Models module

`OSRModule`(n_known[, lower_stack, verbose])
`Openmax`([lower_stack, n_known, epsilon, ...])	Implementation of Openmax method

class torchosr.models.OSRModule(n_known, lower_stack=None, verbose=False)

Bases: Module

forward(X)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

train(dataloader, loss_fn, optimizer)

Sets the module in training mode.

This has any effect only on certain modules. See documentations of particular modules for details of their behaviors in training/evaluation mode, if they are affected, e.g. Dropout, BatchNorm, etc.

Args:

mode (bool): whether to set training mode (True) or evaluation: mode (False). Default: True.

Returns:

Module: self

class torchosr.models.Openmax(lower_stack=None, n_known=3, epsilon=0.5, tail=20, alpha=2)

Bases: TSoftmax

Implementation of Openmax method

Parameters:

lower_stack (nn.Sequential) – Network architecture of lower_stack
n_known (int) – Number of known classes
epsilon (float) – Threshold for prediction probability
tail (int) – Tail size for estimating the parameters of Weibull distribution
alpha (int) – Alpha (number of most significant classes to revise)