mkldnn_quantizer.cc 25.0 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
// Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

#include "paddle/fluid/inference/api/mkldnn_quantizer.h"
#include <algorithm>
17
#include <limits>
18 19 20 21 22 23 24 25 26 27 28 29
#include <map>
#include <numeric>
#include <unordered_map>
#include <utility>
#include "paddle/fluid/framework/eigen.h"
#include "paddle/fluid/framework/ir/fuse_pass_base.h"
#include "paddle/fluid/framework/ir/graph.h"
#include "paddle/fluid/framework/ir/pass.h"
#include "paddle/fluid/framework/operator.h"
#include "paddle/fluid/framework/type_defs.h"
#include "paddle/fluid/inference/analysis/analyzer.h"
#include "paddle/fluid/inference/api/analysis_predictor.h"
30
#include "paddle/fluid/platform/mkldnn_helper.h"
31 32 33 34 35 36 37
#include "paddle/fluid/platform/place.h"
#include "paddle/fluid/string/pretty_log.h"

namespace paddle {

using platform::CPUPlace;
using framework::LoDTensor;
38
using framework::Variable;
39 40 41
using framework::ir::Graph;
using ConstEigenVectorArrayMap =
    Eigen::Map<const Eigen::Array<float, Eigen::Dynamic, 1>>;
M
Michał Gallus 已提交
42 43 44 45 46
using EigenMatrixDoubleArray =
    Eigen::Array<double, Eigen::Dynamic, Eigen::Dynamic, Eigen::RowMajor>;
using EigenMatrixArray =
    Eigen::Array<float, Eigen::Dynamic, Eigen::Dynamic, Eigen::RowMajor>;
using ConstEigenMatrixArrayMap = Eigen::Map<const EigenMatrixArray>;
47
using string::PrettyLogH1;
48
using VariableNameMap = std::map<std::string, std::vector<std::string>>;
49
static LoDTensor CreateScaleTensor(int64_t channels_num = 1);
50

51 52 53 54 55 56 57 58 59 60 61 62 63
static void check_var(const Variable* var, const std::string& var_name) {
  PADDLE_ENFORCE_NOT_NULL(var, platform::errors::PreconditionNotMet(
                                   "%s is not in the scope", var_name));
  PADDLE_ENFORCE_EQ(
      var->IsType<LoDTensor>(), true,
      platform::errors::PreconditionNotMet("Only support lod tensor now."));
}

static void check_tensor(const LoDTensor& tensor) {
  PADDLE_ENFORCE_GT(tensor.dims().size(), 0, platform::errors::InvalidArgument(
                                                 "Tensor dimension is empty."));
}

64 65
void AnalysisPredictor::MkldnnQuantizer::CalculateScalesForRNNWeights(
    const paddle::framework::OpDesc* op, bool gru) {
66 67 68 69 70 71 72 73 74 75 76
  const auto& wx_names = op->Input("WeightX");
  const auto& wh_names = op->Input("WeightH");
  for (size_t i = 0; i < wx_names.size(); ++i) {
    const auto& wx_name = wx_names[i];
    const auto& wh_name = wh_names[i];
    auto* wx_var = predictor_.sub_scope_->FindVar(wx_name);
    auto* wh_var = predictor_.sub_scope_->FindVar(wh_name);
    check_var(wx_var, wx_name);
    check_var(wh_var, wh_name);
    LoDTensor* wx_tensor = wx_var->GetMutable<LoDTensor>();
    LoDTensor* wh_tensor = wh_var->GetMutable<LoDTensor>();
77 78 79 80 81
    if (gru) {
      scales_[wx_name] = GetMaxChGRUScalingFactor(*wx_tensor, *wh_tensor);
    } else {
      scales_[wx_name] = GetMaxChLSTMScalingFactor(*wx_tensor, *wh_tensor);
    }
82 83 84 85 86 87
  }
}

void AnalysisPredictor::MkldnnQuantizer::CalculateScalesForOpInputs(
    const paddle::framework::OpDesc* op) {
  if (op->Type() == "fusion_gru" || op->Type() == "multi_gru") {
88 89 90
    CalculateScalesForRNNWeights(op, true);
  } else if (op->Type() == "fusion_lstm") {
    CalculateScalesForRNNWeights(op, false);
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132
  }
  for (auto const& input : op->Inputs()) {
    for (const auto& var_name : input.second) {
      // skip if scale already computed
      if (scales_.find(var_name) != scales_.end()) continue;
      auto* var = predictor_.sub_scope_->FindVar(var_name);
      check_var(var, var_name);
      LoDTensor* var_tensor = var->GetMutable<LoDTensor>();
      // force unsigned type if already know it
      bool is_unsigned = false;
      CalculateSingleScale(op->Type(), input.first, var_name, *var_tensor,
                           is_unsigned);
    }
  }
}

void AnalysisPredictor::MkldnnQuantizer::CalculateScalesForOpOutputs(
    const paddle::framework::OpDesc* op) {
  for (auto const& output : op->Outputs()) {
    for (const auto& var_name : output.second) {
      // skip if scale already computed
      if (scales_.find(var_name) != scales_.end()) continue;
      auto* var = predictor_.sub_scope_->FindVar(var_name);
      check_var(var, var_name);
      LoDTensor* var_tensor = var->GetMutable<LoDTensor>();
      // force unsigned type if already know it
      bool is_unsigned = false;
      bool compute_scale = true;
      if (op->Type() == "conv2d" || op->Type() == "fc") {
        // output of conv2d with relu must be unsigned
        std::string fuse_activation =
            op->GetAttrIfExists<std::string>("fuse_activation");
        is_unsigned = (fuse_activation == "relu" || fuse_activation == "relu6");
      } else if (op->Type() == "relu") {
        is_unsigned = true;
      } else if (op->Type() == "transpose2" || op->Type() == "reshape2" ||
                 op->Type() == "pool2d") {
        auto input_var_name = op->Input("X")[0];
        PADDLE_ENFORCE_NE(scales_.find(input_var_name), scales_.end(),
                          platform::errors::PreconditionNotMet(
                              "Input scales must be calculated before the "
                              "output scales to infer if output is unsigned."));
Z
Zuza 已提交
133 134 135 136 137 138 139 140 141 142
        if (scales_.find(input_var_name) != scales_.end()) {
          scales_[var_name] = scales_[input_var_name];
        }
        compute_scale = false;
      } else if (op->Type() == "slice") {
        auto input_var_name = op->Input("Input")[0];
        PADDLE_ENFORCE_NE(scales_.find(input_var_name), scales_.end(),
                          platform::errors::PreconditionNotMet(
                              "Input scales must be calculated before the "
                              "output scales to infer if output is unsigned."));
143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173
        if (scales_.find(input_var_name) != scales_.end()) {
          scales_[var_name] = scales_[input_var_name];
        }
        compute_scale = false;
      } else if (op->Type() == "concat") {
        // output of ops with unsigned input must be unsigned
        is_unsigned = true;
        double min_scale = std::numeric_limits<double>::max();
        for (auto input_var_name : op->Input("X")) {
          PADDLE_ENFORCE_NE(
              scales_.find(input_var_name), scales_.end(),
              platform::errors::PreconditionNotMet(
                  "Input scales must be calculated before the "
                  "output scales to infer if output is unsigned."));
          is_unsigned = is_unsigned && scales_[input_var_name].first;
          min_scale = std::min(
              min_scale, scales_[input_var_name].second.data<double>()[0]);
        }
        auto scale_tensor = CreateScaleTensor();
        scale_tensor.data<double>()[0] = min_scale;
        scales_[var_name] = {is_unsigned, scale_tensor};
        compute_scale = false;
      }
      if (compute_scale) {
        CalculateSingleScale(op->Type(), output.first, var_name, *var_tensor,
                             is_unsigned);
      }
    }
  }
}

174 175 176 177
bool AnalysisPredictor::MkldnnQuantizer::CalculateScales() {
  PrettyLogH1("--- Calculating scales for quantization");
  std::map<std::string, std::map<std::string, LoDTensor>> gathered_data;
  for (const auto* op : predictor_.inference_program_->Block(0).AllOps()) {
178
    if (platform::HasOpINT8DataType(op)) {
179
      // handle inputs first to let is_unsigned be inferred for the outputs
180 181
      CalculateScalesForOpInputs(op);
      CalculateScalesForOpOutputs(op);
182 183 184 185 186 187 188 189 190 191 192 193
    }
  }
  return true;
}

void AnalysisPredictor::MkldnnQuantizer::CalculateSingleScale(
    const std::string& op_type_name, const std::string& conn_name,
    const std::string& var_name, const LoDTensor& var_tensor,
    bool is_unsigned) {
  auto rule = qconfig_->scale_algo(op_type_name, conn_name);
  if (rule == ScaleAlgo::NONE) return;

194 195 196 197 198 199
  PADDLE_ENFORCE_GT(
      var_tensor.numel(), 0,
      platform::errors::InvalidArgument(
          "MkldnnQuantizer: LoDTensor of variable %s for quantization of op "
          "%s of connection %s should not be empty.",
          var_name, op_type_name, conn_name));
200 201 202 203 204 205

  switch (rule) {
    case ScaleAlgo::MAX:
      scales_[var_name] = GetMaxScalingFactor(var_tensor, is_unsigned);
      break;
    case ScaleAlgo::MAX_CH:
M
Michał Gallus 已提交
206 207 208 209 210 211
      scales_[var_name] = GetMaxChScalingFactor(var_tensor, is_unsigned,
                                                /*is_transposed*/ false);
      break;
    case ScaleAlgo::MAX_CH_T:
      scales_[var_name] = GetMaxChScalingFactor(var_tensor, is_unsigned,
                                                /*is_transposed*/ true);
212 213 214 215 216 217 218 219 220 221
      break;
    case ScaleAlgo::KL:
      scales_[var_name] = GetKLScalingFactor(var_tensor, is_unsigned);
      break;
    default:
      throw std::runtime_error(
          "MkldnnQuantizer: Unexpected ScaleAlgo specified.");
  }
}

222 223 224 225 226 227 228
static LoDTensor CreateScaleTensor(int64_t channels_num) {
  LoDTensor scale_tensor;
  scale_tensor.Resize({channels_num});
  scale_tensor.mutable_data<double>(CPUPlace());
  return scale_tensor;
}

229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267
std::vector<int> AnalysisPredictor::MkldnnQuantizer::ExpandQuantizedBins(
    std::vector<int> quantized_bins, std::vector<int> reference_bins) const {
  std::vector<int> expanded_quantized_bins(reference_bins.size(), 0);
  int num_merged_bins = reference_bins.size() / quantized_bins.size();
  int j_start = 0;
  int j_end = num_merged_bins;
  for (size_t idx = 0; idx < quantized_bins.size(); idx++) {
    int zero_count =
        std::count(&reference_bins[j_start], &reference_bins[j_end], 0);
    num_merged_bins = j_end - j_start;
    int avg_bin_ele;
    if (zero_count == num_merged_bins) {
      avg_bin_ele = 0;
    } else {
      avg_bin_ele = quantized_bins[idx] / (num_merged_bins - zero_count + 0.0);
    }
    for (int idx1 = j_start; idx1 < j_end; idx1++) {
      expanded_quantized_bins[idx1] =
          (reference_bins[idx1] == 0) ? 0 : avg_bin_ele;
    }
    j_start += num_merged_bins;
    j_end += num_merged_bins;
    if ((idx + 1) == quantized_bins.size() - 1) {
      j_end = reference_bins.size();
    }
  }
  return expanded_quantized_bins;
}

std::pair<bool, LoDTensor>
AnalysisPredictor::MkldnnQuantizer::GetKLScalingFactor(
    const LoDTensor& var_tensor, bool is_unsigned) const {
  ConstEigenVectorArrayMap eigen_tensor{var_tensor.data<float>(),
                                        var_tensor.numel(), 1};
  int precision_hist_num_bins = 2048;
  float max_val = eigen_tensor.maxCoeff();
  float min_val = eigen_tensor.minCoeff();
  bool is_positive = min_val >= 0.0f;
  if (is_unsigned)
268 269 270 271 272
    PADDLE_ENFORCE_EQ(
        is_positive, true,
        platform::errors::InvalidArgument(
            "Tensor is claimed to be unsigned, but its min value (%f) is < 0.0",
            min_val));
273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318

  int num_quantized_bins = 255;

  std::vector<int> hist;
  float bin_width;
  int starting_iter;
  int ending_iter = precision_hist_num_bins - 1;
  if (is_positive) {
    std::tie(hist, bin_width) =
        Histogram(var_tensor, min_val, max_val, precision_hist_num_bins);
    starting_iter = static_cast<int>(ending_iter * 0.7);
  } else {
    float th = std::max(std::abs(max_val), std::abs(min_val));
    std::tie(hist, bin_width) =
        Histogram(var_tensor, -th, th, precision_hist_num_bins);
    starting_iter = 0;
    if (std::abs(max_val) > std::abs(min_val)) {
      while (starting_iter < ending_iter) {
        if (hist[starting_iter] == 0) {
          ++starting_iter;
          continue;
        } else {
          break;
        }
      }
      starting_iter += static_cast<int>((ending_iter - starting_iter) * 0.6);
    } else {
      while (ending_iter > 0) {
        if (hist[ending_iter] == 0) {
          --ending_iter;
          continue;
        } else {
          break;
        }
      }
      starting_iter = static_cast<int>(0.6 * ending_iter);
    }
  }
  auto P_sum = eigen_tensor.size();
  int min_kl_divergence = 0;
  int min_kl_index = 0;
  bool kl_inited = false;
  for (int i = starting_iter; i <= ending_iter; i++) {
    std::vector<int> reference_distr_P(&hist[0], &hist[i]);
    auto outliers_count =
        std::accumulate(&hist[i], &hist[precision_hist_num_bins], 0);
319
    if (i <= 0 || reference_distr_P[i - 1] == 0) {
320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365
      continue;
    }
    reference_distr_P[i - 1] += outliers_count;
    auto reference_distr_bins = reference_distr_P;
    std::vector<int> candidate_distr_Q(&hist[0], &hist[i]);
    int num_merged_bins = i / num_quantized_bins;
    std::vector<int> candidate_distr_Q_quantized(num_quantized_bins, 0);
    int j_start = 0;
    int j_end = num_merged_bins;
    for (int idx = 0; idx < num_quantized_bins; idx++) {
      candidate_distr_Q_quantized[idx] = std::accumulate(
          &candidate_distr_Q[j_start], &candidate_distr_Q[j_end], 0);
      j_start += num_merged_bins;
      j_end += num_merged_bins;
      if ((idx + 1) == num_quantized_bins - 1) {
        j_end = i;
      }
    }
    candidate_distr_Q =
        ExpandQuantizedBins(candidate_distr_Q_quantized, reference_distr_bins);
    int Q_sum =
        std::accumulate(candidate_distr_Q.begin(), candidate_distr_Q.end(), 0);
    auto kl_divergence =
        SafeEntropy(reference_distr_P, P_sum, candidate_distr_Q, Q_sum);
    if (!kl_inited) {
      min_kl_divergence = kl_divergence;
      min_kl_index = i;
      kl_inited = true;
    } else if (kl_divergence < min_kl_divergence) {
      min_kl_divergence = kl_divergence;
      min_kl_index = i;
    } else {
    }
  }
  if (min_kl_index == 0) {
    while (starting_iter > 0) {
      if (hist[starting_iter] == 0) {
        starting_iter -= 1;
        continue;
      } else {
        break;
      }
    }
    min_kl_index = starting_iter;
  }

366 367
  LoDTensor scale_tensor = CreateScaleTensor();
  scale_tensor.data<double>()[0] = 1.0 / ((min_kl_index + 0.5) * bin_width);
368 369 370 371 372 373 374 375 376 377 378 379

  return std::make_pair(is_unsigned, scale_tensor);
}

std::pair<bool, LoDTensor>
AnalysisPredictor::MkldnnQuantizer::GetMaxScalingFactor(
    const LoDTensor& var_tensor, bool is_unsigned) const {
  ConstEigenVectorArrayMap eigen_tensor{var_tensor.data<float>(),
                                        var_tensor.numel(), 1};
  float max_abs = eigen_tensor.abs().maxCoeff();
  float min_val = eigen_tensor.minCoeff();
  if (is_unsigned)
380 381 382 383 384
    PADDLE_ENFORCE_GE(
        min_val, 0.0f,
        platform::errors::InvalidArgument(
            "Tensor is claimed to be unsigned, but its min value (%f) is < 0.0",
            min_val));
385

386 387
  LoDTensor scale_tensor = CreateScaleTensor();
  scale_tensor.data<double>()[0] = 1.0 / max_abs;
388 389 390 391 392 393

  return std::make_pair(is_unsigned, scale_tensor);
}

std::pair<bool, LoDTensor>
AnalysisPredictor::MkldnnQuantizer::GetMaxChScalingFactor(
M
Michał Gallus 已提交
394
    const LoDTensor& var_tensor, bool is_unsigned, bool is_transposed) const {
395
  check_tensor(var_tensor);
396 397 398 399 400

  ConstEigenVectorArrayMap eigen_tensor{var_tensor.data<float>(),
                                        var_tensor.numel(), 1};
  float min_val = eigen_tensor.minCoeff();
  if (is_unsigned)
401 402 403 404 405
    PADDLE_ENFORCE_GE(
        min_val, 0.0f,
        platform::errors::InvalidArgument(
            "Tensor is claimed to be unsigned, but its min value (%f) is < 0.0",
            min_val));
406

M
Michał Gallus 已提交
407 408 409 410 411
  auto dims = var_tensor.dims();
  constexpr int num_col_dims = 1;
  auto flattened_dims = framework::flatten_to_2d(dims, num_col_dims);
  ConstEigenMatrixArrayMap eigen_tensor_mat{
      var_tensor.data<float>(), flattened_dims[0], flattened_dims[1]};
412

M
Michał Gallus 已提交
413 414 415 416 417
  EigenMatrixDoubleArray scales;
  if (is_transposed) {
    scales = 1.0 / eigen_tensor_mat.cast<double>().abs().colwise().maxCoeff();
  } else {
    scales = 1.0 / eigen_tensor_mat.cast<double>().abs().rowwise().maxCoeff();
418
  }
M
Michał Gallus 已提交
419 420 421 422 423
  int output_channel_axis = is_transposed;
  int channels = dims[output_channel_axis];
  LoDTensor scale_tensor = CreateScaleTensor(channels);
  auto* scale_ptr = scale_tensor.mutable_data<double>(CPUPlace());
  std::copy(scales.data(), scales.data() + scales.size(), scale_ptr);
424 425 426 427

  return std::make_pair(is_unsigned, scale_tensor);
}

428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482
std::pair<bool, LoDTensor>
AnalysisPredictor::MkldnnQuantizer::GetMaxChGRUScalingFactor(
    const LoDTensor& wx_tensor, const LoDTensor& wh_tensor) const {
  check_tensor(wx_tensor);
  check_tensor(wh_tensor);

  int OC = wh_tensor.dims()[0];
  std::vector<float> scale_ur(2 * OC);
  std::vector<float> scale_o(OC);

  for (int row_id = 0; row_id < wx_tensor.dims()[0]; row_id++) {
    for (int col_id = 0; col_id < 2 * OC; col_id++) {
      int idx = (row_id * wx_tensor.dims()[1]) + col_id;
      auto abs_value = std::abs(wx_tensor.data<float>()[idx]);
      if (row_id == 0) {
        scale_ur[col_id] = abs_value;
      } else {
        if (abs_value > scale_ur[col_id]) scale_ur[col_id] = abs_value;
      }
    }
  }

  for (int i = 0; i < 2 * OC * OC; i++) {
    int col_id = i % (2 * OC);
    auto abs_value = std::abs(wh_tensor.data<float>()[i]);
    if (abs_value > scale_ur[col_id]) scale_ur[col_id] = abs_value;
  }

  for (int row_id = 0; row_id < wx_tensor.dims()[0]; row_id++) {
    for (int col_id = 2 * OC; col_id < wx_tensor.dims()[1]; col_id++) {
      int idx = (row_id * wx_tensor.dims()[1]) + col_id;
      auto abs_value = std::abs(wx_tensor.data<float>()[idx]);
      if (row_id == 0) {
        scale_o[col_id % OC] = abs_value;
      } else {
        if (abs_value > scale_o[col_id]) scale_o[col_id % OC] = abs_value;
      }
    }
  }

  for (int i = 2 * OC * OC; i < OC * wh_tensor.dims()[1]; i++) {
    int col_id = i % OC;
    auto abs_value = std::abs(wh_tensor.data<float>()[i]);
    if (abs_value > scale_o[col_id]) scale_o[col_id] = abs_value;
  }
  scale_ur.insert(scale_ur.end(), scale_o.begin(), scale_o.end());
  transform(scale_ur.begin(), scale_ur.end(), scale_ur.begin(),
            [](float& c) { return 1 / c; });
  LoDTensor scale_tensor = CreateScaleTensor(scale_ur.size());
  auto* scale_ptr = scale_tensor.mutable_data<double>(CPUPlace());
  std::copy(scale_ur.begin(), scale_ur.end(), scale_ptr);
  bool is_unsigned = false;
  return std::make_pair(is_unsigned, scale_tensor);
}

483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517
std::pair<bool, LoDTensor>
AnalysisPredictor::MkldnnQuantizer::GetMaxChLSTMScalingFactor(
    const LoDTensor& wx_tensor, const LoDTensor& wh_tensor) const {
  check_tensor(wx_tensor);
  check_tensor(wh_tensor);

  std::vector<float> scale(wx_tensor.dims()[1]);

  for (int row_id = 0; row_id < wx_tensor.dims()[0]; row_id++) {
    for (int col_id = 0; col_id < wx_tensor.dims()[1]; col_id++) {
      int idx = (row_id * wx_tensor.dims()[1]) + col_id;
      auto abs_value = std::abs(wx_tensor.data<float>()[idx]);
      if (row_id == 0) {
        scale[col_id] = abs_value;
      } else {
        if (abs_value > scale[col_id]) scale[col_id] = abs_value;
      }
    }
  }
  for (int row_id = 0; row_id < wh_tensor.dims()[0]; row_id++) {
    for (int col_id = 0; col_id < wh_tensor.dims()[1]; col_id++) {
      int idx = (row_id * wh_tensor.dims()[1]) + col_id;
      auto abs_value = std::abs(wh_tensor.data<float>()[idx]);
      if (abs_value > scale[col_id]) scale[col_id] = abs_value;
    }
  }
  transform(scale.begin(), scale.end(), scale.begin(),
            [](float& c) { return 1 / c; });
  LoDTensor scale_tensor = CreateScaleTensor(scale.size());
  auto* scale_ptr = scale_tensor.mutable_data<double>(CPUPlace());
  std::copy(scale.begin(), scale.end(), scale_ptr);
  bool is_unsigned = false;
  return std::make_pair(is_unsigned, scale_tensor);
}

518 519 520 521 522
std::pair<std::vector<int>, float>
AnalysisPredictor::MkldnnQuantizer::Histogram(
    const framework::LoDTensor& var_tensor, float min_val, float max_val,
    size_t num_bins) const {
  PADDLE_ENFORCE_GT(num_bins, 0,
523 524 525 526 527 528 529 530 531 532 533 534 535
                    platform::errors::InvalidArgument(
                        "MkldnnQuantizer: To calculate Histogram, num_bins (" +
                        std::to_string(num_bins) + ") must be positive."));
  PADDLE_ENFORCE_GT(var_tensor.numel(), 0,
                    platform::errors::InvalidArgument(
                        "MkldnnQuantizer: To calculate Histogram, the tensor "
                        "must not be empty."));
  PADDLE_ENFORCE_GE(max_val, min_val,
                    platform::errors::InvalidArgument(
                        "MkldnnQuantizer: To calculate Histogram, max_val (" +
                        std::to_string(max_val) + ") must be greater or equal"
                                                  "to min_val (" +
                        std::to_string(min_val) + ")."));
536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
  ConstEigenVectorArrayMap eigen_tensor{var_tensor.data<float>(),
                                        var_tensor.numel(), 1};
  auto bin_width = std::abs(max_val - min_val) / num_bins;
  std::vector<int> hist(num_bins);

  for (int i = 0; i < eigen_tensor.size(); i++) {
    int bin = std::min(
        num_bins - 1,
        static_cast<size_t>(floor((eigen_tensor[i] - min_val) / bin_width)));
    ++hist[bin];
  }

  return std::make_pair(std::move(hist), std::move(bin_width));
}

551 552 553 554
void AnalysisPredictor::MkldnnQuantizer::ClearDeviceContext() const {
  platform::DeviceContextPool& pool = platform::DeviceContextPool::Instance();
  platform::MKLDNNDeviceContext* dev_ctx =
      (platform::MKLDNNDeviceContext*)pool.Get(predictor_.place_);
555 556
  dev_ctx->ResetBlobMap(
      paddle::platform::MKLDNNDeviceContext::tls().get_curr_exec());
557 558
}

559 560 561 562 563 564
void AnalysisPredictor::MkldnnQuantizer::PrepareArgument() const {
  auto& arg = predictor_.argument_;
  if (!arg.scope_valid()) arg.SetScope(new framework::Scope);
  arg.SetMainProgramNotOwned(predictor_.inference_program_.get());
  auto graph = std::unique_ptr<Graph>(new Graph(arg.main_program()));
  arg.SetMainGraph(graph.release());
565
  auto* scope_ptr = arg.scope_ptr();
566 567
  PADDLE_ENFORCE_NOT_NULL(scope_ptr, platform::errors::PreconditionNotMet(
                                         "The scope should not be nullptr."));
568
  arg.main_graph().SetNotOwned(framework::ir::kParamScopeAttr, scope_ptr);
569 570 571

  auto* builder = predictor_.config_.pass_builder();
  builder->SetPasses({
572
      "cpu_quantize_pass", "cpu_quantize_squash_pass",
573 574 575 576 577
  });
  if (predictor_.config_.ir_debug_) builder->TurnOnDebug();
  auto passes = builder->AllPasses();
  predictor_.argument_.SetIrAnalysisPasses(passes);
  predictor_.argument_.SetAnalysisPasses(
578 579
      {"ir_graph_clean_pass", "ir_analysis_pass", "memory_optimize_pass",
       "ir_graph_to_program_pass"});
580 581 582 583 584 585
  predictor_.argument_.SetQuantVarScales(scales_);
}

bool AnalysisPredictor::MkldnnQuantizer::Quantize() {
  if (!RunWarmup()) return false;
  if (!CalculateScales()) return false;
586
  ClearDeviceContext();
587 588 589 590 591 592 593 594 595 596 597 598 599 600
  predictor_.PrepareScope(predictor_.scope_);
  predictor_.CreateExecutor();
  if (!RunQuantizePasses()) return false;
  predictor_.PrepareExecutor();
  predictor_.PrepareFeedFetch();
  return true;
}

bool AnalysisPredictor::MkldnnQuantizer::RunQuantizePasses() const {
  predictor_.executor_->CreateVariables(*predictor_.inference_program_, 0, true,
                                        predictor_.sub_scope_);
  PrepareArgument();
  auto& arg = predictor_.argument_;
  Analyzer().Run(&arg);
601 602 603
  PADDLE_ENFORCE_EQ(
      arg.scope_valid(), true,
      platform::errors::PreconditionNotMet("The scope should be valid."));
604 605 606 607 608 609 610 611 612 613 614 615 616 617
  VLOG(5) << "to prepare executor";
  ARGUMENT_CHECK_FIELD((&arg), ir_analyzed_program);
  predictor_.inference_program_.reset(
      new framework::ProgramDesc(arg.ir_analyzed_program()));
  LOG(INFO) << "== optimize 2 end ==";
  predictor_.executor_->CreateVariables(*predictor_.inference_program_, 0,
                                        false, predictor_.sub_scope_);
  return true;
}

bool AnalysisPredictor::MkldnnQuantizer::RunWarmup() const {
  VLOG(3) << "Predictor: run a quantization warmup iteration";
  auto warmup_data = qconfig_->warmup_data();
  PADDLE_ENFORCE_NOT_NULL(warmup_data,
618 619
                          platform::errors::PreconditionNotMet(
                              "Warmup data cannot be NULL in the config."));
620 621 622 623 624 625 626 627 628 629 630 631
  PrettyLogH1("--- Running warmup iteration for quantization");

  // Run the inference program
  std::vector<PaddleTensor> output_slots;
  predictor_.Run(*warmup_data, &output_slots, qconfig_->warmup_batch_size());

  return true;
}

float AnalysisPredictor::MkldnnQuantizer::SafeEntropy(
    std::vector<int> reference_distr_P, int P_sum,
    std::vector<int> candidate_distr_Q, int Q_sum) const {
632 633 634 635
  PADDLE_ENFORCE_EQ(reference_distr_P.size(), candidate_distr_Q.size(),
                    platform::errors::InvalidArgument(
                        "The P size %d should be equal to Q size %d",
                        reference_distr_P.size(), candidate_distr_Q.size()));
636 637 638 639 640 641 642 643 644
  float tmp_sum1 = 0;
  float tmp_sum2 = 0;
  for (size_t idx = 0; idx < reference_distr_P.size(); idx++) {
    int p_idx = reference_distr_P[idx];
    int q_idx = candidate_distr_Q[idx];
    if (p_idx == 0) {
      tmp_sum1 += 0;
      tmp_sum2 += 0;
    } else {
645 646 647 648 649
      PADDLE_ENFORCE_NE(
          q_idx, 0,
          platform::errors::PreconditionNotMet(
              "MkldnnQuantizer: Fatal error!, idx = " + std::to_string(idx) +
              " qindex = 0! p_idx = " + std::to_string(p_idx)));
650 651 652 653 654 655 656 657
    }
    tmp_sum1 += p_idx * (log(Q_sum * p_idx));
    tmp_sum2 += p_idx * (log(P_sum * q_idx));
  }
  return (tmp_sum1 - tmp_sum2) / P_sum;
}

}  // namespace paddle