oneflow.nn.Threshold

class oneflow.nn.Threshold(threshold: float, value: float)

The Threshold Activation. Return x if x is greater than threshold, else return value.

The interface is consistent with PyTorch. The documentation is referenced from https://pytorch.org/docs/1.10/generated/torch.nn.Threshold.html.

The formula is:

\[\begin{split}\text{Threshold}(x) = \begin{cases} x, & \text{ if } x > \text{ threshold } \\ \text{value }, & \text{ otherwise } \end{cases}\end{split}\]
Parameters
  • threshold (float) – The threshold value for the Threshold formulation

  • value (float) – The value value for the Threshold formulation

Shapes:
  • Input: \((N, *)\) where * means, any number of additional dimensions

  • Output: \((N, *)\), same shape as the input

Returns

The result tensor

Return type

Oneflow.Tensor

For example:

>>> import oneflow as flow
>>> import numpy as np
>>> x = np.array([-1, 0, 0.5, 1]).astype(np.float32)
>>> input = flow.Tensor(x)
>>> th = flow.nn.Threshold(threshold=0.5, value=0.2)
>>> out = th(input)
>>> out
tensor([0.2000, 0.2000, 0.2000, 1.0000], dtype=oneflow.float32)