IntenseWebs/openvino
3
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
1,911 Commits 38 Branches 54 Tags 1.1 GiB
42478ff0ce
Commit Graph

1911 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Jedrzej Hajduczenia
fbae10a235
[IE CLDNN] Add asymmetric quantization support to fsv16 imad general convolution kernel (#2778) 2020-11-04 17:31:40 +03:00
Sergey Shlyapnikov
9c509e5f41
[IE CLDNN] Convolutition optimizations for small channels count (#2849) 2020-11-04 14:41:12 +03:00
Tomasz Socha
7c6794344c
[ONNX][PyTests] Fix result datatype missmatch (#2596) 2020-11-04 12:19:40 +01:00
Michał Karzyński
23188e1b04
Expand ONNX functions to sub-graphs before import (#2733) 
Co-authored-by: Bartosz Sledz <bartosz.sledz@intel.com>
 2020-11-04 10:48:34 +01:00
Alexander Zhogov
df49a2b987
Azure CL: Improve yaml (#2951) 
* Azure CL: Improve yaml

* Fix installing requirements

* Add run onnxImporterUnitTests
 2020-11-03 21:20:53 +03:00
Rafik Saliev
d225ba6e53
[IE] Add batched blob support (#2203) 
* [IE] Add batched blob support

New `class BatchedBlob : public CompoundBlob` defined to allow to pass multiple blobs as 1 InferRequest input.

Motivation: There is the special user case when a number of plain images (e.g. `NV12Blob`) should be passed as one input for network which batch size > 1.

`class CompoundBlob` is not applicable for such cases due to:
1. `NV12Blob` is `CompoundBlob` which prevents to combine multiple NV12 images to a CompoundBlob
2. The default behavior in most of plugins - do not accept generic CompoundBlob as `SetBlob()` argument

Adding `SetBlob(name, vector<Blob::Ptr>...)` to `class IInferRequest`, `class InferRequest`, `class IInferRequestInternal`, ...  - is not effective solution due to limited and specific use cases for `batched inputs`.

+ Apply rule-of-zero to CompoundBlob and inherited classes.

* Add "BATCHED_BLOB" optimization capability metric

* Add BatchedBlob usage to hello_nv12_input_classification

* Apply offline code review outcome:

1. Revert CompoundBlob public .ctors signatures
2. Remove 'workaround' .ctor for `BatchedBlob`
3. Revert tensor descriptors of `I420Blob` `NV12Blob` back to the 'fake' value.

* Code review fix

* Add functional tests for CPU, GPU, MULTI, HETERO

* update doc comment

* Apply code review change requests.
 2020-11-03 21:19:26 +03:00
Evgeny Lazarev
d682950e64
Updated requirements dependency (#2950) 2020-11-03 18:26:11 +03:00
Piotr Szmelczynski
c8bd92ac0d
Sigmoid revise (#2942) 
* remove sigmoid_backprop

* Update Sigmoid spec

* Update Sigmoid spec
 2020-11-03 18:16:16 +03:00
Irina Efode
a66c728a46
[IE TESTS] Fix reg exp for issue 42029 (#2957) 2020-11-03 18:12:24 +03:00
Mikhail Ryzhov
1144130b50
Aligned python read_network() with C/C++ behaviour (#2916) 
Aligned python read_network() with C/C++ behaviour

Excluded mandatory check of weights file, it is redundant

* Added unit tests
 2020-11-03 16:18:43 +03:00
Andrey Zaytsev
cf5055745a
Added info on DockerHub CI Framework (#2918) 2020-11-03 16:14:27 +03:00
Andrey Dmitriev
976f1f84a9
[GNA] Set default compact_mode (#2895) 2020-11-03 16:10:54 +03:00
iliya mironov
1e67eaf030
Add IR reader test for hsigmoid op (#2825) 
* Add IR reader test for hsigmoid op

* update ir

* Fix decomposition

* Update hsigmoid ir reader test

* Update unit test

* fix test
 2020-11-03 14:13:08 +03:00
Mikhail Letavin
78bf78660f
[IE CLDNN] Fix X/Y swap in NV12 shared surface input layout (#2949) 2020-11-03 13:29:36 +03:00
Rafal Blaczkowski
4085785d73
Abort previous builds on PR update (#2877) 2020-11-03 10:54:31 +01:00
Andrew Bakalin
1c6772a9fc
[IE][VPU]: Support for Bidirectional mode in Broadcast DTS (#2873) 
* [VPU][DTS] Update broadcast DTS to support BIDIRECTIONAL mode
* [VPU][Tests] Update tests with inference
* [VPU][Tests] Extend DTS tests
 2020-11-03 11:59:06 +03:00
Andrey Sokolov
f3ac97e9f4
[IE][VPU]: Interpolate - reuse "interp", "resample" layers (#2932) 
Reuse existing "interp", "resample" layers
task: #-29955
 2020-11-03 11:56:55 +03:00
Ilya Churaev
665fd1b773
Fixed README (#2948) 2020-11-03 10:56:28 +03:00
Kamil Magierski
3f4d8b49ff
Fix cases when identities inserted identites are separated by nonFunctional layers (#2771) 2020-11-03 10:49:58 +03:00
Krzysztof Bruniecki
6a2c209e58
Handle multiple GNA devices and modes HW/SW (#2720) 
* Introduce Gna2DeviceGetCount

* Check for the number of GNA devices

* Add Gna2DeviceGetCount() mock
 2020-11-03 10:49:09 +03:00
Andrey Dmitriev
d79263fcd0
[GNA] Added export input/output names (#2601) 
* [GNA] Added export input/output names

* fix comment

* Change version
 2020-11-03 10:45:08 +03:00
Mateusz Tabaka
3f5584534a
Remove obsoleted v0::Reshape operator (#2878) 
* Remove obsoleted v0::Reshape operator

* Fix handling negative dims on 32 bit platforms

Change-Id: Ia18e20df9a79fd0b7b38c434bee7140599833952

* apply code format

Change-Id: I4ddfb9e7033cbc97e167922f6ee5873d5487e551
 2020-11-03 07:01:34 +03:00
Anton Romanov
50e6d0e31e
Added check in hello sample (#2896) 
* Added check in hello sample

* move check before inference
 2020-11-03 06:34:06 +03:00
Ilya Lavrenov
1a1451ed8f
Compilation with MSVC for ARM (#2928) 
* Enable compilation with MSVC for ARM / ARM64

* Condition fix
 2020-11-03 00:55:49 +03:00
Evgeny Lazarev
c8dd831fc3
Added transformation config to support automl efficientdet models (#2894) 
* Added transformation config to support automl efficientdet-4 model

* Added configuration file to convert Automl EfficientDet model

* Updated unit test for Pack

* Added instruction on how to convert EfficientDet Tensorflow model

* Updated documentation on how to convert EfficientDet model

* Updated a documentation with instruction on how to convert Automl EfficientDet.
 2020-11-02 19:21:05 +03:00
Alexandra Sidorova
5596c7c841
[CPU] Added eltwise Round-5 (#2347) 2020-11-02 18:44:50 +03:00
Ilya Lavrenov
f2963713d0
Fixed FAST build mode (#2926) 2020-11-02 17:58:52 +03:00
Aleksandr Korolev
5ffa2cd5e3
[IE TESTS] Split Layer Test refactoring (#2915) 
Co-authored-by: kora6 <kora6@github.com>
 2020-11-02 16:29:59 +03:00
Andrey Kamaev
8425cd492c
Remove remaining Myriad-2 references (#2853) 
* Fix Myriad firmware copying command for multi-target cmake generators

* Remove remaining Myriad-2 references

* Remove remaining Myriad-2 compute stick from documentation
 2020-11-02 15:10:50 +03:00
Anton Romanov
67386294c0
Change pip install documentation (#2884) 2020-11-02 14:09:54 +03:00
Bartek Szmelczynski
8e01022252
Visitor API deserialization - create models for ops (#2889) 
* add tests for 11 operators, which compare V7 and V10 version of the model optimizer

* add test for space_to_depth model
 2020-11-02 13:48:53 +03:00
Vladimir Paramuzov
fea2595aed
[IE CLDNN] Fixed unitialized parameters in resample primitive which led to kernels recompilation (#2923) 2020-11-02 13:31:07 +03:00
Alexander Peskov
6fe05ecda7 Unmark passed ONNX test 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
0a4509914a [TEST] One more trivival loop test 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
045d26ff31 [CPU] Fix case with unused port in Loop body 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
5ac26ff3b6 [TEST] One more trivial test case on Loop 
Also fixed compilation with gcc4.8

Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
7d7021269f [CPU] Improve replication of passthrough Loop body 
There was some problem with replicatioon of simple loop body where
input was used as output as is.

Also was voided usage of special prefixes like "in_" for Loop body inputs.

Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
fe51b7cb8b Min. Additional assert in copyTIBoody 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
d7e3e92b64 [TEST] Several more Loop test with static shapes 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
3160290e13 [CPU] implement Loop specification of TI node 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
124bfc34b3 [NG] Keep special purposes port info in Loop after conversion to legacy 
Also:
Simplified logic of data object name restoring. Avoid duplicatin of input ports
in case of multiple consumers. Provided code has WA comment in corresponding
naming restore section. Also added WA section with restore U8 precision for ouputs.
Avoid to eliminate limitation of CNNNetwork converter.

Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
e770f7b0cf [TEST] improve data utils 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
56dc06d09f Avoid '-' symbol for test name 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
bb28eb7209 Loop tests enabled back 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
05651b18af [IE] Improve Network topological sort in case of disconnected graph 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Alexander Peskov
ec1561a234 Improve CopyTIBody util to cover disconnected graphs 
Signed-off-by: Alexander Peskov <alexander.peskov@intel.com>
 2020-11-02 12:37:48 +03:00
Mateusz Tabaka
16bb29f29c
Remove obsoleted v0::StopGradient op (#2935) 2020-11-02 12:22:01 +03:00
Tomasz Jankowski
2d9a47127d
refactor one-hot nGraph backend test (#2578) 2020-11-02 10:12:39 +01:00
Tomasz Jankowski
1b94682f86
Avoid verbose return-type warnings (#2420) 2020-11-02 10:11:35 +01:00
Tomasz Jankowski
1550e94156
nGraph backend tests refactor: gather, numeric (#2367) 2020-11-02 10:10:25 +01:00