#include <nn_f16.h>
Tensor0D_F16
A 0D tensor (scalar) with a half-precision floating-point data type.