* IntelGPU backend: Enable nBench statistic

* PR1848. Combine counters type time

* remove unused setting

* cleanup

* cleanup

* cleanup

* more cleanup

* update sphinx reqs

* update sphinx reqs

* [Py] Apply PEP8 code style to setup.py

* Remove unused var

* Update version to 0.9.0

Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

Partial Shapes and Types, Part 4a: Implement partial shape/type validation for some existing ops (#1756)

* Enums should be moved to public to access

* fix style check

* Added the capability of directly compiling cuda and ptx to CMake.

* Added comment to show example use case for cuda source file.

* Minimum compute/sm is 35 in order to support PTX intrinsics

* Updated CMake and added example nvcc kernel.

* Made cudart link statically.

* Added compute/sm capabilities for code gen in nvcc based on either cuda 8 or 9.

* Header update.

* Fix cmake indentation.

IntelGPU backend: GNMT training enabling: Added Negative operation functionality for i32 data type (#1845)

* Adapt Tensor class to have partial shapes

* Add PartialShapes to Input, Output, Function, Node classes

* Terminological cleanup

* Add PartialShape propagation for Parameter and Result

* Implement partial-shape propagation for elementwise ops

* More comments

* One more comment tweak

* Add tests for the merge functions

* Add merging of undetermined element types

* Fix a goophup in deserializer implementation

* Implement fallback for ops that do not support partial shape/type validation

* Updates for some older unit tests, now that operator[] exists

* Add missing validate_punt_if_incomplete to AllReduce

* Better docstrings for the stuff introduced in #1692; remove prototype for unimplemented, unused PartialShape::append()

* One more docstring thing I forgot to save

* Switch terminology from 'determined/undetermined' to 'static/dynamic'

* Switch terminology from 'complete/incomplete' to 'static/dynamic' for shapes; fix up some mushily worded comments

* Fix overzealous edits from the last commit

* Rename one test that escaped the Great Renaming

* Remove unnecessary validate_punt_if_dynamic from Reshape

* Show argument types/shapes in long NodeDescription; tank unit tests to block merge

* Fix dynamic element type propagation for elementwise ops, add some unit tests for same

* Roll 'Not' back to existing behavior (non-boolean input types allowed)

* Add a TODO tag to a todo item

* onnx: fix cmake commands list
Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

* onnx: change exception name to UnknownOperator
Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

* onnx: rename ops_bridge class
Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

* onnx: initial opset versions
Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

* onnx: check operator availability
Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

* onnx: opset versions - after reivew changes
Signed-off-by: Artur Wojcik <artur.wojcik@intel.com>

* onnx: fix doxygen comment fromat

* add stub files for quant dequant

* quantize / dequantize doc

* Small doxygen fixes

* Update quantize.rst

* quant docs

* Constant folding with Quantize

* updated with review comments

* [ONNX] Global Pool ops

* Remove is_global flag

* Style fix

* Review fix pt. 1

* Change comments style

IntelGPU backend: ArgMin/ArgMax functionality was fixed; Tests on unsupported functionality were disabled (#1834)

* IntelGPU backend: ArgMin/ArgMax functionality was fixed; Tests on unsupported functionality were disabled

* Additional improvement of ArgMin/ArgMax algorithm

* WIP

* More WIP

* More chiseling

* Move conv validation utils to a separate file; update unit tests

* Fix invalid attributes in pattern containing ConvolutionBackpropFilters

* Remove zero_const_conv test (it's no longer possible to construct the graph being tested)

* Rename infer_convolution_output_item_shape to infer_windowed_reduction_output_shape and add a boolean flag to control whether window-all-in-padding is allowed

* Add generalized function for inferring pooling fprop, use it in AvgPool/AvgPoolBackprop

* Update MaxPool to use new utility functions

* Fix comment

* Remove faulty and redundant check for window shape relative to pre-padding data shape

* Revert change to pattern construction in cpu_fusion

* Update unit test for maxpool

* Restore unjustly eliminated tests; move some computation to ptrdiff_t for safety; fix wording on some error messages

* Formatting

* reset buffer size, use original input size for memcpy

* resolve comment and add test

* update comment

* Update email notification method

- Link to pull request
- Jenkins build links to Blue Ocean

* Update prepare_environment.sh

- remove artifacts from previous build

* Add type_prop unit tests for Quantize

* Fix naming consistency; add test for unsupported RoundMode fail

* Add type-prop unit tests for Dequantize

* [ONNX] Tool to serialize ONNX models

* Review fix.

* Add back dependencies to onnx_import_interface target.

* Review fix pt. 2

* Check also for errors when opening file stream.

* Switch to scale and offset design from min and max for Quantization

* Remove offset and make the quantize ops a single o/p op

* move cpu QuantOps to core and create builders

* rebase to HEAD

* remove convbias and convbiasrelu ctors which take conv

* remove mistakenly added quantize.rst

* remove offset

* Compute scale, move quantization ops to experimental dir and some PR feedback

* Normalize license headers

* dex core quantize/de

* more types

* codegen

* remove comments

* remove more dead code

* [ONNX] Assert all op types supported

* Apply clang-format

* Address code review comments

* Fix #include statements

* Update ONNX Squeeze Op implementation to conform with doc. Add unit test.

* Apply code-format.

* Correct attribute value type.

* Change used loop structure.

* Modified version of loops.

- Without erase and with minimal computation time complexity.

* Run CI

* Reshape with no permutation optimized out

* PR1671. Apply optimization to the same shapes only

* PR1671. Reuse Reshape optimization from common passes

* PR1671. Remove eliminate_reshape from nop_elimination procedure

Add support for an environment variable `NGRAPH_ONNX_IMPORT_ENABLE` to setup.py
This variable controls whether the Python API is build with or without support for ONNX import functionality built into core nGraph.

* Improved AvgPool unit test coverage. Fixed small bug that was revealed.

* Renamed disabled unit tests to reflect new names.

* Ran clang-format on backend_test.in.cpp to fix format.

* Renamed cpu_results->backend_results in two unit tests.

* Added Jenkins Job No and URL

* Update prepare_environment.sh

- clone onnx models
- remove wheel build artifacts

* Link onnx models

* Update Jenkinsfile

- Only send email on fail,
- Change ONNX branch warning to info field
- Disable cloning after tox tests

* Update prepare_environment.sh

- Added comments
- link ONNX models directory

* Update cleanup