[Function] Add slice function (#719)

* fix math functions

* add slice function
This commit is contained in:
Jack Zhou
2022-11-28 15:33:33 +08:00
committed by GitHub
parent dd18471b41
commit d0307192f9
5 changed files with 282 additions and 3 deletions

View File

@@ -17,7 +17,6 @@
#include "fastdeploy/function/cast.h"
#include "fastdeploy/function/clip.h"
#include "fastdeploy/function/concat.h"
#include "fastdeploy/function/cuda_cast.h"
#include "fastdeploy/function/cumprod.h"
#include "fastdeploy/function/elementwise.h"
#include "fastdeploy/function/full.h"
@@ -28,6 +27,7 @@
#include "fastdeploy/function/pad.h"
#include "fastdeploy/function/quantile.h"
#include "fastdeploy/function/reduce.h"
#include "fastdeploy/function/slice.h"
#include "fastdeploy/function/softmax.h"
#include "fastdeploy/function/sort.h"
#include "fastdeploy/function/split.h"

View File

@@ -28,11 +28,13 @@ namespace function {
template <typename T, typename Functor>
void ActivationImpl(const FDTensor& X, FDTensor* Out, const Functor& functor) {
FDASSERT(Out != nullptr, "Output Out should not be nullptr");
FDTensor out_tmp;
auto x = EigenVector<T>::Flatten(X);
Out->Allocate(X.Shape(), X.Dtype());
auto out = EigenVector<T>::Flatten(*Out);
out_tmp.Allocate(X.Shape(), X.Dtype());
auto out = EigenVector<T>::Flatten(out_tmp);
const auto& dev = *EigenDeviceWrapper::GetInstance()->GetDevice();
functor(dev, x, out);
*Out = std::move(out_tmp);
}
DEFINE_ACTIVATION_KERNEL(Sqrt, SqrtFunctor)

View File

@@ -0,0 +1,167 @@
// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include "fastdeploy/function/slice.h"
#include "fastdeploy/function/eigen.h"
#include <algorithm>
namespace fastdeploy {
namespace function {
std::vector<int64_t> GetSliceDims(const std::vector<int64_t>& in_dims,
const std::vector<int64_t>& axes,
const std::vector<int64_t>& starts,
const std::vector<int64_t>& ends,
std::vector<int64_t>* steps = nullptr) {
std::vector<int64_t> slice_dims(in_dims);
for (size_t i = 0; i < axes.size(); ++i) {
int64_t axis = axes[i];
if (in_dims[axis] == -1) {
continue;
}
int64_t start = starts[i];
int64_t end = ends[i];
int64_t step = steps == nullptr ? 1 : (*steps)[i];
if (step > 0) {
slice_dims[axis] = (end - start + step - 1) / step;
} else {
slice_dims[axis] = (end - start + step + 1) / step;
}
}
return slice_dims;
}
void CheckAndUpdateSliceAttrs(const std::vector<int64_t>& in_dims,
const std::vector<int64_t>& axes,
std::vector<int64_t>* starts,
std::vector<int64_t>* ends,
std::vector<int64_t>* steps = nullptr) {
for (size_t i = 0; i < axes.size(); ++i) {
int64_t axis = axes[i];
FDASSERT(axis < in_dims.size(),
"The axis value should be less than the rank of input, "
"but received axes[%d] = %d, rank of input is %d.",
i, axis, in_dims.size());
int64_t dim_value = in_dims[axis];
if (dim_value > 0) {
int64_t step = steps == nullptr ? 1 : (*steps)[i];
FDASSERT(step != 0, "Step should not be 0, but received step = %d.",
step);
int64_t start =
(*starts)[i] < 0 ? ((*starts)[i] + dim_value) : (*starts)[i];
start = std::max(start, static_cast<int64_t>(0));
int64_t end =
0 < step && (*ends)[i] < 0 ? ((*ends)[i] + dim_value) : (*ends)[i];
end = std::min(end, dim_value);
if (step > 0) {
start = std::min(start, dim_value);
end = std::max(end, static_cast<int64_t>(0));
FDASSERT(end > start,
"When step > 0, end should be greater than start, but "
"received end = %d, start = %d.",
end, start)
} else {
start = std::min(start, dim_value - 1);
if (end < -1) {
end += dim_value;
}
end = std::max(end, static_cast<int64_t>(-1));
FDASSERT(start >= end,
"When step < 0, start should be greater than end, but "
"received start = %d, end = %d.",
start, end);
}
(*starts)[i] = start;
(*ends)[i] = end;
} else if (dim_value == 0) {
(*starts)[i] = 0;
(*ends)[i] = 0;
}
}
}
template <typename T, size_t D>
void SliceKernel(const FDTensor& x, const std::vector<int64_t>& axes,
const std::vector<int64_t>& starts,
const std::vector<int64_t>& ends, FDTensor* out) {
FDASSERT(starts.size() == axes.size(),
"The size of starts must be equal to the size of axes.");
FDASSERT(ends.size() == axes.size(),
"The size of ends must be equal to the size of axes.");
auto starts_idx = starts;
auto end_idx = ends;
auto in_dims = x.Shape();
CheckAndUpdateSliceAttrs(in_dims, axes, &starts_idx, &end_idx);
auto slice_dims = GetSliceDims(in_dims, axes, starts, ends);
auto offsets = Eigen::DSizes<Eigen::DenseIndex, D>();
auto extents = Eigen::DSizes<Eigen::DenseIndex, D>();
for (size_t i = 0; i < D; ++i) {
offsets[i] = 0;
extents[i] = slice_dims[i];
}
for (size_t i = 0; i < axes.size(); ++i) {
offsets[axes[i]] = starts[i];
}
out->Allocate(slice_dims, x.Dtype());
auto in_t = EigenTensor<T, D>::From(x, in_dims);
auto out_t = EigenTensor<T, D>::From(*out, slice_dims);
const auto& dev = *EigenDeviceWrapper::GetInstance()->GetDevice();
out_t.device(dev) = in_t.slice(offsets, extents);
}
void Slice(const FDTensor& x, const std::vector<int64_t>& axes,
const std::vector<int64_t>& starts, const std::vector<int64_t>& ends,
FDTensor* out) {
FD_VISIT_ALL_TYPES(
x.dtype, "SliceKernel", ([&] {
int rank = x.Shape().size();
switch (rank) {
case 1:
SliceKernel<data_t, 1>(x, axes, starts, ends, out);
break;
case 2:
SliceKernel<data_t, 2>(x, axes, starts, ends, out);
break;
case 3:
SliceKernel<data_t, 3>(x, axes, starts, ends, out);
break;
case 4:
SliceKernel<data_t, 4>(x, axes, starts, ends, out);
break;
case 5:
SliceKernel<data_t, 5>(x, axes, starts, ends, out);
break;
case 6:
SliceKernel<data_t, 6>(x, axes, starts, ends, out);
break;
default:
FDASSERT(false,
"The rank of input should be less than 7, but received %d.",
rank);
}
}));
}
} // namespace function
} // namespace fastdeploy

View File

@@ -0,0 +1,41 @@
// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#pragma once
#include "fastdeploy/core/fd_tensor.h"
namespace fastdeploy {
namespace function {
/** This operator produces a slice of input along multiple axes.
@param x The input tensor.
@param axes Axes that starts and ends apply to.
@param starts If starts is a list or tuple, the elements of it should be
integers or Tensors with shape [1]. If starts is an Tensor, it should
be an 1-D Tensor. It represents starting indices of corresponding axis
in axes
@param ends If ends is a list or tuple, the elements of it should be
integers or Tensors with shape [1]. If ends is an Tensor, it should
be an 1-D Tensor . It represents ending indices of corresponding axis
in axes.
@param out The output tensor which stores the result.
*/
FASTDEPLOY_DECL void Slice(const FDTensor& x, const std::vector<int64_t>& axes,
const std::vector<int64_t>& starts,
const std::vector<int64_t>& ends, FDTensor* out);
} // namespace function
} // namespace fastdeploy

View File

@@ -0,0 +1,69 @@
// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include "fastdeploy/core/fd_tensor.h"
#include "fastdeploy/function/slice.h"
#include "glog/logging.h"
#include "gtest_utils.h"
#include "gtest/gtest.h"
#include <array>
#include <vector>
namespace fastdeploy {
namespace function {
std::vector<float> CreateTestData() {
// Shape: [2, 3, 4]
std::vector<float> x_data = {
1.8428625, 0.6461913, 0.13740455, 0.11430702, 0.659926, 0.535816,
0.7429162, 0.8456049, -1.21228176, 0.29970083, 0.8621713, 0.40894133,
0.12684688, 2.1566195, -9.42884097, 20.8476526, 0.2458633, 0.669046,
0.87888306, 0.6762589, 0.666453, 0.32523027, 0.4139388, 0.8341406};
return x_data;
}
TEST(fastdeploy, slice) {
CheckShape check_shape;
CheckData check_data;
FDTensor x, y;
auto test_data = CreateTestData();
x.SetExternalData({2, 3, 4}, FDDataType::FP32, test_data.data());
// x[0:1]
Slice(x, {0}, {0}, {1}, &y);
std::vector<float> result = {1.842862, 0.646191, 0.137405, 0.114307,
0.659926, 0.535816, 0.742916, 0.845605,
-1.212282, 0.299701, 0.862171, 0.408941};
check_shape(y.shape, {1, 3, 4});
check_data(reinterpret_cast<const float*>(y.Data()), result.data(),
result.size());
// x[:, 1:2]
Slice(x, {1}, {1}, {2}, &y);
result = {0.659926, 0.535816, 0.742916, 0.845605,
0.245863, 0.669046, 0.878883, 0.676259};
check_shape(y.shape, {2, 1, 4});
check_data(reinterpret_cast<const float*>(y.Data()), result.data(),
result.size());
// x[:, 0:1, 2:4]
Slice(x, {1, 2}, {0, 2}, {1, 4}, &y);
result = {0.137405, 0.114307, -9.428841, 20.847652};
check_shape(y.shape, {2, 1, 2});
check_data(reinterpret_cast<const float*>(y.Data()), result.data(),
result.size());
}
} // namespace function
} // namespace fastdeploy