Gitiles
Code Review
Sign In
review.mlplatform.org
/
ml
/
armnn
/
refs/tags/v22.05.01
/
src
/
backends
/
neon
tag
66abb4bc5d94e6d54c8ffd4009ccb00a295529cf
tagger
Nikhil Raj <nikhil.raj@arm.com>
Fri Jun 17 12:48:19 2022 +0100
object
4a1a47544c9a0f3ae01c8e416451982cac16fdc8
v22.05.01
21fe06f
IVGCVSW-6929 Support for models with implicit expanded dimensions
by Mike Kelly
· 2 years, 1 month ago
f6ae983
IVGCVSW-6126 ConstTensorsAsInput: Conv2d - Backends
by Cathal Corbett
· 2 years, 2 months ago
2cddc72
IVGCVSW-6124 ConstTensorsAsInput: Conv2d - FrontEnd
by Keith Davis
· 2 years, 3 months ago
19e7942
IVGCVSW-6175 Add Pooling3d to Neon
by Ryan OShea
· 2 years, 2 months ago
4452baf
IVGCVSW-6260 ConstTensorsAsInput: Fully Connected Cl and Neon support.
by Cathal Corbett
· 2 years, 2 months ago
4b19d22
IVGCVSW-6940 ConstTensorsAsInput: DepthwiseConvolution2d - Complete ACL
by Cathal Corbett
· 2 years, 2 months ago
0302723
IVGCVSW-6936 Sqrt for CpuRef, CpuAcc and GpuAcc
by Teresa Charlin
· 2 years, 2 months ago
b190da2
IVGCVSW-6862 Use same datatype for all containers of indices in NeonGatherNd
by Teresa Charlin
· 2 years, 2 months ago
aa80510
IVGCVSW-6862 Modify GATHERNd Neon workload
by Teresa Charlin
· 2 years, 2 months ago
06145cc
IVGCVSW-6936 Add SQRT support to Neon
by Teresa Charlin
· 2 years, 2 months ago
0690265
IVGCVSW-6127 ConstTensorsAsInput: DepthwiseConvolution2d
by Cathal Corbett
· 2 years, 3 months ago
bd22c7d
IVGCVSW-6862 Add GATHERNd Neon workload
by Teresa Charlin
· 2 years, 2 months ago
1299496
IVGCVSW-6806 Add Unidirectional Sequence Lstm support to Neon
by Mike Kelly
· 2 years, 2 months ago
2776183
IVGCVSW-6706 Move headers to profiling/client/include
by Jim Flynn
· 2 years, 3 months ago
d83ffe0
IVGCVSW-6839 Unidirectional sequence LSTM test failing in Android S
by Cathal Corbett
· 2 years, 3 months ago
4952a3e
Revert "Revert "IVGCVSW-6267 Add support of Unidirectional Sequence Lstm fp32/fp16 to Cl""
by Cathal Corbett
· 2 years, 4 months ago
fd5bec4
Revert "Revert "IVGCVSW-6268 Add support of Unidirectional Sequence Lstm fp32/fp16 to Neon""
by Cathal Corbett
· 2 years, 4 months ago
3826ab6
IVGCVSW-6700 Disable importing on ClConv2d when datalayout is NCHW
by David Monahan
· 2 years, 4 months ago
f87b90e
Revert "IVGCVSW-6268 Add support of Unidirectional Sequence Lstm fp32/fp16 to Neon"
by Cathal Corbett
· 2 years, 4 months ago
79cef69
Revert "IVGCVSW-6267 Add support of Unidirectional Sequence Lstm fp32/fp16 to Cl"
by Cathal Corbett
· 2 years, 4 months ago
ad91717
IVGCVSW-6267 Add support of Unidirectional Sequence Lstm fp32/fp16 to Cl
by Cathal Corbett
· 2 years, 5 months ago
b0baff7
IVGCVSW-6268 Add support of Unidirectional Sequence Lstm fp32/fp16 to Neon
by Cathal Corbett
· 2 years, 5 months ago
79bb653
Fix unit test where NeonMemoryManager was used instead of RefMemoryManager
by Matthew Bentham
· 2 years, 5 months ago
ec81999
IVGCVSW-6700 Add override functions to all typed Neon/CL workloads
by David Monahan
· 2 years, 5 months ago
67d6390
Handle optional biases better in Neon/Cl FullyConnected workloads
by Matthew Bentham
· 2 years, 5 months ago
c42a987
IVGCVSW-6635 Move MemCopyTestImpl from acl to armnnTestUtils.
by Colm Donelan
· 2 years, 5 months ago
646bc8a
IVGCVSW-6696 Add Forced Import EndToEnd tests to Ref, Neon, and CL
by David Monahan
· 2 years, 5 months ago
e2af6f4
IVGCVSW-6552 Add support of aligned host memory
by Narumol Prangnawarat
· 2 years, 5 months ago
588cbdf
IVGCVSW-6685-6686 Modify workloads to extend Neon/Cl BaseWorkload
by Teresa Charlin
· 2 years, 5 months ago
98b0dcb
IVGCVSW-6683-6684 Add ClBaseWorkload and NeonBaseWorkload
by Teresa Charlin
· 2 years, 5 months ago
7740045
IVGCVSW-6678 Register CopyAndImportFactoryPairs to NeonBackend and unit tests
by Narumol Prangnawarat
· 2 years, 6 months ago
faf2966
Bugfix: Set removal date for Backend API stability items to 22.11
by Francis Murtagh
· 2 years, 5 months ago
34b429c
IVGCVSW-6629 Stabilize the ILayerSupport interface with unified strategy.
by Cathal Corbett
· 2 years, 6 months ago
611c7fb
IVGCVSW-6641 Stabilize the IWorkloadFactory interface with unified strategy
by Teresa Charlin
· 2 years, 6 months ago
56ccf68
IVGCVSW-6633 SubgraphView uses IConnectableLayer rather than Layer in its m_Layers
by Francis Murtagh
· 2 years, 7 months ago
0fa1050
IVGCVSW-6674 Implement CanBeImported function in NeonTensorHandle
by David Monahan
· 2 years, 6 months ago
cbfd718
IVGCVSW-6632 OptimizationViews: has INetwork rather than Graph for holding layers
by Cathal Corbett
· 2 years, 7 months ago
0c47974
IVGCVSW-6626 Promote backend headers in backendCommon to armnn/backends
by Colm Donelan
· 2 years, 7 months ago
a097d2a
IVGCVSW-6453 'Move the ArmNN Test Utils code to a physically separate directory'
by Sadik Armagan
· 2 years, 7 months ago
f9de771
IVGCVSW-6520 Extend pad workload for CpuAcc to support mirror padding
by Teresa Charlin
· 2 years, 8 months ago
019da94
Fix typo errors from ticket IVGCVSW-6420
by Cathal Corbett
· 2 years, 8 months ago
5b8093c
IVGCVSW-6420: Constant flag in tensor info is not set correctly
by Cathal Corbett
· 2 years, 8 months ago
b1aad42
IVGCVSW-6527 Support the new memory API in loaded network
by Finn Williams
· 2 years, 8 months ago
2ea38c7
Bug Fix in CpuAcc and GpuAcc workload: Gather has 2 inputs
by Teresa Charlin
· 2 years, 8 months ago
2e5d0b2
IVGCVSW-6469 Add MirrorPad FrontEnd and Ref Support
by Matthew Sloyan
· 2 years, 8 months ago
ec5f7d1
IVGCVSW-6170 Add CpuAcc Conv3d Workload
by Teresa Charlin
· 2 years, 8 months ago
e8d7ccb
IVGCVSW-6310 Update ICustomAllocator and add GetDefaultAllocator function to backends
by Francis Murtagh
· 2 years, 9 months ago
cff135b
Revert "Disable Sin support in Neon"
by ryan.oshea3
· 2 years, 9 months ago
b8a26d8
IVGCVSW-6300 'IMemoryOptimizerStrategy Add strategy library and add support in BackendRegistry'
by Sadik Armagan
· 2 years, 9 months ago
1b2654f
IVGCVSW-5985 Remove deprecated code
by Jan Eilers
· 2 years, 9 months ago
eb852bb
IVGCVSW-6293 Add Unit test int8 Channel Shuffle
by Teresa Charlin
· 2 years, 9 months ago
e89dd69
IVGCVSW-3707 Add Channel Shuffle Workload to CpuAcc backend
by Teresa Charlin
· 2 years, 10 months ago
c840263
IVGCVSW-6339 'IMemoryOptimizerStrategy Add backend capabilities'
by Sadik Armagan
· 2 years, 10 months ago
4e3e831
IVGCVSW-6262 Add support for Reduce Prod
by Teresa Charlin
· 2 years, 11 months ago
dbf5f31
IVGCVSW-6311 Add the pre-import backend capability
by Finn Williams
· 2 years, 10 months ago
2d0679f
IVGCVSW-6249 Add ProfilingDetails Macros to all workloads in Neon
by Keith Davis
· 2 years, 11 months ago
5a64f22
IVGCVSW-5980 Add Descriptor, TensorInfo and Convolution algorithm to JSON
by Keith Davis
· 2 years, 11 months ago
554fa09
IVGCVSW-5980 JSON profiling output
by Keith Davis
· 3 years ago
7af2fd7
Fix Unittest failures in CL Master
by Nikhil Raj
· 3 years ago
89fd793
Disable Sin support in Neon
by Nikhil Raj
· 3 years ago
50de4fa
IVGCVSW-6088 Add Sin and Log to ElementWiseUnary
by Teresa Charlin
· 3 years, 1 month ago
53ef795
IVGCVSW-5826 Change weights layout for depthwise to [1,H,W,I*M]
by Jan Eilers
· 3 years, 1 month ago
1625efc
IVGCVSW-5963 'Move unit tests to new framework'
by Sadik Armagan
· 3 years, 1 month ago
b9af86e
IVGCVSW-5855 Refactor the reporting of capabilities from backends
by Finn Williams
· 3 years, 1 month ago
483c811
IVGCVSW-5962 Remove boost::multi_array
by Sadik Armagan
· 3 years, 1 month ago
5fc0fd6
MLCE-418 Reduce layer does not support multiple axes
by Matthew Sloyan
· 3 years, 2 months ago
25ab3a8
IVGCVSW-5964 Removing some remaining boost utility usages from tests.
by Colm Donelan
· 3 years, 1 month ago
e5f0b24
IVGCVSW-5818 Enable import on GPU
by Narumol Prangnawarat
· 3 years, 2 months ago
ae12306
Revert "MLCE-418 Reduce layer does not support multiple axes"
by Matthew Sloyan
· 3 years, 2 months ago
d905dec
MLCE-418 Reduce layer does not support multiple axes
by Matthew Sloyan
· 3 years, 2 months ago
1f58f03
IVGCVSW-5815 Generalise ConstCpuTensorHandle
by James Conroy
· 3 years, 2 months ago
73d3e2e
IVGCVSW-5819 5820 5821 Add MemorySourceFlags to TensorHandleFactoryRegistry::GetFactory
by Francis Murtagh
· 3 years, 2 months ago
48f011e
IVGCVSW-5419 'Add ACL Support for CAST Operator'
by Sadik Armagan
· 3 years, 2 months ago
7c67fab
IVGCVSW-5803 Delegate Unit Tests Failure on Android: Normalization & Softmax
by Keith Davis
· 3 years, 3 months ago
aede8ca
IVGCVSW-5783 'Add AsyncExecution Capability'
by Sadik Armagan
· 3 years, 3 months ago
92c5072
IVGCVSW-5676 Fixing build failure in backends jenkins job.
by Colm Donelan
· 3 years, 3 months ago
ade8c1d
IVGCVSW-5676 Eliminating some Neon unit tests for debug builds.
by Colm Donelan
· 3 years, 3 months ago
f0a6dec
IVGCVSW-5736 and IVGCVSW-5743 'NonConstWeights: Update front-end and TfLiteDelegate support for FullyConnected Operator'
by Sadik Armagan
· 3 years, 3 months ago
3d2b4b2
IVGCVSW-4873 Implement Pimpl Idiom for INetwork and IOptimizedNetwork
by Francis Murtagh
· 3 years, 4 months ago
0a7dc6b
IVGCVSW-5685 Add CpuAcc specific configuration option numberOfThreads
by Matthew Sloyan
· 3 years, 5 months ago
a274748
MLCE-347 'REDUCE_MIN, REDUCE_MAX, REDUCE_SUM Support'
by Sadik Armagan
· 3 years, 5 months ago
d92a6e4
IVGCVSW-4873 Implement Pimpl Idiom for IRuntime
by Kevin May
· 3 years, 5 months ago
e561795
Update ACL pin to 6a4ebe1305b544aec1ba0bfc67ed65d94fcc8c2e
by Nikhil Raj
· 3 years, 5 months ago
5dc64fe
IVGCVSW-5525 Handle Neon optionality on 32 bit linux platforms
by Francis Murtagh
· 3 years, 5 months ago
d672f5d
IVGCVSW-5616 Don't fuse activation if quantization parameters are different
by Teresa Charlin
· 3 years, 5 months ago
13816cd
Renaming NEActivationLayerKernel to CpuActivationKernel
by Nikhil Raj
· 3 years, 6 months ago
c11ba46
IVGCVSW-4625 Add CL Rank Workload
by David Monahan
· 3 years, 7 months ago
97451b4
IVGCVSW-4626 Add Neon Rank Workload
by David Monahan
· 3 years, 7 months ago
31c39be
IVGCVSW-5568 Revert "IVGCVSW-5563 Fix Crash on model with FullyConnected Sigmoid Activation"
by Teresa Charlin
· 3 years, 7 months ago
58f8ecd
IVGCVSW-5093 Remove redundant LogicalUnary functions
by James Conroy
· 3 years, 7 months ago
be25d94
IVGCVSW-5563 Fix Crash on model with Fullyconnected Sigmoid Activation
by Kevin May
· 3 years, 7 months ago
177df1e
IVGCVSW-5093 Add NEON Logical workload
by James Conroy
· 3 years, 8 months ago
0c95f4c
Fix logical vts skip
by Narumol Prangnawarat
· 3 years, 7 months ago
1ac690a
MLCE-278-IVGCVSW-5530 FusedActivation issues
by Mike Kelly
· 3 years, 7 months ago
07810fc
IVGCVSW-5328-5329 Fuse Activation
by Mike Kelly
· 3 years, 8 months ago
30dfc54
Update ACL pin to d7341fb9e3b24b904edf7ac9d83e1e063bc77765
by Teresa Charlin
· 3 years, 8 months ago
265e53e
IVGCVSW-5322 Fix segfault between Neon and Cl layers
by Narumol Prangnawarat
· 3 years, 8 months ago
f9f33a0
IVGCVSW-5335 Added Documentation for fast_math
by Mike Kelly
· 3 years, 9 months ago
f290d88
IVGCVSW-4489 Remove remaining occurrence of boost::format
by Matthew Sloyan
· 3 years, 9 months ago
3255d77
Remove Resize from list of layers that need padding in Neon
by Teresa Charlin
· 3 years, 9 months ago
Next »