Commit Graph
6585 Commits
Author SHA1 Message Date
Aleksandr Korolev 8b7aeb7f52 [VPU] Coverity fixes (#10090) 2022-02-10 15:31:00 +03:00
Evgenya Stepyreva 9ad09f2120 Shape inference adoption for dimension tracking (#10016)
* Shape inference adoption for dimension tracking

* Style

* test adj

* tests fixed
2022-02-10 15:30:18 +03:00
Anton Voronov d5c837cc1b [CPU] added some legacy parallel methods to fix perf issues (#9758) 2022-02-10 15:13:05 +03:00
Andrei Molotkov 80be557605 [GPU] Fix Backpropagation issue with BS >= 16 (#10228) 2022-02-10 14:53:27 +03:00
Ilya Lavrenov ea26ec32b3 Removed runtime namespace (#10231) 2022-02-10 14:53:13 +03:00
Alexey Lebedev d484411f39 [tools] Fix image_info detection in benchmark app (#10192)
* Fix image_info detection

* exception instead warning in case input data is not compatible with input
2022-02-10 14:32:56 +03:00
Andrew Kwangwoong Park 51c89dff26 [GPU] Fix detection output stage-0 kernel (#10262)
- Change the constant value to the maximum work group size
- Add CLK_GLOBAL_MEM_FENCE barrier to synchronize storing result in intermediate buffer
- Add condition to prevent access local array out of range

Signed-off-by: Andrew Kwangwoong Park <andrew.kwangwoong.park@intel.com>
2022-02-10 19:43:16 +09:00
Evgenya Stepyreva 89c3a18f83 Fix TensorIterator dynamic rank output (#10247)
* Fix TensorIterator dynamic rank output

* style
2022-02-10 13:03:16 +03:00
Ivan Tikhonov 3f0e532dce Fix the issue in values_from_const method in OVTF integration with TF FE (#10225)
* Fix the issue in values_from_const method in OVTF integration with TF FE

* fix comment
2022-02-10 11:33:42 +03:00
Min, Byungil 334e9e994e Revert WA for onednn first conv (#9783)
+ Reverted WA for fsv32 format first conv
+ Applied blocked input format bsv8fsv4 & bsv8fsv2 for onednn first conv
+ Implemented onednn usage for first conv of feature size 1
+ Added new weight format ABcd16a4b
+ Bugfix in fetch_weight
+ Updated thirdparty onednn_gpu
+ Known issue : AcdB16a4b is not supported

Signed-off-by: Min, Byungil <byungil.min@intel.com>
2022-02-10 12:12:09 +09:00
Bartek Szmelczynski 36de4e8e28 [Model Enablement] fix default onnx domain (#10106) 2022-02-10 03:16:24 +03:00
Gleb Kazantaev 87c6e09cae Fix Add/MulFQFusion transformations (#10252) 2022-02-10 01:22:16 +03:00
Maxim Andronov 36afedd93d [CPU] Increase executor cache capacity (#10232) 2022-02-09 21:49:37 +03:00
Alexandra Sidorova fce49e6d80 [Transformations] Added interchangeable reshape elimination (#9691)
* [Transformations] Added interchangeable reshape elimination

* Applied comments #2

* returned Reshape in condition

* applied comments #3

* applied comments #4

* added comment in plugin with reason about transformation
2022-02-09 21:11:49 +03:00
Mikhail Ryzhov a002b26294 Fixed import for the new api 2.0 (#10175) 2022-02-09 20:51:49 +03:00
Vladislav Golubev d28f8b7857 [LPT] Security fixes (#10243) 2022-02-09 20:46:39 +03:00
Tomasz Dołbniak aedd902cd8 Use double quotes in JSON analysis (#10237) 2022-02-09 20:41:49 +03:00
Egor Shulman 840d2fb80d [CPU] Coverity fixes (#10207) 2022-02-09 20:39:50 +03:00
Gorokhov Dmitriy 6ea20340d1 [CPU] Fixed out of bounds read in JIT planar convolution (#7025) 2022-02-09 20:26:57 +03:00
Irina Efode a37195492c [IE TESTS] Add exception to comparation to provide correct conformance results (#10197)
* [IE TESTS] Add exception to comparation to provide correct conformance results

* Apply comments
2022-02-09 19:15:00 +03:00
Nikolay Tyukaev e81ca9f975 DOCS: change doc tests (#10213)
* change doc tests

* fixes

* fixes

* fixes

* fixes

* fixes
2022-02-09 18:28:54 +03:00
Maxim Shevtsov c0a375f844 adding I64/U64 support to the auto-batching (#10234)
* adding I64/U64/etc support

* inputs precisions tests instantiations for the GPU and BATCH:GPU
2022-02-09 18:28:13 +03:00
Mikhail Nosov f56c640550 SmartReshape: support Param->Convert->Reshape->Proposal pattern (#10204)
Current SmartReshape finds matched to Param->Reshape->Proposal patterns

    For FP16 models, there is additional 'Convert' is inserted after 'Parameter'.

    It causes transformation is not applied and 'ov::set_batch' or CNNNetwork::set_batch will throw

    Proposal1Scales and Proposal4Scales transformations were updated to handle these conditions
2022-02-09 17:44:54 +03:00
Tomasz Dołbniak a60c110b96 Use i64 for ONNX Split attribute (#10203) 2022-02-09 17:30:00 +03:00
Vitaliy Urusovskij c186069025 Fix several coverity issues (#10205)
* Update def value for GetParamAsBool() in legacy parseParams()

* Remove extra check from legacy convertFunctionToICNNNetwork()
2022-02-09 15:43:04 +03:00
Chen Xu c93c9ec3d5 [CPU] Fix bug in topk_bubble_BLK_on_channel_horiz method (#10218) 2022-02-09 14:40:46 +03:00
Victor Kuznetsov 21601398d6 Remove dynamism from time_tests (API 1.0) (#10193) 2022-02-09 19:15:16 +08:00
Vladislav Golubev 051724f0d5 [LPT][Dynamic shapes] MoveFakeQuantize trasformation fix (#10178)
* [LPT] MoveFQ fix

* [LPT] MoveFQ: added check on dynamic channel in case of per-channel fq

* [LPT] MoveFQ: tests extending
2022-02-09 13:55:50 +03:00
Taylor Yeonbok Lee 603ea50277 Fix max batch size to respect available virtual memory in linux environment. (#10201) 2022-02-09 19:40:29 +09:00
Ilya Churaev 79fceddd7e Fixed some coverity issues (#10165) 2022-02-09 12:37:19 +03:00
Gleb Kazantaev 60011b6eb2 Fix EltwiseBroadcastFusion pass (#10214) 2022-02-09 12:35:38 +03:00
Pavel Esir 654b025a26 [MO] set explicitly argument dtype to int for np.split (#9988)
* forced split argument dtype to int

* added unit-test

* fixed typo in split_test.py

* set explicitly np.int64 instead of np.int

* use split_length's dtype
2022-02-09 12:16:33 +03:00
Anton Chetverikov 25ca17e789 [MO IR Reader] Update *Sequence backend_attrs (#10041)
* Update LSTMSequence backend_attrs

* Add missed attribute clip

* Update backend_attrs for all *sequence operations

* Add extender for GRUSequence

* Add GRUSequence to custom ops list

* use has_and_set instead if direct acces to attributes
2022-02-09 12:13:23 +03:00
Gleb Kazantaev 4fdf71cdc1 Preserve RTInfo in output ports (#10114)
* Automation for preserving rt info in output ports; Update FunctionComparator to compare rt info correctly

* Update LPT tests to use real rt_info attributes, so they can be compared

* Fix tests
2022-02-09 12:09:23 +03:00
Daniil Lyakhov 0168bda833 [Offline Transformations] Reshape Layer Pruning Transformation Support (#9350)
* Reshape op pruning support

* Minor reshape fix

* GroupConv reshape extended support

* Comment ir test

* Fix: reshape can only work with constants

* Apply comments

* Fix output shape computing for reshape op

* Fix comment
2022-02-09 12:03:56 +03:00
Maxim Shevtsov 320c64de24 disabling auto-batching when batch<4 (as batch1 kernels are heavily optimized) (#10188) 2022-02-09 12:02:30 +03:00
Anastasia Kuporosova 04194b292d [Python API] Add if for yocto cross-compilation (#10216) 2022-02-09 11:56:42 +03:00
Maxim Vafin 52374a4b8b Write runtime version and how IR was genarated (legacy path or not) (#10196) 2022-02-09 11:41:49 +03:00
Daria Mityagina 6dd6fb6c12 [VPU][XLink] Printf over XLink fails on OpenVINO 2021.4.2 - fix (#10099)
The XLinkReadDataWithTimeout() is used with and incorrect value for timeout parameter.
2022-02-09 11:30:29 +03:00
Edward Shogulin c4e54d882b [LPT] StridedSlice extending (#10148)
* [LPT] StridedSlice extending

* [LPT] tests
2022-02-09 11:23:18 +03:00
Ilya Churaev 9d40c5184f Removed legacy names and environment variables from the code (#10195)
* Removed legacy names and environment variables from the code

* Support documented legacy variables

* Fixed core unit tests

* Fixed some test
2022-02-09 11:04:25 +03:00
Vitaliy Urusovskij 532a1da548 Fix "Error handling issue" (#10119)
* Fix coverity 'error handling issue' in ~CacheGuard()

* Fix coverity 'error handling issue' in reshape()
2022-02-09 11:04:02 +03:00
Sergey Lyubimtsev acf8cacfbc requirements markers clean up (#10179)
* requirements markers clean up

* formatting & comments

* typos
2022-02-09 10:18:24 +03:00
Roman Lyamin 0d64afc2c8 [GPU] program_helpers::are_layouts_identical fix (#10109) 2022-02-09 09:27:15 +03:00
Sergey Shlyapnikov 8f0e974ee6 [GPU] Add new properties and fix bechmark_app (#10149) 2022-02-09 09:18:54 +03:00
Maxim Vafin 1970baeb1c Apply RIC for dynamic dimension in legacy MO (#10130)
* Apply RIC for dynamic dimension in legacy MO and fail if RIC wasn't applied to any input

* Fix moc tests
2022-02-08 22:17:19 +03:00
Smirnov Grigorii d951433b12 fix bug in Serialize (#74447) (#9840)
* fix bug in Serialize (#74447)

add simple serialization test to check pads changes

clang fix

add check and change pads in conv

refactor ov::clone_model

fix

check in test

* fix FrameworkNode and add test

* fix assert in identiry.cpp

* fix clone_nodes

* remove for node and constructor for node_input.cpp

add spaces

add space
2022-02-08 22:00:20 +03:00
Evgenya Stepyreva a18069926e Partial Value propagation from partial to static value (#10162)
* Partial Value propagation from partial to static value

* Style

* Tests ajustment
2022-02-08 21:55:17 +03:00
Yury Gaydaychuk 0dfdadb531 [CPU] Clamp reduces float boundaries in the case of integer data (#6668) 2022-02-08 19:50:45 +03:00
Ivan Novoselov b47b8ad4bf [CPU] Snippets throughput mode fixes (#9488) 2022-02-08 17:58:42 +03:00