Model Zoo
-
Transformer as introduced by Attention Is All You Need.
The transformer model shares an embedding matrix between the encoder and decoder and reuses the embedding weights to compute the decoder output distribution. Outputs of the transformer are normalized using log softmax.
See moreDeclaration
Swift
public struct Transformer<Element, Device> : LayerType, Codable where Element : RandomizableType, Device : DeviceType
-
AlexNet image classification network.
Batch normalization has been added to this implementation.
See moreDeclaration
Swift
public struct AlexNet<Element, Device> : LayerType, Codable where Element : RandomizableType, Device : DeviceType
-
Residual neural network with 18 layers (17 convolutional, 1 dense)
See moreDeclaration
Swift
public struct ResNet18<Element, Device> : LayerType where Element : RandomizableType, Device : DeviceType
-
Residual block with two convolution and batch normalization layers as well as an optional downsampling block.
See moreDeclaration
Swift
public struct ResidualBlock<Element, Device> : LayerType where Element : RandomizableType, Device : DeviceType
-
Undocumented
See moreDeclaration
Swift
public struct VGG11<E, D> : VGGBase where E : RandomizableType, D : DeviceType
-
Undocumented
See moreDeclaration
Swift
public struct VGG13<E, D> : VGGBase where E : RandomizableType, D : DeviceType
-
Undocumented
See moreDeclaration
Swift
public struct VGG16<E, D> : VGGBase where E : RandomizableType, D : DeviceType
-
Undocumented
See moreDeclaration
Swift
public struct VGG19<E, D> : VGGBase where E : RandomizableType, D : DeviceType