diff options
author | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2017-06-23 10:38:25 +0100 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-09-17 14:15:39 +0100 |
commit | 5cb4c42cb5d781a44409ebc97a408e1379ce182d (patch) | |
tree | dbb544322eacee38f9719225e037aca90ba6fbf3 /arm_compute/runtime/CL/functions | |
parent | 0a8334cb78dae66fdc31257a96ba15f7c41bde50 (diff) | |
download | ComputeLibrary-5cb4c42cb5d781a44409ebc97a408e1379ce182d.tar.gz |
COMPMID-414 - Port CLConvolutionLayer to support 8 bit fixed point - CLWeightsReshapeKernel
Change-Id: Ie32e6bdd557a8243eb9988aa7eab4e4ca2291e79
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/78701
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Moritz Pflanzer <moritz.pflanzer@arm.com>
Diffstat (limited to 'arm_compute/runtime/CL/functions')
-rw-r--r-- | arm_compute/runtime/CL/functions/CLConvolutionLayer.h | 14 | ||||
-rw-r--r-- | arm_compute/runtime/CL/functions/CLLocallyConnectedLayer.h | 18 |
2 files changed, 16 insertions, 16 deletions
diff --git a/arm_compute/runtime/CL/functions/CLConvolutionLayer.h b/arm_compute/runtime/CL/functions/CLConvolutionLayer.h index 6a40396f9a..8030b40a71 100644 --- a/arm_compute/runtime/CL/functions/CLConvolutionLayer.h +++ b/arm_compute/runtime/CL/functions/CLConvolutionLayer.h @@ -53,7 +53,7 @@ public: CLConvolutionLayerReshapeWeights(); /** Set the input and output tensors. * - * @param[in] weights Weights tensor. Weights are 4D tensor with dimensions [kernel_x, kernel_y, IFM, OFM]. Data type supported: F32. + * @param[in] weights Weights tensor. Weights are 4D tensor with dimensions [kernel_x, kernel_y, IFM, OFM]. Data type supported: QS8/F16/F32. * @param[in] biases Biases tensor. Shared biases supported. Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p weights. * @param[out] output Destination tensor. Data types supported: Same as @p weights. * @param[in] transpose1xW True if the weights are to undergo a 1xW transposition after reshaping (in case of GEMM operation), false otherwise. @@ -64,16 +64,16 @@ public: void run() override; private: - CLConvolutionLayerWeightsReshapeKernel _weights_reshape_kernel; - CLGEMMTranspose1xWKernel _weights_transposed_kernel; - CLTensor _weights_reshaped; - bool _transpose1xW; + CLWeightsReshapeKernel _weights_reshape_kernel; + CLGEMMTranspose1xWKernel _weights_transposed_kernel; + CLTensor _weights_reshaped; + bool _transpose1xW; }; /** Basic function to compute the convolution layer. This function calls the following OpenCL kernels: * - * -# @ref CLConvolutionLayerWeightsReshapeKernel (executed only once for each configuration) - * -# @ref CLGEMMTranspose1xWKernel (executed only once for each configuration) + * -# @ref CLWeightsReshapeKernel (executed only once for each configuration) + * -# @ref CLGEMMTranspose1xWKernel (executed only once for each configuration) * -# @ref CLIm2ColKernel * -# @ref CLGEMMInterleave4x4Kernel * -# @ref CLGEMMMatrixMultiplyKernel diff --git a/arm_compute/runtime/CL/functions/CLLocallyConnectedLayer.h b/arm_compute/runtime/CL/functions/CLLocallyConnectedLayer.h index b4e469196e..5f4f1ba1d7 100644 --- a/arm_compute/runtime/CL/functions/CLLocallyConnectedLayer.h +++ b/arm_compute/runtime/CL/functions/CLLocallyConnectedLayer.h @@ -39,7 +39,7 @@ class ICLTensor; /** Basic function to compute the locally connected layer. This function calls the following OpenCL kernels: * - * -# @ref CLLocallyConnectedLayerWeightsReshapeKernel (executed only once for each configuration) + * -# @ref CLWeightsReshapeKernel (executed only once for each configuration) * -# @ref CLIm2ColKernel * -# @ref CLLocallyConnectedMatrixMultiplyKernel * -# @ref CLCol2ImKernel @@ -66,14 +66,14 @@ public: void run() override; private: - CLIm2ColKernel _input_im2col_kernel; - CLLocallyConnectedLayerWeightsReshapeKernel _weights_reshape_kernel; - CLLocallyConnectedMatrixMultiplyKernel _mm_kernel; - CLCol2ImKernel _output_col2im_kernel; - CLTensor _input_im2col_reshaped; - CLTensor _weights_reshaped; - CLTensor _gemm_output; - bool _is_first_run; + CLIm2ColKernel _input_im2col_kernel; + CLWeightsReshapeKernel _weights_reshape_kernel; + CLLocallyConnectedMatrixMultiplyKernel _mm_kernel; + CLCol2ImKernel _output_col2im_kernel; + CLTensor _input_im2col_reshaped; + CLTensor _weights_reshaped; + CLTensor _gemm_output; + bool _is_first_run; }; } #endif /* __ARM_COMPUTE_CLLOCALLYCONNECTEDLAYER_H__ */ |