54 Commits (ae0f88a988a1b1e53168a6108484d0cfdcb58003)

Author	SHA1	Message	Date
Jacek Czaja	9eff1a674f	Added missing format of oneDNN (#29670)	5 years ago
Jacek Czaja	f6cca62575	[oneDNN] Making ThreadID info in caching key optional (#29272)	5 years ago
jakpiase	57a4f16d9e	added internal and external reorders to profiler (#29443) ... * added external reorder to profiler * added external and internal reorders to profiler * added internal and external reorder to profiler * added formatting to int/ext reorder commit * removed unnecessary comment	5 years ago
arlesniak	bc902044a4	Fixes mkldnn dygraph learning rate scheduler crashes (#28988)	5 years ago
Jacek Czaja	bd1d6d3b30	extends oneDNN caching keys so caching objects are unique to executor/predictor (#28758)	5 years ago
Jacek Czaja	c11d9b3035	[oneDNN ] conv2d fwd&bwd optimization (#27871)	5 years ago
Jacek Czaja	b9fda2ff09	Fix to issue #25537 (#27546) ... * - condidate fix to issue #25537 test=develop * - UT for transpose NHWC test=develop	5 years ago
joanna.wozna.intel	1483ea2304	Add bfloat16 passes (#26999)	5 years ago
joanna.wozna.intel	95e1434bb2	Add bfloat16 data type (#25402)	5 years ago
Adam	f3909020de	Add mechanism for blocking oneDNN cache clearing (#26502) ... * Add mechanism for blocking oneDNN cache clearing * Review changes and Add thread guards	5 years ago
joanna.wozna.intel	734cf1c3e9	Change use_quantizer attribute name and data type (#25838) ... * Change use_quantizer attribute name and data type * Fix problem with setting attribute * Add changes due to review * Small change in function * Restore use_quantizer attr for compatibility	5 years ago
Adam	68c6160e63	Add oneDNN fusion_gru kernel (#25594) ... * Add oneDNN fusion_gru kernel and fix fc+gru pass test=develop * Formatting changes test=develop * Lint fixes test=develop * Add memory::format_tag::any to GRU weights test=develop * Fix build with CUDA * Fix build with CUDA v2	5 years ago
arlesniak	e52df3b125	Added DNNL cache management for DyGraph (#25624) ... * Added DNNL cache management for DyGraph * move FLAGS_use_mkldnn to more general CMakeLists, getu use of the flag in ClearGradients * missing file * Fixes after review * Bringing back original idea of place for 'use_mkldnn' flag to be accessible from platform nad imperative. * Removed duplicate and added docs * Fixes for CI	5 years ago
Jacek Czaja	a5d1592f6c	Added missing oneDNN format (#25450) ... test=develop	5 years ago
Jacek Czaja	c6c65c65c7	[DNNL] Added elementwise_add mkl-dnn inplace (#23477)	5 years ago
Michał Gallus	a63bcf9ae7	[DNNL][INT8][FP32] MatMul (#23395) ... * Initial FP32 DNNL MatMul Implementation * Implement int8 DNNL MatMul * Unify in-kernel-naming, clean UTs * MatmuL: Introduce op caching * Final adjustments test=develop * Remove dy_graph disablement test=develop * Change dnnl header name to new one test=develop * Contrain multi head check to prevent fails test=develop * Resolve dnnl header problems on MAC CI * Variable namings to kernel and skip_grad_ci added test=develop * Prevent MAC CI from failing * Prevent windows build from failing test=develop * Modify UTs to conform to the rules * Modify MatMul aux functions namings test=develop	5 years ago
bingyanghuang	7872d06ff4	Add explanation on conv grad for dims<3 (#22125)	6 years ago
Jacek Czaja	b0b27ff699	[MKL-DNN] Conv grad and Batch Norm grad NHWC support (#22088)	6 years ago
Jacek Czaja	ad8a9cb82c	[MKL-DNN] Pool & LRN Grad Ops NHWC support (#21747)	6 years ago
Adam	e81f0228df	MKL-DNN 1.0 Update (#20162) ... * MKLDNN v1.0 rebase to Paddle 1.6 test=develop * Add hacky paddle::string::to_string() implementation * vectorize<int64-t>() -> vectorize() cleanup test=develop * PADDLE_ENFORCE and void_cast fixes test=develop * Rebase changes test=develop * Cosmetics test=develop * Delete MKL from mkldnn.cmake test=develop * CMake debug commands test=develop * Delete MKLDNN_VERBOSE and rebase fixes test=develop * Rebase fixes test=develop * Temporarily disable int8 resnet101 vgg16 and vgg19 tests test=develop * Add libmkldnn.so.1 to python setup test=develop * Add libmkldnn.so.1 to inference_lib cmake after rebase test=develop * Post rebase fixes + FC int8 changes test=develop * Fix LRN NHWC test=develop * Fix NHWC conv3d test=develop * Windows build fix + next conv3d fix test=develop * Fix conv2d on AVX2 machines test=develop	6 years ago
Adam	3fda695bb0	Add support for asymetric padding in MKLDNN pool, conv and conv_transpose (#21062) ... * Add asymetric padding support for mkldnn pooling test=develop * Add asymetric padding support for mkldnn conv test=develop * Add asymetric padding support for mkldnn conv_transpose test=develop	6 years ago
Adam	67b59ddb38	Minor MKL-DNN conv int8 performance fixes (#20753) ... test=develop	6 years ago
Adam	dfdd73cbc0	Add MKLDNNhandlerT templatized class (#19801) ... test=develop	6 years ago
Adam	d4413a54bc	Add common CreateKey for mkldnn handlers (#19767) ... test=develop	6 years ago
Adam	428b2b9e17	MKLDNN handler cleanup (#19713) ... * MKLDNN handler cleanup * MKLDNN handler cleanup test=develop	6 years ago
Adam	e94b26daf5	using MKLDNNMemoryFormat = mkldnn::memory::format changes (#19568) ... * using MKLDNNMemoryFormat = mkldnn::memory::format changes test=develop * PADDLE_ENFORCE update test=develop	6 years ago
Jacek Czaja	ecd9f330c9	[MKL-DNN] Fix to face model on AVX512 platforms (#19282) ... - Refactor step 1 - Compilation fix - Yet another compilation fix - Even more compilation fix - Lint fixes test=develop - Removed deprectaed PADDLE_ENFORCE occurance test=develop - Candidate fix to BN forward - Lint fixes test=develop - Refactoring in data_layout_transform - compilation fix - Another comppilation fix - Step further into darkness - Yet another compilation fix - Yet another compilation fix - missing header - compilation fix - Added MKLDNN -> Paddle conversion in fetch op test=develop - Compilation fix test=develop - Lint test=develop - Mul fix - Fix to MKLDNN MUL op and Elementwise MUL UT test=develop - Workaround for diffrent weights with groups representation Paddle vs MKL-DNN. test=develop - Candidate fix for 5D convolution with groups - Refactor of fix for conv3d and conv2d in fetch op test=develop - Compilation fix - Still same compilation fix - Compilation fix - Compilation fix - Reverted refactoring of fixes - Adapted test_conv2d_int8_mkldnn so it exects data in NCHW format not NHWC test=develop - minor fix in UT test=develop - Lint fixes test=develop	6 years ago
lidanqing	9ecd8ee789	change ComputeINT8 to template version to remove checking dst_datatype code (#18756) ... * change INT8 to template so that checking dst_dt with if-else could be removed. CI will be enabled after fixing reviews * reverse user_residual_memory_p and user_bias_memory_p declaration scope test=develop	6 years ago
Physher	0caa08ea40	Add mkldnn int8 mul-op kernel (#17834)	6 years ago
Yihua Xu	669191c9cc	Implement conv3d with mkldnn library (test=develop)	7 years ago
Tao Luo	20120d9c97	Merge pull request #14608 from jczaja/prv-conv2d-transpose-mkldnn ... [MKL-DNN]conv2d transpose	7 years ago
Michal Gallus	9455be0ba5	EltwiseMul: Extract StringToFormat to MKLDNN helper ... test=develop	7 years ago
Jacek Czaja	fb24690a58	- conv2d transpose MKL-DNN ... test=develop - Added new header for MKLDNN reuse functionality - Extended conv2d_transpose GetExpectedKernelType for MKL-DNN supporrt - Buildable conv transpose mkldnn and conv mkldnn using conv template - Conv2d transpose roughlt implemented and buildable - Added modifications conv2d transpose MKLDNN unit tests - Fix to UT of conv2d transpose mkldnn op - Wrong type of MKLDNN primitive was chosen for conv2d transpose - HAcks for conv2d transpose - UT enalbed - Replaced copying loop with memcpy - Draft of passing lambda into AcquireMemory - Made reorder (IOHW->OIHW) to be called only once	7 years ago
Tomasz Patejko	8899d42265	MKLDNN conv residual data: primitive reuse interface used. Reorder done when formats are different ... test=develop	7 years ago
Krzysztof Binias	accdecc681	Correcting Lint errors	7 years ago
Krzysztof Binias	1ce9e9dc30	Renaming decision variable	7 years ago
Krzysztof Binias	1658958fe6	Reusing converted weights	7 years ago
Michał Gallus	cd32ddac12	Fuse Convolution and Eltwise Add into MKLDNN's Conv+Bias (#12669) ... * Fuse Convolution and Eltwise Add into Conv+Bias * Reduce bias branching at conv_mkldnn_op * Add MKLDNN build checks for Conv Bias * Conv-bias: check if bias input exist befor assignment * Conv-bias: Remove Bias dim check from infershape It was causing conv3d test to crash upon\ncalling HasInput(Bias)	7 years ago
Sylwester Fraczek	d74bb6ab9c	fix ut for mkldnn 0.15 - added forcing layout NCHW in mkldnn conv tests	7 years ago
Jacek Czaja	fbe25ef510	MKLDNN: Extending Conv MKLDNN op to reuse MKLDNN primitives (#11750) ... * - Rebase of conv reuse - clag formatter fixes - Fix to conv reuse - Yet another fix - Fix - Fix - clagn format * - comment update	7 years ago
gongweibao	e2b1c5d925	fix code style (#11862)	7 years ago
mozga-intel	b8a04c2fa1	Duplicated code was moved to common function	7 years ago
Jacek Czaja	98f3ad3ba1	- MKLDNN Softmax Grad Op ... - Added hash function inside of MKLDNN softmax op to be used as handle for primitives stroing in a context - Style fixes to softmax mkldnn op - Fixes after review - Coding style - Fix to style - style fixes - style fix - style fixes - Fix to cody style check - Rephrasing a comment fix t obroken merge Fixes to rebase Conflicts: benchmark/fluid/models/machine_translation.py cmake/external/mkldnn.cmake paddle/fluid/operators/softmax_mkldnn_op.cc - Bumped revision of MKL-DNN up to have softmax backward primitive - Added choosing MKLDNN softmax grad operator - First reuse of softmax backward - Reinvented reusing for softmax - Fix to crash in reinvented reuse - Clang format fixes - Clang format fixes - Improved softmax mkldnn reuse mechanism - clang format fixes - Fix to broken merge - Fix	7 years ago
tensor-tang	d5fb8fa778	Revert "Merge pull request #11628 from PaddlePaddle/revert-11102-mozga-intel/Sum_mkldnn_layout" ... This reverts commit 4d8e8ee226, reversing changes made to d6a9f005c8.	7 years ago
tensor-tang	90780e22ce	Revert "MKLDNN layout: Support for sum operator"	7 years ago
mozga-intel	6512be59ec	MKLDNN layout: the code-review changes	7 years ago
mozga-intel	3ff9ba0e6b	Mkldnn layout (#11040) ... * Add MKLDNN layout support in Paddle Add MKLDNN layout in Paddle so that MKLDNN friendly memory layout can be used in MKLDNN enabled OP kernel. Before this commit, NCHW is hardcode to be used in all MKLDNN op kernels. As a result, non-optimized execution path is selected in MKLDNN primitive which bring worse performance. Besides framework change, three MKLDNN OP kernels were updated for using new MKLDNN layout. They are conv/pool2d/batch_norm. Other MKLDNN OP kernels need be also updated in similar way to achieve best performance. * Add MKLDNN layout support in activation OP * Don't populate layout from input to output when kMKLDNN in * Refine pool mkldnn op kernel * MKLDNN layout * Remove the inferitance from tensor file * MKLDNN layout: refactoring * Remove additional #define to register new operator * Prepare mkldnn tests to work with layout	7 years ago
Krzysztof Binias	0aa01929c1	Add backward	7 years ago
Jacek Czaja	5f1333058c	- Draft of reuse of pooling mkldnn operator ... - Finished draft of pooling reusing of operators - Using gethash in PoolGrad added - Removed diagnostic - Added pool mkldnn grad reusing of primitives - Added diagnostic - Removed diagnostic - added dependency to mkldnn data type for pooling mkldnn - Added mkldnn memory data type determining based on template type of op - Compilation warning fix - codying style fixes	7 years ago
mozga-intel	6e7b883bdd	Initial implementation of multiplication operator for MKLDNN	7 years ago