mindspore.ops.AccumulateNV2

class mindspore.ops.AccumulateNV2(*args, **kwargs)[source]

Computes accumulation of all input tensors element-wise.

AccumulateNV2 is similar to AddN, but there is a significant difference among them: AccumulateNV2 will not wait for all of its inputs to be ready before summing. That is to say, AccumulateNV2 is able to save memory when inputs are ready at different time since the minimum temporary storage is proportional to the output size rather than the input size.

Inputs:

input_x (Union(tuple[Tensor], list[Tensor])) - The input tuple or list is made up of multiple tensors whose dtype is number to be added together.

Outputs:

Tensor, has the same shape and dtype as each entry of the input_x.

Raises: TypeError – If input_x is neither tuple nor list.

Supported Platforms:: Ascend

Examples

>>> class NetAccumulateNV2(nn.Cell):
...     def __init__(self):
...         super(NetAccumulateNV2, self).__init__()
...         self.accumulateNV2 = ops.AccumulateNV2()
...
...     def construct(self, *z):
...         return self.accumulateNV2(z)
...
>>> net = NetAccumulateNV2()
>>> input_x = Tensor(np.array([1, 2, 3]), mindspore.float32)
>>> input_y = Tensor(np.array([4, 5, 6]), mindspore.float32)
>>> output = net(input_x, input_y, input_x, input_y)
>>> print(output)
[10. 14. 18.]