diff options
| author | Manuel Bottini <manuel.bottini@arm.com> | 2019-10-25 10:28:28 +0100 |
|---|---|---|
| committer | Manuel Bottini <manuel.bottini@arm.com> | 2019-10-25 16:06:05 +0000 |
| commit | 279814bfdc3e2ec3ed6c4e248356b4e0b2b2abc0 (patch) | |
| tree | 2ee493496aaa44d05c6accaa20cd3eba1cc91960 /src | |
| parent | 7736389c268766efa53ff85dc7260f36fa25d2e7 (diff) | |
| download | ComputeLibrary-279814bfdc3e2ec3ed6c4e248356b4e0b2b2abc0.tar.gz | |
COMPMID-2481: Add InputQuantizationInfo!=OutputQuantization Info tests for DeconvolutionLayer
Fix on CLGEMMDeconvolutionLayer to use actual output quantization
Change-Id: Id9ab2d9a9c839425b5ac54a0501f7bea5a809237
Signed-off-by: Manuel Bottini <manuel.bottini@arm.com>
Reviewed-on: https://review.mlplatform.org/c/2164
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Diffstat (limited to 'src')
| -rw-r--r-- | src/runtime/CL/functions/CLGEMMDeconvolutionLayer.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/runtime/CL/functions/CLGEMMDeconvolutionLayer.cpp b/src/runtime/CL/functions/CLGEMMDeconvolutionLayer.cpp index 78e1ae7f05..4671be5b61 100644 --- a/src/runtime/CL/functions/CLGEMMDeconvolutionLayer.cpp +++ b/src/runtime/CL/functions/CLGEMMDeconvolutionLayer.cpp @@ -279,7 +279,7 @@ void CLGEMMDeconvolutionLayer::configure(const ICLTensor *input, const ICLTensor { const UniformQuantizationInfo iq_info = input->info()->quantization_info().uniform(); const UniformQuantizationInfo wq_info = weights->info()->quantization_info().uniform(); - const UniformQuantizationInfo oq_info = _gemmlowp_final.info()->quantization_info().uniform(); + const UniformQuantizationInfo oq_info = output->info()->quantization_info().uniform(); float multiplier = iq_info.scale * wq_info.scale / oq_info.scale; int output_multiplier(0); |