oneflow.addcdiv¶
-
oneflow.
addcdiv
(input, tensor1, tensor2, *, value=1) → Tensor¶ This function is equivalent to PyTorch’s addcdiv function. The documentation is referenced from: https://pytorch.org/docs/1.10/generated/torch.addcdiv.html.
Performs the element-wise division of
tensor1
bytensor2
, multiply the result by the scalarvalue
and add it toinput
.\[\text{out}_i = \text{input}_i + \text{value} \times \frac{\text{tensor1}_i}{\text{tensor2}_i}\]The shapes of
input
,tensor1
, andtensor2
must be broadcastable.For inputs of type FloatTensor or DoubleTensor,
value
must be a real number, otherwise an integer.- Parameters
- Keyword Arguments
value (Number, optional) – multiplier for \(\text{{tensor1}} / \text{{tensor2}}\)
Example:
>>> import oneflow as flow >>> input = flow.tensor([ 0.3810, 1.2774, -0.2972, -0.3719]) >>> tensor1 = flow.tensor([0.8032, 0.2930, -0.8113, -0.2308]) >>> tensor2 = flow.tensor([[0.5], [1]]) >>> output = flow.addcdiv(input, tensor1, tensor2) >>> output.shape oneflow.Size([2, 4])