Commit Graph

84 Commits

Author SHA1 Message Date
Maksim Kutakov 35696ca1db
[CPU] Unified memory descriptor (#6328) 2021-08-02 09:19:15 +03:00
Yury Gaydaychuk 7ab92b5845
[CPU] Adaptive pooling implementation (#6594) 2021-07-29 20:41:24 +03:00
Egor Shulman ab8d046642
[CPU] NormalizeL2 - unsorted axes are not supported (#6602) 2021-07-29 17:30:38 +03:00
Yury Gaydaychuk e5a990e356
[CPU] RoiPooling handles border proposals correctly (#6199) 2021-07-26 23:35:27 +03:00
Szymon Durawa fa057627ec
Ref implementation mvn revise (#6600)
* Add visitor and backend tests.

* Add reduction_axes to SLTs, serialization and backend tests.

* Update backend tests.

* Move backend tests to template plugin, remove old ones.

* Apply correct format.

* Add BF16 to SLT, add comment regarding deprecated class.

* Remove BF16 precision as it is not supported in windows, no default values for attributes.

* Reuse Tensor from base_reference_test.hpp
2021-07-23 07:47:08 +03:00
Ilya Lavrenov bc36425381
Added suggest-override flag (#6631) 2021-07-16 18:55:05 +03:00
Bartek Szmelczynski 13de29db97
Revise sinh (#6432)
* update sinh spec

* add rounding for integral type

* refactor ngraph sinh class, add backend, visitors and type_prop tests

* refactor backend test for int32, remove redundant includes, refactor spec, add SLT for ints

* fix style

* add sinh to list of trusted ops

* re-add Sinh to int SLT

* update python modules

* revert update of python modules, and fix the depracated typed_test formula

* fix deprecated typed_test formula
2021-07-05 10:11:02 +03:00
Szymon Durawa 055a70bdf9
Add output_padding and output_shape into GroupConvolutionBackprop SLTs. (#6302)
* Add output_padding and output_shape into GroupConvolutionBackprop SLTs.

* Add legacy API for arm-plugin dependency.

* Replaced templated input, output and filter with one template.

* Apply correct format.

* Change to use INSTANTIATE_TEST_SUITE_P macro.
2021-06-29 15:34:29 +03:00
Egor Shulman fcc07f80cd
[IE CPU] Reimplement extension nodes via MKLDNNNode API (#5784) 2021-06-23 14:22:10 +03:00
Ilya Lavrenov fa2f9c5201
Migrated to official GoogleTest repo (#6286)
* Move gtest to <root>/wqthirdparty

* Fixed inference engine tests

* Fixed unit-tests

* Fixed GPU

* Fixed clDNN_unit_tests compilation

* Reverted ngraph changes

* Fixed VPU tests

* Fixed deprecated tests compilation

* Removed OpenVINO clone

* Added public submodule

* Removed

* Fixed Windows

* More updates for INSTANTIATE_TEST_SUITE_P

* Use release-1.10.0

* Removed ngraph copy of gtest

* Revert "Removed ngraph copy of gtest"

This reverts commit ec9fe08d79.

* Fixes for TYPED_TEST_CASE

* Fixed GNA tests

* Updated submodule

* Updaed index

* TMP disabled tests

* Revert changes and fix compilation errors

* Fixed caching tests

* Revert back

* Revert back all

Co-authored-by: Michael Nosov <mikhail.nosov@intel.com>
2021-06-22 17:39:39 +03:00
Alexandra Sidorova b05977a536
[CPU][IE TESTS] Added more input shapes for Pooling tests (#6083) 2021-06-16 14:14:50 +03:00
Szymon Durawa 772465da1e
Add output shape and output padding for Convolution Backprop SLTs. (#5576)
* Create output shape for Convoution Backprop SLTs.

* Add output_padding attribute to SLT scope.

* Introduce SLT for Serializaton.

* Introduce new test layer class ConvolutionBackpropLayerTest which contains output_padding attribute and output_shape input. Old one is deprecated, but cannot be removed due to kmb plugin dependency.

* Add ConvolutionBackpropDataLayerTest into TEST_P.

* ConvolutionBackpropDataLayerTest left as legacy class used by kmb_plugin.

* Remove redundant variables.

* Switch to new API for gpu SLTs.

* Remove legacy API.

* Introduce legacy API to match dependency for KMB and ARM plugins.

* Create test cases for output_padding attribute.

* Fixing smoke_Deconv tests.
2021-06-15 07:08:10 +03:00
Alexandra Sidorova 9214fa72e2
[CPU] Fixed AvgPooling and FQ fusing (#5994) 2021-06-08 14:04:31 +03:00
Egor Shulman b8b6b4d8b6
[CPU] Reduced number of tests for BatchToSpace and SpaceToBatch (#5385) 2021-06-04 15:59:44 +03:00
Maksim Kutakov 7fb9bac24a
[CPU] Extend Concat node logic to avoid fallback on slow ref implementation. (#4129) 2021-05-31 18:49:57 +03:00
Chenhu Wang d2003095dc
[CPU] MVN_accuracy_fix_on_avx512 (#5787) 2021-05-28 17:56:04 +03:00
Egor Duplensky e754a6b46f
[CPU] Fix RegionYolo (v3) node and Ngraph reference implementations (#5748) 2021-05-26 11:44:36 +03:00
Maxim Andronov 06fb16d799
[CPU] Winograd convolution support (#5699) 2021-05-26 09:19:11 +03:00
Egor Duplensky bae7f4b036
[CPU] Add bf16 support for ROI pooling (#5187) 2021-05-25 21:20:30 +03:00
Maksim Kutakov 617636693a
[CPU] Nspc layout enabling in the FP32/BF16 convolutions (#5292) 2021-05-25 11:41:23 +03:00
Maxim Andronov cec8ad417f
[CPU] Ngraph migration leftovers: performance fixes (#5680) 2021-05-24 08:43:18 +03:00
Alexandra Sidorova 57d49f3215
[CPU] Added MVN fusion for case with constants inside (#5644) 2021-05-21 14:35:56 +03:00
Daria Mityagina 5ea8437c91
[VPU] Batch support for interpolate (#5520)
* batch support for interpolate

* hash update

* compilation errors solved

* changes in tests, so not to break other plugins

* changes in tests, so not to break other plugins

* Revert "changes in tests, so not to break other plugins"

This reverts commit 67e5c1e4ddba4f0cc858eb1293e50ce9d9ae0227.

* Revert "changes in tests, so not to break other plugins"

This reverts commit f6a537d2236f0b64f0ff028627af9e0e388a2d23.

* review corrections

* Firmware update
2021-05-21 12:54:58 +03:00
Irina Efode ae637702d9
[IE TESTS][IE CPU] Fix BF16 threshold in ROI tests (#5703) 2021-05-20 17:11:43 +03:00
Irina Efode 9568f8120e
[IE TESTS] Fix Comparation issue in `LayerTestCommon` class (#5624)
* [IE TESTS] Fix comparation in LayerTestUtils

* Fixes

* Small fix

* Int4 fixes

* remove extra

* Fix NMS

* Some fixes for tests

* Add small fix

* [IE TESTS] Remove const folding as a result engine

* Remove extra

* Revert remove constant folding (DSR test) & fix some cases for cpu

* Fix GNA
2021-05-19 13:14:35 +03:00
Alexandra Sidorova b737e6ec96
[CPU] Fixed blocked layout conditions for DepthToSpace, SpaceToDepth (#5569) 2021-05-14 13:22:03 +03:00
Chenhu Wang 516479ae20
[CPU]shuffle_channel optimization (#4978) 2021-05-13 13:43:42 +03:00
Ilya Lavrenov 07a49184c0
Fixed typo in length (#5611) 2021-05-13 07:52:55 +03:00
Gorokhov Dmitriy a19413c0c0
[CPU] Plugin migration on ngraph (#4344) 2021-05-06 19:49:24 +03:00
Chenhu Wang 0808975a37
[CPU] windows_Interpolate_fused-FQ_nearest-mode_nspc-layout_fix (#5317) 2021-05-03 14:48:42 +03:00
Alexandra Sidorova 03ca3d1ef7
[CPU] Fixed SoftPlus for large positive values (#4932) 2021-04-30 13:34:33 +03:00
Maksim Kutakov 9f2a3d0edc
[CPU] Split supported PD initialization refactoring plus tests for multiply port connections. (#4521) 2021-04-26 15:18:26 +03:00
Egor Duplensky 39e1a21c42
[CPU] Enable bf16 RNN primitives (#4942) 2021-04-25 22:18:38 +03:00
Ivan Novoselov 594c4882e6
[CPU] optimize ExtractImagePatches (#4898) 2021-04-25 19:52:07 +03:00
Chenhu Wang 1426dc7d68
[CPU] Activations are fused with MVN (#5272) 2021-04-23 11:38:18 +03:00
Chenhu Wang 528d4b1bac
contain only some typical fusion, reduce tensor size, not skip on traget without avx512 (#5361) 2021-04-23 11:29:53 +03:00
Alexandra Sidorova 0d3baee1f3
[CPU] [IE TESTS] Added improvements for DepthToSpace and SpaceToDepth (#3897) 2021-04-21 16:45:07 +03:00
Piotr Szmelczynski 6910a737e0
Revise floor mod (#5058)
* add floor_mod type prop test to CMakeList

* create floor_mod type prop tests

* create op_eval tests for floor mod

* Update copyright

* add op to summarize.py

* update copyright

* remove unused variable

* add tests for int64 and int32

* clear unit test file

* Change second input generation of single layer test of cpu plugin to also generate negative numbers for floorMod

* Update test results and remove one test

* fix bug with secondaryInput shape

* add backend tests

* add new line

* create additional backend tests

* Update backend tests
2021-04-21 11:33:48 +03:00
Egor Shulman 211c5773fc
[CPU] BatchToSpace, SpaceToBatch layers optimizations (#4029)
* BatchToSpace, SpaceToBatch layers optimizations

* Fix according to review

* Fix tests for oneDNN (v1.6)

* Fix of support BF16

* Fix in/out config

* Add support of Blocked layout

* Fix of support U8

* Add support for in/out channels of different sizes

* Removed xfail

* Removed xfail_issue_34327

* BatchToSpace, SpaceToBatch layers optimizations

* Fix according to review

* Fix according to review

* 3 different implementations -> 1 implementation

* Fix THROW

* Fix Exception

* Revert Batch_to_Space to parallel_for

* Fix remarks

* parallel_for -> parallel_nt

* Fix remarks

* Fix data separation

* Fix according to review
2021-04-19 10:15:59 +03:00
Chenhu Wang 1f145923f5
fusion_on_sse41_fix (#5114) 2021-04-16 11:04:49 +03:00
Alexandra Sidorova 37893caa36
[CPU] Added improvements for StridedSlice (#4248) 2021-04-08 13:17:29 +03:00
Alexandra Sidorova 325b5564f3
[CPU] Added GELU-7 support (#4800) 2021-04-07 16:01:47 +03:00
Alexandra Sidorova 0cfd007206
[CPU][IE TESTS] Covered fusing of HSwish, Mish to Conv by tests (#4903) 2021-03-26 10:48:54 +03:00
Alexandra Sidorova 54f60ee761
[CPU] Fixed fusing Convolution with Sum (#4617) 2021-03-26 10:26:57 +03:00
Alexey Suhov a748c26fee
Align copyright notice in cpp and cmake source files (CVS-51320) (#4950) 2021-03-25 02:40:09 +03:00
Anton Pankratv 3b3d9a0989
Used IE_THROW macro (#4869)
* Added ie throw macro

* Used IE_THROW macro
2021-03-23 18:57:12 +03:00
Gorokhov Dmitriy 48144860d5
[CPU] Fixed CPU specific Interpolate test for BF16 precision (#4420) 2021-02-19 15:47:50 +03:00
Gorokhov Dmitriy bd431b2565
[CPU] [WA] Enforce linear_onnx mode in Intepolate node to avoid perf regressions (#4395)
- Fixed incorrect Interpolate fusing behavior for unsupported configurations
2021-02-19 10:13:35 +03:00
Maksim Kutakov 24aeb16fd1
[CPU BF16] OneHot layer extended with bf16 support. (#4171) 2021-02-17 10:52:23 +03:00
Maksim Kutakov 274a70bd14
[CPU BF16] BF16 support was added to the SoftMax primitive. (#4279) 2021-02-16 09:41:21 +03:00