Commit Graph

522 Commits

Author SHA1 Message Date
Andrey Somsikov
2ed596c87a
Use default thread sanitizer linkage (#833)
GCC and CLang *default* sanitizer linkage differs (static vs. dynamic).
Prefer default behavior as alternate seen having issues.

Default (GN)U linker fails with unresolved symbols linking Clang built
binaries with sanitizer enabled. Force use LLVM linker lld for Clang
builds.

Sanitizer instrumentation and link flags should be retained for all
binaries. Updating samples cmake configuration to keep those flags
after unset logic at the ie_build_samples().
2020-06-09 20:11:44 +03:00
Evgeny Lazarev
17ca27e4ab
Fixed StridedSlice to Crop transformation (#836)
* Fixed StridedSlice to Crop transformation to not apply when rank of data is changed

* Added unit test for StridedSlice to Crop transformation

Co-authored-by: Evgeny Lazarev <elazarev.nnov@gmail.com>
2020-06-09 20:00:43 +03:00
Roman Lyamin
3b4990ed30
[IE CLDNN] Added batch_to_space operation (#753) 2020-06-09 19:19:24 +03:00
Gleb Kazantaev
4c62786499
Fix divide conversion for integer input type (#841) 2020-06-09 19:13:35 +03:00
Anastasia Kuporosova
f278509236
[Python API] Fixate requirements (#834) 2020-06-09 18:06:57 +03:00
Ilya Lavrenov
f518fbb971
Fixed default args for Android build (#827) 2020-06-09 18:02:03 +03:00
Maxim Vafin
074266bf73
Fix onnx slice by clipping ends to int32 domain (#603) 2020-06-09 17:50:38 +03:00
Andrey Dmitriev
0bf1f53356
[GNA] Added support permute layer (#723)
* [GNA] Added GNA natively supported permute layer cases.
2020-06-09 16:43:01 +03:00
Maksim Doronin
18004bdb5e
[IE VPU] Dynamic Broadcast tests (#737)
* [IE VPU] Enable StaticShapeBroadcast tests

* [IE VPU] DSR: support case when shape is output and input for stage

* [IE VPU] Enable Broadcast and Transpose tests

* [IE VPU] DSR: fix typo

* [IE VPU] Add assertion for numConsumer in DSR

* [IE VPU] Added CheckMyriadX helper method

* [IE VPU] New DSR assert for input->getInputTo

* [IE VPU] Fix myriad2 tests bug
2020-06-09 16:10:12 +03:00
Ivan Tikhonov
9fc818478a
Const folding and reference implementation for PriorBox(Clustered) ops (#785)
* Constant folding for PriorBox, PriorBoxClustered; Deleted PriorBoxIE, PriorBoxClusteredIE and transformations; Added unit tests; codestyle

* Delete debug info

* delete unnecessary convert_prior_to_ie_prior.hpp file

* fix ngraph reader tests; delete PriorBoxIE functional test

* fix for ngraph reader tests

* Apply review comment

* apply ngraph codestyle

* restore PriorBoxClustered tests in disabled state
2020-06-09 14:47:49 +03:00
Denis Orlov
ef8a8dd309
add support for multiple scale factors in speech sample (#835)
Co-authored-by: Anna Alberska <anna.alberska@intel.com>
2020-06-09 14:36:28 +03:00
Andrey Sokolov
d4e880de3d
[IE VPU] Update firmware; enable convolution VPU OCL tests (#802) 2020-06-09 14:34:10 +03:00
Vladimir Paramuzov
fe198dd544
[IE CLDNN] Added 6d tensor support in eltwise/scale primitives (#826) 2020-06-09 14:29:36 +03:00
Anton Zaytsev
b0eb3e67ee
[ci-skip][IE MKLDNN] Add Precision U16 in MKLDNN (#783) 2020-06-09 14:20:43 +03:00
dmitrygo
434361cea9 [TESTS] fixes after rebase 2020-06-09 14:11:18 +03:00
dmitrygo
aa30580109 [CPU] mkldnn submodule up 2020-06-09 14:11:18 +03:00
dmitrygo
051a429c31 [LPT] Fixed quantizeBlob routine for 3D case 2020-06-09 14:11:18 +03:00
Edward Shogulin
8eb88d51f2 [LPT] GPU tests were fixed 2020-06-09 14:11:18 +03:00
Edward Shogulin
971811c8c8 [LPT] [TEST] LayerTransformation test threshold was updated 2020-06-09 14:11:18 +03:00
Anton Voronov
629ca3a5d8 [CPU] Gemm node: supported precisions U8 and I8 and added tests 2020-06-09 14:11:18 +03:00
Edward Shogulin
92e5e010b9 [LPT] FullyConnected & Gemm tests 2020-06-09 14:11:18 +03:00
dmitrygo
c7313bab7f [CPU] Fixed weights candidate initialization in FC node 2020-06-09 14:11:18 +03:00
Edward Shogulin
d798831c95 [LPT] Gemm and FullyConnected 3D improvement 2020-06-09 14:11:18 +03:00
Edward Shogulin
4d01adbe01 [LPT] tests extending 2020-06-09 14:11:18 +03:00
Edward Shogulin
1d51d2185a [LPT] [Test] Low precision transformations functional tests infrastructure improvement 2020-06-09 14:11:18 +03:00
Edward Shogulin
65b00c1dfb [LPT] FullyConnected transformation fix 2020-06-09 14:11:18 +03:00
Edward Shogulin
9758305b32 [nGraph] Remove Reshape for 3D FullyConnected 2020-06-09 14:11:18 +03:00
Edward Shogulin
d7c77212b8 [IE COMMON] [LPT] Concat asymmetric quantization with signed interval fix 2020-06-09 14:11:18 +03:00
Edward Shogulin
e544dd1e28 [IE COMMON] [LPT] Support 3D layout for FullyConnected transformation 2020-06-09 14:11:18 +03:00
dmitry-gorokhov
bc98d17121 [CPU] Added custom implementations (power=0.5, power=-1.0) for Power node 2020-06-09 14:11:18 +03:00
dmitry-gorokhov
bcd38100db [CPU][WA] Supported 3D layout for FullyConnected primitive
Extended jit uni depthwise primitive to support 3D inputs
2020-06-09 14:11:18 +03:00
Nikolay Shchegolev
b6f2c06b26
[Common] Static analyzed issues. (#804) 2020-06-09 13:49:50 +03:00
Vladimir Paramuzov
b4546ad1e0
[IE CLDNN] Better error message when output is not found (#824) 2020-06-09 12:26:28 +03:00
Edward Shogulin
d02b9a9b81
[LPT] [TEST] LayerTransformation test threshold was updated (#828) 2020-06-09 10:34:17 +03:00
Maxim Andronov
d8e82d56d2
[CPU] fix set up config for bin conv fused (#608) 2020-06-09 09:59:29 +03:00
Anastasia Kuporosova
e91453e006
[Python API] Fixate requirements versions (#830) 2020-06-09 08:49:49 +03:00
Anastasia Kuporosova
6a60f93af0
[Python API] Fix deprecation warnings (#812) 2020-06-09 08:48:08 +03:00
Edward Shogulin
ca643edb1b
[LPT] [CPU] NormalizeL2 transformation (#662)
* [LPT] NormalizeL2 transformation

* [LPT] NormalizeL2 transformation tests improvement

* [CPU] Fixed depthwise injector aux_vec_count for broadcasting case

* [LPT] Normalize on GPU enabling

Co-authored-by: Zinoviev, Vladimir <vladimir.zinoviev@intel.com>
Co-authored-by: dmitrygo <dmitry.gorokhov@intel.com>
2020-06-08 22:42:50 +03:00
Pavel Esir
7a11e36eeb
Add fixedscale(bias) components to Kaldi (#725)
* Added fixed scale(bias) components

* Successfully converted after adding fixed bias,scale components

* Added unittests
2020-06-08 21:37:44 +03:00
Mikhail Letavin
155916acde
[IE CLDNN] Fix variable initialization issues (#816) 2020-06-08 21:07:50 +03:00
Nikita Kudriavtsev
ac65ea30fd
[ICV] Watchdog switch + ddr initialization (#554)
* [IE Myriad] Added XLinkBootFirmware method in XLink API for booting firmware buffer

* [IE Myriad] Patch firmware in mvnc. Added test to check device reset without connecting.

* [IE Myriad] Added option MOVIDIUS_DDR_TYPE for Myriad plugin

* [IE Myriad] Added tests for new option MOVIDIUS_DDR_TYPE

* [IE Myriad] Update firmware 1201 -> 1212

* [IE Myriad] Convolution3x3 tests are disabled due to firmware issue. #-32921
2020-06-08 20:51:45 +03:00
Irina Efode
3b5de94a09
[IE TEST] Eltwise tests refactoring (#726)
* [IE TEST] Eltwise tests refactoring

* [IE TESTS] Fix comments
2020-06-08 18:44:42 +03:00
Denis Orlov
ff00817bb7
[GNA] Support changing the execution mode in runtime (#801) 2020-06-08 18:43:12 +03:00
iliya mironov
eefaf56075
Fix unit tests for select layer. (#638)
* Fix unit tests for select layer.
2020-06-08 18:39:40 +03:00
Maxim Vafin
f1811ad060
Implement support for opset3 EmbeddingBag ops (#546)
* [MO] Implement EmbeddingBag_3

* Transform dynamic sub-graph of Wide and Deep into EmbeddingSegmentsSum

- Expressed SparseWeightedSum sub-graph through EmbeddingSegmentsSum
- Removed experimental SparseWeightedSum layer
- Implemented tests for the transformation

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix EmbeddingBag shape infer

* Fix EmbeddingSegmentsSum transformation for Wide and Deep

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix EmbeddingSegmentSum replacer after ports swap

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Update package_BOM.txt

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Add unit tests for EmbeddingXXX shape infer

* Fix ATen resolver

* Remove deleted files from BOM

* Add opset version to embedding_bag

* Use base class for EmbeddingBag

* Fix per_sample_weights case

* Fix EmbeddingSegmentsSum transformation

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix EmbeddingBag checks

* Fix ATen front transformation and merge conflicts

* Fix BOM

* Work around limitation for I64 input of W&D model

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Cleanup where operation to fix affect of WhereDecomposition transform

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix BOM

* Correct EmbeddingSegmentSum transform for Wide and Deep

Add casting segment ids to i32 and remove ConstToResult sub-graph.

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Update BOM with RemoveConstToResult transform

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Add more comments for RemoveConstToResult transformation

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Remove useless logging in EmbeddingSegmentsSum transformation

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Small fixes

* Move EmbeddingBag resolving back to front phase

* Improve error messages

* Fix typo in unittests

* Reimplement sparse_reshape middle transform

Avoid deprecated API.

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Clean-up graph after sparse_reshape and ConstToResult transformation

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix clean-up for transformations

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix clean-up for transformation #2

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>
2020-06-08 18:06:40 +03:00
Konrad Dobros
d155483573
[IE CLDNN] Optimize 1x1 imad convolution kernel (#757) 2020-06-08 16:44:50 +03:00
Andrey Somsikov
626bc4f3d4
Add commit links to memcheck report (#820) 2020-06-08 15:08:58 +03:00
Tomasz Dołbniak
60d4d62536
Disable warnings-as-errors for ONNX target (#749)
* Disable warnings-as-errors for ONNX target

* Disable warnigs-as-errors for windows too

* Change WIN32 -> MSVC
2020-06-08 13:52:45 +03:00
Evgenya Stepyreva
e7f5f53f92
[ MO ] Groupped conv fusion (#797)
Fixed the group convolution fusion pass to properly get the feature dim in NCHW layout case.
2020-06-08 13:00:54 +03:00
Edward Shogulin
a224078c5c
[LPT] [Test] DepthToSpace sporadic fail fix (#815) 2020-06-08 12:55:37 +03:00