cavis/libnd4j/include/system/play.h

173 lines
7.4 KiB
C
Raw Normal View History

2021-02-01 13:31:45 +01:00
/* ******************************************************************************
*
2019-06-06 14:21:15 +02:00
*
* 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.
*
2021-02-01 13:31:45 +01:00
* See the NOTICE file distributed with this work for additional
* information regarding copyright ownership.
2019-06-06 14:21:15 +02:00
* 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 21.08.16.
//
#ifndef LIBND4J_PLAY_H
#define LIBND4J_PLAY_H
cuDNN integration (#150) * initial commit Signed-off-by: raver119 <raver119@gmail.com> * one file Signed-off-by: raver119 <raver119@gmail.com> * few more includes Signed-off-by: raver119 <raver119@gmail.com> * m? Signed-off-by: raver119 <raver119@gmail.com> * const Signed-off-by: raver119 <raver119@gmail.com> * cudnn linkage in tests Signed-off-by: raver119 <raver119@gmail.com> * culibos Signed-off-by: raver119 <raver119@gmail.com> * static reminder Signed-off-by: raver119 <raver119@gmail.com> * platform engine tag Signed-off-by: raver119 <raver119@gmail.com> * HAVE_CUDNN moved to config.h.in Signed-off-by: raver119 <raver119@gmail.com> * include Signed-off-by: raver119 <raver119@gmail.com> * include Signed-off-by: raver119 <raver119@gmail.com> * skip cudnn handle creation if there's not cudnn Signed-off-by: raver119 <raver119@gmail.com> * meh Signed-off-by: raver119 <raver119@gmail.com> * target device in context Signed-off-by: raver119 <raver119@gmail.com> * platform engines Signed-off-by: raver119 <raver119@gmail.com> * platform engines Signed-off-by: raver119 <raver119@gmail.com> * allow multiple -h args Signed-off-by: raver119 <raver119@gmail.com> * allow multiple -h args Signed-off-by: raver119 <raver119@gmail.com> * move mkldnn out of CPU block Signed-off-by: raver119 <raver119@gmail.com> * link to mkldnn on cuda Signed-off-by: raver119 <raver119@gmail.com> * less prints Signed-off-by: raver119 <raver119@gmail.com> * minor tweaks Signed-off-by: raver119 <raver119@gmail.com> * next step Signed-off-by: raver119 <raver119@gmail.com> * conv2d NCHW draft Signed-off-by: raver119 <raver119@gmail.com> * conv2d biasAdd Signed-off-by: raver119 <raver119@gmail.com> * test for MKL/CUDNN combined use Signed-off-by: raver119 <raver119@gmail.com> * - provide additional code for conv2d ff based on cudnn api, not tested yet Signed-off-by: Yurii <iuriish@yahoo.com> * - further work on conv2d helper based on using cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - fixing several cuda bugs which appeared after cudnn lib had been started to use Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of conv2d backprop op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - implementaion of conv3d and conv3d_bp ops based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - bugs fixing in conv3d/conv3d_bp ops (cudnn in use) Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of depthwiseConv2d (ff/bp) op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of batchnorm ff op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - disable cudnn batchnorm temporary Signed-off-by: Yurii <iuriish@yahoo.com> * - add minor change in cmake Signed-off-by: Yurii <iuriish@yahoo.com> * engine for depthwise mkldnn Signed-off-by: raver119 <raver119@gmail.com> * couple of includes Signed-off-by: raver119 <raver119@gmail.com> * - provide permutation to cudnn batchnorm ff when format is NHWC Signed-off-by: Yurii <iuriish@yahoo.com> * lgamma fix Signed-off-by: raver119 <raver119@gmail.com> * - eliminate memory leak in two tests Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: Yurii Shyrma <iuriish@yahoo.com>
2020-01-20 19:32:46 +01:00
//#include <type_boilerplate.h>
#include <system/platform_boilerplate.h>
cuDNN integration (#150) * initial commit Signed-off-by: raver119 <raver119@gmail.com> * one file Signed-off-by: raver119 <raver119@gmail.com> * few more includes Signed-off-by: raver119 <raver119@gmail.com> * m? Signed-off-by: raver119 <raver119@gmail.com> * const Signed-off-by: raver119 <raver119@gmail.com> * cudnn linkage in tests Signed-off-by: raver119 <raver119@gmail.com> * culibos Signed-off-by: raver119 <raver119@gmail.com> * static reminder Signed-off-by: raver119 <raver119@gmail.com> * platform engine tag Signed-off-by: raver119 <raver119@gmail.com> * HAVE_CUDNN moved to config.h.in Signed-off-by: raver119 <raver119@gmail.com> * include Signed-off-by: raver119 <raver119@gmail.com> * include Signed-off-by: raver119 <raver119@gmail.com> * skip cudnn handle creation if there's not cudnn Signed-off-by: raver119 <raver119@gmail.com> * meh Signed-off-by: raver119 <raver119@gmail.com> * target device in context Signed-off-by: raver119 <raver119@gmail.com> * platform engines Signed-off-by: raver119 <raver119@gmail.com> * platform engines Signed-off-by: raver119 <raver119@gmail.com> * allow multiple -h args Signed-off-by: raver119 <raver119@gmail.com> * allow multiple -h args Signed-off-by: raver119 <raver119@gmail.com> * move mkldnn out of CPU block Signed-off-by: raver119 <raver119@gmail.com> * link to mkldnn on cuda Signed-off-by: raver119 <raver119@gmail.com> * less prints Signed-off-by: raver119 <raver119@gmail.com> * minor tweaks Signed-off-by: raver119 <raver119@gmail.com> * next step Signed-off-by: raver119 <raver119@gmail.com> * conv2d NCHW draft Signed-off-by: raver119 <raver119@gmail.com> * conv2d biasAdd Signed-off-by: raver119 <raver119@gmail.com> * test for MKL/CUDNN combined use Signed-off-by: raver119 <raver119@gmail.com> * - provide additional code for conv2d ff based on cudnn api, not tested yet Signed-off-by: Yurii <iuriish@yahoo.com> * - further work on conv2d helper based on using cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - fixing several cuda bugs which appeared after cudnn lib had been started to use Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of conv2d backprop op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - implementaion of conv3d and conv3d_bp ops based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - bugs fixing in conv3d/conv3d_bp ops (cudnn in use) Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of depthwiseConv2d (ff/bp) op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of batchnorm ff op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - disable cudnn batchnorm temporary Signed-off-by: Yurii <iuriish@yahoo.com> * - add minor change in cmake Signed-off-by: Yurii <iuriish@yahoo.com> * engine for depthwise mkldnn Signed-off-by: raver119 <raver119@gmail.com> * couple of includes Signed-off-by: raver119 <raver119@gmail.com> * - provide permutation to cudnn batchnorm ff when format is NHWC Signed-off-by: Yurii <iuriish@yahoo.com> * lgamma fix Signed-off-by: raver119 <raver119@gmail.com> * - eliminate memory leak in two tests Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: Yurii Shyrma <iuriish@yahoo.com>
2020-01-20 19:32:46 +01:00
/*
2019-06-06 14:21:15 +02:00
#define DATA_TYPES \
(DATA_FLOAT, float) ,\
(DATA_DOUBLE, double)
#define Y_TYPES \
(DATA_INT8, int8_t) ,\
(DATA_INT16, int16_t)
#define Z_TYPES \
(DATA_UINT8, uint8_t) ,\
(DATA_UINT16, uint16_t)
#define PWT_LIST \
(float, long, float),\
(float, long, long)
[WIP] multi-device support (#80) * fix pad javadoc and @see links. (#72) Signed-off-by: Robert Altena <Rob@Ra-ai.com> * [WIP] More fixes (#73) * special tests for ConstantTadHelper/ConstantShapeHelper Signed-off-by: raver119 <raver119@gmail.com> * release methods for data buffers Signed-off-by: raver119 <raver119@gmail.com> * delete temporary buffer Java side Signed-off-by: raver119 <raver119@gmail.com> * delete temporary buffer Java side Signed-off-by: raver119 <raver119@gmail.com> * delete temporary TadPack C++/Java side (#74) Signed-off-by: raver119 <raver119@gmail.com> * Zoo model TF import test updates (#75) * argLine fix, update compression_gru comment * updated comment for xception * undid but commented argLine change * updated xlnet comment * copyright headers * - new NDArray methods like()/ulike() (#77) - fix for depthwise_conv2d_bp + special test Signed-off-by: raver119 <raver119@gmail.com> * upsampling2d fix CUDA Signed-off-by: raver119 <raver119@gmail.com> * DL4J trace logging (#79) * MLN/CG trace logging for debugging Signed-off-by: AlexDBlack <blacka101@gmail.com> * Tiny tweak Signed-off-by: AlexDBlack <blacka101@gmail.com> * strided_slice_bp shape fn leak fix Signed-off-by: raver119 <raver119@gmail.com> * SameDiff fixes and naming (#78) * remove SDVariable inplace methods * import methods * npe fix in OpVal * removed SameDiff inplace ops from tests * Naming updates, moved to centralized methods in SameDiff, should use op_#:# for everything * quick fixes * javadoc * SDVariable eval with placeholders * use regex match * better matching * initial commit Signed-off-by: raver119 <raver119@gmail.com> * initial commit Signed-off-by: raver119 <raver119@gmail.com> * fix javadoc. (#76) * fix javadoc. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * replace most @see with @link s. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * 4 additional tests Signed-off-by: raver119 <raver119@gmail.com> * launch context reorganization Signed-off-by: raver119 <raver119@gmail.com> * LaunchContext reorganization Signed-off-by: raver119 <raver119@gmail.com> * per-device LaunchContext Signed-off-by: raver119 <raver119@gmail.com> * Various DL4J/ND4J fixes (#81) * #7954 Force refresh of UI when switching tabs on overview page Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8017 Concurrent modification exception (synchronize) fix Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8033 Don't initialize updater in middle of writing memory crash dump Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8208 Fix shape checks for ND4J int[] creator methods Signed-off-by: AlexDBlack <blacka101@gmail.com> * #6385 #7992 Keras import naming fixes + cleanup Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8016 Upsampling3D - add NDHWC format support Signed-off-by: AlexDBlack <blacka101@gmail.com> * ContextBuffers as separate entity Signed-off-by: raver119 <raver119@gmail.com> * Refactor NativeOps.h to export C functions * Actually export functions from NativeOps.h * Adapt the Java wrappers in ND4J generated with JavaCPP * Create C wrappers for some of the C++ classes currently used by ND4J * ContextBuffers as separate entity Signed-off-by: raver119 <raver119@gmail.com> * remove duplicate code in createBufferDetached. (#83) Signed-off-by: Robert Altena <Rob@Ra-ai.com> * Keras model import - updater lr fix (#84) * Keras model import - updater lr fix Signed-off-by: eraly <susan.eraly@gmail.com> * Keras model import - updater lr fix, cleanup Signed-off-by: eraly <susan.eraly@gmail.com> * ContextBuffers as separate entity Signed-off-by: raver119 <raver119@gmail.com> * ContextBuffers as separate entity Signed-off-by: raver119 <raver119@gmail.com> * Fix functions of OpaqueVariablesSet * thread-local buffers/affinity Signed-off-by: raver119 <raver119@gmail.com> * thread safety for LaunchContext Signed-off-by: raver119 <raver119@gmail.com> * more of thread safety Signed-off-by: raver119 <raver119@gmail.com> * one more multi threaded test Signed-off-by: raver119 <raver119@gmail.com> * SameDiff Convolution Config validation, better output methods (#82) * Conv Config validation & tests Signed-off-by: Ryan Nett <rnett@skymind.io> * stackOutputs utility method Signed-off-by: Ryan Nett <rnett@skymind.io> * use constructor for validation, support negative kernel sizes (infered from weights) Signed-off-by: Ryan Nett <rnett@skymind.io> * better output methods Signed-off-by: Ryan Nett <rnett@skymind.io> * move output to be with fit and evaluate Signed-off-by: Ryan Nett <rnett@skymind.io> * fixes Signed-off-by: Ryan Nett <rnett@skymind.io> * more fixes Signed-off-by: Ryan Nett <rnett@skymind.io> * refactor duplicate code from pad methods. (#86) * refactor duplicate code from pad methods. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * replace switch with if. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * Various ND4J/DL4J fixes and improvements (#87) * Reshape and reallocate - small fixes Signed-off-by: AlexDBlack <blacka101@gmail.com> * Reshape and reallocate - small fixes Signed-off-by: AlexDBlack <blacka101@gmail.com> * #6488 ElementWiseVertex broadcast support Signed-off-by: AlexDBlack <blacka101@gmail.com> * Constructors and broadcast supported it Transforms.max/min Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8054 ElementWiseVertex now supports broadcast inputs Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8057 Nd4j.create overload dtype fix Signed-off-by: AlexDBlack <blacka101@gmail.com> * #7551 ND4J Shape validation fix Signed-off-by: AlexDBlack <blacka101@gmail.com> * [WIP] Numpy boolean import (#91) * numpy bool type Signed-off-by: raver119 <raver119@gmail.com> * numpy bool java side Signed-off-by: raver119 <raver119@gmail.com> * remove create method with unused parameter. (#89) * remove create method with unused parameter. * removed more unused methods. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * removing more unused code. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * last removal of unused code. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * remove createSparse methods. (#92) Signed-off-by: Robert Altena <Rob@Ra-ai.com> * Various ND4J/DL4J fixes (#90) * Deprecate Old*Op instances Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8063 #8054 Broadcast exceptions + cleanup inplace ops Signed-off-by: AlexDBlack <blacka101@gmail.com> * Small fix Signed-off-by: AlexDBlack <blacka101@gmail.com> * Remove bad test condition Signed-off-by: AlexDBlack <blacka101@gmail.com> * #7993 Fix shape function issue in crop_and_resize op Signed-off-by: AlexDBlack <blacka101@gmail.com> * DL4J SameDiff lambda layer fix Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8029 Fix for pnorm backprop math Signed-off-by: AlexDBlack <blacka101@gmail.com> * #8038 Fix Op profiler NaN/Inf triggering + add tests (#93) Signed-off-by: AlexDBlack <blacka101@gmail.com> * createUninitializedDetached refactoring. (#94) * wip * update interface, add null implementations. * Breaking one test in a weird way. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * createUninitializedDetached refactored. Signed-off-by: Robert Altena <Rob@Ra-ai.com> * cuda build fix for issues introduced by recent refactoring Signed-off-by: raver119 <raver119@gmail.com> * [WIP] More of CUDA (#95) * initial commit Signed-off-by: raver119 <raver119@gmail.com> * Implementation of hashcode cuda helper. Working edition. * Fixed parallel test input arangements. * Fixed tests for hashcode op. * Fixed shape calculation for image:crop_and_resize op and test. * NativeOps tests. Initial test suite. * Added tests for indexReduce methods. * Added test on execBroadcast with NDArray as dimensions. * Added test on execBroadcastBool with NDArray as dimensions. * Added tests on execPairwiseTransform and execPairwiseTransofrmBool. * Added tests for execReduce with scalar results. * Added reduce tests for non-empty dims array. * Added tests for reduce3. * Added tests for execScalar. * Added tests for execSummaryStats. * - provide cpu/cuda code for batch_to_space - testing it Signed-off-by: Yurii <yurii@skymind.io> * - remove old test for batch_to_space (had wrong format and numbers were not checked) Signed-off-by: Yurii <yurii@skymind.io> * Fixed complilation errors with test. * Added test for execTransformFloat. * Added test for execTransformSame. * Added test for execTransformBool. * Added test for execTransformStrict. * Added tests for execScalar/execScalarBool with TADs. * Added test for flatten. * - provide cpu/cuda code for space_to_Batch operaion Signed-off-by: Yurii <yurii@skymind.io> * Added test for concat. * comment unnecessary stuff in s_t_b Signed-off-by: Yurii <yurii@skymind.io> * Added test for specialConcat. * Added tests for memcpy/set routines. * Fixed pullRow cuda test. * Added pullRow test. * Added average test. * - correct typo in NDArray::applyPairwiseTransform(nd4j::pairwise::BoolOps op...) Signed-off-by: Yurii <yurii@skymind.io> * - debugging and fixing cuda tests in JavaInteropTests file Signed-off-by: Yurii <yurii@skymind.io> * - correct some tests Signed-off-by: Yurii <yurii@skymind.io> * Added test for shuffle. * Fixed ops declarations. * Restored omp and added shuffle test. * Added convertTypes test. * Added tests for execRandom. Eliminated usage of RandomBuffer with NativeOps. * Added sort tests. * Added tests for execCustomOp. * - further debuging and fixing tests terminated with crash Signed-off-by: Yurii <yurii@skymind.io> * Added tests for calculateOutputShapes. * Addded Benchmarks test. * Commented benchmark tests. * change assertion Signed-off-by: raver119 <raver119@gmail.com> * Added tests for apply_sgd op. Added cpu helper for that op. * Implement cuda helper for aplly_sgd op. Fixed tests for NativeOps. * Added test for assign broadcastable. * Added tests for assign_bp op. * Added tests for axpy op. * - assign/execScalar/execTransformAny signature change - minor test fix Signed-off-by: raver119 <raver119@gmail.com> * Fixed axpy op. * meh Signed-off-by: raver119 <raver119@gmail.com> * - fix tests for nativeOps::concat Signed-off-by: Yurii <yurii@skymind.io> * sequential transform/scalar Signed-off-by: raver119 <raver119@gmail.com> * allow nested parallelism Signed-off-by: raver119 <raver119@gmail.com> * assign_bp leak fix Signed-off-by: raver119 <raver119@gmail.com> * block setRNG fix Signed-off-by: raver119 <raver119@gmail.com> * enable parallelism by default Signed-off-by: raver119 <raver119@gmail.com> * enable nested parallelism by default Signed-off-by: raver119 <raver119@gmail.com> * Added cuda implementation for row_count helper. * Added implementation for tnse gains op helper. * - take into account possible situations when input arrays are empty in reduce_ cuda stuff Signed-off-by: Yurii <yurii@skymind.io> * Implemented tsne/edge_forces op cuda-based helper. Parallelized cpu-based helper for edge_forces. * Added kernel for tsne/symmetrized op heleper. * Implementation of tsne/symmetrized op cuda helper. Working edition. * Eliminated waste printfs. * Added test for broadcastgradientargs op. * host-only fallback for empty reduce float Signed-off-by: raver119 <raver119@gmail.com> * - some tests fixes Signed-off-by: Yurii <yurii@skymind.io> * - correct the rest of reduce_ stuff Signed-off-by: Yurii <yurii@skymind.io> * - further correction of reduce_ stuff Signed-off-by: Yurii <yurii@skymind.io> * Added test for Cbow op. Also added cuda implementation for cbow helpers. * - improve code of stack operation for scalar case Signed-off-by: Yurii <yurii@skymind.io> * - provide cuda kernel for gatherND operation Signed-off-by: Yurii <yurii@skymind.io> * Implementation of cbow helpers with cuda kernels. * minor tests tweaks Signed-off-by: raver119 <raver119@gmail.com> * minor tests tweaks Signed-off-by: raver119 <raver119@gmail.com> * - further correction of cuda stuff Signed-off-by: Yurii <yurii@skymind.io> * Implementatation of cbow op helper with cuda kernels. Working edition. * Skip random testing for cudablas case. * lstmBlockCell context fix Signed-off-by: raver119 <raver119@gmail.com> * Added tests for ELU and ELU_BP ops. * Added tests for eq_scalar, gt_scalar, gte_scalar and lte_scalar ops. * Added tests for neq_scalar. * Added test for noop. * - further work on clipbynorm_bp Signed-off-by: Yurii <yurii@skymind.io> * - get rid of concat op call, use instead direct concat helper call Signed-off-by: Yurii <yurii@skymind.io> * lstmBlockCell context fix Signed-off-by: raver119 <raver119@gmail.com> * Added tests for lrelu and lrelu_bp. * Added tests for selu and selu_bp. * Fixed lrelu derivative helpers. * - some corrections in lstm Signed-off-by: Yurii <yurii@skymind.io> * operator * result shape fix Signed-off-by: raver119 <raver119@gmail.com> * - correct typo in lstmCell Signed-off-by: Yurii <yurii@skymind.io> * few tests fixed Signed-off-by: raver119 <raver119@gmail.com> * CUDA inverse broadcast bool fix Signed-off-by: raver119 <raver119@gmail.com> * disable MMAP test for CUDA Signed-off-by: raver119 <raver119@gmail.com> * BooleanOp syncToDevice Signed-off-by: raver119 <raver119@gmail.com> * meh Signed-off-by: raver119 <raver119@gmail.com> * additional data types for im2col/col2im Signed-off-by: raver119 <raver119@gmail.com> * Added test for firas_sparse op. * one more RandomBuffer test excluded Signed-off-by: raver119 <raver119@gmail.com> * Added tests for flatten op. * Added test for Floor op. * bunch of tests fixed Signed-off-by: raver119 <raver119@gmail.com> * mmulDot tests fixed Signed-off-by: raver119 <raver119@gmail.com> * more tests fixed Signed-off-by: raver119 <raver119@gmail.com> * Implemented floordiv_bp op and tests. * Fixed scalar case with cuda implementation for bds. * - work on cuda kernel for clip_by_norm backprop op is completed Signed-off-by: Yurii <yurii@skymind.io> * Eliminate cbow crach. * more tests fixed Signed-off-by: raver119 <raver119@gmail.com> * more tests fixed Signed-off-by: raver119 <raver119@gmail.com> * Eliminated abortion with batched nlp test. * more tests fixed Signed-off-by: raver119 <raver119@gmail.com> * Fixed shared flag initializing. * disabled bunch of cpu workspaces tests Signed-off-by: raver119 <raver119@gmail.com> * scalar operators fix: missing registerSpecialUse call Signed-off-by: raver119 <raver119@gmail.com> * Fixed logdet for cuda and tests. * - correct clipBynorm_bp Signed-off-by: Yurii <yurii@skymind.io> * Fixed crop_and_resize shape datatype. * - correct some mmul tests Signed-off-by: Yurii <yurii@skymind.io> * build fix Signed-off-by: raver119 <raver119@gmail.com> * exclude two methods for JNI Signed-off-by: raver119 <raver119@gmail.com> * exclude two methods for JNI Signed-off-by: raver119 <raver119@gmail.com> * exclude two methods for JNI (#97) Signed-off-by: raver119 <raver119@gmail.com> * temporary stack fix Signed-off-by: raver119 <raver119@gmail.com> * round robin affinity test Signed-off-by: raver119 <raver119@gmail.com> * get rid of legacy CudaContext methods Signed-off-by: raver119 <raver119@gmail.com> * get rid of legacy ContextPool classes/methods Signed-off-by: raver119 <raver119@gmail.com> * one legacy test removed Signed-off-by: raver119 <raver119@gmail.com> * few more fields rearranged Signed-off-by: raver119 <raver119@gmail.com> * OpaqueLaunchContext Signed-off-by: raver119 <raver119@gmail.com> * OpaqueLaunchContext++ Signed-off-by: raver119 <raver119@gmail.com> * more of OpaqueLaunchContext methods Signed-off-by: raver119 <raver119@gmail.com> * LaunchContext -> CudaContext Signed-off-by: raver119 <raver119@gmail.com> * AffinityManger changes Signed-off-by: raver119 <raver119@gmail.com> * AffinityManger changes Signed-off-by: raver119 <raver119@gmail.com> * cusolver handles Signed-off-by: raver119 <raver119@gmail.com> * typo Signed-off-by: raver119 <raver119@gmail.com> * cusolver method Signed-off-by: raver119 <raver119@gmail.com> * cusolver handle propagated Signed-off-by: raver119 <raver119@gmail.com> * blas/solver handles Signed-off-by: raver119 <raver119@gmail.com> * one more test Signed-off-by: raver119 <raver119@gmail.com> * legacy concat implementations replaced with new CustomOp Signed-off-by: raver119 <raver119@gmail.com> * one more test Signed-off-by: raver119 <raver119@gmail.com> * concat now uses way more blocks Signed-off-by: raver119 <raver119@gmail.com> * print Signed-off-by: raver119 <raver119@gmail.com> * no more triple template mmul Signed-off-by: raver119 <raver119@gmail.com> * bunch of kernels have dtypes reconsidered Signed-off-by: raver119 <raver119@gmail.com> * bunch of kernels have dtypes reconsidered Signed-off-by: raver119 <raver119@gmail.com> * bitonic sort reorganized Signed-off-by: raver119 <raver119@gmail.com> * bunch of cpu stuff removed from cuda scope Signed-off-by: raver119 <raver119@gmail.com> * bunch of cpu stuff removed from cuda scope Signed-off-by: raver119 <raver119@gmail.com> * type conversions moved to generic impl Signed-off-by: raver119 <raver119@gmail.com> * cpu data types pass Signed-off-by: raver119 <raver119@gmail.com> * non_max_suppression Signed-off-by: raver119 <raver119@gmail.com> * sortByValue fix Signed-off-by: raver119 <raver119@gmail.com> * ignore all mixed datatype tests for mmul Signed-off-by: raver119 <raver119@gmail.com> * special handling of OpProfiler exceptions Signed-off-by: raver119 <raver119@gmail.com> * - one failing concat test in cpp - Nd4j.tile now uses op internally Signed-off-by: raver119 <raver119@gmail.com> * get back dtype exception for legacy arrays deserialization Signed-off-by: raver119 <raver119@gmail.com>
2019-08-14 15:52:34 +02:00
BUILD_SINGLE_TEMPLATE_TWICE(template class functionName, , DATA_TYPES)
cuDNN integration (#150) * initial commit Signed-off-by: raver119 <raver119@gmail.com> * one file Signed-off-by: raver119 <raver119@gmail.com> * few more includes Signed-off-by: raver119 <raver119@gmail.com> * m? Signed-off-by: raver119 <raver119@gmail.com> * const Signed-off-by: raver119 <raver119@gmail.com> * cudnn linkage in tests Signed-off-by: raver119 <raver119@gmail.com> * culibos Signed-off-by: raver119 <raver119@gmail.com> * static reminder Signed-off-by: raver119 <raver119@gmail.com> * platform engine tag Signed-off-by: raver119 <raver119@gmail.com> * HAVE_CUDNN moved to config.h.in Signed-off-by: raver119 <raver119@gmail.com> * include Signed-off-by: raver119 <raver119@gmail.com> * include Signed-off-by: raver119 <raver119@gmail.com> * skip cudnn handle creation if there's not cudnn Signed-off-by: raver119 <raver119@gmail.com> * meh Signed-off-by: raver119 <raver119@gmail.com> * target device in context Signed-off-by: raver119 <raver119@gmail.com> * platform engines Signed-off-by: raver119 <raver119@gmail.com> * platform engines Signed-off-by: raver119 <raver119@gmail.com> * allow multiple -h args Signed-off-by: raver119 <raver119@gmail.com> * allow multiple -h args Signed-off-by: raver119 <raver119@gmail.com> * move mkldnn out of CPU block Signed-off-by: raver119 <raver119@gmail.com> * link to mkldnn on cuda Signed-off-by: raver119 <raver119@gmail.com> * less prints Signed-off-by: raver119 <raver119@gmail.com> * minor tweaks Signed-off-by: raver119 <raver119@gmail.com> * next step Signed-off-by: raver119 <raver119@gmail.com> * conv2d NCHW draft Signed-off-by: raver119 <raver119@gmail.com> * conv2d biasAdd Signed-off-by: raver119 <raver119@gmail.com> * test for MKL/CUDNN combined use Signed-off-by: raver119 <raver119@gmail.com> * - provide additional code for conv2d ff based on cudnn api, not tested yet Signed-off-by: Yurii <iuriish@yahoo.com> * - further work on conv2d helper based on using cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - fixing several cuda bugs which appeared after cudnn lib had been started to use Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of conv2d backprop op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - implementaion of conv3d and conv3d_bp ops based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - bugs fixing in conv3d/conv3d_bp ops (cudnn in use) Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of depthwiseConv2d (ff/bp) op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - implementation of batchnorm ff op based on cudnn api Signed-off-by: Yurii <iuriish@yahoo.com> * - disable cudnn batchnorm temporary Signed-off-by: Yurii <iuriish@yahoo.com> * - add minor change in cmake Signed-off-by: Yurii <iuriish@yahoo.com> * engine for depthwise mkldnn Signed-off-by: raver119 <raver119@gmail.com> * couple of includes Signed-off-by: raver119 <raver119@gmail.com> * - provide permutation to cudnn batchnorm ff when format is NHWC Signed-off-by: Yurii <iuriish@yahoo.com> * lgamma fix Signed-off-by: raver119 <raver119@gmail.com> * - eliminate memory leak in two tests Signed-off-by: Yurii <iuriish@yahoo.com> Co-authored-by: Yurii Shyrma <iuriish@yahoo.com>
2020-01-20 19:32:46 +01:00
*/
DECLARE_PLATFORM(conv2d, ENGINE_CPU)
2019-06-06 14:21:15 +02:00
//BUILD_PAIRWISE_SELECTOR(xType, yType, zType, functionName, (signature), DATA_TYPES, Y_TYPES);
//BUILD_SINGLE_UNCHAINED_TEMPLATE(functionName , (signature), Y_TYPES);
//BUILD_TRIPLE_SELECTOR(xType, yType, zType, functionName, (signature), DATA_TYPES, Y_TYPES, Z_TYPES)
//BUILD_TRIPLE_TEMPLATE(functionName, (signature), DATA_TYPES, Y_TYPES, Z_TYPES)
//BUILD_ENUMERATION(DATA_TYPES)
//BUILD_SINGLE_SELECTOR(xType, functions::IndexReduce, ::op(a, b, c, d, e), DATA_TYPES)
//BUILD_DOUBLE_SELECTOR(xType, yType, functions::IndexReduce, ::op(a, b, c, d, e), DATA_TYPES, DATA_TYPES)
//BUILD_SINGLE_TEMPLATE(template class Alpha, (signature), DATA_TYPES);
//BUILD_DOUBLE_TEMPLATE(template class Alpha, (signature) , DATA_TYPES, DATA_TYPES);
/*
#define SCALAR_OPS \
(0, simdOps::Identity) ,\
(1, simdOps::ReLU)
*/
/*
#define NATIVE_LAYERS \
(0, sd::layers::DenseLayer)
// (1, sd::layers::ConvolutionLayer) ,\
// (2, sd::layers::Pooling2DLayer) ,\
// (3, sd::layers::LSTMLayer)
2019-06-06 14:21:15 +02:00
*/
/*
#define PAIRWISE_TRANSFORM_OPS \
(0, simdOps::Add),\
(1, simdOps::Copy),\
(2, simdOps::Divide),\
(3, simdOps::EqualTo),\
(4, simdOps::GreaterThan),\
(5, simdOps::LessThan),\
(6, simdOps::Multiply),\
(7, simdOps::Pow),\
(8, simdOps::ReverseSubtract),\
(9, simdOps::Subtract),\
(10,simdOps::Epsilon),\
(11,simdOps::GreaterThanOrEqual),\
(12,simdOps::LessThanOrEqual),\
(13,simdOps::Max),\
(14,simdOps::Min),\
(15,simdOps::NotEqualTo),\
(16,simdOps::Copy),\
(17,simdOps::Axpy),\
(18,simdOps::ReverseDivide),\
(45,simdOps::CompareAndSet),\
(46,simdOps::CompareAndReplace),\
(56,simdOps::And),\
(57,simdOps::Or),\
(58,simdOps::Xor),\
(59,simdOps::Remainder),\
(60,simdOps::FMod),\
(69,simdOps::Atan2)
EXECUTE_NOE((x, y, extras), OPS_A(PAIRWISE_TRANSFORM_OPS))
*/
//EXECUTE_NOE((x, extras), OPS_A(SCALAR_OPS))
//BUILD_CALL_1(template void sd::NDArray<float16>::applyTransform, float16, (NDArray<float16>* a, float16* b), TRANSFORM_OPS)
2019-06-06 14:21:15 +02:00
//BUILD_CALL_1(template void sd::NDArray<float16>::applyPairwiseTransform, float16, (NDArray<float16>* other, float16* extraParams), PAIRWISE_TRANSFORM_OPS)
2019-06-06 14:21:15 +02:00
//BUILD_TRACKER(TRANSFORM, ACTIVATIONS)
//BUILD_CALL_1(template void sd::NDArray<float16>::applyScalar, float16, (float16 scalar, NDArray<float16>* target, float16 *extraParams) , ACTIVATIONS);
2019-06-06 14:21:15 +02:00
/*
#define DECLARE_OP(NAME, NIN, NOUT) DECLARE_OP_UNIQ(__COUNTER__, NAME, NIN, NOUT)
#define DECLARE_OP_UNIQ(CTR, NAME, NIN, NOUT) template <typename T> \
class NAME: public sd::ops::DeclarableOp<T> { \
2019-06-06 14:21:15 +02:00
public:\
NAME() : sd::ops::DeclarableOp<T>(NIN, NOUT, #NAME) { } \
2019-06-06 14:21:15 +02:00
protected: \
Nd4jStatus validateAndExecute(Block<T>& block); \
};\
template <typename T> \
Nd4jStatus sd::ops::NAME<T>::validateAndExecute(Block<T>& block)
2019-06-06 14:21:15 +02:00
*/
//#define END_OP(NAME) }; static sd::ops::__registrator<NAME<float>> register_op##Name;
2019-06-06 14:21:15 +02:00
//#DECLARE_OP(Concat, -1, 1)
//END_OP(Concat)
//BUILD_LAYERS_FACTORY(float, OPS_A(NATIVE_LAYERS), OPS_B(ACTIVATIONS))
//DISPATCH_SIMPLE(scalarAlongDimension_, float, PARAMS(x, xShapeInfo, extraParamx, z, zShapeInfo, scalars, tadShapeInfo, tadOffsets, tadShapeInfoZ, tadOffsetsZ), OPS_A(SCALAR_OPS))
//_EXEC_KERNEL_F(scalarAlongDimension_, scalarAlongDimensionGeneric, float, (float inputA, float inputB), (paramA, paramB), (10, SCALAR::Add), (11, SCALAR::Subtract), (12, SCALAR::Multiply))
//DISPATCH_KERNEL_SIMPLE(scalarAlongDimension_, scalarAlongDimensionGeneric, float, INPUT(float inputA, float inputB), PARAMS(paramA, paramB), OPS_A(SCALAR_OPS))
// original version
// DISPATCH_METAOP(functions::pairwise_transforms::PairWiseTransform<T>::template transformCuda, PARAMS(N, dx, dy, xStride, yStride, paramsPtr, dz, zStride, nullptr, nullptr, nullptr), InvertedMetaOp, OPS_A(SCALAR_OPS), OPS_B(PAIRWISE_TRANSFORM_OPS))
/*
DISPATCH_METAOP(invertedMetaPairwiseShaped_Pairwise_Scalar, PARAMS(opTypeA, opTypeB, N, x, xShape, y, yShape, z, zShape, extrasA, extrasB, scalarA, scalarB), float, OPS_A(PAIRWISE_TRANSFORM_OPS), OPS_B(SCALAR_OPS));*/
//DISPATCH_KERNEL_META(invertedMetaPairwiseShaped_Pairwise_Scalar_, invertedMetaPairwiseShapedGeneric, float, simdOps::InvertedMetaOp, INPUT(const int opTypeA, const int opTypeB, long N, float *dx, int *xShapeInfo, float *dy, int *yShapeInfo, float *dz, int *zShapeInfo, float *extraA, float *extraB, float scalarA, float scalarB), PARAMS(opTypeA, opTypeB, N, dx, xShapeInfo, dy, yShapeInfo, dz, zShapeInfo, extraA, extraB, scalarA, scalarB), OPS_A(PAIRWISE_TRANSFORM_OPS), OPS_B(SCALAR_OPS))
//_EXPAND_KERNEL_CALL(invertedMetaPairwiseShaped_Pairwise_Scalar_, invertedMetaPairwiseShapedGeneric, float, simdOps::InvertedMetaOp, INPUT(const int opTypeA, const int opTypeB, long N, float *dx, int *xShapeInfo, float *dy, int *yShapeInfo, float *dz, int *zShapeInfo, float *extraA, float *extraB, float scalarA, float scalarB), PARAMS(N, dx, dy, xStride, yStride, paramsPtr, dz, zStride, nullptr, nullptr, nullptr), 66, simdOps::SomeOpA, 99, simdOps::SomeOpB)
/*
extern "C" __global__ void invertedMetaOpKernel_Pairwise_Scalar_16_1_float(const int opTypeA, const int opTypeB, long N, float *dx, int *xShapeInfo, float *dy, int *yShapeInfo, float *dz, int *zShapeInfo, float *extraA, float *extraB, float scalarA, float scalarB) {
invertedMetaPairwiseShapedGeneric<float, simdOps::InvertedMetaOp<float, simdOps::Copy<float>, simdOps::Multiply<float>>>(opTypeA, opTypeB, N, dx, xShapeInfo, dy, yShapeInfo, dz, zShapeInfo, extraA, extraB, scalarA, scalarB);
}
*/
#endif //LIBND4J_PLAY_H