diff options
author | Sheri Zhang <sheri.zhang@arm.com> | 2020-02-26 10:30:15 +0000 |
---|---|---|
committer | Sheri Zhang <sheri.zhang@arm.com> | 2020-03-12 13:50:31 +0000 |
commit | a14817a7eee8b8cb7e5ccb6186ca01c23eec2629 (patch) | |
tree | 791eebc4b5f47ab055e5911a3fa0d2a0dafa7f7a /tests/validation/reference/DeconvolutionLayer.h | |
parent | c7b183ab741650653289f8ce3bdeb4926521fdbd (diff) | |
download | ComputeLibrary-a14817a7eee8b8cb7e5ccb6186ca01c23eec2629.tar.gz |
COMPMID-2756: Add support for QASYMM8_SIGNED in CLDeconvolutionLayer
Tests cover for CLGEMMDeconvolution and CLDirectDeconvolution.
Change-Id: I9a26d0adef1d177ffad39a8992a2dc65327f07e5
Signed-off-by: Sheri Zhang <sheri.zhang@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/2853
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'tests/validation/reference/DeconvolutionLayer.h')
-rw-r--r-- | tests/validation/reference/DeconvolutionLayer.h | 8 |
1 files changed, 5 insertions, 3 deletions
diff --git a/tests/validation/reference/DeconvolutionLayer.h b/tests/validation/reference/DeconvolutionLayer.h index db394faa68..fff529a719 100644 --- a/tests/validation/reference/DeconvolutionLayer.h +++ b/tests/validation/reference/DeconvolutionLayer.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2017-2019 ARM Limited. + * Copyright (c) 2017-2020 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -37,9 +37,11 @@ namespace reference { /** Deconvolution reference implementation. * - * src Input tensor. 3 lower dimensions represent a single input, and an optional 4th dimension for batch of inputs. Data types supported: F32. + * src Input tensor. 3 lower dimensions represent a single input, and an optional 4th dimension for batch of inputs. + * Data types supported: QASYMM8/QASYMM8_SIGNED/F32/F16. * weights The 4d weights with dimensions [width, height, OFM, IFM]. Data type supported: Same as @p input. - * bias Optional, ignored if NULL. The biases have one dimension. Data type supported: Same as @p input. + * bias Optional, ignored if NULL. The biases have one dimension. + * Data type supported: Same as @p input, except for input of QASYMM8 and QASYMM8_SIGNED type where biases should be of S32 type * output_shape Output tensor shape. The output has the same number of dimensions as the @p input. * info Contains padding and policies to be used in the deconvolution, this is decribed in @ref PadStrideInfo. * a The number of zeros added to right and top edges of the input. |