Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[FrontEnd]enable pdpd ops conversion part4: deformable_convolution, multiclass_nms and matrix_nms. #6619

Closed
wants to merge 122 commits into from
Closed

[FrontEnd]enable pdpd ops conversion part4: deformable_convolution, multiclass_nms and matrix_nms. #6619

wants to merge 122 commits into from

Conversation

ceciliapeng2011
Copy link
Contributor

Details:

enable 3 op conversions for PDPD frontend. They are opset8 operations asked by PaddlePaddle detection NN.

  • matrix_nms
  • multiclass_nms
  • deformable_convolution

Tickets:

  • 54689
  • 53847
  • 57133

@ceciliapeng2011 ceciliapeng2011 requested a review from a team July 13, 2021 08:56
@ceciliapeng2011
Copy link
Contributor Author

ceciliapeng2011 commented Jul 13, 2021 •
edited
Loading

@itikhono @nosovmik @ilya-lavrenov @ilyachur @jane-intel @luo-cheng2021 @zhangYiIntel request your review.

@openvino-pushbot openvino-pushbot added the category: Core OpenVINO Core (aka ngraph) label Jul 13, 2021
@ceciliapeng2011 ceciliapeng2011 changed the title Cecilia/pr/dcn and nms op conversion for deformable_convolution, multiclass_nms and matrix_nms. Jul 13, 2021
@ceciliapeng2011 ceciliapeng2011 changed the title op conversion for deformable_convolution, multiclass_nms and matrix_nms. PDPD frontend op conversion for deformable_convolution, multiclass_nms and matrix_nms. Jul 13, 2021
Jan Iwaszkiewicz and others added 20 commits July 13, 2021 12:49
[nGraph] Return Node's operation version (openvinotoolkit#6589)
01eebba 
* Add python test

* expose node version to python api
@andrew-k-park
[IE CLDNN] Add common comparator for ngraph reference and cldnn detec…
2edb0e0 
…tion output primitive (openvinotoolkit#6530)

- Add common comparator for sort function to be aligned with others(cldnn/mkldnn DO primitive)

Signed-off-by: Andrew Kwangwoong Park <[email protected]>

Apply clang-format to do ngraph reference code

Signed-off-by: Andrew Kwangwoong Park <[email protected]>

Apply code review

Signed-off-by: Andrew Kwangwoong Park <[email protected]>
Zero point optimization (openvinotoolkit#6617)
46a20fd 
Ticket: #59063
Summary: try to fuse zero point to the weights in order to make the zero point == 0.
@generalova-kate
add a new label to track 3d party changes (openvinotoolkit#6610)
27be552
@evkotov
- add InsertTransposeAfterConvOrPool tranformation tests (openvinotoo…
b29ee60 
…lkit#6462)

- fix InsertTransposeAfterConvOrPool transformation copying runtime info
@evkotov
GNA add SwapInputMatMul transformation unit tests (openvinotoolkit#6458)
a386869 
* - add SwapInputMatMul transformation unit tests
- fix SwapInputMatMul matcher pattern

* code review fixes: remove unused piece of code

* use clone function instead of creating reference function code duplicate
@ZlobinGM
[IE][VPU]: Completion of Configuration options refactoring in VPU plu…
3d13424 
…gins (openvinotoolkit#6176)

* [IE][VPU]: Refactors PROTOCOL configuration option

* [IE][VPU]: Refactors POWER_CONFIG configuration option

* [IE][VPU]: Refactors HW_OPTIMIZATION configuration option

* [IE][VPU]: Refactors HW_EXTRA_SPLIT configuration option

* [IE][VPU]: Refactors HW_BLACK_LIST configuration option

* [IE][VPU]: Refactors TILING_CMX_LIMIT_KB configuration option

* [IE][VPU]: Refactors WATCHDOG_INTERVAL configuration option

* [IE][VPU]: Refactors PRINT_RECEIVE_TENSOR_TIME configuration option

* [IE][VPU]: Refactors PERF_REPORT_MODE configuration option

* [IE][VPU]: Refactors PERF_COUNT configuration option

* [IE][VPU]: Refactors PACK_DATA_IN_CMX configuration option

* [IE][VPU]: Refactors NUMBER_OF_SHAVES configuration option

* [IE][VPU]: Refactors NUM_EXECUTORS configuration option

* [IE][VPU]: Refactors NUM_CMX_SLICES configuration option

* [IE][VPU]: Refactors HW_POOL_CONV_MERGE configuration option

* [IE][VPU]: Refactors TENSOR_STRIDES configuration option

* [IE][VPU]: Refactors INJECT_SW_OPS configuration option

* [IE][VPU]: Refactors IGNORE_UNKNOWN_LAYERS configuration option

* [IE][VPU]: Remove HW_WHITE_LIST configuration option

* [IE][VPU]: Refactors HW_DILATION configuration option

* [IE][VPU]: Refactors FORCE_PURE_TENSOR_ITERATOR configuration option

* [IE][VPU]: Remove FORCE_DEPRECATED_CNN_CONVERSION configuration option

* [IE][VPU]: Refactors EXCLUSIVE_ASYNC_REQUESTS configuration option

* [IE][VPU]: Refactors ENABLE_WEIGHTS_ANALYSIS configuration option

* [IE][VPU]: Refactors ENABLE_TENSOR_ITERATOR_UNROLLING configuration option

* [IE][VPU]: Refactors ENABLE_REPL_WITH_SCRELU configuration option

* [IE][VPU]: Remove ENABLE_REPLACE_WITH_REDUCE_MEAN configuration option

* [IE][VPU]: Refactors ENABLE_PERMUTE_MERGING configuration option

* [IE][VPU]: Refactors ENABLE_MEMORY_TYPES_ANNOTATION configuration options

* [IE][VPU]: Refactors DUMP_INTERNAL_GRAPH_FILE_NAME configuration options

* [IE][VPU]: Refactors DUMP_ALL_PASSES_DIRECTORY configuration options

* [IE][VPU]: Refactors DUMP_ALL_PASSES configuration options

* [IE][VPU]: Refactors DISABLE_REORDER configuration option

* [IE][VPU]: Refactors DISABLE_CONVERT_STAGES configuration option

* [IE][VPU]: Refactors DEVICE_ID configuration option

* [IE][VPU]: Refactors DEVICE_CONNECT_TIMEOUT configuration option

* [IE][VPU]: Refactors DETECT_NETWORK_BATCH configuration option

* [IE][VPU]: Refactors CUSTOM_LAYERS configuration option

* [IE][VPU]: Refactors MEMORY_TYPE configuration option

* [IE][VPU]: Refactors FORCE_RESET configuration option

* [IE][VPU]: Refactors PLATFORM configuration options

* [IE][VPU]: Remove PLUGIN_LOG_FILE_PATH configuration options

* [IE][VPU]: Remove COMPILER_LOG_FILE_PATH configuration options

* [IE][VPU]: Refactors CHECK_PREPROCESSING_INSIDE_MODEL configuration option

* [IE][VPU]: Refactors ENABLE_EARLY_ELTWISE_RELU_FUSION configuration option

* [IE][VPU]: Refactors ENABLE_CUSTOM_RESHAPE_PARAM configuration option

* [IE][VPU]: Refactors NONE_LAYERS configuration option

* [IE][VPU]: Remove INPUT_SCALE and INPUT_BIAS configuration option

* [IE][VPU]: Refactors ENABLE_ASYNC_DMA configuration options

* [IE][VPU]: Delete old configuration API
@ceciliapeng2011
pdpd frontend: enable deformable_conv.
b9eb916
Optimize ShapeOf sub-graphs (openvinotoolkit#6627)
3710c0e 
1. Include shape sub-graphs optimization in the model optimizer nGraph pipeline
2. Extend shape sub-graph optimizations with useless Concat and useless Gather optimization
@ceciliapeng2011
pdpd frontend: enable multiclass_nms and matrix_nms.
068d7b6
@luo-cheng2021
MulticlassNms/MatrixNms: ngraph python api (openvinotoolkit#6573)
2378593 
* nms python api

* fix python code style

* fix python code style

* apply review comments

* apply review comments
@dpigasin
[IE Samples Docs] Add rows with model download and conversion (openvi…
1cfe27a 
…notoolkit#6626)

* Add information about model preparation

* Add the missed file extension for the onnx model

* [c] Add information about model preparation

* [cpp] Add information about model preparation

* Update the string about model converter

* Specify a model name at steps 1 and 2 of the running example
@vladimir-paramuzov
[GPU] Moved memory pool to network (openvinotoolkit#6599)
cbad1b2
@pelszkow
remove **Default value**: None when **Required**: *yes* (openvino…
b9d0293 
…toolkit#6638)
@antonvor
[CPU] gemm inner product - memory access fix (openvinotoolkit#6072)
ae25f5f
@sirzabek
Add fake quantize support for convolution padding (openvinotoolkit#6577)
34c20ad 
* [GNA] Add fake quantize support for convolution padding

Combine seven ngraph matcher passes into two.
Remove max pool size checking.
Add ngraph reference tests for subgraphs which were processed by POT.
Fix remaining issues with Max Pooling output calculations.
Add setting of default compile target based on execution target.

* [GNA] Remove redundant subgraph matcher

* [GNA] Remove redundant subgraph matcher
@slyubimt @andrew-zaytsev
Install guides improvements (openvinotoolkit#6418)
c8c515e 
* Install guides improvements

* add bullet to conda  System Requirements

* fix formating

* - Add conda install command for Ubuntu20

- fix typo /tmp

* added conda prerequisites

* Update installing-openvino-apt.md

* Update installing-openvino-conda.md

* Update installing-openvino-conda.md

CentOS 7.6

* Update installing-openvino-apt.md

APT Repository

* Update installing-openvino-conda.md

Added Introduction & notice about runtime package

* Update installing-openvino-conda.md

Co-authored-by: Andrey Zaytsev <[email protected]>
@EgorDuplensky
[CPU] Graph's RemoveEdge refactoring (openvinotoolkit#6221)
3617a7e 
- Move RemoveEdge from graph optimizer to graph
- Remove extra removeEdge lambdas
@a-sidorova
[CPU][IE TESTS] Added Gather-7 I8 test support (openvinotoolkit#6225)
864b3ba
Revise PriorBoxClustered Op (openvinotoolkit#6525)
c14edd4 
* Move type prop tests to separate file, add cases with exceptions thrown

* Add visitor test, add step to pbc attributes

* SLT for CPU, adjust GPU and SSLT to serve the step attr

* Add backend test

* Fix variance issue, verify by SLT

* Formatting fix

* Add another backend testcase

* Add PriorBoxClustered to trusted ops list

* Add step attr description, remove obsolete comments from ref impl

* Test multiple step values in SLT
@ceciliapeng2011 ceciliapeng2011 changed the title PDPD frontend op conversion for deformable_convolution, multiclass_nms and matrix_nms. [FrontEnd]enable pdpd ops conversion part4: deformable_convolution, multiclass_nms and matrix_nms. Jul 27, 2021
@ilyachur
Copy link
Contributor

Please fix CI

@ceciliapeng2011 ceciliapeng2011 requested a review from a team July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested a review from a team as a code owner July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested review from a team July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested review from a team as code owners July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested review from a team July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested a review from a team as a code owner July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested review from a team July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested a review from a team as a code owner July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 requested review from a team July 28, 2021 06:22
@ceciliapeng2011 ceciliapeng2011 mentioned this pull request Jul 28, 2021
Merged
@ceciliapeng2011
Copy link
Contributor Author

recreated and replaced by #6833. This one is unintentionally closed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@mvafin mvafin mvafin left review comments

@jane-intel jane-intel jane-intel approved these changes

@ilyachur ilyachur ilyachur approved these changes

@ilya-lavrenov ilya-lavrenov ilya-lavrenov approved these changes

@itikhono itikhono Awaiting requested review from itikhono

@nosovmik nosovmik Awaiting requested review from nosovmik

@GlebKazantaev GlebKazantaev Awaiting requested review from GlebKazantaev

Assignees
No one assigned
Labels
category: Core OpenVINO Core (aka ngraph)
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.