diff options
author | Georgios Pinitas <georgios.pinitas@arm.com> | 2018-09-18 13:42:51 +0100 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:54:54 +0000 |
commit | 64f1a908841913049ccc0eb941b5b213290d7bf7 (patch) | |
tree | 4c9ecadbfd901ae4adfd1c2f2fd188f10540e3aa | |
parent | bf28a3cde6f77cbac3e3409d5597085ccbc71499 (diff) | |
download | ComputeLibrary-64f1a908841913049ccc0eb941b5b213290d7bf7.tar.gz |
COMPMID-1591: Fix NEPoolingLayer for NHWC
Restore window step across width to 4 for FP32 instead of the whole row
as the kernel code was inconsistent with this decision.
Change-Id: I7c4dcdf960b8cbc970a36fa1df39df2c6f000c86
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/148908
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: bsgcomp <bsgcomp@arm.com>
-rw-r--r-- | src/core/NEON/kernels/NEPoolingLayerKernel.cpp | 2 | ||||
-rw-r--r-- | tests/datasets/ShapeDatasets.h | 4 |
2 files changed, 3 insertions, 3 deletions
diff --git a/src/core/NEON/kernels/NEPoolingLayerKernel.cpp b/src/core/NEON/kernels/NEPoolingLayerKernel.cpp index a8e3be28fe..1fa8f47332 100644 --- a/src/core/NEON/kernels/NEPoolingLayerKernel.cpp +++ b/src/core/NEON/kernels/NEPoolingLayerKernel.cpp @@ -242,7 +242,7 @@ std::pair<Status, Window> validate_and_configure_window(ITensorInfo *input, ITen case DataType::F32: if(is_nhwc) { - num_elems_processed_per_iteration = std::max(4, ceil_to_multiple<int>(input->dimension(0), 2)); + num_elems_processed_per_iteration = 4; break; } switch(pool_size_x) diff --git a/tests/datasets/ShapeDatasets.h b/tests/datasets/ShapeDatasets.h index 5ec3d4c5cb..483fada5d8 100644 --- a/tests/datasets/ShapeDatasets.h +++ b/tests/datasets/ShapeDatasets.h @@ -164,8 +164,8 @@ public: { // Batch size 1 TensorShape{ 11U, 11U }, - TensorShape{ 27U, 13U, 2U }, - TensorShape{ 128U, 64U, 1U, 3U }, + TensorShape{ 27U, 13U, 7U }, + TensorShape{ 31U, 27U, 17U, 2U }, // Batch size 4 TensorShape{ 11U, 11U, 3U, 4U }, TensorShape{ 27U, 13U, 2U, 4U }, |