diff options
author | Manuel Bottini <manuel.bottini@arm.com> | 2020-04-08 10:15:51 +0100 |
---|---|---|
committer | Manuel Bottini <manuel.bottini@arm.com> | 2020-04-17 14:10:38 +0000 |
commit | 4c6bd514a8d424a29b776754f1b3426fa3a8c339 (patch) | |
tree | ed20ea238ae069ab138b62ea6879e0c1b955b606 /arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h | |
parent | d2f6d96cdc2ad2169c4abd0a8c4884f61ed6d186 (diff) | |
download | ComputeLibrary-4c6bd514a8d424a29b776754f1b3426fa3a8c339.tar.gz |
COMPMID-3280: Make all ML primitives for CL use the new interface - Part 1
- Only CLKernels have been updated
Change-Id: Ife55b847c2e39e712a186eb6ca452503d5b66937
Signed-off-by: Manuel Bottini <manuel.bottini@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/3001
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h')
-rw-r--r-- | arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h | 19 |
1 files changed, 18 insertions, 1 deletions
diff --git a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h index 82c4091c6e..fe34735fe4 100644 --- a/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h +++ b/arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2017-2019 ARM Limited. + * Copyright (c) 2017-2020 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -68,6 +68,23 @@ public: */ void configure(const ICLTensor *input0, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float alpha, float beta = 0.f, bool is_interleaved_transposed = true, const GEMMReshapeInfo &reshape_info = GEMMReshapeInfo(), bool fp_mixed_precision = false, const ActivationLayerInfo &activation_info = ActivationLayerInfo()); + /** Initialise the kernel's input, output and alpha + * + * @param[in] compile_context The compile context to be used. + * @param[in] input0 Input tensor containing the Matrix A. Data types supported: F16/F32 + * @param[in] input1 Input tensor containing the Matrix B. Data type supported: same as @p input0 + * @param[in] input2 Input tensor containing the Matrix C (bias). Can be nullptr. Data type supported: same as @p input0 + * @param[out] output Output tensor to store the result of matrix multiplication. Data type supported: same as @p input0 + * @param[in] alpha Weight of the matrix product + * @param[in] beta (Optional) Weight of vector C. Default value is 0. Only beta = 1 is currently supported. + * @param[in] is_interleaved_transposed (Optional) True if input0 and input1 have been reshaped respectively using @ref CLGEMMReshapeLHSMatrixKernel and @ref CLGEMMReshapeRHSMatrixKernel + * @param[in] reshape_info (Optional) GEMM reshape info. If is_interleaved_transposed = true, this object must contain the information to understand how the matrix A and matrix B have been reshaped + * @param[in] fp_mixed_precision (Optional) Use wider accumulators (32 bit instead of 16 for FP16) to improve accuracy + * @param[in] activation_info (Optional) Activation to apply after the matrix multiplication + * + */ + void configure(CLCompileContext &compile_context, const ICLTensor *input0, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float alpha, float beta = 0.f, + bool is_interleaved_transposed = true, const GEMMReshapeInfo &reshape_info = GEMMReshapeInfo(), bool fp_mixed_precision = false, const ActivationLayerInfo &activation_info = ActivationLayerInfo()); /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMMatrixMultiplyKernel * * @param[in] input0 Input tensor containing the Matrix A info. Data types supported: F16/F32 |