/howto/usage/cluster/k8s_distributed_cn.md
浏览文件 @ ba8660a1
# Kubernetes分布式训练
前一篇文章介绍了如何在Kubernetes集群上启动一个单机PaddlePaddle训练作业 (Job)。在这篇文章里，我们介绍如何在Kubernetes集群上进行分布式PaddlePaddle训练作业。关于PaddlePaddle的分布式训练，文章 [Cluster Training](https://github.com/baidu/Paddle/blob/develop/doc/cluster/opensource/cluster_train.md)介绍了一种通过SSH远程分发任务，进行分布式训练的方法，与此不同的是，本文将介绍在Kubernetes容器管理平台上快速构建PaddlePaddle容器集群，进行分布式训练的方案。
有关Kubernetes相关概念以及如何搭建和配置Kubernetes集群，可以参考[k8s_basis](./k8s_basis_cn.md)。
前一篇文章介绍了如何在Kubernetes集群上启动一个单机PaddlePaddle训练作业 (Job)。在这篇文章里，我们介绍如何在Kubernetes集群上进行分布式PaddlePaddle训练作业。关于PaddlePaddle的分布式训练，文章 [Cluster Training](http://www.paddlepaddle.org/docs/develop/documentation/zh/howto/usage/cluster/cluster_train_cn.html)介绍了一种通过SSH远程分发任务，进行分布式训练的方法，与此不同的是，本文将介绍在Kubernetes容器管理平台上快速构建PaddlePaddle容器集群，进行分布式训练的方案。
## 整体方案
......@@ -28,7 +26,7 @@ PaddlePaddle镜像需要提供`paddle pserver`与`paddle train`进程的运行
- 拷贝训练文件到容器内
- 生成`paddle pserver`与`paddle train`进程的启动参数，并且启动训练
因为官方镜像 `paddledev/paddle:cpu-latest` 内已经包含PaddlePaddle的执行程序但是还没上述功能，所以我们可以在这个基础上，添加启动脚本，制作新镜像来完成以上的工作。参考镜像的[*Dockerfile*](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/howto/usage/cluster/k8s/src/k8s_train/Dockerfile)。
因为官方镜像 `paddlepaddle/paddle:latest` 内已经包含PaddlePaddle的执行程序但是还没上述功能，所以我们可以在这个基础上，添加启动脚本，制作新镜像来完成以上的工作。参考镜像的[*Dockerfile*](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/howto/usage/cluster/src/k8s_train/Dockerfile)。
```bash
$ cd doc/howto/usage/k8s/src/k8s_train
......@@ -62,7 +60,7 @@ spec:
hostNetwork: true
containers:
- name: paddle-data
image: paddledev/paddle-tutorial:k8s_data
image: paddlepaddle/paddle-tutorial:k8s_data
imagePullPolicy: Always
volumeMounts:
- mountPath: "/mnt"
......@@ -149,20 +147,19 @@ spec:
文件中，`metadata`下的`name`表示这个job的名字。`parallelism，completions`字段表示这个job会同时开启3个PaddlePaddle节点，成功训练且退出的pod数目为3时，这个job才算成功结束。然后申明一个存储卷`jobpath`，代表宿主机目录`/home/work/mfs`，在对容器的描述`containers`字段中，将此目录挂载为容器的`/home/jobpath`目录，这样容器的`/home/jobpath`目录就成为了共享存储，放在这个目录里的文件其实是保存到了MFS上。
`env`字段表示容器的环境变量，我们将`paddle`运行的一些参数通过这种方式传递到容器内。
`env`字段表示容器的环境变量，我们将`paddle`运行的一些参数通过这种方式传递到容器内：
环境变量 | 说明
--- | ---
JOB_PATH | 共享存储挂在的路径
JOB_NAME | Job的名字
TRAIN_CONFIG_DIR | 本次训练文件所在目录，与JOB_PATH,JOB_NAME组合可以找到本次训练需要的文件路径
CONF_PADDLE_NIC | `paddle pserver`进程需要的`--nics`参数，即网卡名
CONF_PADDLE_PORT | `paddle paserver`的`--port`参数
CONF_PADDLE_PORTS_NUM | 稠密更新的端口数量，即`--ports_num`参数
CONF_PADDLE_PORTS_NUM_SPARSE | 稀疏更新的端口数量，即`--ports_num_for_sparse`参数
CONF_PADDLE_GRADIENT_NUM | 训练节点数量，即`--num_gradient_servers参数`
- JOB_PATH：共享存储挂在的路径
- JOB_NAME：Job的名字
- TRAIN_CONFIG_DIR：本次训练文件所在目录，与JOB_PATH,JOB_NAME组合可以找到本次训练需要的文件路径
- CONF_PADDLE_NIC：`paddle pserver`进程需要的`--nics`参数，即网卡名
- CONF_PADDLE_PORT：`paddle paserver`的`--port`参数
- CONF_PADDLE_PORTS_NUM：稠密更新的端口数量，即`--ports_num`参数
- CONF_PADDLE_PORTS_NUM_SPARSE：稀疏更新的端口数量，即`--ports_num_for_sparse`参数
- CONF_PADDLE_GRADIENT_NUM：训练节点数量，即`--num_gradient_servers参数`
这些参数的具体描述，读者可以查看[这里](http://www.paddlepaddle.org/doc/ui/cmd_argument/detail_introduction.html#parameter-server-and-distributed-communication)。
这些参数的具体描述，读者可以查看[这里](http://www.paddlepaddle.org/docs/develop/documentation/zh/howto/usage/cmd_parameter/detail_introduction_cn.html)。
编写完YAML文件后，可以使用Kubernetes的命令行工具创建job。
......
doc/howto/usage/k8s/k8s_en.md → doc/howto/usage/cluster/k8s_en.md
浏览文件 @ ba8660a1
# Paddle On Kubernetes
# PaddlePaddle On Kubernetes
>In this article, we will introduce how to run Paddle training job on single CPU machine using Kubernetes. In next article, we will introduce how to run Paddle training job on distributed cluster.
In this article, we will introduce how to run PaddlePaddle training job on single CPU machine using Kubernetes. In next article, we will introduce how to run PaddlePaddle training job on distributed cluster.
## Build Docker Image
In distributed Kubernetes cluster, we will use Ceph or other shared storage system for storing training related data so that all processes in Paddle training can retrieve data from Ceph. In this example, we will only demo training job on single machine. In order to simplify the requirement of the environment, we will directly put training data into Paddle's Docker Image, so we need to create a Paddle Docker image that already includes the training data.
In distributed Kubernetes cluster, we will use Ceph or other distributed
storage system for storing training related data so that all processes in
PaddlePaddle training can retrieve data from Ceph. In this example, we will
only demo training job on single machine. In order to simplify the requirement
of the environment, we will directly put training data into the PaddlePaddle Docker Image,
so we need to create a PaddlePaddle Docker image that includes the training data.
The production Docker Image `paddlepaddle/paddle:cpu-demo-latest` has the PaddlePaddle
source code and demo. (Caution: Default PaddlePaddle Docker Image `paddlepaddle/paddle:latest` doesn't include
the source code, PaddlePaddle's different versions of Docker Image can be referred here:
[Docker Installation Guide](http://paddlepaddle.org/docs/develop/documentation/zh/getstarted/build_and_install/docker_install_en.html)),
so we run this Docker Image and download the training data, and then commit the whole
Container to be a new Docker Image.
Paddle's [Quick Start Tutorial](http://www.paddlepaddle.org/doc/demo/quick_start/index_en.html) introduces how to download and train data by using script from Paddle's source code.
And `paddledev/paddle:cpu-demo-latest` image has the Paddle source code and demo. (Caution: Default Paddle image `paddledev/paddle:cpu-latest` doesn't include the source code, Paddle's different versions of image can be referred here: [Docker installation guide](http://www.paddlepaddle.org/doc/build/docker_install.html)), so we run this container and download the training data, and then commit the whole container to be a new Docker image.
### Run Docker Container
```
$ docker run --name quick_start_data -it paddledev/paddle:cpu-demo-latest
$ docker run --name quick_start_data -it paddlepaddle/paddle:cpu-demo-latest
```
### Download Training Data
......@@ -67,7 +76,7 @@ $ docker commit quick_start_data mypaddle/paddle:quickstart
## Use Kubernetes For Training
>We will use Kubernetes job for training process, following steps shows how to do the training with Kubernetes.
We will use Kubernetes job for training process, following steps shows how to do the training with Kubernetes.
### Create Yaml Files
......@@ -99,7 +108,7 @@ spec:
restartPolicy: Never
```
### Start Paddle Job
### Start PaddlePaddle Job
Using the above yaml file to start the Kubernetes job.
......
doc/howto/usage/cluster/openmpi_cn.md 0 → 100644
浏览文件 @ ba8660a1
# 在OpenMPI集群中提交训练作业
## 准备OpenMPI集群
执行下面的命令以启动3个节点的OpenMPI集群和一个"head"节点：
```bash
paddle/scripts/cluster_train_v2/openmpi/docker_cluster
kubectl create -f head.yaml
kubectl create -f mpi-nodes.yaml
```
然后可以从head节点ssh无密码登录到OpenMPI的每个节点上。
## 启动集群作业
您可以按照下面的步骤在OpenMPI集群中提交paddle训练任务：
```bash
# 获得head和node节点的IP地址
kubectl get po -o wide
# 将node节点的IP地址保存到machines文件中
kubectl get po -o wide | grep nodes | awk '{print $6}' > machines
# 拷贝必要的文件到head节点
scp -i ssh/id_rsa.mpi.pub machines prepare.py train.py start_mpi_train.sh tutorial@[headIP]:~
# ssh 登录到head节点
ssh -i ssh/id_rsa.mpi.pub tutorial@[headIP]
# --------------- 以下操作均在head节点中执行 ---------------
# 准备训练数据
python prepare.py
# 拷贝训练程序和字典文件到每台MPI节点
cat machines | xargs -i scp word_dict.pickle train.py start_mpi_train.sh machines {}:/home/tutorial
# 创建日志目录
mpirun -hostfile machines -n 3 mkdir /home/tutorial/logs
# 拷贝训练数据到各自的节点
scp train.txt-00000 test.txt-00000 [node1IP]:/home/tutorial
scp train.txt-00001 test.txt-00001 [node2IP]:/home/tutorial
scp train.txt-00002 test.txt-00002 [node3IP]:/home/tutorial
# 启动训练任务
mpirun -hostfile machines -n 3 /home/tutorial/start_mpi_train.sh
```
doc/howto/usage/cluster/openmpi_en.md 0 → 100644
浏览文件 @ ba8660a1
# Cluster Training Using OpenMPI
## Prepare an OpenMPI cluster
Run the following command to start a 3-node MPI cluster and one "head" node.
```bash
cd paddle/scripts/cluster_train_v2/openmpi/docker_cluster
kubectl create -f head.yaml
kubectl create -f mpi-nodes.yaml
```
Then you can log in to every OpenMPI node using ssh without input any passwords.
## Launching Cluster Job
Follow the steps to launch a PaddlePaddle training job in OpenMPI cluster:\
```bash
# find out node IP addresses
kubectl get po -o wide
# generate a "machines" file containing node IP addresses
kubectl get po -o wide | grep nodes | awk '{print $6}' > machines
# copy necessary files onto "head" node
scp -i ssh/id_rsa.mpi.pub machines prepare.py train.py start_mpi_train.sh tutorial@[headIP]:~
# login to head node using ssh
ssh -i ssh/id_rsa.mpi.pub tutorial@[headIP]
# --------------- in head node ---------------
# prepare training data
python prepare.py
# copy training data and dict file to MPI nodes
cat machines | xargs -i scp word_dict.pickle train.py start_mpi_train.sh machines {}:/home/tutorial
# creat a directory for storing log files
mpirun -hostfile machines -n 3 mkdir /home/tutorial/logs
# copy training data to every node
scp train.txt-00000 test.txt-00000 [node1IP]:/home/tutorial
scp train.txt-00001 test.txt-00001 [node2IP]:/home/tutorial
scp train.txt-00002 test.txt-00002 [node3IP]:/home/tutorial
# start the job
mpirun -hostfile machines -n 3 /home/tutorial/start_mpi_train.sh
```
doc/howto/usage/k8s/src/Dockerfile → doc/howto/usage/cluster/src/Dockerfile
浏览文件 @ ba8660a1
FROM paddledev/paddle:cpu-latest
FROM paddlepaddle/paddle:latest
MAINTAINER zjsxzong89@gmail.com
......
doc/howto/usage/k8s/src/k8s_train/Dockerfile → doc/howto/usage/cluster/src/k8s_train/Dockerfile
浏览文件 @ ba8660a1
FROM paddledev/paddle:cpu-latest
FROM paddlepaddle/paddle:latest
COPY start.sh /root/
COPY start_paddle.py /root/
......
doc/howto/usage/k8s/k8s_basis_cn.md 已删除 100644 → 0
浏览文件 @ 544ff784
# Kubernetes 简介
[*Kubernetes*](http://kubernetes.io/)是Google开源的容器集群管理系统，其提供应用部署、维护、扩展机制等功能，利用Kubernetes能方便地管理跨机器运行容器化的应用。Kubernetes可以在物理机或虚拟机上运行，且支持部署到[AWS](http://kubernetes.io/docs/getting-started-guides/aws)，[Azure](http://kubernetes.io/docs/getting-started-guides/azure/)，[GCE](http://kubernetes.io/docs/getting-started-guides/gce)等多种公有云环境。介绍分布式训练之前，需要对[Kubernetes](http://kubernetes.io/)有一个基本的认识，下面先简要介绍一下本文用到的几个Kubernetes概念。
- [*Node*](http://kubernetes.io/docs/admin/node/) 表示一个Kubernetes集群中的一个工作节点，这个节点可以是物理机或者虚拟机，Kubernetes集群就是由node节点与master节点组成的。
- [*Pod*](http://kubernetes.io/docs/user-guide/pods/) 是一组(一个或多个)容器，pod是Kubernetes的最小调度单元，一个pod中的所有容器会被调度到同一个node上。Pod中的容器共享NET，PID，IPC，UTS等Linux namespace。由于容器之间共享NET namespace，所以它们使用同一个IP地址，可以通过*localhost*互相通信。不同pod之间可以通过IP地址访问。
- [*Job*](http://kubernetes.io/docs/user-guide/jobs/) 描述Kubernetes上运行的作业，一次作业称为一个job，通常每个job包括一个或者多个pods，job启动后会创建这些pod并开始执行一个程序，等待这个程序执行成功并返回0则成功退出，如果执行失败，也可以配置不同的重试机制。
- [*Volume*](http://kubernetes.io/docs/user-guide/volumes/) 存储卷，是pod内的容器都可以访问的共享目录，也是容器与node之间共享文件的方式，因为容器内的文件都是暂时存在的，当容器因为各种原因被销毁时，其内部的文件也会随之消失。通过volume，就可以将这些文件持久化存储。Kubernetes支持多种volume，例如hostPath(宿主机目录)，gcePersistentDisk，awsElasticBlockStore等。
- [*Namespaces*](https://kubernetes.io/docs/user-guide/namespaces/) 命名空间，在kubernetes中创建的所有资源对象(例如上文的pod，job)等都属于一个命名空间，在同一个命名空间中，资源对象的名字是唯一的，不同空间的资源名可以重复，命名空间主要为了对象进行逻辑上的分组便于管理。本文只使用了默认命名空间。
- [*PersistentVolume*](https://kubernetes.io/docs/user-guide/persistent-volumes/): 和[*PersistentVolumeClaim*](https://kubernetes.io/docs/user-guide/persistent-volumes/#persistentvolumeclaims)结合，将外部的存储服务在Kubernetes中描述成为统一的资源形式，便于存储资源管理和Pod引用。
## 部署Kubernetes集群
Kubernetes提供了多种集群部署的方案，本文档内不重复介绍。这里给出集中常见的部署方法：
- [*minikube*](https://kubernetes.io/docs/getting-started-guides/minikube/): 快速在本地启动一个单机的kubernetes服务器，便于本地验证和测试。
- [*kubeadm*](http://kubernetes.io/docs/getting-started-guides/kubeadm/): 在不同操作系统，不同主机(Bare-Metal, AWS, GCE)条件下，快速部署集群。
- [*AWS EC2*](https://kubernetes.io/docs/getting-started-guides/aws/): 在aws上快速部署集群。
- [*Bare-Metal*](https://kubernetes.io/docs/getting-started-guides/centos/centos_manual_config/): 在物理机上手动部署。
可以参考[这个表格](https://kubernetes.io/docs/getting-started-guides/#table-of-solutions)选择适合您的场景的合适方案。
## 选择存储方案
容器不会保留在运行时生成的数据，job或者应用程序在容器中运行时生成的数据会在容器销毁时消失。为了完成分布式机器学习训练任务，需要有一个外部的存储服务来保存训练所需数据和训练输出。
常见的可选存储服务包括：
- [*NFS*](https://github.com/kubernetes/kubernetes/tree/master/examples/volumes/nfs): 可以将磁盘上某个目录共享给网络中其他机器访问。部署和配置比较简单，可以用于小量数据的验证。不提供分布式存储，高可用，冗余等功能。NFS的部署方法可以参考[这里](http://www.tecmint.com/how-to-setup-nfs-server-in-linux/)。
- [*GlusterFS*](http://gluster.readthedocs.io/en/latest/Quick-Start-Guide/Quickstart/): 网络分布式文件系统，可以在Kubernetes中按照[这个](https://github.com/kubernetes/kubernetes/tree/master/examples/volumes/glusterfs)例子使用。
- [*Ceph*](http://docs.ceph.com/docs/master/): 分布式文件系统，支持rbd，POSIX API接口(ceph fs)和对象存储API，参考[这里](https://kubernetes.io/docs/user-guide/volumes/#rbd)。
- [*MooseFS*](https://moosefs.com/documentation.html): 一个分布式的存储系统。需要先挂载到服务器Node上再通过kubernetes hostPath Volume挂载到容器中。
## 配置kubectl
### 安装kubectl
```
# OS X
curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/darwin/amd64/kubectl
# Linux
curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/linux/amd64/kubectl
# Windows
curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/windows/amd64/kubectl.exe
```
### 配置kubectl访问你的kubernetes集群
编辑`~/.kube/config`这个配置文件，修改`Master-IP`的地址。如果使用SSL认证，则需要配置`certificate-authority`和`users`中的用户证书。如果是使用非SSL方式访问（比如通过8080端口），也可以去掉这些证书的配置。
```
apiVersion: v1
clusters:
- cluster:
certificate-authority: /path/to/ca.crt
server: https://[Master-IP]:443
name: minikube
contexts:
- context:
cluster: minikube
user: minikube
name: minikube
current-context: minikube
kind: Config
preferences: {}
users:
- name: minikube
user:
client-certificate: /path/to/apiserver.crt
client-key: /Users/wuyi/.minikube/apiserver.key
```
paddle/framework/attribute.cc
浏览文件 @ ba8660a1
......@@ -19,42 +19,42 @@ limitations under the License. */
namespace paddle {
namespace framework {
Attribute GetAttrValue(const OpDesc::Attr& attr_desc) {
Attribute GetAttrValue(const proto::OpDesc::Attr& attr_desc) {
switch (attr_desc.type()) {
case framework::AttrType::BOOLEAN: {
case proto::AttrType::BOOLEAN: {
return attr_desc.b();
}
case framework::AttrType::INT: {
case proto::AttrType::INT: {
return attr_desc.i();
}
case framework::AttrType::FLOAT: {
case proto::AttrType::FLOAT: {
return attr_desc.f();
}
case framework::AttrType::STRING: {
case proto::AttrType::STRING: {
return attr_desc.s();
}
case framework::AttrType::BOOLEANS: {
case proto::AttrType::BOOLEANS: {
std::vector<bool> val(attr_desc.bools_size());
for (int i = 0; i < attr_desc.bools_size(); ++i) {
val[i] = attr_desc.bools(i);
}
return val;
}
case framework::AttrType::INTS: {
case proto::AttrType::INTS: {
std::vector<int> val(attr_desc.ints_size());
for (int i = 0; i < attr_desc.ints_size(); ++i) {
val[i] = attr_desc.ints(i);
}
return val;
}
case framework::AttrType::FLOATS: {
case proto::AttrType::FLOATS: {
std::vector<float> val(attr_desc.floats_size());
for (int i = 0; i < attr_desc.floats_size(); ++i) {
val[i] = attr_desc.floats(i);
}
return val;
}
case framework::AttrType::STRINGS: {
case proto::AttrType::STRINGS: {
std::vector<std::string> val(attr_desc.strings_size());
for (int i = 0; i < attr_desc.strings_size(); ++i) {
val[i] = attr_desc.strings(i);
......
paddle/framework/attribute.h
浏览文件 @ ba8660a1
......@@ -27,12 +27,12 @@ limitations under the License. */
namespace paddle {
namespace framework {
template <typename T>
inline AttrType AttrTypeID() {
inline proto::AttrType AttrTypeID() {
Attribute tmp = T();
return static_cast<AttrType>(tmp.which() - 1);
return static_cast<proto::AttrType>(tmp.which() - 1);
}
Attribute GetAttrValue(const OpDesc::Attr& attr_desc);
Attribute GetAttrValue(const proto::OpDesc::Attr& attr_desc);
class AttrReader {
public:
......
paddle/framework/backward.cc
浏览文件 @ ba8660a1
......@@ -42,7 +42,7 @@ static std::unordered_set<std::string>& CtrlFlowOps() {
static inline std::unique_ptr<OperatorBase> CreateGradOp(
const OperatorBase& op, const std::unordered_set<std::string>& no_grad_set,
std::unordered_map<std::string, std::string>* grad_to_var) {
OpDescBind op_desc;
OpDesc op_desc;
op_desc.SetInputMap(op.Inputs());
op_desc.SetOutputMap(op.Outputs());
op_desc.SetType(op.Type());
......@@ -53,7 +53,7 @@ static inline std::unique_ptr<OperatorBase> CreateGradOp(
grad_ops.reserve(grad_descs.size());
std::transform(grad_descs.begin(), grad_descs.end(),
std::back_inserter(grad_ops),
[](const std::unique_ptr<OpDescBind>& grad_desc) {
[](const std::unique_ptr<OpDesc>& grad_desc) {
return OpRegistry::CreateOp(*grad_desc);
});
PADDLE_ENFORCE(!grad_ops.empty());
......@@ -217,7 +217,7 @@ static std::unique_ptr<OperatorBase> BackwardRecursive(
// If part of input gradient of that operator is not calculated, fill
// zero variables to that input gradient.
net->AppendOp(OpRegistry::CreateOp("fill_zeros_like", {{"X", {prefix}}},
{{"Y", {grad_input}}},
{{"Out", {grad_input}}},
AttributeMap{}));
}
return false;
......@@ -296,7 +296,7 @@ static std::string FwdName(const std::string& grad_name) {
static void CreateGradVarInBlock(
size_t grad_op_start_index,
const std::unordered_map<std::string, std::string>& param_name_map,
BlockDescBind* block_desc,
BlockDesc* block_desc,
std::unordered_map<std::string, GradVarInfo>* grad_var_record) {
auto ops = block_desc->AllOps();
for (size_t op_index = grad_op_start_index; op_index < ops.size();
......@@ -341,7 +341,7 @@ static void CreateGradVarInBlock(
auto* param = block_desc->FindVarRecursive(pname);
auto* grad = block_desc->FindVar(arg);
if (param == nullptr) {
grad->SetDataType(DataType::FP32);
grad->SetDataType(proto::DataType::FP32);
} else {
grad->SetDataType(param->GetDataType());
}
......@@ -350,12 +350,11 @@ static void CreateGradVarInBlock(
}
}
std::vector<std::unique_ptr<OpDescBind>> MakeOpGrad(
const OpDescBind* op_desc, std::unordered_set<std::string>* no_grad_vars,
std::vector<std::unique_ptr<OpDesc>> MakeOpGrad(
const OpDesc* op_desc, std::unordered_set<std::string>* no_grad_vars,
std::unordered_map<std::string, std::string>* grad_to_var,
const std::vector<BlockDescBind*>& grad_block =
std::vector<BlockDescBind*>()) {
std::vector<std::unique_ptr<OpDescBind>> grad_op_descs;
const std::vector<BlockDesc*>& grad_block = std::vector<BlockDesc*>()) {
std::vector<std::unique_ptr<OpDesc>> grad_op_descs;
// All input gradients of forwarding operator do not need to calculate.
const std::vector<std::string>& inputs = op_desc->InputArgumentNames();
if (AllGradInSet(inputs, *no_grad_vars)) {
......@@ -386,7 +385,7 @@ std::vector<std::unique_ptr<OpDescBind>> MakeOpGrad(
.Get(op_desc->Type())
.GradOpMaker()(*op_desc, *no_grad_vars, grad_to_var, grad_block);
std::list<std::unique_ptr<OpDescBind>> pending_fill_zeros_ops;
std::list<std::unique_ptr<OpDesc>> pending_fill_zeros_ops;
for (auto& desc : grad_op_descs) {
for (const std::string& in_name : desc->InputArgumentNames()) {
if (no_grad_vars->count(in_name)) {
......@@ -394,9 +393,9 @@ std::vector<std::unique_ptr<OpDescBind>> MakeOpGrad(
0, in_name.size() - sizeof(kGradVarSuffix) / sizeof(char) + 1);
std::string new_name = prefix + kZeroVarSuffix;
desc->Rename(in_name, new_name);
std::unique_ptr<OpDescBind> fill_zeros_op(
new OpDescBind("fill_zeros_like", {{"X", {prefix}}},
{{"Y", {new_name}}}, AttributeMap{}));
std::unique_ptr<OpDesc> fill_zeros_op(
new OpDesc("fill_zeros_like", {{"X", {prefix}}},
{{"Out", {new_name}}}, AttributeMap{}));
pending_fill_zeros_ops.push_back(std::move(fill_zeros_op));
}
}
......@@ -408,34 +407,33 @@ std::vector<std::unique_ptr<OpDescBind>> MakeOpGrad(
return grad_op_descs;
}
static BlockDescBind* CreateStepBlock(
ProgramDescBind& program_desc,
std::unordered_set<std::string>* no_grad_vars,
static BlockDesc* CreateStepBlock(
ProgramDesc& program_desc, std::unordered_set<std::string>* no_grad_vars,
std::unordered_map<std::string, std::string>* grad_to_var,
int step_block_idx);
std::vector<std::unique_ptr<OpDescBind>> MakeBlockBackward(
ProgramDescBind& program_desc, int block_idx,
std::vector<std::unique_ptr<OpDesc>> MakeBlockBackward(
ProgramDesc& program_desc, int block_idx,
std::unordered_set<std::string>* no_grad_vars,
std::unordered_map<std::string, std::string>* grad_to_var) {
VLOG(5) << "MakeBlockBackward";
BlockDescBind* cur_block = program_desc.MutableBlock(block_idx);
std::vector<OpDescBind*> op_descs = cur_block->AllOps();
BlockDesc* cur_block = program_desc.MutableBlock(block_idx);
std::vector<OpDesc*> op_descs = cur_block->AllOps();
std::unordered_map<std::string, std::vector<size_t>> dup_out_ops;
size_t grad_desc_idx = 0;
std::vector<std::unique_ptr<OpDescBind>> backward_descs;
std::vector<std::unique_ptr<OpDesc>> backward_descs;
for (auto it = op_descs.rbegin(); it != op_descs.rend(); ++it) {
VLOG(5) << "Making backward " << (*it)->Type() << " op";
std::vector<std::unique_ptr<OpDescBind>> op_grads;
std::vector<std::unique_ptr<OpDesc>> op_grads;
if ((*it)->Type() == "recurrent" || (*it)->Type() == "while") {
int step_block_idx = (*it)->GetBlockAttr("sub_block");
BlockDescBind* backward_block = CreateStepBlock(
program_desc, no_grad_vars, grad_to_var, step_block_idx);
BlockDesc* backward_block = CreateStepBlock(program_desc, no_grad_vars,
grad_to_var, step_block_idx);
op_grads = MakeOpGrad(*it, no_grad_vars, grad_to_var, {backward_block});
} else if ((*it)->Type() == "conditional_block") {
BlockDescBind* backward_block =
BlockDesc* backward_block =
CreateStepBlock(program_desc, no_grad_vars, grad_to_var,
(*it)->GetBlockAttr("sub_block"));
op_grads = MakeOpGrad(*it, no_grad_vars, grad_to_var, {backward_block});
......@@ -463,14 +461,14 @@ std::vector<std::unique_ptr<OpDescBind>> MakeBlockBackward(
}
++grad_desc_idx;
}
std::transform(
op_grads.begin(), op_grads.end(), std::back_inserter(backward_descs),
[](std::unique_ptr<OpDescBind>& ptr) { return std::move(ptr); });
std::transform(op_grads.begin(), op_grads.end(),
std::back_inserter(backward_descs),
[](std::unique_ptr<OpDesc>& ptr) { return std::move(ptr); });
}
VLOG(5) << "Appending Sums";
// Check whether some variables are written more than once
std::list<std::pair<size_t, std::unique_ptr<OpDescBind>>> pending_sum_ops;
std::list<std::pair<size_t, std::unique_ptr<OpDesc>>> pending_sum_ops;
for (const auto& dup : dup_out_ops) {
const std::string& out_name = dup.first;
const std::vector<size_t> dup_op = dup.second;
......@@ -486,18 +484,17 @@ std::vector<std::unique_ptr<OpDescBind>> MakeBlockBackward(
sum_op_inputs.emplace_back(new_name);
next_g_name = sum_op_inputs.back();
}
std::unique_ptr<OpDescBind> sum_op(
new OpDescBind("sum", {{"X", sum_op_inputs}}, {{"Out", {out_name}}},
AttributeMap{}));
std::unique_ptr<OpDesc> sum_op(new OpDesc("sum", {{"X", sum_op_inputs}},
{{"Out", {out_name}}},
AttributeMap{}));
pending_sum_ops.push_back({dup_op.back(), std::move(sum_op)});
}
}
pending_sum_ops.sort(
[](const std::pair<size_t, std::unique_ptr<OpDescBind>>& a,
const std::pair<size_t, std::unique_ptr<OpDescBind>>& b) {
return a.first > b.first;
});
pending_sum_ops.sort([](const std::pair<size_t, std::unique_ptr<OpDesc>>& a,
const std::pair<size_t, std::unique_ptr<OpDesc>>& b) {
return a.first > b.first;
});
for (auto& p : pending_sum_ops) {
backward_descs.insert(backward_descs.begin() + p.first + 1,
std::move(p.second));
......@@ -508,14 +505,13 @@ std::vector<std::unique_ptr<OpDescBind>> MakeBlockBackward(
return backward_descs;
}
static BlockDescBind* CreateStepBlock(
ProgramDescBind& program_desc,
std::unordered_set<std::string>* no_grad_vars,
static BlockDesc* CreateStepBlock(
ProgramDesc& program_desc, std::unordered_set<std::string>* no_grad_vars,
std::unordered_map<std::string, std::string>* grad_to_var,
int step_block_idx) {
auto backward_block_op_descs = MakeBlockBackward(program_desc, step_block_idx,
no_grad_vars, grad_to_var);
BlockDescBind* backward_block =
BlockDesc* backward_block =
program_desc.AppendBlock(*program_desc.MutableBlock(step_block_idx));
for (auto& ptr : backward_block_op_descs) {
backward_block->AppendAllocatedOp(move(ptr));
......@@ -524,7 +520,7 @@ static BlockDescBind* CreateStepBlock(
}
ParamGradInfoMap AppendBackward(
ProgramDescBind& program_desc, const VarDescBind& target,
ProgramDesc& program_desc, const VarDesc& target,
const std::unordered_set<std::string>& no_grad_vars) {
std::unordered_set<std::string> no_grad_var_names;
no_grad_var_names.reserve(no_grad_vars.size() + 1);
......@@ -541,11 +537,11 @@ ParamGradInfoMap AppendBackward(
PADDLE_ENFORCE(is_scalar, "target should be scalar");
VLOG(3) << "backward from loss=" << target.Name()
<< " data_type=" << target.GetDataType();
std::unique_ptr<OpDescBind> fill_one_op(
new OpDescBind("fill_constant", {}, {{"Out", {fill_one_op_out}}},
{{"shape", std::vector<int>{1}},
{"value", static_cast<float>(1.0)},
{"dtype", target.GetDataType()}}));
std::unique_ptr<OpDesc> fill_one_op(
new OpDesc("fill_constant", {}, {{"Out", {fill_one_op_out}}},
{{"shape", std::vector<int>{1}},
{"value", static_cast<float>(1.0)},
{"dtype", target.GetDataType()}}));
// infer var type of fill_one_op
fill_one_op->InferVarType(root_block);
......
paddle/framework/backward.h
浏览文件 @ ba8660a1
......@@ -49,7 +49,7 @@ using ParamGradInfoMap = std::unordered_map<std::string /*fwd_var_name*/,
GradVarInfo /*grad_var_info*/>;
ParamGradInfoMap AppendBackward(
ProgramDescBind& program_desc, const VarDescBind& target,
ProgramDesc& program_desc, const VarDesc& target,
const std::unordered_set<std::string>& no_grad_vars);
} // namespace framework
......
paddle/framework/backward_test.cc
浏览文件 @ ba8660a1
......@@ -58,13 +58,13 @@ class RowWiseAddGradMaker : public SingleGradOpDescMaker {
using SingleGradOpDescMaker::SingleGradOpDescMaker;
protected:
std::unique_ptr<OpDescBind> Apply() const override {
auto grad_op = new OpDescBind();
std::unique_ptr<OpDesc> Apply() const override {
auto grad_op = new OpDesc();
grad_op->SetInput(GradVarName("Out"), OutputGrad("Out"));
grad_op->SetOutput(GradVarName("X"), InputGrad("X"));
grad_op->SetOutput(GradVarName("b"), InputGrad("b"));
grad_op->SetType("rowwise_add_grad");
return std::unique_ptr<OpDescBind>(grad_op);
return std::unique_ptr<OpDesc>(grad_op);
}
};
......@@ -159,14 +159,14 @@ class FillZeroOpMaker : public OpProtoAndCheckerMaker {
FillZeroOpMaker(OpProto *proto, OpAttrChecker *op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("X", "x");
AddOutput("Y", "out");
AddOutput("Out", "out");
AddComment("");
}
};
class SumOpMaker : public framework::OpProtoAndCheckerMaker {
public:
SumOpMaker(framework::OpProto *proto, framework::OpAttrChecker *op_checker)
SumOpMaker(OpProto *proto, OpAttrChecker *op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("X", "the input tensors of sum operator.").AsDuplicable();
AddOutput("Out", "the output tensor of sum operator.");
......@@ -190,11 +190,11 @@ class MinusGradOpDescMaker : public GradOpDescMakerBase {
public:
using GradOpDescMakerBase::GradOpDescMakerBase;
std::vector<std::unique_ptr<OpDescBind>> operator()() const override {
std::vector<std::unique_ptr<OpDescBind>> retv;
std::vector<std::unique_ptr<OpDesc>> operator()() const override {
std::vector<std::unique_ptr<OpDesc>> retv;
auto x_g = InputGrad("X");
if (!x_g.empty()) {
auto *op_desc = new OpDescBind();
auto *op_desc = new OpDesc();
op_desc->SetType("scale");
op_desc->SetInput("X", OutputGrad("Out"));
op_desc->SetOutput("Out", x_g);
......@@ -204,7 +204,7 @@ class MinusGradOpDescMaker : public GradOpDescMakerBase {
auto y_g = InputGrad("Y");
if (!y_g.empty()) {
auto *op_desc = new OpDescBind();
auto *op_desc = new OpDesc();
op_desc->SetType("scale");
op_desc->SetInput("X", OutputGrad("Out"));
op_desc->SetOutput("Out", y_g);
......@@ -430,8 +430,8 @@ TEST(Backward, op_part_of_output_are_not_need) {
ASSERT_EQ("fill_zeros_like", fill_zero.Type());
ASSERT_EQ(1UL, fill_zero.Inputs("X").size());
ASSERT_EQ("Z", fill_zero.Input("X"));
ASSERT_EQ(1UL, fill_zero.Outputs("Y").size());
ASSERT_EQ(std::string("Z") + f::kZeroVarSuffix, fill_zero.Output("Y"));
ASSERT_EQ(1UL, fill_zero.Outputs("Out").size());
ASSERT_EQ(std::string("Z") + f::kZeroVarSuffix, fill_zero.Output("Out"));
auto &d_many_out = *net->ops_[1];
ASSERT_EQ("many_output_op_grad", d_many_out.Type());
......@@ -505,25 +505,25 @@ TEST(Backward, linear_net_intermediate_variable_has_no_grad) {
}
TEST(Backward, simple_single_op) {
f::ProgramDescBind program;
f::BlockDescBind *block = program.MutableBlock(0);
f::ProgramDesc program;
f::BlockDesc *block = program.MutableBlock(0);
f::OpDescBind *op = block->AppendOp();
f::OpDesc *op = block->AppendOp();
op->SetType("rowwise_add");
op->SetInput("X", {"x"});
op->SetInput("b", {"b"});
op->SetOutput("Out", {"out"});
auto target = f::VarDescBind("out");
auto target = f::VarDesc("out");
target.SetShape({1});
auto var_to_grad =
AppendBackward(program, target, std::unordered_set<std::string>{});
ASSERT_EQ(block->AllOps().size(), 3UL);
f::OpDescBind *fill_op = block->AllOps()[1];
f::OpDesc *fill_op = block->AllOps()[1];
EXPECT_EQ(fill_op->Type(), "fill_constant");
f::OpDescBind *grad_op = block->AllOps()[2];
f::OpDesc *grad_op = block->AllOps()[2];
EXPECT_EQ(grad_op->Type(), "rowwise_add_grad");
ASSERT_EQ(grad_op->InputNames().size(), 1UL);
ASSERT_EQ(grad_op->OutputNames().size(), 2UL);
......@@ -543,16 +543,16 @@ TEST(Backward, simple_single_op) {
}
TEST(Backward, default_attribute) {
f::ProgramDescBind program;
f::BlockDescBind *block = program.MutableBlock(0);
f::OpDescBind *op = block->AppendOp();
f::ProgramDesc program;
f::BlockDesc *block = program.MutableBlock(0);
f::OpDesc *op = block->AppendOp();
op->SetType("mul");
op->SetInput("X", {"x"});
op->SetInput("Y", {"y"});
op->SetOutput("Out", {"out"});
op->CheckAttrs();
auto target = f::VarDescBind("out");
auto target = f::VarDesc("out");
target.SetShape({1});
AppendBackward(program, target, std::unordered_set<std::string>{});
......@@ -560,47 +560,47 @@ TEST(Backward, default_attribute) {
EXPECT_EQ(boost::get<int>(op->GetAttr("x_num_col_dims")), 1);
EXPECT_EQ(boost::get<int>(op->GetAttr("y_num_col_dims")), 1);
f::OpDescBind *fill_op = block->AllOps()[1];
f::OpDesc *fill_op = block->AllOps()[1];
EXPECT_EQ(fill_op->Type(), "fill_constant");
f::OpDescBind *grad_op = block->AllOps()[2];
f::OpDesc *grad_op = block->AllOps()[2];
ASSERT_EQ(grad_op->Type(), "mul_grad");
EXPECT_EQ(boost::get<int>(grad_op->GetAttr("x_num_col_dims")), 1);
EXPECT_EQ(boost::get<int>(grad_op->GetAttr("y_num_col_dims")), 1);
}
TEST(Backward, simple_mult_op) {
f::ProgramDescBind program;
f::BlockDescBind *block = program.MutableBlock(0);
f::OpDescBind *op1 = block->AppendOp();
f::ProgramDesc program;
f::BlockDesc *block = program.MutableBlock(0);
f::OpDesc *op1 = block->AppendOp();
op1->SetType("rowwise_add");
op1->SetInput("X", {"x1"});
op1->SetInput("b", {"b1"});
op1->SetOutput("Out", {"out1"});
f::OpDescBind *op2 = block->AppendOp();
f::OpDesc *op2 = block->AppendOp();
op2->SetType("mul");
op2->SetInput("X", {"out1"});
op2->SetInput("Y", {"y2"});
op2->SetOutput("Out", {"out2"});
f::OpDescBind *op3 = block->AppendOp();
f::OpDesc *op3 = block->AppendOp();
op3->SetType("rowwise_add");
op3->SetInput("X", {"out2"});
op3->SetInput("b", {"b3"});
op3->SetOutput("Out", {"out3"});
auto target = f::VarDescBind("out3");
auto target = f::VarDesc("out3");
target.SetShape({1});
size_t forward_len = block->AllOps().size();
auto var_to_grad =
AppendBackward(program, target, std::unordered_set<std::string>{});
ASSERT_EQ(block->AllOps().size(), 6UL + 1);
f::OpDescBind *fill_op = block->AllOps()[forward_len];
f::OpDesc *fill_op = block->AllOps()[forward_len];
EXPECT_EQ(fill_op->Type(), "fill_constant");
f::OpDescBind *grad_op1 = block->AllOps()[6];
f::OpDesc *grad_op1 = block->AllOps()[6];
EXPECT_EQ(grad_op1->Type(), "rowwise_add_grad");
ASSERT_EQ(grad_op1->InputNames().size(), 1UL);
ASSERT_EQ(grad_op1->OutputNames().size(), 2UL);
......@@ -611,7 +611,7 @@ TEST(Backward, simple_mult_op) {
EXPECT_EQ(grad_op1->Output(f::GradVarName("b")),
std::vector<std::string>({f::GradVarName("b1")}));
f::OpDescBind *grad_op2 = block->AllOps()[5];
f::OpDesc *grad_op2 = block->AllOps()[5];
EXPECT_EQ(grad_op2->Type(), "mul_grad");
ASSERT_EQ(grad_op2->InputNames().size(), 4UL);
ASSERT_EQ(grad_op2->OutputNames().size(), 2UL);
......@@ -625,7 +625,7 @@ TEST(Backward, simple_mult_op) {
EXPECT_EQ(grad_op2->Output(f::GradVarName("Y")),
std::vector<std::string>({f::GradVarName("y2")}));
f::OpDescBind *grad_op3 = block->AllOps()[4];
f::OpDesc *grad_op3 = block->AllOps()[4];
EXPECT_EQ(grad_op3->Type(), "rowwise_add_grad");
ASSERT_EQ(grad_op3->InputNames().size(), 1UL);
ASSERT_EQ(grad_op3->OutputNames().size(), 2UL);
......@@ -655,42 +655,42 @@ TEST(Backward, simple_mult_op) {
}
TEST(Backward, intermedia_var_no_grad) {
f::ProgramDescBind program;
f::BlockDescBind *block = program.MutableBlock(0);
f::OpDescBind *op1 = block->AppendOp();
f::ProgramDesc program;
f::BlockDesc *block = program.MutableBlock(0);
f::OpDesc *op1 = block->AppendOp();
op1->SetType("rowwise_add");
op1->SetInput("X", {"x1"});
op1->SetInput("b", {"b1"});
op1->SetOutput("Out", {"out1"});
f::OpDescBind *op2 = block->AppendOp();
f::OpDesc *op2 = block->AppendOp();
op2->SetType("mul");
op2->SetInput("X", {"x2"});
op2->SetInput("Y", {"y2"});
op2->SetOutput("Out", {"out2"});
f::OpDescBind *op3 = block->AppendOp();
f::OpDesc *op3 = block->AppendOp();
op3->SetType("rowwise_add");
op3->SetInput("X", {"out2"});
op3->SetInput("b", {"b3"});
op3->SetOutput("Out", {"out3"});
f::OpDescBind *op4 = block->AppendOp();
f::OpDesc *op4 = block->AppendOp();
op4->SetType("mul");
op4->SetInput("X", {"out1"});
op4->SetInput("Y", {"out3"});
op4->SetOutput("Out", {"out4"});
auto target = f::VarDescBind("out4");
auto target = f::VarDesc("out4");
target.SetShape({1});
size_t forward_len = block->AllOps().size();
auto var_to_grad = AppendBackward(program, target, {"out3"});
ASSERT_EQ(block->AllOps().size(), 7UL);
f::OpDescBind *fill_op = block->AllOps()[forward_len];
f::OpDesc *fill_op = block->AllOps()[forward_len];
EXPECT_EQ(fill_op->Type(), "fill_constant");
f::OpDescBind *grad_op1 = block->AllOps()[6];
f::OpDesc *grad_op1 = block->AllOps()[6];
EXPECT_EQ(grad_op1->Type(), "rowwise_add_grad");
ASSERT_EQ(grad_op1->InputNames().size(), 1UL);
ASSERT_EQ(grad_op1->OutputNames().size(), 2UL);
......@@ -701,7 +701,7 @@ TEST(Backward, intermedia_var_no_grad) {
EXPECT_EQ(grad_op1->Output(f::GradVarName("b")),
std::vector<std::string>({f::GradVarName("b1")}));
f::OpDescBind *grad_op4 = block->AllOps()[5];
f::OpDesc *grad_op4 = block->AllOps()[5];
EXPECT_EQ(grad_op4->Type(), "mul_grad");
ASSERT_EQ(grad_op4->InputNames().size(), 4UL);
ASSERT_EQ(grad_op4->OutputNames().size(), 2UL);
......@@ -726,32 +726,32 @@ TEST(Backward, intermedia_var_no_grad) {
}
TEST(Backward, var_no_grad) {
f::ProgramDescBind program;
f::BlockDescBind *block = program.MutableBlock(0);
f::OpDescBind *op1 = block->AppendOp();
f::ProgramDesc program;
f::BlockDesc *block = program.MutableBlock(0);
f::OpDesc *op1 = block->AppendOp();
op1->SetType("mult_in_out");
op1->SetInput("X", {"x1"});
op1->SetInput("H", {"h1"});
op1->SetOutput("Y", {"y1"});
op1->SetOutput("Z", {"z1"});
f::OpDescBind *op2 = block->AppendOp();
f::OpDesc *op2 = block->AppendOp();
op2->SetType("mult_in_out");
op2->SetInput("X", {"y1"});
op2->SetInput("H", {"z1"});
op2->SetOutput("Y", {"y2"});
op2->SetOutput("Z", {"z2"});
auto target = f::VarDescBind("z2");
auto target = f::VarDesc("z2");
target.SetShape({1});
size_t forward_len = block->AllOps().size();
auto var_to_grad = AppendBackward(program, target, {"z1"});
ASSERT_EQ(block->AllOps().size(), 6UL);
f::OpDescBind *fill_op = block->AllOps()[forward_len];
f::OpDesc *fill_op = block->AllOps()[forward_len];
EXPECT_EQ(fill_op->Type(), "fill_constant");
f::OpDescBind *grad_op2 = block->AllOps()[3];
f::OpDesc *grad_op2 = block->AllOps()[3];
ASSERT_EQ(grad_op2->Type(), "mult_in_out_grad");
ASSERT_EQ(grad_op2->InputNames().size(), 6UL);
ASSERT_EQ(grad_op2->OutputNames().size(), 2UL);
......@@ -767,15 +767,15 @@ TEST(Backward, var_no_grad) {
std::vector<std::string>({f::GradVarName("y1")}));
EXPECT_EQ(grad_op2->Output(f::GradVarName("H")), std::vector<std::string>());
f::OpDescBind *fill_zero_op = block->AllOps()[4];
f::OpDesc *fill_zero_op = block->AllOps()[4];
ASSERT_EQ(fill_zero_op->Type(), "fill_zeros_like");
ASSERT_EQ(fill_zero_op->InputNames().size(), 1UL);
ASSERT_EQ(fill_zero_op->OutputNames().size(), 1UL);
EXPECT_EQ(fill_zero_op->Input("X"), std::vector<std::string>({"z1"}));
EXPECT_EQ(fill_zero_op->Output("Y"),
EXPECT_EQ(fill_zero_op->Output("Out"),
std::vector<std::string>({std::string("z1") + f::kZeroVarSuffix}));
f::OpDescBind *grad_op1 = block->AllOps()[5];
f::OpDesc *grad_op1 = block->AllOps()[5];
ASSERT_EQ(grad_op1->Type(), "mult_in_out_grad");
ASSERT_EQ(grad_op1->InputNames().size(), 6UL);
ASSERT_EQ(grad_op1->OutputNames().size(), 2UL);
......@@ -803,37 +803,37 @@ TEST(Backward, var_no_grad) {
}
TEST(Backward, shared_var) {
f::ProgramDescBind program;
f::BlockDescBind *block = program.MutableBlock(0);
f::OpDescBind *op1 = block->AppendOp();
f::ProgramDesc program;
f::BlockDesc *block = program.MutableBlock(0);
f::OpDesc *op1 = block->AppendOp();
op1->SetType("rowwise_add");
op1->SetInput("X", {"x1"});
op1->SetInput("b", {"b1"});
op1->SetOutput("Out", {"out1"});
f::OpDescBind *op2 = block->AppendOp();
f::OpDesc *op2 = block->AppendOp();
op2->SetType("mul");
op2->SetInput("X", {"out1"});
op2->SetInput("Y", {"y2"});
op2->SetOutput("Out", {"out2"});
f::OpDescBind *op3 = block->AppendOp();
f::OpDesc *op3 = block->AppendOp();
op3->SetType("rowwise_add");
op3->SetInput("X", {"out1"});
op3->SetInput("b", {"b3"});
op3->SetOutput("Out", {"out3"});
auto target = f::VarDescBind("out3");
auto target = f::VarDesc("out3");
target.SetShape({1});
size_t forward_len = block->AllOps().size();
auto var_to_grad =
AppendBackward(program, target, std::unordered_set<std::string>{});
ASSERT_EQ(block->AllOps().size(), 8UL);
f::OpDescBind *fill_op = block->AllOps()[forward_len];
f::OpDesc *fill_op = block->AllOps()[forward_len];
EXPECT_EQ(fill_op->Type(), "fill_constant");
f::OpDescBind *grad_op3 = block->AllOps()[4];
f::OpDesc *grad_op3 = block->AllOps()[4];
ASSERT_EQ(grad_op3->Type(), "rowwise_add_grad");
ASSERT_EQ(grad_op3->InputNames().size(), 1UL);
ASSERT_EQ(grad_op3->OutputNames().size(), 2UL);
......@@ -844,7 +844,7 @@ TEST(Backward, shared_var) {
EXPECT_EQ(grad_op3->Output(f::GradVarName("b")),
std::vector<std::string>({f::GradVarName("b3")}));
f::OpDescBind *grad_op4 = block->AllOps()[5];
f::OpDesc *grad_op4 = block->AllOps()[5];
ASSERT_EQ(grad_op4->Type(), "mul_grad");
ASSERT_EQ(grad_op4->InputNames().size(), 4UL);
ASSERT_EQ(grad_op4->OutputNames().size(), 2UL);
......@@ -858,7 +858,7 @@ TEST(Backward, shared_var) {
EXPECT_EQ(grad_op4->Output(f::GradVarName("Y")),
std::vector<std::string>({f::GradVarName("y2")}));
f::OpDescBind *sum_op = block->AllOps()[6];
f::OpDesc *sum_op = block->AllOps()[6];
ASSERT_EQ(sum_op->Type(), "sum");
ASSERT_EQ(sum_op->InputNames().size(), 1UL);
ASSERT_EQ(sum_op->OutputNames().size(), 1UL);
......@@ -868,7 +868,7 @@ TEST(Backward, shared_var) {
EXPECT_EQ(sum_op->Output("Out"),
std::vector<std::string>({f::GradVarName("out1")}));
f::OpDescBind *grad_op1 = block->AllOps()[7];
f::OpDesc *grad_op1 = block->AllOps()[7];
ASSERT_EQ(grad_op1->Type(), "rowwise_add_grad");
ASSERT_EQ(grad_op1->InputNames().size(), 1UL);
ASSERT_EQ(grad_op1->OutputNames().size(), 2UL);
......@@ -895,19 +895,19 @@ TEST(Backward, shared_var) {
}
TEST(Backward, half_backward) {
f::ProgramDescBind program;
f::BlockDescBind *block = program.MutableBlock(0);
f::ProgramDesc program;
f::BlockDesc *block = program.MutableBlock(0);
auto *op1 = block->AppendOp();
op1->SetType("minus");
op1->SetInput("X", {"a"});
op1->SetInput("Y", {"b"});
op1->SetOutput("Out", {"out"});
auto target = f::VarDescBind("out");
auto target = f::VarDesc("out");
target.SetShape({1});
size_t forward_len = block->AllOps().size();
auto var_to_grad = AppendBackward(program, target, {"b"});
f::OpDescBind *fill_op = block->AllOps()[forward_len];
f::OpDesc *fill_op = block->AllOps()[forward_len];
EXPECT_EQ(fill_op->Type(), "fill_constant");
auto ops = block->AllOps();
ASSERT_EQ(3UL, ops.size());
......
paddle/framework/block_desc.cc
浏览文件 @ ba8660a1
......@@ -19,18 +19,18 @@ limitations under the License. */
namespace paddle {
namespace framework {
VarDescBind *BlockDescBind::Var(const std::string &name) {
VarDesc *BlockDesc::Var(const std::string &name) {
auto it = vars_.find(name);
if (it != vars_.end()) {
return it->second.get();
}
need_update_ = true;
auto *var = new VarDescBind(name);
auto *var = new VarDesc(name);
vars_[name].reset(var);
return var;
}
VarDescBind *BlockDescBind::FindVar(const std::string &name) const {
VarDesc *BlockDesc::FindVar(const std::string &name) const {
auto it = vars_.find(name);
if (it == vars_.end()) {
return nullptr;
......@@ -38,11 +38,11 @@ VarDescBind *BlockDescBind::FindVar(const std::string &name) const {
return it->second.get();
}
bool BlockDescBind::HasVar(const std::string &name) const {
bool BlockDesc::HasVar(const std::string &name) const {
return vars_.find(name) != vars_.end();
}
VarDescBind *BlockDescBind::FindVarRecursive(const std::string &name) const {
VarDesc *BlockDesc::FindVarRecursive(const std::string &name) const {
if (name == kEmptyVarName) return nullptr;
auto it = vars_.find(name);
......@@ -53,53 +53,67 @@ VarDescBind *BlockDescBind::FindVarRecursive(const std::string &name) const {
return it->second.get();
}
VarDescBind *BlockDescBind::FindRecursiveOrCreateVar(
const std::string &name_bytes) {
VarDescBind *res = FindVarRecursive(name_bytes);
VarDesc *BlockDesc::FindRecursiveOrCreateVar(const std::string &name_bytes) {
VarDesc *res = FindVarRecursive(name_bytes);
if (res == nullptr) {
res = Var(name_bytes);
}
return res;
}
bool BlockDescBind::HasVarRecursive(const std::string &name) const {
bool BlockDesc::HasVarRecursive(const std::string &name) const {
return FindVarRecursive(name) != nullptr;
}
std::vector<VarDescBind *> BlockDescBind::AllVars() const {
std::vector<VarDescBind *> res;
std::vector<VarDesc *> BlockDesc::AllVars() const {
std::vector<VarDesc *> res;
for (const auto &p : vars_) {
res.push_back(p.second.get());
}
return res;
}
OpDescBind *BlockDescBind::AppendOp() {
OpDesc *BlockDesc::AppendOp() {
need_update_ = true;
ops_.emplace_back(new OpDescBind());
ops_.emplace_back(new OpDesc());
return ops_.back().get();
}
void BlockDescBind::AppendAllocatedOp(std::unique_ptr<OpDescBind> &&op_desc) {
void BlockDesc::AppendAllocatedOp(std::unique_ptr<OpDesc> &&op_desc) {
need_update_ = true;
ops_.emplace_back(std::move(op_desc));
}
OpDescBind *BlockDescBind::PrependOp() {
OpDesc *BlockDesc::PrependOp() {
need_update_ = true;
ops_.emplace_front(new OpDescBind());
ops_.emplace_front(new OpDesc());
return ops_.front().get();
}
std::vector<OpDescBind *> BlockDescBind::AllOps() const {
std::vector<OpDescBind *> res;
void BlockDesc::RemoveOp(size_t s, size_t e) {
if (ops_.begin() + s == ops_.end() || ops_.begin() + e == ops_.end()) {
return;
}
need_update_ = true;
for (auto it = ops_.begin() + s; it != ops_.begin() + e; it++) {
auto names = (*it)->InputArgumentNames();
for (auto n : names) {
// TODO(typhoonzero): delete vars if no other op use it.
VLOG(3) << "deleting var " << n;
}
}
ops_.erase(ops_.begin() + s, ops_.begin() + e);
}
std::vector<OpDesc *> BlockDesc::AllOps() const {
std::vector<OpDesc *> res;
for (const auto &op : ops_) {
res.push_back(op.get());
}
return res;
}
void BlockDescBind::Flush() {
void BlockDesc::Flush() {
for (auto &op_desc : ops_) {
op_desc->Flush();
}
......@@ -121,43 +135,43 @@ void BlockDescBind::Flush() {
}
}
BlockDescBind *BlockDescBind::ParentBlock() const {
BlockDesc *BlockDesc::ParentBlock() const {
if (this->desc_->parent_idx() == kNoneBlockIndex) {
return nullptr;
}
return prog_->MutableBlock(static_cast<size_t>(this->desc_->parent_idx()));
}
BlockDesc *BlockDescBind::Proto() {
proto::BlockDesc *BlockDesc::Proto() {
Flush();
return desc_;
}
BlockDescBind::BlockDescBind(ProgramDescBind *prog, BlockDesc *desc)
BlockDesc::BlockDesc(ProgramDesc *prog, proto::BlockDesc *desc)
: prog_(prog), desc_(desc), need_update_(false) {
for (const VarDesc &var_desc : desc_->vars()) {
vars_[var_desc.name()].reset(new VarDescBind(var_desc));
for (const proto::VarDesc &var_desc : desc_->vars()) {
vars_[var_desc.name()].reset(new VarDesc(var_desc));
}
for (const OpDesc &op_desc : desc_->ops()) {
ops_.emplace_back(new OpDescBind(op_desc, prog));
for (const proto::OpDesc &op_desc : desc_->ops()) {
ops_.emplace_back(new OpDesc(op_desc, prog));
}
}
BlockDescBind::BlockDescBind(const BlockDescBind &other, BlockDesc *desc,
ProgramDescBind *prog)
BlockDesc::BlockDesc(const BlockDesc &other, proto::BlockDesc *desc,
ProgramDesc *prog)
: prog_(prog), desc_(desc) {
need_update_ = true;
for (auto &op : other.ops_) {
ops_.emplace_back(new OpDescBind(*op));
ops_.emplace_back(new OpDesc(*op));
}
for (auto &it : other.vars_) {
auto *var = new VarDescBind(*it.second);
auto *var = new VarDesc(*it.second);
vars_[it.first].reset(var);
}
}
void BlockDescBind::ClearPBOps() {
void BlockDesc::ClearPBOps() {
auto ops = this->desc_->mutable_ops();
while (!ops->empty()) {
// we do not own the OpDesc, so release the ownership.
......@@ -165,7 +179,7 @@ void BlockDescBind::ClearPBOps() {
}
}
void BlockDescBind::ClearPBVars() {
void BlockDesc::ClearPBVars() {
auto vars = this->desc_->mutable_vars();
while (!vars->empty()) {
// we do not own the VarDesc, so release the ownership.
......
paddle/framework/block_desc.h
浏览文件 @ ba8660a1
......@@ -28,20 +28,19 @@ limitations under the License. */
namespace paddle {
namespace framework {
class ProgramDescBind;
class ProgramDesc;
// Each Protobuf Message, we provide a XXXBind class. In that class, we optimize
// read/write speed. Only when we want the protobuf message, the local changes
// will be synchronized (by `Sync` method).
class BlockDescBind {
class BlockDesc {
public:
BlockDescBind(ProgramDescBind *prog, BlockDesc *desc);
BlockDesc(ProgramDesc *prog, proto::BlockDesc *desc);
BlockDescBind(const BlockDescBind &other, BlockDesc *desc,
ProgramDescBind *prog);
BlockDesc(const BlockDesc &other, proto::BlockDesc *desc, ProgramDesc *prog);
~BlockDescBind() {
~BlockDesc() {
this->ClearPBVars();
this->ClearPBOps();
}
......@@ -50,15 +49,15 @@ class BlockDescBind {
int32_t Parent() const { return desc_->parent_idx(); }
VarDescBind *Var(const std::string &name_bytes);
VarDesc *Var(const std::string &name_bytes);
VarDescBind *FindVar(const std::string &name_bytes) const;
VarDesc *FindVar(const std::string &name_bytes) const;
bool HasVar(const std::string &var_name) const;
VarDescBind *FindVarRecursive(const std::string &name_bytes) const;
VarDesc *FindVarRecursive(const std::string &name_bytes) const;
VarDescBind *FindRecursiveOrCreateVar(const std::string &name_bytes);
VarDesc *FindRecursiveOrCreateVar(const std::string &name_bytes);
bool HasVarRecursive(const std::string &var_name) const;
......@@ -70,41 +69,43 @@ class BlockDescBind {
return var_names;
}
std::vector<VarDescBind *> AllVars() const;
std::vector<VarDesc *> AllVars() const;
BlockDescBind *ParentBlock() const;
BlockDesc *ParentBlock() const;
OpDescBind *AppendOp();
OpDesc *AppendOp();
void AppendAllocatedOp(std::unique_ptr<OpDescBind> &&op_desc);
void AppendAllocatedOp(std::unique_ptr<OpDesc> &&op_desc);
OpDescBind *PrependOp();
OpDesc *PrependOp();
std::vector<OpDescBind *> AllOps() const;
void RemoveOp(size_t s, size_t e);
std::vector<OpDesc *> AllOps() const;
size_t OpSize() const { return ops_.size(); }
OpDescBind *Op(int idx) { return ops_.at(idx).get(); }
OpDesc *Op(int idx) { return ops_.at(idx).get(); }
void Flush();
BlockDesc *Proto();
proto::BlockDesc *Proto();
ProgramDescBind *Program() { return this->prog_; }
ProgramDesc *Program() { return this->prog_; }
private:
void ClearPBOps();
void ClearPBVars();
private:
ProgramDescBind *prog_; // not_own
BlockDesc *desc_; // not_own
ProgramDesc *prog_; // not_own
proto::BlockDesc *desc_; // not_own
bool need_update_;
std::deque<std::unique_ptr<OpDescBind>> ops_;
std::unordered_map<std::string, std::unique_ptr<VarDescBind>> vars_;
std::deque<std::unique_ptr<OpDesc>> ops_;
std::unordered_map<std::string, std::unique_ptr<VarDesc>> vars_;
DISABLE_COPY_AND_ASSIGN(BlockDescBind);
DISABLE_COPY_AND_ASSIGN(BlockDesc);
};
} // namespace framework
} // namespace paddle
paddle/framework/data_layout.h 0 → 100644
浏览文件 @ ba8660a1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
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. */
#pragma once
namespace paddle {
namespace framework {
enum DataLayout {
kNHWC = 0,
kNCHW = 1,
kAnyLayout = 2,
};
inline DataLayout StringToDataLayout(const std::string& str) {
if (str == "NHWC" || str == "nhwc") {
return DataLayout::kNHWC;
} else if (str == "NCHW" || str == "nchw") {
return DataLayout::kNCHW;
} else {
PADDLE_THROW("Unknown storage order string: %s", str);
}
}
} // namespace framework
} // namespace paddle
paddle/framework/data_type.h
浏览文件 @ ba8660a1
......@@ -20,7 +20,8 @@
namespace paddle {
namespace framework {
inline DataType ToDataType(std::type_index type) {
inline proto::DataType ToDataType(std::type_index type) {
using namespace paddle::framework::proto;
if (typeid(float).hash_code() == type.hash_code()) {
return DataType::FP32;
} else if (typeid(double).hash_code() == type.hash_code()) {
......@@ -36,7 +37,8 @@ inline DataType ToDataType(std::type_index type) {
}
}
inline std::type_index ToTypeIndex(DataType type) {
inline std::type_index ToTypeIndex(proto::DataType type) {
using namespace paddle::framework::proto;
switch (type) {
case DataType::FP32:
return typeid(float);
......@@ -54,7 +56,8 @@ inline std::type_index ToTypeIndex(DataType type) {
}
template <typename Visitor>
inline void VisitDataType(DataType type, Visitor visitor) {
inline void VisitDataType(proto::DataType type, Visitor visitor) {
using namespace paddle::framework::proto;
switch (type) {
case DataType::FP32:
visitor.template operator()<float>();
......
paddle/framework/details/op_registry.h
浏览文件 @ ba8660a1
......@@ -90,7 +90,7 @@ struct OpInfoFiller<T, kOperator> {
template <typename T>
struct OpInfoFiller<T, kOpProtoAndCheckerMaker> {
void operator()(const char* op_type, OpInfo* info) const {
info->proto_ = new OpProto;
info->proto_ = new proto::OpProto;
info->checker_ = new OpAttrChecker();
auto maker = T(info->proto_, info->checker_);
maker.Validate();
......@@ -106,10 +106,10 @@ template <typename T>
struct OpInfoFiller<T, kGradOpDescMaker> {
void operator()(const char* op_type, OpInfo* info) const {
info->grad_op_maker_ = [](
const OpDescBind& fwd_op,
const OpDesc& fwd_op,
const std::unordered_set<std::string>& no_grad_set,
std::unordered_map<std::string, std::string>* grad_to_var,
const std::vector<BlockDescBind*>& grad_block) {
const std::vector<BlockDesc*>& grad_block) {
T maker(fwd_op, no_grad_set, grad_to_var, grad_block);
return maker();
};
......@@ -119,7 +119,7 @@ struct OpInfoFiller<T, kGradOpDescMaker> {
template <typename T>
struct OpInfoFiller<T, kVarTypeInference> {
void operator()(const char* op_type, OpInfo* info) const {
info->infer_var_type_ = [](const OpDescBind& fwd_op, BlockDescBind* block) {
info->infer_var_type_ = [](const OpDesc& fwd_op, BlockDesc* block) {
T inference;
inference(fwd_op, block);
};
......
paddle/framework/executor.cc
浏览文件 @ ba8660a1
......@@ -41,20 +41,20 @@ Executor::Executor(const std::vector<platform::Place>& places) {
device_contexts_.swap(borrowed_contexts);
}
static void CreateTensor(Variable* var, VarDesc::VarType var_type) {
if (var_type == VarDesc::LOD_TENSOR) {
static void CreateTensor(Variable* var, proto::VarDesc::VarType var_type) {
if (var_type == proto::VarDesc::LOD_TENSOR) {
var->GetMutable<LoDTensor>();
} else if (var_type == VarDesc::SELECTED_ROWS) {
} else if (var_type == proto::VarDesc::SELECTED_ROWS) {
var->GetMutable<SelectedRows>();
} else if (var_type == VarDesc::FEED_MINIBATCH) {
} else if (var_type == proto::VarDesc::FEED_MINIBATCH) {
var->GetMutable<FeedFetchList>();
} else if (var_type == VarDesc::FETCH_LIST) {
} else if (var_type == proto::VarDesc::FETCH_LIST) {
var->GetMutable<FeedFetchList>();
} else if (var_type == VarDesc::STEP_SCOPES) {
} else if (var_type == proto::VarDesc::STEP_SCOPES) {
var->GetMutable<std::vector<framework::Scope>>();
} else if (var_type == VarDesc::LOD_RANK_TABLE) {
} else if (var_type == proto::VarDesc::LOD_RANK_TABLE) {
var->GetMutable<LoDRankTable>();
} else if (var_type == VarDesc::LOD_TENSOR_ARRAY) {
} else if (var_type == proto::VarDesc::LOD_TENSOR_ARRAY) {
var->GetMutable<LoDTensorArray>();
} else {
PADDLE_THROW(
......@@ -64,8 +64,8 @@ static void CreateTensor(Variable* var, VarDesc::VarType var_type) {
}
}
void Executor::Run(const ProgramDescBind& pdesc, Scope* scope, int block_id,
bool create_local_scope) {
void Executor::Run(const ProgramDesc& pdesc, Scope* scope, int block_id,
bool create_local_scope, bool create_vars) {
// TODO(tonyyang-svail):
// - only runs on the first device (i.e. no interdevice communication)
// - will change to use multiple blocks for RNN op and Cond Op
......@@ -74,33 +74,35 @@ void Executor::Run(const ProgramDescBind& pdesc, Scope* scope, int block_id,
auto& device = device_contexts_[0];
Scope* local_scope = scope;
if (create_local_scope) {
local_scope = &scope->NewScope();
for (auto& var : block.AllVars()) {
if (var->Name() == framework::kEmptyVarName) {
continue;
if (create_vars) {
if (create_local_scope) {
local_scope = &scope->NewScope();
for (auto& var : block.AllVars()) {
if (var->Name() == framework::kEmptyVarName) {
continue;
}
if (var->Persistable()) {
auto* ptr = scope->Var(var->Name());
CreateTensor(ptr, var->GetType());
VLOG(3) << "Create Variable " << var->Name()
<< " global, which pointer is " << ptr;
} else {
auto* ptr = local_scope->Var(var->Name());
CreateTensor(ptr, var->GetType());
VLOG(3) << "Create Variable " << var->Name()
<< " locally, which pointer is " << ptr;
}
}
if (var->Persistable()) {
auto* ptr = scope->Var(var->Name());
CreateTensor(ptr, var->GetType());
VLOG(3) << "Create Variable " << var->Name()
<< " global, which pointer is " << ptr;
} else {
} else {
for (auto& var : block.AllVars()) {
auto* ptr = local_scope->Var(var->Name());
CreateTensor(ptr, var->GetType());
VLOG(3) << "Create Variable " << var->Name()
<< " locally, which pointer is " << ptr;
VLOG(3) << "Create variable " << var->Name() << ", which pointer is "
<< ptr;
}
}
} else {
for (auto& var : block.AllVars()) {
auto* ptr = local_scope->Var(var->Name());
CreateTensor(ptr, var->GetType());
VLOG(3) << "Create variable " << var->Name() << ", which pointer is "
<< ptr;
}
}
} // if (create_local_scope)
} // if (create_vars)
for (auto& op_desc : block.AllOps()) {
auto op = paddle::framework::OpRegistry::CreateOp(*op_desc);
......
paddle/framework/executor.h
浏览文件 @ ba8660a1
......@@ -40,6 +40,16 @@ class DeviceContextPool {
return *pool;
}
const platform::DeviceContext* Borrow(const platform::Place& place) {
auto range = device_contexts_.equal_range(place);
if (range.first == range.second) {
PADDLE_THROW(
"'Place' is not supported, Please re-compile with WITH_GPU "
"option");
}
return range.first->second;
}
std::vector<const platform::DeviceContext*> Borrow(
const std::vector<platform::Place>& places) {
PADDLE_ENFORCE_GT(places.size(), 0);
......@@ -114,7 +124,8 @@ class Executor {
* ProgramDesc
* Scope
*/
void Run(const ProgramDescBind&, Scope*, int, bool create_local_scope = true);
void Run(const ProgramDesc&, Scope*, int, bool create_local_scope = true,
bool create_vars = true);
private:
std::vector<const platform::DeviceContext*> device_contexts_;
......
paddle/framework/framework.proto
浏览文件 @ ba8660a1
......@@ -14,7 +14,7 @@ limitations under the License. */
syntax = "proto2";
option optimize_for = LITE_RUNTIME;
package paddle.framework;
package paddle.framework.proto;
enum AttrType {
INT = 0;
......
paddle/framework/grad_op_desc_maker.h
浏览文件 @ ba8660a1
......@@ -22,21 +22,27 @@
namespace paddle {
namespace framework {
/*
This functor class is responsible for creating the gradient ops for the given
operator fwd_op. After it is called (through operator()), the pairs of
(gradient variable, corresponding input variable of fwd_op) will be added to
grad_to_var. If an input variable of fwd_op is contained in no_grad_set, its
gradient varialbe will be ignored or kEmptyVarName depending on the template
argument DropEmptyIG in the derived classes.
*/
class GradOpDescMakerBase {
public:
explicit GradOpDescMakerBase(
const OpDescBind& fwd_op,
const std::unordered_set<std::string>& no_grad_set,
const OpDesc& fwd_op, const std::unordered_set<std::string>& no_grad_set,
std::unordered_map<std::string, std::string>* grad_to_var,
const std::vector<BlockDescBind*>& grad_block =
std::vector<BlockDescBind*>())
const std::vector<BlockDesc*>& grad_block = std::vector<BlockDesc*>())
: fwd_op_(fwd_op),
no_grad_set_(no_grad_set),
grad_to_var_(grad_to_var),
grad_block_(grad_block) {}
virtual ~GradOpDescMakerBase() = default;
virtual std::vector<std::unique_ptr<OpDescBind>> operator()() const = 0;
virtual std::vector<std::unique_ptr<OpDesc>> operator()() const = 0;
protected:
std::vector<std::string> InputGrad(const std::string& name,
......@@ -58,6 +64,16 @@ class GradOpDescMakerBase {
if (!drop_empty_grad) {
return ret_val;
}
PADDLE_ENFORCE_LE(var_names.size(), 1UL,
"BUG from operator developer:"
" for input argument with a list of variables, "
" drop_empty_grad is not allowed because it makes"
" the correspondence bewteen a variable and its gradient"
" ambiguous. Use REGISTER_OP_EX to register the op"
" or call InputGrad(?,false) in GradOpDescMaker."
" Op type %s",
fwd_op_.Type());
std::vector<std::string> dropped_ret_val;
dropped_ret_val.reserve(ret_val.size());
std::copy_if(ret_val.begin(), ret_val.end(),
......@@ -105,26 +121,26 @@ class GradOpDescMakerBase {
std::string ForwardOpType() const { return this->fwd_op_.Type(); }
private:
const OpDescBind& fwd_op_;
const OpDesc& fwd_op_;
const std::unordered_set<std::string>& no_grad_set_;
std::unordered_map<std::string, std::string>* grad_to_var_;
protected:
std::vector<BlockDescBind*> grad_block_;
std::vector<BlockDesc*> grad_block_;
};
class SingleGradOpDescMaker : public GradOpDescMakerBase {
public:
using GradOpDescMakerBase::GradOpDescMakerBase;
std::vector<std::unique_ptr<OpDescBind>> operator()() const {
std::vector<std::unique_ptr<OpDescBind>> retv;
std::vector<std::unique_ptr<OpDesc>> operator()() const {
std::vector<std::unique_ptr<OpDesc>> retv;
retv.emplace_back(this->Apply());
return retv;
}
protected:
virtual std::unique_ptr<OpDescBind> Apply() const = 0;
virtual std::unique_ptr<OpDesc> Apply() const = 0;
};
template <bool DropEmptyIG = true>
......@@ -133,8 +149,8 @@ class DefaultGradOpDescMaker : public SingleGradOpDescMaker {
using SingleGradOpDescMaker::SingleGradOpDescMaker;
protected:
virtual std::unique_ptr<OpDescBind> Apply() const {
auto* grad = new OpDescBind();
virtual std::unique_ptr<OpDesc> Apply() const {
auto* grad = new OpDesc();
grad->SetType(this->GradOpType());
for (auto& input_param : this->InputNames()) {
......@@ -150,7 +166,7 @@ class DefaultGradOpDescMaker : public SingleGradOpDescMaker {
grad->SetAttrMap(this->Attrs());
return std::unique_ptr<OpDescBind>(grad);
return std::unique_ptr<OpDesc>(grad);
}
virtual std::string GradOpType() const {
......@@ -161,7 +177,7 @@ class DefaultGradOpDescMaker : public SingleGradOpDescMaker {
class EmptyGradOpMaker : public GradOpDescMakerBase {
public:
using GradOpDescMakerBase::GradOpDescMakerBase;
std::vector<std::unique_ptr<OpDescBind>> operator()() const override {
std::vector<std::unique_ptr<OpDesc>> operator()() const override {
return {};
}
};
......
paddle/framework/library_type.h 0 → 100644
浏览文件 @ ba8660a1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
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. */
#pragma once
namespace paddle {
namespace framework {
// For more details about the design of LibraryType, Please refer to
// https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/operator_kernel_type.md#library
enum LibraryType { kPlain = 0; kMKLDNN = 1; kCUDNN = 2; }
} // namespace
} // framework
paddle/framework/lod_rank_table.cc
浏览文件 @ ba8660a1
......@@ -46,4 +46,13 @@ void LoDRankTable::Reset(const LoD& lod, size_t level) {
}
} // namespace framework
std::ostream& operator<<(std::ostream& out,
const framework::LoDRankTable& table) {
out << "NumOfSequence " << table.items().size() << "\n";
for (auto& each_item : table.items()) {
out << "\tSeq #" << each_item.index << ", Len=" << each_item.length << "\n";
}
return out;
}
} // namespace paddle
paddle/framework/lod_rank_table.h
浏览文件 @ ba8660a1
......@@ -13,6 +13,7 @@
limitations under the License. */
#pragma once
#include <iosfwd>
#include "paddle/framework/lod_tensor.h"
namespace paddle {
......@@ -52,4 +53,8 @@ class LoDRankTable {
};
} // namespace framework
std::ostream& operator<<(std::ostream& out,
const framework::LoDRankTable& table);
} // namespace paddle
paddle/framework/lod_tensor.cc
浏览文件 @ ba8660a1
......@@ -197,7 +197,7 @@ void SerializeToStream(std::ostream &os, const LoDTensor &tensor,
{ // the 2nd field, tensor description
// int32_t size
// void* protobuf message
framework::TensorDesc desc;
proto::TensorDesc desc;
desc.set_data_type(framework::ToDataType(tensor.type()));
auto dims = framework::vectorize(tensor.dims());
auto *pb_dims = desc.mutable_dims();
......@@ -262,7 +262,7 @@ void DeserializeFromStream(std::istream &is, LoDTensor *tensor) {
uint32_t version;
is.read(reinterpret_cast<char *>(&version), sizeof(version));
PADDLE_ENFORCE_EQ(version, 0U, "Only version 0 is supported");
framework::TensorDesc desc;
proto::TensorDesc desc;
{ // int32_t size
// proto buffer
int32_t size;
......@@ -281,16 +281,16 @@ void DeserializeFromStream(std::istream &is, LoDTensor *tensor) {
void *buf;
platform::Place cpu = platform::CPUPlace();
switch (desc.data_type()) {
case framework::FP32:
case proto::FP32:
buf = tensor->mutable_data<float>(cpu);
break;
case framework::FP64:
case proto::FP64:
buf = tensor->mutable_data<double>(cpu);
break;
case framework::INT32:
case proto::INT32:
buf = tensor->mutable_data<int>(cpu);
break;
case framework::INT64:
case proto::INT64:
buf = tensor->mutable_data<int64_t>(cpu);
break;
default:
......
paddle/framework/lod_tensor.h
浏览文件 @ ba8660a1
......@@ -184,6 +184,18 @@ LoDTensor LodExpand(const LoDTensor& source, const LoD& lod, size_t level,
return tensor;
}
// Get the absolute offset of a lod[start_level][start_idx:end_idx] and
// relative length of details for every levels(i.e., [start_level: ]).
//
// For example,
// lod = [[0, 3, 4, 8], [0, 9, 10, 11, 13, 17, 19, 22, 24]]
// start_level = 0
// start_idx = 1
// end_idx = 3
//
// Returns:
// LoD = [[1, 4], [2, 4, 2, 3, 2]]
// pair<size_t, size_t> = {11, 24}
std::pair<LoD, std::pair<size_t, size_t>> GetSubLoDAndAbsoluteOffset(
const LoD& lod, size_t start_idx, size_t end_idx, size_t start_level);
......
paddle/framework/op_desc.cc
浏览文件 @ ba8660a1
......@@ -25,12 +25,11 @@ limitations under the License. */
namespace paddle {
namespace framework {
class OpDescBind;
class BlockDescBind;
class OpDesc;
class BlockDesc;
class CompileTimeInferShapeContext : public InferShapeContext {
public:
CompileTimeInferShapeContext(const OpDescBind &op,
const BlockDescBind &block);
CompileTimeInferShapeContext(const OpDesc &op, const BlockDesc &block);
bool HasInput(const std::string &name) const override;
......@@ -58,11 +57,11 @@ class CompileTimeInferShapeContext : public InferShapeContext {
PADDLE_ENFORCE_LT(j, Outputs(out).size());
auto *in_var = block_.FindVarRecursive(Inputs(in)[i]);
auto *out_var = block_.FindVarRecursive(Outputs(out)[j]);
if (in_var->GetType() != VarDesc::LOD_TENSOR) {
if (in_var->GetType() != proto::VarDesc::LOD_TENSOR) {
VLOG(3) << "input " << in << " is not LodTensor";
return;
}
PADDLE_ENFORCE_EQ(in_var->GetType(), VarDesc::LOD_TENSOR,
PADDLE_ENFORCE_EQ(in_var->GetType(), proto::VarDesc::LOD_TENSOR,
"The %d-th output of Output(%s) must be LoDTensor.", j,
out);
out_var->SetLoDLevel(in_var->GetLodLevel());
......@@ -70,19 +69,18 @@ class CompileTimeInferShapeContext : public InferShapeContext {
bool IsRuntime() const override;
protected:
VarDesc::VarType GetVarType(const std::string &name) const override;
proto::VarDesc::VarType GetVarType(const std::string &name) const override;
DDim GetDim(const std::string &name) const override;
void SetDim(const std::string &name, const DDim &dim) override;
const OpDescBind &op_;
const BlockDescBind &block_;
const OpDesc &op_;
const BlockDesc &block_;
};
OpDescBind::OpDescBind(const std::string &type, const VariableNameMap &inputs,
const VariableNameMap &outputs,
const AttributeMap &attrs) {
OpDesc::OpDesc(const std::string &type, const VariableNameMap &inputs,
const VariableNameMap &outputs, const AttributeMap &attrs) {
desc_.set_type(type);
inputs_ = inputs;
outputs_ = outputs;
......@@ -90,12 +88,12 @@ OpDescBind::OpDescBind(const std::string &type, const VariableNameMap &inputs,
need_update_ = true;
}
OpDescBind::OpDescBind(const OpDesc &desc, ProgramDescBind *prog)
OpDesc::OpDesc(const proto::OpDesc &desc, ProgramDesc *prog)
: desc_(desc), need_update_(false) {
// restore inputs_
int input_size = desc_.inputs_size();
for (int i = 0; i < input_size; ++i) {
const OpDesc::Var &var = desc_.inputs(i);
const proto::OpDesc::Var &var = desc_.inputs(i);
std::vector<std::string> &args = inputs_[var.parameter()];
int argu_size = var.arguments_size();
args.reserve(argu_size);
......@@ -106,7 +104,7 @@ OpDescBind::OpDescBind(const OpDesc &desc, ProgramDescBind *prog)
// restore outputs_
int output_size = desc_.outputs_size();
for (int i = 0; i < output_size; ++i) {
const OpDesc::Var &var = desc_.outputs(i);
const proto::OpDesc::Var &var = desc_.outputs(i);
std::vector<std::string> &args = outputs_[var.parameter()];
int argu_size = var.arguments_size();
args.reserve(argu_size);
......@@ -115,9 +113,9 @@ OpDescBind::OpDescBind(const OpDesc &desc, ProgramDescBind *prog)
}
}
// restore attrs_
for (const OpDesc::Attr &attr : desc_.attrs()) {
for (const proto::OpDesc::Attr &attr : desc_.attrs()) {
std::string attr_name = attr.name();
if (attr.type() != AttrType::BLOCK) {
if (attr.type() != proto::AttrType::BLOCK) {
attrs_[attr_name] = GetAttrValue(attr);
} else {
auto bid = attr.block_idx();
......@@ -126,20 +124,19 @@ OpDescBind::OpDescBind(const OpDesc &desc, ProgramDescBind *prog)
}
}
OpDesc *OpDescBind::Proto() {
proto::OpDesc *OpDesc::Proto() {
Flush();
return &desc_;
}
const std::vector<std::string> &OpDescBind::Input(
const std::string &name) const {
const std::vector<std::string> &OpDesc::Input(const std::string &name) const {
auto it = inputs_.find(name);
PADDLE_ENFORCE(it != inputs_.end(), "Input %s cannot be found in Op %s", name,
Type());
return it->second;
}
std::vector<std::string> OpDescBind::InputArgumentNames() const {
std::vector<std::string> OpDesc::InputArgumentNames() const {
std::vector<std::string> retv;
for (auto &ipt : this->inputs_) {
retv.insert(retv.end(), ipt.second.begin(), ipt.second.end());
......@@ -147,21 +144,20 @@ std::vector<std::string> OpDescBind::InputArgumentNames() const {
return retv;
}
void OpDescBind::SetInput(const std::string &param_name,
const std::vector<std::string> &args) {
void OpDesc::SetInput(const std::string &param_name,
const std::vector<std::string> &args) {
need_update_ = true;
inputs_[param_name] = args;
}
const std::vector<std::string> &OpDescBind::Output(
const std::string &name) const {
const std::vector<std::string> &OpDesc::Output(const std::string &name) const {
auto it = outputs_.find(name);
PADDLE_ENFORCE(it != outputs_.end(), "Output %s cannot be found in Op %s",
name, Type());
return it->second;
}
std::vector<std::string> OpDescBind::OutputArgumentNames() const {
std::vector<std::string> OpDesc::OutputArgumentNames() const {
std::vector<std::string> retv;
for (auto &ipt : this->outputs_) {
retv.insert(retv.end(), ipt.second.begin(), ipt.second.end());
......@@ -169,19 +165,19 @@ std::vector<std::string> OpDescBind::OutputArgumentNames() const {
return retv;
}
void OpDescBind::SetOutput(const std::string &param_name,
const std::vector<std::string> &args) {
void OpDesc::SetOutput(const std::string &param_name,
const std::vector<std::string> &args) {
need_update_ = true;
this->outputs_[param_name] = args;
}
AttrType OpDescBind::GetAttrType(const std::string &name) const {
proto::AttrType OpDesc::GetAttrType(const std::string &name) const {
auto it = attrs_.find(name);
PADDLE_ENFORCE(it != attrs_.end(), "Attribute %s is not found", name);
return static_cast<AttrType>(it->second.which() - 1);
return static_cast<proto::AttrType>(it->second.which() - 1);
}
std::vector<std::string> OpDescBind::AttrNames() const {
std::vector<std::string> OpDesc::AttrNames() const {
std::vector<std::string> retv;
retv.reserve(attrs_.size());
for (auto &attr : attrs_) {
......@@ -190,41 +186,39 @@ std::vector<std::string> OpDescBind::AttrNames() const {
return retv;
}
void OpDescBind::SetAttr(const std::string &name, const Attribute &v) {
void OpDesc::SetAttr(const std::string &name, const Attribute &v) {
this->attrs_[name] = v;
need_update_ = true;
}
void OpDescBind::SetBlockAttr(const std::string &name, BlockDescBind &block) {
void OpDesc::SetBlockAttr(const std::string &name, BlockDesc &block) {
this->attrs_[name] = &block;
need_update_ = true;
}
void OpDescBind::SetAttrMap(
void OpDesc::SetAttrMap(
const std::unordered_map<std::string, Attribute> &attr_map) {
attrs_ = attr_map;
need_update_ = true;
}
Attribute OpDescBind::GetAttr(const std::string &name) const {
Attribute OpDesc::GetAttr(const std::string &name) const {
auto it = attrs_.find(name);
PADDLE_ENFORCE(it != attrs_.end(), "Attribute %s is not found", name);
return it->second;
}
int OpDescBind::GetBlockAttr(const std::string &name) const {
int OpDesc::GetBlockAttr(const std::string &name) const {
auto it = attrs_.find(name);
PADDLE_ENFORCE(it != attrs_.end(), "Attribute %s is not found", name);
return boost::get<BlockDescBind *>(it->second)->ID();
return boost::get<BlockDesc *>(it->second)->ID();
}
const std::unordered_map<std::string, Attribute> &OpDescBind::GetAttrMap()
const {
const std::unordered_map<std::string, Attribute> &OpDesc::GetAttrMap() const {
return attrs_;
}
void OpDescBind::Rename(const std::string &old_name,
const std::string &new_name) {
void OpDesc::Rename(const std::string &old_name, const std::string &new_name) {
for (auto &input : inputs_) {
std::replace(input.second.begin(), input.second.end(), old_name, new_name);
}
......@@ -235,8 +229,8 @@ void OpDescBind::Rename(const std::string &old_name,
need_update_ = true;
}
void OpDescBind::RenameOutput(const std::string &old_name,
const std::string &new_name) {
void OpDesc::RenameOutput(const std::string &old_name,
const std::string &new_name) {
for (auto &output : outputs_) {
std::replace(output.second.begin(), output.second.end(), old_name,
new_name);
......@@ -244,8 +238,8 @@ void OpDescBind::RenameOutput(const std::string &old_name,
need_update_ = true;
}
void OpDescBind::RenameInput(const std::string &old_name,
const std::string &new_name) {
void OpDesc::RenameInput(const std::string &old_name,
const std::string &new_name) {
for (auto &input : inputs_) {
std::replace(input.second.begin(), input.second.end(), old_name, new_name);
}
......@@ -253,8 +247,8 @@ void OpDescBind::RenameInput(const std::string &old_name,
}
struct SetAttrDescVisitor : public boost::static_visitor<void> {
explicit SetAttrDescVisitor(OpDesc::Attr *attr) : attr_(attr) {}
mutable OpDesc::Attr *attr_;
explicit SetAttrDescVisitor(proto::OpDesc::Attr *attr) : attr_(attr) {}
mutable proto::OpDesc::Attr *attr_;
void operator()(int v) const { attr_->set_i(v); }
void operator()(float v) const { attr_->set_f(v); }
void operator()(const std::string &v) const { attr_->set_s(v); }
......@@ -272,11 +266,13 @@ struct SetAttrDescVisitor : public boost::static_visitor<void> {
void operator()(const std::vector<bool> &v) const {
VectorToRepeated(v, attr_->mutable_bools());
}
void operator()(BlockDesc *desc) const { attr_->set_block_idx(desc->idx()); }
void operator()(proto::BlockDesc *desc) const {
attr_->set_block_idx(desc->idx());
}
void operator()(boost::blank) const { PADDLE_THROW("Unexpected branch"); }
};
void OpDescBind::Flush() {
void OpDesc::Flush() {
if (need_update_) {
this->desc_.mutable_inputs()->Clear();
for (auto &ipt : inputs_) {
......@@ -297,7 +293,7 @@ void OpDescBind::Flush() {
auto *attr_desc = desc_.add_attrs();
attr_desc->set_name(attr.first);
attr_desc->set_type(
static_cast<framework::AttrType>(attr.second.which() - 1));
static_cast<proto::AttrType>(attr.second.which() - 1));
SetAttrDescVisitor visitor(attr_desc);
boost::apply_visitor(visitor, attr.second);
}
......@@ -328,7 +324,7 @@ static void InitInferShapeFuncs() {
});
}
void OpDescBind::CheckAttrs() {
void OpDesc::CheckAttrs() {
PADDLE_ENFORCE(!Type().empty(),
"CheckAttr() can not be called before type is setted.");
auto *checker = OpInfoMap::Instance().Get(Type()).Checker();
......@@ -340,7 +336,7 @@ void OpDescBind::CheckAttrs() {
checker->Check(attrs_);
}
void OpDescBind::InferShape(const BlockDescBind &block) const {
void OpDesc::InferShape(const BlockDesc &block) const {
VLOG(3) << "CompileTime infer shape on " << Type();
InitInferShapeFuncs();
auto &infer_shape = OpInfoMap::Instance().Get(this->Type()).infer_shape_;
......@@ -363,7 +359,7 @@ void OpDescBind::InferShape(const BlockDescBind &block) const {
infer_shape(&ctx);
}
void OpDescBind::InferVarType(BlockDescBind *block) const {
void OpDesc::InferVarType(BlockDesc *block) const {
auto &info = OpInfoMap::Instance().Get(this->Type());
if (info.infer_var_type_) {
info.infer_var_type_(*this, block);
......@@ -375,14 +371,14 @@ void OpDescBind::InferVarType(BlockDescBind *block) const {
for (auto &out_pair : this->outputs_) {
for (auto &out_var_name : out_pair.second) {
block->FindRecursiveOrCreateVar(out_var_name)
->SetType(VarDesc::LOD_TENSOR);
->SetType(proto::VarDesc::LOD_TENSOR);
}
}
}
}
CompileTimeInferShapeContext::CompileTimeInferShapeContext(
const OpDescBind &op, const BlockDescBind &block)
const OpDesc &op, const BlockDesc &block)
: op_(op), block_(block) {}
bool CompileTimeInferShapeContext::HasInput(const std::string &name) const {
......@@ -484,7 +480,7 @@ void CompileTimeInferShapeContext::SetDim(const std::string &name,
}
bool CompileTimeInferShapeContext::IsRuntime() const { return false; }
VarDesc::VarType CompileTimeInferShapeContext::GetVarType(
proto::VarDesc::VarType CompileTimeInferShapeContext::GetVarType(
const std::string &name) const {
return block_.FindVarRecursive(name)->GetType();
}
......
paddle/framework/op_desc.h
浏览文件 @ ba8660a1
......@@ -23,19 +23,19 @@ limitations under the License. */
namespace paddle {
namespace framework {
class BlockDescBind;
class ProgramDescBind;
class BlockDesc;
class ProgramDesc;
class OpDescBind {
class OpDesc {
public:
OpDescBind() {}
OpDesc() {}
OpDescBind(const std::string &type, const VariableNameMap &inputs,
const VariableNameMap &outputs, const AttributeMap &attrs);
OpDesc(const std::string &type, const VariableNameMap &inputs,
const VariableNameMap &outputs, const AttributeMap &attrs);
OpDescBind(const OpDesc &desc, ProgramDescBind *prog);
OpDesc(const proto::OpDesc &desc, ProgramDesc *prog);
OpDesc *Proto();
proto::OpDesc *Proto();
std::string Type() const { return desc_.type(); }
......@@ -59,13 +59,13 @@ class OpDescBind {
return attrs_.find(name) != attrs_.end();
}
AttrType GetAttrType(const std::string &name) const;
proto::AttrType GetAttrType(const std::string &name) const;
std::vector<std::string> AttrNames() const;
void SetAttr(const std::string &name, const Attribute &v);
void SetBlockAttr(const std::string &name, BlockDescBind &block);
void SetBlockAttr(const std::string &name, BlockDesc &block);
Attribute GetAttr(const std::string &name) const;
......@@ -107,9 +107,9 @@ class OpDescBind {
void CheckAttrs();
void InferShape(const BlockDescBind &block) const;
void InferShape(const BlockDesc &block) const;
void InferVarType(BlockDescBind *block) const;
void InferVarType(BlockDesc *block) const;
void MarkAsTarget() { desc_.set_is_target(true); }
......@@ -126,8 +126,10 @@ class OpDescBind {
return ret_val;
}
OpDesc desc_;
proto::OpDesc desc_;
// input arg name => output variable names
VariableNameMap inputs_;
// output arg name => output variable names
VariableNameMap outputs_;
AttributeMap attrs_;
......
paddle/framework/op_info.h
浏览文件 @ ba8660a1
......@@ -34,7 +34,7 @@ class InferShapeBase {
struct OpInfo {
OpCreator creator_;
GradOpMakerFN grad_op_maker_;
OpProto* proto_{nullptr};
proto::OpProto* proto_{nullptr};
OpAttrChecker* checker_{nullptr};
InferVarTypeFN infer_var_type_;
InferShapeFN infer_shape_;
......@@ -43,7 +43,7 @@ struct OpInfo {
return proto_ != nullptr && checker_ != nullptr;
}
const OpProto& Proto() const {
const proto::OpProto& Proto() const {
PADDLE_ENFORCE_NOT_NULL(proto_, "Operator Proto has not been registered");
PADDLE_ENFORCE(proto_->IsInitialized(),
"Operator Proto must be initialized in op info");
......
paddle/framework/op_proto_maker.h
浏览文件 @ ba8660a1
......@@ -22,6 +22,8 @@ namespace framework {
// this class not only make proto but also init attribute checkers.
class OpProtoAndCheckerMaker {
public:
using OpProto = proto::OpProto;
using OpAttrChecker = framework::OpAttrChecker;
OpProtoAndCheckerMaker(OpProto* proto, OpAttrChecker* op_checker)
: proto_(proto), op_checker_(op_checker) {}
......@@ -80,7 +82,7 @@ class OpProtoAndCheckerMaker {
class NOPMaker : public OpProtoAndCheckerMaker {
public:
NOPMaker(framework::OpProto* proto, framework::OpAttrChecker* op_checker)
NOPMaker(OpProto* proto, framework::OpAttrChecker* op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {}
};
......
paddle/framework/op_proto_maker_test.cc
浏览文件 @ ba8660a1
......@@ -18,7 +18,7 @@ limitations under the License. */
class TestAttrProtoMaker : public paddle::framework::OpProtoAndCheckerMaker {
public:
TestAttrProtoMaker(paddle::framework::OpProto* proto,
TestAttrProtoMaker(paddle::framework::proto::OpProto* proto,
paddle::framework::OpAttrChecker* op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddAttr<float>("scale", "scale of test op");
......@@ -27,7 +27,7 @@ class TestAttrProtoMaker : public paddle::framework::OpProtoAndCheckerMaker {
};
TEST(ProtoMaker, DuplicatedAttr) {
paddle::framework::OpProto op_proto;
paddle::framework::proto::OpProto op_proto;
paddle::framework::OpAttrChecker op_checker;
auto proto_maker = TestAttrProtoMaker(&op_proto, &op_checker);
ASSERT_THROW(proto_maker.Validate(), paddle::platform::EnforceNotMet);
......@@ -35,7 +35,7 @@ TEST(ProtoMaker, DuplicatedAttr) {
class TestInOutProtoMaker : public paddle::framework::OpProtoAndCheckerMaker {
public:
TestInOutProtoMaker(paddle::framework::OpProto* proto,
TestInOutProtoMaker(paddle::framework::proto::OpProto* proto,
paddle::framework::OpAttrChecker* op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("input", "input of test op");
......@@ -44,7 +44,7 @@ class TestInOutProtoMaker : public paddle::framework::OpProtoAndCheckerMaker {
};
TEST(ProtoMaker, DuplicatedInOut) {
paddle::framework::OpProto op_proto;
paddle::framework::proto::OpProto op_proto;
paddle::framework::OpAttrChecker op_checker;
auto proto_maker = TestInOutProtoMaker(&op_proto, &op_checker);
ASSERT_THROW(proto_maker.Validate(), paddle::platform::EnforceNotMet);
......
paddle/framework/op_registry.cc
浏览文件 @ ba8660a1
......@@ -31,7 +31,8 @@ std::unique_ptr<OperatorBase> OpRegistry::CreateOp(
}
static VariableNameMap ConvertOpDescVarsToVarNameMap(
const google::protobuf::RepeatedPtrField<OpDesc::Var>& op_desc_vars) {
const google::protobuf::RepeatedPtrField<proto::OpDesc::Var>&
op_desc_vars) {
VariableNameMap ret_val;
for (auto& var : op_desc_vars) {
auto& var_names = ret_val[var.parameter()];
......@@ -43,9 +44,10 @@ static VariableNameMap ConvertOpDescVarsToVarNameMap(
return ret_val;
}
std::unique_ptr<OperatorBase> OpRegistry::CreateOp(const OpDesc& op_desc) {
std::unique_ptr<OperatorBase> OpRegistry::CreateOp(
const proto::OpDesc& op_desc) {
VLOG(1) << "CreateOp directly from OpDesc is deprecated. It should only be"
"used in unit tests. Use CreateOp(const OpDescBind& op_desc) "
"used in unit tests. Use CreateOp(const OpDesc& op_desc) "
"instead.";
VariableNameMap inputs = ConvertOpDescVarsToVarNameMap(op_desc.inputs());
VariableNameMap outputs = ConvertOpDescVarsToVarNameMap(op_desc.outputs());
......@@ -57,7 +59,7 @@ std::unique_ptr<OperatorBase> OpRegistry::CreateOp(const OpDesc& op_desc) {
return CreateOp(op_desc.type(), inputs, outputs, attrs);
}
std::unique_ptr<OperatorBase> OpRegistry::CreateOp(const OpDescBind& op_desc) {
std::unique_ptr<OperatorBase> OpRegistry::CreateOp(const OpDesc& op_desc) {
return CreateOp(op_desc.Type(), op_desc.Inputs(), op_desc.Outputs(),
op_desc.GetAttrMap());
}
......
paddle/framework/op_registry.h
浏览文件 @ ba8660a1
......@@ -77,9 +77,9 @@ class OpRegistry {
const VariableNameMap& outputs,
AttributeMap attrs);
static std::unique_ptr<OperatorBase> CreateOp(const OpDesc& op_desc);
static std::unique_ptr<OperatorBase> CreateOp(const proto::OpDesc& op_desc);
static std::unique_ptr<OperatorBase> CreateOp(const OpDescBind& op_desc);
static std::unique_ptr<OperatorBase> CreateOp(const OpDesc& op_desc);
};
template <typename PlaceType, bool at_end, size_t I, typename... KernelType>
......@@ -126,6 +126,14 @@ class OpKernelRegistrar : public Registrar {
__test_global_namespace_##uniq_name##__>::value, \
msg)
/*
The variadic arguments should be class types derived from one of the
following classes:
OpProtoAndCheckerMaker
GradOpDescMakerBase
VarTypeInference
InferShapeBase
*/
#define REGISTER_OPERATOR(op_type, op_class, ...) \
STATIC_ASSERT_GLOBAL_NAMESPACE( \
__reg_op__##op_type, \
......@@ -144,20 +152,29 @@ class OpKernelRegistrar : public Registrar {
}
/**
* Macro to register Operator.
* Macro to register Operator. When the input is duplicable, you should
* use REGISTER_OP_EX with deop_empty_grad=false instead.
*/
#define REGISTER_OP(op_type, op_class, op_maker_class, grad_op_type, \
grad_op_class) \
REGISTER_OPERATOR(grad_op_type, grad_op_class); \
class _GradOpDescMaker_##grad_op_type##_ \
: public ::paddle::framework::DefaultGradOpDescMaker<true> { \
using ::paddle::framework::DefaultGradOpDescMaker< \
true>::DefaultGradOpDescMaker; \
\
protected: \
virtual std::string GradOpType() const { return #grad_op_type; } \
}; \
REGISTER_OPERATOR(op_type, op_class, _GradOpDescMaker_##grad_op_type##_, \
#define REGISTER_OP(op_type, op_class, op_maker_class, grad_op_type, \
grad_op_class) \
REGISTER_OP_EX(op_type, op_class, op_maker_class, grad_op_type, \
grad_op_class, true)
// When an argument is duplicable, we need to use this version.
// Perhaps we can omit DropEmptyIG template parameter and
// only have one version of REGISTER_OP.
#define REGISTER_OP_EX(op_type, op_class, op_maker_class, grad_op_type, \
grad_op_class, drop_empty_grad) \
REGISTER_OPERATOR(grad_op_type, grad_op_class); \
class _GradOpDescMaker_##grad_op_type##_ \
: public ::paddle::framework::DefaultGradOpDescMaker<drop_empty_grad> { \
using ::paddle::framework::DefaultGradOpDescMaker< \
drop_empty_grad>::DefaultGradOpDescMaker; \
\
protected: \
virtual std::string GradOpType() const { return #grad_op_type; } \
}; \
REGISTER_OPERATOR(op_type, op_class, _GradOpDescMaker_##grad_op_type##_, \
op_maker_class);
#define REGISTER_OP_WITH_KERNEL(op_type, ...) \
......
paddle/framework/op_registry_test.cc
浏览文件 @ ba8660a1
......@@ -51,7 +51,7 @@ class MyTestOpProtoAndCheckerMaker : public OpProtoAndCheckerMaker {
static void BuildVar(const std::string& param_name,
std::initializer_list<const char*> arguments,
paddle::framework::OpDesc::Var* var) {
paddle::framework::proto::OpDesc::Var* var) {
var->set_parameter(param_name);
for (auto& arg_name : arguments) {
var->add_arguments(arg_name);
......@@ -63,7 +63,7 @@ REGISTER_OP_WITHOUT_GRADIENT(my_test_op, paddle::framework::MyTestOp,
paddle::framework::MyTestOpProtoAndCheckerMaker);
TEST(OpRegistry, CreateOp) {
paddle::framework::OpDesc op_desc;
paddle::framework::proto::OpDesc op_desc;
op_desc.set_type("cos_sim");
BuildVar("input", {"aa"}, op_desc.add_inputs());
BuildVar("output", {"bb"}, op_desc.add_outputs());
......@@ -71,7 +71,7 @@ TEST(OpRegistry, CreateOp) {
float scale = 3.3;
auto attr = op_desc.mutable_attrs()->Add();
attr->set_name("scale");
attr->set_type(paddle::framework::AttrType::FLOAT);
attr->set_type(paddle::framework::proto::AttrType::FLOAT);
attr->set_f(scale);
auto op = paddle::framework::OpRegistry::CreateOp(op_desc);
......@@ -83,14 +83,14 @@ TEST(OpRegistry, CreateOp) {
}
TEST(OpRegistry, IllegalAttr) {
paddle::framework::OpDesc op_desc;
paddle::framework::proto::OpDesc op_desc;
op_desc.set_type("cos_sim");
BuildVar("input", {"aa"}, op_desc.add_inputs());
BuildVar("output", {"bb"}, op_desc.add_outputs());
auto attr = op_desc.mutable_attrs()->Add();
attr->set_name("scale");
attr->set_type(paddle::framework::AttrType::FLOAT);
attr->set_type(paddle::framework::proto::AttrType::FLOAT);
attr->set_f(-2.0);
bool caught = false;
......@@ -108,7 +108,7 @@ TEST(OpRegistry, IllegalAttr) {
}
TEST(OpRegistry, DefaultValue) {
paddle::framework::OpDesc op_desc;
paddle::framework::proto::OpDesc op_desc;
op_desc.set_type("cos_sim");
BuildVar("input", {"aa"}, op_desc.add_inputs());
BuildVar("output", {"bb"}, op_desc.add_outputs());
......@@ -123,7 +123,7 @@ TEST(OpRegistry, DefaultValue) {
}
TEST(OpRegistry, CustomChecker) {
paddle::framework::OpDesc op_desc;
paddle::framework::proto::OpDesc op_desc;
op_desc.set_type("my_test_op");
BuildVar("input", {"ii"}, op_desc.add_inputs());
BuildVar("output", {"oo"}, op_desc.add_outputs());
......@@ -145,7 +145,7 @@ TEST(OpRegistry, CustomChecker) {
// set 'test_attr' set to an illegal value
auto attr = op_desc.mutable_attrs()->Add();
attr->set_name("test_attr");
attr->set_type(paddle::framework::AttrType::INT);
attr->set_type(paddle::framework::proto::AttrType::INT);
attr->set_i(3);
caught = false;
try {
......@@ -164,7 +164,7 @@ TEST(OpRegistry, CustomChecker) {
op_desc.mutable_attrs()->Clear();
attr = op_desc.mutable_attrs()->Add();
attr->set_name("test_attr");
attr->set_type(paddle::framework::AttrType::INT);
attr->set_type(paddle::framework::proto::AttrType::INT);
attr->set_i(4);
auto op = paddle::framework::OpRegistry::CreateOp(op_desc);
paddle::platform::CPUDeviceContext dev_ctx;
......
paddle/framework/operator.cc
浏览文件 @ ba8660a1
......@@ -377,7 +377,7 @@ class RuntimeInferShapeContext : public InferShapeContext {
}
}
VarDesc::VarType GetVarType(const std::string& name) const override {
proto::VarDesc::VarType GetVarType(const std::string& name) const override {
auto* var = scope_.FindVar(name);
return ToVarType(var->Type());
}
......@@ -417,7 +417,7 @@ OpKernelType OperatorWithKernel::GetKernelType(
const ExecutionContext& ctx) const {
return OpKernelType(IndicateDataType(ctx), ctx.GetPlace());
}
DataType OperatorWithKernel::IndicateDataType(
proto::DataType OperatorWithKernel::IndicateDataType(
const ExecutionContext& ctx) const {
auto& scope = ctx.scope();
int data_type = -1;
......@@ -443,7 +443,7 @@ DataType OperatorWithKernel::IndicateDataType(
}
}
PADDLE_ENFORCE(data_type != -1, "DataType should be indicated by input");
return static_cast<DataType>(data_type);
return static_cast<proto::DataType>(data_type);
}
} // namespace framework
......
paddle/framework/operator.h
浏览文件 @ ba8660a1
......@@ -358,12 +358,13 @@ struct OpKernelType {
};
platform::Place place_;
DataType data_type_;
proto::DataType data_type_;
OpKernelType(DataType data_type, platform::Place place)
OpKernelType(proto::DataType data_type, platform::Place place)
: place_(place), data_type_(data_type) {}
OpKernelType(DataType data_type, const platform::DeviceContext& dev_ctx)
OpKernelType(proto::DataType data_type,
const platform::DeviceContext& dev_ctx)
: place_(dev_ctx.GetPlace()), data_type_(data_type) {}
bool operator==(const OpKernelType& o) const {
......@@ -409,7 +410,7 @@ class OperatorWithKernel : public OperatorBase {
private:
// indicate kernel DataType by input data. Defaultly all input data must be
// same.
DataType IndicateDataType(const ExecutionContext& ctx) const;
proto::DataType IndicateDataType(const ExecutionContext& ctx) const;
};
std::ostream& operator<<(std::ostream& os, const OpKernelType& kernel_key);
......
paddle/framework/operator_test.cc
浏览文件 @ ba8660a1
......@@ -58,7 +58,7 @@ class OpeWithoutKernelTestProtoAndCheckerMaker : public OpProtoAndCheckerMaker {
static void BuildVar(const std::string& param_name,
std::initializer_list<const char*> arguments,
paddle::framework::OpDesc::Var* var) {
paddle::framework::proto::OpDesc::Var* var) {
var->set_parameter(param_name);
for (auto& arg_name : arguments) {
*var->mutable_arguments()->Add() = arg_name;
......@@ -70,14 +70,14 @@ REGISTER_OP_WITHOUT_GRADIENT(
paddle::framework::OpeWithoutKernelTestProtoAndCheckerMaker);
TEST(OperatorBase, all) {
paddle::framework::OpDesc op_desc;
paddle::framework::proto::OpDesc op_desc;
op_desc.set_type("test_operator");
BuildVar("input", {"IN1"}, op_desc.add_inputs());
BuildVar("output", {"OUT1"}, op_desc.add_outputs());
auto attr = op_desc.mutable_attrs()->Add();
attr->set_name("scale");
attr->set_type(paddle::framework::AttrType::FLOAT);
attr->set_type(paddle::framework::proto::AttrType::FLOAT);
attr->set_f(3.14);
paddle::platform::CPUDeviceContext device_context;
......@@ -115,7 +115,7 @@ class OpWithKernelTest : public OperatorWithKernel {
protected:
void InferShape(framework::InferShapeContext* ctx) const override {}
OpKernelType GetKernelType(const ExecutionContext& ctx) const override {
return OpKernelType(DataType::FP32, ctx.GetPlace());
return OpKernelType(proto::DataType::FP32, ctx.GetPlace());
}
};
......@@ -195,14 +195,14 @@ REGISTER_OP_CPU_KERNEL(op_with_kernel,
// test with single input
TEST(OpKernel, all) {
paddle::framework::OpDesc op_desc;
paddle::framework::proto::OpDesc op_desc;
op_desc.set_type("op_with_kernel");
BuildVar("x", {"IN1"}, op_desc.add_inputs());
BuildVar("y", {"OUT1"}, op_desc.add_outputs());
auto attr = op_desc.mutable_attrs()->Add();
attr->set_name("scale");
attr->set_type(paddle::framework::AttrType::FLOAT);
attr->set_type(paddle::framework::proto::AttrType::FLOAT);
attr->set_f(3.14);
paddle::platform::CPUDeviceContext cpu_device_context;
......@@ -224,7 +224,7 @@ REGISTER_OP_CPU_KERNEL(op_multi_inputs_with_kernel,
TEST(OpKernel, multi_inputs) {
using namespace paddle::framework;
OpDesc op_desc;
proto::OpDesc op_desc;
op_desc.set_type("op_multi_inputs_with_kernel");
BuildVar("xs", {"x0", "x1", "x2"}, op_desc.add_inputs());
BuildVar("k", {"k0"}, op_desc.add_inputs());
......@@ -232,7 +232,7 @@ TEST(OpKernel, multi_inputs) {
auto attr = op_desc.mutable_attrs()->Add();
attr->set_name("scale");
attr->set_type(paddle::framework::AttrType::FLOAT);
attr->set_type(paddle::framework::proto::AttrType::FLOAT);
attr->set_f(3.14);
paddle::platform::CPUDeviceContext cpu_device_context;
......
paddle/framework/program_desc.cc
浏览文件 @ ba8660a1
......@@ -18,49 +18,49 @@ limitations under the License. */
namespace paddle {
namespace framework {
BlockDescBind *ProgramDescBind::AppendBlock(const BlockDescBind &parent) {
BlockDesc *ProgramDesc::AppendBlock(const BlockDesc &parent) {
auto *b = desc_.add_blocks();
b->set_parent_idx(parent.ID());
b->set_idx(desc_.blocks_size() - 1);
blocks_.emplace_back(new BlockDescBind(this, b));
blocks_.emplace_back(new BlockDesc(this, b));
return blocks_.back().get();
}
ProgramDesc *ProgramDescBind::Proto() {
proto::ProgramDesc *ProgramDesc::Proto() {
for (auto &block : blocks_) {
block->Flush();
}
return &desc_;
}
ProgramDescBind::ProgramDescBind() {
ProgramDesc::ProgramDesc() {
auto *block = desc_.mutable_blocks()->Add();
block->set_idx(kRootBlockIndex);
block->set_parent_idx(kNoneBlockIndex);
blocks_.emplace_back(new BlockDescBind(this, block));
blocks_.emplace_back(new BlockDesc(this, block));
}
ProgramDescBind::ProgramDescBind(const ProgramDescBind &o) {
ProgramDesc::ProgramDesc(const ProgramDesc &o) {
desc_ = o.desc_;
for (int i = 0; i < desc_.blocks_size(); ++i) {
auto *block = desc_.mutable_blocks(i);
blocks_.emplace_back(new BlockDescBind(*o.blocks_[i], block, this));
blocks_.emplace_back(new BlockDesc(*o.blocks_[i], block, this));
}
}
ProgramDescBind::ProgramDescBind(const ProgramDesc &desc) {
ProgramDesc::ProgramDesc(const proto::ProgramDesc &desc) {
desc_ = desc;
for (auto &block_desc : *desc_.mutable_blocks()) {
blocks_.emplace_back(new BlockDescBind(this, &block_desc));
blocks_.emplace_back(new BlockDesc(this, &block_desc));
}
}
ProgramDescBind::ProgramDescBind(const std::string &binary_str) {
ProgramDesc::ProgramDesc(const std::string &binary_str) {
PADDLE_ENFORCE(desc_.ParseFromString(binary_str),
"Fail to parse program_desc from binary string.");
for (auto &block_desc : *desc_.mutable_blocks()) {
blocks_.emplace_back(new BlockDescBind(this, &block_desc));
blocks_.emplace_back(new BlockDesc(this, &block_desc));
}
}
......
paddle/framework/program_desc.h
浏览文件 @ ba8660a1
......@@ -23,32 +23,32 @@ limitations under the License. */
namespace paddle {
namespace framework {
class BlockDescBind;
class BlockDesc;
class ProgramDescBind {
class ProgramDesc {
public:
ProgramDescBind();
ProgramDesc();
explicit ProgramDescBind(const ProgramDesc &desc);
explicit ProgramDesc(const proto::ProgramDesc &desc);
ProgramDescBind(const ProgramDescBind &o);
ProgramDesc(const ProgramDesc &o);
explicit ProgramDescBind(const std::string &binary_str);
explicit ProgramDesc(const std::string &binary_str);
BlockDescBind *AppendBlock(const BlockDescBind &parent);
BlockDesc *AppendBlock(const BlockDesc &parent);
BlockDescBind *MutableBlock(size_t idx) { return blocks_[idx].get(); }
BlockDesc *MutableBlock(size_t idx) { return blocks_[idx].get(); }
const BlockDescBind &Block(size_t idx) const { return *blocks_[idx]; }
const BlockDesc &Block(size_t idx) const { return *blocks_[idx]; }
size_t Size() const { return blocks_.size(); }
ProgramDesc *Proto();
proto::ProgramDesc *Proto();
private:
ProgramDesc desc_;
proto::ProgramDesc desc_;
std::vector<std::unique_ptr<BlockDescBind>> blocks_;
std::vector<std::unique_ptr<BlockDesc>> blocks_;
};
} // namespace framework
} // namespace paddle
paddle/framework/program_desc_test.cc
浏览文件 @ ba8660a1
......@@ -19,18 +19,18 @@
namespace paddle {
namespace framework {
TEST(ProgramDesc, copy_ctor) {
ProgramDescBind program;
ProgramDesc program;
auto* global_block = program.MutableBlock(0);
auto* x = global_block->Var("X");
x->SetType(VarDesc_VarType_LOD_TENSOR);
x->SetType(proto::VarDesc_VarType_LOD_TENSOR);
x->SetLoDLevel(0);
x->SetDataType(FP32);
x->SetDataType(proto::FP32);
x->SetShape({1000, 784});
auto* y = global_block->Var("Y");
y->SetType(VarDesc_VarType_LOD_TENSOR);
y->SetType(proto::VarDesc_VarType_LOD_TENSOR);
y->SetLoDLevel(0);
y->SetDataType(FP32);
y->SetDataType(proto::FP32);
y->SetShape({784, 100});
auto* op = global_block->AppendOp();
......@@ -39,15 +39,15 @@ TEST(ProgramDesc, copy_ctor) {
op->SetInput("Y", {y->Name()});
auto* out = global_block->Var("Out");
out->SetType(VarDesc_VarType_LOD_TENSOR);
out->SetType(proto::VarDesc_VarType_LOD_TENSOR);
op->SetOutput("Y", {out->Name()});
ProgramDescBind program_copy(program);
ProgramDesc program_copy(program);
auto* global_block_copy = program_copy.MutableBlock(0);
ASSERT_NE(global_block, global_block_copy);
auto assert_same_var = [&](const std::string& name, VarDescBind* var_before) {
auto assert_same_var = [&](const std::string& name, VarDesc* var_before) {
ASSERT_TRUE(global_block_copy->HasVar(name));
auto* copy = global_block_copy->Var(name);
ASSERT_NE(copy, var_before);
......@@ -81,18 +81,18 @@ TEST(ProgramDesc, copy_ctor) {
}
TEST(ProgramDescBind, serialize_and_deserialize) {
ProgramDescBind program_origin;
ProgramDesc program_origin;
auto* global_block = program_origin.MutableBlock(0);
auto* x = global_block->Var("X");
x->SetType(VarDesc_VarType_LOD_TENSOR);
x->SetType(proto::VarDesc_VarType_LOD_TENSOR);
x->SetLoDLevel(0);
x->SetDataType(FP32);
x->SetDataType(proto::FP32);
x->SetShape({1000, 784});
auto* y = global_block->Var("Y");
y->SetType(VarDesc_VarType_LOD_TENSOR);
y->SetType(proto::VarDesc_VarType_LOD_TENSOR);
y->SetLoDLevel(0);
y->SetDataType(FP32);
y->SetDataType(proto::FP32);
y->SetShape({784, 100});
auto* op = global_block->AppendOp();
......@@ -101,17 +101,17 @@ TEST(ProgramDescBind, serialize_and_deserialize) {
op->SetInput("Y", {y->Name()});
auto* out = global_block->Var("Out");
out->SetType(VarDesc_VarType_LOD_TENSOR);
out->SetType(proto::VarDesc_VarType_LOD_TENSOR);
op->SetOutput("Y", {out->Name()});
std::string binary_str;
program_origin.Proto()->SerializeToString(&binary_str);
ProgramDescBind program_restored(binary_str);
ProgramDesc program_restored(binary_str);
auto* global_block_restored = program_restored.MutableBlock(0);
ASSERT_NE(global_block, global_block_restored);
auto assert_same_var = [&](const std::string& name, VarDescBind* var_before) {
auto assert_same_var = [&](const std::string& name, VarDesc* var_before) {
ASSERT_TRUE(global_block_restored->HasVar(name));
auto* restored = global_block_restored->Var(name);
ASSERT_NE(restored, var_before);
......
paddle/framework/prune.cc
浏览文件 @ ba8660a1
......@@ -29,7 +29,7 @@ const std::string kFetchOpType = "fetch";
const std::string kDropOutOpType = "dropout";
const std::string kBatchNormOpType = "batch_norm";
bool HasDependentVar(const OpDesc& op_desc,
bool HasDependentVar(const proto::OpDesc& op_desc,
const std::set<std::string>& dependent_vars) {
for (auto& var : op_desc.outputs()) {
for (auto& argu : var.arguments()) {
......@@ -41,14 +41,15 @@ bool HasDependentVar(const OpDesc& op_desc,
return false;
}
bool IsTarget(const OpDesc& op_desc) {
bool IsTarget(const proto::OpDesc& op_desc) {
if (op_desc.has_is_target()) {
return op_desc.is_target();
}
return false;
}
void prune_impl(const ProgramDesc& input, ProgramDesc* output, int block_id) {
void prune_impl(const proto::ProgramDesc& input, proto::ProgramDesc* output,
int block_id) {
// TODO(tonyyang-svail):
// - will change to use multiple blocks for RNN op and Cond Op
......@@ -104,12 +105,12 @@ void prune_impl(const ProgramDesc& input, ProgramDesc* output, int block_id) {
}
// TODO(fengjiayi): Prune() could be inplaced to avoid unnecessary copies
void Prune(const ProgramDesc& input, ProgramDesc* output) {
void Prune(const proto::ProgramDesc& input, proto::ProgramDesc* output) {
prune_impl(input, output, 0);
}
void inference_optimize_impl(const ProgramDesc& input, ProgramDesc* output,
int block_id) {
void inference_optimize_impl(const proto::ProgramDesc& input,
proto::ProgramDesc* output, int block_id) {
*output = input;
auto* op_field = output->mutable_blocks(block_id)->mutable_ops();
for (auto& op_desc : *op_field) {
......@@ -125,7 +126,8 @@ void inference_optimize_impl(const ProgramDesc& input, ProgramDesc* output,
}
}
void InferenceOptimize(const ProgramDesc& input, ProgramDesc* output) {
void InferenceOptimize(const proto::ProgramDesc& input,
proto::ProgramDesc* output) {
inference_optimize_impl(input, output, 0);
}
......
paddle/framework/prune.h
浏览文件 @ ba8660a1
......@@ -20,9 +20,10 @@ limitations under the License. */
namespace paddle {
namespace framework {
void Prune(const ProgramDesc& input, ProgramDesc* output);
void Prune(const proto::ProgramDesc& input, proto::ProgramDesc* output);
void InferenceOptimize(const ProgramDesc& input, ProgramDesc* output);
void InferenceOptimize(const proto::ProgramDesc& input,
proto::ProgramDesc* output);
} // namespace framework
} // namespace paddle
paddle/framework/prune_test.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/framework/shape_inference.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/framework/shape_inference.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/framework/type_defs.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/framework/var_desc.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/framework/var_desc.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/framework/var_type.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/memory/memcpy.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/accuracy_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/accuracy_op.cu
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/activation_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/adadelta_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/adagrad_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/adam_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/adamax_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/assign_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/auc_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/batch_norm_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/batch_norm_op.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/beam_search_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/cast_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/cast_op.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/chunk_eval_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/clip_by_norm_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/clip_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/compare_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/concat_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/cond_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/conv_cudnn_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/conv_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/conv_op.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/conv_shift_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/cos_sim_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/crf_decoding_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/crop_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/dropout_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/dropout_op.cu
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/dropout_op.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/elementwise_op.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/expand_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/feed_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/fetch_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/fill_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/ftrl_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/gather_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/gru_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/gru_unit_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/hinge_loss_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/huber_loss_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/increment_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/is_empty_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/l1_norm_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/load_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/lod_reset_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/log_loss_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/logical_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/lookup_table_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/lrn_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/lstm_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/lstm_unit_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/math/im2col.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/matmul_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/maxout_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/mean_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/minus_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/momentum_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/mul_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/multiplex_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/nccl_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/nccl_op_test.cu.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/nce_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/op_documentation/op_markdown_format.md 0 → 100644
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/pad_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/pool_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/pool_op.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/prelu_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/proximal_gd_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/rank_loss_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/recurrent_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/recv_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/reduce_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/reorder_lod_tensor_by_rank_op.cc 0 → 100644
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/reshape_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/rmsprop_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/roi_pool_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/row_conv_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/save_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/scale_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/scatter_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/send_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/sgd_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/sign_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/softmax_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/split_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/spp_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/sum_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/top_k_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/transpose_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/unpool_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/operators/while_op.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/platform/device_context.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/platform/device_context.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/platform/gpu_info.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/platform/gpu_info.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/platform/transform_test.cu
浏览文件 @ ba8660a1
此差异已折叠。
paddle/pybind/CMakeLists.txt
浏览文件 @ ba8660a1
此差异已折叠。
paddle/pybind/const_value.cc 0 → 100644
浏览文件 @ ba8660a1
此差异已折叠。
paddle/pybind/const_value.h 0 → 100644
浏览文件 @ ba8660a1
此差异已折叠。
paddle/pybind/protobuf.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/pybind/pybind.cc
浏览文件 @ ba8660a1
此差异已折叠。
paddle/pybind/tensor_py.h
浏览文件 @ ba8660a1
此差异已折叠。
paddle/scripts/travis/build_doc.sh
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/__init__.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/clip.py 0 → 100644
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/distribute_transpiler.py 0 → 100644
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/evaluator.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/executor.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/framework.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/layers/io.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/layers/nn.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/nets.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/optimizer.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/registry.py
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/tests/test_const_value.py 0 → 100644
浏览文件 @ ba8660a1
此差异已折叠。
python/paddle/v2/fluid/tests/test_reorder_lod_tensor.py 0 → 100644
浏览文件 @ ba8660a1
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