diff options
| author | Georgios Pinitas <georgios.pinitas@arm.com> | 2018-07-25 17:57:49 +0100 |
|---|---|---|
| committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:54:54 +0000 |
| commit | ef776a844741826fc4740ad24714866aaeb35b8e (patch) | |
| tree | 55e79221c564df4e0c5fab5a684597e065667c13 /arm_compute/runtime/NEON/functions | |
| parent | a855af10a486c53c2271361cb87f349eca64b749 (diff) | |
| download | ComputeLibrary-ef776a844741826fc4740ad24714866aaeb35b8e.tar.gz | |
COMPMID-1386: Add FC convert weights on NEON
Change-Id: I7a3c6db9285e3899494f496b2562d80cec1b6521
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/141407
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'arm_compute/runtime/NEON/functions')
| -rw-r--r-- | arm_compute/runtime/NEON/functions/NEFullyConnectedLayer.h | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/arm_compute/runtime/NEON/functions/NEFullyConnectedLayer.h b/arm_compute/runtime/NEON/functions/NEFullyConnectedLayer.h index 92ca17a3a4..fe0f2f03f7 100644 --- a/arm_compute/runtime/NEON/functions/NEFullyConnectedLayer.h +++ b/arm_compute/runtime/NEON/functions/NEFullyConnectedLayer.h @@ -30,6 +30,7 @@ #include "arm_compute/core/NEON/kernels/NEIm2ColKernel.h" #include "arm_compute/core/NEON/kernels/NETransposeKernel.h" #include "arm_compute/runtime/MemoryGroup.h" +#include "arm_compute/runtime/NEON/functions/NEConvertFullyConnectedWeights.h" #include "arm_compute/runtime/NEON/functions/NEGEMM.h" #include "arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h" #include "arm_compute/runtime/NEON/functions/NEGEMMLowpOutputStage.h" @@ -129,6 +130,7 @@ private: MemoryGroup _memory_group; NEIm2ColKernel _im2col_kernel; + NEConvertFullyConnectedWeights _convert_weights; NEFullyConnectedLayerReshapeWeights _reshape_weights_function; NEGEMM _mm_gemm; NEGEMMLowpMatrixMultiplyCore _mm_gemmlowp; @@ -136,8 +138,10 @@ private: NEGEMMMatrixAccumulateBiasesKernel _accumulate_biases_kernel; Tensor _im2col_output; Tensor _gemmlowp_output; + Tensor _converted_weights_output; Tensor _reshape_weights_output; const ITensor *_original_weights; + bool _are_weights_converted; bool _are_weights_reshaped; bool _is_fc_after_conv; bool _accumulate_biases; |