XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 1 | // Copyright (c) Facebook, Inc. and its affiliates. |
| 2 | // All rights reserved. |
| 3 | // |
| 4 | // Copyright 2019 Google LLC |
| 5 | // |
| 6 | // This source code is licensed under the BSD-style license found in the |
| 7 | // LICENSE file in the root directory of this source tree. |
| 8 | |
| 9 | #pragma once |
| 10 | |
| 11 | #include <gtest/gtest.h> |
| 12 | |
| 13 | #include <cstddef> |
| 14 | #include <cstdlib> |
| 15 | #include <algorithm> |
| 16 | #include <cmath> |
| 17 | #include <functional> |
Marat Dukhan | 5ce30d9 | 2020-04-14 03:31:26 -0700 | [diff] [blame] | 18 | #include <limits> |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 19 | #include <random> |
| 20 | #include <vector> |
| 21 | |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 22 | #include <fp16.h> |
| 23 | |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 24 | #include <xnnpack.h> |
| 25 | |
| 26 | |
| 27 | class FullyConnectedOperatorTester { |
| 28 | public: |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 29 | enum class WeightsType { |
| 30 | Default, |
| 31 | FP32, |
| 32 | }; |
| 33 | |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 34 | inline FullyConnectedOperatorTester& input_channels(size_t input_channels) { |
| 35 | assert(input_channels >= 1); |
| 36 | this->input_channels_ = input_channels; |
| 37 | return *this; |
| 38 | } |
| 39 | |
| 40 | inline size_t input_channels() const { |
| 41 | return this->input_channels_; |
| 42 | } |
| 43 | |
| 44 | inline FullyConnectedOperatorTester& output_channels(size_t output_channels) { |
| 45 | assert(output_channels >= 1); |
| 46 | this->output_channels_ = output_channels; |
| 47 | return *this; |
| 48 | } |
| 49 | |
| 50 | inline size_t output_channels() const { |
| 51 | return this->output_channels_; |
| 52 | } |
| 53 | |
| 54 | inline FullyConnectedOperatorTester& batch_size(size_t batch_size) { |
| 55 | assert(batch_size >= 1); |
| 56 | this->batch_size_ = batch_size; |
| 57 | return *this; |
| 58 | } |
| 59 | |
| 60 | inline size_t batch_size() const { |
| 61 | return this->batch_size_; |
| 62 | } |
| 63 | |
| 64 | inline FullyConnectedOperatorTester& input_stride(size_t input_stride) { |
| 65 | assert(input_stride >= 1); |
| 66 | this->input_stride_ = input_stride; |
| 67 | return *this; |
| 68 | } |
| 69 | |
| 70 | inline size_t input_stride() const { |
| 71 | if (this->input_stride_ == 0) { |
| 72 | return input_channels(); |
| 73 | } else { |
| 74 | assert(this->input_stride_ >= input_channels()); |
| 75 | return this->input_stride_; |
| 76 | } |
| 77 | } |
| 78 | |
| 79 | inline FullyConnectedOperatorTester& output_stride(size_t output_stride) { |
| 80 | assert(output_stride >= 1); |
| 81 | this->output_stride_ = output_stride; |
| 82 | return *this; |
| 83 | } |
| 84 | |
| 85 | inline size_t output_stride() const { |
| 86 | if (this->output_stride_ == 0) { |
| 87 | return output_channels(); |
| 88 | } else { |
| 89 | assert(this->output_stride_ >= output_channels()); |
| 90 | return this->output_stride_; |
| 91 | } |
| 92 | } |
| 93 | |
| 94 | inline FullyConnectedOperatorTester& qmin(uint8_t qmin) { |
| 95 | this->qmin_ = qmin; |
| 96 | return *this; |
| 97 | } |
| 98 | |
| 99 | inline uint8_t qmin() const { |
| 100 | return this->qmin_; |
| 101 | } |
| 102 | |
| 103 | inline FullyConnectedOperatorTester& qmax(uint8_t qmax) { |
| 104 | this->qmax_ = qmax; |
| 105 | return *this; |
| 106 | } |
| 107 | |
| 108 | inline uint8_t qmax() const { |
| 109 | return this->qmax_; |
| 110 | } |
| 111 | |
Marat Dukhan | c4f0ff9 | 2019-12-03 14:59:08 -0800 | [diff] [blame] | 112 | inline FullyConnectedOperatorTester& transpose_weights(bool transpose_weights) { |
| 113 | this->transpose_weights_ = transpose_weights; |
| 114 | return *this; |
| 115 | } |
| 116 | |
| 117 | inline bool transpose_weights() const { |
| 118 | return this->transpose_weights_; |
| 119 | } |
| 120 | |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 121 | inline FullyConnectedOperatorTester& has_bias(bool has_bias) { |
| 122 | this->has_bias_ = has_bias; |
| 123 | return *this; |
| 124 | } |
| 125 | |
| 126 | inline bool has_bias() const { |
| 127 | return this->has_bias_; |
| 128 | } |
| 129 | |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 130 | inline FullyConnectedOperatorTester& weights_type(WeightsType weights_type) { |
| 131 | this->weights_type_ = weights_type; |
| 132 | return *this; |
| 133 | } |
| 134 | |
| 135 | inline WeightsType weights_type() const { |
| 136 | return this->weights_type_; |
| 137 | } |
| 138 | |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 139 | inline FullyConnectedOperatorTester& iterations(size_t iterations) { |
| 140 | this->iterations_ = iterations; |
| 141 | return *this; |
| 142 | } |
| 143 | |
| 144 | inline size_t iterations() const { |
| 145 | return this->iterations_; |
| 146 | } |
| 147 | |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 148 | void TestQS8() const { |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 149 | ASSERT_EQ(weights_type(), WeightsType::Default); |
| 150 | |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 151 | std::random_device random_device; |
| 152 | auto rng = std::mt19937(random_device()); |
Marat Dukhan | e7991e7 | 2021-08-10 22:30:03 -0700 | [diff] [blame] | 153 | auto i32rng = std::bind(std::uniform_int_distribution<int32_t>(-10000, 10000), std::ref(rng)); |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 154 | auto i8rng = std::bind(std::uniform_int_distribution<int32_t>( |
Marat Dukhan | e7991e7 | 2021-08-10 22:30:03 -0700 | [diff] [blame] | 155 | std::numeric_limits<int8_t>::min(), std::numeric_limits<int8_t>::max()), std::ref(rng)); |
| 156 | auto w8rng = std::bind(std::uniform_int_distribution<int32_t>( |
| 157 | -std::numeric_limits<int8_t>::max(), std::numeric_limits<int8_t>::max()), std::ref(rng)); |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 158 | |
| 159 | std::vector<int8_t> input(XNN_EXTRA_BYTES / sizeof(int8_t) + |
| 160 | (batch_size() - 1) * input_stride() + input_channels()); |
| 161 | std::vector<int8_t> kernel(output_channels() * input_channels()); |
| 162 | std::vector<int32_t> bias(output_channels()); |
| 163 | std::vector<int8_t> output((batch_size() - 1) * output_stride() + output_channels()); |
| 164 | std::vector<int32_t> accumulators(batch_size() * output_channels()); |
| 165 | std::vector<double> output_ref(batch_size() * output_channels()); |
| 166 | |
| 167 | const int8_t input_zero_point = 127; |
| 168 | |
| 169 | for (size_t iteration = 0; iteration < iterations(); iteration++) { |
| 170 | std::generate(input.begin(), input.end(), std::ref(i8rng)); |
Marat Dukhan | e7991e7 | 2021-08-10 22:30:03 -0700 | [diff] [blame] | 171 | std::generate(kernel.begin(), kernel.end(), std::ref(w8rng)); |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 172 | std::generate(bias.begin(), bias.end(), std::ref(i32rng)); |
| 173 | std::fill(output.begin(), output.end(), 0xA5); |
| 174 | |
| 175 | // Compute reference results, without renormalization. |
| 176 | if (has_bias()) { |
| 177 | for (size_t i = 0; i < batch_size(); i++) { |
| 178 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 179 | accumulators[i * output_channels() + oc] = bias[oc]; |
| 180 | } |
| 181 | } |
| 182 | } else { |
| 183 | std::fill(accumulators.begin(), accumulators.end(), 0); |
| 184 | } |
| 185 | if (transpose_weights()) { |
| 186 | for (size_t i = 0; i < batch_size(); i++) { |
| 187 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 188 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 189 | accumulators[i * output_channels() + oc] += |
| 190 | (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * |
| 191 | int32_t(kernel[ic * output_channels() + oc]); |
| 192 | } |
| 193 | } |
| 194 | } |
| 195 | } else { |
| 196 | for (size_t i = 0; i < batch_size(); i++) { |
| 197 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 198 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 199 | accumulators[i * output_channels() + oc] += |
| 200 | (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * |
| 201 | int32_t(kernel[oc * input_channels() + ic]); |
| 202 | } |
| 203 | } |
| 204 | } |
| 205 | } |
| 206 | |
| 207 | // Compute renormalization parameters. |
| 208 | const int32_t accumulated_min = *std::min_element(accumulators.cbegin(), accumulators.cend()); |
| 209 | const int32_t accumulated_max = *std::max_element(accumulators.cbegin(), accumulators.cend()); |
| 210 | |
| 211 | const double output_scale = double(uint32_t(accumulated_max - accumulated_min)) / 255.0; |
| 212 | const int8_t output_zero_point = int8_t(std::max(std::min( |
| 213 | lrint(-0.5 - 0.5 * double(accumulated_min + accumulated_max) / output_scale), |
| 214 | long(std::numeric_limits<int8_t>::max())), long(std::numeric_limits<int8_t>::min()))); |
| 215 | |
| 216 | // Renormalize reference results. |
| 217 | std::transform(accumulators.cbegin(), accumulators.cend(), output_ref.begin(), |
| 218 | [this, output_scale, output_zero_point](int32_t x) -> double { |
| 219 | return std::max<double>(std::min<double>(double(x) / output_scale, double(qmax() - 0x80) - output_zero_point), double(qmin() - 0x80) - output_zero_point); |
| 220 | }); |
| 221 | |
| 222 | // Create, setup, run, and destroy Fully Connected operator. |
| 223 | ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); |
| 224 | xnn_operator_t fully_connected_op = nullptr; |
| 225 | |
| 226 | const xnn_status status = xnn_create_fully_connected_nc_qs8( |
| 227 | input_channels(), output_channels(), |
| 228 | input_stride(), output_stride(), |
| 229 | input_zero_point, 1.0f /* input scale */, |
| 230 | 1.0f /* kernel scale */, |
| 231 | kernel.data(), has_bias() ? bias.data() : nullptr, |
| 232 | output_zero_point, output_scale, int8_t(qmin() - 0x80), int8_t(qmax() - 0x80), |
| 233 | transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, |
| 234 | &fully_connected_op); |
| 235 | if (status == xnn_status_unsupported_hardware) { |
| 236 | GTEST_SKIP(); |
| 237 | } |
| 238 | ASSERT_EQ(xnn_status_success, status); |
| 239 | ASSERT_NE(nullptr, fully_connected_op); |
| 240 | |
| 241 | // Smart pointer to automatically delete fully_connected_op. |
| 242 | std::unique_ptr<xnn_operator, decltype(&xnn_delete_operator)> auto_fully_connected_op(fully_connected_op, xnn_delete_operator); |
| 243 | |
| 244 | ASSERT_EQ(xnn_status_success, |
| 245 | xnn_setup_fully_connected_nc_qs8( |
| 246 | fully_connected_op, |
| 247 | batch_size(), |
| 248 | input.data(), output.data(), |
| 249 | nullptr /* thread pool */)); |
| 250 | |
| 251 | ASSERT_EQ(xnn_status_success, |
| 252 | xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); |
| 253 | |
| 254 | // Verify results. |
| 255 | for (size_t i = 0; i < batch_size(); i++) { |
| 256 | for (size_t c = 0; c < output_channels(); c++) { |
| 257 | ASSERT_LE(int32_t(output[i * output_stride() + c]), int32_t(qmax() - 0x80)) |
| 258 | << "batch index = " << i << ", channel = " << c; |
| 259 | ASSERT_GE(int32_t(output[i * output_stride() + c]), int32_t(qmin() - 0x80)) |
| 260 | << "batch index = " << i << ", channel = " << c; |
| 261 | ASSERT_NEAR( |
| 262 | output_ref[i * output_channels() + c], |
| 263 | double(output[i * output_stride() + c]) - double(output_zero_point), |
| 264 | 0.9) |
| 265 | << "batch index = " << i << ", channel = " << c; |
| 266 | } |
| 267 | } |
| 268 | } |
| 269 | } |
| 270 | |
Marat Dukhan | 08b7a97 | 2020-07-14 18:17:29 -0700 | [diff] [blame] | 271 | void TestQU8() const { |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 272 | ASSERT_EQ(weights_type(), WeightsType::Default); |
| 273 | |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 274 | std::random_device random_device; |
| 275 | auto rng = std::mt19937(random_device()); |
Marat Dukhan | e7991e7 | 2021-08-10 22:30:03 -0700 | [diff] [blame] | 276 | auto i32rng = std::bind(std::uniform_int_distribution<int32_t>(-10000, 10000), std::ref(rng)); |
| 277 | auto u8rng = std::bind( |
| 278 | std::uniform_int_distribution<uint32_t>(0, std::numeric_limits<uint8_t>::max()), std::ref(rng)); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 279 | |
| 280 | std::vector<uint8_t> input(XNN_EXTRA_BYTES / sizeof(uint8_t) + |
| 281 | (batch_size() - 1) * input_stride() + input_channels()); |
| 282 | std::vector<uint8_t> kernel(output_channels() * input_channels()); |
| 283 | std::vector<int32_t> bias(output_channels()); |
| 284 | std::vector<uint8_t> output((batch_size() - 1) * output_stride() + output_channels()); |
| 285 | std::vector<int32_t> accumulators(batch_size() * output_channels()); |
| 286 | std::vector<double> output_ref(batch_size() * output_channels()); |
| 287 | |
| 288 | const uint8_t input_zero_point = 127; |
| 289 | const uint8_t kernel_zero_point = 127; |
| 290 | |
| 291 | for (size_t iteration = 0; iteration < iterations(); iteration++) { |
| 292 | std::generate(input.begin(), input.end(), std::ref(u8rng)); |
| 293 | std::generate(kernel.begin(), kernel.end(), std::ref(u8rng)); |
Marat Dukhan | ecd8311 | 2020-08-03 21:50:28 -0700 | [diff] [blame] | 294 | std::generate(bias.begin(), bias.end(), std::ref(i32rng)); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 295 | std::fill(output.begin(), output.end(), 0xA5); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 296 | |
| 297 | // Compute reference results, without renormalization. |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 298 | if (has_bias()) { |
| 299 | for (size_t i = 0; i < batch_size(); i++) { |
| 300 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 301 | accumulators[i * output_channels() + oc] = bias[oc]; |
| 302 | } |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 303 | } |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 304 | } else { |
| 305 | std::fill(accumulators.begin(), accumulators.end(), 0); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 306 | } |
Marat Dukhan | c4f0ff9 | 2019-12-03 14:59:08 -0800 | [diff] [blame] | 307 | if (transpose_weights()) { |
| 308 | for (size_t i = 0; i < batch_size(); i++) { |
| 309 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 310 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 311 | accumulators[i * output_channels() + oc] += |
| 312 | (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * |
| 313 | (int32_t(kernel[ic * output_channels() + oc]) - int32_t(kernel_zero_point)); |
| 314 | } |
| 315 | } |
| 316 | } |
| 317 | } else { |
| 318 | for (size_t i = 0; i < batch_size(); i++) { |
| 319 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 320 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 321 | accumulators[i * output_channels() + oc] += |
| 322 | (int32_t(input[i * input_stride() + ic]) - int32_t(input_zero_point)) * |
| 323 | (int32_t(kernel[oc * input_channels() + ic]) - int32_t(kernel_zero_point)); |
| 324 | } |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 325 | } |
| 326 | } |
| 327 | } |
| 328 | |
| 329 | // Compute renormalization parameters. |
| 330 | const int32_t accumulated_min = *std::min_element(accumulators.cbegin(), accumulators.cend()); |
| 331 | const int32_t accumulated_max = *std::max_element(accumulators.cbegin(), accumulators.cend()); |
| 332 | |
| 333 | const double output_scale = double(uint32_t(accumulated_max - accumulated_min)) / 255.0; |
| 334 | const uint8_t output_zero_point = uint8_t(std::max(std::min( |
| 335 | lrint(127.5 - 0.5 * double(accumulated_min + accumulated_max) / output_scale), |
| 336 | long(std::numeric_limits<uint8_t>::max())), long(std::numeric_limits<uint8_t>::min()))); |
| 337 | |
| 338 | // Renormalize reference results. |
| 339 | std::transform(accumulators.cbegin(), accumulators.cend(), output_ref.begin(), |
| 340 | [this, output_scale, output_zero_point](int32_t x) -> double { |
| 341 | return std::max<double>(std::min<double>(double(x) / output_scale, double(qmax()) - output_zero_point), double(qmin()) - output_zero_point); |
| 342 | }); |
| 343 | |
| 344 | // Create, setup, run, and destroy Fully Connected operator. |
Marat Dukhan | 04f03be | 2019-11-19 12:36:47 -0800 | [diff] [blame] | 345 | ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 346 | xnn_operator_t fully_connected_op = nullptr; |
| 347 | |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 348 | const xnn_status status = xnn_create_fully_connected_nc_qu8( |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 349 | input_channels(), output_channels(), |
| 350 | input_stride(), output_stride(), |
| 351 | input_zero_point, 1.0f /* input scale */, |
| 352 | kernel_zero_point, 1.0f /* kernel scale */, |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 353 | kernel.data(), has_bias() ? bias.data() : nullptr, |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 354 | output_zero_point, output_scale, qmin(), qmax(), |
Marat Dukhan | c4f0ff9 | 2019-12-03 14:59:08 -0800 | [diff] [blame] | 355 | transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 356 | &fully_connected_op); |
| 357 | if (status == xnn_status_unsupported_hardware) { |
| 358 | GTEST_SKIP(); |
| 359 | } |
| 360 | ASSERT_EQ(xnn_status_success, status); |
| 361 | ASSERT_NE(nullptr, fully_connected_op); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 362 | |
| 363 | // Smart pointer to automatically delete fully_connected_op. |
| 364 | std::unique_ptr<xnn_operator, decltype(&xnn_delete_operator)> auto_fully_connected_op(fully_connected_op, xnn_delete_operator); |
| 365 | |
| 366 | ASSERT_EQ(xnn_status_success, |
Marat Dukhan | 08b7a97 | 2020-07-14 18:17:29 -0700 | [diff] [blame] | 367 | xnn_setup_fully_connected_nc_qu8( |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 368 | fully_connected_op, |
| 369 | batch_size(), |
| 370 | input.data(), output.data(), |
| 371 | nullptr /* thread pool */)); |
| 372 | |
| 373 | ASSERT_EQ(xnn_status_success, |
| 374 | xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); |
| 375 | |
| 376 | // Verify results. |
| 377 | for (size_t i = 0; i < batch_size(); i++) { |
| 378 | for (size_t c = 0; c < output_channels(); c++) { |
| 379 | ASSERT_LE(int32_t(output[i * output_stride() + c]), int32_t(qmax())) |
| 380 | << "batch index = " << i << ", channel = " << c; |
| 381 | ASSERT_GE(int32_t(output[i * output_stride() + c]), int32_t(qmin())) |
| 382 | << "batch index = " << i << ", channel = " << c; |
| 383 | ASSERT_NEAR( |
| 384 | output_ref[i * output_channels() + c], |
| 385 | double(output[i * output_stride() + c]) - double(output_zero_point), |
| 386 | 0.9) |
| 387 | << "batch index = " << i << ", channel = " << c; |
| 388 | } |
| 389 | } |
| 390 | } |
| 391 | } |
| 392 | |
| 393 | void TestF32() const { |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 394 | ASSERT_EQ(weights_type(), WeightsType::Default); |
| 395 | |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 396 | std::random_device random_device; |
| 397 | auto rng = std::mt19937(random_device()); |
Marat Dukhan | e7991e7 | 2021-08-10 22:30:03 -0700 | [diff] [blame] | 398 | auto f32rng = std::bind(std::uniform_real_distribution<float>(0.1f, 1.0f), std::ref(rng)); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 399 | |
| 400 | std::vector<float> input(XNN_EXTRA_BYTES / sizeof(float) + |
| 401 | (batch_size() - 1) * input_stride() + input_channels()); |
| 402 | std::vector<float> kernel(output_channels() * input_channels()); |
| 403 | std::vector<float> bias(output_channels()); |
| 404 | std::vector<float> output((batch_size() - 1) * output_stride() + output_channels()); |
| 405 | std::vector<float> output_ref(batch_size() * output_channels()); |
| 406 | |
| 407 | for (size_t iteration = 0; iteration < iterations(); iteration++) { |
| 408 | std::generate(input.begin(), input.end(), std::ref(f32rng)); |
| 409 | std::generate(kernel.begin(), kernel.end(), std::ref(f32rng)); |
| 410 | std::generate(bias.begin(), bias.end(), std::ref(f32rng)); |
| 411 | std::fill(output.begin(), output.end(), nanf("")); |
| 412 | |
| 413 | // Compute reference results, without renormalization. |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 414 | if (has_bias()) { |
| 415 | for (size_t i = 0; i < batch_size(); i++) { |
| 416 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 417 | output_ref[i * output_channels() + oc] = bias[oc]; |
| 418 | } |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 419 | } |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 420 | } else { |
| 421 | std::fill(output_ref.begin(), output_ref.end(), 0.0f); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 422 | } |
Marat Dukhan | c4f0ff9 | 2019-12-03 14:59:08 -0800 | [diff] [blame] | 423 | if (transpose_weights()) { |
| 424 | for (size_t i = 0; i < batch_size(); i++) { |
| 425 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 426 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 427 | output_ref[i * output_channels() + oc] += |
| 428 | input[i * input_stride() + ic] * kernel[ic * output_channels() + oc]; |
| 429 | } |
| 430 | } |
| 431 | } |
| 432 | } else { |
| 433 | for (size_t i = 0; i < batch_size(); i++) { |
| 434 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 435 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 436 | output_ref[i * output_channels() + oc] += |
| 437 | input[i * input_stride() + ic] * kernel[oc * input_channels() + ic]; |
| 438 | } |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 439 | } |
| 440 | } |
| 441 | } |
| 442 | |
| 443 | // Compute clamping parameters. |
Marat Dukhan | c6edf92 | 2019-10-03 15:08:04 -0700 | [diff] [blame] | 444 | const float accumulated_min = *std::min_element(output_ref.cbegin(), output_ref.cend()); |
| 445 | const float accumulated_max = *std::max_element(output_ref.cbegin(), output_ref.cend()); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 446 | |
Marat Dukhan | 869c62d | 2020-04-09 17:17:55 -0700 | [diff] [blame] | 447 | const float output_min = qmin() == 0 ? -std::numeric_limits<float>::infinity() : |
| 448 | accumulated_min + (accumulated_max - accumulated_min) / 255.0f * float(qmin()); |
| 449 | const float output_max = qmax() == 255 ? std::numeric_limits<float>::infinity() : |
| 450 | accumulated_max - (accumulated_max - accumulated_min) / 255.0f * float(255 - qmax()); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 451 | |
| 452 | // Clamp reference results. |
| 453 | for (float& value : output_ref) { |
| 454 | value = std::max(std::min(value, output_max), output_min); |
| 455 | } |
| 456 | |
| 457 | // Create, setup, run, and destroy Fully Connected operator. |
Marat Dukhan | 04f03be | 2019-11-19 12:36:47 -0800 | [diff] [blame] | 458 | ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 459 | xnn_operator_t fully_connected_op = nullptr; |
| 460 | |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 461 | const xnn_status status = xnn_create_fully_connected_nc_f32( |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 462 | input_channels(), output_channels(), |
| 463 | input_stride(), output_stride(), |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 464 | kernel.data(), has_bias() ? bias.data() : nullptr, |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 465 | output_min, output_max, |
Marat Dukhan | c4f0ff9 | 2019-12-03 14:59:08 -0800 | [diff] [blame] | 466 | transpose_weights() ? XNN_FLAG_TRANSPOSE_WEIGHTS : 0, |
Marat Dukhan | d23cb6e | 2021-04-01 01:18:58 -0700 | [diff] [blame] | 467 | &fully_connected_op); |
| 468 | if (status == xnn_status_unsupported_hardware) { |
| 469 | GTEST_SKIP(); |
| 470 | } |
| 471 | ASSERT_EQ(xnn_status_success, status); |
| 472 | ASSERT_NE(nullptr, fully_connected_op); |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 473 | |
| 474 | // Smart pointer to automatically delete fully_connected_op. |
| 475 | std::unique_ptr<xnn_operator, decltype(&xnn_delete_operator)> auto_fully_connected_op(fully_connected_op, xnn_delete_operator); |
| 476 | |
| 477 | ASSERT_EQ(xnn_status_success, |
| 478 | xnn_setup_fully_connected_nc_f32( |
| 479 | fully_connected_op, |
| 480 | batch_size(), |
| 481 | input.data(), output.data(), |
| 482 | nullptr /* thread pool */)); |
| 483 | |
| 484 | ASSERT_EQ(xnn_status_success, |
| 485 | xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); |
| 486 | |
| 487 | // Verify results. |
| 488 | for (size_t i = 0; i < batch_size(); i++) { |
| 489 | for (size_t c = 0; c < output_channels(); c++) { |
| 490 | ASSERT_LE(output[i * output_stride() + c], output_max) |
| 491 | << "batch index = " << i << ", channel = " << c; |
| 492 | ASSERT_GE(output[i * output_stride() + c], output_min) |
| 493 | << "batch index = " << i << ", channel = " << c; |
| 494 | ASSERT_NEAR( |
| 495 | output_ref[i * output_channels() + c], |
| 496 | output[i * output_stride() + c], |
| 497 | 1.0e-4 * std::abs(output_ref[i * output_channels() + c])) |
| 498 | << "batch index = " << i << ", channel = " << c; |
| 499 | } |
| 500 | } |
| 501 | } |
| 502 | } |
| 503 | |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 504 | void TestF16() const { |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 505 | switch (weights_type()) { |
| 506 | case WeightsType::Default: |
| 507 | break; |
| 508 | case WeightsType::FP32: |
| 509 | break; |
| 510 | default: |
| 511 | GTEST_FAIL() << "unexpected weights type"; |
| 512 | } |
| 513 | |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 514 | std::random_device random_device; |
| 515 | auto rng = std::mt19937(random_device()); |
| 516 | auto f32rng = std::bind(std::uniform_real_distribution<float>(0.1f, 1.0f), std::ref(rng)); |
| 517 | auto f16rng = std::bind(fp16_ieee_from_fp32_value, f32rng); |
| 518 | |
| 519 | std::vector<uint16_t> input(XNN_EXTRA_BYTES / sizeof(uint16_t) + |
| 520 | (batch_size() - 1) * input_stride() + input_channels()); |
| 521 | std::vector<uint16_t> kernel(output_channels() * input_channels()); |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 522 | std::vector<float> kernel_as_float(kernel.size()); |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 523 | std::vector<uint16_t> bias(output_channels()); |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 524 | std::vector<float> bias_as_float(bias.size()); |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 525 | std::vector<uint16_t> output((batch_size() - 1) * output_stride() + output_channels()); |
| 526 | std::vector<float> output_ref(batch_size() * output_channels()); |
| 527 | |
| 528 | for (size_t iteration = 0; iteration < iterations(); iteration++) { |
| 529 | std::generate(input.begin(), input.end(), std::ref(f16rng)); |
| 530 | std::generate(kernel.begin(), kernel.end(), std::ref(f16rng)); |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 531 | std::transform(kernel.cbegin(), kernel.cend(), kernel_as_float.begin(), fp16_ieee_to_fp32_value); |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 532 | std::generate(bias.begin(), bias.end(), std::ref(f16rng)); |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 533 | std::transform(bias.cbegin(), bias.cend(), bias_as_float.begin(), fp16_ieee_to_fp32_value); |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 534 | std::fill(output.begin(), output.end(), UINT16_C(0x7C00)); |
| 535 | |
| 536 | // Compute reference results, without renormalization. |
| 537 | if (has_bias()) { |
| 538 | for (size_t i = 0; i < batch_size(); i++) { |
| 539 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 540 | output_ref[i * output_channels() + oc] = fp16_ieee_to_fp32_value(bias[oc]); |
| 541 | } |
| 542 | } |
| 543 | } else { |
| 544 | std::fill(output_ref.begin(), output_ref.end(), 0.0f); |
| 545 | } |
| 546 | if (transpose_weights()) { |
| 547 | for (size_t i = 0; i < batch_size(); i++) { |
| 548 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 549 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 550 | output_ref[i * output_channels() + oc] += |
| 551 | fp16_ieee_to_fp32_value(input[i * input_stride() + ic]) * fp16_ieee_to_fp32_value(kernel[ic * output_channels() + oc]); |
| 552 | } |
| 553 | } |
| 554 | } |
| 555 | } else { |
| 556 | for (size_t i = 0; i < batch_size(); i++) { |
| 557 | for (size_t oc = 0; oc < output_channels(); oc++) { |
| 558 | for (size_t ic = 0; ic < input_channels(); ic++) { |
| 559 | output_ref[i * output_channels() + oc] += |
| 560 | fp16_ieee_to_fp32_value(input[i * input_stride() + ic]) * fp16_ieee_to_fp32_value(kernel[oc * input_channels() + ic]); |
| 561 | } |
| 562 | } |
| 563 | } |
| 564 | } |
| 565 | |
| 566 | // Compute clamping parameters. |
| 567 | const float accumulated_min = *std::min_element(output_ref.cbegin(), output_ref.cend()); |
| 568 | const float accumulated_max = *std::max_element(output_ref.cbegin(), output_ref.cend()); |
| 569 | const float accumulated_range = accumulated_max - accumulated_min; |
| 570 | const float scaled_min = fp16_ieee_to_fp32_value(fp16_ieee_from_fp32_value(accumulated_min + accumulated_range / 255.0f * float(qmin()))); |
| 571 | const float scaled_max = fp16_ieee_to_fp32_value(fp16_ieee_from_fp32_value(accumulated_max - accumulated_range / 255.0f * float(255 - qmax()))); |
| 572 | const float output_min = scaled_min == scaled_max ? -std::numeric_limits<float>::infinity() : scaled_min; |
| 573 | const float output_max = scaled_min == scaled_max ? +std::numeric_limits<float>::infinity() : scaled_max; |
| 574 | |
| 575 | // Clamp reference results. |
| 576 | for (float& value : output_ref) { |
| 577 | value = std::max(std::min(value, output_max), output_min); |
| 578 | } |
| 579 | |
| 580 | // Create, setup, run, and destroy Fully Connected operator. |
| 581 | ASSERT_EQ(xnn_status_success, xnn_initialize(nullptr /* allocator */)); |
| 582 | xnn_operator_t fully_connected_op = nullptr; |
| 583 | |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 584 | const void* kernel_data = kernel.data(); |
| 585 | const void* bias_data = bias.data(); |
| 586 | if (weights_type() == WeightsType::FP32) { |
| 587 | kernel_data = kernel_as_float.data(); |
| 588 | bias_data = bias_as_float.data(); |
| 589 | } |
| 590 | uint32_t flags = 0; |
| 591 | if (transpose_weights()) { |
| 592 | flags |= XNN_FLAG_TRANSPOSE_WEIGHTS; |
| 593 | } |
| 594 | if (weights_type() == WeightsType::FP32) { |
| 595 | flags |= XNN_FLAG_FP32_STATIC_WEIGHTS; |
| 596 | } |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 597 | const xnn_status status = xnn_create_fully_connected_nc_f16( |
| 598 | input_channels(), output_channels(), |
| 599 | input_stride(), output_stride(), |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 600 | kernel_data, has_bias() ? bias_data : nullptr, |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 601 | output_min, output_max, |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 602 | flags, |
Marat Dukhan | ddb3d16 | 2021-10-25 17:05:51 -0700 | [diff] [blame] | 603 | &fully_connected_op); |
| 604 | if (status == xnn_status_unsupported_hardware) { |
| 605 | GTEST_SKIP(); |
| 606 | } |
| 607 | ASSERT_EQ(xnn_status_success, status); |
| 608 | ASSERT_NE(nullptr, fully_connected_op); |
| 609 | |
| 610 | // Smart pointer to automatically delete fully_connected_op. |
| 611 | std::unique_ptr<xnn_operator, decltype(&xnn_delete_operator)> auto_fully_connected_op(fully_connected_op, xnn_delete_operator); |
| 612 | |
| 613 | ASSERT_EQ(xnn_status_success, |
| 614 | xnn_setup_fully_connected_nc_f16( |
| 615 | fully_connected_op, |
| 616 | batch_size(), |
| 617 | input.data(), output.data(), |
| 618 | nullptr /* thread pool */)); |
| 619 | |
| 620 | ASSERT_EQ(xnn_status_success, |
| 621 | xnn_run_operator(fully_connected_op, nullptr /* thread pool */)); |
| 622 | |
| 623 | // Verify results. |
| 624 | for (size_t i = 0; i < batch_size(); i++) { |
| 625 | for (size_t c = 0; c < output_channels(); c++) { |
| 626 | ASSERT_LE(fp16_ieee_to_fp32_value(output[i * output_stride() + c]), output_max) |
| 627 | << "batch index = " << i << ", channel = " << c; |
| 628 | ASSERT_GE(fp16_ieee_to_fp32_value(output[i * output_stride() + c]), output_min) |
| 629 | << "batch index = " << i << ", channel = " << c; |
| 630 | ASSERT_NEAR( |
| 631 | output_ref[i * output_channels() + c], |
| 632 | fp16_ieee_to_fp32_value(output[i * output_stride() + c]), |
| 633 | 1.0e-2f * std::abs(output_ref[i * output_channels() + c])) |
| 634 | << "batch index = " << i << ", channel = " << c; |
| 635 | } |
| 636 | } |
| 637 | } |
| 638 | } |
| 639 | |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 640 | private: |
| 641 | size_t input_channels_{1}; |
| 642 | size_t input_stride_{0}; |
| 643 | size_t output_channels_{1}; |
| 644 | size_t output_stride_{0}; |
| 645 | size_t batch_size_{1}; |
| 646 | uint8_t qmin_{0}; |
| 647 | uint8_t qmax_{255}; |
Marat Dukhan | c4f0ff9 | 2019-12-03 14:59:08 -0800 | [diff] [blame] | 648 | bool transpose_weights_{false}; |
Marat Dukhan | f568f08 | 2019-10-30 09:47:07 -0700 | [diff] [blame] | 649 | bool has_bias_{true}; |
Marat Dukhan | 1d6b7c9 | 2022-01-14 21:18:44 -0800 | [diff] [blame] | 650 | WeightsType weights_type_{WeightsType::Default}; |
XNNPACK Team | b455b12 | 2019-09-27 18:10:33 -0700 | [diff] [blame] | 651 | size_t iterations_{1}; |
| 652 | }; |