Automatic Differentiation
 
Loading...
Searching...
No Matches
softmax.hpp
Go to the documentation of this file.
1#ifndef STAN_MATH_OPENCL_REV_SOFTMAX_HPP
2#define STAN_MATH_OPENCL_REV_SOFTMAX_HPP
3#ifdef STAN_OPENCL
4
10
11namespace stan {
12namespace math {
13
22template <typename T,
23 require_all_kernel_expressions_and_none_scalar_t<T>* = nullptr>
25 if (A.size() == 0) {
26 return A;
27 }
28 return make_callback_var(
29 softmax(A.val()), [A](vari_value<matrix_cl<double>>& res) mutable {
30 A.adj() += elt_multiply(
31 res.val(), (res.adj() - dot_product(res.adj(), res.val())));
32 });
33}
34
35} // namespace math
36} // namespace stan
37
38#endif
39#endif
Represents an arithmetic matrix on the OpenCL device.
Definition matrix_cl.hpp:47
var_value< plain_type_t< T > > make_callback_var(T &&value, F &&functor)
Creates a new var initialized with a callback_vari with a given value and reverse-pass callback funct...
auto softmax(const ColVec &alpha)
Definition softmax.hpp:16
The lgamma implementation in stan-math is based on either the reentrant safe lgamma_r implementation ...