* Update Swish OP description.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Use RTTI to declare/define NGraph Swish OP.
Add input element type check when constructing Swish OP.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Add Swish into activation serialization test list.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Add Swish into IE CPU plugin activation single layer test suit.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Add Swish NGraph backend and visitor API tests.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Add Swish unsupported parameter data type test cases.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Update the Swish OP visistor API to use typed test.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* change threshold to 20 instead of max limit of data type
* add invalid input type test case
* add invalid input data check
* add input type dynamic check
* add backend test case
* add more clarity on spec and align with real implementation
* Revert "[CPU] Fix for CoreThreadingTestsWithIterations tests (#5892)"
This reverts commit 115aa143ef.
* Revert "Revert "[CPU] Fix for CoreThreadingTestsWithIterations tests (#5892)""
This reverts commit 95afa50d94.
* Revert "change threshold to 20 instead of max limit of data type"
This reverts commit 91af825056.
* Revert "add more clarity on spec and align with real implementation"
This reverts commit a3b232a8fb.
* add visitor attribute test case
* Revert "add visitor attribute test case"
This reverts commit 610728f1ab.
* add attribute test case
* revise the attribute visitor test per parametrized visitor API PR: #6181
* update spec
* add RTTI macro
* create visitor test
* clean backend test file
* add negative to type_prop tests of unary_ops
* add validate_and_infer_types method
* remove negative from type_propp unary_ops
* create negative type_prop tests
* remove boolean check from validate_and_infer_types
* update supported types in spec
* Remove check for signed numeric types
* Remove type_prop test cases with non-signed type
* Add zero sized negative test to interpreter manifest
* Revert "update supported types in spec"
This reverts commit a7ac3e4dc8.
* Revert "update spec"
This reverts commit 4c431d753d.
* Add minor changes to review op class
* Add validation of inputs and outputs in evaluate
* Use shape from host tensor to calculate element count
* Minor changes in variable names of backend tests
* Add SLT case with integer precision
* Add operation to list of trusted ops
* Address review comments:
* Add bf16 precision to evaluate method
* Use unary ops typed case tests for type_prop tests
* Add end of line to interpreter manifest
* Refactored visitor unit test
* Fix typo in visitor test name
Co-authored-by: ggalieroc <gabriele.galiero.casay@intel.com>
* [NG Frontend] Introduce classes and macros for Frontend-specific exceptions
* [nGraph Frontend] Python bindings
Exposing Python API for nGraph FrontEndManager introduced in PR #5470
* [nGraph Frontend] Python bindings
Exposing Python API for nGraph FrontEndManager introduced in PR #5470
* Added docs + moved sources to 'pyngraph/frontend'
* Fix issues after rebase
* Fix clang
* Moving ngraph/python/tests_cpp to ngraph/python/tests/mock
* Fix cmakelists-related comments
* Apply review comments
* Added dots to docstrings
* Cmake - Set target properties for ARCHIVE/PDB/COMPILE_PDB as well as for LIBRARY_OUTPUT_DIRECTORY
* Fix unknown command line flag errors if -sample-options is used
* Fix raw escape sequences output with dashes
* Fix raw escape sequences output with run_again message
* try change sampleoptions to array
* additional checks are added to prevent errors from occurring on the GNA library side
* fixing after review
* move AreLayersSupported from Cnn2D namespace
* GNA_DEBUG is commented
* batch size <= 8 test was added
* Correct output nodes naming in scope of NMS transformation
Details:
NMS transformation changes name of NMS node output (adds '/convert' suffix)
In case NMS transformation is the last node before Output, it leads to
the situation when InferRequest's network Outputs do not match any
graph's Output nodes.
Tickets:
- 57695
* Add test
* Migrate the unary OP visitor to typed template test.
The unaray OP list is :Floor, Log, Result, Sqrt, Squeeze.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Add element precision type into GTEST typed parameter.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Update the license to new version.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Update the OP Ceiling visitor test to use unary typed template.
Signed-off-by: Luwei Zhou <luwei.zhou@intel.com>
* Specification of a new operator called MatrixNMS.
This Op functionally equivalent to NonMaxSuppression-5, with an non max suppression algorithm called matrix_nms, and perform more post-processing phases, and lay out
the detection outputs in the way of PaddlePaddle detection.
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update docs/ops/sort/MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* Update MatrixNMS_8.md
* Update MatrixNMS_8.md
Co-authored-by: Anastasiya Ageeva <anastasiya.ageeva@intel.com>
* [GNA] [WIP] Introduce an option to invoke the QoS feature 56759
* Apply remaining revew (typos) for PR 5741
* Introduce HW_WITH_SW_FBACK
* Add unit test for HW_WITH_SW_FBACK
* Enable HW_WITH_SW_FBACK in speech_sample cpp
* Use perf counters to report number of HW delivered frames to the user (eg speech_sample)
* Update speech_sample.hpp
based on review
* Update gna_config.hpp
Describe special perf counter
* lint fix
* Apply review
* Remove special performance counter
* Add GNA frequency for 6/151 CPU family/model
* Update inference-engine/samples/speech_sample/main.cpp
Co-authored-by: Mikhail Ryzhov <mikhail.ryzhov@intel.com>
* Update main.cpp
Co-authored-by: Mikhail Ryzhov <mikhail.ryzhov@intel.com>
* MulticlassNms-8 spec.
This Op functionally extends NonMaxSuppression-5, to perform more post-processing phases, and lay out
the detection outputs in the way of PaddlePaddle detection.
* Update MulticlassNMS_8.md
Clarify the meaning of "nms_top_k".
* Update MulticlassNMS_8.md
* Update MulticlassNMS_8.md
* Update MulticlassNMS_8.md
* Update MulticlassNMS_8.md
* Update MulticlassNMS_8.md
* Deprecated the get_default_value method of Node class
* Cleanup of the obsolete get_version methods
* Deprecation warnings suppression
* Deprecation of get_default_value for v0 ops
