pub fn prepare_input_owned<T: Scalar>(
arr: StridedArray<T>,
_n_batch: usize,
n_group1: usize,
n_group2: usize,
conj: bool,
) -> Result<ContiguousOperand<T>>Expand description
Prepare an owned input array for GEMM.
Expects batch-last canonical order: [group1..., group2..., batch...].
If already contiguous after dimension grouping, transfers ownership without copying.
Otherwise, copies to a new col-major buffer.