Initial Commit - GPU LinSpace

comment fix

docString fix

added asserts in test file atop np checks

lint

lint-2

lint3
pull/8928/head
danishnxt 4 years ago
parent adc8e3e707
commit a17f76dd1d

@ -0,0 +1,32 @@
/**
* Copyright 2020 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/gpu/cuda_impl/linspace.cuh"
#include <iostream>
template <typename T>
__global__ void LinSpaceKernel(const T *start, const T *stop, const size_t value_count, T *output) {
T add_value = ((*stop - *start) / (value_count - 1));
for (size_t i = blockIdx.x * blockDim.x + threadIdx.x; i < value_count; i += gridDim.x * blockDim.x) {
output[i] = *start + (add_value * i);
}
}
template <typename T>
void calLinSpace(const T *start, const T *stop, const size_t value_count, T *output, cudaStream_t cuda_stream) {
LinSpaceKernel<<<GET_BLOCKS(value_count), GET_THREADS, 0, cuda_stream>>>(start, stop, value_count, output);
}
template void calLinSpace<float>(const float *start, const float *stop, const size_t value_count, float *output,
cudaStream_t cuda_stream);

@ -0,0 +1,23 @@
/**
* Copyright 2020 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_KERNEL_GPU_CUDA_LINSPACE_IMPL_CU_H_
#define MINDSPORE_CCSRC_KERNEL_GPU_CUDA_LINSPACE_IMPL_CU_H_
#include "runtime/device/gpu/cuda_common.h"
template <typename T>
void calLinSpace(const T *start, const T *stop, const size_t value_count, T *output, cudaStream_t cuda_stream);
#endif // MINDSPORE_CCSRC_KERNEL_GPU_CUDA_LINSPACE_IMPL_CU_H_

@ -0,0 +1,29 @@
/**
* Copyright 2020 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/gpu/math/linspace.h"
namespace mindspore {
namespace kernel {
MS_REG_GPU_KERNEL_ONE(LinSpace,
KernelAttr()
.AddInputAttr(kNumberTypeFloat32)
.AddInputAttr(kNumberTypeFloat32)
.AddInputAttr(kNumberTypeInt64)
.AddOutputAttr(kNumberTypeFloat32),
LinSpaceGpuKernel, float)
} // namespace kernel
} // namespace mindspore

@ -0,0 +1,102 @@
/**
* Copyright 2020 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_GPU_MATH_LINSPACE_GPU_KERNEL_H_
#define MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_GPU_MATH_LINSPACE_GPU_KERNEL_H_
#include <vector>
#include <memory>
#include <iostream>
#include "backend/kernel_compiler/gpu/gpu_kernel.h"
#include "backend/kernel_compiler/gpu/gpu_kernel_factory.h"
#include "backend/kernel_compiler/gpu/cuda_impl/linspace.cuh"
#include "backend/kernel_compiler/gpu/kernel_constants.h"
namespace mindspore {
namespace kernel {
template <typename T>
class LinSpaceGpuKernel : public GpuKernel {
public:
LinSpaceGpuKernel() { ResetResource(); }
~LinSpaceGpuKernel() = default;
const std::vector<size_t> &GetInputSizeList() const override { return input_size_list_; }
const std::vector<size_t> &GetOutputSizeList() const override { return output_size_list_; }
const std::vector<size_t> &GetWorkspaceSizeList() const override { return workspace_size_list_; }
bool Launch(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs, void *stream_ptr) override {
VARIABLE_NOT_USED(workspace);
T *start_addr = GetDeviceAddress<T>(inputs, 0);
T *stop_addr = GetDeviceAddress<T>(inputs, 1);
T *output_addr = GetDeviceAddress<T>(outputs, 0);
calLinSpace(start_addr, stop_addr, value_count_, output_addr, reinterpret_cast<cudaStream_t>(stream_ptr));
return true;
}
bool Init(const CNodePtr &kernel_node) override {
size_t input_num = AnfAlgo::GetInputTensorNum(kernel_node);
if (input_num != 3) {
MS_LOG(ERROR) << "Input number is " << input_num << ", but DynamicLinSpace needs 3 inputs.";
return false;
}
size_t output_num = AnfAlgo::GetOutputTensorNum(kernel_node);
if (output_num != 1) {
MS_LOG(ERROR) << "Output number is " << output_num << ", but DynamicLinSpace needs 1 output.";
return false;
}
auto input_1 = AnfAlgo::GetInputRealDeviceShapeIfExist(kernel_node, 0);
auto input_2 = AnfAlgo::GetInputRealDeviceShapeIfExist(kernel_node, 1);
// error checking input data
if ((input_1.size() != 0) || (input_2.size() != 0)) {
MS_LOG(ERROR) << "For LinShape "
<< "both start and end must be 0-D Tensors. Got " << input_1.size() << " and " << input_2.size()
<< ".";
return false;
}
auto value_count = AnfAlgo::GetOutputRealDeviceShapeIfExist(kernel_node, 0);
if (value_count.size() != 1) {
MS_LOG(ERROR) << "For LinShape, output shape incorrect rank. Expect Rank: 1, got Rank: " << value_count.size()
<< ".";
}
value_count_ = value_count[0];
InitSizeLists();
return true;
}
void ResetResource() noexcept override {
value_count_ = 0;
input_size_list_.clear();
output_size_list_.clear();
workspace_size_list_.clear();
}
protected:
void InitSizeLists() override {
input_size_list_.push_back(sizeof(T)); // Scalar tensor
input_size_list_.push_back(sizeof(T)); // Scalar tensor
output_size_list_.push_back(value_count_ * sizeof(T));
}
private:
size_t value_count_;
int num_input_;
std::vector<size_t> input_size_list_;
std::vector<size_t> output_size_list_;
std::vector<size_t> workspace_size_list_;
};
} // namespace kernel
} // namespace mindspore
#endif // MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_GPU_MATH_LINSPACE_GPU_KERNEL_H_

@ -247,6 +247,8 @@ AbstractBasePtr InferImplMinimum(const AnalysisEnginePtr &, const PrimitivePtr &
const AbstractBasePtrList &args_spec_list);
AbstractBasePtr InferImplDivNoNan(const AnalysisEnginePtr &, const PrimitivePtr &primitive,
const AbstractBasePtrList &args_spec_list);
AbstractBasePtr InferImplLinSpace(const AnalysisEnginePtr &, const PrimitivePtr &primitive,
const AbstractBasePtrList &args_spec_list);
AbstractBasePtr InferImplExpandDims(const AnalysisEnginePtr &, const PrimitivePtr &primitive,
const AbstractBasePtrList &args_spec_list);
AbstractBasePtr InferImplGpuConvertToDynamicShape(const AnalysisEnginePtr &, const PrimitivePtr &primitive,

@ -167,5 +167,47 @@ AbstractBasePtr InferImplDivNoNan(const AnalysisEnginePtr &engine_ptr, const Pri
const AbstractBasePtrList &args_spec_list) {
return InferImplBinaryBase(engine_ptr, primitive, args_spec_list);
}
AbstractBasePtr InferImplLinSpace(const AnalysisEnginePtr &, const PrimitivePtr &primitive,
const AbstractBasePtrList &args_spec_list) {
const std::string op_name = primitive->name();
CheckArgsSize(op_name, args_spec_list, 3);
auto start = CheckArg<AbstractTensor>(op_name, args_spec_list, 0);
MS_EXCEPTION_IF_NULL(start);
MS_EXCEPTION_IF_NULL(start->shape());
auto stop = CheckArg<AbstractTensor>(op_name, args_spec_list, 1);
MS_EXCEPTION_IF_NULL(stop);
MS_EXCEPTION_IF_NULL(stop->shape());
(void)CheckTensorDType(start, {kFloat32}, "Input 0 (start) for LinSpace should be %s");
(void)CheckTensorDType(stop, {kFloat32}, "Input 1 (stop) for LinSpace should be %s");
ShapeVector shape;
ShapeVector max_shape;
ShapeVector min_shape;
int64_t num_val = 0;
// 3rd input is a Tensor when LinSpace is a dynamic shape operator
if (args_spec_list[2]->isa<AbstractTensor>()) {
auto num = args_spec_list[2]->cast<AbstractTensorPtr>();
MS_EXCEPTION_IF_NULL(num);
auto num_value_ptr = num->BuildValue();
MS_EXCEPTION_IF_NULL(num_value_ptr);
auto num_tensor = num_value_ptr->cast<tensor::TensorPtr>();
MS_EXCEPTION_IF_NULL(num_tensor);
num_val = *static_cast<int64_t *>(num_tensor->data_c());
} else if (args_spec_list[2]->isa<AbstractScalar>()) {
auto num = args_spec_list[2]->cast<AbstractScalarPtr>();
num_val = GetValue<int64_t>(num->BuildValue());
} else {
MS_LOG(EXCEPTION) << "Invalid abstract type:" << args_spec_list[2]->type_name();
}
shape.emplace_back(num_val);
if (shape[0] < 0) {
MS_LOG(EXCEPTION) << "num must be >= 0 in LinSpace";
}
max_shape.emplace_back(num_val);
min_shape.emplace_back(num_val);
AbstractTensorPtr ret =
std::make_shared<AbstractTensor>(start->element(), std::make_shared<Shape>(shape, min_shape, max_shape));
return ret;
}
} // namespace abstract
} // namespace mindspore

@ -45,6 +45,7 @@ PrimitiveEvalImplMap &GetPrimitiveToEvalImplMap() {
{prim::kPrimEqual, {InferImplEqual, true}},
{prim::kPrimMinimum, {InferImplMinimum, true}},
{prim::kPrimDivNoNan, {InferImplDivNoNan, true}},
{prim::kPrimLinSpace, {InferImplLinSpace, true}},
// Array
{prim::kPrimScalarToArray, {InferImplScalarToArray, true}},
{prim::kPrimArrayToScalar, {InferImplArrayToScalar, true}},

@ -241,6 +241,7 @@ inline const PrimitivePtr kPrimExp = std::make_shared<Primitive>("Exp");
inline const PrimitivePtr kPrimLog = std::make_shared<Primitive>("Log");
inline const PrimitivePtr kPrimRsqrt = std::make_shared<Primitive>("Rsqrt");
inline const PrimitivePtr kPrimSplitV = std::make_shared<Primitive>("SplitV");
inline const PrimitivePtr kPrimLinSpace = std::make_shared<Primitive>("LinSpace");
// Statements
inline const PrimitivePtr kPrimReturn = std::make_shared<Primitive>("return");

@ -3946,15 +3946,19 @@ class Eps(PrimitiveWithInfer):
class LinSpace(PrimitiveWithInfer):
r"""
Generates values in an interval and returns the corresponding interpolation accroding to assist.
Generates values in an interval (inclusive of start and stop) and returns the corresponding
interpolated array with **num** number of ticks.
Inputs:
- **start** (Tensor[float32]) - The start of interval, With shape of 0-D.
- **stop** (Tensor[float32]) - The end of interval, With shape of 0-D.
- **num** (int) - Ticks number in the interval, the ticks include start and stop value.
- **start** (Tensor[float32]) - Start value of interval, With shape of 0-D.
- **stop** (Tensor[float32]) - Last value of interval, With shape of 0-D.
- **num** (int) - Number of ticks in the interval, inclusive of start and stop.
Outputs:
Tensor, has the same shape as `assist`.
Tensor, has the same shape as `start`.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> linspace = P.LinSpace()

@ -0,0 +1,99 @@
# Copyright 2020 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.common.dtype as mstype
import mindspore.context as context
from mindspore.common.tensor import Tensor
from mindspore.nn import Cell
from mindspore.ops import operations as P
class LinSpaceNet(Cell):
def __init__(self, num):
super(LinSpaceNet, self).__init__()
self.ls_op = P.LinSpace()
self.num = num
def construct(self, start, stop):
output = self.ls_op(start, stop, self.num)
return output
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_lin_space_1():
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
start_np = 5
stop_np = 150
num_np = 12
start = Tensor(start_np, dtype=mstype.float32)
stop = Tensor(stop_np, dtype=mstype.float32)
num = num_np
ls_op = P.LinSpace()
result_ms = ls_op(start, stop, num).asnumpy()
result_np = np.linspace(start_np, stop_np, num_np)
assert np.allclose(result_ms, result_np)
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_lin_shape_2():
context.set_context(mode=context.PYNATIVE_MODE, device_target='GPU')
start_np = -25
stop_np = 147
num_np = 10
start = Tensor(start_np, dtype=mstype.float32)
stop = Tensor(stop_np, dtype=mstype.float32)
num = num_np
ls_op = P.LinSpace()
result_ms = ls_op(start, stop, num).asnumpy()
result_np = np.linspace(start_np, stop_np, num_np)
assert np.allclose(result_ms, result_np)
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_lin_shape_3():
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
start_np = 25
stop_np = -147
num_np = 20
start = Tensor(start_np, dtype=mstype.float32)
stop = Tensor(stop_np, dtype=mstype.float32)
net = LinSpaceNet(num_np)
result_ms = net(start, stop).asnumpy()
result_np = np.linspace(start_np, stop_np, num_np)
assert np.allclose(result_ms, result_np)
@pytest.mark.level0
@pytest.mark.platform_x86_gpu_training
@pytest.mark.env_onecard
def test_lin_shape_4():
context.set_context(mode=context.GRAPH_MODE, device_target='GPU')
start_np = -25.3
stop_np = -147
num_np = 36
start = Tensor(start_np, dtype=mstype.float32)
stop = Tensor(stop_np, dtype=mstype.float32)
net = LinSpaceNet(num_np)
result_ms = net(start, stop).asnumpy()
result_np = np.linspace(start_np, stop_np, num_np)
assert np.allclose(result_ms, result_np)
Loading…
Cancel
Save