Ilya Lavrenov [Tue, 21 Jul 2020 15:58:20 +0000 (18:58 +0300)]
CMAKE: added ie_libraries target (#1411)
Ilya Znamenskiy [Tue, 21 Jul 2020 14:45:42 +0000 (17:45 +0300)]
[IE CLDNN] Gemm int8 with slm optimization. Fused ops fix (#1319)
Roman Lyamin [Tue, 21 Jul 2020 12:11:05 +0000 (15:11 +0300)]
[IE TESTS] Added single layer test Mish (#1401)
Anna Khakimova [Tue, 21 Jul 2020 11:19:15 +0000 (14:19 +0300)]
Pre-processing(GAPI): ARM(NEON) integration + Split, Merge, Color conversion kernels on NEON (#1315)
Ilya Lavrenov [Tue, 21 Jul 2020 10:57:56 +0000 (13:57 +0300)]
Split CNNlayer validators (#1403)
* Simplified CNNLayer validators: keep only parsing of arguments
* Added validators as is
* Removed parsing arguments
* Removed Backetize validator
Nikita Kudriavtsev [Tue, 21 Jul 2020 08:36:35 +0000 (11:36 +0300)]
[IE][VPU]: Added Gelu layer (#1195)
* Implement Gelu stage in GraphTransformer
* Disable GeLU decomposition in the nGraph transformations
Ilya Lavrenov [Tue, 21 Jul 2020 03:36:18 +0000 (06:36 +0300)]
Removed addLayer implementation from ngraph impl (#1400)
Ilya Churaev [Tue, 21 Jul 2020 03:02:00 +0000 (06:02 +0300)]
Remove redundant node methods (#1324)
* Remove placement
* Removed validate and infer eltwise
* Remove is eltwise
* Remove support broadcast and decompose
* Removed is_op, is_parameter, is_pattern
* Fixed code style
* Added is_constant and is_output
* Removed is_communicative and is_null
* Fixed code style
* Fixed typo
* Fixed comments
* Fixed typo
* Revert is_parameter, is_output, is_result for OpenCV build
Anna Khakimova [Mon, 20 Jul 2020 17:09:23 +0000 (20:09 +0300)]
Fix for issue (#1335)
Vladislav Volkov [Mon, 20 Jul 2020 15:11:22 +0000 (18:11 +0300)]
Header length for supported model detection is increased (#1363)
iliya mironov [Mon, 20 Jul 2020 14:43:32 +0000 (17:43 +0300)]
Add mish op to ngraph (#1187)
* Add mish op to ngraph
* Update mish op
* Set v4 namespase for tests
* Add mish to cmake
* Add comments for mish op.
* Refactoring code style
* Update version to v1 for Mish op
* Add value propogation test for Mish op
* Refactoring mish op according to review
* Fix mish version
* Update cmake file
* Fix mish value propogation unit test
* Add unit test for mish op
Co-authored-by: Your Name <you@example.com>
Nikolay Tyukaev [Mon, 20 Jul 2020 14:36:08 +0000 (17:36 +0300)]
Doc Migration (master) (#1377)
* Doc Migration from Gitlab (#1289)
* doc migration
* fix
* Update FakeQuantize_1.md
* Update performance_benchmarks.md
* Updates graphs for FPGA
* Update performance_benchmarks.md
* Change DL Workbench structure (#1)
* Changed DL Workbench structure
* Fixed tags
* fixes
* Update ie_docs.xml
* Update performance_benchmarks_faq.md
* Fixes in DL Workbench layout
* Fixes for CVS-31290
* [DL Workbench] Minor correction
* Fix for CVS-30955
* Added nGraph deprecation notice as requested by Zoe
* fix broken links in api doxy layouts
* CVS-31131 fixes
* Additional fixes
* Fixed POT TOC
* Update PAC_Configure.md
PAC DCP 1.2.1 install guide.
* Update inference_engine_intro.md
* fix broken link
* Update opset.md
* fix
* added opset4 to layout
* added new opsets to layout, set labels for them
* Update VisionAcceleratorFPGA_Configure.md
Updated from 2020.3 to 2020.4
Co-authored-by: domi2000 <domi2000@users.noreply.github.com>
Ivan Tikhonov [Mon, 20 Jul 2020 11:15:37 +0000 (14:15 +0300)]
Added default constructor for RNNCellBase, fix conversions (#1370)
Rafal Blaczkowski [Mon, 20 Jul 2020 10:42:52 +0000 (12:42 +0200)]
ONNX CI: add docker image cleanup (#1394)
* change stage name
* add docker image cleanup
Jan Iwaszkiewicz [Mon, 20 Jul 2020 08:31:49 +0000 (10:31 +0200)]
Add getting/setting friendly name for Node wrapper Py API (#1286)
Tomasz Dołbniak [Mon, 20 Jul 2020 08:20:05 +0000 (10:20 +0200)]
Reference implementations for ReduceLogicalAnd & ReduceLogicalOr (#1333)
Jozef Daniecki [Mon, 20 Jul 2020 08:18:03 +0000 (10:18 +0200)]
Add Acosh operator to ONNX importer opset4 (#1351)
Vitaliy Urusovskij [Mon, 20 Jul 2020 07:23:51 +0000 (10:23 +0300)]
[Stress] Integrate compare_memcheck_2_runs call in run_memcheck.py (#1036)
* [Stress] Define Database constant arguments in memcheck_upload.py only
* [Stress] Simplify computations using HashableDict in `compare_memcheck_2_runs`
* [Stress] Add comparison using pandas
Ilya Lavrenov [Sat, 18 Jul 2020 09:42:40 +0000 (12:42 +0300)]
Removed tests for old IR reader (#1368)
Ilya Churaev [Fri, 17 Jul 2020 16:51:04 +0000 (19:51 +0300)]
Removed v0 fused operations (#1263)
Anton Chetverikov [Fri, 17 Jul 2020 13:13:19 +0000 (16:13 +0300)]
ReduceLp operation specification (#1205)
* Add ReduceL2 operation specification
* Add example
* Update operation specification
* Update operation specification
* Update operation specification
* Update operation specification
* Update types in operation specification
Tomasz Socha [Fri, 17 Jul 2020 13:12:11 +0000 (15:12 +0200)]
[nGraph][Py][Tests] Remove backup of old models (#1349)
Alexander Zhogov [Fri, 17 Jul 2020 12:07:58 +0000 (15:07 +0300)]
CODEOWNERS: Add .ci & docs
Andrew Bakalin [Fri, 17 Jul 2020 08:47:37 +0000 (11:47 +0300)]
[IE][VPU][GT]: Use topological order in shape allocation (#1281)
* Some pass creates datas duplicate with a different order from time to time (because of unordered_set usage). It leads to a different order in model->datas() list and affects the shape allocation process which relies on this order.
* Make shape allocation be relied on topological order of datas which is stable and doesn't depend on order datas creation during different passes.
Ilya Lavrenov [Fri, 17 Jul 2020 07:51:56 +0000 (10:51 +0300)]
Fix unsafe use of CPU_ISSET_S macro. (#1357)
Don't increment mapped_idx via prefix increment within the argument of the
potentially unsafe CPU_ISSET_S macro. If the macro is expanded so that the
increment expression is evaluated multiple times, it will return unexpected
results.
While the glibc implementation of CPU_ISSET_S macro seems to be safe, the musl
libc (v1.1.23) version is unsafe and will evaluate the first argument of
CPU_ISSET_S three times.
Co-authored-by: Christian Priebe <cp3213@ic.ac.uk>
Ilya Churaev [Fri, 17 Jul 2020 06:50:06 +0000 (09:50 +0300)]
Remove get_arguments (#1323)
* Removed get_arguments
* Fixed code style
Nikita Kudriavtsev [Thu, 16 Jul 2020 16:54:53 +0000 (19:54 +0300)]
[IE][VPU]: CMX limit compile option (#1268)
In some networks, mvTensor would request a large CMX-DMA transfer (<512K). That starves DMA for other timing critical tasks such as SIPP. Limit CMX-DMA request size as an option in myriad_compile:
* Add compile option TILING_CMX_LIMIT_KB
* Declare compile option TILING_CMX_LIMIT_KB in IE tools (compile_tool and vpu_compile)
* Add tests for compile option TILING_CMX_LIMIT_KB. Small fix for naming behavior tests.
Alexander Zhogov [Thu, 16 Jul 2020 16:29:54 +0000 (19:29 +0300)]
Azure CI: Fix cloning git submodules (#1356)
Andrey Somsikov [Thu, 16 Jul 2020 15:58:26 +0000 (18:58 +0300)]
fix: inference-engine/ie_bridges/python/sample/requirements.txt to reduce vulnerabilities (#1007)
The following vulnerabilities are fixed by pinning transitive dependencies:
- https://snyk.io/vuln/SNYK-PYTHON-NUMPY-73513
Co-authored-by: snyk-bot <snyk-bot@snyk.io>
Irina Efode [Thu, 16 Jul 2020 15:16:57 +0000 (18:16 +0300)]
[IE TESTS] Add PReLu single layer test (#1306)
* [IE TESTS] Add PReLu single layer test
* [IE TESTS] Small refactoring of test infrastructure
* [IE TESTS] Add shape for activation
Rafal Blaczkowski [Thu, 16 Jul 2020 15:09:45 +0000 (17:09 +0200)]
Revert "Header length for supported model detection is increased (#1340)" (#1354)
This reverts commit
5b2ec7840ac3c5c1d1ec107e50e499f22c9fe7a6.
Ilya Lavrenov [Thu, 16 Jul 2020 13:44:48 +0000 (16:44 +0300)]
Remove implicit conversion from getInputTo, getLayerCreator (#1274)
* Added ctor for CNNNetworkImpl to convert from ngraphImpl
* Re-use in all places instead of manual conversion
* Hide convertToCNNNetworkImpl usage
* Removed conversion from getCreatorLayer
* Fixes 2
* Fixes 3
* Fixes 4
* Fixed ieFuncTests
* Fixed more tests
* Fixed LPT tests
* Remove useless test
* Fixed GNA
* Fixed Gleb's comments
* Fixed Core integration tests
* Trying to fix python
* Fixed GPU tests
* Small fixes
* Fixed QueryNetwork after removing implicit conversion
* Fixed Core integration tests for QueryNetwork
* Fixed python; MULTI device QueryNetwork
* Fixed MULTI QueryNetwork
* Removed unused methods
* Enabled LPT FullyConnectedTestModel test
* Fixed typo in python
Vitaliy Urusovskij [Thu, 16 Jul 2020 12:28:25 +0000 (15:28 +0300)]
[Stress] Specify OMZ `develop` branch in get_testdata.py (#1318)
Since in validation OMZ is used from a package (and OMZ is from develop branch), align OMZ default for get_testdata.py
Irina Efode [Thu, 16 Jul 2020 11:20:08 +0000 (14:20 +0300)]
[IE TESTS] Small refactoring of test infrastructure (#1332)
Nadezhda Ageeva [Thu, 16 Jul 2020 10:53:43 +0000 (13:53 +0300)]
Allow python benchmark_app load onnx model (#1283)
Roman Kazantsev [Thu, 16 Jul 2020 09:36:15 +0000 (14:36 +0500)]
Specify operation CTCLoss-4 (#1189)
* Specify operation CTCLoss-4
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Correct documentation for CTCLoss after #1 review
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Correct documentation for CTCLoss after #2 review
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Correct documentation for CTCLoss after #3 review
* Correct documentation for CTCLoss after #4 review
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Correct layout for logits and add more description for unique attribute
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Correct types for length and indices tensors
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Correct formulas and punctuation
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
Alexander Chaiko [Thu, 16 Jul 2020 08:54:24 +0000 (10:54 +0200)]
[IE CLDNN] Adjustment of layouts to choose optimal deconvolution (#781)
Vladislav Volkov [Thu, 16 Jul 2020 08:12:40 +0000 (11:12 +0300)]
Header length for supported model detection is increased (#1340)
Anton Pankratv [Thu, 16 Jul 2020 07:33:22 +0000 (10:33 +0300)]
Added default multi threaded configuration (#1310)
* Added default multythreaded configuration
* Fixed typo
Nikita Kudriavtsev [Thu, 16 Jul 2020 07:29:56 +0000 (10:29 +0300)]
[IE Myriad] Remove Myriad 2 from supported devices in XLink (#1331)
Ilya Churaev [Thu, 16 Jul 2020 03:03:59 +0000 (06:03 +0300)]
remove nGraph deprecated methods (part 1) (#1314)
* Remove remove_goe
* Remove traverse_nodes
* Removed deprecated constructors
* Removed deprecated tensor methods
* Fixed IE build
* Fixed code style
Gladilov, Gleb [Wed, 15 Jul 2020 19:21:19 +0000 (22:21 +0300)]
[IE][VPU][nGraph]: Enables merging subsequent DSR operations (#1326)
Myriad plugin treats DSR operation in a way removing such operations
and connecting inputs with each other (replacing output with one of them).
Semantic of connection is one inputs contains shape of another.
Since the same data object can have exactly one shape it's prohibited
to have DSR inputs connected with another data objects
(the only allowed exception is inputs that are already connected between
each other).
As a result of nGraph -> CNN conversion some operations could be optimized
out which in turn could lead to subsequent DSR operations where each has
its own shape sub-graph. Even if shape sub-graphs are identical it's not
visible to plugin that sees incorrect inputs (inputs of DSR are already
connected, but now with each other, when second DSR is parsed).
To overcome such issue (the reason is when operations are optimized out,
their shape sub-graphs are still there), additional ngraph
transformation should be introduced to merge subsequent DSR into single
DSR operation.
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
Gladilov, Gleb [Wed, 15 Jul 2020 19:17:36 +0000 (22:17 +0300)]
[IE][VPU][nGraph]: Fixes Reshape's shape infer method (#1327)
Previously, if Reshape had input pattern with values [0, -1] - it
propagated dynamic shape through a function. At the same time,
taking "0" and "-1" interpretation into consideration, it turns out
in such cases we could just propagate the same input shape in case of
2D input.
For Faster-RCNN this fix makes static dimensions on dynamic paths static.
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
Gladilov, Gleb [Wed, 15 Jul 2020 16:43:31 +0000 (19:43 +0300)]
[IE][VPU][nGraph]: Fixes StridedSlice DTS (#1328)
* In case of Begin/End/Stride inputs of StridedSlice have rank less
than input data rank - remaining dimensions must be kept unchanged.
* Previous, implementation had UB in such cases - out of bound
vector element access
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
Jan Iwaszkiewicz [Wed, 15 Jul 2020 13:32:24 +0000 (15:32 +0200)]
Add Input and Output class to Py API (#1284)
Adam Osewski [Wed, 15 Jul 2020 12:02:18 +0000 (14:02 +0200)]
[ONNX] Exception handling refinements. (#1266)
Andrew Bakalin [Wed, 15 Jul 2020 10:48:57 +0000 (13:48 +0300)]
[IE Common][Tests] saturated_cast: refactoring & tests (#1304)
* [IE Common] Refactor saturated_cast
* [IE Common][Tests] Add tests for saturated casts
* [IE Common] Review fixes
* [IE Common] Make enable_if check a template parameter
Jan Iwaszkiewicz [Wed, 15 Jul 2020 10:48:53 +0000 (12:48 +0200)]
Reverse Sequence code clean up (#1303)
iliya mironov [Wed, 15 Jul 2020 07:30:33 +0000 (10:30 +0300)]
Added mish layer doc opset (#1149)
* Added mish layer doc opset
* Refactoring mish spec
* Update mish spec
* Change output description of Mish layer
* Fix Mish according to review
* Refactoring Mish and GELU spec according to code review
* Update formula for ops in spec
* Refactoring spec text
* Update Mish opset
* Change Mish version from 1 to 4
* Sort opset4
Co-authored-by: Your Name <you@example.com>
Vladimir Gavrilov [Wed, 15 Jul 2020 07:27:56 +0000 (10:27 +0300)]
Specify, review and approve operation Interpolate-4 (#1035)
* Added documentation for Interpolate-3.
* Some fixes.
* Fixed some typos.
* Now Interpolate-3 is Interpolate-4.
* Fixed typo.
* DEleted unused 'mode' 'area'.
* Fixed some typos.
* Now 'axes' attribute is an input of Interpolate.
* Added description of variants of nearest_mode.
* Added descriptions of coordinate transformation modes.
* Now 'axes' is an optional input.
* Fixed typo.
Anton Zaytsev [Tue, 14 Jul 2020 20:55:32 +0000 (23:55 +0300)]
[IE TESTS] Add ShapeOf SingleLayerTest (#1285)
* [IE TESTS] add single layer test ShapeOf
* [IE TESTS] update for master
* [IE TESTS] add subgraph test
* [IE TESTS] update todo in skip_tests_config
* [IE TESTS] update skip_tests_config
* [IE TESTS] update skip_tests_config
* [IE TESTS] update opset3
Alexey Suhov [Tue, 14 Jul 2020 19:06:55 +0000 (22:06 +0300)]
update system requirements (#1321) (#1322)
* update system requirements
* update release version in readme
Maxim Shevtsov [Tue, 14 Jul 2020 16:15:52 +0000 (19:15 +0300)]
Fix that brings back the MULTI's ability to add/remove devices (to the priorities list) on the fly. Presumably was lost during refactoring. (#1309)
the point is that we should check the ORIGINALLY (largest) list of the devices (actually ExecutableNetworks for them) to see if the device is just added back
Edward Shogulin [Tue, 14 Jul 2020 15:55:06 +0000 (18:55 +0300)]
[LPT] FuseFakeQuantizeAndScaleShift transformation for last layer fix (#1291)
* [LPT] FuseFakeQuantizeAndScaleShift transformation for last layer fix
* [LPT] refactoring
* [LPT] FuseFakeQuantizeAndScaleShift test: last layer name validation was added
Jedrzej Hajduczenia [Tue, 14 Jul 2020 15:00:51 +0000 (17:00 +0200)]
[IE CLDNN] WA to use bfyx format if applicable for bf(w)zyx Gather input (#1056)
Jedrzej Hajduczenia [Tue, 14 Jul 2020 14:59:20 +0000 (16:59 +0200)]
[IE CLDNN] Disable inserting reorders if num_dims mismatch (#1023)
Rafal Blaczkowski [Tue, 14 Jul 2020 12:57:27 +0000 (14:57 +0200)]
Add OpenVINO-ONNX CI check (#688)
Andrew Bakalin [Tue, 14 Jul 2020 10:13:56 +0000 (13:13 +0300)]
[IE][VPU][Tests]: Fix M2 on deprecated tests for StridedSlice (#1300)
* Disable reorder in Myriad2 cases in StridedSlice deprecated_tests
Adam Osewski [Tue, 14 Jul 2020 08:55:07 +0000 (10:55 +0200)]
[ONNX] Quantize linear using FakeQuantize (#1169)
Adam Osewski [Tue, 14 Jul 2020 08:54:39 +0000 (10:54 +0200)]
Enable importing of TF_NASNet_Mobile (#1252)
Ilya Churaev [Tue, 14 Jul 2020 07:27:51 +0000 (10:27 +0300)]
Removed max pool v0 (#1277)
* Removed MaxPool v0
* Removed atan2
* Removed and operation
Ilya Churaev [Tue, 14 Jul 2020 07:27:10 +0000 (10:27 +0300)]
Added U32 precision (#1297)
Mikhail Letavin [Mon, 13 Jul 2020 15:33:32 +0000 (18:33 +0300)]
[IE CLDNN] Optimize kernel cache memory usage in GPU plugin (#1233)
Gladilov, Gleb [Mon, 13 Jul 2020 15:19:05 +0000 (18:19 +0300)]
[IE][VPU][nGraph]: Enables dynamic Reshape with non-const pattern support in myriad plugin (#1159)
* [IE][nGraph]: Introduces PartialShape ctor from values vector
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
* [IE][VPU][nGraph]: Moves evaluateTargetShape to common utilities
The same functionality - get upper-bound shape estimation for dynamic
input - is needed in dynamic Reshape along with dynamic Broadcast.
Return value type has been changed from PartialShape to vector<int64_t>.
The reason is Reshape encodes special values (0, -1) into input values
that define output shape. Representing those values (which upper-bound
provides evaluateTargetShape) as PartialShape leads to incorrect
representation vector with -1 as dynamic shape - which is not expected.
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
* [IE][VPU][nGraph]: Introduces StaticShapeReshape
In comparison with original Reshape StaticShapeReshape propagates
upper-bound shape through a function in case of dynamic input. To do so,
shape inference method gets upper-bound shape from evaluateTargetShape,
decodes special values (0, -1) in it and then propagate the result.
Output shape processing happens only once, because if shape inference
were called after ShapeOf operations have been optimized out on dynamic
path, then evaluateTargetShape will require evaluate method for all
operations that appear in function before current Reshape. Since
evaluate method is implemented not for all operations it lead to
Faster-RCNN compilation error.
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
* [IE][VPU][nGraph]: Updates Reshape DTS on StaticShapeReshape
In case of non-const Reshape input that defines output shape DTS uses
StaticShapeReshape which propagates upper-bound shape evaluated from
this input through a function.
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
* [IE][VPU][nGraph][Tests]: Refactoring DTS Reshape tests
The only changes are:
* header files include reordering
* indentation/wrapping fixing
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
* [IE][VPU][nGraph]: Moves ShapeOf transformation out of DTS scope
In comparison with DTS ShapeOf transformation needs to work on whole
function. Separating these 2 transformations makes testing easier since
now it's possible to call specific DTS without ShapeOf transformation
and vice versa.
Also DynamicToStaticShapeOf has been renamed into
EliminateShapeOfAfterDSR since transformation doesn't introduce new DSR
operations.
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
* [VPU][Tests]: Introduces DTS Reshape tests with non-const pattern
New StaticShapeReshape constructor has been added as well, since test
fixture should create it from reshape parameters, not reshape itself.
Signed-off-by: Gladilov, Gleb <gleb.gladilov@intel.com>
Maxim Andronov [Mon, 13 Jul 2020 14:40:24 +0000 (17:40 +0300)]
fix strided slice neg out of bounds ends (#1177)
Anastasia Kuporosova [Mon, 13 Jul 2020 11:48:40 +0000 (14:48 +0300)]
[IE Samples] Add api arg to classification sample (#943)
Egor Churaev [Mon, 13 Jul 2020 10:25:15 +0000 (03:25 -0700)]
[IE CLDNN] LRN int8 fsv16 optimizations (#814)
JIRA: 32367
Maxim Andronov [Mon, 13 Jul 2020 09:48:00 +0000 (12:48 +0300)]
[CPU] Add check quantize ranges (#850)
Andrey Sokolov [Mon, 13 Jul 2020 08:49:20 +0000 (11:49 +0300)]
[IE][VPU]: adjust batch - support dynamic number of iterations (#1114)
* support dynamic number of iterations in "AdjustBatch" pass
* add unit tests for this case
Ilya Lavrenov [Mon, 13 Jul 2020 03:13:59 +0000 (06:13 +0300)]
Throw special exception if IR v7 is passed, but no IR v7 reader (#1293)
Vladislav Volkov [Mon, 13 Jul 2020 03:11:02 +0000 (06:11 +0300)]
Optimisations for binary operations broadcast. Phase 2. (#1295)
Ilya Churaev [Mon, 13 Jul 2020 03:02:20 +0000 (06:02 +0300)]
Remove old Scatter operations (#1265)
Andrew Bakalin [Fri, 10 Jul 2020 11:32:49 +0000 (14:32 +0300)]
[IE][VPU][GT]: Process StridedSlice stage on device as one kernel (#1244)
* Remove replacement of StridedSlice with other stages and execute it on device as one kernel.
* Refactor strided slice tests to be able to parametrize it by precision.
* Update firmware.
Gorokhov Dmitriy [Fri, 10 Jul 2020 10:56:22 +0000 (13:56 +0300)]
[TESTS] Added Comparison and Logical single layer tests (#1242)
Ilya Churaev [Fri, 10 Jul 2020 10:49:43 +0000 (13:49 +0300)]
Removed adjoints (#1269)
Bartosz Sochacki [Fri, 10 Jul 2020 10:39:29 +0000 (12:39 +0200)]
[GNA] Support in GNA plugin for power layer with non-1 exponents (#997)
* added support for power layer with non-1 exponents to GNA plugin
* reverted a change caused by merge issue
* fixes for review comments (typo fix - lrelu instead of leru, unnamed structure instead of of named one in union with arguments of activation function, name fix - input instead of inputs),
scale-shift implementation based on affine layer instead of PWL,
* fixed code style
* fixes for coding style in scale_factor_calc.hpp
* added domain for power function
* fixed review comment - power function specific methods
* added check if dynamic casting was successful
* removed I16 as it is not supported by ngraph
* fixed initialization per review comment
Kamil Magierski [Fri, 10 Jul 2020 10:37:12 +0000 (12:37 +0200)]
[GNA] MemoryStateImpl (#839)
* [GNA] fix query state for GNAMemoryState
* [GNA] MemoryState implementation:
Fix shared tests
Fix smoke tests
fix SetState
Implement LastState
Fix Reset
* Move getPrecision() to GNAMemoryState
Change Smoke Reset() test to check resetting one state
* [GNA] add dequantize to getLastState()
* code refactor
Co-authored-by: Anna Alberska <anna.alberska@intel.com>
Co-authored-by: kmagiers <kmagiers@intel.com>
Michał Karzyński [Fri, 10 Jul 2020 09:53:56 +0000 (11:53 +0200)]
Add nGraph-ONNX tests (#1215)
Roman Kazantsev [Fri, 10 Jul 2020 09:29:15 +0000 (14:29 +0500)]
Implement ScatterND operation in MO and transform for SparseToDense (#584)
SparseToDense used in Wide and Deep model is expressed through ScatterND operation.
ScatterND is more functional than SparseToDense. Hence, it was decided to replace SparseToDense
with ScatterND. ScatterND is more useful for other models.
Remove SparseToDense from the previous opset
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
Ilya Lavrenov [Fri, 10 Jul 2020 08:22:49 +0000 (11:22 +0300)]
Simplified usage of CNNNetworkIterator (#1260)
Andrew Bakalin [Fri, 10 Jul 2020 08:20:17 +0000 (11:20 +0300)]
[IE Common] Replace static_cast with saturated one (#1257)
Chenhu Wang [Fri, 10 Jul 2020 08:19:23 +0000 (16:19 +0800)]
[CPU] ScatterUpdate ScatterElementsUpdate and ScatterNDUpdate support (#909)
* scatter_update_series_enable
* scatter_update_series_enable
* add single layer tests
Roman Lyamin [Fri, 10 Jul 2020 07:49:50 +0000 (10:49 +0300)]
[IE TESTS] Added single layer test "Ceiling" (#1271)
Maksim Doronin [Thu, 9 Jul 2020 16:21:18 +0000 (19:21 +0300)]
[IE][VPU]: Remove dataToShape edges from unsued data (#1213)
* Check equality of shape data for the replaced and replacement input/output data in the model
* Connect data with shape in duplicateData method
* Disconnect shape with data which is being removed as unsued.
* Check that disconnected shape still have child dataToShape edges or consumers
* Refactor cleanUp to use removeUnsuedData and not duplicate code
Anastasia Kuporosova [Thu, 9 Jul 2020 15:09:06 +0000 (18:09 +0300)]
Add python API codeowners (#1254)
Roman Lyamin [Thu, 9 Jul 2020 13:57:59 +0000 (16:57 +0300)]
[IE CLDNN] Added Mish operation (#1125)
Tomasz Dołbniak [Thu, 9 Jul 2020 13:52:25 +0000 (15:52 +0200)]
Fix the __repr__ function for dynamic shapes (#1210)
Anton Chetverikov [Thu, 9 Jul 2020 13:50:15 +0000 (16:50 +0300)]
Swish operation specification (#1200)
* Add Swish operation specification
Anton Chetverikov [Thu, 9 Jul 2020 13:48:59 +0000 (16:48 +0300)]
SoftPlus operation specification (#1128)
* Add SoftPlus operation specification
Tomasz Socha [Thu, 9 Jul 2020 13:31:04 +0000 (15:31 +0200)]
[nGraph] Fix reshape builder if target shape is scalar (#1206)
Tomasz Socha [Thu, 9 Jul 2020 13:30:17 +0000 (15:30 +0200)]
[nGraph] Allow to use protobuf lite in onnx importer (#687)
Tomasz Dołbniak [Thu, 9 Jul 2020 10:19:40 +0000 (12:19 +0200)]
ONNX Imagescaler op (#1262)
* ONNX ImageScaler op
* UT for ImageScaler op
Mateusz Bencer [Thu, 9 Jul 2020 09:26:00 +0000 (11:26 +0200)]
Enable importing TinyYOLOv2 (#1121)
* Fixed bugs in TinyYOLOv2 ops
* Added tests
* styles applied
* styles applied
* code review remarks introduced
* code review remarks (unit tests added)
iliya mironov [Thu, 9 Jul 2020 08:28:47 +0000 (11:28 +0300)]
Move arithmetics spec from v1 to v4 spec docs (#1230)
* Move atanh asinh acosh to opset4
* Update opset3 tables
Roman Lyamin [Thu, 9 Jul 2020 07:22:34 +0000 (10:22 +0300)]
[IE TESTS] Added single layer tests (#1137)
Vladimir Zinoviev [Thu, 9 Jul 2020 07:19:52 +0000 (10:19 +0300)]
[LPT] Eltwise Prod transformation fix (#1135)
* [LPT] Eltwise Prod transformation fix
* [LPT] ngraph Multiply lp transformation test
Ilya Lavrenov [Thu, 9 Jul 2020 03:13:20 +0000 (06:13 +0300)]
Fixed RTTI issues for ngraph::Variant (#1258)
Ilya Churaev [Thu, 9 Jul 2020 03:09:28 +0000 (06:09 +0300)]
Use ngraph.hpp in samples and documentation (#1240)
Anton Zaytsev [Wed, 8 Jul 2020 17:59:24 +0000 (20:59 +0300)]
[IE TESTS] dynavic batch for mvn layer (#1010)
* [ci-skip][IE TESTS] dynavic batch for mvn layer
* update instance v0
* [ci-skip][IE TESTS] update instance for mvn layer
* [ci-skip][IE TESTS] fix
* [ci-skip][IE TESTS] add dynamic batch for singleLayer basic class
* [ci-skip][IE TESTS] update dynamic batch for singleLayer basic class
* [ci-skip][IE TESTS] removing bathFlag
* [IE TESTS] removing bathSize
* [IE TESTS] refactor dynamic batch in basic class
* [IE TESTS] refactor dynamic batch in basic class
Ilya Lavrenov [Wed, 8 Jul 2020 14:26:37 +0000 (17:26 +0300)]
Conversion via CNNNetworkImpl ctor (#1222)
* Added ctor for CNNNetworkImpl to convert from ngraphImpl
* Re-use in all places instead of manual conversion
* Hide convertToCNNNetworkImpl usage
* Remove useless test
* Fixed Gleb's comments