k1lib.knn module

Some nice utils to complement torch.nn. This is exposed automatically with:

from k1lib.imports import *
knn.Lambda # exposed

class k1lib.knn.Lambda(f: Callable[[Any], Any])

Bases: torch.nn.modules.module.Module

__init__(f: Callable[[Any], Any])

Creates a simple module with a specified forward() function.

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.

training: bool

class k1lib.knn.Identity

Bases: k1lib.knn.Lambda

Creates a module that returns the input in forward()

training: bool

class k1lib.knn.LinBlock(inC, outC)

Bases: torch.nn.modules.module.Module

__init__(inC, outC)

Linear layer with relu behind it

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.

training: bool