For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo function

transpose_strided

def transpose_strided[rank: Int, dtype: DType](output: TileTensor[dtype, Storage=output.Storage, linear_idx_type=output.linear_idx_type], input: TileTensor[dtype, Storage=input.Storage, linear_idx_type=input.linear_idx_type], perms: Pointer[Int, _safe=False], ctx: Optional[DeviceContext] = None)

Transposes a tensor with arbitrary strides via a generic strided copy.

Computes row-major strides for the input, permutes them according to perms, and recursively copies data into the row-major output buffer.

Parameters:

rank (Int): Number of dimensions in the tensor shape.
dtype (DType): Element type of the input and output buffers.

Args:

output (TileTensor[dtype, Storage=output.Storage, linear_idx_type=output.linear_idx_type]): The output buffer.
input (TileTensor[dtype, Storage=input.Storage, linear_idx_type=input.linear_idx_type]): The input buffer.
perms (Pointer[Int, _safe=False]): Permutation of the input axes.
ctx (Optional[DeviceContext]): The context to execute the work on.