Sum#

  • Domain: ai.onnx

  • Since version: 13

Element-wise sum of each of the input tensors (with Numpy-style broadcasting support). All inputs and outputs must have the same data type.

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check the doc.

Inputs

  • data_0 (T): List of tensors for sum.

Outputs

  • sum (T): Output tensor.

Type Constraints

  • T: Constrain input and output types to float tensors. Allowed types: tensor(bfloat16), tensor(double), tensor(float), tensor(float16).

Examples#

test_cc_sum_bcast

Node:
  Sum(data_0, data_1) -> (sum)

Inputs:
  data_0: shape=(2, 2), dtype=float32
    [[1., 2.],
     [3., 4.]]
  data_1: shape=(), dtype=float32
    10.

Outputs:
  sum: shape=(2, 2), dtype=float32
    [[11., 12.],
     [13., 14.]]

test_cc_sum_example

Node:
  Sum(data_0, data_1, data_2) -> (sum)

Inputs:
  data_0: shape=(3,), dtype=float32
    [1., 0., 1.]
  data_1: shape=(3,), dtype=float32
    [3., 4., 5.]
  data_2: shape=(3,), dtype=float32
    [6., 0., 5.]

Outputs:
  sum: shape=(3,), dtype=float32
    [10.,  4., 11.]

test_cc_sum_one_input

Node:
  Sum(data_0) -> (sum)

Inputs:
  data_0: shape=(3,), dtype=float32
    [1., 2., 3.]

Outputs:
  sum: shape=(3,), dtype=float32
    [1., 2., 3.]

test_cc_sum_two_inputs

Node:
  Sum(data_0, data_1) -> (sum)

Inputs:
  data_0: shape=(2, 3), dtype=float32
    [[1., 2., 3.],
     [4., 5., 6.]]
  data_1: shape=(2, 3), dtype=float32
    [[10., 20., 30.],
     [40., 50., 60.]]

Outputs:
  sum: shape=(2, 3), dtype=float32
    [[11., 22., 33.],
     [44., 55., 66.]]

test_sum_example

Node:
  Sum(data_0, data_1, data_2) -> (sum)

Inputs:
  data_0: shape=(3, 4, 5), dtype=float32
    [[[-1.6207618 , -0.26159498, -1.6894207 , -0.5941726 ,  0.23757638],
      [-0.77598375,  1.7195183 ,  0.87334126, -0.47817174,  0.16062957],
      [-1.1535512 , -1.0489278 , -0.8616083 , -0.25931633,  0.1662263 ],
      [-1.19579   , -0.09554294, -0.7415066 ,  1.8345209 ,  0.66055524]],

     [[-0.37864083, -1.2763816 , -0.20821808,  0.79001987, -0.305968  ],
      [-0.13338329, -0.26947752, -0.68312395, -0.02752227,  1.2929565 ],
      [-2.3823667 , -1.3751754 ,  0.08531325, -1.0451319 ,  0.5773258 ],
      [ 0.5335717 ,  1.7940854 , -0.25193027, -0.42498377,  0.62358594]],

     [[-0.21925525,  0.9713564 , -0.12102764,  0.72468317, -0.30977774],
      [ 1.0066974 ,  0.20055684,  0.35697085, -0.9946885 ,  1.5686967 ],
      [-2.015447  , -0.80032647, -1.3453012 , -1.3160369 ,  0.09626406],
      [ 0.00318402,  1.1699535 , -1.1138605 , -0.72566646, -0.04194619]]]
  data_1: shape=(3, 4, 5), dtype=float32
    [[[-3.2327610e-01,  5.3205568e-01,  1.5432492e+00, -6.9176567e-01,
        1.9503884e+00],
      [-1.7234032e+00, -1.0676117e+00, -1.6856503e+00, -3.3474345e-02,
       -9.3504351e-01],
      [ 8.5077822e-01,  4.7314724e-01, -3.6031015e-02,  1.1229317e+00,
        1.0024427e-01],
      [-1.3745983e-01, -4.5334585e-02,  4.9413046e-01,  5.0943595e-01,
        2.7962857e-01]],

     [[ 1.6754329e+00, -1.1071100e+00,  3.6536074e-01,  9.5587057e-01,
        9.5825948e-02],
      [-3.4379008e+00, -2.1401776e-01,  1.0207568e+00,  1.5230697e-01,
        1.7962818e-01],
      [ 1.9142120e-01,  1.2306712e-04, -1.4188983e+00, -1.1545553e+00,
       -3.6328000e-01],
      [-1.1123656e+00, -2.7028066e-01,  1.0528580e+00, -1.3669981e+00,
       -1.9538646e+00]],

     [[-8.5295737e-01,  1.2235709e+00, -9.7294033e-02,  8.7098098e-01,
        3.7720674e-01],
      [ 1.0523090e+00, -5.4001397e-01,  5.1259655e-01,  3.2776645e-01,
       -2.8144464e-01],
      [ 2.0450303e-01, -5.3457934e-01,  7.0617035e-02, -4.4991007e-01,
        7.2952157e-01],
      [-1.1638440e+00,  6.6377270e-01, -1.4313440e+00, -1.1646674e+00,
       -1.5334731e+00]]]
  data_2: shape=(3, 4, 5), dtype=float32
    [[[-1.9504728 , -0.384003  ,  0.6175111 ,  0.4860437 ,  0.14837483],
      [ 0.4243168 ,  0.7555941 ,  1.2209445 , -0.8658253 , -0.2577181 ],
      [-1.1415162 , -0.40477493, -0.87081665,  0.88169414,  0.38555267],
      [-1.5191493 , -1.1147851 , -1.6198577 ,  0.18040422, -0.61582077]],

     [[ 1.0728145 , -0.32273647, -1.4748397 ,  1.3682464 ,  0.69247085],
      [ 1.4945487 , -1.8275837 ,  1.1269883 , -1.1244485 ,  0.19961263],
      [ 0.8090156 ,  0.8177859 , -0.01459381, -1.4260765 ,  1.5820233 ],
      [-1.1409763 ,  0.3658983 ,  0.5160113 , -1.0337888 ,  0.6509004 ]],

     [[-0.2102342 , -0.90819484, -0.16215591, -0.6403306 , -0.68816286],
      [ 2.1505942 ,  0.07699525, -0.6226731 , -0.6401388 ,  0.8192693 ],
      [-1.3715283 , -0.58853245,  0.3524632 ,  0.0612135 ,  0.20318082],
      [-0.23185334,  0.7497727 ,  0.7727014 ,  0.7761635 , -0.32464558]]]

Outputs:
  sum: shape=(3, 4, 5), dtype=float32
    [[[-3.8945107 , -0.11354232,  0.47133964, -0.7998946 ,  2.3363397 ],
      [-2.0750701 ,  1.4075007 ,  0.40863544, -1.3774714 , -1.032132  ],
      [-1.4442892 , -0.9805554 , -1.768456  ,  1.7453096 ,  0.6520232 ],
      [-2.8523993 , -1.2556626 , -1.8672338 ,  2.5243611 ,  0.32436305]],

     [[ 2.3696065 , -2.706228  , -1.317697  ,  3.1141367 ,  0.4823288 ],
      [-2.0767355 , -2.311079  ,  1.4646212 , -0.99966383,  1.6721972 ],
      [-1.3819299 , -0.5572664 , -1.348179  , -3.625764  ,  1.7960691 ],
      [-1.7197702 ,  1.889703  ,  1.3169391 , -2.8257706 , -0.6793782 ]],

     [[-1.2824467 ,  1.2867324 , -0.3804776 ,  0.95533353, -0.62073386],
      [ 4.2096004 , -0.26246187,  0.2468943 , -1.307061  ,  2.1065214 ],
      [-3.1824722 , -1.9234383 , -0.9222209 , -1.7047335 ,  1.0289664 ],
      [-1.3925134 ,  2.583499  , -1.7725033 , -1.1141703 , -1.900065  ]]]

Differences with previous version (8)#

SchemaDiff: Sum (domain 'ai.onnx')

  • old version: 8

  • new version: 13

  • breaking: no

Type constraints:

  • changed ‘T’: added types: [‘tensor(bfloat16)’]

Version History#