Age | Commit message (Collapse) | Author |
|
* Adds Reference FP16 support and unit tests for layers not already supported
!referencetests:202156
Signed-off-by: Matthew Jackson <matthew.jackson@arm.com>
Change-Id: I6fc9b9ce2809e163f72e27e877025c8fb85d9fbe
|
|
* Wrap CL workload creation methods into MakeWorkload to
ensure proper handling of ClError exceptions
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: Id6b743b4682c092316b4775baa065921780dd24a
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: If54f3a285b288e7ff0b2674829c116df70d6934c
|
|
Signed-off-by: Kevin May <kevin.may@arm.com>
Change-Id: I99d8a3d53551ce5c293462d847c22ae4f912331b
|
|
* Working on layers required by FSRCNN, FCRN and DeepSpeaker
* Updates RefLayerSupport and RefWorkloadFactory methods
* Adds RefPadFloat16Workload
* Tested by successful execution of these networks on Reference FP16 backend
Signed-off-by: Matthew Jackson <matthew.jackson@arm.com>
Change-Id: I4817dca0a89bba6902f0feffc494b27a26a0ab2d
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: Idf40ed0ebf2af3170137bf3a3cb5a9635a203d4a
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: Ida590b7d28dbee40e496794c42b682d740427cca
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I802bd882da4adcc1e7e52c772771e743143745fa
|
|
* Implement Float 16 Encoder and Decoder
* Add Stack Float 16 layer and create workload tests
Signed-off-by: Matthew Jackson <matthew.jackson@arm.com>
Change-Id: Ice4678226f4d22c06ebcc6db3052d42ce0c1bd67
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I0c2ead004200e053d748ea39937f2f9ed35a636b
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I807a57af6dc2555a5a422388558a02e958ed610d
|
|
* Implemented RefAbsWorkload and unit tests
Signed-off-by: Sadik Armagan <sadik.armagan@arm.com>
Change-Id: Ibfcdb2b37fd8d240c181f96856e2c997a4b88914
|
|
Change-Id: I31c5616bea3097f30cde68442d3222e0b0fe2235
Signed-off-by: Nikhil Raj <nikhil.raj@arm.com>
|
|
* Add Abs Layer
* Add no-op factory implementations for CpuRef, CpuAcc, GpuAcc
* Add Queue Descriptor in WorkloadData
* Add IsAbsLayerSupported to LayerSupport
* Add LayerVisitor tests
Signed-off-by: Kevin May <kevin.may@arm.com>
Change-Id: Ib0992571d1c80de851cea466291be904e9bdb430
|
|
* Generic layer parameters now show up in dot file
* Convolution layer parameters have also been added to dot file
* ExecucteNetwork has an additional -v flag which generated dot file if there
Change-Id: I210bb19b45384eb3639b7e488c7a89049fa6f18d
Signed-off-by: Andre Ghattas <andre.ghattas@arm.com>
Signed-off-by: Szilard Papp <szilard.papp@arm.com>
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: Icf39434f09fd340ad664cb3b97b8bee6d9da4838
|
|
* Renamed ARMNN_COMPUTE_REF_ENABLED to ARMNN_REF_ENABLED in the
Android makefiles
Change-Id: I12af2ca00c719568274585482d428b035674737c
Signed-off-by: Matteo Martincigh <matteo.martincigh@arm.com>
|
|
ElementwiseTestHelper
* Refactored tests for element-wise operations to use ElementwiseTestHelper
* Moved the etasts for each operation in a separate file under backendsCommon/test/layerTests
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: Icefb6b35df78b9619f69378229789505bf92670e
|
|
* Renamed ARMCOMPUTEREF to ARMNNREF (and relative symbols) as the
previous name was misleading
!android-nn-driver:1819
Change-Id: I912e9f521750b91539e48f8a2fb4de73fb6f7425
Signed-off-by: Matteo Martincigh <matteo.martincigh@arm.com>
|
|
* Adds compatibility between Cl and Neon
* Adds related unit test
Signed-off-by: Jan Eilers <jan.eilers@arm.com>
Signed-off-by: Mike Kelly <mike.kelly@arm.com>
Change-Id: Iaa454ea2a6392c89c0a83836c3156d4ddb354879
|
|
Failing VTS tests were "NeuralnetworksHidlTest.depthwise_conv2d_*"
In depthwise convolution there was a difference in weight tensor channel
order between the reference and ACL implementations. This specifically related
to NCHW. This commit:
* Adds ReorderWeightChannelsForAcl to WorkloadUtils which will correct the weight tensor channel order.
* Add unit tests to detect this problem.
Signed-off-by: Colm Donelan <Colm.Donelan@arm.com>
Change-Id: Icaeac08e14b3d5da9e222ad2f118db55ebb15d09
|
|
Signed-off-by: Ferran Balaguer <ferran.balaguer@arm.com>
Change-Id: I6f16ea0dca359283a3b187e2f046f82a7dc2ff7c
|
|
* CMake "install" commands require a RUNTIME argument for platforms with
DLLs (e.g. Windows).
* Replace use of non-standard variable length array with vector
* Remove unnecessary #include of unistd.h
* Add #ifdefs to dynamic backend code to disable for non-Unix platforms
where you can't use dlopen etc. We could implement this properly for Windows
later using LoadLibrary etc., but for now erroring is fine.
* Add missing #include of <algorithm>
Change-Id: Ic8ef5fd599b37bf8772510157b6e479819f6a1eb
|
|
Signed-off-by: Ferran Balaguer <ferran.balaguer@arm.com>
Change-Id: I87d003ae14575d7d7be67b2a7d00d69ef6201849
|
|
* Made the build of the reference backend depend on a new ARMCOMPUTEREF
macro
* Made the relevant targets dependent on the ref backend
* Moved Cl and Neon static registry initializers to separate files
* Wrapped some of the unit tests into proper ifdefs where necessary
Change-Id: I7f2c42699682630233a4c4b6aed2f005083de189
Signed-off-by: Matteo Martincigh <matteo.martincigh@arm.com>
|
|
Signed-off-by: Ferran Balaguer <ferran.balaguer@arm.com>
Change-Id: I64f82c0e3c5cddcaebe0d5001020880a759ec444
|
|
Signed-off-by: James Conroy <james.conroy@arm.com>
Change-Id: Ibfd5a81b2dbf9eeef49f6209885798804adee5a4
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I3da43122240fda7864e0932d47b3efeea8bbed40
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I9d19ae4559a98b9e740c1d4d05add0c5788093d6
|
|
* Fixed Quant8 VTS test failures where an output value should have been greater than 255 before a negative bias was applied.
Signed-off-by: Mike Kelly <mike.kelly@arm.com>
Change-Id: Ie0084c1c8c4a4b5f069112d523765ed338277e58
|
|
* NEON support for TransposeConolution2d
* Indentation fix on SpaceToDepth LayerSupport method
Signed-off-by: Keith Davis <keith.davis@arm.com>
Signed-off-by: Sadik Armagan <sadik.armagan@arm.com>
Signed-off-by: Aron Virginas-Tar <aron.virginas-tar@arm.com>
Change-Id: Ie3a01a22e05f62f475f768dfe05ce561d2f3b0ad
|
|
Signed-off-by: Ferran Balaguer <ferran.balaguer@arm.com>
Change-Id: I94bd191f88e0911ad4e4727610e81cd7afa95512
|
|
* Fixed bug in multi-channel test and reference workload implementation
* Enabled multi-channel tests on CL backend
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I08bb523bc68d9c93a1012b4f487a5bce36a085b1
|
|
* Adds ClTensorHandleFactory
* Includes some refactoring of NeonTensorHandleFactory
Signed-off-by: Jan Eilers <jan.eilers@arm.com>
Change-Id: I9f0874b1945050267c5ece970e6d9b200ed8a865
|
|
* Now the utility function RegisterDynamicBackends returns a list of
the backend ids that have been registered
* The list of registered ids is added to the list of supported backends
in the Runtime
* Added unit tests
Change-Id: I97bbe1f680920358f5baba5a4666e4983b849cac
Signed-off-by: Matteo Martincigh <matteo.martincigh@arm.com>
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: Idebe0c3f956ccac39ef8bf139d693b76e839af7a
|
|
The following NeuralNetworkTests tests were failing on GpuAcc and CpuAcc:
GeneratedTests.softmax_v1_2_relaxed
GeneratedTests.softmax_v1_2_quant8
GeneratedTests.softmax_v1_2_2
GeneratedTests.softmax_v1_2_relaxed_2
GeneratedTests.softmax_v1_2_quant8_2
The default value for Softmax axis parameter in Android is -1 but is 1 in ACL.
Detect and handle this in ArmComputeUtils.ComputeSoftmaxAclAxis.
Signed-off-by: Colm Donelan <Colm.Donelan@arm.com>
Change-Id: Ibb0660e4cb0dc6bd4c804c4397fbd61f38acdd9c
|
|
* Increased MaxNumOfTensorDimensions and fixed issues related to its use
* Fixed issues caused by assuming 5d tensors are invalid
* Updated ArmComputeTensorUtils for 5d tensors
* Added 5d tensor unit tests for add, mul, stack and reshape (needed by IVGCVSW-3527)
Signed-off-by: Matthew Jackson <matthew.jackson@arm.com>
Change-Id: I5bcd64942d0d04efcc6c5acb240ad4b88e010743
|
|
Signed-off-by: James Conroy <james.conroy@arm.com>
Signed-off-by: Narumol Prangnawarat <narumol.prangnawarat@arm.com>
Change-Id: I9f24f7d97c1e7d37ee7d58ff85c6c3f8496e52ec
|
|
* Fixed possible driver crash due to supportedTypes arrays with the wrong size.
Signed-off-by: Mike Kelly <mike.kelly@arm.com>
Change-Id: I52702b84089d02d3c63ff7aa5923a9f5628bec91
|
|
Signed-off-by: Ferran Balaguer <ferran.balaguer@arm.com>
Change-Id: I634aa3b1d609ca33b196fd68ce7fb7881be73e6e
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I496f95ab464f3a38e0601799a6ff6ba5ffe3a776
|
|
* Change Getter Signatures to follow coding guidelines
Change-Id: Ic02621e834dbf79b9df63f8b4c6339f71651e944
Signed-off-by: Francis Murtagh <francis.murtagh@arm.com>
|
|
* Mentioned the TestDynamicBackend source files
* Added the RefDynamicBackend example
Change-Id: I395c9f6c3107a5a5b16f658327d6048ff2ae68d8
Signed-off-by: Matteo Martincigh <matteo.martincigh@arm.com>
|
|
RefLayerSupport::IsTransposeConvolution2dSupported()
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I85a9563e2d12a5eb9dbcac1b70563a303905b67b
|
|
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: Ib9271920b39fb9d730006c79a322e32008096b56
|
|
* Now supports channel multiplier != 1 in Ref TransposeConvolution2d
Signed-off-by: Mike Kelly <mike.kelly@arm.com>
Change-Id: Ice95444025311b8b952c8f7f5fa0f1ecf483df41
|
|
* Fixed Max and Min VTS test failures.
Signed-off-by: Sadik Armagan <sadik.armagan@arm.com>
Change-Id: I133f5ff43774f74ad6e1c698a3570066ad45527c
|
|
in IBackendInternal
Signed-off-by: Narumol Prangnawarat <narumol.prangnawarat@arm.com>
Change-Id: Id40b825649d0447058273b8bc4f7133a3ac3c96c
|
|
* Silently considering the dynamic backend loading feature as
disabled if an empty list of paths is passed to ArmNN
* This was causing problems during testing as ArmNN's output
was containing extra unnecessary warning messages
Change-Id: I4ebda412004179817cd54494ccddc964b04b8232
Signed-off-by: Matteo Martincigh <matteo.martincigh@arm.com>
|