skainet-lang-core/sk.ainet.lang.nn/GroupedConv2d

GroupedConv2d

class GroupedConv2d<T : DType, V>(val inChannels: Int, val outChannels: Int, val kernelSize: Pair<Int, Int>, val groups: Int, val stride: Pair<Int, Int> = 1 to 1, val padding: Pair<Int, Int> = 0 to 0, val dilation: Pair<Int, Int> = 1 to 1, val bias: Boolean = true, val name: String = "GroupedConv2d", initWeights: Tensor<T, V>, initBias: Tensor<T, V>? = null) : Module<T, V> , ModuleParameters<T, V> (source)

Grouped Convolution layer.

In grouped convolution, the input channels are divided into groups, and each group is convolved separately with its own set of filters. This reduces the number of parameters and computational complexity while potentially improving feature learning by encouraging different groups to learn different types of features.

When groups = 1, this is equivalent to standard convolution. When groups = inChannels, this is equivalent to depthwise convolution.

This is essentially a convenience wrapper around Conv2d with explicit group handling and additional utility methods for grouped convolution operations.

Parameters

inChannels

Number of input channels

outChannels

Number of output channels/filters

kernelSize

Size of the convolving kernel (height, width)

groups

Number of groups to divide channels into

stride

Stride of the convolution (default: 1, 1)

padding

Padding added to all sides of the input (default: 0, 0)

dilation

Spacing between kernel elements (default: 1, 1)

bias

Whether to add a learnable bias to the output (default: true)

name

Name of the module

initWeights

Initial weights tensor

initBias

Initial bias tensor (if bias is true)

Constructors

GroupedConv2d

constructor(inChannels: Int, outChannels: Int, kernelSize: Pair<Int, Int>, groups: Int, stride: Pair<Int, Int> = 1 to 1, padding: Pair<Int, Int> = 0 to 0, dilation: Pair<Int, Int> = 1 to 1, bias: Boolean = true, name: String = "GroupedConv2d", initWeights: Tensor<T, V>, initBias: Tensor<T, V>? = null)

Properties

bias

val bias: Boolean

dilation

val dilation: Pair<Int, Int>

groups

val groups: Int

inChannels

val inChannels: Int

kernelSize

val kernelSize: Pair<Int, Int>

modules

open override val modules: List<Module<T, V>>

name

open override val name: String

outChannels

val outChannels: Int

padding

val padding: Pair<Int, Int>

params

open override val params: List<ModuleParameter<T, V>>

Parameters owned by this node (weights, biases, etc.).

stride

val stride: Pair<Int, Int>

Functions

computationalReduction

fun computationalReduction(): Double

Returns the computational complexity reduction ratio compared to standard convolution. This is the same as parameter reduction for convolution operations.

forward