Autoencoder CIFAR Example

Trains a compact vector autoencoder on a real CIFAR-10 minibatch.

def NN.Examples.Models.Generative.Autoencoder.exeName : String

CLI subcommand name used in terminal banners and error messages.

def NN.Examples.Models.Generative.Autoencoder.defaultLogJson : System.FilePath

Default JSON loss-curve path for this command.

def NN.Examples.Models.Generative.Autoencoder.cfg : TorchLean.nn.models.VectorGenerativeConfig

Shared vector-image configuration.

The compact config fixes the CIFAR batch size, flattened image dimension, and latent width used by the vector generative examples, so autoencoder/VAE/VQ-VAE/GAN runs use the same data boundary.

@[reducible, inline]

abbrev NN.Examples.Models.Generative.Autoencoder.σ : Shape

Input shape: a batch of flattened CIFAR image vectors.

@[reducible, inline]

abbrev NN.Examples.Models.Generative.Autoencoder.τ : Shape

Target shape: the same flattened image-vector batch, because this is reconstruction.

def NN.Examples.Models.Generative.Autoencoder.model : TorchLean.nn.M (TorchLean.nn.Sequential σ τ)

Trainable vector autoencoder.

The architecture is defined in the public model API. The command chooses the dataset, optimizer, runtime options, and logging path.

def NN.Examples.Models.Generative.Autoencoder.data (flags : RealData.CifarModelTrainFlags) : TorchLean.Trainer.Dataset σ τ

Public singleton dataset for compact CIFAR reconstruction.

def NN.Examples.Models.Generative.Autoencoder.train (opts : TorchLean.Options) (flags : RealData.CifarModelTrainFlags) : IO (TorchLean.Trainer.TrainResult σ τ)

Train the compact autoencoder with the public Trainer surface.

def NN.Examples.Models.Generative.Autoencoder.main (args : List String) : IO UInt32

Executable entrypoint for CIFAR reconstruction.

The command loads one real CIFAR minibatch, builds the supervised reconstruction sample x -> x, trains the autoencoder for --steps, and writes the standard TorchLean training summary/log.