activation_op.kps 40.3 KB
Newer Older
1
/* Copyright (c) 2016 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. */
L
liaogang 已提交
11

Y
Yi Wang 已提交
12
#include "paddle/fluid/operators/activation_op.h"
13 14
#include "paddle/fluid/operators/amp/fp16_type_traits.h"
#include "paddle/fluid/operators/elementwise/elementwise_op_impl.cu.h"
15
#include "paddle/fluid/platform/bfloat16.h"
16
#include "paddle/fluid/platform/device/gpu/gpu_device_function.h"
17

18 19
#include "paddle/phi/kernels/funcs/activation_functor.h"

20 21 22
namespace paddle {
namespace operators {

23 24 25 26 27
template <typename T>
struct CudaReciprocalFunctor : public BaseActivationFunctor<T> {
  T one = static_cast<T>(1.0f);

  // reciprocal(x) = 1 / x
28
  __device__ __forceinline__ T operator()(const T x) const { return one / x; }
29
};
30

31
template <typename T>
32 33
struct CudaReciprocalGradFunctor : public BaseActivationFunctor<T> {
  // dx = -dout * out^2
34
  __device__ __forceinline__ T operator()(const T dout, const T out) const {
35
    return -dout * out * out;
36
  }
37

38 39 40
  static constexpr ActBwdOpFwdDeps FwdDeps() {
    return ActBwdOpFwdDeps::kDepOut;
  }
41
};
42

43 44 45 46 47
template <typename T>
struct CudaExpFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;

  // exp(x) = exp(x)
48
  __device__ __forceinline__ T operator()(const T arg_x) const {
49
    MPType x = static_cast<MPType>(arg_x);
50 51 52
    return static_cast<T>(exp(x));
  }
};
53 54

template <typename T>
55 56
struct CudaExpGradFunctor : public BaseActivationFunctor<T> {
  // dx = dout * out
57
  __device__ __forceinline__ T operator()(const T dout, const T out) const {
58
    return dout * out;
59
  }
60

61 62 63
  static constexpr ActBwdOpFwdDeps FwdDeps() {
    return ActBwdOpFwdDeps::kDepOut;
  }
64
};
65

R
ronnywang 已提交
66 67 68 69 70
template <typename T>
struct CudaExpm1Functor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;

  // expm1(x) = expm1(x)
71
  __device__ __forceinline__ T operator()(const T arg_x) const {
72
    MPType x = static_cast<MPType>(arg_x);
R
ronnywang 已提交
73 74 75 76 77 78 79
    return static_cast<T>(expm1(x));
  }
};

template <typename T>
struct CudaExpm1GradFunctor : public BaseActivationFunctor<T> {
  // dx = dout * out
80
  __device__ __forceinline__ T operator()(const T dout, const T out) const {
81
    return dout * out + dout;
R
ronnywang 已提交
82 83
  }

84 85 86
  static constexpr ActBwdOpFwdDeps FwdDeps() {
    return ActBwdOpFwdDeps::kDepOut;
  }
R
ronnywang 已提交
87 88
};

89 90 91
template <typename T>
struct CudaSquareFunctor : public BaseActivationFunctor<T> {
  // square(x) = x * x
92
  __device__ __forceinline__ T operator()(const T x) const { return x * x; }
93
};
94

95 96 97 98 99
template <typename T>
struct CudaSquareGradFunctor : public BaseActivationFunctor<T> {
  T two = static_cast<T>(2.0f);

  // dx = dout * 2 * x
100
  __device__ __forceinline__ T operator()(const T dout, const T x) const {
101
    return dout * two * x;
102 103
  }

104
  static constexpr ActBwdOpFwdDeps FwdDeps() { return ActBwdOpFwdDeps::kDepX; }
105 106
};

107 108 109 110 111
template <typename T>
struct CudaSqrtFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;

  // sqrt(x) = sqrt(x)
112
  __device__ __forceinline__ T operator()(const T arg_x) const {
113
    MPType x = static_cast<MPType>(arg_x);
114 115 116
    return static_cast<T>(sqrt(x));
  }
};
117

118 119 120 121 122
template <typename T>
struct CudaSqrtGradFunctor : public BaseActivationFunctor<T> {
  T one_half = static_cast<T>(0.5f);

  // dx = dout * 0.5 / out
123
  __device__ __forceinline__ T operator()(const T dout, const T out) const {
124
    return one_half * dout / out;
125 126
  }

127 128 129
  static constexpr ActBwdOpFwdDeps FwdDeps() {
    return ActBwdOpFwdDeps::kDepOut;
  }
130
};
131

132 133 134 135 136
template <typename T>
struct CudaRsqrtFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;

  // rsqrt(x) = rsqrt(x)
137
  __device__ __forceinline__ T operator()(const T arg_x) const {
138
    MPType x = static_cast<MPType>(arg_x);
139 140 141 142 143 144 145 146
    return static_cast<T>(rsqrt(x));
  }
};

template <typename T>
struct CudaRsqrtGradFunctor : public BaseActivationFunctor<T> {
  T minus_one_half = static_cast<T>(-0.5f);

147
  // dx = -0.5 * dout * out^3
148
  __device__ __forceinline__ T operator()(const T dout, const T out) const {
149
    return minus_one_half * dout * out * out * out;
150 151
  }

152 153 154
  static constexpr ActBwdOpFwdDeps FwdDeps() {
    return ActBwdOpFwdDeps::kDepOut;
  }
155
};
156

157 158 159 160 161 162 163 164 165 166 167 168
template <typename T>
struct CudaSoftReluFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType one = static_cast<MPType>(1.0f);
  float threshold;

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

  // soft_relu(x) = log(1 + exp(max(min(x, threshold), -threshold)))
  // threshold should not be negative
169
  __device__ __forceinline__ T operator()(const T arg_x) const {
170
    MPType x = static_cast<MPType>(arg_x);
171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189
    MPType t = static_cast<MPType>(threshold);
    MPType temp_min = x < t ? x : t;
    MPType temp_max = temp_min > -t ? temp_min : -t;
    return static_cast<T>(log(one + exp(temp_max)));
  }
};

template <typename T>
struct CudaSoftReluGradFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType one = static_cast<MPType>(1.0f);
  float threshold;

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

  // dx = (out > -threshold && out < threshold) ? dout * (1 - exp(-out)) : 0
  // threshold should not be negative
190 191
  __device__ __forceinline__ T operator()(const T arg_dout,
                                          const T arg_out) const {
192 193
    MPType dout = static_cast<MPType>(arg_dout);
    MPType out = static_cast<MPType>(arg_out);
194 195 196 197 198
    MPType t = static_cast<MPType>(threshold);
    return (out > -t && out < t) ? static_cast<T>(dout * (one - exp(-out)))
                                 : static_cast<T>(0.0f);
  }

199 200 201
  static constexpr ActBwdOpFwdDeps FwdDeps() {
    return ActBwdOpFwdDeps::kDepOut;
  }
202 203 204 205 206 207 208 209 210 211 212 213 214
};

template <typename T>
struct CudaSTanhFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  float scale_a;
  float scale_b;

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

  // stanh(x) = b * tanh(a * x)
215
  __device__ __forceinline__ T operator()(const T arg_x) const {
216
    MPType x = static_cast<MPType>(arg_x);
217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234
    MPType a = static_cast<MPType>(scale_a);
    MPType b = static_cast<MPType>(scale_b);
    return static_cast<T>(b * tanh(a * x));
  }
};

template <typename T>
struct CudaSTanhGradFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType one = static_cast<MPType>(1.0f);
  float scale_a;
  float scale_b;

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

  // dx = dout * a * b * (1 - tanh(a * x) * tanh(a * x))
235 236
  __device__ __forceinline__ T operator()(const T arg_dout,
                                          const T arg_x) const {
237 238
    MPType dout = static_cast<MPType>(arg_dout);
    MPType x = static_cast<MPType>(arg_x);
239 240 241 242 243 244
    MPType a = static_cast<MPType>(scale_a);
    MPType b = static_cast<MPType>(scale_b);
    MPType temp = tanh(a * x);
    return static_cast<T>(dout * a * b * (one - temp * temp));
  }

245
  static constexpr ActBwdOpFwdDeps FwdDeps() { return ActBwdOpFwdDeps::kDepX; }
246 247 248 249 250 251 252 253 254 255 256 257 258 259
};

template <typename T>
struct CudaSoftplusFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType one = static_cast<MPType>(1.0f);
  float beta;
  float threshold;

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

  // softplus(x) = beta * x > threshold ? x : log(1 + exp(beta * x)) / beta
260
  __device__ __forceinline__ T operator()(const T arg_x) const {
261
    MPType x = static_cast<MPType>(arg_x);
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280
    MPType b = static_cast<MPType>(beta);
    MPType t = static_cast<MPType>(threshold);
    MPType x_beta = x * beta;
    return static_cast<T>(x_beta > t ? x : log(one + exp(x_beta)) / b);
  }
};

template <typename T>
struct CudaSoftplusGradFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType one = static_cast<MPType>(1.0f);
  float beta;
  float threshold;

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

  // dx = x * beta > threshold ? dout : dout / (1 + exp(-beta * x))
281 282
  __device__ __forceinline__ T operator()(const T arg_dout,
                                          const T arg_x) const {
283 284
    MPType dout = static_cast<MPType>(arg_dout);
    MPType x = static_cast<MPType>(arg_x);
285 286 287
    MPType b = static_cast<MPType>(beta);
    MPType t = static_cast<MPType>(threshold);
    MPType x_beta = x * beta;
288
    return x_beta > t ? arg_dout : static_cast<T>(dout / (one + exp(-x_beta)));
289 290
  }

291
  static constexpr ActBwdOpFwdDeps FwdDeps() { return ActBwdOpFwdDeps::kDepX; }
292 293 294 295 296 297 298
};

template <typename T>
struct CudaSoftsignFunctor : public BaseActivationFunctor<T> {
  T one = static_cast<T>(1.0f);

  // softsign(x) = x / (1 + abs(x))
299
  __device__ __forceinline__ T operator()(const T x) const {
300
    return x / (one + abs(x));
301 302 303 304 305 306 307 308
  }
};

template <typename T>
struct CudaSoftsignGradFunctor : public BaseActivationFunctor<T> {
  T one = static_cast<T>(1.0f);

  // dx = dout / (1 + abs(x))^2
309
  __device__ __forceinline__ T operator()(const T dout, const T x) const {
310 311
    T temp = one + abs(x);
    return dout / (temp * temp);
312 313
  }

314
  static constexpr ActBwdOpFwdDeps FwdDeps() { return ActBwdOpFwdDeps::kDepX; }
315 316 317 318 319 320 321 322 323 324 325 326
};

template <typename T>
struct CudaRelu6Functor : public BaseActivationFunctor<T> {
  T zero = static_cast<T>(0.0f);
  float threshold;

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

  // relu6(x) = min(max(0, x), 6)
327
  __device__ __forceinline__ T operator()(const T x) const {
328
    T t = static_cast<T>(threshold);
329
    return x <= zero ? zero : (x < t ? x : t);
330 331 332 333 334 335 336 337 338 339 340 341 342
  }
};

template <typename T>
struct CudaRelu6GradFunctor : public BaseActivationFunctor<T> {
  T zero = static_cast<T>(0.0f);
  float threshold;

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

  // dx = (out > 0 && out < t) ? dout : 0
343
  __device__ __forceinline__ T operator()(const T dout, const T out) const {
344
    T t = static_cast<T>(threshold);
345
    return (out > zero && out < t) ? dout : zero;
346 347
  }

348 349 350
  static constexpr ActBwdOpFwdDeps FwdDeps() {
    return ActBwdOpFwdDeps::kDepOut;
  }
351 352
};

353 354 355 356 357 358 359 360 361 362 363 364 365 366
template <typename T>
struct CudaMishFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType one = static_cast<MPType>(1.0f);
  float threshold;

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

  // mish(x) = x * tanh(softplus(x))
  // softplus(x) = x, if x > threshold
  //             = ln(1 + exp(x)), otherwise
  // Inputs: args[0], the input x
367
  __device__ __forceinline__ T operator()(const T arg_x) const {
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
    MPType x = static_cast<MPType>(arg_x);
    MPType sp = (x > static_cast<MPType>(threshold)) ? x : log(one + exp(x));
    return static_cast<T>(x * tanh(sp));
  }
};

template <typename T>
struct CudaMishGradFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType one = static_cast<MPType>(1.0f);
  float threshold;

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

  // dx = dout * (tanh(sp) + x * (1 - tanh(sp) ** 2) * (1 - exp(-sp)))
  // sp = softplus(x)
  // Inputs: args[0], the input dout
  //         args[1], the input x
388 389
  __device__ __forceinline__ T operator()(const T arg_dout,
                                          const T arg_x) const {
390 391 392 393 394 395 396 397 398
    MPType dout = static_cast<MPType>(arg_dout);
    MPType x = static_cast<MPType>(arg_x);
    MPType sp = (x > static_cast<MPType>(threshold)) ? x : log(one + exp(x));
    MPType gsp =
        (x > static_cast<MPType>(threshold)) ? one : one / (one + exp(-x));
    MPType tsp = tanh(sp);
    return static_cast<T>(dout * (tsp + x * (one - tsp * tsp) * gsp));
  }

399
  static constexpr ActBwdOpFwdDeps FwdDeps() { return ActBwdOpFwdDeps::kDepX; }
400 401
};

402 403 404 405 406 407 408 409 410 411 412 413
template <typename T>
struct CudaCELUFunctor : public BaseActivationFunctor<T> {
  using CT = typename details::MPTypeTrait<T>::Type;
  CT zero = static_cast<CT>(0.0f);
  CT one = static_cast<CT>(1.0f);
  float alpha;

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

  // celu(x) = max(0, x) + min(0, alpha * (exp(x/alpha) - 1))
414
  __device__ __forceinline__ T operator()(const T arg_x) const {
415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436
    CT x = static_cast<CT>(arg_x);
    CT temp = static_cast<CT>(alpha) * (exp(x / static_cast<CT>(alpha)) - one);
    CT res = (x > zero ? x : zero) + (temp > zero ? zero : temp);
    return static_cast<T>(res);
  }
};

template <typename T>
struct CudaCELUGradFunctor : public BaseActivationFunctor<T> {
  using MPType = typename details::MPTypeTrait<T>::Type;
  MPType zero = static_cast<MPType>(0.0f);
  MPType one = static_cast<MPType>(1.0f);
  float alpha;

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

  // dx = dout, if alpha > 0 and x > 0
  // dx = dout * (x/alpha).exp(), if alpha > 0 and x <= 0
  // dx = dout , if alpha < 0 and x > 0
  // dx = dout * (x/alpha).exp(), if alpha < 0 and x <=0
437 438
  __device__ __forceinline__ T operator()(const T arg_dout,
                                          const T arg_x) const {
439 440 441 442 443 444 445 446 447 448 449 450 451
    MPType dout = static_cast<MPType>(arg_dout);
    MPType x = static_cast<MPType>(arg_x);
    MPType a = static_cast<MPType>(alpha);
    MPType temp_a_pos = static_cast<MPType>(alpha > 0.0f);
    MPType temp_a_neg = static_cast<MPType>(alpha <= 0.0f);
    MPType temp_x_pos = static_cast<MPType>(x > zero);
    MPType temp_x_neg = static_cast<MPType>(x <= zero);
    return static_cast<T>(
        dout *
        (temp_a_pos * temp_x_pos + temp_a_pos * temp_x_neg * exp(x / a) +
         temp_a_neg * temp_x_pos + exp(x / a) * temp_a_neg * temp_x_neg));
  }

452
  static constexpr ActBwdOpFwdDeps FwdDeps() { return ActBwdOpFwdDeps::kDepX; }
453 454
};

455
template <typename DeviceContext, typename Functor>
456
class ActivationCudaKernel
457 458 459
    : public framework::OpKernel<typename Functor::ELEMENT_TYPE> {
 public:
  using T = typename Functor::ELEMENT_TYPE;
460 461
  void Compute(const framework::ExecutionContext& ctx) const override {
    const framework::Tensor* x = nullptr;
462
    framework::Tensor* out = nullptr;
463 464 465 466 467 468
    ExtractActivationTensor(ctx, &x, &out);
    out->mutable_data<T>(ctx.GetPlace());
    auto& dev_ctx = ctx.template device_context<DeviceContext>();
    std::vector<const framework::Tensor*> ins = {x};
    std::vector<framework::Tensor*> outs = {out};
    auto functor = Functor();
469 470
    auto attrs = functor.GetAttrs();
    for (auto& attr : attrs) {
471
      *attr.second = ctx.Attr<float>(attr.first);
472
    }
473 474
    paddle::operators::LaunchSameDimsElementwiseCudaKernel<T>(dev_ctx, ins,
                                                              &outs, functor);
475 476 477 478
  }
};

template <typename DeviceContext, typename Functor>
479
class ActivationGradCudaKernel
480 481 482
    : public framework::OpKernel<typename Functor::ELEMENT_TYPE> {
 public:
  using T = typename Functor::ELEMENT_TYPE;
483
  void Compute(const framework::ExecutionContext& ctx) const override {
484 485 486
    const framework::Tensor *x, *out, *d_out;
    framework::Tensor* d_x = nullptr;
    x = out = d_out = nullptr;
487
    ExtractActivationGradTensor<Functor::FwdDeps()>(ctx, &x, &out, &d_out,
488
                                                    &d_x);
489 490 491 492 493 494 495 496 497 498
    d_x->mutable_data<T>(ctx.GetPlace());
    auto& dev_ctx = ctx.template device_context<DeviceContext>();
    auto functor = Functor();
    auto attrs = functor.GetAttrs();
    for (auto& attr : attrs) {
      *attr.second = ctx.Attr<float>(attr.first);
    }

    std::vector<const framework::Tensor*> ins = {d_out};
    std::vector<framework::Tensor*> outs = {d_x};
499

500 501
    if (static_cast<int>(Functor::FwdDeps()) ==
        static_cast<int>(ActBwdOpFwdDeps::kDepOut)) {
502
      // Only need forward output Out
503
      ins.push_back(out);
504 505
      paddle::operators::LaunchSameDimsElementwiseCudaKernel<T>(dev_ctx, ins,
                                                                &outs, functor);
506
    } else if (static_cast<int>(Functor::FwdDeps()) ==
507
               static_cast<int>(ActBwdOpFwdDeps::kDepX)) {
508
      // Only need forward input X
509
      ins.push_back(x);
510 511
      paddle::operators::LaunchSameDimsElementwiseCudaKernel<T>(dev_ctx, ins,
                                                                &outs, functor);
512
    } else {
513 514
      paddle::operators::LaunchSameDimsElementwiseCudaKernel<T>(dev_ctx, ins,
                                                                &outs, functor);
515 516 517 518
    }
  }
};

519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
USE_PHI_FUNCTOR(CudaCos)
USE_PHI_FUNCTOR(CudaTan)
USE_PHI_FUNCTOR(CudaAcos)
USE_PHI_FUNCTOR(CudaSin)
USE_PHI_FUNCTOR(CudaAsin)
USE_PHI_FUNCTOR(CudaAtan)
USE_PHI_FUNCTOR(CudaSinh)
USE_PHI_FUNCTOR(CudaCosh)
USE_PHI_FUNCTOR(CudaAsinh)
USE_PHI_FUNCTOR(CudaAcosh)
USE_PHI_FUNCTOR(CudaAtanh)
USE_PHI_FUNCTOR(CudaTanh)
USE_PHI_FUNCTOR(CudaBRelu)
USE_PHI_FUNCTOR(CudaLeakyRelu)
USE_PHI_FUNCTOR(CudaThresholdedRelu)
Y
YuanRisheng 已提交
534 535 536 537 538
USE_PHI_FUNCTOR(CudaHardShrink)
USE_PHI_FUNCTOR(CudaSoftShrink)
USE_PHI_FUNCTOR(CudaTanhShrink)
USE_PHI_FUNCTOR(CudaSilu)
USE_PHI_FUNCTOR(CudaELU)
Y
YuanRisheng 已提交
539 540 541
USE_PHI_FUNCTOR(CudaSigmoid)
USE_PHI_FUNCTOR(CudaLogSigmoid)
USE_PHI_FUNCTOR(CudaHardSigmoid)
542 543 544 545
USE_PHI_FUNCTOR(CudaLog)
USE_PHI_FUNCTOR(CudaLog2)
USE_PHI_FUNCTOR(CudaLog10)
USE_PHI_FUNCTOR(CudaLog1p)
Y
YuanRisheng 已提交
546 547 548 549 550 551 552 553 554 555 556 557 558 559
USE_PHI_FUNCTOR(CudaSwish)
USE_PHI_FUNCTOR(CudaHardSwish)

template <typename T>
using CudaRoundFunctor = phi::funcs::CudaRoundFunctor<T>;

template <typename T>
using CudaFloorFunctor = phi::funcs::CudaFloorFunctor<T>;

template <typename T>
using CudaCeilFunctor = phi::funcs::CudaCeilFunctor<T>;

template <typename T>
using CudaZeroGradFunctor = phi::funcs::CudaZeroGradFunctor<T>;
Y
YuanRisheng 已提交
560 561 562 563

template <typename T>
using CudaELUGradNegativeAlphaFunctor =
    phi::funcs::CudaELUGradNegativeAlphaFunctor<T>;
564

565 566 567
}  // namespace operators
}  // namespace paddle

568
namespace ops = paddle::operators;
569 570
namespace plat = paddle::platform;

571 572
#define REGISTER_ACTIVATION_CUDA_KERNEL(act_type, op_name, functor,            \
                                        grad_functor)                          \
573
  REGISTER_OP_CUDA_KERNEL(                                                     \
574 575 576 577 578
      act_type, ops::ActivationCudaKernel<paddle::platform::CUDADeviceContext, \
                                          ops::functor<float>>,                \
      ops::ActivationCudaKernel<paddle::platform::CUDADeviceContext,           \
                                ops::functor<double>>,                         \
      ops::ActivationCudaKernel<plat::CUDADeviceContext,                       \
579 580 581
                                ops::functor<plat::float16>>,                  \
      ops::ActivationCudaKernel<plat::CUDADeviceContext,                       \
                                ops::functor<plat::bfloat16>>);                \
582
  REGISTER_OP_CUDA_KERNEL(                                                     \
583 584 585 586 587 588
      act_type##_grad,                                                         \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<float>>,                 \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<double>>,                \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
589 590 591
                                    ops::grad_functor<plat::float16>>,         \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<plat::bfloat16>>);
592

593 594 595 596 597 598 599 600 601 602 603 604
#define REGISTER_ACTIVATION_CUDA_KERNEL_INT(act_type, op_name, functor,        \
                                            grad_functor)                      \
  REGISTER_OP_CUDA_KERNEL(                                                     \
      act_type, ops::ActivationCudaKernel<paddle::platform::CUDADeviceContext, \
                                          ops::functor<float>>,                \
      ops::ActivationCudaKernel<paddle::platform::CUDADeviceContext,           \
                                ops::functor<double>>,                         \
      ops::ActivationCudaKernel<paddle::platform::CUDADeviceContext,           \
                                ops::functor<int>>,                            \
      ops::ActivationCudaKernel<paddle::platform::CUDADeviceContext,           \
                                ops::functor<int64_t>>,                        \
      ops::ActivationCudaKernel<plat::CUDADeviceContext,                       \
605 606 607
                                ops::functor<plat::float16>>,                  \
      ops::ActivationCudaKernel<plat::CUDADeviceContext,                       \
                                ops::functor<plat::bfloat16>>);                \
608 609 610 611 612 613 614 615 616 617 618
  REGISTER_OP_CUDA_KERNEL(                                                     \
      act_type##_grad,                                                         \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<float>>,                 \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<double>>,                \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<int>>,                   \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<int64_t>>,               \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
619 620 621
                                    ops::grad_functor<plat::float16>>,         \
      ops::ActivationGradCudaKernel<plat::CUDADeviceContext,                   \
                                    ops::grad_functor<plat::bfloat16>>);
622

D
Double_V 已提交
623 624
/* ========================================================================== */

625 626 627 628 629 630 631 632 633 634 635 636 637
/* ======================== celu register  ============================ */
REGISTER_ACTIVATION_CUDA_KERNEL(celu, CELU, CudaCELUFunctor,
                                CudaCELUGradFunctor);

REGISTER_OP_CUDA_KERNEL(
    celu_grad_grad, ops::CELUDoubleGradKernel<plat::CUDADeviceContext,
                                              ops::CELUGradGradFunctor<float>>,
    ops::CELUDoubleGradKernel<plat::CUDADeviceContext,
                              ops::CELUGradGradFunctor<double>>,
    ops::CELUDoubleGradKernel<plat::CUDADeviceContext,
                              ops::CELUGradGradFunctor<plat::float16>>);
/* ========================================================================== */

L
lvmengsi 已提交
638
/* ===========================   sqrt register  ============================= */
639 640
REGISTER_ACTIVATION_CUDA_KERNEL(sqrt, Sqrt, CudaSqrtFunctor,
                                CudaSqrtGradFunctor);
L
lvmengsi 已提交
641 642 643 644 645 646 647 648

REGISTER_OP_CUDA_KERNEL(
    sqrt_grad_grad,
    ops::SqrtDoubleGradKernel<paddle::platform::CUDADeviceContext,
                              ops::SqrtGradGradFunctor<float>>,
    ops::SqrtDoubleGradKernel<paddle::platform::CUDADeviceContext,
                              ops::SqrtGradGradFunctor<double>>,
    ops::SqrtDoubleGradKernel<paddle::platform::CUDADeviceContext,
649 650 651
                              ops::SqrtGradGradFunctor<plat::float16>>,
    ops::SqrtDoubleGradKernel<paddle::platform::CUDADeviceContext,
                              ops::SqrtGradGradFunctor<plat::bfloat16>>);
L
lvmengsi 已提交
652 653
/* ========================================================================== */

W
whs 已提交
654 655
/* ===========================   rsqrt register  =============================
 */
656 657
REGISTER_ACTIVATION_CUDA_KERNEL(rsqrt, Rsqrt, CudaRsqrtFunctor,
                                CudaRsqrtGradFunctor);
W
whs 已提交
658 659 660 661 662 663 664 665 666 667 668

REGISTER_OP_CUDA_KERNEL(
    rsqrt_grad_grad,
    ops::RsqrtDoubleGradKernel<paddle::platform::CUDADeviceContext,
                               ops::RsqrtGradGradFunctor<float>>,
    ops::RsqrtDoubleGradKernel<paddle::platform::CUDADeviceContext,
                               ops::RsqrtGradGradFunctor<double>>,
    ops::RsqrtDoubleGradKernel<paddle::platform::CUDADeviceContext,
                               ops::RsqrtGradGradFunctor<plat::float16>>);
/* ========================================================================== */

669
/* ===========================  square register  ============================ */
670 671
REGISTER_ACTIVATION_CUDA_KERNEL_INT(square, Square, CudaSquareFunctor,
                                    CudaSquareGradFunctor);
672 673 674 675 676 677 678 679

REGISTER_OP_CUDA_KERNEL(
    square_grad_grad,
    ops::SquareDoubleGradKernel<paddle::platform::CUDADeviceContext,
                                ops::SquareGradGradFunctor<float>>,
    ops::SquareDoubleGradKernel<paddle::platform::CUDADeviceContext,
                                ops::SquareGradGradFunctor<double>>,
    ops::SquareDoubleGradKernel<plat::CUDADeviceContext,
680
                                ops::SquareGradGradFunctor<plat::float16>>,
681 682
    ops::SquareDoubleGradKernel<plat::CUDADeviceContext,
                                ops::SquareGradGradFunctor<plat::bfloat16>>,
683 684 685 686
    ops::SquareDoubleGradKernel<paddle::platform::CUDADeviceContext,
                                ops::SquareGradGradFunctor<int>>,
    ops::SquareDoubleGradKernel<paddle::platform::CUDADeviceContext,
                                ops::SquareGradGradFunctor<int64_t>>);
687
/* ========================================================================== */
688

W
wangzhen38 已提交
689 690 691 692 693 694 695 696 697 698 699 700 701 702 703
/* ==========================   logit register  ============================ */
namespace ops = paddle::operators;
REGISTER_OP_CUDA_KERNEL(
    logit, ops::LogitKernel<paddle::platform::CUDADeviceContext, float>,
    ops::LogitKernel<paddle::platform::CUDADeviceContext, double>,
    ops::LogitKernel<paddle::platform::CUDADeviceContext,
                     paddle::platform::float16>);
REGISTER_OP_CUDA_KERNEL(
    logit_grad,
    ops::LogitGradKernel<paddle::platform::CUDADeviceContext, float>,
    ops::LogitGradKernel<paddle::platform::CUDADeviceContext, double>,
    ops::LogitGradKernel<paddle::platform::CUDADeviceContext,
                         paddle::platform::float16>);
/* ========================================================================== */

704 705
/* ==========================   exp register  ============================ */
REGISTER_OP_CUDA_KERNEL(
706 707 708 709
    exp, ops::ActivationCudaKernel<plat::CUDADeviceContext,
                                   ops::CudaExpFunctor<float>>,
    ops::ActivationCudaKernel<plat::CUDADeviceContext,
                              ops::CudaExpFunctor<double>>,
710 711
    ops::ActivationKernel<plat::CUDADeviceContext, ops::ExpFunctor<int>>,
    ops::ActivationKernel<plat::CUDADeviceContext, ops::ExpFunctor<int64_t>>,
712 713
    ops::ActivationCudaKernel<plat::CUDADeviceContext,
                              ops::CudaExpFunctor<plat::float16>>);
714
REGISTER_OP_CUDA_KERNEL(
715 716 717 718 719 720 721 722 723 724
    exp_grad, ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                            ops::CudaExpGradFunctor<float>>,
    ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                  ops::CudaExpGradFunctor<double>>,
    ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                  ops::CudaExpGradFunctor<int>>,
    ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                  ops::CudaExpGradFunctor<int64_t>>,
    ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                  ops::CudaExpGradFunctor<plat::float16>>);
725 726
/* ========================================================================== */

R
ronnywang 已提交
727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744
/* ==========================   expm1 register  ============================ */

REGISTER_OP_CUDA_KERNEL(
    expm1, ops::ActivationCudaKernel<plat::CUDADeviceContext,
                                     ops::CudaExpm1Functor<float>>,
    ops::ActivationCudaKernel<plat::CUDADeviceContext,
                              ops::CudaExpm1Functor<double>>,
    ops::ActivationCudaKernel<plat::CUDADeviceContext,
                              ops::CudaExpm1Functor<plat::float16>>);
REGISTER_OP_CUDA_KERNEL(
    expm1_grad, ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                              ops::CudaExpm1GradFunctor<float>>,
    ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                  ops::CudaExpm1GradFunctor<double>>,
    ops::ActivationGradCudaKernel<plat::CUDADeviceContext,
                                  ops::CudaExpm1GradFunctor<plat::float16>>);
/* ========================================================================== */

745 746 747 748 749 750 751 752 753 754 755 756 757 758
#define FOR_EACH_ACTIVATION_CUDA_OP(__macro)                                  \
  __macro(softshrink, SoftShrink, CudaSoftShrinkFunctor,                      \
          CudaSoftShrinkGradFunctor);                                         \
  __macro(reciprocal, Reciprocal, CudaReciprocalFunctor,                      \
          CudaReciprocalGradFunctor);                                         \
  __macro(soft_relu, SoftRelu, CudaSoftReluFunctor, CudaSoftReluGradFunctor); \
  __macro(stanh, STanh, CudaSTanhFunctor, CudaSTanhGradFunctor);              \
  __macro(softplus, Softplus, CudaSoftplusFunctor, CudaSoftplusGradFunctor);  \
  __macro(softsign, Softsign, CudaSoftsignFunctor, CudaSoftsignGradFunctor);  \
  __macro(relu6, Relu6, CudaRelu6Functor, CudaRelu6GradFunctor);              \
  __macro(tanh_shrink, TanhShrink, CudaTanhShrinkFunctor,                     \
          CudaTanhShrinkGradFunctor);                                         \
  __macro(hard_shrink, HardShrink, CudaHardShrinkFunctor,                     \
          CudaHardShrinkGradFunctor);                                         \
Y
YuanRisheng 已提交
759
  __macro(mish, Mish, CudaMishFunctor, CudaMishGradFunctor);
760
FOR_EACH_ACTIVATION_CUDA_OP(REGISTER_ACTIVATION_CUDA_KERNEL)
761 762

#ifdef PADDLE_WITH_XPU_KP
763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978
REGISTER_OP_KERNEL(
    brelu, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              phi::funcs::CudaBReluFunctor<float>>);
REGISTER_OP_KERNEL(
    brelu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  phi::funcs::CudaBReluGradFunctor<float>>);

REGISTER_OP_KERNEL(ceil, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaCeilFunctor<float>>);
REGISTER_OP_KERNEL(
    ceil_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaZeroGradFunctor<float>>);

REGISTER_OP_KERNEL(celu, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaCELUFunctor<float>>);
REGISTER_OP_KERNEL(
    celu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaCELUGradFunctor<float>>);

REGISTER_OP_KERNEL(elu, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaELUFunctor<float>>);
REGISTER_OP_KERNEL(
    elu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaELUGradFunctor<float>>);

REGISTER_OP_KERNEL(exp, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaExpFunctor<float>>);
REGISTER_OP_KERNEL(
    exp_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaExpGradFunctor<float>>);

REGISTER_OP_KERNEL(floor, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaFloorFunctor<float>>);
REGISTER_OP_KERNEL(
    floor_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaZeroGradFunctor<float>>);

REGISTER_OP_KERNEL(
    hard_shrink, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              ops::CudaHardShrinkFunctor<float>>);
REGISTER_OP_KERNEL(
    hard_shrink_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaHardShrinkGradFunctor<float>>);

REGISTER_OP_KERNEL(
    hard_sigmoid, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              ops::CudaHardSigmoidFunctor<float>>);
REGISTER_OP_KERNEL(
    hard_sigmoid_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaHardSigmoidGradFunctor<float>>);

REGISTER_OP_KERNEL(hard_swish, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaHardSwishFunctor<float>>);
REGISTER_OP_KERNEL(
    hard_swish_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaHardSwishGradFunctor<float>>);

REGISTER_OP_KERNEL(
    leaky_relu, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              phi::funcs::CudaLeakyReluFunctor<float>>);
REGISTER_OP_KERNEL(
    leaky_relu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  phi::funcs::CudaLeakyReluGradFunctor<float>>);

REGISTER_OP_KERNEL(log, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaLogFunctor<float>>);
REGISTER_OP_KERNEL(
    log_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaLogGradFunctor<float>>);

REGISTER_OP_KERNEL(log1p, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaLog1pFunctor<float>>);
REGISTER_OP_KERNEL(
    log1p_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaLog1pGradFunctor<float>>);

REGISTER_OP_KERNEL(
    logsigmoid, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              ops::CudaLogSigmoidFunctor<float>>);
REGISTER_OP_KERNEL(
    logsigmoid_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaLogSigmoidGradFunctor<float>>);

REGISTER_OP_KERNEL(
    reciprocal, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              ops::CudaReciprocalFunctor<float>>);
REGISTER_OP_KERNEL(
    reciprocal_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaReciprocalGradFunctor<float>>);

REGISTER_OP_KERNEL(
    relu, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              phi::funcs::CudaReluFunctor<float>>);
REGISTER_OP_KERNEL(
    relu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  phi::funcs::CudaReluGradFunctor<float>>);

REGISTER_OP_KERNEL(relu6, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaRelu6Functor<float>>);
REGISTER_OP_KERNEL(
    relu6_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaRelu6GradFunctor<float>>);

REGISTER_OP_KERNEL(sigmoid, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSigmoidFunctor<float>>);
REGISTER_OP_KERNEL(
    sigmoid_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSigmoidGradFunctor<float>>);

REGISTER_OP_KERNEL(silu, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSiluFunctor<float>>);
REGISTER_OP_KERNEL(
    silu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSiluGradFunctor<float>>);

REGISTER_OP_KERNEL(soft_relu, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSoftReluFunctor<float>>);
REGISTER_OP_KERNEL(
    soft_relu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSoftReluGradFunctor<float>>);

REGISTER_OP_KERNEL(softplus, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSoftplusFunctor<float>>);
REGISTER_OP_KERNEL(
    softplus_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSoftplusGradFunctor<float>>);

REGISTER_OP_KERNEL(
    softshrink, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              ops::CudaSoftShrinkFunctor<float>>);
REGISTER_OP_KERNEL(
    softshrink_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSoftShrinkGradFunctor<float>>);

REGISTER_OP_KERNEL(softsign, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSoftsignFunctor<float>>);
REGISTER_OP_KERNEL(
    softsign_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSoftsignGradFunctor<float>>);

REGISTER_OP_KERNEL(sqrt, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSqrtFunctor<float>>);
REGISTER_OP_KERNEL(
    sqrt_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSqrtGradFunctor<float>>);

REGISTER_OP_KERNEL(square, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSquareFunctor<float>>);
REGISTER_OP_KERNEL(
    square_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSquareGradFunctor<float>>);

REGISTER_OP_KERNEL(swish, KP, plat::XPUPlace,
                   ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                                             ops::CudaSwishFunctor<float>>);
REGISTER_OP_KERNEL(
    swish_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaSwishGradFunctor<float>>);

REGISTER_OP_KERNEL(
    thresholded_relu, KP, plat::XPUPlace,
    ops::ActivationCudaKernel<paddle::platform::XPUDeviceContext,
                              ops::CudaThresholdedReluFunctor<float>>);
REGISTER_OP_KERNEL(
    thresholded_relu_grad, KP, plat::XPUPlace,
    ops::ActivationGradCudaKernel<paddle::platform::XPUDeviceContext,
                                  ops::CudaThresholdedReluGradFunctor<float>>);
979 980

#endif  // PADDLE_WITH_XPU_KP