create_const_float

Applicable Products

All processors

Function Usage

Creates a float constant tensor.

Prototype

1
create_const_float(value: Union[float, List[float]], shape: Optional[List[int]] = None) -> TensorHolder

Parameters

Parameter

Input/Output

Description

value

Input

A single floating-point number or a list of floating-point numbers.

shape

Input

(Optional) Shape dimension.

Returns

(TensorHolder) Tensor holder of the constant

Restrictions

  • If it is a list and the shape is provided, the number of elements must match the product of the shape dimensions.
  • If the value is not of the float, float list, or int type, TypeError is thrown.
  • If the number of values does not match the shape, ValueError is thrown.
  • If the constant fails to be created, RuntimeError is thrown.