* auto-batching- bare min of the info
* renaming BATCH.MD to the automatic_batching.md, also aligned the link to the new naming convention
* more info and brushed
* added openvino_docs_OV_UG_Automatic_Batching to the main TOC
* Apply suggestions from code review
Co-authored-by: Tatiana Savina <tatiana.savina@intel.com>
* close on the comments, added the code examples
* Apply suggestions from code review
Co-authored-by: Tatiana Savina <tatiana.savina@intel.com>
* Update example
* Update format
* Update docs format
* added couple of more perf considerations
* more code examples
* Apply suggestions from code review
* Apply the rest from code review
* Update header
Co-authored-by: Tatiana Savina <tatiana.savina@intel.com>
* Remove fp16 of Convert layer test from skip_tests.config.cpp as it works now
* update repo
* add initial op impl check tests
* add op imple check tests
* add op impl check tests
* add rnn cell based ops
* modify lstmsequence
* update rnn cell base op test
* add priorbox, priorboxclustered, proposal
* add ROIAlign to ReverseSequence
* add Roll to ScatterElementsUpdate
* add select to swish tests
* add tensoriterator to variadicsplit test
* temporary block of LSTMCell v1 due to crash in mkldnn
* use ov namespace instead of ngraph as possible
* update indexing of vector array
* update multiple parameter vector
* add loop test
* fix cpplint errors
* fix build error
* Fix in Preprocessing python bindings - add correct default arguments for:
- PreProcessSteps::convert_element_type
- PostProcessSteps::convert_element_type
- InputTensorInfo::set_color_format
Otherwise, python users must always specify optional params
E.g. instead of writing `tensor().set_color_format(ColorFormat.RGB)` python users will have to write `tensor().set_color_format(ColorFormat.RGB, [])`
* Corrected 'help' output
* Exposing 'openvino.runtime.Type.undefined' and use it in 'convert_element_type' documentation
* Fixed Apple install
* Update path to libs in setupvars.sh
* Fix IE_CPACK_RUNTIME_PATH for Apple
* Fix wheels packaging
Co-authored-by: Alexey Suhov <alexey.suhov@intel.com>
* [DOCS] hddl update
include info on hddl and myriad working at the same time
* Update docs/OV_Runtime_UG/supported_plugins/MYRIAD.md
Co-authored-by: Andrey Zaytsev <andrey.zaytsev@intel.com>
* Update HDDL.md
* Update MYRIAD.md
Co-authored-by: Ilya Lavrenov <ilya.lavrenov@intel.com>
Co-authored-by: Andrey Zaytsev <andrey.zaytsev@intel.com>
* Added inputs argument to all compare() function overloads
* Rewritten compare() function for NMS
* Implemented sorting by name of expected outputs
* Implemented sorting by name of actual outputs
* Added accounting for simultaneous dynamism and the need to convert outputs in Template plugin
* Added a separate case to the GetBlob function for correct dimensions
* Rewritten Expected outputs sorting to work correctly on cpuFuncTests
* Fixing code style problems
* Implemented sorting by name of actual outputs for functional tests
* Debug prints removed
* Replacing a raw pointer with a vector
* Fixing code style problems
* Shifting the sorting place Expected outputs
* Added sorting of Expected exits in one more place
* Quality transition to SLT2.0
* Removing unnecessary code after SLT2.0
* Fix soft_nms_sigma argument
* Removing unnecessary parts after SLT2.0
* Remove unnecessary outputs sorting
* Removing parts from the code for debugging
* Fix for NMS
* Trying to make CI green
* Checking test passage without adding convert precision
* Checking CI
* There is an algorithm that adds Convert only if there is f16, fp16 in inputs
* Add Convert Op in cases where inputs are not already installed f32
* Check that the CI will go away if you put everything back
* Revert changes, validate f32 change on ci
* Adding Convert f16-f32 only if there is a function parameter of type f16
* The presence of f16/bf16 as a parameter type is now mandatory to add Convert
* Added prints for params, inputs, outputs
* Logic checking the absence of Convert
* Cosmetic fixes
* Setting the correct value for selected_scores_type NMS-5
* Fix bf
* Increased readability
* Missing parts added
* Removed the static for the vector
* [MO] Clean-up MO cmd-line options
Remove the following Model Optimizer deprecated options that are no longer used for several releases: disable_fusing, disable_gfusing, generate_deprecated_IR_V7,
legacy_ir_generation, keep_shape_ops, move_to_preprocess
Deprecate through CLI the following options for which functionality triggered from POT or automatically: disable_weights_compression, disable_nhwc_to_nchw,
disable_resnet_optimization, finegrain_fusing.
Correct and extend description of each MO option to be printed during model conversion.
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Correct documentation about input shapes
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Perform final corrections in documentation
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Remove legacy_ir_generation overall
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Clean-up tests from deprecated options
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Recover disable_fusing option as deprecated
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Fix keys for static_shape and extensions
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Remove extension key that does not work
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Apply feedback: remove disable_gfusing, correct docs
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Recover disable_fusing option for unit-tests
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Apply feedback for documentation
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Apply feedback about parameters use_legacy_frontend and use_new_frontend
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* DO minor fixes for indentation of MO logs
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Revert log.error for fallback message
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Revert disable_weights_compression parameter for tests
Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
* Fixed some comments about transformations
* Changed transformation guide
* Fixed typo
* Moved transformation doc to extensibility
* Moved images to Extensibility_UG
* Added separate document for each pass
* Added see also section
* Fixed comments
* Checking compatibility between 'pyopenvino' and 'libopenvino' on 'import phase'
This fix is to prevent undefined behavior when user loads OpenVINO from python, but pyopenvino loads different version of 'libopenvino'
This may happen if user has several releases installed and played around PATH/PYTHONPATH environment variables.
In such case, user may have undefined behavior - application may crash in the middle of the usage or use incorrect release.
Fix checks build versions for pyopenvino and ov::get_openvino_version. If mismatch occurs, exception is thrown.
This logic is disabled if user has built OpenVINO locally, experienced developers probably know what they're doing, so if version has 'custom_' prefix - this logic is disabled
* Removed custom logic for CI_BUILD_NUMBER, it is reused from already included version.cmake
* Use addVersionDefines macro
* Update samples and samplers with the new DataLoader format
* Update with utils
* Pylint updates
* Update metric with the exception
* Pylint
* Update with the exception
* Pylint
* Revert index sampler changes
* Update ImageLoader & SimplifiedEngine
* Update with the different solution
* Remove utils
* Pylint
* Remove list wrapping
* Remove list from meta_data
