cavis/libnd4j/include/ops/declarable/generic/transforms/slice.cpp

230 lines
9.6 KiB
C++
Raw Normal View History

2019-06-06 14:21:15 +02:00
/*******************************************************************************
* Copyright (c) 2015-2018 Skymind, Inc.
*
* This program and the accompanying materials are made available under the
* terms of the Apache License, Version 2.0 which is available at
* https://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.
*
* SPDX-License-Identifier: Apache-2.0
******************************************************************************/
//
// Created by raver119 on 02.11.2017.
//
#include <system/op_boilerplate.h>
2019-06-06 14:21:15 +02:00
//#if NOT_EXCLUDED(OP_slice)
#include <ops/declarable/CustomOperations.h>
#include <helpers/ShapeUtils.h>
namespace sd {
2019-06-06 14:21:15 +02:00
namespace ops {
CUSTOM_OP_IMPL(slice, 1, 1, false, 0, -2) {
auto input = INPUT_VARIABLE(0);
auto output = OUTPUT_VARIABLE(0);
int x_rank = input->rankOf();
std::vector<int> begin;
std::vector<int> sz;
if (block.width() == 3) {
auto b = INPUT_VARIABLE(1);
auto e = INPUT_VARIABLE(2);
begin = b->template asVectorT<int>();
sz = e->template asVectorT<int>();
} else {
REQUIRE_TRUE(block.numI() >= x_rank * 2, 0, "Number of IArgs should be equal to [%i] but got [%i] instead", x_rank * 2, block.numI());
ShapeUtils::copyVectorPart(begin, *(block.getIArguments()), x_rank, 0);
ShapeUtils::copyVectorPart(sz, *(block.getIArguments()), x_rank, x_rank);
}
REQUIRE_TRUE(begin.size() == x_rank, 0, "begin array should have length of [%i] but got [%i] instead", x_rank, begin.size());
REQUIRE_TRUE(sz.size() == x_rank, 0, "size array should have length of [%i] but got [%i] instead", x_rank, sz.size());
std::vector<Nd4jLong> indices(2 * x_rank);
auto empty = false;
for (int e = 0; e < x_rank; e++) {
int size = sz[e];
int start = begin[e];
REQUIRE_TRUE(start >= 0, 0, "Slice: start index should not be negative");
REQUIRE_TRUE(start <= input->sizeAt(e), 0, "Index %i is invalid for dimension %i with size %i.", start, e, input->shapeInfo()[e + 1]);
if (size == -1){
size = input->sizeAt(e) - start;
}
REQUIRE_TRUE(size >= 0, 0, "Slice: interval for dimension %i is less then 1");
REQUIRE_TRUE(start + size <= input->sizeAt(e), 0, "Slice: interval [%i, %i] is out of bounds for dimension %i with size %i", start, start + size, e, input->sizeAt(e));
if(start == input->sizeAt(e) || size == 0 ){
empty = true;
//Don't break to perform input validation on other dims
}
profiling of stack and unstack ops (#261) * - profiling of stack and unstack ops Signed-off-by: Yurii <iuriish@yahoo.com> * - fix bug in cpu concat op Signed-off-by: Yurii <iuriish@yahoo.com> * - correction of cuda stack and unstack Signed-off-by: Yurii <iuriish@yahoo.com> * - change shape.h method which operates with unity dimensions strides Signed-off-by: Yurii <iuriish@yahoo.com> * - rearrange stack tests Signed-off-by: Yurii <iuriish@yahoo.com> * - correct evaluation of smallest stride for moving through contiguous axis Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to update signature of function strideOverContigAxis in cuda concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - remove ShapeUtils::shapeAsString method applied before input arrays validations Signed-off-by: Yurii <iuriish@yahoo.com> * - further removing of ShapeUtils::shapeAsString Signed-off-by: Yurii <iuriish@yahoo.com> * - take sub-array shapeIndo/offset calculation out of NDArray class - add possibility of contiguous memory copy in execTransformAny op if opNum == assign Signed-off-by: Yurii <iuriish@yahoo.com> * - correct test_empty_scatter_2 in EmptyTests.cpp Signed-off-by: Yurii <iuriish@yahoo.com> * - profiling of slice op Signed-off-by: Yurii <iuriish@yahoo.com> * - get rid of contiguous memcpy for some cases in concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to declare oid nd4j::SpecialMethods<T>::splitCpuGeneric Signed-off-by: Yurii <iuriish@yahoo.com> * - correct typo in calculation of threads in cuda split op Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to correct another set of threads variables in split cuda ops Signed-off-by: Yurii <iuriish@yahoo.com> * - further conflicts resolving Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: raver119 <raver119@gmail.com>
2020-03-03 05:32:37 +01:00
2019-06-06 14:21:15 +02:00
indices[2*e] = start;
indices[2*e+1] = start + size;
}
if(empty){
REQUIRE_TRUE(output->isEmpty(), 0, "Slice: empty array indices requested, but output array is not empty");
return Status::OK();
}
profiling of stack and unstack ops (#261) * - profiling of stack and unstack ops Signed-off-by: Yurii <iuriish@yahoo.com> * - fix bug in cpu concat op Signed-off-by: Yurii <iuriish@yahoo.com> * - correction of cuda stack and unstack Signed-off-by: Yurii <iuriish@yahoo.com> * - change shape.h method which operates with unity dimensions strides Signed-off-by: Yurii <iuriish@yahoo.com> * - rearrange stack tests Signed-off-by: Yurii <iuriish@yahoo.com> * - correct evaluation of smallest stride for moving through contiguous axis Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to update signature of function strideOverContigAxis in cuda concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - remove ShapeUtils::shapeAsString method applied before input arrays validations Signed-off-by: Yurii <iuriish@yahoo.com> * - further removing of ShapeUtils::shapeAsString Signed-off-by: Yurii <iuriish@yahoo.com> * - take sub-array shapeIndo/offset calculation out of NDArray class - add possibility of contiguous memory copy in execTransformAny op if opNum == assign Signed-off-by: Yurii <iuriish@yahoo.com> * - correct test_empty_scatter_2 in EmptyTests.cpp Signed-off-by: Yurii <iuriish@yahoo.com> * - profiling of slice op Signed-off-by: Yurii <iuriish@yahoo.com> * - get rid of contiguous memcpy for some cases in concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to declare oid nd4j::SpecialMethods<T>::splitCpuGeneric Signed-off-by: Yurii <iuriish@yahoo.com> * - correct typo in calculation of threads in cuda split op Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to correct another set of threads variables in split cuda ops Signed-off-by: Yurii <iuriish@yahoo.com> * - further conflicts resolving Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: raver119 <raver119@gmail.com>
2020-03-03 05:32:37 +01:00
Nd4jLong* subArrShapeInfo = nullptr;
ALLOCATE(subArrShapeInfo, block.getWorkspace(), shape::shapeInfoLength(input->rankOf()), Nd4jLong);
Nd4jLong offset;
shape::calcSubArrShapeInfoAndOffset(indices.data(), input->getShapeInfo(), subArrShapeInfo, offset, true);
auto subArrShapeInfoPack = ConstantShapeHelper::getInstance()->bufferForShapeInfo(subArrShapeInfo);
NDArray::prepareSpecialUse({output}, {input});
NativeOpExecutioner::execTransformAny(block.launchContext(), sd::transform::Assign,
input->bufferWithOffset(offset), reinterpret_cast<Nd4jLong *>(subArrShapeInfoPack.primary()),
input->specialBufferWithOffset(offset), reinterpret_cast<Nd4jLong *>(subArrShapeInfoPack.special()),
output->buffer(), output->shapeInfo(), output->specialBuffer(), output->specialShapeInfo(),
nullptr, nullptr, nullptr, true);
NDArray::registerSpecialUse({output}, {input});
RELEASE(subArrShapeInfo, block.getWorkspace());
// auto sub = (*input)(indices, true);
// output->assign(sub);
2019-06-06 14:21:15 +02:00
STORE_RESULT(output);
return Status::OK();
}
DECLARE_TYPES(slice) {
getOpDescriptor()
->setAllowedInputTypes(sd::DataType::ANY)
2019-06-06 14:21:15 +02:00
->setSameMode(true);
}
DECLARE_SHAPE_FN(slice) {
auto inShape = inputShape->at(0);
auto x_rank = shape::rank(inShape);
std::vector<int> begin;
std::vector<int> sz;
if (block.width() == 3) {
auto b = INPUT_VARIABLE(1);
auto e = INPUT_VARIABLE(2);
begin = b->template asVectorT<int>();
sz = e->template asVectorT<int>();
} else {
REQUIRE_TRUE(block.numI() >= x_rank * 2, 0, "Number of IArgs should be equal to [%i] but got [%i] instead", x_rank * 2, block.numI());
ShapeUtils::copyVectorPart(begin, *(block.getIArguments()), x_rank, 0);
ShapeUtils::copyVectorPart(sz, *(block.getIArguments()), x_rank, x_rank);
}
REQUIRE_TRUE(begin.size() == x_rank, 0, "Begin array should have length of [%i] but got [%i] instead", x_rank, begin.size());
REQUIRE_TRUE(sz.size() == x_rank, 0, "Size array should have length of [%i] but got [%i] instead", x_rank, sz.size());
profiling of stack and unstack ops (#261) * - profiling of stack and unstack ops Signed-off-by: Yurii <iuriish@yahoo.com> * - fix bug in cpu concat op Signed-off-by: Yurii <iuriish@yahoo.com> * - correction of cuda stack and unstack Signed-off-by: Yurii <iuriish@yahoo.com> * - change shape.h method which operates with unity dimensions strides Signed-off-by: Yurii <iuriish@yahoo.com> * - rearrange stack tests Signed-off-by: Yurii <iuriish@yahoo.com> * - correct evaluation of smallest stride for moving through contiguous axis Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to update signature of function strideOverContigAxis in cuda concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - remove ShapeUtils::shapeAsString method applied before input arrays validations Signed-off-by: Yurii <iuriish@yahoo.com> * - further removing of ShapeUtils::shapeAsString Signed-off-by: Yurii <iuriish@yahoo.com> * - take sub-array shapeIndo/offset calculation out of NDArray class - add possibility of contiguous memory copy in execTransformAny op if opNum == assign Signed-off-by: Yurii <iuriish@yahoo.com> * - correct test_empty_scatter_2 in EmptyTests.cpp Signed-off-by: Yurii <iuriish@yahoo.com> * - profiling of slice op Signed-off-by: Yurii <iuriish@yahoo.com> * - get rid of contiguous memcpy for some cases in concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to declare oid nd4j::SpecialMethods<T>::splitCpuGeneric Signed-off-by: Yurii <iuriish@yahoo.com> * - correct typo in calculation of threads in cuda split op Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to correct another set of threads variables in split cuda ops Signed-off-by: Yurii <iuriish@yahoo.com> * - further conflicts resolving Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: raver119 <raver119@gmail.com>
2020-03-03 05:32:37 +01:00
2019-06-06 14:21:15 +02:00
std::vector<Nd4jLong> shape;
auto empty = false;
for (int e = 0; e < x_rank; e++) {
auto size = sz[e];
auto start = begin[e];
if(size == -1){
size = inShape[e+1] - start;
}
//Bounds checking. Note that begin[i] == size[i] means empty array
REQUIRE_TRUE(start >= 0 && start <= inShape[e+1], 0, "Invalid begin[%i] value: Begin must satisfy 0 <= begin <= size[i], got begin=%i for dimension size %i", e, start, inShape[e+1]);
REQUIRE_TRUE(size == -1 || size >= 0, 0, "Invalid size[%i] value: must be positive (or -1 for 'all remaining'), got %i", e, size, inShape[e+1]);
REQUIRE_TRUE(start >= 0 && start <= inShape[e+1], 0, "Invalid begin[%i] value: Begin must satisfy 0 <= begin <= size[i], got begin=%i for dimension size %i", e, start, inShape[e+1]);
REQUIRE_TRUE(start + size <= inShape[e+1], 0, "Slice: interval [%i, %i] is out of bounds for dimension %i with size %i", start, start + size, e, inShape[e+1]);
Dev branch merge: dev_20190606 (#7904) * correct logsoftmax looss (#2) * Small SameDiff listener fix (#4) * Various fixes (#6) * #7839 Fix for asXMatrix and tests * #7866 EmbeddingSequenceLayer dtype fix + test * #7856 SameDiff save/load stream methods * #7859 RegressionEvaluation rank 4 fix + tests + axis configuration * EvaluationBinary 3d/4d * More evaluation 3d/4d tests * #7847 Evaluation empty checks * Small test ifx * #7848 Fix median edge case * Improve DL4J samediff layer tests * [WIP] FastText wrapper implemented (#8) * FastText implemented * Some fixes * Fix shapes for wordsNearest * Validation of input vectors * Fixes * Fixed test * Thread tagged * Some tweaks * setContextClassLoader for DeallocatorServiceThread * Numpy format tests (#1) * Various fixes (#11) * #7852 SameDiff gather fix * #7892 SameDiff placeholder to constant conversion * #7890 validate input rank for MLN/CG init methods * Fix broken permute shape calculation * Permute and gather fixes * Tests * #7850 LogSumExp fix + test * Handful of test fixes * Empty arrays with non-scalar shapes (#10) * minor rearrangements for lambdas * empty tensors with non-scalar shapes * numpy empty tensors with non-scalar shapes * few more empty tweaks * Small fixes * conv3d signature update * micro fix in batchnorm mkldnn * Import fixes * Fix * MKL-DNN update * Small fill fix * fill with empty input + test * Fixes * Small error improvement * Fix * one special test * couple of fixes for lstm * Rewrite TFGraphMapper.getNDArrayFromTensor to be maintainable and less error prone * Fixes * FP16 * Unsigned * BFloat16 * Fill op - empty tweaks * - couple of fixes for empty arrays construction - stack updated * strided slice fix * one transform test * provide method for reducing shapeInfo in case of input array is empty * Fixed reduceAlongDimensions to use empty input properly. * couple of broadcast tests * couple of tests broadcast tests + tweak to make them pass * add check of non-empty to methods producing sub-arrays * Fixed reshapeC with zeros in shape. * complete empty check in reduce_... legacy ops * Concat and cumsum/prod * Tweak to empty shape inference on import * add empty check to the rest of reduce legacy ops * one more test * correct typo in evalReduceShapeInfoEmpty * Added tests for reduce_* ops to tests with zero shapes. * few more tests for empty reductions * Fixed strided_slice op with empty case and tests. * one more empty reduction test * Fixed strided_slice test. * add empty check to NDArray::reshapei * infOrMax * empty min/max with infinity tests * made unstack working correctly with empty arrays * few IndexReduce tests + tweaks for empty shapes * add test for empty concat * few tests fixed * Validation fix for reductions on empty shapes * Reverse fix * Reduction shape calc fixes * SameDiff.generateOutputVariable: don't use shape function to determine number of outputs * Range fix * - NDArray constructor updated for scalars/empty arrays - few tests fixed * More fixes * Empty creator fixes * concat fix * concat fix * TF import tests: allow 'both all NaN' and 'both all inf' to pass * Slice, zero fraction, and reshape fixes * transpose, gather * Zero fraction * scalar cast fix * Empty reduction axis support * few more tests fixed * Fixed input checks conforming with TF for concat op and tests. * few tests fixed * matmul scalar shape fix * Fixed checkout for data type and scalarity with concat to allow non-empty scalars with vector concats. * broadcast bool fix * few more tests * few more tests * correct evalReduceShapeInfoEmpty * argmax/argmin + tests * one more empty edge case + one more test * argmax/argmin/realdiv_bp tweaks * empty reshape test + fix * Helper fixes * Small fixes * Gather test fix * Gather test fix * Small fixes * reduce scalar zero values * scalar mean workaround * Remove debug code * along dim mean workaround * one more test * - equalsTo() tweak for empty arrays - one more test * broadcast tweaks
2019-06-15 13:34:34 +02:00
if(start == inShape[e+1] ){
size = 0;
2019-06-06 14:21:15 +02:00
}
shape.emplace_back(size);
}
auto newShape = ConstantShapeHelper::getInstance()->createShapeInfo(ArrayOptions::dataType(inShape), 'c', shape);
return SHAPELIST(newShape);
}
DECLARE_TYPES(slice_bp) {
getOpDescriptor()
->setAllowedInputTypes(sd::DataType::ANY)
2019-06-06 14:21:15 +02:00
->setAllowedOutputTypes({ALL_FLOATS});
}
CUSTOM_OP_IMPL(slice_bp, 2, 1, false, 0, -2) {
auto input = INPUT_VARIABLE(0);
auto epsNext = block.width() == 4 ? INPUT_VARIABLE(3) : INPUT_VARIABLE(1);
auto output = OUTPUT_VARIABLE(0);
output->assign(0.);
int x_rank = input->rankOf();
std::vector<int> begin;
std::vector<int> end;
if (block.width() == 4) {
auto b = INPUT_VARIABLE(1);
auto e = INPUT_VARIABLE(2);
begin = b->template asVectorT<int>();
end = e->template asVectorT<int>();
} else {
REQUIRE_TRUE(block.numI() >= x_rank * 2, 0, "Number of IArgs should be equal to [%i] but got [%i] instead", x_rank * 2, block.numI());
ShapeUtils::copyVectorPart(begin, *(block.getIArguments()), x_rank, 0);
ShapeUtils::copyVectorPart(end, *(block.getIArguments()), x_rank, x_rank);
}
REQUIRE_TRUE(begin.size() == x_rank, 0, "begin array should have length of [%i] but got [%i] instead", x_rank, begin.size());
REQUIRE_TRUE(end.size() == x_rank, 0, "end array should have length of [%i] but got [%i] instead", x_rank, end.size());
std::vector<Nd4jLong> indices(2 * x_rank);
for (int e = 0; e < x_rank; e++) {
int size = end[e];
int start = begin[e];
if (size == -1){ //-1 means all remaining values
size = input->sizeAt(e) - start;
}
REQUIRE_TRUE(size > 0, 0, "Slice: interval for dimension %i is less then 1", e);
profiling of stack and unstack ops (#261) * - profiling of stack and unstack ops Signed-off-by: Yurii <iuriish@yahoo.com> * - fix bug in cpu concat op Signed-off-by: Yurii <iuriish@yahoo.com> * - correction of cuda stack and unstack Signed-off-by: Yurii <iuriish@yahoo.com> * - change shape.h method which operates with unity dimensions strides Signed-off-by: Yurii <iuriish@yahoo.com> * - rearrange stack tests Signed-off-by: Yurii <iuriish@yahoo.com> * - correct evaluation of smallest stride for moving through contiguous axis Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to update signature of function strideOverContigAxis in cuda concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - remove ShapeUtils::shapeAsString method applied before input arrays validations Signed-off-by: Yurii <iuriish@yahoo.com> * - further removing of ShapeUtils::shapeAsString Signed-off-by: Yurii <iuriish@yahoo.com> * - take sub-array shapeIndo/offset calculation out of NDArray class - add possibility of contiguous memory copy in execTransformAny op if opNum == assign Signed-off-by: Yurii <iuriish@yahoo.com> * - correct test_empty_scatter_2 in EmptyTests.cpp Signed-off-by: Yurii <iuriish@yahoo.com> * - profiling of slice op Signed-off-by: Yurii <iuriish@yahoo.com> * - get rid of contiguous memcpy for some cases in concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to declare oid nd4j::SpecialMethods<T>::splitCpuGeneric Signed-off-by: Yurii <iuriish@yahoo.com> * - correct typo in calculation of threads in cuda split op Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to correct another set of threads variables in split cuda ops Signed-off-by: Yurii <iuriish@yahoo.com> * - further conflicts resolving Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: raver119 <raver119@gmail.com>
2020-03-03 05:32:37 +01:00
2019-06-06 14:21:15 +02:00
indices[2*e] = start;
indices[2*e + 1] = start + size;
}
auto sub = (*output)(indices, true);
profiling of stack and unstack ops (#261) * - profiling of stack and unstack ops Signed-off-by: Yurii <iuriish@yahoo.com> * - fix bug in cpu concat op Signed-off-by: Yurii <iuriish@yahoo.com> * - correction of cuda stack and unstack Signed-off-by: Yurii <iuriish@yahoo.com> * - change shape.h method which operates with unity dimensions strides Signed-off-by: Yurii <iuriish@yahoo.com> * - rearrange stack tests Signed-off-by: Yurii <iuriish@yahoo.com> * - correct evaluation of smallest stride for moving through contiguous axis Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to update signature of function strideOverContigAxis in cuda concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - remove ShapeUtils::shapeAsString method applied before input arrays validations Signed-off-by: Yurii <iuriish@yahoo.com> * - further removing of ShapeUtils::shapeAsString Signed-off-by: Yurii <iuriish@yahoo.com> * - take sub-array shapeIndo/offset calculation out of NDArray class - add possibility of contiguous memory copy in execTransformAny op if opNum == assign Signed-off-by: Yurii <iuriish@yahoo.com> * - correct test_empty_scatter_2 in EmptyTests.cpp Signed-off-by: Yurii <iuriish@yahoo.com> * - profiling of slice op Signed-off-by: Yurii <iuriish@yahoo.com> * - get rid of contiguous memcpy for some cases in concat and split ops Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to declare oid nd4j::SpecialMethods<T>::splitCpuGeneric Signed-off-by: Yurii <iuriish@yahoo.com> * - correct typo in calculation of threads in cuda split op Signed-off-by: Yurii <iuriish@yahoo.com> * - forgot to correct another set of threads variables in split cuda ops Signed-off-by: Yurii <iuriish@yahoo.com> * - further conflicts resolving Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: raver119 <raver119@gmail.com>
2020-03-03 05:32:37 +01:00
sub.assign(epsNext);
2019-06-06 14:21:15 +02:00
return Status::OK();
}
DECLARE_SHAPE_FN(slice_bp) {
auto inShape = inputShape->at(0);
Nd4jLong *newShape;
COPY_SHAPE(inShape, newShape);
return SHAPELIST(CONSTANT(newShape));
}
}
}
//#endif