forked from pytorch/pytorch
-
Notifications
You must be signed in to change notification settings - Fork 0
/
ActivationHardswishKernel.cu
63 lines (54 loc) · 2.03 KB
/
ActivationHardswishKernel.cu
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
#define TORCH_ASSERT_NO_OPERATORS
#define _USE_MATH_DEFINES
#include <ATen/native/Activation.h>
#include <cmath>
#include <thrust/tuple.h>
#include <ATen/AccumulateType.h>
#include <ATen/Dispatch.h>
#include <ATen/core/TensorBase.h>
#include <c10/core/Scalar.h>
#include <c10/cuda/CUDAMathCompat.h>
#include <ATen/cuda/ApplyGridUtils.cuh>
#include <ATen/cuda/detail/OffsetCalculator.cuh>
#include <ATen/native/cuda/Loops.cuh>
namespace at::native {
namespace {
void hardswish_kernel(TensorIterator& iter) {
AT_DISPATCH_FLOATING_TYPES_AND2(at::ScalarType::Half, at::ScalarType::BFloat16, iter.dtype(), "hardswish_cuda", [&]() {
using opmath_t = at::opmath_type<scalar_t>;
const opmath_t zero(0.0f);
const opmath_t one_sixth(1.0f / 6.0f);
const opmath_t three(3.0f);
const opmath_t six(6.0f);
gpu_kernel(iter, [zero, one_sixth, three, six]GPU_LAMBDA(scalar_t self_val) -> scalar_t {
opmath_t x = static_cast<opmath_t>(self_val);
return x * std::min(std::max(x + three, zero), six) * one_sixth;
});
});
}
void hardswish_backward_kernel(TensorIterator& iter) {
AT_DISPATCH_FLOATING_TYPES_AND2(at::ScalarType::Half, at::ScalarType::BFloat16, iter.dtype(), "hardswish_backward_cuda", [&]() {
using opmath_t = at::opmath_type<scalar_t>;
const opmath_t zero(0.0f);
const opmath_t three(3.0f);
const opmath_t neg_three(-3.0f);
const opmath_t one_half(0.5f);
gpu_kernel(
iter,
[zero, three, neg_three, one_half]GPU_LAMBDA(scalar_t grad_val_, scalar_t self_val_) -> scalar_t {
opmath_t grad_val = static_cast<opmath_t>(grad_val_);
opmath_t self_val = static_cast<opmath_t>(self_val_);
if (self_val < neg_three) {
return zero;
} else if (self_val <= three) {
return grad_val * ((self_val / three) + one_half);
} else {
return grad_val;
}
});
});
}
} // namespace
REGISTER_DISPATCH(hardswish_stub, &hardswish_kernel);
REGISTER_DISPATCH(hardswish_backward_stub, &hardswish_backward_kernel);
} // namespace at::native