diff options
Diffstat (limited to 'tests/validation/fixtures/DirectConvolutionLayerFixture.h')
| -rw-r--r-- | tests/validation/fixtures/DirectConvolutionLayerFixture.h | 150 |
1 files changed, 115 insertions, 35 deletions
diff --git a/tests/validation/fixtures/DirectConvolutionLayerFixture.h b/tests/validation/fixtures/DirectConvolutionLayerFixture.h index 614aa20753..6f204642ca 100644 --- a/tests/validation/fixtures/DirectConvolutionLayerFixture.h +++ b/tests/validation/fixtures/DirectConvolutionLayerFixture.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2017-2021 Arm Limited. + * Copyright (c) 2017-2023 Arm Limited. * * SPDX-License-Identifier: MIT * @@ -21,6 +21,10 @@ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE * SOFTWARE. */ + +#ifndef ACL_TESTS_VALIDATION_FIXTURES_DIRECTCONVOLUTIONLAYERFIXTURE_H +#define ACL_TESTS_VALIDATION_FIXTURES_DIRECTCONVOLUTIONLAYERFIXTURE_H + #include "arm_compute/core/Helpers.h" #include "arm_compute/core/TensorShape.h" #include "arm_compute/core/Types.h" @@ -51,11 +55,52 @@ class DirectConvolutionValidationGenericFixture : public framework::Fixture public: using TBias = typename std::conditional < std::is_same<T, uint8_t>::value || std::is_same<T, int8_t>::value, int32_t, T >::type; - template <typename...> + void setup_quantization(const TensorShape &input_shape, const TensorShape &weights_shape, QuantizationInfo &input_q_info, + QuantizationInfo &weights_q_info, DataType data_type) + { + const int32_t t_max = static_cast<int32_t>(std::numeric_limits<T>::max()); + const int32_t t_min = static_cast<int32_t>(std::numeric_limits<T>::min()); + + std::mt19937 generator(library->seed() + _hash); + std::uniform_real_distribution<float> distribution_float(-5.0f, 3.0f); + std::uniform_int_distribution<int32_t> distribution_t(t_min, t_max); + + const float scale_lhs = pow(2, distribution_float(generator)); // [2^-5, 2^3] + const float scale_rhs = pow(2, distribution_float(generator)); // [2^-5, 2^3] + + const int32_t offset_lhs = distribution_t(generator); + const int32_t offset_rhs = distribution_t(generator); + + input_q_info = QuantizationInfo(scale_lhs, offset_lhs); + weights_q_info = QuantizationInfo(scale_rhs, offset_rhs); + + QuantizationHint q_hint = suggest_conv_dst_q_info_and_bias(input_q_info, weights_q_info, + weights_shape.y() /* heights */, weights_shape.x() /* width */, input_shape.z() /* channels */, + data_type, 0.5f /* bias_fraction */); + + _dst_q_info = q_hint.q_info; + _min_bias = q_hint.bias_min; + _max_bias = q_hint.bias_max; + + // Do not change here as these limits are the natural limits of the associated data types and + // are embeded in the computation of the dst quantization info. + _min_u8 = 0; + _max_u8 = 255; + _min_s8 = -128; + _max_s8 = 127; + } + void setup(TensorShape input_shape, int stride_x, int stride_y, int pad_x, int pad_y, unsigned int kernel_size, unsigned int num_kernels, DataType data_type, QuantizationInfo quantization_info, ActivationLayerInfo act_info, DataLayout data_layout, bool mixed_layout = false) { - _quantization_info = quantization_info; + // This hash is used by random generators. There may be hash collisions but + // this is intentional as it's a very easy way to make the the current + // random generation process almost different for many test configurations, + // which were using the same set of values before. + _hash = input_shape[0] + input_shape[1] + input_shape[2] + input_shape[3] + + stride_x + stride_y + pad_x + pad_y + kernel_size + num_kernels + mixed_layout + + (data_layout == DataLayout::NHWC); + _data_type = data_type; _mixed_layout = mixed_layout; @@ -69,24 +114,48 @@ public: const TensorShape output_shape = compute_deep_convolution_shape(input_info, weights_info, info); - _target = compute_target(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, quantization_info, act_info, data_layout); - _reference = compute_reference(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, quantization_info, act_info); + QuantizationInfo input_q_info = quantization_info; + QuantizationInfo weights_q_info = quantization_info; + _dst_q_info = quantization_info; + + if(is_data_type_quantized(data_type) && (!act_info.enabled() || act_info.activation() == ActivationFunction::IDENTITY)) + { + setup_quantization(input_shape, weights_shape, input_q_info, weights_q_info, data_type); + } + + _target = compute_target(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, input_q_info, weights_q_info, act_info, data_layout); + _reference = compute_reference(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, input_q_info, weights_q_info, act_info); } - template <typename...> void setup(TensorShape input_shape, TensorShape weights_shape, TensorShape bias_shape, TensorShape output_shape, PadStrideInfo info, Size2D dilation, DataType data_type, QuantizationInfo quantization_info, ActivationLayerInfo act_info, DataLayout data_layout) { ARM_COMPUTE_ERROR_ON(data_layout == DataLayout::UNKNOWN); ARM_COMPUTE_UNUSED(dilation); - _quantization_info = quantization_info; + // This hash is used by random generators. There may be hash collisions but + // this is intentional as it's a very easy way to make the the current + // random generation process almost different for many test configurations, + // which were using the same set of values before. + _hash = input_shape[0] + input_shape[1] + input_shape[2] + input_shape[3] + + weights_shape[0] + weights_shape[1] + weights_shape[2] + weights_shape[3] + dilation.x() + + dilation.y() + info.pad_bottom() + info.pad_left() + info.pad_right() + info.pad_top(); + _data_type = data_type; const DataType bias_data_type = is_data_type_quantized_asymmetric(data_type) ? DataType::S32 : data_type; - _target = compute_target(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, quantization_info, act_info, data_layout); - _reference = compute_reference(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, quantization_info, act_info); + QuantizationInfo input_q_info = quantization_info; + QuantizationInfo weights_q_info = quantization_info; + _dst_q_info = quantization_info; + + if(is_data_type_quantized(data_type) && (!act_info.enabled() || act_info.activation() == ActivationFunction::IDENTITY)) + { + setup_quantization(input_shape, weights_shape, input_q_info, weights_q_info, data_type); + } + + _target = compute_target(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, input_q_info, weights_q_info, act_info, data_layout); + _reference = compute_reference(input_shape, weights_shape, bias_shape, output_shape, info, data_type, bias_data_type, input_q_info, weights_q_info, act_info); } protected: @@ -112,14 +181,14 @@ protected: { case DataType::QASYMM8: { - std::uniform_int_distribution<uint8_t> distribution(0, 50); + std::uniform_int_distribution<uint32_t> distribution(_min_u8, _max_u8); library->fill(tensor, distribution, i); break; } case DataType::QASYMM8_SIGNED: { // Use small input range to avoid all the test results being saturated at the end. - std::uniform_int_distribution<int8_t> distribution(-25, 25); + std::uniform_int_distribution<int32_t> distribution(_min_s8, _max_s8); library->fill(tensor, distribution, i); break; } @@ -137,7 +206,7 @@ protected: } case DataType::S32: { - std::uniform_int_distribution<int32_t> distribution(-5, 5); + std::uniform_int_distribution<int32_t> distribution(_min_bias, _max_bias); library->fill(tensor, distribution, i); break; } @@ -147,7 +216,7 @@ protected: } TensorType compute_target(TensorShape input_shape, TensorShape weights_shape, const TensorShape &bias_shape, TensorShape output_shape, const PadStrideInfo &info, - DataType data_type, DataType bias_data_type, QuantizationInfo quantization_info, ActivationLayerInfo act_info, const DataLayout &data_layout) + DataType data_type, DataType bias_data_type, QuantizationInfo input_q_info, QuantizationInfo weights_q_info, ActivationLayerInfo act_info, const DataLayout &data_layout) { if(data_layout == DataLayout::NHWC) { @@ -157,10 +226,10 @@ protected: } // Create tensors - TensorType src = create_tensor<TensorType>(input_shape, data_type, 1, quantization_info, data_layout); - TensorType weights = create_tensor<TensorType>(weights_shape, data_type, 1, quantization_info, data_layout); - TensorType bias = create_tensor<TensorType>(bias_shape, bias_data_type, 1, quantization_info); - TensorType dst = create_tensor<TensorType>(output_shape, data_type, 1, quantization_info, data_layout); + TensorType src = create_tensor<TensorType>(input_shape, data_type, 1, input_q_info, data_layout); + TensorType weights = create_tensor<TensorType>(weights_shape, data_type, 1, weights_q_info, data_layout); + TensorType bias = create_tensor<TensorType>(bias_shape, bias_data_type, 1, QuantizationInfo()); + TensorType dst = create_tensor<TensorType>(output_shape, data_type, 1, _dst_q_info, data_layout); add_padding_x({ &src, &bias, &dst }, data_layout); add_padding_x({ &weights }, data_layout, input_shape[0] % 4 == 0); // Don't add left padding if cl image will be used @@ -186,9 +255,9 @@ protected: ARM_COMPUTE_ASSERT(!dst.info()->is_resizable()); // Fill tensors - fill(AccessorType(src), 0); - fill(AccessorType(weights), 1); - fill(AccessorType(bias), 2); + fill(AccessorType(src), 0 + _hash); + fill(AccessorType(weights), 1 + _hash); + fill(AccessorType(bias), 2 + _hash); if(_mixed_layout) { @@ -204,33 +273,45 @@ protected: } SimpleTensor<T> compute_reference(const TensorShape &input_shape, const TensorShape &weights_shape, const TensorShape &bias_shape, const TensorShape &output_shape, const PadStrideInfo &info, - DataType data_type, DataType bias_data_type, QuantizationInfo quantization_info, ActivationLayerInfo act_info) + DataType data_type, DataType bias_data_type, QuantizationInfo input_q_info, QuantizationInfo weights_q_info, ActivationLayerInfo act_info) { // Create reference - SimpleTensor<T> src{ input_shape, data_type, 1, quantization_info }; - SimpleTensor<T> weights{ weights_shape, data_type, 1, quantization_info }; - SimpleTensor<TBias> bias{ bias_shape, bias_data_type, 1, quantization_info }; + SimpleTensor<T> src{ input_shape, data_type, 1, input_q_info }; + SimpleTensor<T> weights{ weights_shape, data_type, 1, weights_q_info }; + SimpleTensor<TBias> bias{ bias_shape, bias_data_type, 1, QuantizationInfo() }; // Fill reference - fill(src, 0); - fill(weights, 1); - fill(bias, 2); - - SimpleTensor<T> dst = reference::convolution_layer<T>(src, weights, bias, output_shape, info); - return (act_info.enabled()) ? reference::activation_layer<T>(dst, act_info) : dst; + fill(src, 0 + _hash); + fill(weights, 1 + _hash); + fill(bias, 2 + _hash); + + SimpleTensor<T> dst = reference::convolution_layer<T>(src, weights, bias, output_shape, info, + Size2D(1U, 1U) /* dilation */, 1 /* num_groups */, _dst_q_info); + SimpleTensor<T> dst2 = (act_info.enabled()) ? reference::activation_layer<T>(dst, act_info) : dst; + return dst2; } TensorType _target{}; SimpleTensor<T> _reference{}; - QuantizationInfo _quantization_info{}; + QuantizationInfo _dst_q_info{}; DataType _data_type{}; bool _mixed_layout{ false }; + int32_t _hash{0}; + + // Random initialization limits + // Default values are previously handcrafted limits + // that sould be used when we don't use dynamic quantization + int32_t _min_bias{-5}; + int32_t _max_bias{5}; + int32_t _min_u8{0}; + int32_t _max_u8{50}; + int32_t _min_s8{-25}; + int32_t _max_s8{25}; }; template <typename TensorType, typename AccessorType, typename FunctionType, typename T, bool mixed_layout = false> class DirectConvolutionValidationFixture : public DirectConvolutionValidationGenericFixture<TensorType, AccessorType, FunctionType, T> { public: - template <typename...> void setup(TensorShape input_shape, int stride_x, int stride_y, int pad_x, int pad_y, unsigned int kernel_size, unsigned int num_kernels, DataType data_type, ActivationLayerInfo act_info, DataLayout data_layout) { @@ -243,7 +324,6 @@ template <typename TensorType, typename AccessorType, typename FunctionType, typ class DirectConvolutionValidationQuantizedFixture : public DirectConvolutionValidationGenericFixture<TensorType, AccessorType, FunctionType, T> { public: - template <typename...> void setup(TensorShape input_shape, int stride_x, int stride_y, int pad_x, int pad_y, unsigned int kernel_size, unsigned int num_kernels, DataType data_type, QuantizationInfo quantization_info, ActivationLayerInfo act_info, DataLayout data_layout) { @@ -256,7 +336,6 @@ template <typename TensorType, typename AccessorType, typename FunctionType, typ class DirectConvolutionValidationWithTensorShapesQuantizedFixture : public DirectConvolutionValidationGenericFixture<TensorType, AccessorType, FunctionType, T> { public: - template <typename...> void setup(TensorShape input_shape, TensorShape weights_shape, TensorShape bias_shape, TensorShape output_shape, PadStrideInfo info, Size2D dilation, DataType data_type, QuantizationInfo quantization_info, ActivationLayerInfo act_info, DataLayout data_layout) { @@ -269,7 +348,6 @@ template <typename TensorType, typename AccessorType, typename FunctionType, typ class DirectConvolutionValidationWithTensorShapesFixture : public DirectConvolutionValidationGenericFixture<TensorType, AccessorType, FunctionType, T> { public: - template <typename...> void setup(TensorShape input_shape, TensorShape weights_shape, TensorShape bias_shape, TensorShape output_shape, PadStrideInfo info, Size2D dilation, DataType data_type, ActivationLayerInfo act_info) { @@ -281,3 +359,5 @@ public: } // namespace validation } // namespace test } // namespace arm_compute + +#endif // ACL_TESTS_VALIDATION_FIXTURES_DIRECTCONVOLUTIONLAYERFIXTURE_H |