Pooling Layers#

Classes#

class AdaptivePool(shape, shapes, *, channels=True, mode='average', **kwargs)[source]#

Uses pooling to downscale the input to the desired shape.

Attributes:

groupint: Layer group, if 0 it will always be used, else it will only be used if its group matches the Networks
descriptionstr: Description of the layer
layersSequential: Layers to loop through in the forward pass

Parameters:

shapeint | list[int]: Output shape of the layer
shapeslist[list[int]]: Shape of the outputs from each layer
channelsbool, Optional: If the input includes a channels dimension, default = True
mode{‘average’, ‘max’}: Whether to use ‘max’ or ‘average’ pooling
**kwargs: Leftover parameters to pass to base layer for checking

extra_repr()[source]#

Displays layer parameters when printing the network

Returns:

class Pool(shapes, *, kernel=2, stride=2, padding=0, mode='max', pool_kwargs=None, **kwargs)[source]#

Constructs a max or average pooling layer.

Supports 1D, 2D, and 3D pooling.

Attributes:

groupint: Layer group, if 0 it will always be used, else it will only be used if its group matches the Networks
descriptionstr: Description of the layer
layersSequential: Layers to loop through in the forward pass

Parameters:

shapeslist[list[int]]: Shape of the outputs from each layer
kernelint | list[int], Optional: Size of the kernel, default = 2
strideint | list[int], Optional: Stride of the kernel, default = 2
paddingint | {‘same’} | list[int], Optional: Input padding, can an int or ‘same’ where ‘same’ preserves the input shape, default = 0
mode{‘max’, ‘average’}: Whether to use ‘max’ or ‘average’ pooling
pool_kwargsdict[str, Any] | NoneType, Optional: Additional keyword arguments to pass to the pooling layer
**kwargs: Leftover parameters to pass to base layer for checking

forward(x, *args, **kwargs)[source]#

Forward pass of the pool layer.

Parameters:

xTensor: Input tensor of shape (N,…) and type float, where N is the batch size
*args: Optional arguments to pass to the parent forward method
**kwargs: Optional keyword arguments to pass to the parent forward method

Returns:

class PoolDownscale(scale, shapes, *, mode='max', **kwargs)[source]#

Bases: Pool

Downscales the input using pooling.

Supports 1D, 2D, and 3D pooling.

Attributes:

groupint: Layer group, if 0 it will always be used, else it will only be used if its group matches the Networks
descriptionstr: Description of the layer
layersSequential: Layers to loop through in the forward pass

Parameters:

scaleint: Stride and size of the kernel, which acts as the downscaling factor
shapeslist[list[int]]: Shape of the outputs from each layer
mode{‘max’, ‘average’}: Whether to use ‘max’ or ‘average’ pooling
**kwargs: Leftover parameters to pass to base layer for checking