Metalhead

Metalhead.jl provides standard machine learning vision models for use with Flux.jl. The architectures in this package make use of pure Flux layers, and they represent the best-practices for creating modules like residual blocks, inception blocks, etc. in Flux. Metalhead also provides some building blocks for more complex models in the Layers module.

Installation

Metalhead.jl requires Julia 1.10 or later. Install it from the Julia package manager:

julia> ]add Metalhead

Getting Started

You can find the Metalhead.jl getting started guide here.

Available models

To contribute new models, see our contributing docs.

Image Classification

Model Name	Constructor	Pre-trained?
AlexNet	`AlexNet`	N
ConvMixer	`ConvMixer`	N
ConvNeXt	`ConvNeXt`	N
DenseNet	`DenseNet`	N
EfficientNet	`EfficientNet`	N
EfficientNetv2	`EfficientNetv2`	N
gMLP	`gMLP`	N
GoogLeNet	`GoogLeNet`	N
Inception-v3	`Inceptionv3`	N
Inception-v4	`Inceptionv4`	N
InceptionResNet-v2	`InceptionResNetv2`	N
MLPMixer	`MLPMixer`	N
MobileNetv1	`MobileNetv1`	N
MobileNetv2	`MobileNetv2`	N
MobileNetv3	`MobileNetv3`	N
MNASNet	`MNASNet`	N
Res2Net	`Res2Net`	N
Res2NeXt	`Res2NeXt`	N
ResMLP	`ResMLP`	N
ResNet	`ResNet`	Y
ResNeXt	`ResNeXt`	Y
SE-ResNet	`SEResNet`	N
SE-ResNeXt	`SEResNeXt`	N
SqueezeNet	`SqueezeNet`	Y
Xception	`Xception`	N
WideResNet	`WideResNet`	Y
VGG	`VGG`	Y
Vision Transformer	`ViT`	Y

Other Models

Model Name	Constructor	Pre-trained?
UNet	`UNet`	N