// Copyright (c) Facebook, Inc. and its affiliates. // All rights reserved. // // Copyright 2019 Google LLC // // This source code is licensed under the BSD-style license found in the // LICENSE file in the root directory of this source tree. #pragma once #include #include #include #include #include #include #include #include #include #include #include #include class FullyConnectedOperatorTester { public: enum class WeightsType { Default, FP32, }; inline FullyConnectedOperatorTester& input_channels(size_t input_channels) { assert(input_channels >= 1); this->input_channels_ = input_channels; return *this; } inline size_t input_channels() const { return this->input_channels_; } inline FullyConnectedOperatorTester& output_channels(size_t output_channels) { assert(output_channels >= 1); this->output_channels_ = output_channels; return *this; } inline size_t output_channels() const { return this->output_channels_; } inline FullyConnectedOperatorTester& batch_size(size_t batch_size) { assert(batch_size >= 1); this->batch_size_ = batch_size; return *this; } inline size_t batch_size() const { return this->batch_size_; } inline FullyConnectedOperatorTester& input_stride(size_t input_stride) { assert(input_stride >= 1); this->input_stride_ = input_stride; return *this; } inline size_t input_stride() const { if (this->input_stride_ == 0) { return input_channels(); } else { assert(this->input_stride_ >= input_channels()); return this->input_stride_; } } inline FullyConnectedOperatorTester& output_stride(size_t output_stride) { assert(output_stride >= 1); this->output_stride_ = output_stride; return *this; } inline size_t output_stride() const { if (this->output_stride_ == 0) { return output_channels(); } else { assert(this->output_stride_ >= output_channels()); return this->output_stride_; } } inline FullyConnectedOperatorTester& qmin(uint8_t qmin) { this->qmin_ = qmin; return *this; } inline uint8_t qmin() const { return this->qmin_; } inline FullyConnectedOperatorTester& qmax(uint8_t qmax) { this->qmax_ = qmax; return *this; } inline uint8_t qmax() const { return this->qmax_; } inline FullyConnectedOperatorTester& transpose_weights(bool transpose_weights) { this->transpose_weights_ = transpose_weights; return *this; } inline bool transpose_weights() const { return this->transpose_weights_; } inline FullyConnectedOperatorTester& has_bias(bool has_bias) { this->has_bias_ = has_bias; return *this; } inline bool has_bias() const { return this->has_bias_; } inline FullyConnectedOperatorTester& weights_type(WeightsType weights_type) { this->weights_type_ = weights_type; return *this; } inline WeightsType weights_type() const { return this->weights_type_; } inline FullyConnectedOperatorTester& use_weights_cache(bool use_weights_cache) { this->use_weights_cache_ = use_weights_cache; return *this; } inline bool use_weights_cache() const { return this->use_weights_cache_; } inline FullyConnectedOperatorTester& iterations(size_t iterations) { this->iterations_ = iterations; return *this; } inline size_t iterations() const { return this->iterations_; } void TestQS8() const { ASSERT_EQ(weights_type(), WeightsType::Default); std::random_device random_device; auto rng = std::mt19937(random_device()); std::uniform_int_distribution i32dist(-10000, 10000); std::uniform_int_distribution i8dist( std::numeric_limits::min(), std::numeric_limits::max()); std::uniform_int_distribution w8dist( -std::numeric_limits::max(), std::numeric_limits::max()); std::vector input(XNN_EXTRA_BYTES / sizeof(int8_t) + (batch_size() - 1) * input_stride() + input_channels()); std::vector kernel(output_channels() * input_channels()); std::vector bias(output_channels()); std::vector output((batch_size() - 1) * output_stride() + output_channels()); std::vector accumulators(batch_size() * output_channels()); std::vector output_ref(batch_size() * output_channels()); const int8_t input_zero_point = 127; for (size_t iteration = 0; iteration < iterations(); iteration++) { std::generate(input.begin(), input.end(), [&]() { return i8dist(rng); }); std::generate(kernel.begin(), kernel.end(), [&]() { return w8dist(rng); }); std::generate(bias.begin(), bias.end(), [&]() { return i32dist(rng); }); std::fill(output.begin(), output.end(), INT8_C(0xA5)); // Compute reference results, without renormalization. if (has_bias()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { accumulators[i * output_channels() + oc] = bias[oc]; } } } else { std::fill(accumulators.begin(), accumulators.end(), 0); } if (transpose_weights()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { accumulators[i * output_channels() + oc] += (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * int32_t(kernel[ic * output_channels() + oc]); } } } } else { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { accumulators[i * output_channels() + oc] += (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * int32_t(kernel[oc * input_channels() + ic]); } } } } // Compute renormalization parameters. const int32_t accumulated_min = *std::min_element(accumulators.cbegin(), accumulators.cend()); const int32_t accumulated_max = *std::max_element(accumulators.cbegin(), accumulators.cend()); const double output_scale = double(uint32_t(accumulated_max - accumulated_min)) / 255.0; const int8_t output_zero_point = int8_t(std::max(std::min( lrint(-0.5 - 0.5 * double(accumulated_min + accumulated_max) / output_scale), long(std::numeric_limits::max())), long(std::numeric_limits::min()))); // Renormalize reference results. std::transform(accumulators.cbegin(), accumulators.cend(), output_ref.begin(), [this, output_scale, output_zero_point](int32_t x) -> double { return std::max(std::min(double(x) / output_scale, double(qmax() - 0x80) - output_zero_point), double(qmin() - 0x80) - output_zero_point); }); // Create, setup, run, and destroy Fully Connected operator. ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); xnn_operator_t fully_connected_op = nullptr; xnn_caches caches = { .code_cache = NULL, .weights_cache = NULL, }; xnn_weights_cache weights_cache; if (use_weights_cache()) { xnn_init_weights_cache(&weights_cache); caches.weights_cache = &weights_cache; } const xnn_status status = xnn_create_fully_connected_nc_qs8( input_channels(), output_channels(), input_stride(), output_stride(), input_zero_point, 1.0f /* input scale */, 1.0f /* kernel scale */, kernel.data(), has_bias() ? bias.data() : nullptr, output_zero_point, output_scale, int8_t(qmin() - 0x80), int8_t(qmax() - 0x80), transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, &caches, &fully_connected_op); if (status == xnn_status_unsupported_hardware) { GTEST_SKIP(); } ASSERT_EQ(xnn_status_success, status); ASSERT_NE(nullptr, fully_connected_op); if (use_weights_cache()) { ASSERT_EQ(xnn_status_success, xnn_finalize_weights_cache(&weights_cache, xnn_weights_cache_finalization_kind_soft)); } // Smart pointer to automatically delete fully_connected_op. std::unique_ptr auto_fully_connected_op(fully_connected_op, xnn_delete_operator); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_qs8( fully_connected_op, batch_size(), input.data(), output.data(), nullptr /* thread pool */)); ASSERT_EQ(xnn_status_success, xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); // Verify results. VerifyQS8(output, output_ref, double(output_zero_point)); if (use_weights_cache()) { // Create another operator with the same weights cache. xnn_operator_t fully_connected_op2 = nullptr; size_t old_weights_cache_size = weights_cache.cache.weights.size; ASSERT_EQ(xnn_status_success, xnn_create_fully_connected_nc_qs8( input_channels(), output_channels(), input_stride(), output_stride(), input_zero_point, 1.0f /* input scale */, 1.0f /* kernel scale */, kernel.data(), has_bias() ? bias.data() : nullptr, output_zero_point, output_scale, int8_t(qmin() - 0x80), int8_t(qmax() - 0x80), transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, &caches, &fully_connected_op2)); ASSERT_NE(nullptr, fully_connected_op2); // Smart pointer to automatically delete fully_connected_op. std::unique_ptr auto_fully_connected_op(fully_connected_op2, xnn_delete_operator); std::vector output2(output.size(), INT8_C(0xA5)); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_qs8( fully_connected_op2, batch_size(), input.data(), output2.data(), nullptr /* thread pool */)); ASSERT_EQ( xnn_status_success, xnn_run_operator(fully_connected_op2, nullptr /* thread pool */)); VerifyWeightsCache(weights_cache, old_weights_cache_size); xnn_release_weights_cache(&weights_cache); VerifyQS8(output, output_ref, double(output_zero_point)); } } } void VerifyQS8(const std::vector& output, const std::vector& output_ref, double output_zero_point) const { for (size_t i = 0; i < batch_size(); i++) { for (size_t c = 0; c < output_channels(); c++) { ASSERT_LE(int32_t(output[i * output_stride() + c]), int32_t(qmax() - 0x80)) << "batch index = " << i << ", channel = " << c; ASSERT_GE(int32_t(output[i * output_stride() + c]), int32_t(qmin() - 0x80)) << "batch index = " << i << ", channel = " << c; ASSERT_NEAR(output_ref[i * output_channels() + c], double(output[i * output_stride() + c]) - output_zero_point, 0.9) << "batch index = " << i << ", channel = " << c; } } } void TestQU8() const { ASSERT_EQ(weights_type(), WeightsType::Default); std::random_device random_device; auto rng = std::mt19937(random_device()); std::uniform_int_distribution i32dist(-10000, 10000); std::uniform_int_distribution u8dist( std::numeric_limits::min(), std::numeric_limits::max()); std::vector input(XNN_EXTRA_BYTES / sizeof(uint8_t) + (batch_size() - 1) * input_stride() + input_channels()); std::vector kernel(output_channels() * input_channels()); std::vector bias(output_channels()); std::vector output((batch_size() - 1) * output_stride() + output_channels()); std::vector accumulators(batch_size() * output_channels()); std::vector output_ref(batch_size() * output_channels()); const uint8_t input_zero_point = 127; const uint8_t kernel_zero_point = 127; for (size_t iteration = 0; iteration < iterations(); iteration++) { std::generate(input.begin(), input.end(), [&]() { return u8dist(rng); }); std::generate(kernel.begin(), kernel.end(), [&]() { return u8dist(rng); }); std::generate(bias.begin(), bias.end(), [&]() { return i32dist(rng); }); std::fill(output.begin(), output.end(), UINT8_C(0xA5)); // Compute reference results, without renormalization. if (has_bias()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { accumulators[i * output_channels() + oc] = bias[oc]; } } } else { std::fill(accumulators.begin(), accumulators.end(), 0); } if (transpose_weights()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { accumulators[i * output_channels() + oc] += (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * (int32_t(kernel[ic * output_channels() + oc]) - int32_t(kernel_zero_point)); } } } } else { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { accumulators[i * output_channels() + oc] += (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * (int32_t(kernel[oc * input_channels() + ic]) - int32_t(kernel_zero_point)); } } } } // Compute renormalization parameters. const int32_t accumulated_min = *std::min_element(accumulators.cbegin(), accumulators.cend()); const int32_t accumulated_max = *std::max_element(accumulators.cbegin(), accumulators.cend()); const double output_scale = double(uint32_t(accumulated_max - accumulated_min)) / 255.0; const uint8_t output_zero_point = uint8_t(std::max(std::min( lrint(127.5 - 0.5 * double(accumulated_min + accumulated_max) / output_scale), long(std::numeric_limits::max())), long(std::numeric_limits::min()))); // Renormalize reference results. std::transform(accumulators.cbegin(), accumulators.cend(), output_ref.begin(), [this, output_scale, output_zero_point](int32_t x) -> double { return std::max(std::min(double(x) / output_scale, double(qmax()) - output_zero_point), double(qmin()) - output_zero_point); }); // Create, setup, run, and destroy Fully Connected operator. ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); xnn_operator_t fully_connected_op = nullptr; xnn_caches caches = { .code_cache = NULL, .weights_cache = NULL, }; xnn_weights_cache weights_cache; if (use_weights_cache()) { xnn_init_weights_cache(&weights_cache); caches.weights_cache = &weights_cache; } const xnn_status status = xnn_create_fully_connected_nc_qu8( input_channels(), output_channels(), input_stride(), output_stride(), input_zero_point, 1.0f /* input scale */, kernel_zero_point, 1.0f /* kernel scale */, kernel.data(), has_bias() ? bias.data() : nullptr, output_zero_point, output_scale, qmin(), qmax(), transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, &caches, &fully_connected_op); if (status == xnn_status_unsupported_hardware) { GTEST_SKIP(); } ASSERT_EQ(xnn_status_success, status); ASSERT_NE(nullptr, fully_connected_op); if (use_weights_cache()) { ASSERT_EQ(xnn_status_success, xnn_finalize_weights_cache(&weights_cache, xnn_weights_cache_finalization_kind_soft)); } // Smart pointer to automatically delete fully_connected_op. std::unique_ptr auto_fully_connected_op(fully_connected_op, xnn_delete_operator); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_qu8( fully_connected_op, batch_size(), input.data(), output.data(), nullptr /* thread pool */)); ASSERT_EQ(xnn_status_success, xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); VerifyQU8(output, output_ref, double(output_zero_point)); if (use_weights_cache()) { // Create another operator with the same weights cache. xnn_operator_t fully_connected_op2 = nullptr; size_t old_weights_cache_size = weights_cache.cache.weights.size; ASSERT_EQ(xnn_status_success, xnn_create_fully_connected_nc_qu8( input_channels(), output_channels(), input_stride(), output_stride(), input_zero_point, 1.0f /* input scale */, kernel_zero_point, 1.0f /* kernel scale */, kernel.data(), has_bias() ? bias.data() : nullptr, output_zero_point, output_scale, qmin(), qmax(), transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, &caches, &fully_connected_op2)); ASSERT_NE(nullptr, fully_connected_op2); // Smart pointer to automatically delete fully_connected_op. std::unique_ptr auto_fully_connected_op(fully_connected_op2, xnn_delete_operator); std::vector output2(output.size(), UINT8_C(0xA5)); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_qu8( fully_connected_op2, batch_size(), input.data(), output2.data(), nullptr /* thread pool */)); ASSERT_EQ( xnn_status_success, xnn_run_operator(fully_connected_op2, nullptr /* thread pool */)); VerifyWeightsCache(weights_cache, old_weights_cache_size); xnn_release_weights_cache(&weights_cache); VerifyQU8(output2, output_ref, double(output_zero_point)); } } } void VerifyQU8(const std::vector& output, const std::vector& output_ref, double output_zero_point) const { for (size_t i = 0; i < batch_size(); i++) { for (size_t c = 0; c < output_channels(); c++) { ASSERT_LE(int32_t(output[i * output_stride() + c]), int32_t(qmax())) << "batch index = " << i << ", channel = " << c; ASSERT_GE(int32_t(output[i * output_stride() + c]), int32_t(qmin())) << "batch index = " << i << ", channel = " << c; ASSERT_NEAR(output_ref[i * output_channels() + c], double(output[i * output_stride() + c]) - output_zero_point, 0.9) << "batch index = " << i << ", channel = " << c; } } } void TestF32() const { ASSERT_EQ(weights_type(), WeightsType::Default); std::random_device random_device; auto rng = std::mt19937(random_device()); std::uniform_real_distribution f32dist(0.1f, 1.0f); std::vector input(XNN_EXTRA_BYTES / sizeof(float) + (batch_size() - 1) * input_stride() + input_channels()); std::vector kernel(output_channels() * input_channels()); std::vector bias(output_channels()); std::vector output((batch_size() - 1) * output_stride() + output_channels()); std::vector output_ref(batch_size() * output_channels()); for (size_t iteration = 0; iteration < iterations(); iteration++) { std::generate(input.begin(), input.end(), [&]() { return f32dist(rng); }); std::generate(kernel.begin(), kernel.end(), [&]() { return f32dist(rng); }); std::generate(bias.begin(), bias.end(), [&]() { return f32dist(rng); }); std::fill(output.begin(), output.end(), nanf("")); // Compute reference results, without renormalization. if (has_bias()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { output_ref[i * output_channels() + oc] = bias[oc]; } } } else { std::fill(output_ref.begin(), output_ref.end(), 0.0f); } if (transpose_weights()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { output_ref[i * output_channels() + oc] += input[i * input_stride() + ic] * kernel[ic * output_channels() + oc]; } } } } else { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { output_ref[i * output_channels() + oc] += input[i * input_stride() + ic] * kernel[oc * input_channels() + ic]; } } } } // Compute clamping parameters. const float accumulated_min = *std::min_element(output_ref.cbegin(), output_ref.cend()); const float accumulated_max = *std::max_element(output_ref.cbegin(), output_ref.cend()); const float output_min = qmin() == 0 ? -std::numeric_limits::infinity() : accumulated_min + (accumulated_max - accumulated_min) / 255.0f * float(qmin()); const float output_max = qmax() == 255 ? std::numeric_limits::infinity() : accumulated_max - (accumulated_max - accumulated_min) / 255.0f * float(255 - qmax()); // Clamp reference results. for (float& value : output_ref) { value = std::max(std::min(value, output_max), output_min); } // Create, setup, run, and destroy Fully Connected operator. ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); xnn_operator_t fully_connected_op = nullptr; xnn_caches caches = { .code_cache = NULL, .weights_cache = NULL, }; xnn_weights_cache weights_cache; if (use_weights_cache()) { xnn_init_weights_cache(&weights_cache); caches.weights_cache = &weights_cache; } const xnn_status status = xnn_create_fully_connected_nc_f32( input_channels(), output_channels(), input_stride(), output_stride(), kernel.data(), has_bias() ? bias.data() : nullptr, output_min, output_max, transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, &caches, &fully_connected_op); if (status == xnn_status_unsupported_hardware) { GTEST_SKIP(); } ASSERT_EQ(xnn_status_success, status); ASSERT_NE(nullptr, fully_connected_op); if (use_weights_cache()) { ASSERT_EQ(xnn_status_success, xnn_finalize_weights_cache(&weights_cache, xnn_weights_cache_finalization_kind_soft)); } // Smart pointer to automatically delete fully_connected_op. std::unique_ptr auto_fully_connected_op(fully_connected_op, xnn_delete_operator); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_f32( fully_connected_op, batch_size(), input.data(), output.data(), nullptr /* thread pool */)); ASSERT_EQ(xnn_status_success, xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); VerifyF32(output, output_ref, output_max, output_min); if (use_weights_cache()) { // Create another operator with the same weights cache. xnn_operator_t fully_connected_op2 = nullptr; size_t old_weights_cache_size = weights_cache.cache.weights.size; ASSERT_EQ(xnn_status_success, xnn_create_fully_connected_nc_f32( input_channels(), output_channels(), input_stride(), output_stride(), kernel.data(), has_bias() ? bias.data() : nullptr, output_min, output_max, transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, &caches, &fully_connected_op2)); ASSERT_NE(nullptr, fully_connected_op2); std::unique_ptr auto_fully_connected_op(fully_connected_op2, xnn_delete_operator); std::vector output2(output.size(), nanf("")); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_f32( fully_connected_op2, batch_size(), input.data(), output2.data(), nullptr /* thread pool */)); ASSERT_EQ(xnn_status_success, xnn_run_operator(fully_connected_op2, nullptr /* thread pool */)); VerifyWeightsCache(weights_cache, old_weights_cache_size); xnn_release_weights_cache(&weights_cache); VerifyF32(output, output_ref, output_max, output_min); } } } void VerifyF32(const std::vector& output, const std::vector& output_ref, float output_max, float output_min) const { // Verify results. for (size_t i = 0; i < batch_size(); i++) { for (size_t c = 0; c < output_channels(); c++) { ASSERT_LE(output[i * output_stride() + c], output_max) << "batch index = " << i << ", channel = " << c; ASSERT_GE(output[i * output_stride() + c], output_min) << "batch index = " << i << ", channel = " << c; ASSERT_NEAR(output_ref[i * output_channels() + c], output[i * output_stride() + c], 1.0e-4 * std::abs(output_ref[i * output_channels() + c])) << "batch index = " << i << ", channel = " << c; } } } void TestF16() const { switch (weights_type()) { case WeightsType::Default: break; case WeightsType::FP32: break; default: GTEST_FAIL() << "unexpected weights type"; } std::random_device random_device; auto rng = std::mt19937(random_device()); std::uniform_real_distribution f32dist(0.1f, 1.0f); std::vector input(XNN_EXTRA_BYTES / sizeof(uint16_t) + (batch_size() - 1) * input_stride() + input_channels()); std::vector kernel(output_channels() * input_channels()); std::vector kernel_as_float(kernel.size()); std::vector bias(output_channels()); std::vector bias_as_float(bias.size()); std::vector output((batch_size() - 1) * output_stride() + output_channels()); std::vector output_ref(batch_size() * output_channels()); for (size_t iteration = 0; iteration < iterations(); iteration++) { std::generate(input.begin(), input.end(), [&]() { return fp16_ieee_from_fp32_value(f32dist(rng)); }); std::generate(kernel.begin(), kernel.end(), [&]() { return fp16_ieee_from_fp32_value(f32dist(rng)); }); std::transform(kernel.cbegin(), kernel.cend(), kernel_as_float.begin(), fp16_ieee_to_fp32_value); std::generate(bias.begin(), bias.end(), [&]() { return fp16_ieee_from_fp32_value(f32dist(rng)); }); std::transform(bias.cbegin(), bias.cend(), bias_as_float.begin(), fp16_ieee_to_fp32_value); std::fill(output.begin(), output.end(), UINT16_C(0x7E00) /* NaN */); // Compute reference results, without renormalization. if (has_bias()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { output_ref[i * output_channels() + oc] = fp16_ieee_to_fp32_value(bias[oc]); } } } else { std::fill(output_ref.begin(), output_ref.end(), 0.0f); } if (transpose_weights()) { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { output_ref[i * output_channels() + oc] += fp16_ieee_to_fp32_value(input[i * input_stride() + ic]) * fp16_ieee_to_fp32_value(kernel[ic * output_channels() + oc]); } } } } else { for (size_t i = 0; i < batch_size(); i++) { for (size_t oc = 0; oc < output_channels(); oc++) { for (size_t ic = 0; ic < input_channels(); ic++) { output_ref[i * output_channels() + oc] += fp16_ieee_to_fp32_value(input[i * input_stride() + ic]) * fp16_ieee_to_fp32_value(kernel[oc * input_channels() + ic]); } } } } // Compute clamping parameters. const float accumulated_min = *std::min_element(output_ref.cbegin(), output_ref.cend()); const float accumulated_max = *std::max_element(output_ref.cbegin(), output_ref.cend()); const float accumulated_range = accumulated_max - accumulated_min; const float scaled_min = fp16_ieee_to_fp32_value(fp16_ieee_from_fp32_value(accumulated_min + accumulated_range / 255.0f * float(qmin()))); const float scaled_max = fp16_ieee_to_fp32_value(fp16_ieee_from_fp32_value(accumulated_max - accumulated_range / 255.0f * float(255 - qmax()))); const float output_min = scaled_min == scaled_max ? -std::numeric_limits::infinity() : scaled_min; const float output_max = scaled_min == scaled_max ? +std::numeric_limits::infinity() : scaled_max; // Clamp reference results. for (float& value : output_ref) { value = std::max(std::min(value, output_max), output_min); } // Create, setup, run, and destroy Fully Connected operator. ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); xnn_operator_t fully_connected_op = nullptr; xnn_caches caches = { .code_cache = NULL, .weights_cache = NULL, }; xnn_weights_cache weights_cache; if (use_weights_cache()) { xnn_init_weights_cache(&weights_cache); caches.weights_cache = &weights_cache; } const void* kernel_data = kernel.data(); const void* bias_data = bias.data(); if (weights_type() == WeightsType::FP32) { kernel_data = kernel_as_float.data(); bias_data = bias_as_float.data(); } uint32_t flags = 0; if (transpose_weights()) { flags |= XNN_FLAG_TRANSPOSE_WEIGHTS; } if (weights_type() == WeightsType::FP32) { flags |= XNN_FLAG_FP32_STATIC_WEIGHTS; } const xnn_status status = xnn_create_fully_connected_nc_f16( input_channels(), output_channels(), input_stride(), output_stride(), kernel_data, has_bias() ? bias_data : nullptr, output_min, output_max, flags, &caches, &fully_connected_op); if (status == xnn_status_unsupported_hardware) { GTEST_SKIP(); } ASSERT_EQ(xnn_status_success, status); ASSERT_NE(nullptr, fully_connected_op); if (use_weights_cache()) { ASSERT_EQ(xnn_status_success, xnn_finalize_weights_cache(&weights_cache, xnn_weights_cache_finalization_kind_soft)); } // Smart pointer to automatically delete fully_connected_op. std::unique_ptr auto_fully_connected_op(fully_connected_op, xnn_delete_operator); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_f16( fully_connected_op, batch_size(), input.data(), output.data(), nullptr /* thread pool */)); ASSERT_EQ(xnn_status_success, xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); // Verify results. VerifyF16(output, output_ref, output_max, output_min); if (use_weights_cache()) { xnn_operator_t fully_connected_op2 = nullptr; size_t old_weights_cache_size = weights_cache.cache.weights.size; ASSERT_EQ(xnn_status_success, xnn_create_fully_connected_nc_f16( input_channels(), output_channels(), input_stride(), output_stride(), kernel_data, has_bias() ? bias_data : nullptr, output_min, output_max, flags, &caches, &fully_connected_op2)); if (status == xnn_status_unsupported_hardware) { GTEST_SKIP(); } ASSERT_EQ(xnn_status_success, status); ASSERT_NE(nullptr, fully_connected_op2); // Smart pointer to automatically delete fully_connected_op2. std::unique_ptr auto_fully_connected_op(fully_connected_op2, xnn_delete_operator); std::vector output2(output.size(), UINT16_C(0x7E00) /* NaN */); ASSERT_EQ(xnn_status_success, xnn_setup_fully_connected_nc_f16( fully_connected_op2, batch_size(), input.data(), output2.data(), nullptr /* thread pool */)); ASSERT_EQ(xnn_status_success, xnn_run_operator(fully_connected_op2, nullptr /* thread pool */)); // Verify results. VerifyF16(output2, output_ref, output_max, output_min); VerifyWeightsCache(weights_cache, old_weights_cache_size); xnn_release_weights_cache(&weights_cache); } } } void VerifyF16(const std::vector& output, const std::vector& output_ref, const float output_max, const float output_min) const { for (size_t i = 0; i < batch_size(); i++) { for (size_t c = 0; c < output_channels(); c++) { ASSERT_LE(fp16_ieee_to_fp32_value(output[i * output_stride() + c]), output_max) << "batch index = " << i << ", channel = " << c; ASSERT_GE(fp16_ieee_to_fp32_value(output[i * output_stride() + c]), output_min) << "batch index = " << i << ", channel = " << c; ASSERT_NEAR( output_ref[i * output_channels() + c], fp16_ieee_to_fp32_value(output[i * output_stride() + c]), 1.0e-2f * std::abs(output_ref[i * output_channels() + c])) << "batch index = " << i << ", channel = " << c; } } } void VerifyWeightsCache(const xnn_weights_cache& weights_cache, size_t old_size) const { ASSERT_EQ(weights_cache.cache.hits, 1); // Ensure that we did not write more weights to the cache because it was a cache hit. ASSERT_EQ(old_size, weights_cache.cache.weights.size); }; private: size_t input_channels_{1}; size_t input_stride_{0}; size_t output_channels_{1}; size_t output_stride_{0}; size_t batch_size_{1}; uint8_t qmin_{0}; uint8_t qmax_{255}; bool transpose_weights_{false}; bool has_bias_{true}; WeightsType weights_type_{WeightsType::Default}; bool use_weights_cache_{false}; size_t iterations_{1}; };