提交 e1da6c06 编写于 作者: W wanghaoshuang

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

# Design Doc: Operation Graph Based Parameter Server
## Abstract
We propose an approach to implement the parameter server. In this
approach, there is no fundamental difference between the trainer and
the parameter server: they both run subgraphs, but subgraphs of
different purposes.
## Background
The previous implementations of the parameter server does not run a
subgraph. parameter initialization, optimizer computation, network
communication and checkpointing are implemented twice on both the
trainer and the parameter server.
It would be great if we can write code once and use them on both the
trainer and the parameter server: reduces code duplication and
improves extensibility. Given that after the current refactor, we are
representing everything as a computing graph on the
trainer. Representing everything as a computing graph on the parameter
server becomes a natural extension.
## Design
### Graph Converter
The *graph converter* converts the user-defined operation (OP) graph
into subgraphs to be scheduled on different nodes with the following
steps:
1. OP placement: the OPs will be placed on different nodes according
to heuristic that minimizes estimated total computation
time. Currently we will use a simple heuristic that puts parameter
varable on parameter server workers and everything else on trainer
workers.
1. Add communication OPs to enable the communication between nodes.
We will need these OPs: *Send*, *Recv*, *Enqueue*, *Dequeue*.
Below is an example of converting the user defined graph to the
subgraphs for the trainer and the parameter server:
<img src="src/local-graph.png" width="300"/>
After converting:
<img src="src/dist-graph.png" width="700"/>
1. The parameter variable W and it's optimizer subgraph are placed on the parameter server.
1. Operators are added to the subgraphs.
- *Send* sends data to the connected *Recv* operator. The
scheduler on the receive node will only schedule *Recv* operator
to run when the *Send* operator has ran (the *Send* OP will mark
the *Recv* OP runnable automatically).
- *Enueue* enqueues the input variable, it can block until space
become available in the queue.
- *Dequeue* outputs configurable numbers of tensors from the
queue. It will block until the queue have the required number of
tensors.
### Benefits
- Model parallelism become easier to implement: it's an extension to
the trainer - parameter server approach. we already have the
communication OPs, but need to extend the graph converter's
placement functionality.
- User-defined optimizer is easier to add - user can now express it as
a subgraph.
- No more duplication logic inside the trainer and the parameter
server mentioned in the background section.
### Challenges
- It might be hard for the graph converter to cut a general graph
(without any hint for which subgraph is the optimizer). We may need
to label which subgraph inside the OP graph is the optimizer.
- It's important to balance the parameter shards of on multiple
parameter server. If a single parameter is very big (some
word-embedding, fully connected, softmax layer), we need to
automatically partition the single parameter onto different
parameter servers when possible (only element-wise optimizer depends
on the parameter variable).
### Discussion
- In the "Aync SGD" figure, the "W" variable on the parameter server
could be read and wrote concurrently, what is our locking strategy?
E.g., each variable have a lock cpp method to be invoked by every
OP, or, have a lock OP.
- Can the Enqueue OP be implemented under our current tensor design
(puts the input tensor into the queue tensor)?
- *Dequeue* OP will have variable numbers of output (depends on the
`min_count` attribute), does our current design support it? (similar
question for the *Add* OP)
### References:
[1] [TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems](https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45166.pdf)
...@@ -123,6 +123,15 @@ OperatorBase::OperatorBase(const std::string& type, ...@@ -123,6 +123,15 @@ OperatorBase::OperatorBase(const std::string& type,
CheckAllInputOutputSet(); CheckAllInputOutputSet();
} }
std::vector<std::string> OperatorBase::InputVars() const {
std::vector<std::string> ret_val;
for (auto& o : outputs_) {
ret_val.reserve(ret_val.size() + o.second.size());
ret_val.insert(ret_val.end(), o.second.begin(), o.second.end());
}
return ret_val;
}
std::vector<std::string> OperatorBase::OutputVars(bool has_intermediate) const { std::vector<std::string> OperatorBase::OutputVars(bool has_intermediate) const {
std::vector<std::string> ret_val; std::vector<std::string> ret_val;
if (has_intermediate) { if (has_intermediate) {
......
...@@ -94,11 +94,14 @@ class OperatorBase { ...@@ -94,11 +94,14 @@ class OperatorBase {
const VariableNameMap& Inputs() const { return inputs_; } const VariableNameMap& Inputs() const { return inputs_; }
const VariableNameMap& Outputs() const { return outputs_; } const VariableNameMap& Outputs() const { return outputs_; }
//! Get a input with argument's name described in `op_proto` //! Get a input with argument's name described in `op_proto`
std::string Input(const std::string& name) const; std::string Input(const std::string& name) const;
//! Get a input which has multiple variables. //! Get a input which has multiple variables.
const std::vector<std::string>& Inputs(const std::string& name) const; const std::vector<std::string>& Inputs(const std::string& name) const;
std::vector<std::string> InputVars() const;
//! Get a output with argument's name described in `op_proto` //! Get a output with argument's name described in `op_proto`
std::string Output(const std::string& name) const; std::string Output(const std::string& name) const;
//! Get an output which has multiple variables. //! Get an output which has multiple variables.
...@@ -311,9 +314,9 @@ class InferShapeContext { ...@@ -311,9 +314,9 @@ class InferShapeContext {
} }
template <typename T> template <typename T>
std::vector<const T*> MultiOutput(const std::string& name) const { std::vector<T*> MultiOutput(const std::string& name) const {
auto names = op_.Outputs(name); auto names = op_.Outputs(name);
std::vector<const T*> res; std::vector<T*> res;
res.reserve(names.size()); res.reserve(names.size());
std::transform(names.begin(), names.end(), std::back_inserter(res), std::transform(names.begin(), names.end(), std::back_inserter(res),
[&](const std::string& sub_name) { [&](const std::string& sub_name) {
......
...@@ -62,14 +62,18 @@ void BatchNormBaseLayer::calFeatureMapSize() { ...@@ -62,14 +62,18 @@ void BatchNormBaseLayer::calFeatureMapSize() {
const ImageConfig& conf = config_.inputs(0).image_conf(); const ImageConfig& conf = config_.inputs(0).image_conf();
imageH_ = inputLayers_[0]->getOutput().getFrameHeight(); imageH_ = inputLayers_[0]->getOutput().getFrameHeight();
imageW_ = inputLayers_[0]->getOutput().getFrameWidth(); imageW_ = inputLayers_[0]->getOutput().getFrameWidth();
imageD_ = inputLayers_[0]->getOutput().getFrameDepth();
if (0 == imageD_) imageD_ = conf.img_size_z();
if (imageH_ == 0 && imageW_ == 0) { if (imageH_ == 0 && imageW_ == 0) {
imageH_ = conf.has_img_size_y() ? conf.img_size_y() : conf.img_size(); imageH_ = conf.has_img_size_y() ? conf.img_size_y() : conf.img_size();
imageW_ = conf.img_size(); imageW_ = conf.img_size();
} else { } else {
getOutput().setFrameHeight(imageH_); getOutput().setFrameHeight(imageH_);
getOutput().setFrameWidth(imageW_); getOutput().setFrameWidth(imageW_);
getOutput().setFrameDepth(imageD_);
} }
imgPixels_ = imageH_ * imageW_; imgPixels_ = imageH_ * imageW_ * imageD_;
} }
} // namespace paddle } // namespace paddle
...@@ -80,6 +80,7 @@ protected: ...@@ -80,6 +80,7 @@ protected:
/// Height or width of input image feature. /// Height or width of input image feature.
/// Both of them are 1 if the input is fully-connected layer. /// Both of them are 1 if the input is fully-connected layer.
int imageD_;
int imageH_; int imageH_;
int imageW_; int imageW_;
/// Height * Width. /// Height * Width.
......
...@@ -37,7 +37,7 @@ bool CudnnBatchNormLayer::init(const LayerMap& layerMap, ...@@ -37,7 +37,7 @@ bool CudnnBatchNormLayer::init(const LayerMap& layerMap,
} }
void CudnnBatchNormLayer::reshape(int batchSize) { void CudnnBatchNormLayer::reshape(int batchSize) {
hl_tensor_reshape(ioDesc_, batchSize, channels_, imageH_, imageW_); hl_tensor_reshape(ioDesc_, batchSize, channels_, imageH_ * imageD_, imageW_);
} }
void CudnnBatchNormLayer::forward(PassType passType) { void CudnnBatchNormLayer::forward(PassType passType) {
...@@ -104,7 +104,7 @@ void CudnnBatchNormLayer::forward(PassType passType) { ...@@ -104,7 +104,7 @@ void CudnnBatchNormLayer::forward(PassType passType) {
EPS, EPS,
batchSize, batchSize,
channels_, channels_,
imageH_, imageH_ * imageD_,
imageW_); imageW_);
} }
} }
......
...@@ -24,10 +24,12 @@ bool SwitchOrderLayer::init(const LayerMap& layerMap, ...@@ -24,10 +24,12 @@ bool SwitchOrderLayer::init(const LayerMap& layerMap,
/* Initialize the basic parent class */ /* Initialize the basic parent class */
Layer::init(layerMap, parameterMap); Layer::init(layerMap, parameterMap);
auto& img_conf = config_.inputs(0).image_conf(); auto& img_conf = config_.inputs(0).image_conf();
size_t inD = img_conf.img_size_z();
size_t inH = size_t inH =
img_conf.has_img_size_y() ? img_conf.img_size_y() : img_conf.img_size(); img_conf.has_img_size_y() ? img_conf.img_size_y() : img_conf.img_size();
size_t inW = img_conf.img_size(); size_t inW = img_conf.img_size();
size_t inC = img_conf.channels(); size_t inC = img_conf.channels();
inH = inH * inD;
inDims_ = TensorShape({0, inC, inH, inW}); inDims_ = TensorShape({0, inC, inH, inW});
outDims_ = TensorShape(4); outDims_ = TensorShape(4);
...@@ -64,9 +66,10 @@ void SwitchOrderLayer::setInDims() { ...@@ -64,9 +66,10 @@ void SwitchOrderLayer::setInDims() {
MatrixPtr input = inputLayers_[0]->getOutputValue(); MatrixPtr input = inputLayers_[0]->getOutputValue();
size_t batchSize = input->getHeight(); size_t batchSize = input->getHeight();
inDims_.setDim(0, batchSize); inDims_.setDim(0, batchSize);
int d = inputLayers_[0]->getOutput().getFrameDepth();
d = (d == 0 ? 1 : d);
int h = inputLayers_[0]->getOutput().getFrameHeight(); int h = inputLayers_[0]->getOutput().getFrameHeight();
if (h != 0) inDims_.setDim(2, h); if (h != 0) inDims_.setDim(2, h * d);
int w = inputLayers_[0]->getOutput().getFrameWidth(); int w = inputLayers_[0]->getOutput().getFrameWidth();
if (w != 0) inDims_.setDim(3, w); if (w != 0) inDims_.setDim(3, w);
int totalCount = input->getElementCnt(); int totalCount = input->getElementCnt();
......
...@@ -1703,6 +1703,55 @@ TEST(Layer, BatchNormalizationLayer) { ...@@ -1703,6 +1703,55 @@ TEST(Layer, BatchNormalizationLayer) {
#endif #endif
} }
void testBatchNorm3DLayer(const string& type, bool trans, bool useGpu) {
TestConfig config;
const int CHANNELS = 10;
const int IMG_SIZE = 16;
const int IMG_SIZE_Y = 8;
const int IMG_SIZE_Z = 8;
size_t size = CHANNELS * IMG_SIZE * IMG_SIZE_Y * IMG_SIZE_Z;
config.layerConfig.set_type(type);
config.layerConfig.set_size(size);
config.layerConfig.set_active_type("sigmoid");
config.biasSize = CHANNELS;
config.inputDefs.push_back({INPUT_DATA,
"layer_0",
/* dim= */ size,
/* paraSize= */ CHANNELS});
config.inputDefs.push_back({INPUT_DATA, "layer_1_running_mean", 1, CHANNELS});
config.inputDefs.back().isStatic = true;
config.inputDefs.push_back({INPUT_DATA, "layer_2_running_var", 1, CHANNELS});
config.inputDefs.back().isStatic = true;
LayerInputConfig* input = config.layerConfig.add_inputs();
config.layerConfig.add_inputs();
config.layerConfig.add_inputs();
ImageConfig* img_conf = input->mutable_image_conf();
img_conf->set_channels(CHANNELS);
img_conf->set_img_size(IMG_SIZE);
img_conf->set_img_size_y(IMG_SIZE_Y);
img_conf->set_img_size_z(IMG_SIZE_Z);
testLayerGrad(config,
"batch_norm",
64,
/* trans= */ trans,
useGpu,
/* useWeight */ true);
}
TEST(Layer, testBatchNorm3DLayer) {
testBatchNorm3DLayer("batch_norm", false, false);
#ifndef PADDLE_ONLY_CPU
testBatchNorm3DLayer("batch_norm", false, true);
if (hl_get_cudnn_lib_version() >= int(4000)) {
testBatchNorm3DLayer("cudnn_batch_norm", false, true);
}
#endif
}
void testConvOperator(bool isDeconv) { void testConvOperator(bool isDeconv) {
TestConfig config; TestConfig config;
const int NUM_FILTERS = 16; const int NUM_FILTERS = 16;
......
/* 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. */
#include "paddle/operators/sum_op.h"
#include <vector>
namespace paddle {
namespace operators {
using framework::Tensor;
class SumOp : public framework::OperatorWithKernel {
public:
using framework::OperatorWithKernel::OperatorWithKernel;
protected:
void InferShape(const framework::InferShapeContext &ctx) const override {
auto ins = ctx.MultiInput<framework::Tensor>("X");
auto *out = ctx.Output<framework::Tensor>("Out");
int N = ins.size();
auto in_dim = ins[0]->dims();
PADDLE_ENFORCE_GT(N, 1, "Input tensors count should > 1.");
for (int i = 1; i < N; i++) {
auto dim = ins[i]->dims();
PADDLE_ENFORCE(in_dim == dim, "Input tensors must have same shape");
}
out->Resize(in_dim);
}
};
class SumOpMaker : public framework::OpProtoAndCheckerMaker {
public:
SumOpMaker(framework::OpProto *proto, framework::OpAttrChecker *op_checker)
: OpProtoAndCheckerMaker(proto, op_checker) {
AddInput("X", "the input tensors of sum operator.").AsDuplicable();
AddOutput("Out", "the output tensor of sum operator.");
AddComment(R"DOC(
Sum the input tensors.
)DOC");
}
};
class SumGradOp : public framework::OperatorWithKernel {
public:
using framework::OperatorWithKernel::OperatorWithKernel;
protected:
void InferShape(const framework::InferShapeContext &ctx) const override {
auto outputs = ctx.MultiOutput<Tensor>(framework::GradVarName("X"));
auto dims = ctx.Input<Tensor>(framework::GradVarName("Out"))->dims();
for (auto output : outputs) {
output->Resize(dims);
}
}
};
} // namespace operators
} // namespace paddle
namespace ops = paddle::operators;
REGISTER_OP(sum, ops::SumOp, ops::SumOpMaker, sum_grad, ops::SumGradOp);
REGISTER_OP_CPU_KERNEL(sum, ops::SumKernel<paddle::platform::CPUPlace, float>);
REGISTER_OP_CPU_KERNEL(sum_grad,
ops::SumGradKernel<paddle::platform::CPUPlace, float>);
/* 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. */
#define EIGEN_USE_GPU
#include "paddle/operators/sum_op.h"
namespace ops = paddle::operators;
REGISTER_OP_GPU_KERNEL(sum, ops::SumKernel<paddle::platform::GPUPlace, float>);
REGISTER_OP_GPU_KERNEL(sum_grad,
ops::SumGradKernel<paddle::platform::GPUPlace, float>);
/* 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
#include "paddle/framework/eigen.h"
#include "paddle/framework/op_registry.h"
namespace paddle {
namespace operators {
using Tensor = framework::Tensor;
template <typename T, int MajorType = Eigen::RowMajor,
typename IndexType = Eigen::DenseIndex>
using EigenVector = framework::EigenVector<T, MajorType, IndexType>;
template <typename Place, typename T>
class SumKernel : public framework::OpKernel {
public:
void Compute(const framework::ExecutionContext& context) const override {
auto ins = context.MultiInput<Tensor>("X");
auto* out = context.Output<Tensor>("Out");
out->mutable_data<T>(context.GetPlace());
auto place = context.GetEigenDevice<Place>();
auto result = EigenVector<T>::Flatten(*out);
int N = ins.size();
auto in = EigenVector<T>::Flatten(*(ins[0]));
result.device(place) = in;
for (int i = 1; i < N; i++) {
auto in = EigenVector<T>::Flatten(*(ins[i]));
result.device(place) = result + in;
}
}
};
template <typename Place, typename T>
class SumGradKernel : public framework::OpKernel {
public:
void Compute(const framework::ExecutionContext& context) const override {
auto* input = context.Input<Tensor>(framework::GradVarName("Out"));
auto outs = context.MultiOutput<Tensor>(framework::GradVarName("X"));
for (auto out : outs) {
out->mutable_data<T>(context.GetPlace());
}
auto place = context.GetEigenDevice<Place>();
auto in = EigenVector<T>::Flatten(*input);
for (auto out : outs) {
auto result = EigenVector<T>::Flatten(*out);
result.device(place) = in;
}
}
};
} // namespace operators
} // namespace paddle
...@@ -52,6 +52,7 @@ USE_OP(pad); ...@@ -52,6 +52,7 @@ USE_OP(pad);
USE_CPU_ONLY_OP(scatter); USE_CPU_ONLY_OP(scatter);
USE_OP(top_k); USE_OP(top_k);
USE_OP(squared_l2_distance); USE_OP(squared_l2_distance);
USE_OP(sum);
namespace paddle { namespace paddle {
namespace framework { namespace framework {
...@@ -217,7 +218,10 @@ All parameter, weight, gradient are variables in Paddle. ...@@ -217,7 +218,10 @@ All parameter, weight, gradient are variables in Paddle.
-> std::map<std::string, std::vector<std::string>> { -> std::map<std::string, std::vector<std::string>> {
return op.Outputs(); return op.Outputs();
}) })
.def("output_vars",
[](const OperatorBase &op) { return op.OutputVars(true); })
.def("inputs", [](const OperatorBase &op) { return op.Inputs(); }) .def("inputs", [](const OperatorBase &op) { return op.Inputs(); })
.def("input_vars", [](const OperatorBase &op) { return op.InputVars(); })
.def("__str__", &OperatorBase::DebugString) .def("__str__", &OperatorBase::DebugString)
.def("no_intermediate_outputs", .def("no_intermediate_outputs",
[](const OperatorBase &op) { return op.OutputVars(false); }) [](const OperatorBase &op) { return op.OutputVars(false); })
......
...@@ -271,6 +271,7 @@ message ImageConfig { ...@@ -271,6 +271,7 @@ message ImageConfig {
// The size of input feature map. // The size of input feature map.
required uint32 img_size = 8; required uint32 img_size = 8;
optional uint32 img_size_y = 9; optional uint32 img_size_y = 9;
optional uint32 img_size_z = 10 [ default = 1 ];
} }
message PriorBoxConfig { message PriorBoxConfig {
...@@ -519,6 +520,7 @@ message LayerConfig { ...@@ -519,6 +520,7 @@ message LayerConfig {
// for HuberRegressionLoss // for HuberRegressionLoss
optional double delta = 57 [ default = 1.0 ]; optional double delta = 57 [ default = 1.0 ];
// for 3D data
optional uint64 depth = 58 [ default = 1 ]; optional uint64 depth = 58 [ default = 1 ];
// for switch order layer // for switch order layer
......
...@@ -1332,6 +1332,12 @@ def parse_image(image, input_layer_name, image_conf): ...@@ -1332,6 +1332,12 @@ def parse_image(image, input_layer_name, image_conf):
get_img_size(input_layer_name, image_conf.channels) get_img_size(input_layer_name, image_conf.channels)
def parse_image3d(image, input_layer_name, image_conf):
image_conf.channels = image.channels
image_conf.img_size, image_conf.img_size_y, image_conf.img_size_z = \
get_img3d_size(input_layer_name, image_conf.channels)
def parse_norm(norm, input_layer_name, norm_conf): def parse_norm(norm, input_layer_name, norm_conf):
norm_conf.norm_type = norm.norm_type norm_conf.norm_type = norm.norm_type
config_assert( config_assert(
...@@ -2365,9 +2371,11 @@ class BatchNormLayer(LayerBase): ...@@ -2365,9 +2371,11 @@ class BatchNormLayer(LayerBase):
name, name,
inputs, inputs,
bias=True, bias=True,
img3D=False,
use_global_stats=True, use_global_stats=True,
moving_average_fraction=0.9, moving_average_fraction=0.9,
batch_norm_type=None, batch_norm_type=None,
mean_var_names=None,
**xargs): **xargs):
if inputs is None: if inputs is None:
inputs = [] inputs = []
...@@ -2409,24 +2417,69 @@ class BatchNormLayer(LayerBase): ...@@ -2409,24 +2417,69 @@ class BatchNormLayer(LayerBase):
input_layer = self.get_input_layer(0) input_layer = self.get_input_layer(0)
image_conf = self.config.inputs[0].image_conf image_conf = self.config.inputs[0].image_conf
if img3D:
parse_image3d(self.inputs[0].image, input_layer.name, image_conf)
# Only pass the width and height of input to batch_norm layer
# when either of it is non-zero.
if input_layer.width != 0 or input_layer.height != 0:
self.set_cnn_layer(
input_layer_name=name,
depth=image_conf.img_size_z,
height=image_conf.img_size_y,
width=image_conf.img_size,
channels=image_conf.channels,
is_print=True)
else:
self.set_layer_size(input_layer.size)
else:
parse_image(self.inputs[0].image, input_layer.name, image_conf) parse_image(self.inputs[0].image, input_layer.name, image_conf)
# Only pass the width and height of input to batch_norm layer # Only pass the width and height of input to batch_norm layer
# when either of it is non-zero. # when either of it is non-zero.
if input_layer.width != 0 or input_layer.height != 0: if input_layer.width != 0 or input_layer.height != 0:
self.set_cnn_layer(name, image_conf.img_size_y, image_conf.img_size, self.set_cnn_layer(
image_conf.channels, False) input_layer_name=name,
height=image_conf.img_size_y,
width=image_conf.img_size,
channels=image_conf.channels,
is_print=True)
else: else:
self.set_layer_size(input_layer.size) self.set_layer_size(input_layer.size)
psize = self.calc_parameter_size(image_conf) psize = self.calc_parameter_size(image_conf)
dims = [1, psize] dims = [1, psize]
if mean_var_names is not None:
assert len(mean_var_names) == 2
self.inputs[1].parameter_name = mean_var_names[0]
self.inputs[2].parameter_name = mean_var_names[1]
self.create_input_parameter(0, psize) self.create_input_parameter(0, psize)
self.create_input_parameter(1, psize, dims) self.create_input_parameter(1, psize, dims)
self.create_input_parameter(2, psize, dims) self.create_input_parameter(2, psize, dims)
self.create_bias_parameter(bias, psize) self.create_bias_parameter(bias, psize)
def set_cnn_layer(self,
input_layer_name,
depth=None,
height=None,
width=None,
channels=None,
is_print=True):
depthIsNone = False
if depth is None:
depth = 1
depthIsNone = True
size = depth * height * width * channels
self.set_layer_size(size)
self.set_layer_height_width(height, width)
self.set_layer_depth(depth)
if is_print and depthIsNone:
print("output for %s: c = %d, h = %d, w = %d, size = %d" %
(input_layer_name, channels, height, width, size))
elif is_print:
print("output for %s: c = %d, d = %d, h = %d, w = %d, size = %d" %
(input_layer_name, channels, depth, height, width, size))
def calc_parameter_size(self, image_conf): def calc_parameter_size(self, image_conf):
return image_conf.channels return image_conf.channels
...@@ -2688,9 +2741,20 @@ class AddToLayer(LayerBase): ...@@ -2688,9 +2741,20 @@ class AddToLayer(LayerBase):
super(AddToLayer, self).__init__( super(AddToLayer, self).__init__(
name, 'addto', 0, inputs=inputs, **xargs) name, 'addto', 0, inputs=inputs, **xargs)
config_assert(len(inputs) > 0, 'inputs cannot be empty for AddToLayer') config_assert(len(inputs) > 0, 'inputs cannot be empty for AddToLayer')
if len(self.inputs) > 1:
for input_index in xrange(len(self.inputs)): for input_index in xrange(len(self.inputs)):
input_layer = self.get_input_layer(input_index) assert self.get_input_layer(0).height == self.get_input_layer(
self.set_layer_size(input_layer.size) input_index).height
assert self.get_input_layer(0).width == self.get_input_layer(
input_index).width
assert self.get_input_layer(0).depth == self.get_input_layer(
input_index).depth
self.set_layer_size(self.get_input_layer(0).size)
self.set_layer_height_width(self.get_input_layer(0).height, \
self.get_input_layer(0).width)
self.set_layer_depth(self.get_input_layer(0).depth)
self.create_bias_parameter(bias, self.config.size) self.create_bias_parameter(bias, self.config.size)
...@@ -3370,11 +3434,20 @@ class ConcatenateLayer(LayerBase): ...@@ -3370,11 +3434,20 @@ class ConcatenateLayer(LayerBase):
name, 'concat', 0, inputs=inputs, **xargs) name, 'concat', 0, inputs=inputs, **xargs)
size = 0 size = 0
for input_index in xrange(len(self.inputs)): for input_index in xrange(len(self.inputs)):
assert self.get_input_layer(0).height == self.get_input_layer(
input_index).height
assert self.get_input_layer(0).width == self.get_input_layer(
input_index).width
assert self.get_input_layer(0).depth == self.get_input_layer(
input_index).depth
input_layer = self.get_input_layer(input_index) input_layer = self.get_input_layer(input_index)
input = self.inputs[input_index] input = self.inputs[input_index]
if self.config.size == 0: if self.config.size == 0:
size += input_layer.size size += input_layer.size
self.set_layer_height_width(self.get_input_layer(0).height, \
self.get_input_layer(0).width)
self.set_layer_depth(self.get_input_layer(0).depth)
self.set_layer_size(size) self.set_layer_size(size)
......
...@@ -354,6 +354,10 @@ class LayerOutput(object): ...@@ -354,6 +354,10 @@ class LayerOutput(object):
def height(self): def height(self):
return cp.g_layer_map[self.full_name].height return cp.g_layer_map[self.full_name].height
@property
def depth(self):
return cp.g_layer_map[self.full_name].depth
def set_input(self, input): def set_input(self, input):
""" """
Set the input for a memory layer. Can only be used for memory layer Set the input for a memory layer. Can only be used for memory layer
...@@ -2953,13 +2957,15 @@ def img_cmrnorm_layer(input, ...@@ -2953,13 +2957,15 @@ def img_cmrnorm_layer(input,
def batch_norm_layer(input, def batch_norm_layer(input,
act=None, act=None,
name=None, name=None,
img3D=False,
num_channels=None, num_channels=None,
bias_attr=None, bias_attr=None,
param_attr=None, param_attr=None,
layer_attr=None, layer_attr=None,
batch_norm_type=None, batch_norm_type=None,
moving_average_fraction=0.9, moving_average_fraction=0.9,
use_global_stats=None): use_global_stats=None,
mean_var_names=None):
""" """
Batch Normalization Layer. The notation of this layer as follow. Batch Normalization Layer. The notation of this layer as follow.
...@@ -3026,6 +3032,8 @@ def batch_norm_layer(input, ...@@ -3026,6 +3032,8 @@ def batch_norm_layer(input,
:math:`runningMean = newMean*(1-factor) :math:`runningMean = newMean*(1-factor)
+ runningMean*factor` + runningMean*factor`
:type moving_average_fraction: float. :type moving_average_fraction: float.
:param mean_var_names: [mean name, variance name]
:type mean_var_names: string list
:return: LayerOutput object. :return: LayerOutput object.
:rtype: LayerOutput :rtype: LayerOutput
""" """
...@@ -3039,6 +3047,7 @@ def batch_norm_layer(input, ...@@ -3039,6 +3047,7 @@ def batch_norm_layer(input,
(batch_norm_type == "cudnn_batch_norm") (batch_norm_type == "cudnn_batch_norm")
l = Layer( l = Layer(
name=name, name=name,
img3D=img3D,
inputs=Input( inputs=Input(
input.name, image=Image(channels=num_channels), **param_attr.attr), input.name, image=Image(channels=num_channels), **param_attr.attr),
active_type=act.name, active_type=act.name,
...@@ -3047,6 +3056,7 @@ def batch_norm_layer(input, ...@@ -3047,6 +3056,7 @@ def batch_norm_layer(input,
bias=ParamAttr.to_bias(bias_attr), bias=ParamAttr.to_bias(bias_attr),
moving_average_fraction=moving_average_fraction, moving_average_fraction=moving_average_fraction,
use_global_stats=use_global_stats, use_global_stats=use_global_stats,
mean_var_names=mean_var_names,
**ExtraLayerAttribute.to_kwargs(layer_attr)) **ExtraLayerAttribute.to_kwargs(layer_attr))
return LayerOutput( return LayerOutput(
...@@ -6410,7 +6420,7 @@ def gated_unit_layer(input, ...@@ -6410,7 +6420,7 @@ def gated_unit_layer(input,
@wrap_name_default('switch_order') @wrap_name_default('switch_order')
def switch_order_layer(input, def switch_order_layer(input,
name=None, name=None,
reshape=None, reshape_axis=None,
act=None, act=None,
layer_attr=None): layer_attr=None):
""" """
...@@ -6421,8 +6431,9 @@ def switch_order_layer(input, ...@@ -6421,8 +6431,9 @@ def switch_order_layer(input,
The example usage is: The example usage is:
.. code-block:: python .. code-block:: python
reshape_axis = 3
switch = switch_order(input=layer, name='switch', reshape_axis=reshape_axis)
reshape = {'height':[ 0, 1, 2], 'width':[3]} reshape = {'height':[ 0, 1, 2], 'width':[3]}
switch = switch_order(input=layer, name='switch', reshape=reshape)
:param input: The input layer. :param input: The input layer.
:type input: LayerOutput :type input: LayerOutput
...@@ -6434,6 +6445,11 @@ def switch_order_layer(input, ...@@ -6434,6 +6445,11 @@ def switch_order_layer(input,
:rtype: LayerOutput :rtype: LayerOutput
""" """
assert isinstance(input, LayerOutput) assert isinstance(input, LayerOutput)
assert reshape_axis != None and (reshape_axis > 0 and reshape_axis < 4)
height = [ele for ele in xrange(reshape_axis)]
width = [ele for ele in range(reshape_axis, 4)]
reshape = {'height': height, 'width': width}
l = Layer( l = Layer(
name=name, name=name,
inputs=input.name, inputs=input.name,
......
...@@ -10,6 +10,6 @@ test_prelu_layer test_row_conv test_detection_output_layer test_multibox_loss_la ...@@ -10,6 +10,6 @@ test_prelu_layer test_row_conv test_detection_output_layer test_multibox_loss_la
test_recursive_topology test_gated_unit_layer test_clip_layer test_row_l2_norm_layer test_recursive_topology test_gated_unit_layer test_clip_layer test_row_l2_norm_layer
test_kmax_seq_socre_layer test_sub_nested_seq_select_layer test_scale_shift_layer test_kmax_seq_socre_layer test_sub_nested_seq_select_layer test_scale_shift_layer
test_seq_slice_layer test_cross_entropy_over_beam test_pooling3D_layer test_seq_slice_layer test_cross_entropy_over_beam test_pooling3D_layer
test_conv3d_layer test_deconv3d_layer) test_conv3d_layer test_deconv3d_layer test_BatchNorm3D)
export whole_configs=(test_split_datasource) export whole_configs=(test_split_datasource)
...@@ -62,6 +62,7 @@ layers { ...@@ -62,6 +62,7 @@ layers {
moving_average_fraction: 0.9 moving_average_fraction: 0.9
height: 227 height: 227
width: 227 width: 227
depth: 1
} }
layers { layers {
name: "__crmnorm_0__" name: "__crmnorm_0__"
......
...@@ -62,6 +62,7 @@ layers { ...@@ -62,6 +62,7 @@ layers {
moving_average_fraction: 0.9 moving_average_fraction: 0.9
height: 256 height: 256
width: 256 width: 256
depth: 1
} }
layers { layers {
name: "__crmnorm_0__" name: "__crmnorm_0__"
......
type: "nn"
layers {
name: "data3D"
type: "data"
size: 360
active_type: ""
height: 6
width: 20
depth: 3
}
layers {
name: "__batch_norm_0__"
type: "batch_norm"
size: 360
active_type: "relu"
inputs {
input_layer_name: "data3D"
input_parameter_name: "___batch_norm_0__.w0"
image_conf {
channels: 1
img_size: 20
img_size_y: 6
img_size_z: 3
}
}
inputs {
input_layer_name: "data3D"
input_parameter_name: "___batch_norm_0__.w1"
}
inputs {
input_layer_name: "data3D"
input_parameter_name: "___batch_norm_0__.w2"
}
bias_parameter_name: "___batch_norm_0__.wbias"
moving_average_fraction: 0.9
height: 6
width: 20
depth: 3
}
parameters {
name: "___batch_norm_0__.w0"
size: 1
initial_mean: 1.0
initial_std: 0.0
initial_strategy: 0
initial_smart: false
}
parameters {
name: "___batch_norm_0__.w1"
size: 1
initial_mean: 0.0
initial_std: 0.0
dims: 1
dims: 1
initial_strategy: 0
initial_smart: false
is_static: true
is_shared: true
}
parameters {
name: "___batch_norm_0__.w2"
size: 1
initial_mean: 0.0
initial_std: 0.0
dims: 1
dims: 1
initial_strategy: 0
initial_smart: false
is_static: true
is_shared: true
}
parameters {
name: "___batch_norm_0__.wbias"
size: 1
initial_mean: 0.0
initial_std: 0.0
dims: 1
dims: 1
initial_strategy: 0
initial_smart: false
}
input_layer_names: "data3D"
output_layer_names: "__batch_norm_0__"
sub_models {
name: "root"
layer_names: "data3D"
layer_names: "__batch_norm_0__"
input_layer_names: "data3D"
output_layer_names: "__batch_norm_0__"
is_recurrent_layer_group: false
}
...@@ -74,6 +74,9 @@ layers { ...@@ -74,6 +74,9 @@ layers {
inputs { inputs {
input_layer_name: "__bidirectional_gru_0___bw" input_layer_name: "__bidirectional_gru_0___bw"
} }
height: 0
width: 0
depth: 1
} }
parameters { parameters {
name: "___bidirectional_gru_0___fw_transform.w0" name: "___bidirectional_gru_0___fw_transform.w0"
......
...@@ -16,6 +16,9 @@ layers { ...@@ -16,6 +16,9 @@ layers {
inputs { inputs {
input_layer_name: "data" input_layer_name: "data"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_1__" name: "__addto_1__"
...@@ -28,6 +31,9 @@ layers { ...@@ -28,6 +31,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_0__" input_layer_name: "__addto_0__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_2__" name: "__addto_2__"
...@@ -40,6 +46,9 @@ layers { ...@@ -40,6 +46,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_1__" input_layer_name: "__addto_1__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_3__" name: "__addto_3__"
...@@ -52,6 +61,9 @@ layers { ...@@ -52,6 +61,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_2__" input_layer_name: "__addto_2__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_4__" name: "__addto_4__"
...@@ -64,6 +76,9 @@ layers { ...@@ -64,6 +76,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_3__" input_layer_name: "__addto_3__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_5__" name: "__addto_5__"
...@@ -76,6 +91,9 @@ layers { ...@@ -76,6 +91,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_4__" input_layer_name: "__addto_4__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_6__" name: "__addto_6__"
...@@ -88,6 +106,9 @@ layers { ...@@ -88,6 +106,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_5__" input_layer_name: "__addto_5__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_7__" name: "__addto_7__"
...@@ -100,6 +121,9 @@ layers { ...@@ -100,6 +121,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_6__" input_layer_name: "__addto_6__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_8__" name: "__addto_8__"
...@@ -112,6 +136,9 @@ layers { ...@@ -112,6 +136,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_7__" input_layer_name: "__addto_7__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_9__" name: "__addto_9__"
...@@ -124,6 +151,9 @@ layers { ...@@ -124,6 +151,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_8__" input_layer_name: "__addto_8__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_10__" name: "__addto_10__"
...@@ -136,6 +166,9 @@ layers { ...@@ -136,6 +166,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_9__" input_layer_name: "__addto_9__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_11__" name: "__addto_11__"
...@@ -148,6 +181,9 @@ layers { ...@@ -148,6 +181,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_10__" input_layer_name: "__addto_10__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_12__" name: "__addto_12__"
...@@ -160,6 +196,9 @@ layers { ...@@ -160,6 +196,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_11__" input_layer_name: "__addto_11__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_13__" name: "__addto_13__"
...@@ -172,6 +211,9 @@ layers { ...@@ -172,6 +211,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_12__" input_layer_name: "__addto_12__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_14__" name: "__addto_14__"
...@@ -184,6 +226,9 @@ layers { ...@@ -184,6 +226,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_13__" input_layer_name: "__addto_13__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_15__" name: "__addto_15__"
...@@ -196,6 +241,9 @@ layers { ...@@ -196,6 +241,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_14__" input_layer_name: "__addto_14__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_16__" name: "__addto_16__"
...@@ -208,6 +256,9 @@ layers { ...@@ -208,6 +256,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_15__" input_layer_name: "__addto_15__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_17__" name: "__addto_17__"
...@@ -220,6 +271,9 @@ layers { ...@@ -220,6 +271,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_16__" input_layer_name: "__addto_16__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_18__" name: "__addto_18__"
...@@ -232,6 +286,9 @@ layers { ...@@ -232,6 +286,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_17__" input_layer_name: "__addto_17__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_19__" name: "__addto_19__"
...@@ -244,6 +301,9 @@ layers { ...@@ -244,6 +301,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_18__" input_layer_name: "__addto_18__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_20__" name: "__addto_20__"
...@@ -256,6 +316,9 @@ layers { ...@@ -256,6 +316,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_19__" input_layer_name: "__addto_19__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_21__" name: "__addto_21__"
...@@ -268,6 +331,9 @@ layers { ...@@ -268,6 +331,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_20__" input_layer_name: "__addto_20__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_22__" name: "__addto_22__"
...@@ -280,6 +346,9 @@ layers { ...@@ -280,6 +346,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_21__" input_layer_name: "__addto_21__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_23__" name: "__addto_23__"
...@@ -292,6 +361,9 @@ layers { ...@@ -292,6 +361,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_22__" input_layer_name: "__addto_22__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_24__" name: "__addto_24__"
...@@ -304,6 +376,9 @@ layers { ...@@ -304,6 +376,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_23__" input_layer_name: "__addto_23__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_25__" name: "__addto_25__"
...@@ -316,6 +391,9 @@ layers { ...@@ -316,6 +391,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_24__" input_layer_name: "__addto_24__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_26__" name: "__addto_26__"
...@@ -328,6 +406,9 @@ layers { ...@@ -328,6 +406,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_25__" input_layer_name: "__addto_25__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_27__" name: "__addto_27__"
...@@ -340,6 +421,9 @@ layers { ...@@ -340,6 +421,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_26__" input_layer_name: "__addto_26__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_28__" name: "__addto_28__"
...@@ -352,6 +436,9 @@ layers { ...@@ -352,6 +436,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_27__" input_layer_name: "__addto_27__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_29__" name: "__addto_29__"
...@@ -364,6 +451,9 @@ layers { ...@@ -364,6 +451,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_28__" input_layer_name: "__addto_28__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_30__" name: "__addto_30__"
...@@ -376,6 +466,9 @@ layers { ...@@ -376,6 +466,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_29__" input_layer_name: "__addto_29__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__addto_31__" name: "__addto_31__"
...@@ -388,6 +481,9 @@ layers { ...@@ -388,6 +481,9 @@ layers {
inputs { inputs {
input_layer_name: "__addto_30__" input_layer_name: "__addto_30__"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__fc_layer_0__" name: "__fc_layer_0__"
......
...@@ -22,6 +22,9 @@ layers { ...@@ -22,6 +22,9 @@ layers {
inputs { inputs {
input_layer_name: "b" input_layer_name: "b"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__concat_0__" name: "__concat_0__"
...@@ -34,6 +37,9 @@ layers { ...@@ -34,6 +37,9 @@ layers {
inputs { inputs {
input_layer_name: "b" input_layer_name: "b"
} }
height: 0
width: 0
depth: 1
} }
layers { layers {
name: "__concat_1__" name: "__concat_1__"
......
from paddle.trainer_config_helpers import *
settings(batch_size=1000, learning_rate=1e-4)
#data = data_layer(name='data', size=180, width=30, height=6)
#batchNorm = batch_norm_layer(data, num_channels=1)
#outputs(batchNorm)
data3D = data_layer(name='data3D', size=120 * 3, width=20, height=6, depth=3)
batchNorm3D = batch_norm_layer(data3D, num_channels=1, img3D=True)
outputs(batchNorm3D)
...@@ -142,8 +142,8 @@ def create_op_creation_method(op_proto): ...@@ -142,8 +142,8 @@ def create_op_creation_method(op_proto):
return OpInfo( return OpInfo(
method=__impl__, method=__impl__,
name=op_proto.type, name=op_proto.type,
inputs=[var.name for var in op_proto.inputs], inputs=[(var.name, var.duplicable) for var in op_proto.inputs],
outputs=[var.name for var in op_proto.outputs], outputs=[(var.name, var.duplicable) for var in op_proto.outputs],
attrs=[attr.name for attr in op_proto.attrs]) attrs=[attr.name for attr in op_proto.attrs])
...@@ -180,9 +180,15 @@ class OperatorFactory(object): ...@@ -180,9 +180,15 @@ class OperatorFactory(object):
return self.op_methods.get(t) return self.op_methods.get(t)
def get_op_input_names(self, type): def get_op_input_names(self, type):
return map(lambda x: x[0], self.get_op_info(type).inputs)
def get_op_inputs(self, type):
return self.get_op_info(type).inputs return self.get_op_info(type).inputs
def get_op_output_names(self, type): def get_op_output_names(self, type):
return map(lambda x: x[0], self.get_op_info(type).outputs)
def get_op_outputs(self, type):
return self.get_op_info(type).outputs return self.get_op_info(type).outputs
def get_op_attr_names(self, type): def get_op_attr_names(self, type):
......
...@@ -33,5 +33,6 @@ py_test(test_sgd_op SRCS test_sgd_op.py) ...@@ -33,5 +33,6 @@ py_test(test_sgd_op SRCS test_sgd_op.py)
py_test(test_gradient_checker SRCS test_gradient_checker.py) py_test(test_gradient_checker SRCS test_gradient_checker.py)
py_test(test_lookup_table SRCS test_lookup_table.py) py_test(test_lookup_table SRCS test_lookup_table.py)
py_test(test_scale_and_identity_op SRCS test_scale_and_identity_op.py) py_test(test_scale_and_identity_op SRCS test_scale_and_identity_op.py)
py_test(test_sum_op SRCS test_sum_op.py)
py_test(mnist SRCS mnist.py) py_test(mnist SRCS mnist.py)
py_test(test_squared_l2_distance_op SRCS test_squared_l2_distance_op.py) py_test(test_squared_l2_distance_op SRCS test_squared_l2_distance_op.py)
import unittest
import numpy as np
import itertools
import paddle.v2.framework.core as core
from paddle.v2.framework.op import Operator
def grad_var_name(var_name):
return var_name + "@GRAD"
def create_op(scope, op_type, inputs, outputs, attrs=None):
kwargs = dict()
for in_name, in_dup in Operator.get_op_inputs(op_type):
if in_name in inputs:
kwargs[in_name] = []
if in_dup:
sub_in = inputs[in_name]
for sub_in_name in sub_in:
var = scope.new_var(sub_in_name)
kwargs[in_name].append(sub_in_name)
else:
var = scope.new_var(in_name)
kwargs[in_name].append(in_name)
for out_name, out_dup in Operator.get_op_outputs(op_type):
if out_name in outputs:
kwargs[out_name] = []
if out_dup:
sub_in = outputs[out_name]
for sun_in_name in sub_in:
var = scope.new_var(sun_in_name)
kwargs[out_name].append(sun_in_name)
else:
var = scope.new_var(out_name)
kwargs[out_name].append(out_name)
for attr_name in Operator.get_op_attr_names(op_type):
kwargs[attr_name] = attrs[attr_name]
return Operator(op_type, **kwargs)
def set_input(scope, op, inputs, place):
for in_name, in_dup in Operator.get_op_inputs(op.type()):
if in_name in inputs:
if in_dup:
sub_in = inputs[in_name]
for sub_in_name in sub_in:
var = scope.find_var(sub_in_name)
tensor = var.get_tensor()
arr = sub_in[sub_in_name]
tensor.set_dims(arr.shape)
tensor.set(arr, place)
else:
var = scope.find_var(in_name)
tensor = var.get_tensor()
arr = inputs[in_name]
tensor.set_dims(arr.shape)
tensor.set(arr, place)
def set_output_grad(scope, op, outputs, place):
for out_name, out_dup in Operator.get_op_outputs(op.type()):
if out_name in outputs:
if out_dup:
sub_out = outputs[out_name]
for sub_out_name in sub_out:
out_tensor = scope.find_var(sub_out_name).get_tensor()
grad_tensor = scope.new_var(grad_var_name(
sub_out_name)).get_tensor()
grad_tensor.set_dims(out_tensor.shape())
data = np.ones(out_tensor.shape(), dtype=np.float32)
grad_tensor.set(data, place)
else:
out_tensor = scope.find_var(out_name).get_tensor()
grad_tensor = scope.new_var(grad_var_name(out_name)).get_tensor(
)
grad_tensor.set_dims(out_tensor.shape())
data = np.ones(out_tensor.shape(), dtype=np.float32)
grad_tensor.set(data, place)
def get_numeric_gradient(scope,
op,
inputs,
input_to_check,
output_name,
delta=0.005,
in_place=False):
set_input(scope, op, inputs, core.CPUPlace())
op.infer_shape(scope)
tensor_to_check = scope.find_var(input_to_check).get_tensor()
def product(dim):
return reduce(lambda a, b: a * b, dim, 1)
ctx = core.DeviceContext.create(core.CPUPlace())
def get_output():
op.run(scope, ctx)
return np.array(scope.find_var(output_name).get_tensor()).sum()
tensor_to_check = scope.find_var(input_to_check).get_tensor()
tensor_size = product(tensor_to_check.get_dims())
gradient_flat = np.zeros(shape=(tensor_size, ), dtype='float32')
# we only compute gradient of one element each time.
# we use a for loop to compute the gradient of every element.
for i in xrange(tensor_size):
if in_place:
set_input(op, inputs, core.CPUPlace())
# get one input element throw it's index i.
origin = tensor_to_check.get_float_element(i)
# add delta to it, run op and then get the sum of the result tensor.
x_pos = origin + delta
tensor_to_check.set_float_element(i, x_pos)
y_pos = get_output()
if in_place:
set_input(op, inputs, core.CPUPlace())
x_neg = origin - delta
tensor_to_check.set_float_element(i, x_neg)
y_neg = get_output()
tensor_to_check.set_float_element(i, origin)
gradient_flat[i] = (y_pos - y_neg) / delta / 2
return gradient_flat.reshape(tensor_to_check.get_dims())
def get_backward_op(scope, op, no_grad_set):
backward_op = core.Operator.backward(op, no_grad_set)
for input in backward_op.input_vars():
var = scope.new_var(input)
var.get_tensor()
for output in backward_op.output_vars():
var = scope.new_var(output)
var.get_tensor()
return backward_op
def get_gradient(scope, op, inputs, outputs, grad_name, place,
no_grad_set=None):
ctx = core.DeviceContext.create(place)
set_input(scope, op, inputs, place)
op.infer_shape(scope)
op.run(scope, ctx)
if no_grad_set is None:
no_grad_set = set()
backward_op = get_backward_op(scope, op, no_grad_set)
set_output_grad(scope, op, outputs, place)
backward_op.infer_shape(scope)
backward_op.run(scope, ctx)
out = np.array(scope.find_var(grad_name).get_tensor())
return out
class OpTest(unittest.TestCase):
def check_output_with_place(self, place):
self.scope = core.Scope()
self.op = create_op(self.scope, self.op_type, self.inputs, self.outputs)
if isinstance(place, core.GPUPlace) and not self.op.support_gpu():
return
set_input(self.scope, self.op, self.inputs, place)
self.op.infer_shape(self.scope)
ctx = core.DeviceContext.create(place)
self.op.run(self.scope, ctx)
for out_name, out_dup in Operator.get_op_outputs(self.op.type()):
if out_dup:
sub_out = self.outputs[out_name]
for sub_out_name in sub_out:
actual = np.array(
self.scope.find_var(sub_out_name).get_tensor())
expect = sub_out[sub_out_name]
self.assertTrue(
np.allclose(
actual, expect, atol=1e-05),
"output name: " + out_name + "has diff")
else:
actual = np.array(self.scope.find_var(out_name).get_tensor())
expect = self.outputs[out_name]
self.assertTrue(
np.allclose(
actual, expect, atol=1e-05),
"output name: " + out_name + "has diff")
def check_output(self):
places = [core.CPUPlace()]
if core.is_compile_gpu():
places.append(core.GPUPlace(0))
for place in places:
self.check_output_with_place(place)
def __assert_is_close(self, numeric_grads, analytic_grads, names,
max_relative_error, msg_prefix):
for a, b, name in itertools.izip(numeric_grads, analytic_grads, names):
abs_a = np.abs(a)
abs_a[abs_a < 1e-3] = 1
diff_mat = np.abs(a - b) / abs_a
max_diff = np.max(diff_mat)
def err_msg():
offset = np.argmax(diff_mat > max_relative_error)
return "%s Variable %s max gradient diff %f over limit %f, the first " \
"error element is %d" % (
msg_prefix, name, max_diff, max_relative_error, offset)
self.assertLessEqual(max_diff, max_relative_error, err_msg())
def check_grad(self,
inputs_to_check,
output_name,
no_grad_set=None,
in_place=False,
max_relative_error=0.005):
self.scope = core.Scope()
self.op = create_op(self.scope, self.op_type, self.inputs, self.outputs)
if no_grad_set is None:
no_grad_set = set()
numeric_grads = [
get_numeric_gradient(
self.scope,
self.op,
self.inputs,
input_to_check,
output_name,
in_place=in_place) for input_to_check in inputs_to_check
]
grad_names = [
grad_var_name(input_to_check) for input_to_check in inputs_to_check
]
cpu_place = core.CPUPlace()
cpu_analytic_grads = [
get_gradient(self.scope, self.op, self.inputs, self.outputs,
grad_name, cpu_place, no_grad_set)
for grad_name in grad_names
]
self.__assert_is_close(numeric_grads, cpu_analytic_grads, grad_names,
max_relative_error,
"Gradient Check On %s" % str(cpu_place))
if core.is_compile_gpu() and self.op.support_gpu():
gpu_place = core.GPUPlace(0)
gpu_analytic_grads = [
get_gradient(self.scope, self.op, self.inputs, self.outputs,
grad_name, gpu_place, no_grad_set)
for grad_name in grad_names
]
self.__assert_is_close(numeric_grads, gpu_analytic_grads,
grad_names, max_relative_error,
"Gradient Check On %s" % str(gpu_place))
for c_grad, g_grad, name in itertools.izip(
cpu_analytic_grads, gpu_analytic_grads, grad_names):
self.assertTrue(
np.allclose(
c_grad, g_grad, atol=1e-4),
"output name: " + name + " has diff")
import unittest import unittest
import numpy import numpy
from op_test_util import OpTestMeta from op_test import OpTest
from gradient_checker import GradientChecker, create_op
class TestCrossEntropy(unittest.TestCase): class TestCrossEntropy(OpTest):
__metaclass__ = OpTestMeta
def setUp(self): def setUp(self):
self.type = "onehot_cross_entropy" self.op_type = "onehot_cross_entropy"
batch_size = 30 batch_size = 30
class_num = 10 class_num = 10
X = numpy.random.random((batch_size, class_num)).astype("float32") X = numpy.random.uniform(0.1, 1.0,
label = 5 * numpy.ones(batch_size).astype("int32") [batch_size, class_num]).astype("float32")
label = (class_num / 2) * numpy.ones(batch_size).astype("int32")
self.inputs = {'X': X, 'label': label} self.inputs = {'X': X, 'label': label}
Y = [] Y = []
for i in range(0, batch_size): for i in range(0, batch_size):
Y.append(-numpy.log(X[i][label[i]])) Y.append(-numpy.log(X[i][label[i]]))
self.outputs = {'Y': numpy.array(Y).astype("float32")} self.outputs = {'Y': numpy.array(Y).astype("float32")}
def test_check_output(self):
self.check_output()
class CrossEntropyGradOpTest(GradientChecker):
def test_check_grad(self): def test_check_grad(self):
op = create_op("onehot_cross_entropy") self.check_grad(["X"], "Y")
batch_size = 30
class_num = 10
inputs = {
"X": numpy.random.uniform(
0.1, 1.0, [batch_size, class_num]).astype("float32"),
"label": (class_num / 2) * numpy.ones(batch_size).astype("int32")
}
self.check_grad(op, inputs, set("X"), "Y")
if __name__ == "__main__": if __name__ == "__main__":
......
...@@ -4,7 +4,7 @@ from op_test_util import OpTestMeta ...@@ -4,7 +4,7 @@ from op_test_util import OpTestMeta
from gradient_checker import GradientChecker, create_op from gradient_checker import GradientChecker, create_op
class TestSigmoidOp(unittest.TestCase): class TestLookupTableOp(unittest.TestCase):
__metaclass__ = OpTestMeta __metaclass__ = OpTestMeta
def setUp(self): def setUp(self):
...@@ -15,7 +15,7 @@ class TestSigmoidOp(unittest.TestCase): ...@@ -15,7 +15,7 @@ class TestSigmoidOp(unittest.TestCase):
self.outputs = {'Out': table[ids]} self.outputs = {'Out': table[ids]}
class TestSigmoidGradOp(GradientChecker): class TestLookupTableGradOp(GradientChecker):
def test_grad(self): def test_grad(self):
op = create_op('lookup_table') op = create_op('lookup_table')
table = np.random.random((17, 31)).astype('float32') table = np.random.random((17, 31)).astype('float32')
......
import unittest import unittest
import numpy as np import numpy as np
from op_test_util import OpTestMeta from op_test import OpTest
from gradient_checker import GradientChecker, create_op
class TestSigmoidOp(unittest.TestCase): class TestSigmoid(OpTest):
__metaclass__ = OpTestMeta
def setUp(self): def setUp(self):
self.type = "sigmoid" self.op_type = "sigmoid"
self.inputs = {'X': np.random.random((15, 31)).astype("float32")} self.inputs = {
'X': np.random.uniform(0.1, 1, [11, 17]).astype("float32")
}
self.outputs = {'Y': 1 / (1 + np.exp(-self.inputs['X']))} self.outputs = {'Y': 1 / (1 + np.exp(-self.inputs['X']))}
def test_check_output(self):
self.check_output()
class TestSigmoidGradOp(GradientChecker): def test_check_grad(self):
def test_grad(self): self.check_grad(["X"], "Y", max_relative_error=0.007)
op = create_op("sigmoid")
inputs = {"X": np.random.uniform(0.1, 1, [11, 17]).astype("float32")}
# compare gpu and cpu results for backward op.
# this test will be skiped if only compiling CPU version.
self.compare_grad(op, inputs)
# check gradients
self.check_grad(op, inputs, set("X"), "Y", max_relative_error=0.007)
if __name__ == '__main__': if __name__ == '__main__':
......
import unittest
import numpy as np
from op_test import OpTest
class TestSumOp(OpTest):
def setUp(self):
self.op_type = "sum"
x0 = np.random.random((3, 4)).astype('float32')
x1 = np.random.random((3, 4)).astype('float32')
x2 = np.random.random((3, 4)).astype('float32')
self.inputs = {"X": {"x0": x0, "x1": x1, "x2": x2}}
y = x0 + x1 + x2
self.outputs = {'Out': y}
def test_check_output(self):
self.check_output()
def test_check_grad(self):
self.check_grad(["x0"], "Out")
if __name__ == '__main__':
unittest.main()
Markdown is supported
0% .
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