m53297601
/
mindspore
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

!11636 add bce cpu ops

Browse Source 
From: @zhangzhewei01
Reviewed-by: 
Signed-off-by:
pull/11636/MERGE
mindspore-ci-bot 4 years ago committed by Gitee
parent a84a5215ca 22ca18fe1d
commit e3796b3639
6 changed files with 443 additions and 1 deletions
Show Stats Download Patch File Download Diff File

100
mindspore/ccsrc/backend/kernel_compiler/cpu/binary_cross_entropy_cpu_kernel.cc
Unescape View File

@ -0,0 +1,100 @@
/**
* Copyright 2021 Huawei Technologies Co., Ltd
*
* 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 "backend/kernel_compiler/cpu/binary_cross_entropy_cpu_kernel.h"
namespace mindspore {
namespace kernel {
template <typename T>
void BinaryCrossEntropyCpuKernel::LaunchToScalar(const int &input_size, const int &reduction, T *loss, T *tmp_loss) {
if (input_size % 2 == 1) {
tmp_loss[0] += tmp_loss[input_size - 1];
}
for (int stride = input_size / 2; stride > 0; stride >>= 1) {
for (int i = 0; i < stride; i++) {
tmp_loss[i] += tmp_loss[i + stride];
}
if (stride > 2 && stride % 2 == 1) {
tmp_loss[0] += tmp_loss[stride - 1];
}
}
loss[0] += tmp_loss[0];
if (reduction == 1) {
loss[0] /= static_cast<T>(input_size);
}
}
template <typename T>
void BinaryCrossEntropyCpuKernel::Launchkernel(const std::vector<AddressPtr> &inputs,
const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs) {
T *input_x = reinterpret_cast<T *>(inputs[0]->addr);
T *input_y = reinterpret_cast<T *>(inputs[1]->addr);
T *weight = reinterpret_cast<T *>(inputs[2]->addr);
T *loss = reinterpret_cast<T *>(outputs[0]->addr);
std::vector<T> tmp_loss(input_size_);
T epsilon = static_cast<T>(1e-12);
T one = static_cast<T>(1);
if (reduction_ == 0) {
for (size_t i = 0; i < input_size_; i++) {
T value =
-weight[i] * (input_y[i] * log(input_x[i] + epsilon) + (one - input_y[i]) * log(one - input_x[i] + epsilon));
loss[i] = value;
}
} else {
for (size_t i = 0; i < input_size_; i++) {
T value =
-weight[i] * (input_y[i] * log(input_x[i] + epsilon) + (one - input_y[i]) * log(one - input_x[i] + epsilon));
tmp_loss[i] = value;
}
}
if (reduction_ != 0) {
LaunchToScalar<T>(input_size_, reduction_, loss, tmp_loss.data());
}
}
bool BinaryCrossEntropyCpuKernel::Launch(const std::vector<AddressPtr> &inputs,
const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs) {
if (input_size_ > 0) {
if (dtype_ == kNumberTypeFloat32) {
Launchkernel<float>(inputs, workspace, outputs);
} else if (dtype_ == kNumberTypeFloat16) {
Launchkernel<float16>(inputs, workspace, outputs);
}
}
return true;
}
void BinaryCrossEntropyCpuKernel::InitKernel(const CNodePtr &kernel_node) {
auto input_shape = AnfAlgo::GetPrevNodeOutputInferShape(kernel_node, 0);
for (size_t i = 0; i < input_shape.size(); i++) {
input_size_ *= input_shape[i];
}
string reduction = AnfAlgo::GetNodeAttr<string>(kernel_node, "reduction");
if (reduction == "none") {
reduction_ = 0;
} else if (reduction == "sum") {
reduction_ = 2;
}
dtype_ = AnfAlgo::GetInputDeviceDataType(kernel_node, 0);
}
} // namespace kernel
} // namespace mindspore

62
mindspore/ccsrc/backend/kernel_compiler/cpu/binary_cross_entropy_cpu_kernel.h
Unescape View File

@ -0,0 +1,62 @@
/**
* Copyright 2021 Huawei Technologies Co., Ltd
*
* 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.
*/
#ifndef MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_NN_BINARY_CROSS_ENTROPY_KERNEL_H
#define MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_NN_BINARY_CROSS_ENTROPY_KERNEL_H
#include <vector>
#include <string>
#include "backend/kernel_compiler/cpu/cpu_kernel.h"
#include "backend/kernel_compiler/cpu/cpu_kernel_factory.h"
namespace mindspore {
namespace kernel {
class BinaryCrossEntropyCpuKernel : public CPUKernel {
public:
BinaryCrossEntropyCpuKernel() : input_size_(1), reduction_(1) {}
~BinaryCrossEntropyCpuKernel() override = default;
void InitKernel(const CNodePtr &kernel_node) override;
bool Launch(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs) override;
private:
template <typename T>
void LaunchToScalar(const int &input_size, const int &reduction, T *loss, T *tmp_loss);
template <typename T>
void Launchkernel(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs);
TypeId dtype_{kTypeUnknown};
size_t input_size_;
int reduction_;
};
MS_REG_CPU_KERNEL(BinaryCrossEntropy,
KernelAttr()
.AddInputAttr(kNumberTypeFloat16)
.AddInputAttr(kNumberTypeFloat16)
.AddInputAttr(kNumberTypeFloat16)
.AddOutputAttr(kNumberTypeFloat16),
BinaryCrossEntropyCpuKernel);
MS_REG_CPU_KERNEL(BinaryCrossEntropy,
KernelAttr()
.AddInputAttr(kNumberTypeFloat32)
.AddInputAttr(kNumberTypeFloat32)
.AddInputAttr(kNumberTypeFloat32)
.AddOutputAttr(kNumberTypeFloat32),
BinaryCrossEntropyCpuKernel);
} // namespace kernel
} // namespace mindspore
#endif // MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_NN_BINARY_CROSS_ENTROPY_KERNEL_H

78
mindspore/ccsrc/backend/kernel_compiler/cpu/binary_cross_entropy_grad_kernel.cc
Unescape View File

@ -0,0 +1,78 @@
/**
* Copyright 2021 Huawei Technologies Co., Ltd
*
* 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 "backend/kernel_compiler/cpu/binary_cross_entropy_grad_kernel.h"
namespace mindspore {
namespace kernel {
template <typename T>
void BinaryCrossEntropyGradCpuKernel::Launchkernel(const std::vector<AddressPtr> &inputs,
const std::vector<AddressPtr> &outputs) {
T *input_x = reinterpret_cast<T *>(inputs[0]->addr);
T *input_y = reinterpret_cast<T *>(inputs[1]->addr);
T *dloss = reinterpret_cast<T *>(inputs[2]->addr);
T *weight = reinterpret_cast<T *>(inputs[3]->addr);
T *dx = reinterpret_cast<T *>(outputs[0]->addr);
T epsilon = static_cast<T>(1e-12);
T one = static_cast<T>(1);
if (reduction_ == 0) {
for (size_t i = 0; i < input_size_; i++) {
T denominator = ((input_x[i] * (one - input_x[i])) > epsilon) ? (input_x[i] * (one - input_x[i])) : epsilon;
T value = weight[i] * (input_x[i] - input_y[i]) / denominator;
dx[i] = value * dloss[i];
}
} else {
T dloss1 = dloss[0];
if (reduction_ == 1) {
dloss1 = dloss[0] / static_cast<T>(input_size_);
}
for (size_t i = 0; i < input_size_; i++) {
T denominator = ((input_x[i] * (one - input_x[i])) > epsilon) ? (input_x[i] * (one - input_x[i])) : epsilon;
T value = weight[i] * (input_x[i] - input_y[i]) / denominator;
dx[i] = value * dloss1;
}
}
}
bool BinaryCrossEntropyGradCpuKernel::Launch(const std::vector<AddressPtr> &inputs,
const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs) {
if (input_size_ > 0) {
if (dtype_ == kNumberTypeFloat32) {
Launchkernel<float>(inputs, outputs);
} else if (dtype_ == kNumberTypeFloat16) {
Launchkernel<float16>(inputs, outputs);
}
}
return true;
}
void BinaryCrossEntropyGradCpuKernel::InitKernel(const CNodePtr &kernel_node) {
auto input_shape = AnfAlgo::GetPrevNodeOutputInferShape(kernel_node, 0);
for (size_t i = 0; i < input_shape.size(); i++) {
input_size_ *= input_shape[i];
}
string reduction = AnfAlgo::GetNodeAttr<string>(kernel_node, "reduction");
if (reduction == "none") {
reduction_ = 0;
} else if (reduction == "sum") {
reduction_ = 2;
}
dtype_ = AnfAlgo::GetInputDeviceDataType(kernel_node, 0);
}
} // namespace kernel
} // namespace mindspore

61
mindspore/ccsrc/backend/kernel_compiler/cpu/binary_cross_entropy_grad_kernel.h
Unescape View File

@ -0,0 +1,61 @@
/**
* Copyright 2021 Huawei Technologies Co., Ltd
*
* 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.
*/
#ifndef MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_NN_BINARY_CROSS_ENTROPY_GRAD_KERNEL_H
#define MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_NN_BINARY_CROSS_ENTROPY_GRAD_KERNEL_H
#include <vector>
#include <string>
#include "backend/kernel_compiler/cpu/cpu_kernel.h"
#include "backend/kernel_compiler/cpu/cpu_kernel_factory.h"
namespace mindspore {
namespace kernel {
class BinaryCrossEntropyGradCpuKernel : public CPUKernel {
public:
BinaryCrossEntropyGradCpuKernel() : input_size_(1), reduction_(1) {}
~BinaryCrossEntropyGradCpuKernel() override = default;
void InitKernel(const CNodePtr &kernel_node) override;
bool Launch(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs) override;
private:
template <typename T>
void Launchkernel(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &outputs);
TypeId dtype_{kTypeUnknown};
size_t input_size_;
int reduction_;
};
MS_REG_CPU_KERNEL(BinaryCrossEntropyGrad,
KernelAttr()
.AddInputAttr(kNumberTypeFloat16)
.AddInputAttr(kNumberTypeFloat16)
.AddInputAttr(kNumberTypeFloat16)
.AddInputAttr(kNumberTypeFloat16)
.AddOutputAttr(kNumberTypeFloat16),
BinaryCrossEntropyGradCpuKernel);
MS_REG_CPU_KERNEL(BinaryCrossEntropyGrad,
KernelAttr()
.AddInputAttr(kNumberTypeFloat32)
.AddInputAttr(kNumberTypeFloat32)
.AddInputAttr(kNumberTypeFloat32)
.AddInputAttr(kNumberTypeFloat32)
.AddOutputAttr(kNumberTypeFloat32),
BinaryCrossEntropyGradCpuKernel);
} // namespace kernel
} // namespace mindspore
#endif // MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_NN_BINARY_CROSS_ENTROPY_GRAD_KERNEL_H

2
mindspore/ops/operations/nn_ops.py
Unescape View File

@ -4619,7 +4619,7 @@ class BinaryCrossEntropy(PrimitiveWithInfer):
Otherwise, the output is a scalar.
Supported Platforms:
``Ascend`` ``GPU``
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> import mindspore

141
tests/st/ops/cpu/test_binary_cross_entropy_op.py
Unescape View File

@ -0,0 +1,141 @@
# Copyright 2021 Huawei Technologies Co., Ltd
#
# 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.
# ============================================================================
import numpy as np
import pytest
import mindspore.context as context
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.ops import composite as C
from mindspore.ops import operations as P
context.set_context(mode=context.GRAPH_MODE, device_target="CPU")
class Net(nn.Cell):
def __init__(self, reduction="none"):
super(Net, self).__init__()
self.BinaryCrossEntropy = P.BinaryCrossEntropy(reduction)
def construct(self, x, y, weight):
return self.BinaryCrossEntropy(x, y, weight)
@pytest.mark.level0
@pytest.mark.platform_x86_cpu
@pytest.mark.env_onecard
def test_binary_cross_entropy_loss():
np.random.seed(42)
prediction = np.random.rand(20).astype(np.float32)
target = np.random.rand(20).astype(np.float32)
weight = np.random.rand(20).astype(np.float32)
reduction = "none"
net = Net(reduction)
loss = net(Tensor(prediction), Tensor(target), Tensor(weight))
expect = [0.09555826, 1.2861121, 0.03518666, 0.6969416, 0.24313456, 0.99062896,
0.19205657, 0.5465214, 0.36964455, 0.21999404, 2.2953863, 2.2566645,
1.5803775, 1.3266402, 0.9883408, 1.2997618, 0.05439841, 0.14389999,
0.03405444, 0.23934692]
assert np.allclose(loss.asnumpy(), expect)
def test_binary_cross_entropy_loss_mean():
np.random.seed(42)
prediction = np.random.rand(20).astype(np.float32)
target = np.random.rand(20).astype(np.float32)
weight = np.random.rand(20).astype(np.float32)
reduction = "mean"
net = Net(reduction)
loss = net(Tensor(prediction), Tensor(target), Tensor(weight))
expect = [0.7447324991226196]
assert loss.asnumpy() == expect
def test_binary_cross_entropy_loss_sum():
np.random.seed(42)
prediction = np.random.rand(20).astype(np.float32)
target = np.random.rand(20).astype(np.float32)
weight = np.random.rand(20).astype(np.float32)
reduction = "sum"
net = Net(reduction)
loss = net(Tensor(prediction), Tensor(target), Tensor(weight))
expect = [14.894649505615234]
assert loss.asnumpy() == expect
def test_binary_cross_entropy_loss_16():
np.random.seed(42)
prediction = np.random.rand(20).astype(np.float16)
target = np.random.rand(20).astype(np.float16)
weight = np.random.rand(20).astype(np.float16)
reduction = "none"
net = Net(reduction)
loss = net(Tensor(prediction), Tensor(target), Tensor(weight))
expect = [0.09552, 1.28613, 0.0351868, 0.696777, 0.243164, 0.990234,
0.192139, 0.546875, 0.370117, 0.219971, 2.29492, 2.25391,
1.58105, 1.32812, 0.987305, 1.30078, 0.0544434, 0.143921,
0.0340576, 0.239258]
assert np.allclose(loss.asnumpy(), expect)
def test_binary_cross_entropy_loss_mean_16():
np.random.seed(42)
prediction = np.random.rand(20).astype(np.float16)
target = np.random.rand(20).astype(np.float16)
weight = np.random.rand(20).astype(np.float16)
reduction = "mean"
net = Net(reduction)
loss = net(Tensor(prediction), Tensor(target), Tensor(weight))
expect = [0.74462890625]
assert loss.asnumpy() == expect
def test_binary_cross_entropy_loss_sum_16():
np.random.seed(42)
prediction = np.random.rand(20).astype(np.float16)
target = np.random.rand(20).astype(np.float16)
weight = np.random.rand(20).astype(np.float16)
reduction = "sum"
net = Net(reduction)
loss = net(Tensor(prediction), Tensor(target), Tensor(weight))
expect = [14.890625]
assert loss.asnumpy() == expect
class Grad(nn.Cell):
def __init__(self, network):
super(Grad, self).__init__()
self.grad = C.GradOperation(get_all=True, sens_param=True)
self.network = network
def construct(self, x1, x2, sens, weight):
gout = self.grad(self.network)(x1, x2, sens, weight)
return gout
@pytest.mark.level0
@pytest.mark.platform_x86_cpu
@pytest.mark.env_onecard
def test_binary_cross_entropy_loss_grad():
np.random.seed(42)
prediction = np.random.rand(20).astype(np.float32)
target = np.random.rand(20).astype(np.float32)
sens = np.random.rand(20).astype(np.float32)
weight = np.random.rand(20).astype(np.float32)
reduction = "none"
grad = Grad(Net(reduction))
dx = grad(Tensor(prediction), Tensor(target), Tensor(sens), Tensor(weight))
dx1_expect = [-4.80516590e-02, 2.32625079e+00, 6.38972521e-02, 3.13642323e-01,
-1.65661633e-01, -1.71821892e+00, -1.13685496e-01, 1.26669514e+00,
1.47891801e-03, 5.83921909e-01, -2.17992840e+01, 4.21899414e+00,
2.85430793e-02, -3.21346498e+00, -2.22674108e+00, -2.80453944e+00,
-1.19787852e-04, 2.48514321e-02, -1.66696273e-02, -2.71965731e-02]
assert np.allclose(dx[0].asnumpy(), dx1_expect)
Write Preview
Loading…
Cancel
Save