From 633d30be83a7cc76cf7221d7004d768dc4381742 Mon Sep 17 00:00:00 2001 From: Michele Di Giorgio Date: Tue, 8 Oct 2019 17:17:18 +0100 Subject: COMPMID-2306: CLDepthwiseConvolution: support for QUANT8_PER_CHANNEL_SYMM - Reference This patch modifies the reference implementation and the fixtures of depthwise convolution layer to support QSYMM8_PER_CHANNEL quantization. Change-Id: I28adb5c110308b1024a213bec2d35a89180a46dc Signed-off-by: Michele Di Giorgio Reviewed-on: https://review.mlplatform.org/c/2063 Reviewed-by: Giorgio Arena Comments-Addressed: Arm Jenkins Reviewed-by: Giuseppe Rossini Tested-by: Arm Jenkins --- .../validation/NEON/DepthwiseConvolutionLayer.cpp | 2 +- .../fixtures/DepthwiseConvolutionLayerFixture.h | 82 ++++++++++++++-------- .../reference/DepthwiseConvolutionLayer.cpp | 82 ++++++++++++++++------ .../reference/DepthwiseConvolutionLayer.h | 4 +- 4 files changed, 116 insertions(+), 54 deletions(-) diff --git a/tests/validation/NEON/DepthwiseConvolutionLayer.cpp b/tests/validation/NEON/DepthwiseConvolutionLayer.cpp index c62c07bdfd..6392906037 100644 --- a/tests/validation/NEON/DepthwiseConvolutionLayer.cpp +++ b/tests/validation/NEON/DepthwiseConvolutionLayer.cpp @@ -53,7 +53,7 @@ RelativeTolerance tolerance_f16(half_float::half(0.01)); /**< constexpr float tolerance_num = 0.05f; /**< Tolerance number */ #endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC -const auto depth_multipliers = framework::dataset::make("DepthMultiplier", { 1, 2, 5 }); +const auto depth_multipliers = framework::dataset::make("DepthMultiplier", { 1, 2, 5 }); const auto large_depth_multipliers = framework::dataset::make("DepthMultiplier", { 1, 2, 5, 8 }); //Activation Functions diff --git a/tests/validation/fixtures/DepthwiseConvolutionLayerFixture.h b/tests/validation/fixtures/DepthwiseConvolutionLayerFixture.h index 2c9b31866b..85930eb95e 100644 --- a/tests/validation/fixtures/DepthwiseConvolutionLayerFixture.h +++ b/tests/validation/fixtures/DepthwiseConvolutionLayerFixture.h @@ -48,32 +48,34 @@ namespace validation { using namespace arm_compute::misc::shape_calculator; -template +template class DepthwiseConvolutionLayerValidationGenericFixture : public framework::Fixture { public: - using TBias = typename std::conditional::type, uint8_t>::value, int32_t, T>::type; + using TBias = typename std::conditional < std::is_same::value || std::is_same::value, int32_t, T >::type; public: template - void setup(TensorShape in_shape, Size2D kernel_size, PadStrideInfo pad_stride_info, Size2D dilation, unsigned int depth_multiplier, DataType data_type, - QuantizationInfo input_quantization_info, QuantizationInfo output_quantization_info, DataLayout data_layout, ActivationLayerInfo act_info) + void setup(TensorShape in_shape, Size2D kernel_size, PadStrideInfo pad_stride_info, Size2D dilation, + unsigned int depth_multiplier, DataType input_data_type, DataType weights_data_type, + QuantizationInfo input_quantization_info, QuantizationInfo weights_quantization_info, QuantizationInfo output_quantization_info, + DataLayout data_layout, ActivationLayerInfo act_info) { - const DataType bias_data_type = is_data_type_quantized_asymmetric(data_type) ? DataType::S32 : data_type; + const DataType bias_data_type = is_data_type_quantized(input_data_type) ? DataType::S32 : input_data_type; TensorShape weights_shape(kernel_size.width, kernel_size.height); - const TensorInfo in_info(in_shape, 1, data_type); - const TensorInfo we_info(weights_shape, 1, data_type); + const TensorInfo in_info(in_shape, 1, input_data_type); + const TensorInfo we_info(weights_shape, 1, weights_data_type); const TensorShape out_shape = compute_depthwise_convolution_shape(in_info, we_info, pad_stride_info, depth_multiplier, dilation); weights_shape.set(2, out_shape.z()); const TensorShape biases_shape(weights_shape[2]); _target = compute_target(in_shape, weights_shape, biases_shape, out_shape, pad_stride_info, dilation, depth_multiplier, - data_type, bias_data_type, input_quantization_info, output_quantization_info, data_layout, act_info); + input_data_type, weights_data_type, bias_data_type, input_quantization_info, weights_quantization_info, output_quantization_info, data_layout, act_info); _reference = compute_reference(in_shape, weights_shape, biases_shape, out_shape, pad_stride_info, dilation, depth_multiplier, - data_type, bias_data_type, input_quantization_info, output_quantization_info, act_info); + input_data_type, weights_data_type, bias_data_type, input_quantization_info, weights_quantization_info, output_quantization_info, act_info); } protected: @@ -88,6 +90,12 @@ protected: library->fill(tensor, distribution, i); break; } + case DataType::QSYMM8_PER_CHANNEL: + { + std::uniform_int_distribution distribution(-10, 10); + library->fill(tensor, distribution, i); + break; + } case DataType::F32: case DataType::F16: { @@ -107,9 +115,8 @@ protected: } TensorType compute_target(TensorShape input_shape, TensorShape weights_shape, TensorShape biases_shape, TensorShape output_shape, PadStrideInfo &pad_stride_info, Size2D dilation, - unsigned int depth_multiplier, - const DataType data_type, const DataType bias_data_type, - const QuantizationInfo &input_quantization_info, const QuantizationInfo &output_quantization_info, + unsigned int depth_multiplier, const DataType input_data_type, const DataType weights_data_type, const DataType bias_data_type, + const QuantizationInfo &input_quantization_info, const QuantizationInfo &weights_quantization_info, const QuantizationInfo &output_quantization_info, const DataLayout data_layout, const ActivationLayerInfo &act_info) { if(data_layout == DataLayout::NHWC) @@ -120,10 +127,10 @@ protected: } // Create tensors - TensorType src = create_tensor(input_shape, data_type, 1, input_quantization_info, data_layout); - TensorType weights = create_tensor(weights_shape, data_type, 1, input_quantization_info, data_layout); + TensorType src = create_tensor(input_shape, input_data_type, 1, input_quantization_info, data_layout); + TensorType weights = create_tensor(weights_shape, weights_data_type, 1, weights_quantization_info, data_layout); TensorType biases = create_tensor(biases_shape, bias_data_type, 1, input_quantization_info, data_layout); - TensorType dst = create_tensor(output_shape, data_type, 1, output_quantization_info, data_layout); + TensorType dst = create_tensor(output_shape, input_data_type, 1, output_quantization_info, data_layout); // Create Depthwise Convolution configure function FunctionType dwc; @@ -157,14 +164,13 @@ protected: } SimpleTensor compute_reference(const TensorShape &in_shape, const TensorShape &weights_shape, const TensorShape &biases_shape, const TensorShape &out_shape, - const PadStrideInfo &pad_stride_info, - const Size2D &dilation, unsigned int depth_multiplier, - const DataType data_type, const DataType bias_data_type, - const QuantizationInfo &input_quantization_info, const QuantizationInfo &output_quantization_info, + const PadStrideInfo &pad_stride_info, const Size2D &dilation, unsigned int depth_multiplier, + const DataType input_data_type, const DataType weights_data_type, const DataType bias_data_type, + const QuantizationInfo &input_quantization_info, const QuantizationInfo &weights_quantization_info, const QuantizationInfo &output_quantization_info, const ActivationLayerInfo &act_info) { - SimpleTensor src{ in_shape, data_type, 1, input_quantization_info }; - SimpleTensor weights{ weights_shape, data_type, 1, input_quantization_info }; + SimpleTensor src{ in_shape, input_data_type, 1, input_quantization_info }; + SimpleTensor weights{ weights_shape, weights_data_type, 1, weights_quantization_info }; SimpleTensor biases{ biases_shape, bias_data_type, 1, input_quantization_info }; fill(src, 0); @@ -180,20 +186,21 @@ protected: }; template -class DepthwiseConvolutionLayerValidationFixture : public DepthwiseConvolutionLayerValidationGenericFixture +class DepthwiseConvolutionLayerValidationFixture : public DepthwiseConvolutionLayerValidationGenericFixture { public: template void setup(TensorShape in_shape, Size2D kernel_size, PadStrideInfo pad_stride_info, Size2D dilation, unsigned int depth_multiplier, DataType data_type, DataLayout data_layout, ActivationLayerInfo act_info) { - DepthwiseConvolutionLayerValidationGenericFixture::setup(in_shape, kernel_size, pad_stride_info, dilation, depth_multiplier, - data_type, QuantizationInfo(), QuantizationInfo(), data_layout, act_info); + DepthwiseConvolutionLayerValidationGenericFixture::setup(in_shape, kernel_size, pad_stride_info, dilation, depth_multiplier, + data_type, data_type, QuantizationInfo(), QuantizationInfo(), QuantizationInfo(), + data_layout, act_info); } }; template -class DepthwiseConvolutionLayerNativeValidationFixture : public DepthwiseConvolutionLayerValidationGenericFixture +class DepthwiseConvolutionLayerNativeValidationFixture : public DepthwiseConvolutionLayerValidationGenericFixture { public: template @@ -302,7 +309,7 @@ protected: }; template -class DepthwiseConvolutionLayerNativeConfigurableValidationFixture : public DepthwiseConvolutionLayerValidationGenericFixture +class DepthwiseConvolutionLayerNativeConfigurableValidationFixture : public DepthwiseConvolutionLayerValidationGenericFixture { public: template @@ -423,15 +430,32 @@ protected: }; template -class DepthwiseConvolutionLayerValidationQuantizedFixture : public DepthwiseConvolutionLayerValidationGenericFixture +class DepthwiseConvolutionLayerValidationQuantizedFixture : public DepthwiseConvolutionLayerValidationGenericFixture { public: template void setup(TensorShape in_shape, Size2D kernel_size, PadStrideInfo pad_stride_info, Size2D dilation, unsigned int depth_multiplier, DataType data_type, QuantizationInfo input_quantization_info, QuantizationInfo output_quantization_info, DataLayout data_layout, ActivationLayerInfo act_info) { - DepthwiseConvolutionLayerValidationGenericFixture::setup(in_shape, kernel_size, pad_stride_info, dilation, depth_multiplier, - data_type, input_quantization_info, output_quantization_info, data_layout, act_info); + DepthwiseConvolutionLayerValidationGenericFixture::setup(in_shape, kernel_size, pad_stride_info, dilation, depth_multiplier, data_type, + data_type, input_quantization_info, input_quantization_info, output_quantization_info, + data_layout, act_info); + } +}; + +template +class DepthwiseConvolutionLayerValidationQuantizedPerChannelFixture : public DepthwiseConvolutionLayerValidationGenericFixture +{ +public: + template + void setup(TensorShape in_shape, Size2D kernel_size, PadStrideInfo pad_stride_info, Size2D dilation, unsigned int depth_multiplier, DataType input_data_type, DataType weights_data_type, + QuantizationInfo input_quantization_info, QuantizationInfo weights_quantization_info, QuantizationInfo output_quantization_info, + DataLayout data_layout, ActivationLayerInfo act_info) + { + DepthwiseConvolutionLayerValidationGenericFixture::setup(in_shape, kernel_size, pad_stride_info, dilation, depth_multiplier, + input_data_type, weights_data_type, + input_quantization_info, weights_quantization_info, output_quantization_info, + data_layout, act_info); } }; } // namespace validation diff --git a/tests/validation/reference/DepthwiseConvolutionLayer.cpp b/tests/validation/reference/DepthwiseConvolutionLayer.cpp index b1d2b923f7..7458f815b8 100644 --- a/tests/validation/reference/DepthwiseConvolutionLayer.cpp +++ b/tests/validation/reference/DepthwiseConvolutionLayer.cpp @@ -40,7 +40,9 @@ namespace validation { namespace reference { -/** Perform a depthwise convolution +namespace +{ +/** Perform a depthwise convolution for floating-point types * * - Three dimensions tensors * - Third dimention is number of channels @@ -48,9 +50,9 @@ namespace reference * - Padding, stride and output shape "match" * */ -template -SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, const PadStrideInfo &conv_info, - unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) +template +SimpleTensor depthwise_convolution_fp(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, const PadStrideInfo &conv_info, + unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) { ARM_COMPUTE_UNUSED(out_quant_info); @@ -114,7 +116,7 @@ SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTe } } - dst[out_pos++] = saturate_cast(val + *static_cast(biases(Coordinates(out_z)))); + dst[out_pos++] = saturate_cast(val + *static_cast(biases(Coordinates(out_z)))); } } } @@ -124,26 +126,32 @@ SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTe return dst; } -template <> -SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, - const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) +/** Perform a quantized depthwise convolution + * + * - Three dimensions tensors + * - Third dimention is number of channels + * - Depths of input tensor and filter are equals + * - Padding, stride and output shape "match" + * - QASYMM8 input, output + * - QASYMM8 or QSYMM8_PER_CHANNEL filter + * + */ +template +SimpleTensor depthwise_convolution_quantized(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, + const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) { // if no explicit quantization has been set you the same as src const QuantizationInfo &dst_qinfo = out_quant_info.uniform().empty() ? src.quantization_info() : out_quant_info; - SimpleTensor dst{ dst_shape, src.data_type(), 1, dst_qinfo }; + SimpleTensor dst{ dst_shape, src.data_type(), 1, dst_qinfo }; // Create reference const int input_offset = -src.quantization_info().uniform().offset; const float input_scale = src.quantization_info().uniform().scale; const int weights_offset = -weights.quantization_info().uniform().offset; - const float weights_scale = weights.quantization_info().uniform().scale; const int output_offset = dst_qinfo.uniform().offset; const float output_scale = dst_qinfo.uniform().scale; - int output_multiplier = 0; - int output_shift = 0; - const float multiplier = input_scale * weights_scale / output_scale; - arm_compute::quantization::calculate_quantized_multiplier_less_than_one(multiplier, &output_multiplier, &output_shift); + const std::vector weights_scale_vec = weights.quantization_info().scale(); // Compute reference const int filter_width = weights.shape().x(); @@ -173,11 +181,19 @@ SimpleTensor depthwise_convolution(const SimpleTensor &src, co const int maximum_x = input_width + pad_left + pad_right - static_cast(patch_width); const int maximum_y = input_height + pad_top + pad_bottom - static_cast(patch_height); + const bool is_quantized_per_channel = is_data_type_quantized_per_channel(weights.data_type()); + int out_pos = 0; for(int r = 0; r < num_batches; ++r) { for(int z = 0; z < input_depth; ++z) { + int output_multiplier = 0; + int output_shift = 0; + const float weights_scale = (is_quantized_per_channel) ? weights_scale_vec[z] : weights_scale_vec[0]; + const float multiplier = input_scale * weights_scale / output_scale; + arm_compute::quantization::calculate_quantized_multiplier_less_than_one(multiplier, &output_multiplier, &output_shift); + for(unsigned int m = 0; m < depth_multiplier; ++m) { const int out_z = z * depth_multiplier + m; @@ -197,8 +213,8 @@ SimpleTensor depthwise_convolution(const SimpleTensor &src, co { coords.set(0, i); coords.set(1, j); - const auto in_val = tensor_elem_at(src, coords, BorderMode::CONSTANT, -input_offset); - const uint8_t w_val = *(weights.data() + filter_offset); + const auto in_val = tensor_elem_at(src, coords, BorderMode::CONSTANT, -input_offset); + const TW w_val = *(weights.data() + filter_offset); val += (in_val + input_offset) * (w_val + weights_offset); ++filter_offset; } @@ -206,8 +222,7 @@ SimpleTensor depthwise_convolution(const SimpleTensor &src, co val += bias_val; val = asymm_rounding_divide_by_pow2(asymm_int_mult(val, output_multiplier), output_shift); val += output_offset; - val = std::max(val, 0); - val = std::min(val, 255); + val = utility::clamp(val, 0, 255); // Store the result dst[out_pos++] = val; @@ -219,12 +234,35 @@ SimpleTensor depthwise_convolution(const SimpleTensor &src, co return dst; } +} // namespace + +template <> +SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, + const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) +{ + return depthwise_convolution_fp(src, weights, biases, dst_shape, conv_info, depth_multiplier, dilation, out_quant_info); +} -template SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, - const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info); +template <> +SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, + const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) +{ + return depthwise_convolution_fp(src, weights, biases, dst_shape, conv_info, depth_multiplier, dilation, out_quant_info); +} -template SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, - const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info); +template <> +SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, + const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) +{ + return depthwise_convolution_quantized(src, weights, biases, dst_shape, conv_info, depth_multiplier, dilation, out_quant_info); +} + +template <> +SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, + const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation, const QuantizationInfo &out_quant_info) +{ + return depthwise_convolution_quantized(src, weights, biases, dst_shape, conv_info, depth_multiplier, dilation, out_quant_info); +} } // namespace reference } // namespace validation } // namespace test diff --git a/tests/validation/reference/DepthwiseConvolutionLayer.h b/tests/validation/reference/DepthwiseConvolutionLayer.h index ee323fa8df..38a225a1ae 100644 --- a/tests/validation/reference/DepthwiseConvolutionLayer.h +++ b/tests/validation/reference/DepthwiseConvolutionLayer.h @@ -35,8 +35,8 @@ namespace validation { namespace reference { -template -SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, const PadStrideInfo &conv_info, +template +SimpleTensor depthwise_convolution(const SimpleTensor &src, const SimpleTensor &weights, const SimpleTensor &biases, const TensorShape &dst_shape, const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation = Size2D(1U, 1U), const QuantizationInfo &out_quant_info = QuantizationInfo(0.0f, 0)); } // namespace reference } // namespace validation -- cgit v1.2.1