diff options
author | Michalis Spyrou <michalis.spyrou@arm.com> | 2017-11-30 14:25:57 +0000 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:42:17 +0000 |
commit | afa5d817b1d083837cd7ea30d32f845d82620c12 (patch) | |
tree | 1ca2a27ab7108b7137b96fc1547a8b5ac5d9c8f7 /arm_compute/core | |
parent | 631c41a4e3645a948b0f597caa77e8fa91ca0efc (diff) | |
download | ComputeLibrary-afa5d817b1d083837cd7ea30d32f845d82620c12.tar.gz |
COMPMID-617 Add validation methods to Kernels
- NEActivationLayer
- NESoftmax
- NEDirectConvolutionLayer
- NENormalizationLayer
- NEPoolingLayer
Change-Id: Ib279f1c1b7f9247679b0d6593aed7393da8fe87b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111335
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Diffstat (limited to 'arm_compute/core')
7 files changed, 93 insertions, 4 deletions
diff --git a/arm_compute/core/NEON/kernels/NEActivationLayerKernel.h b/arm_compute/core/NEON/kernels/NEActivationLayerKernel.h index ef51cbe841..e8c032aaeb 100644 --- a/arm_compute/core/NEON/kernels/NEActivationLayerKernel.h +++ b/arm_compute/core/NEON/kernels/NEActivationLayerKernel.h @@ -59,6 +59,16 @@ public: * @param[in] activation_info Activation layer information. */ void configure(ITensor *input, ITensor *output, ActivationLayerInfo activation_info); + /** Static function to check if given info will lead to a valid configuration of @ref NEActivationLayerKernel + * + * @param[in] input Source tensor info. In case of @p output tensor info = nullptr, this tensor will store the result + * of the activation function. Data types supported: QS8/QS16/F16/F32. + * @param[in] output Destination tensor info. Data type supported: same as @p input + * @param[in] act_info Activation layer information. + * + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *output, const ActivationLayerInfo &act_info); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; diff --git a/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerBiasAccumulateKernel.h b/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerBiasAccumulateKernel.h index 14c8e9c7e0..05ade1c5dd 100644 --- a/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerBiasAccumulateKernel.h +++ b/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerBiasAccumulateKernel.h @@ -57,6 +57,16 @@ public: * Data type supported: Same as @p input */ void configure(ITensor *input, const ITensor *bias, ITensor *output = nullptr); + /** Static function to check if given info will lead to a valid configuration of @ref NEDirectConvolutionLayerBiasAccumulateKernel + * + * @param[in] input Input to add the bias to. If @p output is not specified then accumulation is done in-place. + * Data type supported: QS8/QS16/F16/F32 + * @param[in] bias The shared bias tensor to add. It must be 1D Tensor. Data type supported: Same as @p input + * @param[in] output (Optional) If the output tensor is specified the accumulation is done out-of-place. (Defaults to nullptr) + * Data type supported: Same as @p input + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output = nullptr); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; diff --git a/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerKernel.h b/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerKernel.h index 370ddca480..4529120f02 100644 --- a/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerKernel.h +++ b/arm_compute/core/NEON/kernels/NEDirectConvolutionLayerKernel.h @@ -62,6 +62,20 @@ public: * @param[in] conv_info Contains padding and stride information described in @ref PadStrideInfo. */ void configure(const ITensor *input, const ITensor *weights, ITensor *output, const PadStrideInfo &conv_info); + /** Static function to check if given info will lead to a valid configuration of @ref NEDirectConvolutionLayerKernel + * + * @param[in] input The input tensor to convolve. 3 lower dimensions represent a single input [width, height, IFM], + * while every optional dimension from 4 and above represent a batch of inputs. Data types supported: QS8/QS16/F16/F32. + * @param[in] weights Weights tensor. Weights are 4D tensor with dimensions [kernel_x, kernel_y, IFM, OFM]. + * The 3rd dimension must be the same as the input's volume 3rd dimension. + * Data type supported:Same as @p input. + * @param[in] output Output tensor. + * The 3rd dimensions must be equal to the 4th dimension of the @p kernels tensor. Data types supported: Same as @p input. + * @param[in] conv_info Contains padding and stride information described in @ref PadStrideInfo. + * + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *weights, const ITensorInfo *output, const PadStrideInfo &conv_info); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; diff --git a/arm_compute/core/NEON/kernels/NENormalizationLayerKernel.h b/arm_compute/core/NEON/kernels/NENormalizationLayerKernel.h index 40fae3520b..405daf1068 100644 --- a/arm_compute/core/NEON/kernels/NENormalizationLayerKernel.h +++ b/arm_compute/core/NEON/kernels/NENormalizationLayerKernel.h @@ -57,6 +57,18 @@ public: * @param[in] norm_info Normalization layer information like the normalization type, normalization size and other parameters. */ void configure(const ITensor *input, const ITensor *input_squared, ITensor *output, NormalizationLayerInfo norm_info); + /** Static function to check if given info will lead to a valid configuration of @ref NENormalizationLayerKernel + * + * @param[in] input Source tensor. 3 lower dims represent a single input with dimensions [width, height, IFM], + * and an optional 4th dimension for batch of inputs. Data types supported: QS8/QS16/FP16/F32. + * @param[in] input_squared Source with each element has been squared. 3 lower dims represent a single input with dimensions [width, height, IFM], + * Data type supported: same as @p input + * @param[in] output Destination tensor. Output will have the same number of dimensions as input. Data type supported: same as @p input + * @param[in] norm_info Normalization layer information like the normalization type, normalization size and other parameters. + * + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *input_squared, const ITensorInfo *output, NormalizationLayerInfo norm_info); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; diff --git a/arm_compute/core/NEON/kernels/NEPixelWiseMultiplicationKernel.h b/arm_compute/core/NEON/kernels/NEPixelWiseMultiplicationKernel.h index 05eb8d6ddc..10f990e7ef 100644 --- a/arm_compute/core/NEON/kernels/NEPixelWiseMultiplicationKernel.h +++ b/arm_compute/core/NEON/kernels/NEPixelWiseMultiplicationKernel.h @@ -64,9 +64,13 @@ public: void configure(const ITensor *input1, const ITensor *input2, ITensor *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy); /** Static function to check if given info will lead to a valid configuration of @ref NEPixelWiseMultiplicationKernel * - * @param[in] input1 An input tensor info. Data types supported: U8/QS8/QS16/S16/F16/F32 - * @param[in] input2 An input tensor info. Data types supported: U8, QS8 (only if @p input1 is QS8), QS16 (only if @p input1 is QS16), S16/F16 (only if @p input1 is F16), F32 (only if @p input1 is F32). - * @param[in] output The output tensor info. Data types supported: U8 (Only if both inputs are U8), QS8 (only if both inputs are QS8), QS16 (only if both inputs are QS16), S16/F16 (only if @p input1 is F16), F32 (only if both inputs are F32). + * @note For @p scale equal to 1/255 only round to nearest even (implemented as round half up) is supported. + * For all other scale values only round to zero (implemented as round towards minus infinity) is supported. + * For QS8/QS16 scale = 1 is the only supported value. + * + * @param[in] input1 An input tensor. Data types supported: U8/QS8/QS16/S16/F16/F32 + * @param[in] input2 An input tensor. Data types supported: U8, QS8 (only if @p input1 is QS8), QS16 (only if @p input1 is QS16), S16/F16 (only if @p input1 is F16), F32 (only if @p input1 is F32). + * @param[in] output The output tensor. Data types supported: U8 (Only if both inputs are U8), QS8 (only if both inputs are QS8), QS16 (only if both inputs are QS16), S16/F16 (only if @p input1 is F16), F32 (only if both inputs are F32). * @param[in] scale Scale to apply after multiplication. * Scale must be positive and its value must be either 1/255 or 1/2^n where n is between 0 and 15. * @param[in] overflow_policy Overflow policy. diff --git a/arm_compute/core/NEON/kernels/NEPoolingLayerKernel.h b/arm_compute/core/NEON/kernels/NEPoolingLayerKernel.h index 0a57a26f17..87d14e5f91 100644 --- a/arm_compute/core/NEON/kernels/NEPoolingLayerKernel.h +++ b/arm_compute/core/NEON/kernels/NEPoolingLayerKernel.h @@ -55,6 +55,17 @@ public: * @param[in] pool_info Contains pooling operation information described in @ref PoolingLayerInfo. */ void configure(const ITensor *input, ITensor *output, const PoolingLayerInfo &pool_info); + /** Static function to check if given info will lead to a valid configuration of @ref NEPoolingLayerKernel + * + * @note QS8, QS16 and F16 are supported for pool sizes 2 and 3 only + * + * @param[in] input Source tensor. Data types supported: QS8/QS16/F16/F32. + * @param[in] output Destination tensor. Data types supported: Same as @p input. + * @param[in] pool_info Contains pooling operation information described in @ref PoolingLayerInfo. + * + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *output, const PoolingLayerInfo &pool_info); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; @@ -144,7 +155,7 @@ private: const ITensor *_input; ITensor *_output; PoolingLayerInfo _pool_info; - int _num_elems_processed_per_iteration; + unsigned int _num_elems_processed_per_iteration; BorderSize _border_size; }; } // namespace arm_compute diff --git a/arm_compute/core/NEON/kernels/NESoftmaxLayerKernel.h b/arm_compute/core/NEON/kernels/NESoftmaxLayerKernel.h index c3e25181b6..0fecfac15a 100644 --- a/arm_compute/core/NEON/kernels/NESoftmaxLayerKernel.h +++ b/arm_compute/core/NEON/kernels/NESoftmaxLayerKernel.h @@ -43,6 +43,14 @@ public: * @param[out] output Destination tensor. Data types supported: same as @p input */ void configure(const ITensor *input, ITensor *output); + /** Static function to check if given info will lead to a valid configuration of @ref NELogits1DMaxKernel + * + * @param[in] input Source tensor. Data types supported: QS8/QS16/F16/F32 + * @param[in] output Destination tensor. Data types supported: same as @p input + * + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *output); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; @@ -81,6 +89,17 @@ public: * @param[in] beta (Optional) A scaling factor for the exponent. QS8/QS16 only support a beta value of 1. */ void configure(const ITensor *input, const ITensor *max, ITensor *output, ITensor *sum, float beta = 1.0f); + /** Static function to check if given info will lead to a valid configuration of @ref NELogits1DShiftExpSumKernel + * + * @param[in] input Source tensor. Data types supported: QS8/QS16/F16/F32 + * @param[in] max Max values tensor. Data types supported: same as @p input + * @param[in] output Destination tensor. Data types supported: same as @p input. + * @param[in] sum Sum of 1D logits tensor. Data types supported: same as @p input. + * @param[in] beta (Optional) A scaling factor for the exponent. QS8/QS16 only support a beta value of 1. + * + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *max, const ITensorInfo *output, const ITensorInfo *sum, float beta = 1.0f); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; @@ -120,6 +139,15 @@ public: * @param[out] output Destination tensor. Data types supported: same as @p input. */ void configure(const ITensor *input, const ITensor *sum, ITensor *output); + /** Static function to check if given info will lead to a valid configuration of @ref NELogits1DNormKernel + * + * @param[in] input Source tensor. Data types supported: QS8/QS16/S32/F16/F32 + * @param[in] sum Sum tensor. The number of dimensions should be dim(input)-1. Data types supported: same as @p input. + * @param[in] output Destination tensor. Data types supported: same as @p input. + * + * @return a status + */ + static Status validate(const ITensorInfo *input, const ITensorInfo *sum, const ITensorInfo *output); // Inherited methods overridden: void run(const Window &window, const ThreadInfo &info) override; |