“1db1a0dcea98abff0364e3aceb3d4d7d8084ab75”上不存在“git@gitcode.net:BaiXuePrincess/Paddle.git”
fake_quantize_op.cc 21.6 KB
Newer Older
视言's avatar
视言 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
/* Copyright (c) 2016 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/operators/fake_quantize_op.h"
#include <string>
17 18 19
#include "paddle/fluid/framework/eigen.h"
#include "paddle/fluid/operators/clip_op.h"
#include "paddle/fluid/platform/transform.h"
视言's avatar
视言 已提交
20 21 22 23

namespace paddle {
namespace operators {

24 25 26 27 28
template <typename T>
struct Compare {
 public:
  bool operator()(const T a, const T b) { return (std::abs(a) < std::abs(b)); }
};
29 30 31 32 33

template <typename T>
struct FindAbsMaxFunctor<platform::CPUDeviceContext, T> {
  void operator()(const platform::CPUDeviceContext& ctx, const T* in,
                  const int num, T* out) {
34
    *out = std::abs(*(std::max_element(in + 0, in + num, Compare<T>())));
35 36 37 38 39
  }
};

template struct FindAbsMaxFunctor<platform::CPUDeviceContext, float>;

40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
template <typename T>
struct FindChannelAbsMaxFunctor<platform::CPUDeviceContext, T> {
  void operator()(const platform::CPUDeviceContext& ctx, const T* in,
                  const int num, const int channel, T* out) {
    const int channel_size = num / channel;
    for (int i = 0; i < channel; i++) {
      auto* start = in + i * channel_size;
      auto* end = in + (i + 1) * channel_size;
      out[i] = std::abs(*(std::max_element(start, end, Compare<T>())));
    }
  }
};

template struct FindChannelAbsMaxFunctor<platform::CPUDeviceContext, float>;

55 56 57 58 59 60
template <typename T>
struct ClipAndFakeQuantFunctor<platform::CPUDeviceContext, T> {
  void operator()(const platform::CPUDeviceContext& ctx,
                  const framework::Tensor& in, const framework::Tensor& scale,
                  const int bin_cnt, framework::Tensor* out) {
    T s = scale.data<T>()[0];
61
    T inv_s = inverse(s);
62 63 64 65
    platform::Transform<platform::CPUDeviceContext> trans;
    trans(ctx, in.data<T>(), in.data<T>() + in.numel(),
          out->mutable_data<T>(ctx.GetPlace()), ClipFunctor<T>(-s, s));
    auto out_e = framework::EigenVector<T>::Flatten(*out);
66
    out_e.device(*ctx.eigen_device()) = (bin_cnt * inv_s * out_e).round();
67 68 69 70 71
  }
};

template struct ClipAndFakeQuantFunctor<platform::CPUDeviceContext, float>;

72 73 74 75 76 77
template <typename T>
struct ClipAndFakeQuantDequantFunctor<platform::CPUDeviceContext, T> {
  void operator()(const platform::CPUDeviceContext& ctx,
                  const framework::Tensor& in, const framework::Tensor& scale,
                  const int bin_cnt, framework::Tensor* out) {
    T s = scale.data<T>()[0];
78 79
    T inv_s = inverse(s);

80 81 82 83 84
    platform::Transform<platform::CPUDeviceContext> trans;
    trans(ctx, in.data<T>(), in.data<T>() + in.numel(),
          out->mutable_data<T>(ctx.GetPlace()), ClipFunctor<T>(-s, s));
    auto out_e = framework::EigenVector<T>::Flatten(*out);
    out_e.device(*ctx.eigen_device()) =
85
        (s / bin_cnt) * (bin_cnt * inv_s * out_e).round();
86 87 88 89 90
  }
};
template struct ClipAndFakeQuantDequantFunctor<platform::CPUDeviceContext,
                                               float>;

91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110
template <typename T>
struct ChannelClipAndFakeQuantFunctor<platform::CPUDeviceContext, T> {
  void operator()(const platform::CPUDeviceContext& ctx,
                  const framework::Tensor& in, const framework::Tensor& scale,
                  const int bin_cnt, const int channel,
                  framework::Tensor* out) {
    auto* scale_data = scale.data<T>();
    auto* in_data = in.data<T>();
    auto* out_data = out->mutable_data<T>(ctx.GetPlace());
    const int channel_size = in.numel() / channel;
    platform::Transform<platform::CPUDeviceContext> trans;
    for (int i = 0; i < channel; i++) {
      T s = scale_data[i];
      auto* start = in_data + i * channel_size;
      auto* end = in_data + (i + 1) * channel_size;
      trans(ctx, start, end, out_data + i * channel_size,
            ClipFunctor<T>(-s, s));
    }
    for (int i = 0; i < channel; i++) {
      T s = scale_data[i];
111
      T inv_s = inverse(s);
112 113
      framework::Tensor one_channel_out = out->Slice(i, i + 1);
      auto out_e = framework::EigenVector<T>::Flatten(one_channel_out);
114
      out_e.device(*ctx.eigen_device()) = (bin_cnt * inv_s * out_e).round();
115 116 117 118 119 120 121
    }
  }
};

template struct ChannelClipAndFakeQuantFunctor<platform::CPUDeviceContext,
                                               float>;

122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
template <typename T>
struct FindRangeAbsMaxFunctor<platform::CPUDeviceContext, T> {
  void operator()(const platform::CPUDeviceContext& ctx,
                  const framework::Tensor& cur_scale,
                  const framework::Tensor& last_scale,
                  const framework::Tensor& iter, const int window_size,
                  framework::Tensor* scales_arr, framework::Tensor* out_scale) {
    T* scale_arr = scales_arr->mutable_data<T>(ctx.GetPlace());
    int64_t it = iter.data<int64_t>()[0];
    int idx = it % window_size;
    T removed = scale_arr[idx];
    T cur = cur_scale.data<T>()[0];
    scale_arr[idx] = cur;

    T max = last_scale.data<T>()[0];
    if (max < cur) {
      max = cur;
    } else if (fabs(removed - max) < 1e-6) {
      int size = (it > window_size) ? window_size : it;
      FindAbsMaxFunctor<platform::CPUDeviceContext, T>()(ctx, scale_arr, size,
                                                         &max);
    }
    out_scale->mutable_data<T>(ctx.GetPlace())[0] = max;
  }
};

template struct FindRangeAbsMaxFunctor<platform::CPUDeviceContext, float>;

150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173
template <typename T>
struct FindMovingAverageAbsMaxFunctor<platform::CPUDeviceContext, T> {
  void operator()(const platform::CPUDeviceContext& ctx,
                  const framework::Tensor& in_accum,
                  const framework::Tensor& in_state, const T* cur_scale,
                  const float rate, framework::Tensor* out_state,
                  framework::Tensor* out_accum, framework::Tensor* out_scale) {
    T accum = in_accum.data<T>()[0];
    T state = in_state.data<T>()[0];
    T scale = cur_scale[0];

    state = rate * state + 1;
    accum = rate * accum + scale;
    scale = accum / state;

    out_state->mutable_data<T>(ctx.GetPlace())[0] = state;
    out_accum->mutable_data<T>(ctx.GetPlace())[0] = accum;
    out_scale->mutable_data<T>(ctx.GetPlace())[0] = scale;
  }
};

template struct FindMovingAverageAbsMaxFunctor<platform::CPUDeviceContext,
                                               float>;

174
class FakeQuantizeAbsMaxOp : public framework::OperatorWithKernel {
视言's avatar
视言 已提交
175
 public:
176 177 178 179
  FakeQuantizeAbsMaxOp(const std::string& type,
                       const framework::VariableNameMap& inputs,
                       const framework::VariableNameMap& outputs,
                       const framework::AttributeMap& attrs)
视言's avatar
视言 已提交
180 181
      : OperatorWithKernel(type, inputs, outputs, attrs) {}

182
  void InferShape(framework::InferShapeContext* ctx) const override {
183 184 185 186 187
    OP_INOUT_CHECK(ctx->HasInput("X"), "Input", "X", "FakeQuantizeAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("Out"), "Output", "Out",
                   "FakeQuantizeAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("OutScale"), "Output", "OutScale",
                   "FakeQuantizeAbsMax");
视言's avatar
视言 已提交
188
    ctx->SetOutputDim("Out", ctx->GetInputDim("X"));
189
    ctx->SetOutputDim("OutScale", {1});
视言's avatar
视言 已提交
190 191
    ctx->ShareLoD("X", /*->*/ "Out");
  }
192 193 194 195

 protected:
  framework::OpKernelType GetExpectedKernelType(
      const framework::ExecutionContext& ctx) const override {
196 197 198
    return framework::OpKernelType(
        OperatorWithKernel::IndicateVarDataType(ctx, "X"),
        ctx.device_context());
199
  }
视言's avatar
视言 已提交
200 201
};

202
class FakeQuantizeAbsMaxOpMaker : public framework::OpProtoAndCheckerMaker {
视言's avatar
视言 已提交
203 204
 public:
  void Make() override {
205 206 207 208 209
    AddInput("X", "(Tensor) Input is float data type.");
    AddOutput("Out",
              "(Tensor) Output of quantized low level tensor, "
              "but also saved as float data type.");
    AddOutput("OutScale", "(Tensor) Current scale");
视言's avatar
视言 已提交
210 211
    AddAttr<int>("bit_length", "(int, default 8)")
        .SetDefault(8)
212
        .AddCustomChecker([](const int& bit_length) {
213 214 215 216 217
          PADDLE_ENFORCE_EQ(bit_length >= 1 && bit_length <= 16, true,
                            platform::errors::InvalidArgument(
                                "'bit_length' should be between 1 and 16, but "
                                "the received is %d",
                                bit_length));
视言's avatar
视言 已提交
218 219 220 221
        });
    AddComment(R"DOC(
FakeQuantize operator

222
$$scale = max(abs(X))$$
223 224
$$range = 2^{bit_length - 1} - 1$$
$$Out = round(X/scale * range)$$
视言's avatar
视言 已提交
225

226 227 228
)DOC");
  }
};
视言's avatar
视言 已提交
229

Z
Zhen Wang 已提交
230 231 232 233 234
class FakeChannelWiseQuantizeAbsMaxOp : public framework::OperatorWithKernel {
 public:
  using framework::OperatorWithKernel::OperatorWithKernel;

  void InferShape(framework::InferShapeContext* ctx) const override {
235 236 237 238 239 240
    OP_INOUT_CHECK(ctx->HasInput("X"), "Input", "X",
                   "FakeChannelWiseQuantizeAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("Out"), "Output", "Out",
                   "FakeChannelWiseQuantizeAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("OutScale"), "Output", "OutScale",
                   "FakeChannelWiseQuantizeAbsMax");
Z
Zhen Wang 已提交
241
    ctx->SetOutputDim("Out", ctx->GetInputDim("X"));
242
    ctx->SetOutputDim("OutScale", {ctx->GetInputDim("X")[0]});
Z
Zhen Wang 已提交
243 244 245 246 247 248
    ctx->ShareLoD("X", /*->*/ "Out");
  }

 protected:
  framework::OpKernelType GetExpectedKernelType(
      const framework::ExecutionContext& ctx) const override {
249 250
    return framework::OpKernelType(
        OperatorWithKernel::IndicateVarDataType(ctx, "X"), ctx.GetPlace());
Z
Zhen Wang 已提交
251 252 253 254 255 256 257 258 259 260 261
  }
};

class FakeChannelWiseQuantizeAbsMaxOpMaker
    : public framework::OpProtoAndCheckerMaker {
 public:
  void Make() override {
    AddInput("X", "(Tensor) Input is float data type.");
    AddOutput("Out",
              "(Tensor) Output of quantized low level tensor, "
              "but also saved as float data type.");
262
    AddOutput("OutScale", "(Tensor) Current channel wise scale");
Z
Zhen Wang 已提交
263 264 265
    AddAttr<int>("bit_length", "(int, default 8)")
        .SetDefault(8)
        .AddCustomChecker([](const int& bit_length) {
266 267 268 269 270
          PADDLE_ENFORCE_EQ(bit_length >= 1 && bit_length <= 16, true,
                            platform::errors::InvalidArgument(
                                "'bit_length' should be between 1 and 16, but "
                                "the received is %d",
                                bit_length));
Z
Zhen Wang 已提交
271 272 273 274 275 276
        });
    AddComment(R"DOC(
The scale of FakeChannelWiseQuantize operator is a vector.
In detail, each channel of the input X has a scale value.

$$scale_c = max(abs(X_c))$$
Z
Zhen Wang 已提交
277 278
$$range = 2^{bit\_length - 1} - 1$$
$$Out_c = round(\frac{X_c * range} {scale_c})$$
Z
Zhen Wang 已提交
279
In above three formulas, the range value of c is as follow:
Z
Zhen Wang 已提交
280
$$0 \leq c \lt \ the\ channel\ number\ of\ X$$
Z
Zhen Wang 已提交
281 282 283 284
)DOC");
  }
};

285 286 287 288 289 290 291
class FakeQuantizeRangeAbsMaxOp : public framework::OperatorWithKernel {
 public:
  FakeQuantizeRangeAbsMaxOp(const std::string& type,
                            const framework::VariableNameMap& inputs,
                            const framework::VariableNameMap& outputs,
                            const framework::AttributeMap& attrs)
      : OperatorWithKernel(type, inputs, outputs, attrs) {}
视言's avatar
视言 已提交
292

293
  void InferShape(framework::InferShapeContext* ctx) const override {
294 295 296 297 298
    OP_INOUT_CHECK(ctx->HasInput("X"), "Input", "X", "FakeQuantizeRangeAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("Out"), "Output", "Out",
                   "FakeQuantizeRangeAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("OutScale"), "Output", "OutScale",
                   "FakeQuantizeRangeAbsMax");
299 300 301 302 303 304 305 306
    if (ctx->HasOutput("OutScales")) {
      int window_size = ctx->Attrs().Get<int>("window_size");
      ctx->SetOutputDim("OutScales", {window_size});
    }
    ctx->SetOutputDim("Out", ctx->GetInputDim("X"));
    ctx->SetOutputDim("OutScale", {1});
    ctx->ShareLoD("X", /*->*/ "Out");
  }
视言's avatar
视言 已提交
307

308 309 310
 protected:
  framework::OpKernelType GetExpectedKernelType(
      const framework::ExecutionContext& ctx) const override {
311 312 313
    return framework::OpKernelType(
        OperatorWithKernel::IndicateVarDataType(ctx, "X"),
        ctx.device_context());
314 315
  }
};
视言's avatar
视言 已提交
316

317 318 319 320 321 322 323 324 325 326 327 328 329 330 331
class FakeQuantizeRangeAbsMaxOpMaker
    : public framework::OpProtoAndCheckerMaker {
 public:
  void Make() override {
    AddInput("X", "(Tensor) Input is float data type.");
    AddInput("InScale", "Last scale.");
    AddInput("Iter", "Global step iteration.").AsDispensable();
    AddOutput("Out", "(Tensor) Output of quantized low level tensor.");
    AddOutput("OutScale", " Current scale");
    AddOutput("OutScales", "(Tensor) scale buffer.").AsDispensable();
    AddAttr<int>("window_size", "(int, default 10000) window range size.")
        .SetDefault(10000);
    AddAttr<int>("bit_length", "(int, default 8), quantization bit number.")
        .SetDefault(8)
        .AddCustomChecker([](const int& bit_length) {
332 333 334 335 336
          PADDLE_ENFORCE_EQ(bit_length >= 1 && bit_length <= 16, true,
                            platform::errors::InvalidArgument(
                                "'bit_length' should be between 1 and 16, but "
                                "the received is %d",
                                bit_length));
337
        });
338 339 340 341
    AddAttr<bool>("is_test",
                  "(bool, default false) Set to true for inference only, false "
                  "for training. Some layers may run faster when this is true.")
        .SetDefault(false);
342 343
    AddComment(R"DOC(
FakeQuantize operator is used in static quantization.
视言's avatar
视言 已提交
344

345
$$scale = max(max(abs(x)), history_abs_max)$$
346 347
$$range = 2^{bit_length - 1} - 1$$
$$Out = round(X/scale * range)$$
视言's avatar
视言 已提交
348 349 350 351 352

)DOC");
  }
};

353 354
class FakeQuantOrWithDequantMovingAverageAbsMaxOp
    : public framework::OperatorWithKernel {
355
 public:
356 357 358 359
  FakeQuantOrWithDequantMovingAverageAbsMaxOp(
      const std::string& type, const framework::VariableNameMap& inputs,
      const framework::VariableNameMap& outputs,
      const framework::AttributeMap& attrs)
360 361 362
      : OperatorWithKernel(type, inputs, outputs, attrs) {}

  void InferShape(framework::InferShapeContext* ctx) const override {
363 364 365 366 367 368
    OP_INOUT_CHECK(ctx->HasInput("X"), "Input", "X",
                   "FakeQuantOrWithDequantMovingAverageAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("Out"), "Output", "Out",
                   "FakeQuantOrWithDequantMovingAverageAbsMax");
    OP_INOUT_CHECK(ctx->HasOutput("OutScale"), "Output", "OutScale",
                   "FakeQuantOrWithDequantMovingAverageAbsMax");
369 370 371 372 373 374 375 376 377 378 379 380 381 382
    if (ctx->HasOutput("OutState")) {
      ctx->SetOutputDim("OutState", {1});
    }
    if (ctx->HasOutput("OutAccum")) {
      ctx->SetOutputDim("OutAccum", {1});
    }
    ctx->SetOutputDim("Out", ctx->GetInputDim("X"));
    ctx->SetOutputDim("OutScale", {1});
    ctx->ShareLoD("X", /*->*/ "Out");
  }

 protected:
  framework::OpKernelType GetExpectedKernelType(
      const framework::ExecutionContext& ctx) const override {
383 384 385
    return framework::OpKernelType(
        OperatorWithKernel::IndicateVarDataType(ctx, "X"),
        ctx.device_context());
386 387 388
  }
};

389
class FakeQuantOrWithDequantMovingAverageAbsMaxOpMaker
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405
    : public framework::OpProtoAndCheckerMaker {
 public:
  void Make() override {
    AddInput("X", "(Tensor) Input is float data type.");
    AddInput("InScale", "Last scale.");
    AddInput("InAccum", "Last accum.").AsDispensable();
    AddInput("InState", "Last state.").AsDispensable();
    AddOutput("Out", "(Tensor) Output of quantized low level tensor.");
    AddOutput("OutScale", " Current scale");
    AddOutput("OutState", "(Tensor) state buffer.").AsDispensable();
    AddOutput("OutAccum", "(Tensor) accum buffer.").AsDispensable();
    AddAttr<float>("moving_rate", "(float, default 0.9) moving rate.")
        .SetDefault(0.9);
    AddAttr<int>("bit_length", "(int, default 8), quantization bit number.")
        .SetDefault(8)
        .AddCustomChecker([](const int& bit_length) {
406 407 408 409 410
          PADDLE_ENFORCE_EQ(bit_length >= 1 && bit_length <= 16, true,
                            platform::errors::InvalidArgument(
                                "'bit_length' should be between 1 and 16, but "
                                "the received is %d",
                                bit_length));
411 412 413 414 415 416
        });
    AddAttr<bool>("is_test",
                  "(bool, default false) Set to true for inference only, false "
                  "for training. Some layers may run faster when this is true.")
        .SetDefault(false);
    AddComment(R"DOC(
417 418
This is a Base Op which support FakeQuantMovingAverageAbsMaxOp and FakeQuantDequantMovingAverageAbsMaxOp
FakeQuantMovingAverageAbsMaxOp operator is used in static quantization.
419

Z
Zhen Wang 已提交
420 421
$$scale = (moving\_rate*accum+max(abs(x)))/(moving\_rate*state+1)$$
$$range = 2^{bit\_length - 1} - 1$$
422 423
$$Out = round(X/scale * range)$$

424 425 426 427 428 429
FakeQuantDequantMovingAverageAbsMaxOp operator do the moving_average_abs_max op quant and then dequant.

$$scale = (moving\_rate*accum+max(abs(x)))/(moving\_rate*state+1)$$
$$range = 2^{bit\_length - 1} - 1$$
$$Out = round(X/scale * range) * scale / range$$

430 431 432 433
)DOC");
  }
};

Z
Zhen Wang 已提交
434 435 436 437 438
class MovingAverageAbsMaxScaleOp : public framework::OperatorWithKernel {
 public:
  using framework::OperatorWithKernel::OperatorWithKernel;

  void InferShape(framework::InferShapeContext* ctx) const override {
439 440 441 442 443 444
    OP_INOUT_CHECK(ctx->HasInput("X"), "Input", "X",
                   "MovingAverageAbsMaxScale");
    OP_INOUT_CHECK(ctx->HasOutput("Out"), "Output", "Out",
                   "MovingAverageAbsMaxScale");
    OP_INOUT_CHECK(ctx->HasOutput("OutScale"), "Output", "OutScale",
                   "MovingAverageAbsMaxScale");
Z
Zhen Wang 已提交
445 446 447 448 449 450 451 452 453 454 455 456 457 458
    if (ctx->HasOutput("OutState")) {
      ctx->SetOutputDim("OutState", {1});
    }
    if (ctx->HasOutput("OutAccum")) {
      ctx->SetOutputDim("OutAccum", {1});
    }
    ctx->SetOutputDim("Out", ctx->GetInputDim("X"));
    ctx->SetOutputDim("OutScale", {1});
    ctx->ShareLoD("X", /*->*/ "Out");
  }

 protected:
  framework::OpKernelType GetExpectedKernelType(
      const framework::ExecutionContext& ctx) const override {
459 460
    return framework::OpKernelType(
        OperatorWithKernel::IndicateVarDataType(ctx, "X"), ctx.GetPlace());
Z
Zhen Wang 已提交
461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492
  }
};

class MovingAverageAbsMaxScaleOpMaker
    : public framework::OpProtoAndCheckerMaker {
 public:
  void Make() override {
    AddInput("X", "(Tensor) Input is float data type.");
    AddInput("InAccum", "Last accum.").AsDispensable();
    AddInput("InState", "Last state.").AsDispensable();
    AddOutput("Out",
              "(Tensor) Output tensor is just equivalent to the input tensor.");
    AddOutput("OutScale", " Current scale");
    AddOutput("OutState", "(Tensor) state buffer.").AsDispensable();
    AddOutput("OutAccum", "(Tensor) accum buffer.").AsDispensable();
    AddAttr<float>("moving_rate", "(float, default 0.9) moving rate.")
        .SetDefault(0.9);
    AddAttr<bool>("is_test",
                  "(bool, default false) Set true for inference only and false "
                  "for training. Some layers may run faster when this is true.")
        .SetDefault(false);
    AddComment(R"DOC(
MovingAverageAbsMaxScale operator is only used for calculating the quantization scale.
And it will not quantize the input tensor.

$$scale = (moving\_rate*accum+max(abs(x)))/(moving\_rate*state+1)$$
$$Out = X$$

)DOC");
  }
};

视言's avatar
视言 已提交
493 494 495 496
}  // namespace operators
}  // namespace paddle

namespace ops = paddle::operators;
497 498
using CPU = paddle::platform::CPUDeviceContext;

H
hong 已提交
499 500 501 502 503
REGISTER_OPERATOR(
    fake_quantize_abs_max, ops::FakeQuantizeAbsMaxOp,
    ops::FakeQuantizeAbsMaxOpMaker,
    paddle::framework::EmptyGradOpMaker<paddle::framework::OpDesc>,
    paddle::framework::EmptyGradOpMaker<paddle::imperative::OpBase>);
504 505
REGISTER_OP_CPU_KERNEL(fake_quantize_abs_max,
                       ops::FakeQuantizeAbsMaxKernel<CPU, float>);
视言's avatar
视言 已提交
506

H
hong 已提交
507 508 509 510 511
REGISTER_OPERATOR(
    fake_quantize_range_abs_max, ops::FakeQuantizeRangeAbsMaxOp,
    ops::FakeQuantizeRangeAbsMaxOpMaker,
    paddle::framework::EmptyGradOpMaker<paddle::framework::OpDesc>,
    paddle::framework::EmptyGradOpMaker<paddle::imperative::OpBase>);
512 513
REGISTER_OP_CPU_KERNEL(fake_quantize_range_abs_max,
                       ops::FakeQuantizeRangeAbsMaxKernel<CPU, float>);
Z
Zhen Wang 已提交
514

H
hong 已提交
515 516 517 518 519 520
REGISTER_OPERATOR(
    fake_quantize_moving_average_abs_max,
    ops::FakeQuantOrWithDequantMovingAverageAbsMaxOp,
    ops::FakeQuantOrWithDequantMovingAverageAbsMaxOpMaker,
    paddle::framework::EmptyGradOpMaker<paddle::framework::OpDesc>,
    paddle::framework::EmptyGradOpMaker<paddle::imperative::OpBase>);
521

522 523
REGISTER_OP_CPU_KERNEL(fake_quantize_moving_average_abs_max,
                       ops::FakeQuantizeMovingAverageAbsMaxKernel<CPU, float>);
524

H
hong 已提交
525 526 527 528 529 530
REGISTER_OPERATOR(
    fake_quantize_dequantize_moving_average_abs_max,
    ops::FakeQuantOrWithDequantMovingAverageAbsMaxOp,
    ops::FakeQuantOrWithDequantMovingAverageAbsMaxOpMaker,
    paddle::framework::EmptyGradOpMaker<paddle::framework::OpDesc>,
    paddle::framework::EmptyGradOpMaker<paddle::imperative::OpBase>);
531 532 533 534
REGISTER_OP_CPU_KERNEL(
    fake_quantize_dequantize_moving_average_abs_max,
    ops::FakeQuantizeDequantizeMovingAverageAbsMaxKernel<CPU, float>);

H
hong 已提交
535 536 537 538 539
REGISTER_OPERATOR(
    fake_channel_wise_quantize_abs_max, ops::FakeChannelWiseQuantizeAbsMaxOp,
    ops::FakeChannelWiseQuantizeAbsMaxOpMaker,
    paddle::framework::EmptyGradOpMaker<paddle::framework::OpDesc>,
    paddle::framework::EmptyGradOpMaker<paddle::imperative::OpBase>);
Z
Zhen Wang 已提交
540 541
REGISTER_OP_CPU_KERNEL(fake_channel_wise_quantize_abs_max,
                       ops::FakeChannelWiseQuantizeAbsMaxKernel<CPU, float>);
Z
Zhen Wang 已提交
542

H
hong 已提交
543 544 545 546 547
REGISTER_OPERATOR(
    moving_average_abs_max_scale, ops::MovingAverageAbsMaxScaleOp,
    ops::MovingAverageAbsMaxScaleOpMaker,
    paddle::framework::EmptyGradOpMaker<paddle::framework::OpDesc>,
    paddle::framework::EmptyGradOpMaker<paddle::imperative::OpBase>);
Z
Zhen Wang 已提交
548 549
REGISTER_OP_CPU_KERNEL(moving_average_abs_max_scale,
                       ops::MovingAverageAbsMaxScaleKernel<CPU, float>);