* Operation class review
* Remove clamp from FusedOp
* Add node validation checks
* Add type_prop unit tests
* Move backend unit tests into a separate file
* Remove unnecessary includes
* Add support for i8, i16, u8, u16 and bf16 precisions
* Refactor backend unit tests
* Remove instance with f64 precision
* Move dynamic shape tests to op_eval
* Group failing tests into manifest
* Add single layer tests and serialization tests
* Skip tests with conversion due to plugin misbehavior
* Add unit tests fails to manifest with documented issues
* Fix centos7 test to skip slt cpu instances of integer precision
* Enable dynamic test for float element type
* Address review comments
* change docker approach
* Experiment with run onnx runtime tests over MS Azure
* experiment to run bnaries directly
* enable all tests
* Set tests execution as separated scripts
* Configurations update:
- Add new yml file for onnx_runtime check
- Add specific tag to checkout for onnx runtime
* Rename according to conventions
* Change OpenVINO release version
* [Hetero plugin] Model caching support
- Enable IMPORT_EXPORT_SUPPORT metric
- Introduce internal FORCE_DISABLE_CACHE flag to avoid caching of subnetworks
- Added test for hetero with plugins which don't support caching
- Added test for hetero with plugins supported different cache architecture
* Hetero plugin - support DEVICE_ARCHITECTURE
Test setup:
mock.1 - mock.9 returns "one" for DEVICE_ARCHITECTURE
mock.10 - mock.99 returns "two"
Test:
Load "HETERO:mock.1,mock.51".
Load "HETERO:mock.2,mock.52" - cache shall be reused
* Fixed review comments
Covered use case case
ie.SetConfig({{"TARGET_FALLBACK", "CPU"}}, "HETERO");
ie.LoadNetwork(network, "HETERO");
* Fixed more comments and failed tests
Don't propagate FORCE_DISABLE_CACHE to plugins as they can throw exception
Fixed case with set TARGET_FALLBACK from core with different architectures of one plugin
* Fix unit tests
Add 'FORCE_DISABLE_CACHE' config key only for LoadExeNetwork
It is not needed to have in in QueryNetwork and other places
* Attempt to fix failed func test on Windows
* Update benchmark_app to pass precision via command line
* Update vpu_perfcheck
* Update python benchmark_app to support setting precision from cmd
* Review comments
* Address more review comments
* Fixes after rebase
* [MO] Split iteration node for TensorIterator in case multiple consumers
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Implement additional BackEdgeTensorIterator and NonConstBeginStridedSlice transformations
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Finalize implementation of DIEN support by the MO tool
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Add unit test for NonConstBeginStridedSliceReplacement transformation
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Fix retrieving shrink_axis_mask attribute value
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Address the majority of review feedback
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
