aboutsummaryrefslogtreecommitdiff
path: root/arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h
diff options
context:
space:
mode:
authorVidhya Sudhan Loganathan <vidhyasudhan.loganathan@arm.com>2019-11-04 14:42:08 +0000
committerMichele Di Giorgio <michele.digiorgio@arm.com>2019-11-14 16:25:06 +0000
commit951b8a4c01de2810349b6f16cf9bbba7578484fa (patch)
tree8b3ab1c04279da7be3afd6632a9894b6197c1e1b /arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h
parentcd4e9abf7a165f15ccd10ac4541365d4f8a6db19 (diff)
downloadComputeLibrary-951b8a4c01de2810349b6f16cf9bbba7578484fa.tar.gz
COMPMID-2309 : CLConvolutionLayer: support QUANT8_SYMM_PER_CHANNEL filters
Change-Id: I16f6758b768ede404a064db057302ded706e1e8a Signed-off-by: Vidhya Sudhan Loganathan <vidhyasudhan.loganathan@arm.com> Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com> Reviewed-on: https://review.mlplatform.org/c/2215 Tested-by: Arm Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h')
-rw-r--r--arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h63
1 files changed, 37 insertions, 26 deletions
diff --git a/arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h b/arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h
index de06c88d5c..301c67331e 100644
--- a/arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h
+++ b/arm_compute/core/CL/kernels/CLGEMMLowpOffsetContributionOutputStageKernel.h
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -51,39 +51,47 @@ public:
CLGEMMLowpOffsetContributionOutputStageKernel &operator=(CLGEMMLowpOffsetContributionOutputStageKernel &&) = default;
/** Initialise the kernel's input and output.
*
- * @param[in] mm_result Input tensor containing the result of @ref CLGEMMLowpMatrixMultiplyKernel. Data type supported: S32
- * @param[in] vector_sum_col Input row-vector of sums of all the entries in each column of matrix B.
- * Note: vector_sum_col can be a nullptr in case a_offset = 0. Data type supported: same as @p mm_result
- * @param[in] vector_sum_row Input row-vector of sums of all the entries in each row of matrix A.
- * Note: vector_sum_row can be a nullptr in case b_offset = 0. Data type supported: same as @p mm_result
- * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.
- * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.
- * @param[out] output Output tensor. Data type supported: QASYMM8
- * @param[in] k Number of matrix A columns or Matrix B rows
- * @param[in] a_offset Offset to be added to each element of the matrix A.
- * @param[in] b_offset Offset to be added to each element of the matrix B.
- * @param[in] output_stage GEMMLowp output stage info
+ * @param[in] mm_result Input tensor containing the result of @ref CLGEMMLowpMatrixMultiplyKernel. Data type supported: S32
+ * @param[in] vector_sum_col Input row-vector of sums of all the entries in each column of matrix B.
+ * Note: vector_sum_col can be a nullptr in case a_offset = 0. Data type supported: same as @p mm_result
+ * @param[in] vector_sum_row Input row-vector of sums of all the entries in each row of matrix A.
+ * Note: vector_sum_row can be a nullptr in case b_offset = 0. Data type supported: same as @p mm_result
+ * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.
+ * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.
+ * @param[out] output Output tensor. Data type supported: QASYMM8.
+ * @param[in] k Number of matrix A columns or Matrix B rows
+ * @param[in] a_offset Offset to be added to each element of the matrix A.
+ * @param[in] b_offset Offset to be added to each element of the matrix B.
+ * @param[in] output_stage GEMMLowp output stage info
+ * @param[in] output_multipliers Output multipliers tensor. In case of per-channel quantization, the number of multipliers must be equal to the number of filters (OFM).
+ * Supported data types: S32
+ * @param[in] output_shifts Output shifts tensor. In case of per-channel quantization, the number of multipliers must be equal to the number of filters (OFM).
+ * Supported data types: S32
*/
void configure(const ICLTensor *mm_result, const ICLTensor *vector_sum_col, const ICLTensor *vector_sum_row, const ICLTensor *bias, ICLTensor *output, int32_t k, int32_t a_offset, int32_t b_offset,
- const GEMMLowpOutputStageInfo &output_stage);
+ const GEMMLowpOutputStageInfo &output_stage, const ICLTensor *output_multipliers, const ICLTensor *output_shifts);
/** Static function to check if given info will lead to a valid configuration of @ref CLGEMMLowpOffsetContributionKernel
*
- * @param[in] mm_result Input tensor containing the result of @ref CLGEMMLowpOffsetContributionKernel. Data type supported: S32 or QASYMM8 if output_stage != NONE
- * @param[in] vector_sum_col Input row-vector of sums of all the entries in each column of matrix B.
- * Note: vector_sum_col can be a nullptr in case a_offset = 0. Data type supported: same as @p mm_result
- * @param[in] vector_sum_row Input row-vector of sums of all the entries in each row of matrix A.
- * Note: vector_sum_row can be a nullptr in case b_offset = 0. Data type supported: same as @p mm_result
- * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.
- * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.
- * @param[in] output Output tensor. Data type supported: QASYMM8
- * @param[in] a_offset Offset to be added to each element of the matrix A.
- * @param[in] b_offset Offset to be added to each element of the matrix B.
- * @param[in] output_stage GEMMLowp output stage info
+ * @param[in] mm_result Input tensor containing the result of @ref CLGEMMLowpOffsetContributionKernel. Data type supported: S32 or QASYMM8 if output_stage != NONE
+ * @param[in] vector_sum_col Input row-vector of sums of all the entries in each column of matrix B.
+ * Note: vector_sum_col can be a nullptr in case a_offset = 0. Data type supported: same as @p mm_result
+ * @param[in] vector_sum_row Input row-vector of sums of all the entries in each row of matrix A.
+ * Note: vector_sum_row can be a nullptr in case b_offset = 0. Data type supported: same as @p mm_result
+ * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.
+ * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.
+ * @param[in] output Output tensor. Data type supported: QASYMM8.
+ * @param[in] a_offset Offset to be added to each element of the matrix A.
+ * @param[in] b_offset Offset to be added to each element of the matrix B.
+ * @param[in] output_stage GEMMLowp output stage info
+ * @param[in] output_multipliers Output multipliers tensor info. In case of per-channel quantization, the number of multipliers must be equal to the number of filters (OFM).
+ * Supported data types: S32
+ * @param[in] output_shifts Output shifts tensor info. In case of per-channel quantization, the number of multipliers must be equal to the number of filters (OFM).
+ * Supported data types: S32
*
* @return a status
*/
static Status validate(const ITensorInfo *mm_result, const ITensorInfo *vector_sum_col, const ITensorInfo *vector_sum_row, const ITensorInfo *bias, const ITensorInfo *output, int32_t a_offset,
- int32_t b_offset, const GEMMLowpOutputStageInfo &output_stage);
+ int32_t b_offset, const GEMMLowpOutputStageInfo &output_stage, const ITensorInfo *output_multipliers, const ITensorInfo *output_shifts);
// Inherited methods overridden:
void run(const Window &window, cl::CommandQueue &queue) override;
@@ -94,6 +102,9 @@ private:
const ICLTensor *_vector_sum_row;
const ICLTensor *_bias;
ICLTensor *_output;
+ const ICLTensor *_output_multipliers;
+ const ICLTensor *_output_shifts;
+ bool _is_quantized_per_channel;
};
} // namespace arm_compute