xma.cute_dsl_utils.utils

torch_tensor_to_cute_tensor(x: Tensor, leading_dim: int) → cutlass.cute.Tensor