add_n

paddle. add_n ( inputs, name=None ) [source]

This OP is used to sum one or more Tensor of the input.

For example:

Case 1:

    Input:
        input.shape = [2, 3]
        input = [[1, 2, 3],
                 [4, 5, 6]]

    Output:
        output.shape = [2, 3]
        output = [[1, 2, 3],
                  [4, 5, 6]]

Case 2:

    Input:
        First input:
            input1.shape = [2, 3]
            Input1 = [[1, 2, 3],
                      [4, 5, 6]]

        The second input:
            input2.shape = [2, 3]
            input2 = [[7, 8, 9],
                      [10, 11, 12]]

        Output:
            output.shape = [2, 3]
            output = [[8, 10, 12],
                      [14, 16, 18]]
Parameters
  • inputs (Tensor|list(Tensor)) – A Tensor list. The shape and data type of the list elements should be consistent. Input can be multi-dimensional Tensor, and data types can be: float32, float64, int32, int64.

  • name (str, optional) – The default value is None. Normally there is no need for user to set this property. For more information, please refer to Name

Returns

Tensor, the sum of input \(inputs\) , its shape and data types are consistent with \(inputs\).

Examples

import paddle

input0 = paddle.to_tensor([[1, 2, 3], [4, 5, 6]], dtype='float32')
input1 = paddle.to_tensor([[7, 8, 9], [10, 11, 12]], dtype='float32')
output = paddle.add_n([input0, input1])
# [[8., 10., 12.],
#  [14., 16., 18.]]