#define TORCH_ASSERT_NO_OPERATORS

#include <ATen/Dispatch.h>
#include <ATen/native/cuda/JitLoops.cuh>
#include <ATen/native/cuda/Loops.cuh>
#include <ATen/native/BinaryOps.h>
#include <ATen/native/Math.h>
#include <ATen/native/cuda/Math.cuh>
#include <ATen/native/cuda/jit_utils.h>

namespace at::native {
namespace {
constexpr char shifted_chebyshev_polynomial_v_name[] = "shifted_chebyshev_polynomial_v_forward";

void shifted_chebyshev_polynomial_v_kernel_cuda(TensorIteratorBase& iterator) {
#if AT_USE_JITERATOR()
    AT_DISPATCH_FLOATING_TYPES(iterator.common_dtype(), "shifted_chebyshev_polynomial_v_cuda", [&]() {
        opmath_jitted_gpu_kernel_with_scalars<shifted_chebyshev_polynomial_v_name, scalar_t, scalar_t>(iterator, shifted_chebyshev_polynomial_v_string);
    });
#else
    AT_DISPATCH_FLOATING_TYPES(iterator.common_dtype(), "shifted_chebyshev_polynomial_v_cuda", [&]() {
        gpu_kernel_with_scalars(iterator, []GPU_LAMBDA(scalar_t x, scalar_t n) -> scalar_t {
            return shifted_chebyshev_polynomial_v_forward<scalar_t, true>(x, n);
        });
    });
#endif
} // shifted_chebyshev_polynomial_v_kernel_cuda

} // namespace (anonymous)

REGISTER_DISPATCH(shifted_chebyshev_polynomial_v_stub, &shifted_chebyshev_polynomial_v_kernel_cuda)
} // namespace at::native