cavis/libnd4j/include/ops/declarable/generic/nn/convo/upsampling2d.cpp

128 lines
5.1 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
******************************************************************************/
//
// @author raver119, created on 29/10/17.
// @author Yurii Shyrma (iuriish@yahoo.com), changed on 03.05.2018
//
#include <system/op_boilerplate.h>
2019-06-06 14:21:15 +02:00
#if NOT_EXCLUDED(OP_upsampling2d)
#include <ops/declarable/CustomOperations.h>
#include <ops/declarable/helpers/convolutions.h>
namespace sd {
2019-06-06 14:21:15 +02:00
namespace ops {
//////////////////////////////////////////////////////////////////////
CUSTOM_OP_IMPL(upsampling2d, 1, 1, false, 0, 2) {
auto input = INPUT_VARIABLE(0); // [bS, iC, iH, iW] (NCHW) or [bS, iH, iW, iC] (NHWC)
Nullify (#304) * initial commit Signed-off-by: raver119 <raver119@gmail.com> * bunch of tweaks Signed-off-by: raver119 <raver119@gmail.com> * hamming distance nullification Signed-off-by: raver119 <raver119@gmail.com> * Add output array value assignment for testing/debugging Signed-off-by: Alex Black <blacka101@gmail.com> * don't assign empty arrays Signed-off-by: raver119 <raver119@gmail.com> * conv2d/conv3d/depthwise2d nullified Signed-off-by: raver119 <raver119@gmail.com> * conv2d/conv3d/depthwise2d nullified Signed-off-by: raver119 <raver119@gmail.com> * conv2d/conv3d/depthwise2d nullified Signed-off-by: raver119 <raver119@gmail.com> * few more fixes Signed-off-by: raver119 <raver119@gmail.com> * im2col Signed-off-by: raver119 <raver119@gmail.com> * pooling? Signed-off-by: raver119 <raver119@gmail.com> * more nullified Signed-off-by: raver119 <raver119@gmail.com> * ismax nullified Signed-off-by: raver119 <raver119@gmail.com> * rollback ismax nullification Signed-off-by: raver119 <raver119@gmail.com> * synchronized cublas handle use on per-device basis Signed-off-by: raver119 <raver119@gmail.com> * hiding method from jcpp Signed-off-by: raver119 <raver119@gmail.com> * get rid of test assigns in DeclarableOp Signed-off-by: raver119 <raver119@gmail.com> * get rid of assigns Signed-off-by: raver119 <raver119@gmail.com> * proper deviceId is back Signed-off-by: raver119 <raver119@gmail.com> * include fixed Signed-off-by: raver119 <raver119@gmail.com> Co-authored-by: Alex Black <blacka101@gmail.com>
2020-03-20 06:49:28 +01:00
auto output = OUTPUT_NULLIFIED(0); // [bS, iC, factorH*iH, factorW*iW ] (NCHW) or [bS, factorH*iH, factorW*iW, iC] (NHWC)
2019-06-06 14:21:15 +02:00
const int factorH = INT_ARG(0);
const int factorW = INT_ARG(1);
const int isNCHW = block.getIArguments()->size() > 2 ? INT_ARG(2) : 0; // INT_ARG(2): 0-NCHW, 1-NHWC
REQUIRE_TRUE(input->rankOf() == 4, 0, "UPSAMPLING2D op: input should be 4D, but got %i instead!", input->rankOf());
REQUIRE_TRUE(output->rankOf() == 4, 0, "UPSAMPLING2D op: output should be 4D, but got %i instead!", output->rankOf());
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
ConvolutionUtils::upsampling2d(block, *input, *output, factorH, factorW, (bool)isNCHW);
2019-06-06 14:21:15 +02:00
return Status::OK();
}
DECLARE_SYN(upsampling, upsampling2d);
DECLARE_TYPES(upsampling2d) {
getOpDescriptor()
->setAllowedInputTypes(sd::DataType::ANY)
2019-06-06 14:21:15 +02:00
->setAllowedOutputTypes({ALL_FLOATS});
}
DECLARE_SHAPE_FN(upsampling2d) {
auto inputShapeInfo = inputShape->at(0);
REQUIRE_TRUE(inputShapeInfo[0] == 4, 0, "UPSAMPLING2D op: input should be 4D, but got %i instead!", inputShapeInfo[0]);
const int factorH = INT_ARG(0);
const int factorW = INT_ARG(1);
const int isNCHW = block.getIArguments()->size() > 2 ? INT_ARG(2) : 0; // INT_ARG(2): 0-NCHW, 1-NHWC
Nd4jLong *outputShapeInfo = nullptr;
ALLOCATE(outputShapeInfo, block.getWorkspace(), shape::shapeInfoLength(inputShapeInfo[0]), Nd4jLong);
outputShapeInfo[0] = inputShapeInfo[0];
outputShapeInfo[1] = inputShapeInfo[1];
if(isNCHW) {
outputShapeInfo[2] = inputShapeInfo[2];
outputShapeInfo[3] = inputShapeInfo[3] * factorH;
outputShapeInfo[4] = inputShapeInfo[4] * factorW;
}
else {
outputShapeInfo[2] = inputShapeInfo[2] * factorH;
outputShapeInfo[3] = inputShapeInfo[3] * factorW;
outputShapeInfo[4] = inputShapeInfo[4];
}
ShapeUtils::updateStridesAndType(outputShapeInfo, inputShapeInfo, shape::order(inputShapeInfo));
return SHAPELIST(CONSTANT(outputShapeInfo));
}
DECLARE_TYPES(upsampling2d_bp) {
getOpDescriptor()
->setAllowedInputTypes(sd::DataType::ANY)
2019-06-06 14:21:15 +02:00
->setAllowedOutputTypes({ALL_FLOATS});
}
//////////////////////////////////////////////////////////////////////
CUSTOM_OP_IMPL(upsampling2d_bp, 2, 1, false, 0, 0) {
// NDArray<T>* input = INPUT_VARIABLE(0); // [bS, iC, iH, iW] (NCHW) or [bS, iH, iW, iC] (NHWC)
auto gradO = INPUT_VARIABLE(1); // [bS, iC, factorH*iH, factorW*iW ] (NCHW) or [bS, factorH*iH, factorW*iW, iC] (NHWC)
Nullify (#304) * initial commit Signed-off-by: raver119 <raver119@gmail.com> * bunch of tweaks Signed-off-by: raver119 <raver119@gmail.com> * hamming distance nullification Signed-off-by: raver119 <raver119@gmail.com> * Add output array value assignment for testing/debugging Signed-off-by: Alex Black <blacka101@gmail.com> * don't assign empty arrays Signed-off-by: raver119 <raver119@gmail.com> * conv2d/conv3d/depthwise2d nullified Signed-off-by: raver119 <raver119@gmail.com> * conv2d/conv3d/depthwise2d nullified Signed-off-by: raver119 <raver119@gmail.com> * conv2d/conv3d/depthwise2d nullified Signed-off-by: raver119 <raver119@gmail.com> * few more fixes Signed-off-by: raver119 <raver119@gmail.com> * im2col Signed-off-by: raver119 <raver119@gmail.com> * pooling? Signed-off-by: raver119 <raver119@gmail.com> * more nullified Signed-off-by: raver119 <raver119@gmail.com> * ismax nullified Signed-off-by: raver119 <raver119@gmail.com> * rollback ismax nullification Signed-off-by: raver119 <raver119@gmail.com> * synchronized cublas handle use on per-device basis Signed-off-by: raver119 <raver119@gmail.com> * hiding method from jcpp Signed-off-by: raver119 <raver119@gmail.com> * get rid of test assigns in DeclarableOp Signed-off-by: raver119 <raver119@gmail.com> * get rid of assigns Signed-off-by: raver119 <raver119@gmail.com> * proper deviceId is back Signed-off-by: raver119 <raver119@gmail.com> * include fixed Signed-off-by: raver119 <raver119@gmail.com> Co-authored-by: Alex Black <blacka101@gmail.com>
2020-03-20 06:49:28 +01:00
auto gradI = OUTPUT_NULLIFIED(0); // [bS, iC, iH, iW] (NCHW) or [bS, iH, iW, iC] (NHWC)
2019-06-06 14:21:15 +02:00
const int isNCHW = block.getIArguments()->size() > 0 ? INT_ARG(0) : 0; // INT_ARG(0): 0-NCHW, 1-NHWC
// REQUIRE_TRUE(input->rankOf() == 4, 0, "UPSAMPLING2D_BP op: input array must be 4D, but got %i instead!", input->rankOf());
REQUIRE_TRUE(gradO->rankOf() == 4, 0, "UPSAMPLING2D_BP op: output's gradient array must be 4D, but got %i instead!", gradO->rankOf());
REQUIRE_TRUE(gradI->rankOf() == 4, 0, "UPSAMPLING2D_BP op: input's gradient array must be 4D, but got %i instead!", gradI->rankOf());
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
ConvolutionUtils::upsampling2dBP(block, *gradO, *gradI, (bool)isNCHW);
2019-06-06 14:21:15 +02:00
return Status::OK();
}
DECLARE_SYN(upsampling_bp, upsampling2d_bp);
DECLARE_SHAPE_FN(upsampling2d_bp) {
REQUIRE_TRUE(inputShape->at(0)[0] == 4, 0, "UPSAMPLING2D_BP op: input array must be 4D, but got %i instead!", inputShape->at(0)[0]);
REQUIRE_TRUE(inputShape->at(1)[0] == 4, 0, "UPSAMPLING2D_BP op: output's gradient array must be 4D, but got %i instead!", inputShape->at(1)[0]);
auto gradIShapeInfo = ShapeBuilders::copyShapeInfoAndType(inputShape->at(0), inputShape->at(1), false, block.getWorkspace());
return SHAPELIST(CONSTANT(gradIShapeInfo));
}
}
}
#endif