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

index_tensor_shape

def index_tensor_shape[output_rank: Int, input_type: DType, indices_type: DType, batch_dims: Int](input_buf: TileTensor[input_type, address_space=input_buf.address_space, linear_idx_type=input_buf.linear_idx_type, element_size=input_buf.element_size], indices_buf: TileTensor[indices_type, address_space=indices_buf.address_space, linear_idx_type=indices_buf.linear_idx_type, element_size=indices_buf.element_size]) -> IndexList[output_rank]

Compute the output shape of a index_tensor operation, and assert the inputs are compatible.

Parameters:

output_rank (Int): Rank of the output tensor.
input_type (DType): Type of the input tensor.
indices_type (DType): Type of the indices tensor.
batch_dims (Int): Batch dimensions.

Args:

input_buf (TileTensor[input_type, address_space=input_buf.address_space, linear_idx_type=input_buf.linear_idx_type, element_size=input_buf.element_size]): The input tensor.
indices_buf (TileTensor[indices_type, address_space=indices_buf.address_space, linear_idx_type=indices_buf.linear_idx_type, element_size=indices_buf.element_size]): The indices tensor.

Returns:

IndexList[output_rank]: The output shape.