Skip to main content

Mojo function

scatter_nd_shape

scatter_nd_shape[input_type: DType, indices_type: DType](input: TileTensor[input_type, input.LayoutType, input.origin, address_space=input.address_space, linear_idx_type=input.linear_idx_type, element_size=input.element_size], updates: TileTensor[input_type, updates.LayoutType, updates.origin, address_space=updates.address_space, linear_idx_type=updates.linear_idx_type, element_size=updates.element_size], indices: TileTensor[indices_type, indices.LayoutType, indices.origin, address_space=indices.address_space, linear_idx_type=indices.linear_idx_type, element_size=indices.element_size]) -> IndexList[TileTensor[input_type, input.LayoutType, input.origin, address_space=input.address_space, linear_idx_type=input.linear_idx_type, element_size=input.element_size].rank]

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

Parameters:

  • input_type (DType): Type of the input tensor.
  • indices_type (DType): Type of the indices tensor.

Args:

Returns:

IndexList: The output shape.

Was this page helpful?