TorchLean API

NN.Examples.Data.RealPaths

Real Dataset Paths #

Generated tutorial artifacts live in NN.Examples.Data.SamplePaths.

This module names the default paths used by examples that train on datasets prepared by scripts/datasets/download_example_data.py. The examples use small checked artifacts by default; larger public datasets live under data/real after the user runs the preparation script.

Default root for user-downloaded real datasets.

Pass this to TorchLean.CLI.takePathFlag args "data-dir" when an example accepts --data-dir for prepared real datasets.

Instances For

    Directory containing prepared CIFAR-10 .npy arrays.

    Instances For

      Prepared CIFAR-10 training images, shape (N, 3, 32, 32), float32 in $[0,1]$.

      Instances For

        Prepared CIFAR-10 training labels, shape (N,), float32 integer labels 0..9.

        Instances For

          Directory containing a user-prepared ImageNet-style 64x64 subset.

          ImageNet-style runs start from a local image-folder dataset. Users point scripts/datasets/torchlean_data_convert.py image-folder at an ImageNet/ILSVRC, ImageNet-compatible, or Tiny-ImageNet-style directory tree and write the converted arrays here.

          Instances For

            Prepared ImageNet-style training images, shape (N, 3, 64, 64), float32 in $[0,1]$.

            Instances For

              Prepared ImageNet-style training labels, shape (N,), float32 integer class ids.

              The converter assigns ids by sorted subdirectory name when --labels-from-dirs is used.

              Instances For

                Directory containing prepared UCI household-power forecasting windows.

                Instances For

                  Prepared UCI household-power inputs, shape (N, 24, 1), float32 normalized to $[0,1]$.

                  Each row is a 24-hour window of hourly mean Global_active_power.

                  Instances For

                    Prepared UCI household-power targets, shape (N, 24, 1), float32 normalized to $[0,1]$.

                    Each target row is the corresponding input window shifted by one hour.

                    Instances For

                      Directory containing the prepared UCI Auto MPG tabular regression CSV.

                      Instances For

                        Prepared UCI Auto MPG CSV with normalized columns x1..x7,y.

                        Instances For

                          Directory containing downloaded text corpora.

                          Instances For

                            Karpathy tiny-shakespeare corpus.

                            Instances For

                              TinyStories validation split, useful for small local language-model training checks.

                              Instances For

                                Data-preparation hint for commands that only need Tiny Shakespeare.

                                Instances For

                                  Data-preparation hint for commands that accept both Tiny Shakespeare and TinyStories.

                                  Instances For