platform/upstream/dldt.git
4 years agoTensorIterator to RNN/GRU/LSTM Sequence transformation (#2146)
Ivan Tikhonov [Tue, 15 Sep 2020 07:11:51 +0000 (10:11 +0300)]
TensorIterator to RNN/GRU/LSTM Sequence transformation (#2146)

* ti to sequences transformations

* fix sequences to sequences ie conversion

* resolve review marks

* resolve review remarks, fix ti to sequences transformations to support batch > 1 if slice axis == 0

* temporary enable ngraph ti transformations for cpu plugin

* fix includes

* Revert "fix includes"

This reverts commit 6cf15b97bec94168df71606ac959b3cf60634c8a.

* Revert "temporary enable ngraph ti transformations for cpu plugin"

This reverts commit fd528d7216a2e6b11e915a0038a191259dc60551.

* delete todo comments

4 years ago[LPT] Copy constant with several outputs before blob update (cherry-pick to master...
Edward Shogulin [Tue, 15 Sep 2020 06:18:58 +0000 (09:18 +0300)]
[LPT] Copy constant with several outputs before blob update (cherry-pick to master) (#2198)

* [LPT] Copy constant implementation

* [LPT] the same Constant ops as FQ interval boundaries

4 years agosetupvars: update logic for setting path-like vars in case if they are empty. Updated...
Artyom Anokhov [Mon, 14 Sep 2020 16:49:57 +0000 (19:49 +0300)]
setupvars: update logic for setting path-like vars in case if they are empty. Updated copyrights. (#2229)

4 years agoGatherTree description was extended and outdated link fixed (#2167)
Svetlana Dolinina [Mon, 14 Sep 2020 16:49:29 +0000 (19:49 +0300)]
GatherTree description was extended and outdated link fixed (#2167)

* add more alrifications to description

* move clarification to comment

* pseudo code become more accurate

* review changes

4 years agoRevert "update OpenCV version to 4.5.0"
Alexey Suhov [Mon, 14 Sep 2020 15:47:19 +0000 (18:47 +0300)]
Revert "update OpenCV version to 4.5.0"

This reverts commit 5d229a45649091d25c3e85f391f99ea88122e008.

4 years agoupdate OpenCV version to 4.5.0
Alexey Suhov [Mon, 14 Sep 2020 14:58:39 +0000 (17:58 +0300)]
update OpenCV version to 4.5.0

4 years ago[CPU] add check on equality of values IL and IH ranges (#2190)
Maxim Andronov [Mon, 14 Sep 2020 12:28:37 +0000 (15:28 +0300)]
[CPU] add check on equality of values IL and IH ranges (#2190)

4 years agoAdd backend tests to ONNX OpenVino CI (#1972)
Rafal Blaczkowski [Mon, 14 Sep 2020 11:07:47 +0000 (13:07 +0200)]
Add backend tests  to ONNX OpenVino CI (#1972)

4 years agoDefine security policy (#2215)
Andrey Somsikov [Mon, 14 Sep 2020 10:50:03 +0000 (13:50 +0300)]
Define security policy (#2215)

4 years ago[Python Tools] Fix several problems in cross-check-tool (#2170)
Anastasia Kuporosova [Mon, 14 Sep 2020 10:49:01 +0000 (13:49 +0300)]
[Python Tools] Fix several problems in cross-check-tool (#2170)

4 years agoSuper smart reshape: HC Reshape to 2D followed by MatMul (#2183)
Evgenya Stepyreva [Mon, 14 Sep 2020 10:45:27 +0000 (13:45 +0300)]
Super smart reshape: HC Reshape to 2D followed by MatMul (#2183)

* Initial commit

* [SSR] Reshape(2D)->MatMul constrain relaxation

* Moved common pattern mechanics to the common function

* Moving SmartReshape to CNNNetworkNgraphImpl ctors

* Review comment

* Tests

4 years agoSplit time-tests common library (#2173)
Andrey Somsikov [Mon, 14 Sep 2020 06:04:49 +0000 (09:04 +0300)]
Split time-tests common library (#2173)

Split time-tests common library

Add a README.md with workflow description.
Defined "timetest_" suffix for all time tests.
Applied clang-format-9 and added a README.md

Co-authored-by: Alina Alborova <alina.alborova@intel.com>
4 years ago[DOC] Reshape feature (#2023)
Evgenya Stepyreva [Fri, 11 Sep 2020 18:41:42 +0000 (21:41 +0300)]
[DOC] Reshape feature (#2023)

* [DOC] Reshape feature

* Comments adressed.

* Clarifications

* Converting_Model_General.md

* Update ShapeInference.md

* Update ShapeInference.md

* Comments

* Update Convert_Object_Detection_API_Models.md

Co-authored-by: Andrey Zaytsev <andrey.zaytsev@intel.com>
4 years agoAdded type info for transformations (#2116)
Gleb Kazantaev [Fri, 11 Sep 2020 12:38:43 +0000 (15:38 +0300)]
Added type info for transformations (#2116)

4 years agoFix Mish and SoftPlus value propagation functions (#2120)
Anton Chetverikov [Fri, 11 Sep 2020 09:58:14 +0000 (12:58 +0300)]
Fix Mish and SoftPlus value propagation functions (#2120)

* Fix Mish and SoftPlus value propagation functions

* Add unit tests for SoftPlus & Mish operations value propagation functions

4 years ago[PYTHON API TESTS] Add ngraph tests (#2132)
Anastasia Kuporosova [Fri, 11 Sep 2020 09:56:57 +0000 (12:56 +0300)]
[PYTHON API TESTS] Add ngraph tests (#2132)

Co-authored-by: Alexander Zhogov <alexander.zhogov@intel.com>
4 years agoFix for static PartialShape detection algorithm (#2106)
Vladislav Volkov [Fri, 11 Sep 2020 03:15:24 +0000 (06:15 +0300)]
Fix for static PartialShape detection algorithm (#2106)

4 years agoAdded callback to disable PriorBox to PriorBoxIE transformation (#2159)
Ivan Tikhonov [Fri, 11 Sep 2020 03:12:14 +0000 (06:12 +0300)]
Added callback to disable PriorBox to PriorBoxIE transformation (#2159)

* move priorbox to ie transformations to Opset1ToLegacyOpset pipeline

* fix typo

* Revert "fix typo"

This reverts commit 4077a78cbdcc1340b0999a3756ce348dd567e45f.

* Revert "move priorbox to ie transformations to Opset1ToLegacyOpset pipeline"

This reverts commit 910e41ff2099b0516fc9bff2baaa2ec678c8a7bf.

* add functionality to disable prior box to ie transformation

* fix callback

4 years ago[IE CLDNN] Memory allocation optimizations (#2007)
Sergey Shlyapnikov [Thu, 10 Sep 2020 20:18:26 +0000 (23:18 +0300)]
[IE CLDNN] Memory allocation optimizations (#2007)

4 years agoAdd exposing function signatures via Cython (#2024)
Evgeny Talanin [Thu, 10 Sep 2020 16:58:45 +0000 (19:58 +0300)]
Add exposing function signatures via Cython (#2024)

4 years agoImplement `run_executable.py` to run `TimeTests` several times (#2125)
Vitaliy Urusovskij [Thu, 10 Sep 2020 16:41:23 +0000 (19:41 +0300)]
Implement `run_executable.py` to run `TimeTests` several times (#2125)

4 years ago[IE CLDNN] Fix stack-use-after-scope issue (#2150)
Sergey Shlyapnikov [Thu, 10 Sep 2020 15:59:15 +0000 (18:59 +0300)]
[IE CLDNN] Fix stack-use-after-scope issue (#2150)

4 years agoUpdated ConvertPrecision pass not to make extra copy of Constant data (#2108)
Gleb Kazantaev [Thu, 10 Sep 2020 14:11:10 +0000 (17:11 +0300)]
Updated ConvertPrecision pass not to make extra copy of Constant data (#2108)

4 years ago[samples] remove unused ngraph (#2156)
Kate Generalova [Thu, 10 Sep 2020 13:04:13 +0000 (16:04 +0300)]
[samples] remove unused ngraph (#2156)

4 years ago[IE Myriad][#38489] Reduce test scope for Swish layer (#2163)
Nikita Kudriavtsev [Thu, 10 Sep 2020 12:47:20 +0000 (15:47 +0300)]
[IE Myriad][#38489] Reduce test scope for Swish layer (#2163)

4 years agoFixed differences between glib versions (#2153)
Alexey Fadeev [Thu, 10 Sep 2020 12:11:08 +0000 (15:11 +0300)]
Fixed differences between glib versions (#2153)

4 years ago[IE][VPU]: Support ngraph::Function in VPU QueryNetwork (#1929)
Anton Dudchenko [Thu, 10 Sep 2020 11:49:19 +0000 (14:49 +0300)]
[IE][VPU]: Support ngraph::Function in VPU QueryNetwork (#1929)

* Support ngraph::Function in VPU QueryNetwork

4 years ago[IE Myriad][IE Tests] Activation layer's constants parametrization. (#2071)
Nikita Kudriavtsev [Thu, 10 Sep 2020 09:56:21 +0000 (12:56 +0300)]
[IE Myriad][IE Tests] Activation layer's constants parametrization. (#2071)

CI passed: https://gitlab-icv.inn.intel.com/inference-engine/product-configs/merge_requests/870

4 years agofix #38545 remove CPU network layer check for py samples (#2145)
Kate Generalova [Thu, 10 Sep 2020 09:44:13 +0000 (12:44 +0300)]
fix #38545 remove CPU network layer check for py samples (#2145)

4 years agoAdd support for custom onnx operators: DetectionOutput, Normalize and… (#2064)
Mateusz Tabaka [Thu, 10 Sep 2020 09:27:12 +0000 (11:27 +0200)]
Add support for custom onnx operators: DetectionOutput, Normalize and… (#2064)

4 years ago[IE][VPU][Custom CL]: Fix binary convolution3x3 3d transaction (#2144)
Maxim Kurin [Thu, 10 Sep 2020 08:55:32 +0000 (11:55 +0300)]
[IE][VPU][Custom CL]: Fix binary convolution3x3 3d transaction (#2144)

* Fix binary_convolution3x3.cl kernel & test

4 years ago[IE CLDNN] Fully connected MMAD kernel optimizations (#2115)
Ilya Znamenskiy [Thu, 10 Sep 2020 05:56:04 +0000 (08:56 +0300)]
[IE CLDNN] Fully connected MMAD kernel optimizations (#2115)

4 years ago[CPU] statically analyzed issues. (#2139)
Nikolay Shchegolev [Thu, 10 Sep 2020 05:35:32 +0000 (08:35 +0300)]
[CPU] statically analyzed issues. (#2139)

4 years ago[CPU] Supported depthwise 6d, 7d, ..., added test (#971)
Anton Voronov [Thu, 10 Sep 2020 05:33:38 +0000 (08:33 +0300)]
[CPU] Supported depthwise 6d, 7d, ..., added test (#971)

4 years agoFixed addOutput behavior for experimental ops (#2138)
Ilya Churaev [Thu, 10 Sep 2020 03:14:11 +0000 (06:14 +0300)]
Fixed addOutput behavior for experimental ops (#2138)

4 years ago[CPU] Fixed quantization post op memory leak (#2101)
Gorokhov Dmitriy [Wed, 9 Sep 2020 15:59:37 +0000 (18:59 +0300)]
[CPU] Fixed quantization post op memory leak (#2101)

4 years ago[CPU] add const and precision check quantize ranges (#2074)
Maxim Andronov [Wed, 9 Sep 2020 15:57:58 +0000 (18:57 +0300)]
[CPU] add const and precision check quantize ranges (#2074)

4 years ago[ IE ] ExperimentalDetectron shape infer fix (#2143)
Evgenya Stepyreva [Wed, 9 Sep 2020 15:25:53 +0000 (18:25 +0300)]
[ IE ] ExperimentalDetectron shape infer fix (#2143)

4 years ago[JAVA] Code style check added (#1984)
Anna Likholat [Wed, 9 Sep 2020 14:49:23 +0000 (17:49 +0300)]
[JAVA] Code style check added (#1984)

4 years ago[IE CLDNN] Fix problems with loop iterator and parameter check in clDNN (#2141)
Mikhail Letavin [Wed, 9 Sep 2020 14:21:10 +0000 (17:21 +0300)]
[IE CLDNN] Fix problems with loop iterator and parameter check in clDNN (#2141)

4 years ago[IE][TOOLS] compile_tool improvements (#2140)
Vladislav Vinogradov [Wed, 9 Sep 2020 14:14:20 +0000 (17:14 +0300)]
[IE][TOOLS] compile_tool improvements (#2140)

* Remove dead code.
* Protect device specific config options with device checks.
* Add missing space to precision parsing error message.
* Allow to switch FP32 input precision to U8.

4 years agoRelaxed tolerance for fp16_inception_v1 model (#2129)
Katarzyna Mitrus [Wed, 9 Sep 2020 13:47:53 +0000 (15:47 +0200)]
Relaxed tolerance for fp16_inception_v1 model (#2129)

4 years agoExtend MO for support of Interpolate-4 (#2026)
Vladimir Gavrilov [Wed, 9 Sep 2020 13:28:52 +0000 (16:28 +0300)]
Extend MO for support of Interpolate-4 (#2026)

* Commit.

* Added opset4 version in the class Interpolate.

* Added class ONNXResize11Op to read ONNX Resize with opset version >= 11.

* Added support for Interpolate-4 into transformations TestInterpolateReshapeWA and InterpolateConcat.

* Added support for Interpolate-4 into transformation InterpolateWithConcat.

* Deleted redundant checks from the transformation UpsampleToResample.

* Reverted last changes.

* Changed ONNX Resize extractor to support for Interpolate-4.

* Added conversion of ONNXResize11Op into Interpolate-4.

* Added support for Interpolate-4 into the transformation InterpolateSequenceToInterpolate.

* Small fix for formatting.

* Written tests for MO version of Interpolate-4 with shape_calculation_mode = sizes.

* Written tests for infer function of Interpolate-4.

* Now transformations InterpolateWithConcat, InterpolateConcat, InterpolateReshapeWA skip Interpolate-4.

* Used create_op_with_const_inputs in the transformation InterpolateSequenceToInterpolate.

* The transformation ONNXResize11ToInterpolate4 was rewritten using find_and_replace_pattern.

* Now the dictionary infers (dictionary of infer functions of Interpolate) is a class static attribute.

* Deleted unused variable.

* Restored original logic of find_and_replace_pattern method of the class InterpolateReshapeWA.

* Used create_op_with_const_inputs() in the transformation InterpolateSequenceToInterpolate for opset1 case.

* Replaced resize_name by resize.soft_get('name', resize.id).

* Small fixes.

* Added two tests for Interpolate-4 infer function.

* Fixed the transformation ONNXResize11ToInterpolateV4 for the case when ONNXResize11 operation has 3 inputs.

* Added conversion of ONNXResize11 with tf_crop_and_resize_mode to ROIPooling + ONNXResize11.

* Fixed bugs in the transformation ONNXResize11ToInterpolateV4 and in the infer function of the operation ONNXResize11.

* Small changes.

* Renamed transformation that converts ONNXResize11 into ROIPooling + ONNXResize11 and fixed BOM-file.

* Fixed tests for the transformation InterpolateSequenceToInterpolate.

* Small change.

* Now the transformation InterpolateSequenceToInterpolate preserves output layer name.

* Deleted the transformation ONNXResize11ToTFCropAndResize.

4 years agoImplement statistics collection: (#2056)
Vitaliy Urusovskij [Wed, 9 Sep 2020 12:30:23 +0000 (15:30 +0300)]
Implement statistics collection: (#2056)

1. Add `-s` CLI key to get statistics file path
2. Implement `StatisticsWriter` singleton to manage handle to this file

4 years ago[PP GAPI] - Generic precision conversion kernel; support for U8 (#2076)
Anton Potapov [Wed, 9 Sep 2020 12:30:08 +0000 (15:30 +0300)]
[PP GAPI] - Generic precision conversion kernel; support for U8 (#2076)

- added U8 support
- tests are extended

4 years ago[GNA] Fix for concat layer with >2 inputs (#1475)
Bartosz Sochacki [Wed, 9 Sep 2020 11:55:07 +0000 (13:55 +0200)]
[GNA] Fix for concat layer with >2 inputs (#1475)

* Fix for concat layer with more than 2 inputs

Signed-off-by: Bartosz Sochacki <bartosz.sochacki@intel.com>
* Fixed check if affine is used for crop layer

Signed-off-by: Bartosz Sochacki <bartosz.sochacki@intel.com>
* code cleanup for fix affine layer check

Signed-off-by: Bartosz Sochacki <bartosz.sochacki@intel.com>
* added test for concat layer with multiple inputs

* simplified test to use less number of layers

* fixed code style

* fixed coding style

* addressed review comments and one more issue that appeared during testing

* fixed code style errors

* scale factor propagation for concat layer with multiple inputs

* fix for a case when all inputs to concat are activation layers

* fix for linux compilation - C++14 is not enabled and fails on lambda with auto parameters

* corrected current year in headers in concat multi input tests

* fixes for code review issues raised by Denis Orlov

* enabled integer mode computation in GNA concat multi input test

* removed 1 space per review comment

* a fix to fail when not all scale factors are equal

* added GNA_DEVICE_MODE config to concat multi input test

* corrected searching for a next input to concat layer

* changed selection of 2nd candidate for source quant value

* code style fix - else and brackets should be in the same line

* small code improvement

* fix for mixing line endings

* addressed with endless requantization loop and fixed failing tests

4 years agoFixed AddressSanitizer issue (#2122)
Ilya Churaev [Wed, 9 Sep 2020 11:31:12 +0000 (14:31 +0300)]
Fixed AddressSanitizer issue (#2122)

4 years ago[IE CLDNN] fs_b_yx_fsv32 addition in quantize_scale_shift kernel and reorders removal...
Lukasz Debski [Wed, 9 Sep 2020 10:35:31 +0000 (12:35 +0200)]
[IE CLDNN] fs_b_yx_fsv32 addition in quantize_scale_shift kernel and reorders removal fix (#2124)

4 years ago[IE][Tests]: Fixes dangling reference access in nGraph function comparator (#2105)
Gladilov, Gleb [Wed, 9 Sep 2020 10:22:29 +0000 (13:22 +0300)]
[IE][Tests]: Fixes dangling reference access in nGraph function comparator (#2105)

Usage reference to front of the queue after pop is UB

Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
4 years ago[IE TESTS] Remove normilizer tests (#2098)
Irina Efode [Wed, 9 Sep 2020 10:21:55 +0000 (13:21 +0300)]
[IE TESTS] Remove normilizer tests (#2098)

4 years agoTemporarily disable ONNX Loop operator for the 2021.1 release (#2110)
Michał Karzyński [Wed, 9 Sep 2020 10:07:02 +0000 (12:07 +0200)]
Temporarily disable ONNX Loop operator for the 2021.1 release (#2110)

4 years agoClang compilation error fix for Linux (#2126)
Tomasz Dołbniak [Wed, 9 Sep 2020 10:04:33 +0000 (12:04 +0200)]
Clang compilation error fix for Linux (#2126)

4 years ago[TEST] One more ReduceSUM func test
Alexander Peskov [Sun, 6 Sep 2020 23:47:14 +0000 (02:47 +0300)]
[TEST] One more ReduceSUM func test

Special test case with input values which cannot be correctly processed via
decomposition with int AVG pool layer.

Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
4 years ago[TEST] Fix blob util broadcast filler for non fp32 types
Alexander Peskov [Sun, 6 Sep 2020 23:45:15 +0000 (02:45 +0300)]
[TEST] Fix blob util broadcast filler for non fp32 types

Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
4 years ago[NGRAPH] Fix ReduceSum decompose pass
Alexander Peskov [Mon, 31 Aug 2020 12:50:20 +0000 (15:50 +0300)]
[NGRAPH] Fix ReduceSum decompose pass

Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
4 years agoHide itt.hpp to src folders (#2091)
Ilya Lavrenov [Wed, 9 Sep 2020 09:34:16 +0000 (12:34 +0300)]
Hide itt.hpp to src folders (#2091)

4 years agoUse metadata from executable network when fetching results of the infer request;...
Alexander Perepelkin [Wed, 9 Sep 2020 08:47:40 +0000 (11:47 +0300)]
Use metadata from executable network when fetching results of the infer request; Use user provided output precision when it was supplied (#2111)

4 years agoUse element type instead of F32 for constants (#2121)
Katarzyna Mitrus [Wed, 9 Sep 2020 07:50:27 +0000 (09:50 +0200)]
Use element type instead of F32 for constants (#2121)

4 years agoSort memcheck timeline report for failures (#2088)
Andrey Somsikov [Wed, 9 Sep 2020 07:10:22 +0000 (10:10 +0300)]
Sort memcheck timeline report for failures (#2088)

4 years ago[IE CLDNN] Fix eltwise activation types mismatch (#2090)
Vladimir Paramuzov [Wed, 9 Sep 2020 06:57:38 +0000 (09:57 +0300)]
[IE CLDNN] Fix eltwise activation types mismatch (#2090)

4 years ago[CPU] gemm convolution: added post ops JIT implementation (#1817)
Anton Voronov [Wed, 9 Sep 2020 06:17:57 +0000 (09:17 +0300)]
[CPU] gemm convolution: added post ops JIT implementation (#1817)

4 years agoSupport python 3.8 by the Model Optimizer tool in default configuration (#2078)
Roman Kazantsev [Wed, 9 Sep 2020 05:34:43 +0000 (08:34 +0300)]
Support python 3.8 by the Model Optimizer tool in default configuration (#2078)

* Support python 3.8 by the Model Optimizer tool in default configuration

* Fix after review #1

* Fix after the second round review

4 years ago[VPU][OpenCL] Update custom kernels (#2131)
Evgeny Latkin [Wed, 9 Sep 2020 00:50:40 +0000 (03:50 +0300)]
[VPU][OpenCL] Update custom kernels (#2131)

* [Custom CL] Updated OpenCL kernels and tests

* [Custom CL] Update OpenCL compiler

* Update firmware to 1365

* Disable ExpGenerateProposals tests

* VPU: new firmware no. 1370

* Myriad: re-enable ExpGenerateProposals tests

Co-authored-by: Maxim Kurin <maxim.kurin@intel.com>
4 years agoAdd `runPipeline` wrapper to mandatory track full run (#2100)
Vitaliy Urusovskij [Tue, 8 Sep 2020 21:13:07 +0000 (00:13 +0300)]
Add `runPipeline` wrapper to mandatory track full run (#2100)

4 years agoAdded code owners for scripts folder (#2130)
Artyom Anokhov [Tue, 8 Sep 2020 14:23:27 +0000 (17:23 +0300)]
Added code owners for scripts folder (#2130)

4 years agoUpdate tolerance for candy model (#1967)
Mateusz Tabaka [Tue, 8 Sep 2020 14:04:08 +0000 (16:04 +0200)]
Update tolerance for candy model (#1967)

4 years ago[nGraph] Enable u1 data type in Constant PyAPI (#2084)
Jan Iwaszkiewicz [Tue, 8 Sep 2020 11:11:05 +0000 (13:11 +0200)]
[nGraph] Enable u1 data type in Constant PyAPI (#2084)

4 years ago[IE][VPU]: Decompose Swish to Sigmoid + Prod (#2107)
Roman Vyunov (Intel) [Tue, 8 Sep 2020 08:52:05 +0000 (11:52 +0300)]
[IE][VPU]: Decompose Swish to Sigmoid + Prod (#2107)

* Workaround to decompose Swish to Sigmoid + Multiply

4 years ago[IE CLDNN] Switch back to old TI unroller to restore correct matching of TensorIterat...
Mikhail Letavin [Tue, 8 Sep 2020 08:48:11 +0000 (11:48 +0300)]
[IE CLDNN] Switch back to old TI unroller to restore correct matching of TensorIterator output names (#2096)

4 years ago[IE][VPU]: myriad_compile doesn't work without device - fix (#2054)
Daria Mityagina [Tue, 8 Sep 2020 08:21:18 +0000 (11:21 +0300)]
[IE][VPU]: myriad_compile doesn't work without device - fix (#2054)

* Revert setting deprecated PLATFORM config option in compile tool

4 years ago[GNA] add support for NCHW & NHWC layouts for exporting output (#2031)
Anna Alberska [Tue, 8 Sep 2020 07:57:44 +0000 (09:57 +0200)]
[GNA] add support for NCHW & NHWC layouts for exporting output (#2031)

* [GNA] add support for NCHW & NHWC ExportScores

* fix cpplint

4 years ago[GNA] Handling input orientation (#1851)
Andrey Dmitriev [Tue, 8 Sep 2020 07:46:10 +0000 (10:46 +0300)]
[GNA] Handling input orientation (#1851)

Added test

Add fix

4 years agoGRU/RNN/LSTM sequence ops, reference implementations, single layer tests (#1594)
Ivan Tikhonov [Tue, 8 Sep 2020 07:31:44 +0000 (10:31 +0300)]
GRU/RNN/LSTM sequence ops, reference implementations, single layer tests (#1594)

* gru/rnn sequences

* update gru/rnn sequences ops, add unit tests

* enable sequence transformations for cpu plugin

* ngraph codestyle

* update tensor iterator to rnn/gru/lstm sequence transformations, add unit tests

* ngraph codestyle

* add visitors for ngraph ie ops, fix a bug with incorrect axis, fix ngraph to ngraph ie conversion

* update GRUSequence/GRUSequenceIE according to plugin format

* fix ngraph ie implementations according to plugins restricrictions

* fix naming issue

* adapt unit tests to accordance to new changes

* strict checks, additional unit tests

* add descriptions for transformations, fix unit tests

* enable ti to sequnece and unroll transformations in plugins for testing

* disable tensor iterator ngraph reader tests

* delete unnecessary cmake file

* fix includes

* clean up, resolve review comments

* move ti to sequence transformation to ti folder

* validate_and_infer_types() implementation

* input parameter validation for LSTM, GRU and RNN

* style-check applied

* Add LSTMSequence dynamic shape validation and test props for RNNCell, GRUCell, LSTMCell and LSTMSequence.

* recurrent_sequence.hpp moved to ngraph/core/include/ngraph/op/util/

* style check applied

* removed unused variable from LSTMSequence::validate_and_infer_types

* Add missing newline mark at the end of file.

* Add supression macro for FusedOp deprecation.

* Add element type initialization

* transpose,rnn cell reference implementations

* Apply PR review remarks

* reference implementations for cells op, single layer tests, align lstm cell/sequence according to the spec

* lstm/gru/rnn cell decompostion transformations

* ngraph codestyle

* clean up

* ngraph code style

* change inheritance of Cells, fix build

* fix build

* fix build again

* remove Peepholes from LSTMSeq, fix copy_runtime_info in transformations

* Rewrite tests to use gtest exception assertions.

* resolve tests issues

* ngraph codestyle

* add missed files

* fix typeprop tests

* fix lstm sequence checks

* fix arm build

* fix arm again

* delete unnecessary file

* add convert weghts format function, enable lstm test, resolve review comments

* add ngraph builders

* ngraph codestyle

* fix unit tests

* revert transpose reference implementation

* move ti to sequences transformation to another branch, resolve review comments

* resolve review comments

* revert fix in ie_layer_validators

* revert LSTM Cell v0, add LSTMCell v1, update transformation lstm_cell_to_cell_ie

* v1 version of LSTMCell op

* LSTMSequence v1 operation, exclude LSTMSeq from opset4

* fix python api tests

* resolve review comments, tests for decomposition transformations, switch lstm cell to opset4 in mo

* references impl for RNN/GRU/LSTM Sequences, single layer tests, bidirectional transformation

* fix unit tests

* process dynamic ranks of rnn/gru/lstm ops

* remove sequences specifications from opset4

* resolve review comments

* fix validate_and_infer_types of GRU/RNN sequences

Co-authored-by: Szymon Durawa <szymon.durawa@intel.com>
4 years ago[IE CLDNN] Added is_discrete flag into device info and FULL_DEVICE_NAME flag (#2089)
Vladimir Paramuzov [Tue, 8 Sep 2020 06:56:54 +0000 (09:56 +0300)]
[IE CLDNN] Added is_discrete flag into device info and FULL_DEVICE_NAME flag (#2089)

4 years agoONNX Resize fix (#2103)
Tomasz Dołbniak [Tue, 8 Sep 2020 06:51:30 +0000 (08:51 +0200)]
ONNX Resize fix (#2103)

4 years ago[IE][VPU][GT][IE Transformations]: Fixes for post processing model (#2041)
Andrew Bakalin [Mon, 7 Sep 2020 20:23:47 +0000 (23:23 +0300)]
[IE][VPU][GT][IE Transformations]: Fixes for post processing model (#2041)

* Disable StridedSlice to crop convertion in dynamic cases
* NMS: increase shave requirements for some cases
* Update firmware
* Add test cases

4 years ago[LPT] Multiinput with one parent and FQ with three Constant (#2066)
Edward Shogulin [Mon, 7 Sep 2020 17:31:45 +0000 (20:31 +0300)]
[LPT] Multiinput with one parent and FQ with three Constant (#2066)

* [LPT] FakeQuantize with three constants

* [LPT] Dequantization ops on thw inputs with one parent

4 years agoRefactoring: install_openvino_dependencies.sh script (#2112)
Dmitrii Denisov [Mon, 7 Sep 2020 16:54:56 +0000 (19:54 +0300)]
Refactoring: install_openvino_dependencies.sh script (#2112)

4 years ago[IE TESTS] GatherTree single layer test has been created. (#2006)
Maksim Kutakov [Mon, 7 Sep 2020 14:17:14 +0000 (17:17 +0300)]
[IE TESTS] GatherTree single layer test has been created. (#2006)

* [IE TESTS] GatherTree op ref function has been created.

* [IE TESTS] Added GatherTree single layer test

* [IE TESTS] Fixed code styles.

* [IE TESTS] GatherTree test FP32 precion was enabled.

* [IE TESTS] Refactoring of Builder::makeConstatn procedure

The refactoring is aimed at managing the range of random data for the constants initialization procedure.

* [IE TESTS] GatherTree test was extended with constants

* [IE TESTS] GatherTree ref rewritten to non-templated function.

* [IE TESTS] GatherTree test inp shape indx enum removed.

* Revert "[IE TESTS] Refactoring of Builder::makeConstatn procedure"

This reverts commit 2648172e00ccca266d39e8775b890b8a8395f57c.

* [IE TESTS] makeConstant was augmented with random data range parameters.

* [IE TESTS] GatherTree test was rewritten using makeConstant function.

* [IE TESTS] GaterTree test call templated makeConstant

* [IE TESTS] GaterTree test code style fix

4 years agov1::AvgPooling type prop unit tests (#2013)
Tomasz Dołbniak [Mon, 7 Sep 2020 14:05:19 +0000 (16:05 +0200)]
v1::AvgPooling type prop unit tests (#2013)

4 years ago[CPU] fix: supported Tile with more than 5 dims (#2062)
Anton Voronov [Mon, 7 Sep 2020 12:39:19 +0000 (15:39 +0300)]
[CPU] fix: supported Tile with more than 5 dims (#2062)

4 years ago* [CPU] Extend Reduce node to support blocked layouts nC[d]hw8/16C (#580)
Chen Xu [Mon, 7 Sep 2020 12:35:11 +0000 (20:35 +0800)]
* [CPU] Extend Reduce node to support blocked layouts nC[d]hw8/16C (#580)

4 years ago[IE CLDNN] Grouped convolution kernel improvement (#2063)
Mikołaj Życzyński [Mon, 7 Sep 2020 08:52:06 +0000 (10:52 +0200)]
[IE CLDNN] Grouped convolution kernel improvement (#2063)

4 years agoRemove list of the supported ops by ONNX Importer (#2061)
Tomasz Socha [Mon, 7 Sep 2020 08:50:55 +0000 (10:50 +0200)]
Remove list of the supported ops by ONNX Importer (#2061)

4 years ago[IE CLDNN] Fix the order of dimensions in reshapes inserted by Select primitive ...
Mikhail Letavin [Mon, 7 Sep 2020 08:49:56 +0000 (11:49 +0300)]
[IE CLDNN] Fix the order of dimensions in reshapes inserted by Select primitive (#2082)

4 years agoVPU: update firmware version to 1354 (#2087)
Evgeny Latkin [Mon, 7 Sep 2020 07:58:36 +0000 (10:58 +0300)]
VPU: update firmware version to 1354 (#2087)

4 years agoRefactored legacy code for mean_scale_values transformations (#1936)
iliya mironov [Mon, 7 Sep 2020 07:41:47 +0000 (10:41 +0300)]
Refactored legacy code for mean_scale_values transformations (#1936)

* Remove move_to_preproc. Not actual.

* Updated documentation

4 years ago[IE][VPU]: Fix behavior tests for MXpcie (#1879)
Andrey Chekhonin [Mon, 7 Sep 2020 07:38:07 +0000 (10:38 +0300)]
[IE][VPU]: Fix behavior tests for MXpcie (#1879)

* Moved mem type patching under USB protocol switch
* Myriad beh tests fix

4 years agoNgraph improvements (#2058)
Vladislav Volkov [Mon, 7 Sep 2020 07:36:52 +0000 (10:36 +0300)]
Ngraph improvements (#2058)

4 years agoMove FakeOutput resolving to back phase (#2033)
Maxim Vafin [Mon, 7 Sep 2020 07:20:24 +0000 (10:20 +0300)]
Move FakeOutput resolving to back phase (#2033)

4 years ago[IE CLDNN] Fixed fsv16 lrn kernel with fp16 input (#2086)
Vladimir Paramuzov [Mon, 7 Sep 2020 06:04:05 +0000 (09:04 +0300)]
[IE CLDNN] Fixed fsv16 lrn kernel with fp16 input (#2086)

4 years agoHandle Split axes as i64 (#2079)
Tomasz Dołbniak [Mon, 7 Sep 2020 02:48:32 +0000 (04:48 +0200)]
Handle Split axes as i64 (#2079)

4 years ago[IE CLDNN] Fixing blocked format opting for strided_slice (#2073)
Lukasz Debski [Sat, 5 Sep 2020 16:01:43 +0000 (18:01 +0200)]
[IE CLDNN] Fixing blocked format opting for strided_slice (#2073)

4 years ago[IE CLDNN] Disable unsupported blocked formats for softmax (#2049)
Vladimir Paramuzov [Sat, 5 Sep 2020 07:11:18 +0000 (10:11 +0300)]
[IE CLDNN] Disable unsupported blocked formats for softmax (#2049)

4 years ago[IE CLDNN] Reduce performance drop fix (#2065)
Mikołaj Życzyński [Sat, 5 Sep 2020 07:11:00 +0000 (09:11 +0200)]
[IE CLDNN] Reduce performance drop fix (#2065)

4 years ago[IE CLDNN] Enable bfyx fully_connected for large batches (#1996)
Konrad Dobros [Fri, 4 Sep 2020 19:22:46 +0000 (21:22 +0200)]
[IE CLDNN] Enable bfyx fully_connected for large batches (#1996)

This change enables fully_connected to use bfyx format for large
batches, when optimized bf_tiled kernel best performance.

4 years agoFix fusing Multiply node with Convolution in case group != 1 (#1882)
Maxim Vafin [Fri, 4 Sep 2020 17:32:51 +0000 (20:32 +0300)]
Fix fusing Multiply node with Convolution in case group != 1 (#1882)

* Fix fusing Multiply node with Convolution in case group != 1

* Add transformation test

* Do not fuse if not possible to reshape const

* Update fuse_linear_ops.py

4 years agoMake pybind more verbose in debug on windows (#2045)
Tomasz Dołbniak [Fri, 4 Sep 2020 16:03:33 +0000 (18:03 +0200)]
Make pybind more verbose in debug on windows (#2045)

* Make pybind more verbose in debug on windows

* Remove the NDEBUG flag everywhere

* Code complexity reduction...

* Missing colon

* And now the missing empty line...

* Reusable functions

* Now the mood of the sentence was wrong...

* Free functions instead of methods

4 years agoHostTensor2Vector (#2027)
Jozef Daniecki [Fri, 4 Sep 2020 13:02:14 +0000 (15:02 +0200)]
HostTensor2Vector (#2027)

* Add host_tesnor_2_vector() implementation and unit tests. One reference OP refactored to use it.

* Ngraph assertion message refactored.

Co-authored-by: Tomasz Dołbniak <tomasz.dolbniak@intel.com>
* Fix style.

Co-authored-by: Tomasz Dołbniak <tomasz.dolbniak@intel.com>
4 years ago[CPU] Disabled several non-valid static assertions (#2055)
Gorokhov Dmitriy [Fri, 4 Sep 2020 12:38:07 +0000 (15:38 +0300)]
[CPU] Disabled several non-valid static assertions (#2055)