diff options
| author | Michele Di Giorgio <michele.digiorgio@arm.com> | 2019-10-09 15:32:39 +0100 |
|---|---|---|
| committer | Michele Di Giorgio <michele.digiorgio@arm.com> | 2019-10-30 14:44:46 +0000 |
| commit | df4cf57c7394265b27d051cb1cf0152c53659126 (patch) | |
| tree | 87da5d6abeff65b2cee55b63f73bb268776af560 /arm_compute/core/NEON/kernels/NEDepthwiseConvolutionLayerNativeKernel.h | |
| parent | 8b72199f25487040713d1668c998fdde3707413c (diff) | |
| download | ComputeLibrary-df4cf57c7394265b27d051cb1cf0152c53659126.tar.gz | |
COMPMID-2306: CLDepthwiseConvolution: support for QUANT8_PER_CHANNEL_SYMM
Change-Id: I18c886400daa2dcba0b91011bc4e503d807a4732
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/2143
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'arm_compute/core/NEON/kernels/NEDepthwiseConvolutionLayerNativeKernel.h')
| -rw-r--r-- | arm_compute/core/NEON/kernels/NEDepthwiseConvolutionLayerNativeKernel.h | 12 |
1 files changed, 8 insertions, 4 deletions
diff --git a/arm_compute/core/NEON/kernels/NEDepthwiseConvolutionLayerNativeKernel.h b/arm_compute/core/NEON/kernels/NEDepthwiseConvolutionLayerNativeKernel.h index a0205f1ea6..6db1a767d8 100644 --- a/arm_compute/core/NEON/kernels/NEDepthwiseConvolutionLayerNativeKernel.h +++ b/arm_compute/core/NEON/kernels/NEDepthwiseConvolutionLayerNativeKernel.h @@ -58,8 +58,10 @@ public: * @note Supported data layouts: NHWC * * @param[in] input Source tensor. DataType supported: QASYMM8/F16/F32. - * @param[in] weights Weights tensor. This is a 3D tensor with dimensions [IFM, W, H]. Data type supported: Same as @p input. - * @param[in] biases Biases tensor. A 1D tensor with dimensions [IFM]. Must be nullptr if not needed. Data type supported: Same as @p input. + * @param[in] weights Weights tensor. This is a 3D tensor with dimensions [IFM, W, H]. + * Data type supported: Same as @p input or QASYMM8/QSYMM8_PER_CHANNEL when @p input is QASYMM8. + * @param[in] biases Biases tensor. A 1D tensor with dimensions [IFM]. Must be nullptr if not needed. + * Data type supported: Same as @p input, S32 when input is QASYMM8. * @param[out] output Destination tensor. Data type supported: Same as @p input. * @param[in] conv_info Padding and stride information to use for the convolution. * @param[in] depth_multiplier (Optional) Multiplier to apply to the input's depth in order to retrieve the output's depth. Defaults to 1. @@ -73,8 +75,10 @@ public: * @note Supported data layouts: NHWC * * @param[in] input Source tensor info. DataType supported: QASYMM8/F16/F32. - * @param[in] weights Weights tensor info. This is a 3D tensor with dimensions [IFM, W, H]. Data type supported: Same as @p input. - * @param[in] biases Biases tensor info. A 1D tensor with dimensions [IFM]. Must be nullptr if not needed. Data type supported: Same as @p input. + * @param[in] weights Weights tensor info. This is a 3D tensor with dimensions [IFM, W, H]. + * Data type supported: Same as @p input or QASYMM8/QSYMM8_PER_CHANNEL when @p input is QASYMM8. + * @param[in] biases Biases tensor info. A 1D tensor with dimensions [IFM]. Must be nullptr if not needed. + * Data type supported: Same as @p input, S32 when input is QASYMM8. * @param[in] output Destination tensor info. Data type supported: Same as @p input. * @param[in] conv_info Padding and stride information to use for the convolution. * @param[in] depth_multiplier (Optional) Multiplier to apply to the input's depth in order to retrieve the output's depth. Defaults to 1. |