diff options
| author | Aron Virginas-Tar <Aron.Virginas-Tar@arm.com> | 2019-11-26 14:04:54 +0000 |
|---|---|---|
| committer | Aron Virginas-Tar <Aron.Virginas-Tar@arm.com> | 2019-11-26 15:00:28 +0000 |
| commit | 21fc28bb68d7a6232eea3a0e8cfd800d95346ea4 (patch) | |
| tree | 5eca96ac31e3f8c62da3ad7742834180a0f7eeac /src/backends/neon/NeonLayerSupport.cpp | |
| parent | 9b6d4246804d39e5b7570992d63e31a044442385 (diff) | |
| download | armnn-21fc28bb68d7a6232eea3a0e8cfd800d95346ea4.tar.gz | |
IVGCVSW-3866 Add support for per-channel QSymm8 Convolution2d to Neon backend
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: If7d366b2894050c53a651016165f34be1a2ce7a6
Diffstat (limited to 'src/backends/neon/NeonLayerSupport.cpp')
| -rw-r--r-- | src/backends/neon/NeonLayerSupport.cpp | 5 |
1 files changed, 0 insertions, 5 deletions
diff --git a/src/backends/neon/NeonLayerSupport.cpp b/src/backends/neon/NeonLayerSupport.cpp index ed0f41a888..4474b12d37 100644 --- a/src/backends/neon/NeonLayerSupport.cpp +++ b/src/backends/neon/NeonLayerSupport.cpp @@ -274,11 +274,6 @@ bool NeonLayerSupport::IsConvolution2dSupported(const TensorInfo& input, const Optional<TensorInfo>& biases, Optional<std::string&> reasonIfUnsupported) const { - if (weights.HasPerAxisQuantization()) - { - return false; - } - // Multiplier > 1.0f currently not supported in ACL if ((input.GetQuantizationScale() * weights.GetQuantizationScale()) / output.GetQuantizationScale() > 1.0f) { |