activation_op.h 30.9 KB
Newer Older
1
/* Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
L
Luo Tao 已提交
2 3 4 5 6 7 8 9 10
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. */
Q
qijun 已提交
11 12

#pragma once
D
dzhwinter 已提交
13 14 15
#include <glog/logging.h>
#include <string>
#include <unordered_set>
16 17
#include <utility>
#include <vector>
18

Y
Yi Wang 已提交
19 20 21
#include "paddle/fluid/framework/eigen.h"
#include "paddle/fluid/framework/op_registry.h"
#include "paddle/fluid/operators/detail/safe_ref.h"
22
#include "paddle/fluid/platform/float16.h"
Q
qijun 已提交
23

24 25 26 27
#ifdef PADDLE_WITH_MKLDNN
#include "paddle/fluid/platform/mkldnn_helper.h"
#endif

Q
qijun 已提交
28 29 30
namespace paddle {
namespace operators {

D
dzhwinter 已提交
31 32 33 34 35 36 37 38 39 40
/* Use ugly global variable, for the using in python layer side
   Please refer to the layer_helper.py and get the details.
 */
static std::unordered_set<std::string> InplaceOpSet = {
    "sigmoid", "exp",        "relu",  "tanh",      "sqrt",         "ceil",
    "floor",   "reciprocal", "relu6", "soft_relu", "hard_sigmoid",
};

static bool IsInplace(std::string op) { return InplaceOpSet.count(op); }

Q
QI JUN 已提交
41
template <typename DeviceContext, typename Functor>
42 43
class ActivationKernel
    : public framework::OpKernel<typename Functor::ELEMENT_TYPE> {
Q
qijun 已提交
44
 public:
45 46
  using T = typename Functor::ELEMENT_TYPE;

Q
qijun 已提交
47
  void Compute(const framework::ExecutionContext& context) const override {
Y
Update  
Yang Yu 已提交
48 49 50 51 52 53 54 55 56 57
    auto& X = detail::Ref(context.Input<framework::Tensor>("X"),
                          "Cannot get input tensor X, variable name = %s",
                          context.op().Input("X"));

    auto& Out = detail::Ref(context.Output<framework::Tensor>("Out"),
                            "Cannot get output tensor Out, variable name = %s",
                            context.op().Output("Out"));
    Out.mutable_data<T>(context.GetPlace());
    auto x = framework::EigenVector<T>::Flatten(X);
    auto out = framework::EigenVector<T>::Flatten(Out);
Q
QI JUN 已提交
58 59
    auto* place =
        context.template device_context<DeviceContext>().eigen_device();
Q
qijun 已提交
60
    Functor functor;
61 62 63 64 65

    auto attrs = functor.GetAttrs();
    for (auto& attr : attrs) {
      *attr.second = context.Attr<float>(attr.first);
    }
F
fengjiayi 已提交
66
    functor(*place, x, out);
Q
qijun 已提交
67 68 69
  }
};

Q
QI JUN 已提交
70
template <typename DeviceContext, typename Functor>
71 72
class ActivationGradKernel
    : public framework::OpKernel<typename Functor::ELEMENT_TYPE> {
Q
qijun 已提交
73
 public:
74
  using T = typename Functor::ELEMENT_TYPE;
Q
qijun 已提交
75
  void Compute(const framework::ExecutionContext& context) const override {
F
fengjiayi 已提交
76 77 78
    auto* Out = context.Input<framework::Tensor>("Out");
    auto* dOut =
        context.Input<framework::Tensor>(framework::GradVarName("Out"));
Q
qijun 已提交
79 80 81
    auto* dX = context.Output<framework::Tensor>(framework::GradVarName("X"));
    dX->mutable_data<T>(context.GetPlace());

F
fengjiayi 已提交
82 83
    auto dout = framework::EigenVector<T>::Flatten(*dOut);
    auto out = framework::EigenVector<T>::Flatten(*Out);
Q
qijun 已提交
84
    auto dx = framework::EigenVector<T>::Flatten(*dX);
Q
QI JUN 已提交
85 86
    auto* place =
        context.template device_context<DeviceContext>().eigen_device();
Q
qijun 已提交
87
    Functor functor;
88 89 90 91
    auto attrs = functor.GetAttrs();
    for (auto& attr : attrs) {
      *attr.second = context.Attr<float>(attr.first);
    }
D
dzhwinter 已提交
92 93 94 95 96 97
    bool inplace = functor.Inplace();
    if (!inplace) {
      auto* X = context.Input<framework::Tensor>("X");
      auto x = framework::EigenVector<T>::Flatten(*X);
      functor(*place, x, out, dout, dx);
    } else {
98
      VLOG(100) << " Inplace activation ";
D
dzhwinter 已提交
99 100 101
      auto x = framework::EigenVector<T>::Flatten(*dX);
      functor(*place, x, out, dout, dx);
    }
Q
qijun 已提交
102 103 104
  }
};

105 106 107 108 109 110 111
template <typename T>
struct BaseActivationFunctor {
  using ELEMENT_TYPE = T;

  using AttrPair = std::vector<std::pair<const char*, float*>>;

  AttrPair GetAttrs() { return AttrPair(); }
D
dzhwinter 已提交
112 113 114 115 116 117 118 119

  /* NOTE(*): Output reuse X memory if X is not dependented by its Gradient.
     For example, sigmoid op's gradient didn't involve x, so its output can
     reuse
     input memory. But abs op's gradient use x, it can not be inplaced.
     gradient did use x.
   */
  bool Inplace() const { return false; }
120 121
};

122
// sigmoid(x) = 1 / (1 + exp(-x))
Q
qijun 已提交
123
template <typename T>
124
struct SigmoidFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
125 126 127
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = static_cast<T>(1) / (static_cast<T>(1) + (-x).exp());
Q
qijun 已提交
128 129 130
  }
};

131
template <typename T>
132
struct SigmoidGradFunctor : public BaseActivationFunctor<T> {
D
dzhwinter 已提交
133
  bool Inplace() const { return IsInplace("sigmoid"); }
F
fengjiayi 已提交
134 135 136 137
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * out * (static_cast<T>(1) - out);
Q
qijun 已提交
138 139 140
  }
};

141 142 143 144
// Originally: logsigmoid(x) = -log (1 + exp(-x))
// For numerical stability, we can use the log-sum-exp trick:
// https://hips.seas.harvard.edu/blog/2013/01/09/computing-log-sum-exp/
// We can rewrite the above equation as:
F
fengjiayi 已提交
145
// out = -log( exp(0) + exp(-x)) [since exp(0) = 1]
146 147 148 149 150 151 152 153 154 155
//   = -log( exp(max(-x, 0) - max(-x, 0)) + exp(-x + max(-x, 0) - max(-x, 0)))
//   = -log( exp(max(-x, 0)) * exp(-max(-x, 0)) - exp(max(-x, 0)) * exp(-x -
//           max(-x, 0)))
//   = -log( exp(max(-x, 0)) * (exp(-max(-x, 0)) + exp(-x - max(-x, 0))))
//   = -log( exp(max(-x, 0)) - log(exp(-max(-x, 0)) + exp(-x - max(-x, 0)))
//
// Hence, logsigmoid(x) = - (max(-x, 0) + log(exp(-max(-x, 0))
// + exp(-x - max(-x, 0))))
template <typename T>
struct LogSigmoidFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
156 157
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
158
    auto temp = (-x).cwiseMax(static_cast<T>(0));  // temp = max(-x, 0)
F
fengjiayi 已提交
159
    out.device(d) = -temp - (((-temp).exp() + (-x - temp).exp()).log());
160 161 162 163 164 165 166 167
  }
};

// Originally: f' = exp(-x) / (1 + exp(-x))
// For numerical stability: f' = exp(-x - max(-x, 0)) / (exp(-max(-x, 0)) +
// exp(-x - max(-x, 0)))
template <typename T>
struct LogSigmoidGradFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
168 169 170
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
171 172
    auto temp = (-x).cwiseMax(static_cast<T>(0));  // temp = max(-x, 0)
    dx.device(d) =
F
fengjiayi 已提交
173
        dout * ((-x - temp).exp() / ((-temp).exp() + (-x - temp).exp()));
174 175 176
  }
};

Q
qijun 已提交
177
// exp(x) = e^x
178 179
template <typename T>
struct ExpFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
180 181 182
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.exp();
Q
qijun 已提交
183 184 185
  }
};

186 187
template <typename T>
struct ExpGradFunctor : public BaseActivationFunctor<T> {
D
dzhwinter 已提交
188
  bool Inplace() const { return IsInplace("exp"); }
F
fengjiayi 已提交
189 190 191 192
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * out;
Q
qijun 已提交
193 194 195
  }
};

Q
qijun 已提交
196
// relu(x) = max(x, 0)
Q
qijun 已提交
197
template <typename T>
198
struct ReluFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
199 200 201
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.cwiseMax(static_cast<T>(0));
Q
qijun 已提交
202 203
  }
};
Q
qijun 已提交
204

Q
qijun 已提交
205
template <typename T>
206
struct ReluGradFunctor : public BaseActivationFunctor<T> {
D
dzhwinter 已提交
207
  bool Inplace() const { return IsInplace("relu"); }
F
fengjiayi 已提交
208 209 210
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
D
dzhwinter 已提交
211
    dx.device(d) = dout * (out > static_cast<T>(0)).template cast<T>();
Q
qijun 已提交
212 213
  }
};
Q
qijun 已提交
214

215
// tanh(x) = (exp(x) - exp(-x)) / (exp(x) + exp(-x))
216 217
template <typename T>
struct TanhFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
218 219 220
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.tanh();
Q
qijun 已提交
221 222 223 224
  }
};

template <typename T>
225
struct TanhGradFunctor : public BaseActivationFunctor<T> {
D
dzhwinter 已提交
226
  bool Inplace() const { return IsInplace("tanh"); }
F
fengjiayi 已提交
227 228 229 230
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * (static_cast<T>(1) - out * out);
Q
qijun 已提交
231 232 233
  }
};

K
Kavya Srinet 已提交
234 235 236 237
// tanhshrink(x) = x - tanh(x)
// where tanh(x) = (exp(x) - exp(-x)) / (exp(x) + exp(-x))
template <typename T>
struct TanhShrinkFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
238 239 240
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x - x.tanh();
K
Kavya Srinet 已提交
241 242 243 244 245
  }
};

template <typename T>
struct TanhShrinkGradFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
246 247 248 249
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * (x.tanh() * x.tanh());
K
Kavya Srinet 已提交
250 251 252
  }
};

253 254 255 256 257 258 259 260 261
// tanhshrink(x) = x - tanh(x)
// where tanh(x) = (exp(x) - exp(-x)) / (exp(x) + exp(-x))
template <typename T>
struct HardShrinkFunctor : public BaseActivationFunctor<T> {
  float threshold;

  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }
F
fengjiayi 已提交
262 263
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
Y
Yu Yang 已提交
264 265
    auto temp1 = (x < static_cast<T>(threshold * -1)).template cast<T>().eval();
    auto temp2 = (x > static_cast<T>(threshold)).template cast<T>().eval();
F
fengjiayi 已提交
266
    out.device(d) = x * (temp1 + temp2);
267 268 269 270 271 272 273 274 275 276 277
  }
};

template <typename T>
struct HardShrinkGradFunctor : public BaseActivationFunctor<T> {
  float threshold;

  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }

F
fengjiayi 已提交
278 279 280
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
Y
Yu Yang 已提交
281 282
    auto temp1 = (x < static_cast<T>(threshold * -1)).template cast<T>().eval();
    auto temp2 = (x > static_cast<T>(threshold)).template cast<T>().eval();
F
fengjiayi 已提交
283
    dx.device(d) = dout * (temp1 + temp2).template cast<T>();
284 285 286
  }
};

K
Kexin Zhao 已提交
287
// softshrink(x) = x - lambda, if x > lambda; x + lambda, if x < -lambda; 0
288 289 290 291 292 293 294 295
// otherwise
template <typename T>
struct SoftShrinkFunctor : public BaseActivationFunctor<T> {
  float lambda;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"lambda", &lambda}};
  }

F
fengjiayi 已提交
296 297
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
Y
Yu Yang 已提交
298 299 300
    auto lambdaT = static_cast<T>(lambda);
    auto temp1 = (x > lambdaT).template cast<T>().eval();
    auto temp2 = (x < -lambdaT).template cast<T>().eval();
F
fengjiayi 已提交
301
    out.device(d) = temp1 * (x - lambdaT) + temp2 * (x + lambdaT);
302 303 304 305 306 307 308 309 310
  }
};

template <typename T>
struct SoftShrinkGradFunctor : public BaseActivationFunctor<T> {
  float lambda;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"lambda", &lambda}};
  }
F
fengjiayi 已提交
311 312 313
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
Y
Yu Yang 已提交
314 315 316
    auto lambdaT = static_cast<T>(lambda);
    auto temp1 = (x > lambdaT).template cast<T>().eval();
    auto temp2 = (x < -lambdaT).template cast<T>().eval();
F
fengjiayi 已提交
317
    dx.device(d) = dout * (temp1 + temp2).template cast<T>();
318 319 320
  }
};

Q
qijun 已提交
321
// sqrt(x) = x^(1/2)
322 323
template <typename T>
struct SqrtFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
324 325 326
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.sqrt();
Q
qijun 已提交
327 328 329 330
  }
};

template <typename T>
331
struct SqrtGradFunctor : public BaseActivationFunctor<T> {
D
dzhwinter 已提交
332
  bool Inplace() const { return IsInplace("sqrt"); }
F
fengjiayi 已提交
333 334 335
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
C
chengduo 已提交
336
    dx.device(d) = static_cast<T>(0.5) * dout / out;
Q
qijun 已提交
337 338 339
  }
};

D
dzhwinter 已提交
340 341 342
// ceil(x) = ceiling(x)
template <typename T>
struct CeilFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
343 344 345
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.ceil();
D
dzhwinter 已提交
346 347 348 349 350
  }
};

template <typename T>
struct ZeroGradFunctor : public BaseActivationFunctor<T> {
D
dzhwinter 已提交
351
  bool Inplace() const { return IsInplace("ceil"); }
F
fengjiayi 已提交
352 353 354
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
D
dzhwinter 已提交
355
    dx.device(d) = static_cast<T>(0) / out;
D
dzhwinter 已提交
356 357 358 359 360 361
  }
};

// floor(x) = flooring(x)
template <typename T>
struct FloorFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
362 363
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
Q
Qiao Longfei 已提交
364
    out.device(d) = x.floor();
D
dzhwinter 已提交
365 366 367
  }
};

C
add cos  
chengduoZH 已提交
368 369 370 371 372
template <typename T>
struct Sine {
  HOSTDEVICE T operator()(const T& val) const { return sin(val); }
};

373 374 375 376 377 378 379
template <>
struct Sine<platform::float16> {
  HOSTDEVICE platform::float16 operator()(const platform::float16& val) const {
    return platform::float16(sin(static_cast<float>(val)));
  }
};

C
add cos  
chengduoZH 已提交
380 381 382 383 384
template <typename T>
struct Cosine {
  HOSTDEVICE T operator()(const T& val) const { return cos(val); }
};

385 386 387 388 389 390 391
template <>
struct Cosine<platform::float16> {
  HOSTDEVICE platform::float16 operator()(const platform::float16& val) const {
    return platform::float16(cos(static_cast<float>(val)));
  }
};

C
add cos  
chengduoZH 已提交
392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
// cosine'(x) = -sin(x)
template <typename T>
struct CosGradFunctor : public BaseActivationFunctor<T> {
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = -dout * x.unaryExpr(Sine<T>());
  }
};

// cosine(x) = cos(x)
template <typename T>
struct CosFunctor : public BaseActivationFunctor<T> {
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.unaryExpr(Cosine<T>());
  }
};

// sine'(x) = cos(x)
template <typename T>
struct SinGradFunctor : public BaseActivationFunctor<T> {
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * x.unaryExpr(Cosine<T>());
  }
};

// sine(x) = sin(x)
template <typename T>
struct SinFunctor : public BaseActivationFunctor<T> {
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.unaryExpr(Sine<T>());
  }
};

D
dzhwinter 已提交
430 431 432
// round(x) = [x]
template <typename T>
struct RoundFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
433 434 435
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.round();
D
dzhwinter 已提交
436 437 438
  }
};

Q
qijun 已提交
439
// abs(x) = |x|
440 441
template <typename T>
struct AbsFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
442 443 444
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.abs();
Q
qijun 已提交
445 446 447
  }
};

448 449
template <typename T>
struct AbsGradFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
450 451 452 453
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * x.sign();
454 455 456
  }
};

Q
qijun 已提交
457 458
// reciprocal(x) = 1 / x
template <typename T>
459
struct ReciprocalFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
460 461 462
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = static_cast<T>(1) / x;
Q
qijun 已提交
463 464 465
  }
};

466
template <typename T>
467
struct ReciprocalGradFunctor : public BaseActivationFunctor<T> {
D
dzhwinter 已提交
468
  bool Inplace() const { return IsInplace("reciprocal"); }
F
fengjiayi 已提交
469 470 471 472
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * static_cast<T>(-1) * out * out;
Q
qijun 已提交
473 474 475 476
  }
};

// log(x) = natural logarithm of x
477 478
template <typename T>
struct LogFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
479 480 481
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.log();
Q
qijun 已提交
482 483 484
  }
};

485
template <typename T>
486
struct LogGradFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
487 488 489 490
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * (static_cast<T>(1) / x);
Q
qijun 已提交
491 492 493 494
  }
};

// square(x) = x^2
495 496
template <typename T>
struct SquareFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
497 498 499
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.square();
Q
qijun 已提交
500
  }
501
};
Q
qijun 已提交
502

503
template <typename T>
504
struct SquareGradFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
505 506 507 508
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * static_cast<T>(2) * x;
509 510 511
  }
};

512 513 514 515 516 517 518 519 520 521
template <typename T>
struct BReluFunctor : public BaseActivationFunctor<T> {
  float t_min;
  float t_max;

  // NOTE: Explicit hides the `BaseActivationFunctor<T>::GetAttrs`
  // not polymorphism for speed.
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"t_min", &t_min}, {"t_max", &t_max}};
  }
522

F
fengjiayi 已提交
523 524 525
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) =
Y
Yu Yang 已提交
526
        x.cwiseMax(static_cast<T>(t_min)).cwiseMin(static_cast<T>(t_max));
527 528 529
  }
};

530 531 532 533 534 535 536
template <typename T>
struct BReluGradFunctor : public BaseActivationFunctor<T> {
  float t_min;
  float t_max;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"t_min", &t_min}, {"t_max", &t_max}};
  }
F
fengjiayi 已提交
537 538 539 540
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout *
Y
Yu Yang 已提交
541 542
                   ((x > static_cast<T>(t_min)) * (x < static_cast<T>(t_max)))
                       .template cast<T>();
543 544 545
  }
};

546 547 548 549 550 551 552 553 554
// relu6(x) = min(max(0, x), 6)
template <typename T>
struct Relu6Functor : public BaseActivationFunctor<T> {
  float threshold;

  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }

F
fengjiayi 已提交
555 556 557
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) =
Y
Yu Yang 已提交
558
        x.cwiseMax(static_cast<T>(0)).cwiseMin(static_cast<T>(threshold));
559 560 561 562 563 564 565 566 567
  }
};

template <typename T>
struct Relu6GradFunctor : public BaseActivationFunctor<T> {
  float threshold;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }
D
dzhwinter 已提交
568
  bool Inplace() const { return IsInplace("relu6"); }
F
fengjiayi 已提交
569 570 571
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
D
dzhwinter 已提交
572 573 574 575
    dx.device(d) =
        dout *
        ((out > static_cast<T>(0)) * (out < static_cast<T>(threshold)))
            .template cast<T>();
576 577 578
  }
};

K
kexinzhao 已提交
579 580 581 582 583 584 585
// softplus(x) = log(1 + exp(x))
// When x is a very large positive number, exp(x) may explode to inf,
// Using trick below for numerical stability
// https://hips.seas.harvard.edu/blog/2013/01/09/computing-log-sum-exp/
// Then: softplus(x) = max(x, 0) + log(exp(-max(x, 0)) + exp(x - max(x, 0)))
template <typename T>
struct SoftplusFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
586 587
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) {
K
kexinzhao 已提交
588
    auto temp = x.cwiseMax(static_cast<T>(0));  // temp = max(x, 0)
F
fengjiayi 已提交
589
    out.device(d) = temp + (((-temp).exp() + (x - temp).exp()).log());
K
kexinzhao 已提交
590 591 592 593 594 595 596 597 598
  }
};

// d(softplus(x))/dx = exp(x) / (1 + exp(x))
// For numerical stability:
// d(softplus(x))/dx = exp(x - max(x, 0)) / (exp(-max(x, 0)) +
// exp(x - max(x, 0)))
template <typename T>
struct SoftplusGradFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
599 600 601
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) {
K
kexinzhao 已提交
602
    auto temp = x.cwiseMax(static_cast<T>(0));  // temp = max(x, 0)
F
fengjiayi 已提交
603 604
    dx.device(d) =
        dout * ((x - temp).exp() / ((-temp).exp() + (x - temp).exp()));
K
kexinzhao 已提交
605 606 607
  }
};

608 609
// softsign(x) = x / (1 + |x|)
template <typename T>
610
struct SoftsignFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
611 612 613
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) {
    out.device(d) = x / (static_cast<T>(1) + x.abs());
614 615 616 617 618 619
  }
};

// d(softsign(x))/dx = 1 / (1 + |x|)^2
// Taken from https://en.wikipedia.org/wiki/Activation_function
template <typename T>
620
struct SoftsignGradFunctor : public BaseActivationFunctor<T> {
F
fengjiayi 已提交
621 622 623
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) {
624
    dx.device(d) =
F
fengjiayi 已提交
625
        dout * (static_cast<T>(1) / (static_cast<T>(1) + x.abs()).square());
626 627 628
  }
};

629 630 631 632 633 634
template <typename T>
struct SoftReluFunctor : public BaseActivationFunctor<T> {
  float threshold;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }
635

F
fengjiayi 已提交
636 637
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
Y
Yu Yang 已提交
638 639
    auto tmp = static_cast<T>(threshold);
    auto temp = x.cwiseMax(-tmp).cwiseMin(tmp);
F
fengjiayi 已提交
640
    out.device(d) = (static_cast<T>(1) + temp.exp()).log();
641 642 643
  }
};

644 645 646 647 648 649
template <typename T>
struct SoftReluGradFunctor : public BaseActivationFunctor<T> {
  float threshold;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }
D
dzhwinter 已提交
650
  bool Inplace() const { return IsInplace("soft_relu"); }
F
fengjiayi 已提交
651 652 653
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
Y
Yu Yang 已提交
654
    auto tmp = static_cast<T>(threshold);
D
dzhwinter 已提交
655
    auto temp = ((out > -tmp) * (out < tmp)).template cast<T>().eval();
F
fengjiayi 已提交
656
    dx.device(d) = dout * (static_cast<T>(1) - (-out).exp()) * temp;
657 658 659
  }
};

K
Kavya Srinet 已提交
660 661 662 663 664 665
template <typename T>
struct LeakyReluFunctor : public BaseActivationFunctor<T> {
  float alpha;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"alpha", &alpha}};
  }
666

F
fengjiayi 已提交
667 668 669
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.cwiseMax(static_cast<T>(alpha) * x);
670 671 672
  }
};

K
Kavya Srinet 已提交
673 674 675 676 677 678
template <typename T>
struct LeakyReluGradFunctor : public BaseActivationFunctor<T> {
  float alpha;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"alpha", &alpha}};
  }
F
fengjiayi 已提交
679 680 681
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
Y
Yu Yang 已提交
682 683
    auto temp1 = static_cast<T>(alpha) *
                 (x < static_cast<T>(0)).template cast<T>().eval();
K
Kavya Srinet 已提交
684
    auto temp2 = (x >= static_cast<T>(0)).template cast<T>().eval();
F
fengjiayi 已提交
685
    dx.device(d) = dout * (temp1 + temp2).template cast<T>();
686 687 688
  }
};

689 690 691 692 693 694
template <typename T>
struct ELUFunctor : public BaseActivationFunctor<T> {
  float alpha;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"alpha", &alpha}};
  }
695

F
fengjiayi 已提交
696 697 698 699 700
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.cwiseMax(static_cast<T>(0)) +
                    (static_cast<T>(alpha) * (x.exp() - static_cast<T>(1)))
                        .cwiseMin(static_cast<T>(0));
701 702 703
  }
};

704 705 706 707 708 709
template <typename T>
struct ELUGradFunctor : public BaseActivationFunctor<T> {
  float alpha;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"alpha", &alpha}};
  }
F
fengjiayi 已提交
710 711 712 713 714
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * (x > static_cast<T>(0)).template cast<T>() +
                   dout * (out + static_cast<T>(alpha)) *
Y
Yu Yang 已提交
715
                       (x < static_cast<T>(0)).template cast<T>();
716 717 718
  }
};

Q
QI JUN 已提交
719
// FIXME(qijun) https://github.com/PaddlePaddle/Paddle/issues/5198
720 721 722 723 724 725
template <typename T>
struct PowFunctor : public BaseActivationFunctor<T> {
  float factor;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"factor", &factor}};
  }
F
fengjiayi 已提交
726 727 728
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x.pow(static_cast<T>(factor));
729 730 731
  }
};

732 733 734 735 736 737
template <typename T>
struct PowGradFunctor : public BaseActivationFunctor<T> {
  float factor;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"factor", &factor}};
  }
F
fengjiayi 已提交
738 739 740 741
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout * static_cast<T>(factor) *
C
chengduo 已提交
742
                   x.pow(static_cast<T>(factor) - static_cast<T>(1));
743 744 745
  }
};

746 747 748 749 750 751 752
template <typename T>
struct STanhFunctor : public BaseActivationFunctor<T> {
  float scale_a;
  float scale_b;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"scale_a", &scale_a}, {"scale_b", &scale_b}};
  }
753

F
fengjiayi 已提交
754 755 756
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) =
Y
Yu Yang 已提交
757
        static_cast<T>(scale_b) * (static_cast<T>(scale_a) * x).tanh();
758 759 760
  }
};

761 762 763 764 765 766 767
template <typename T>
struct STanhGradFunctor : public BaseActivationFunctor<T> {
  float scale_a;
  float scale_b;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"scale_a", &scale_a}, {"scale_b", &scale_b}};
  }
768

F
fengjiayi 已提交
769 770 771
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
Y
Yu Yang 已提交
772 773 774
    auto a = static_cast<T>(scale_a);
    auto b = static_cast<T>(scale_b);
    auto temp = (a * x).tanh() * (a * x).tanh();
F
fengjiayi 已提交
775
    dx.device(d) = dout * a * b * (static_cast<T>(1) - temp);
Q
qijun 已提交
776 777 778
  }
};

779 780 781 782 783 784 785
template <typename T>
struct ThresholdedReluFunctor : public BaseActivationFunctor<T> {
  float threshold;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }

F
fengjiayi 已提交
786 787
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
Y
Yu Yang 已提交
788
    auto th = static_cast<T>(threshold);
F
fengjiayi 已提交
789
    out.device(d) = (x > th).template cast<T>() * x;
790 791 792 793 794 795 796 797 798 799
  }
};

template <typename T>
struct ThresholdedReluGradFunctor : public BaseActivationFunctor<T> {
  float threshold;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"threshold", &threshold}};
  }

F
fengjiayi 已提交
800 801 802
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
Y
Yu Yang 已提交
803
    auto th = static_cast<T>(threshold);
F
fengjiayi 已提交
804
    dx.device(d) = dout * (x > th).template cast<T>();
805 806 807
  }
};

808 809 810 811 812 813 814 815
template <typename T>
struct HardSigmoidFunctor : public BaseActivationFunctor<T> {
  float slope;
  float offset;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"slope", &slope}, {"offset", &offset}};
  }

F
fengjiayi 已提交
816 817
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
818
    auto temp = x * static_cast<T>(slope) + static_cast<T>(offset);
F
fengjiayi 已提交
819 820
    out.device(d) =
        temp.cwiseMax(static_cast<T>(0)).cwiseMin(static_cast<T>(1));
821 822 823 824 825 826 827 828 829 830
  }
};

template <typename T>
struct HardSigmoidGradFunctor : public BaseActivationFunctor<T> {
  float slope;
  float offset;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"slope", &slope}, {"offset", &offset}};
  }
D
dzhwinter 已提交
831
  bool Inplace() { return IsInplace("hard_sigmoid"); }
F
fengjiayi 已提交
832 833 834 835 836 837 838
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
    dx.device(d) = dout *
                   ((out > static_cast<T>(0)) * (out < static_cast<T>(1)))
                       .template cast<T>() *
                   static_cast<T>(slope);
839 840 841
  }
};

A
Abhinav Arora 已提交
842 843 844 845 846 847 848
template <typename T>
struct SwishFunctor : public BaseActivationFunctor<T> {
  float beta;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"beta", &beta}};
  }

F
fengjiayi 已提交
849 850 851
  template <typename Device, typename X, typename Out>
  void operator()(Device d, X x, Out out) const {
    out.device(d) = x / (static_cast<T>(1) + (static_cast<T>(-beta) * x).exp());
A
Abhinav Arora 已提交
852 853 854 855 856 857 858 859 860 861
  }
};

template <typename T>
struct SwishGradFunctor : public BaseActivationFunctor<T> {
  float beta;
  typename BaseActivationFunctor<T>::AttrPair GetAttrs() {
    return {{"beta", &beta}};
  }

F
fengjiayi 已提交
862 863 864
  template <typename Device, typename X, typename Out, typename dOut,
            typename dX>
  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
A
Abhinav Arora 已提交
865
    auto temp1 = static_cast<T>(1) /
866
                 (static_cast<T>(1) + (static_cast<T>(-beta) * x).exp());
D
dzhwinter 已提交
867 868
    auto temp2 = temp1 * (static_cast<T>(1) - (static_cast<T>(beta) * out));
    dx.device(d) = dout * ((static_cast<T>(beta) * out) + temp2);
A
Abhinav Arora 已提交
869 870 871
  }
};

Q
qijun 已提交
872 873
}  // namespace operators
}  // namespace paddle
874

875 876 877 878
#define FOR_EACH_KERNEL_FUNCTOR(__macro)                             \
  __macro(sigmoid, SigmoidFunctor, SigmoidGradFunctor);              \
  __macro(logsigmoid, LogSigmoidFunctor, LogSigmoidGradFunctor);     \
  __macro(exp, ExpFunctor, ExpGradFunctor);                          \
879
  __macro(relu, ReluFunctor, ReluGradFunctor);                       \
880 881 882 883
  __macro(tanh, TanhFunctor, TanhGradFunctor);                       \
  __macro(softshrink, SoftShrinkFunctor, SoftShrinkGradFunctor);     \
  __macro(sqrt, SqrtFunctor, SqrtGradFunctor);                       \
  __macro(abs, AbsFunctor, AbsGradFunctor);                          \
D
dzhwinter 已提交
884 885
  __macro(ceil, CeilFunctor, ZeroGradFunctor);                       \
  __macro(floor, FloorFunctor, ZeroGradFunctor);                     \
C
add cos  
chengduoZH 已提交
886
  __macro(cos, CosFunctor, CosGradFunctor);                          \
C
add sin  
chengduoZH 已提交
887
  __macro(sin, SinFunctor, SinGradFunctor);                          \
D
dzhwinter 已提交
888
  __macro(round, RoundFunctor, ZeroGradFunctor);                     \
889 890 891 892 893 894 895 896 897 898 899 900 901 902 903
  __macro(reciprocal, ReciprocalFunctor, ReciprocalGradFunctor);     \
  __macro(log, LogFunctor, LogGradFunctor);                          \
  __macro(square, SquareFunctor, SquareGradFunctor);                 \
  __macro(brelu, BReluFunctor, BReluGradFunctor);                    \
  __macro(soft_relu, SoftReluFunctor, SoftReluGradFunctor);          \
  __macro(pow, PowFunctor, PowGradFunctor);                          \
  __macro(stanh, STanhFunctor, STanhGradFunctor);                    \
  __macro(softplus, SoftplusFunctor, SoftplusGradFunctor);           \
  __macro(softsign, SoftsignFunctor, SoftsignGradFunctor);           \
  __macro(relu6, Relu6Functor, Relu6GradFunctor);                    \
  __macro(leaky_relu, LeakyReluFunctor, LeakyReluGradFunctor);       \
  __macro(tanh_shrink, TanhShrinkFunctor, TanhShrinkGradFunctor);    \
  __macro(elu, ELUFunctor, ELUGradFunctor);                          \
  __macro(hard_shrink, HardShrinkFunctor, HardShrinkGradFunctor);    \
  __macro(hard_sigmoid, HardSigmoidFunctor, HardSigmoidGradFunctor); \
A
Abhinav Arora 已提交
904
  __macro(swish, SwishFunctor, SwishGradFunctor);                    \
905
  __macro(thresholded_relu, ThresholdedReluFunctor, ThresholdedReluGradFunctor);