ncdl.nn.Hardtanh

class ncdl.nn.Hardtanh(*args: Any, **kwargs: Any)

Applies the HardTanh function element-wise.

HardTanh is defined as:

\[\begin{split}\text{HardTanh}(x) = \begin{cases} \text{max\_val} & \text{ if } x > \text{ max\_val } \\ \text{min\_val} & \text{ if } x < \text{ min\_val } \\ x & \text{ otherwise } \\ \end{cases}\end{split}\]
Parameters:

  • min_val – minimum value of the linear region range. Default: -1

  • max_val – maximum value of the linear region range. Default: 1

  • inplace – can optionally do the operation in-place. Default: False

Keyword arguments min_value and max_value have been deprecated in favor of min_val and max_val.

Shape:

  • Input: \((*)\), where \(*\) means any number of dimensions.

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

scripts/activation_images/Hardtanh.png

Examples:

>>> m = nn.Hardtanh(-2, 2)
>>> input = torch.randn(2)
>>> output = m(input)
__init__(min_val: float = -1.0, max_val: float = 1.0, inplace: bool = False, min_value: float | None = None, max_value: float | None = None) → None

Methods

__init__([min_val, max_val, inplace, ...])

extra_repr()

forward(input)

Attributes

min_val

max_val

inplace