ResNet Real-Data Example

Runnable torchlean resnet example. It trains a compact ResNet-style classifier built from API.nn.resnetBasicBlock on a prepared CIFAR-10 minibatch.

The reusable model wiring lives in NN.API.Models.Resnet (nn.models.resnet). This file is the runnable wrapper (CIFAR loader construction + multi-epoch training loop).

python3 scripts/datasets/download_example_data.py --cifar10
lake build -R -K cuda=true && lake exe torchlean resnet --cuda --n-total 200 --steps 1

Tip: the defaults are set for a quick sanity run. For a longer run:

lake build -R -K cuda=true
lake exe torchlean resnet --cuda --fast-kernels --n-total 5000 --steps 200

def NN.Examples.Models.Vision.Resnet.exeName : String

def NN.Examples.Models.Vision.Resnet.defaultLogJson : System.FilePath

def NN.Examples.Models.Vision.Resnet.batch : ℕ

def NN.Examples.Models.Vision.Resnet.inC : ℕ

def NN.Examples.Models.Vision.Resnet.inH : ℕ

def NN.Examples.Models.Vision.Resnet.inW : ℕ

def NN.Examples.Models.Vision.Resnet.stemC : ℕ

def NN.Examples.Models.Vision.Resnet.numClasses : ℕ

def NN.Examples.Models.Vision.Resnet.cfg : API.nn.models.ResnetConfig

@[reducible, inline]

abbrev NN.Examples.Models.Vision.Resnet.σ : Shape

@[reducible, inline]

abbrev NN.Examples.Models.Vision.Resnet.τ : Shape

def NN.Examples.Models.Vision.Resnet.mkModel : API.nn.M (API.nn.Sequential σ τ)

def NN.Examples.Models.Vision.Resnet.loadCifarLoader {α : Type} [API.Semantics.Scalar α] [API.Runtime.Scalar α] (xPath yPath : System.FilePath) (nRows seed : ℕ) : IO (API.Data.BatchLoader α batch RealData.CifarImage RealData.CifarTarget)

def NN.Examples.Models.Vision.Resnet.main (args : List String) : IO UInt32