IntenseWebs/openvino
3
0 Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
4,799 Commits 38 Branches 54 Tags
e0cea200e33bb1540022ce43e1b3b88cb1bfb6d4
Commit Graph
2856 Commits
Author SHA1 Message Date
Sungeun KimandKim, Mingyu e0cea200e3 [GPU] Onednn integration for pooling (#7753) 
* [GPU] Onednn integration for pooling primitive

* layout_optimizer for pooling

* Padding code merge for onednn

Co-authored-by: Kim, Mingyu <mingyu.kim@intel.com>
 2021-10-01 15:02:30 +09:00
Gleb Kazantaev a16cc81233 Remove VariantWrapper for simple attr classes (#7771) 
* Remove VariantWrapper for simple attr classes

* Code style
 2021-10-01 08:03:37 +03:00
Ilya Churaev faeaf045a9 Graph comparator to ngraph util (#7729) 
* Moved FrameworkNode to nGraph

* Moved graph comparator to ngraph test util

* Fixed build

* Try to fix centos

* Fix export target

* Moved engine utils to separate library

* Removed ONNX util from common library

* Fixed build

* Fixed code style
 2021-10-01 07:24:28 +03:00
Ilya Znamenskiy f675df625c [GPU] Added post-ops support for OneDNN primitives (#7737) 
[GPU] Memory lock fix
 2021-10-01 12:18:00 +09:00
Ilya Lavrenov 302eb08dc5 Replaced gtest deprecated macro (#7779) 2021-10-01 03:46:32 +03:00
Egor Duplenskyanddmitrygo 778cc2ddbb [CPU] Migration on oneDNN v2.3.2 (#7680) 
binary post
binary post ops support for gemm convolution and gemm inner product
plugin fixes
mkldnn: avx512: fixed prelu post op

Co-authored-by: dmitrygo <dmitry.gorokhov@intel.com>
 2021-09-30 20:25:49 +03:00
Dmitry Pigasin 7fade1e357 Fix a result shape for a model with a 3d tensor at the output (#7741) 2021-09-30 16:58:11 +03:00
Maksim Derbasov 3f0fea5c1b Fix for warning C4297 (#7740) 2021-09-30 16:57:43 +03:00
Mikhail Nosov 6e05cead11 [OV20] Reference implementation for NV12toRGB and NV12toBGR operations (#7601) 
* Reference implementation for NV12toRGB and NV12toBGR operations
Tests:
- ngraph: visitor + type_prop
- template plugin: reference implementation
- inference-engine: shared tests for plugins
- cpu plugin: compare with ref implementation tests

* Fix clang

* Serialization tests

* Fix clang-format

* Changed 'f32' to 'any supported floating-point type'
Added appropriate shape inference tests
Added error test for >2 inputs
Fixed failed CI tests

* Updates after rebase
+ Try to fix Ninja build

* Fix CI

* Support endianness + potential fix of win32 test fails

* Fix review comment

* Fix review comments

* Fix unit test build

* Fix unit test build #2

* Possible build fix 3

* Simplified reference tests
Observed issue with shuffling Y pixels on little-endian systems, added tests
 2021-09-30 16:34:46 +03:00
Evgenya Stepyreva b339bb780b Resolved unexpected -1s in IRs (#7709) 2021-09-30 15:39:32 +03:00
Anton Pankratv fab4056ceb Fixed callback copy count (#6607) 2021-09-30 15:09:47 +03:00
Ilya Lavrenov 757db35528 Dynamism in new API (#7742) 
* Fixed precisions conversion in new API

* Added tests

* Fixed old IR cases

* Disable FP16

* Fixed regex for CentoOS

* Refactored tests to use new API

* Temp

* Fixed tests

* Moved smart reshape related sources to ngraph

* Added tests for invalid names

* Moved reshape to tensor_names

* clang-format

* Fixed CC build

* Removed IEConv, IEDeconv from primitives pririty
 2021-09-30 15:04:24 +03:00
Sergey Shlyapnikov f3c8f2bc49 [GPU] Add use OneDNN impls attribute (#7716) 2021-09-30 19:15:40 +09:00
Katarzyna Mitrus 0eeaf6b2e4 ROIAlign fix - Unify sample_value calculation between max/avg mode (#7710) 
* Unify sample_value between max/avg mode

* Fix mkldnn roi_align impl

* Update test

* Revert missing assert
 2021-09-30 11:50:06 +03:00
Mikhail Nosov 414c3dc133 [OV2.0] Preprocessing - resize (#7583) 
* Initial version

* Added 'network' layout to preprocessing info
Moved existing resize tests to template plugin

* Fix clang

* More tests for 'resize' reference implementation + CPU tests + error cases
Coverage is 100%

* Align with new base_reference_test implementation

* Fixed comments

* Add assert to check that desired size is not out of bounds

* CPU: skip failed test
 2021-09-30 11:41:25 +03:00
Sergey Shlyapnikov c00b0b6ae4 [GPU] Profiling refactoring and fixes (#7714) 2021-09-30 10:54:27 +03:00
Roman Lyamin 17091476d8 [IE CLDNN] Add optional memory access type (#7686) 2021-09-30 10:32:07 +03:00
Mingyu Kim fbed7e13a8 [GPU] debug_config bug fix of not accepting multiple strings (#7754) 
It receives space-separated parameter. Currently, only first argument is
received in case of string argument.
 2021-09-30 10:10:50 +03:00
Dmitrii Khurtin a13b934622 [GNA] fixed failed to inserting transpose after matmul (#7720) 
* fixed failed to inserting transpose after matmul

* added verifier for matmul
 2021-09-30 09:55:34 +03:00
Sergey Shlyapnikov ffd2091477 [GPU] Add new data formats (#7722) 2021-09-30 09:38:48 +03:00
Vladislav Volkov 4bf8d15793 Import/Export tests enabled for the CPU plugin (#7735) 2021-09-30 09:11:01 +03:00
Ilya Lavrenov 1f5355cd07 Enabled several CPU tests (#7724) 2021-09-29 20:51:23 +03:00
Gleb Kazantaev 62ed238f54 RT Info Attributes Serialization/Deserialization (#7603) 
* Init rt info deserialization

* Add RT Info attributes deserialization

* Add RT Info attributes serialization

* Add Serialization and Deserialization tests

* Code style

* Fix AttributeAdapter place

* Extended rt_info section; added support for multi-field attrs; use version for attr serialization/deserialization

* Align attributes and their usage

* Fix LPT

* Add missing #pragma one

* Fix build

* Change version delimiter to have valid xml

* Fix PromitivesPriority

* Fix tests

* Fix test

* Fix IR Reader; remove empty rt_info

* More explicit attribute name and version representation
 2021-09-29 19:17:03 +03:00
Vladislav Golubev 04158afc21 [LPT] Added check for uniqueness of friendly names in tests (#7570) 
* FQDecomposition: added check on unique friendly names after transformation

* NormalizeL2Transformation: added check on unique friendly names after transformation

* ti to sequences conversion: naming fix

* [LPT][TESTS] Check on unique names added to functional tests

* compilation error fix

* compilation error fix
 2021-09-29 18:40:21 +03:00
Eugeny Volosenkov 08897e4c80 Transformation convert_reduce_to_pooling fix (#7705) 
* fix convert_reduce_to_pooling transformation

* Fix codestyle

* add/delete spaces

* Fix code style

* fix code style
 2021-09-29 12:51:10 +03:00
Jade Cho 1e6fd56e01 [GPU] Merge kernel updates (#7699) 
+ fix a bug due to bias type mismatching
  - convolution_gpu_bfyx_to_bfyx_f16

+ refactoring
  - depth_to_space_kernel_base
  - depth_to_space_kernel_ref

+ Adjusting LWS
  - eltwise_kernel_base
 2021-09-29 17:10:53 +09:00
Mingyu Kim ee93823b3a [GPU] New debug config: dry_run (#7689) 
Enabling dry_run will shorten runtime of any application by exiting just
after graph compilation.
 2021-09-29 17:09:52 +09:00
Mikhail Nosov 4fd59f72ea Use 'u8' element type in tests (#7704) 
Use 'float32' output for 'compare with reference' tests to avoid IE precisions undesired conversions
Added GPU preprocessing tests skeleton
 2021-09-29 09:46:37 +03:00
Ilya Lavrenov bd09f70876 Fixed precisions conversion in new API with compile_model(filename) (#7711) 
* Fixed precisions conversion in new API

* Added tests

* Fixed old IR cases

* Disable FP16
 2021-09-29 07:00:13 +03:00
Zhang Yi a5250fd0fc [Transformation]hswish_fusion with clamp mul (#7414) 
* [Transformation]hswish_fusion with clamp

* add const for constant variable

* [Transformation]fix review comments

* [Transformation]fix opset version
 2021-09-28 19:22:53 +03:00
Krzysztof Bruniecki eee864aed6 [GNA] Fix KEY_EXEC_TARGET (cherry pick #7671) (#7701) 
* Use Gna2DeviceCreateForExport when GNA_EXEC_TARGET is != detected

* Update detected GNA device version field in GNA Device helper

   * Use EXEC instead of COMPILE TARGET to append
   CNN Legacy enforcement (GNA1)

* Apply review
 2021-09-28 18:39:49 +03:00
Ilya Lavrenov 068d31511b Improved ov::Tensor behavior (#7683) 
* Improved ov::Tensor behavior

* Fixed python test for setShape on preallocated

* Fixed clang-format
 2021-09-28 12:50:51 +03:00
Sergey Shlyapnikov 6bd0873a40 [GPU] Add CL_QUEUE_INDEX property and queue configuration refactoring (#7633) 2021-09-28 11:37:09 +03:00
Dmitrii Khurtin a3dfa980ad [GNA] Improved accuracy on model after Accuracy Aware (#7576) 
* improved accuracy on model after Accuracy Aware

* refactoring+test

* removed unnessary FakeQuantI8/I16

* added comments, moved fake_quantized from UpdateInputScaleFromNetwork(), removed _Np template param from QuantDescTmpl
 2021-09-28 11:34:09 +03:00
Mingyu Kim acc14c6469 [GPU] Onednn integration for reorder primitive (#7687) 2021-09-28 17:10:35 +09:00
Gleb Kazantaev 204c17cc21 Enable SoftmaxFusion inside MOC Transformations pipeline (#7684) 
* Enable SoftmaxFusion inside MOC Transformations pipeline

* Disable SoftmaxDecomposition by default
 2021-09-28 10:46:31 +03:00
Ilya Lavrenov 476fbee00f Reference template plugin tests to OpenVINO 2.0 (#7685) 2021-09-28 10:46:03 +03:00
Roman Lyamin 6ecfbd47e3 [IE CLDNN] clDNNPlugin size fix (#7655) 2021-09-27 14:32:52 +03:00
Vladislav Golubev 7fa9bbf6fc [LPT] Avoid using std::shared_ptr<Node> when creating a node (#7357) 
* [LPT] Avoid using std::shared_ptr<Node> when creating a node

* [LPT] removed unused files

* [LPT] D2STransformation: transform & isPrecisionPreserved methods are moved to base class

* [LPT] Revert redundant changes
 2021-09-27 11:50:17 +03:00
Irina Efode 818f385398 [IE TESTS] Add example for QueryNetwork in SLT (#7628) 
* [IE TESTS] Add example for QueryNetwork in SLT

* Update mul_conv_fusion.cpp

* Skip

* skip[
 2021-09-27 11:22:38 +03:00
Maxim Andronov 3bf34b1166 [CPU] fix cloneWithUndefStridesAndOffset and cloneWithDefaultStridesA… (#7542) 
* [CPU] fix cloneWithUndefStridesAndOffset and cloneWithDefaultStridesAndOffset

* status fix

* applied comments after review 1

* strides compute was aligned and extracted

* review comments
 2021-09-27 10:32:47 +03:00
Maxim Andronov 5166994750 [CPU] Supporting dynamism into Eltwise and Reorder (#7553) 
* [CPU] Supporting dynamism into Eltwise and Reorder

* reference node fixes

* nms fixes

* part comment fixed

* review comments

* some changes for input node

* comments fix

* monir renaming
 2021-09-27 10:06:09 +03:00
Ilya Lavrenovandapankratovantonp e87cc3fa9e Tensor API in ngraph (#7632) 
* Added OpenVINO Tensor API

* Tensor API improvements

* Moved Tensor to ngraph

* Moved Tensor tests

* Fixed docs and code style

* Trying to fix Windows

* Fixed clang-format

* Moved Tensor to runtime namespace

* Fixed compilation

* Fixed clang-format

* Fixed tests in debug

Co-authored-by: apankratovantonp <anton.pankratov@intel.com>
 2021-09-27 09:57:26 +03:00
Mingyu Kim 95f8544aa6 [GPU] Detect dpas support (#7643) 
* Detect dpas support
* New debug flag DISABLE_ONEDNN is added
 2021-09-27 05:38:57 +03:00
Wilson Seok 8262aed30c Migrate ngraph backend test of arithmetic2 operations (#7500) 
* Remove fp16 of Convert layer test from skip_tests.config.cpp as it works now

* update repo

* add sin asin template plugin reference test

* add cos template plugin reference test and remove asin/sin/cos ngraph backend test

* update CMakeList.txt

* add template plugin reference test for asinh, cosh, sinh, tanh

* remove ngraph backend test for asinh, cosh, sinh, tanh

* update CMakeList.txt

* add ngraph type prop test for tanh

* add ngraph visitor api test for asin, sin

* add atanh type in activation SLT

* remove boolean from sin/cos/tan/asin/sinh/cosh evaluate
 2021-09-27 05:23:52 +03:00
Evgenya Stepyreva ef028a567e Convolution: fast shape inference (#7523) 
* Convolution: fast shape inference

* StaticShape and StaticDimension + static shape infer time test in comparison to Convolution

* Review comments
 2021-09-27 00:14:50 +03:00
Egor Duplensky c92988c8e9 [CPU] Add OV_CPU_VERBOSE env var to print node exec info to cout (#6390) 2021-09-26 22:17:57 +03:00
Yuan Hu b968c7b813 AutoPlugin: support performance Hint (#7621) 
Signed-off-by: Hu, Yuan2 <yuan2.hu@intel.com>
 2021-09-25 12:30:21 +03:00
Pavel Esir 572ab040dc [IE] fix printing dynamic signatures (#7620) 
* [IE] fix printing dynamic signatures

* fixed printing scalars output dimensions as well

* turned off printing dynamic signatures
 2021-09-24 18:50:35 +03:00
Vladimir Paramuzov 3eb1aa59de [MULTI] Fixed devices comparison to respect default device id (#7311) 
* [GPU] Fixed default device id

* [MULTI] Fixed devices comparison to respect default device id
 2021-09-24 12:32:38 +03:00