Model-Zoo Commands

Command-line records used by TorchLean's runnable model examples.

These declarations live with the model zoo because they describe repository commands and local artifact formats, not the tensor, trainer, or runtime API. Applications can use the public trainer directly without importing diffusion schedules, PPM paths, or prepared-dataset defaults.

Generative-model artifacts

structure NN.Examples.ModelZoo.DiffusionScheduleFlags : Type

Diffusion schedule parameters exposed by the model-zoo diffusion command.

• T : ℕ

  Number of diffusion timesteps in the schedule.

• betaStart : Float

  First beta value in the schedule.

• betaEnd : Float

  Final beta value in the schedule.

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprDiffusionScheduleFlags : Repr DiffusionScheduleFlags

def NN.Examples.ModelZoo.instReprDiffusionScheduleFlags.repr : DiffusionScheduleFlags → ℕ → Std.Format

def NN.Examples.ModelZoo.DiffusionScheduleFlags.parse (args : List String) (defaultT : ℕ := 100) (defaultBetaStart : Float := 1e-4) (defaultBetaEnd : Float := 0.12) : Except String (DiffusionScheduleFlags × List String)

Parse --T, --beta-start, and --beta-end.

def NN.Examples.ModelZoo.DiffusionScheduleFlags.trainLogNotes (cfg : DiffusionScheduleFlags) : Array String

Stable TrainLog metadata for a diffusion schedule.

structure NN.Examples.ModelZoo.ImageArtifactFlags : Type

Optional image artifacts emitted by generation and reconstruction examples.

• reconstructStep? : Option ℕ

  Optional timestep used for reconstruction-from-noise artifacts.

• samplePpm? : Option System.FilePath

  Optional path for an unconditional sample image.

• referencePpm? : Option System.FilePath

  Optional path for the clean reference image.

• noisyPpm? : Option System.FilePath

  Optional path for the noised or intermediate image.

• reconstructPpm? : Option System.FilePath

  Optional path for the reconstructed image.

def NN.Examples.ModelZoo.instReprImageArtifactFlags.repr : ImageArtifactFlags → ℕ → Std.Format

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprImageArtifactFlags : Repr ImageArtifactFlags

def NN.Examples.ModelZoo.ImageArtifactFlags.parse (args : List String) : Except String (ImageArtifactFlags × List String)

Parse image-artifact paths and the optional reconstruction timestep.

def NN.Examples.ModelZoo.ImageArtifactFlags.trainLogNotes (cfg : ImageArtifactFlags) : Array String

Stable TrainLog metadata for requested image artifacts.

Prepared-data and diagnostic artifacts

structure NN.Examples.ModelZoo.PairedNpyEvalFlags : Type

Train/test tensor paths and row counts for paired-NPY scientific examples.

• trainRows : ℕ

  Number of rows loaded from the prepared training tensors.

• testRows : ℕ

  Number of rows loaded from the prepared held-out tensors.

• evalRows : ℕ

  Prefix length used for deterministic train/test loss reports.

• trainX : System.FilePath

  Training input .npy path.

• trainY : System.FilePath

  Training target .npy path.

• testX : System.FilePath

  Held-out input .npy path.

• testY : System.FilePath

  Held-out target .npy path.

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprPairedNpyEvalFlags : Repr PairedNpyEvalFlags

def NN.Examples.ModelZoo.instReprPairedNpyEvalFlags.repr : PairedNpyEvalFlags → ℕ → Std.Format

def NN.Examples.ModelZoo.PairedNpyEvalFlags.parse (args : List String) (defaultTrainX defaultTrainY defaultTestX defaultTestY : System.FilePath) (defaultTrainRows defaultTestRows : ℕ) (defaultEvalRows : ℕ := 16) : Except String (PairedNpyEvalFlags × List String)

Parse train/test paths and row counts for paired NPY tensors.

def NN.Examples.ModelZoo.PairedNpyEvalFlags.trainLogNotes (cfg : PairedNpyEvalFlags) : Array String

Stable TrainLog metadata for paired train/test NPY tensors.

structure NN.Examples.ModelZoo.CsvArtifactFlags : Type

Optional CSV artifact path for commands that emit one tabular diagnostic.

• plotCsv : System.FilePath

  CSV path for the diagnostic artifact.

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprCsvArtifactFlags : Repr CsvArtifactFlags

def NN.Examples.ModelZoo.instReprCsvArtifactFlags.repr : CsvArtifactFlags → ℕ → Std.Format

def NN.Examples.ModelZoo.CsvArtifactFlags.parse (args : List String) (defaultPlotCsv : System.FilePath) : Except String (CsvArtifactFlags × List String)

Parse the optional --plot-csv artifact path.

structure NN.Examples.ModelZoo.NpyDataFlags : Type

Prepared NPY feature/target paths and their row budget.

• xPath : System.FilePath

  Prepared feature or image tensor path.

• yPath : System.FilePath

  Prepared label or target tensor path.

• nRows : ℕ

  Number of rows to read from the arrays.

• seed : ℕ

  Data-loader seed.

def NN.Examples.ModelZoo.instReprNpyDataFlags.repr : NpyDataFlags → ℕ → Std.Format

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprNpyDataFlags : Repr NpyDataFlags

def NN.Examples.ModelZoo.NpyDataFlags.parse (args : List String) (defaultX defaultY : System.FilePath) (defaultRows : ℕ) : Except String (NpyDataFlags × List String)

Parse --seed, --n-total, --x, and --y for an NPY-backed example.

def NN.Examples.ModelZoo.NpyDataFlags.trainLogNotes (cfg : NpyDataFlags) (datasetName : String) : Array String

Stable TrainLog metadata for an NPY-backed dataset branch.

inductive NN.Examples.ModelZoo.ImageDatasetChoice : Type

Prepared image datasets understood by the built-in image-model commands.

• imagenet64 : ImageDatasetChoice

  Prepared 64x64 RGB image tensors.

• cifar10 : ImageDatasetChoice

  Prepared CIFAR-10 32x32 RGB tensors.

def NN.Examples.ModelZoo.instReprImageDatasetChoice.repr : ImageDatasetChoice → ℕ → Std.Format

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprImageDatasetChoice : Repr ImageDatasetChoice

@[implicit_reducible]

instance NN.Examples.ModelZoo.instBEqImageDatasetChoice : BEq ImageDatasetChoice

def NN.Examples.ModelZoo.instBEqImageDatasetChoice.beq : ImageDatasetChoice → ImageDatasetChoice → Bool

def NN.Examples.ModelZoo.ImageDatasetChoice.parse (args : List String) : Except String (ImageDatasetChoice × List String)

Parse the dataset selector and reject ambiguous combinations.

structure NN.Examples.ModelZoo.ForecastWindowDataFlags : Type

Prepared forecasting-window paths and report controls.

• xPath : System.FilePath

  Prepared input-window tensor path.

• yPath : System.FilePath

  Prepared target-window tensor path.

• windows : ℕ

  Number of forecasting windows to use.

• reportOffset : ℕ

  Window index used for before/after forecast display.

• seed : ℕ

  Data-loader seed.

def NN.Examples.ModelZoo.instReprForecastWindowDataFlags.repr : ForecastWindowDataFlags → ℕ → Std.Format

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprForecastWindowDataFlags : Repr ForecastWindowDataFlags

def NN.Examples.ModelZoo.ForecastWindowDataFlags.trainLogNotes (cfg : ForecastWindowDataFlags) : Array String

Stable TrainLog metadata for forecasting-window datasets.

Data plus training controls

structure NN.Examples.ModelZoo.SeededTrainFlagsextends TorchLean.CLI.Training.OptimizerOptions : Type

Standard model-training flags plus a reproducibility seed.

• steps : ℕ
• batchSize : ℕ
• log : Training.LogDestination
• logPath : System.FilePath
• cudaMemWatch : ℕ
• lr : Float
• seed : ℕ

  Seed used for initialization, synthetic data, or row order.

def NN.Examples.ModelZoo.instReprSeededTrainFlags.repr : SeededTrainFlags → ℕ → Std.Format

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprSeededTrainFlags : Repr SeededTrainFlags

def NN.Examples.ModelZoo.parseSeededTrainFlags (exeName : String) (args : List String) (defaultLogPath : System.FilePath) (defaultSteps : ℕ := 1) (defaultLr : Float := 1e-3) (allowZeroSteps : Bool := false) : Except String (SeededTrainFlags × List String)

Parse the standard model-training flags together with --seed.

structure NN.Examples.ModelZoo.NpyLoggedTrainFlagsextends TorchLean.CLI.Training.RunOptions, NN.Examples.ModelZoo.NpyDataFlags : Type

Fixed-step training flags paired with an NPY dataset.

• steps : ℕ
• batchSize : ℕ
• log : Training.LogDestination
• logPath : System.FilePath
• cudaMemWatch : ℕ
• xPath : System.FilePath
• yPath : System.FilePath
• nRows : ℕ
• seed : ℕ

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprNpyLoggedTrainFlags : Repr NpyLoggedTrainFlags

def NN.Examples.ModelZoo.instReprNpyLoggedTrainFlags.repr : NpyLoggedTrainFlags → ℕ → Std.Format

def NN.Examples.ModelZoo.NpyLoggedTrainFlags.parse (exeName : String) (args : List String) (defaultLogPath : System.FilePath) (defaultSteps : ℕ) (parseData : List String → Except String (NpyDataFlags × List String)) : Except String NpyLoggedTrainFlags

Parse NPY data and logged-training flags, then reject unused command arguments.

structure NN.Examples.ModelZoo.NpyModelTrainFlagsextends TorchLean.CLI.Training.OptimizerOptions, NN.Examples.ModelZoo.NpyDataFlags : Type

Optimizer/training flags paired with an NPY dataset.

• steps : ℕ
• batchSize : ℕ
• log : Training.LogDestination
• logPath : System.FilePath
• cudaMemWatch : ℕ
• lr : Float
• xPath : System.FilePath
• yPath : System.FilePath
• nRows : ℕ
• seed : ℕ

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprNpyModelTrainFlags : Repr NpyModelTrainFlags

def NN.Examples.ModelZoo.instReprNpyModelTrainFlags.repr : NpyModelTrainFlags → ℕ → Std.Format

def NN.Examples.ModelZoo.NpyModelTrainFlags.parse (exeName : String) (args : List String) (defaultLogPath : System.FilePath) (defaultSteps : ℕ := 1) (defaultLr : Float := 1e-3) (parseData : List String → Except String (NpyDataFlags × List String)) : Except String (NpyModelTrainFlags × List String)

Parse NPY data and the standard model-training flags.

structure NN.Examples.ModelZoo.ForecastWindowModelTrainFlagsextends TorchLean.CLI.Training.OptimizerOptions, NN.Examples.ModelZoo.ForecastWindowDataFlags : Type

Optimizer/training flags paired with a forecasting-window dataset.

• steps : ℕ
• batchSize : ℕ
• log : Training.LogDestination
• logPath : System.FilePath
• cudaMemWatch : ℕ
• lr : Float
• xPath : System.FilePath
• yPath : System.FilePath
• windows : ℕ
• reportOffset : ℕ
• seed : ℕ

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprForecastWindowModelTrainFlags : Repr ForecastWindowModelTrainFlags

def NN.Examples.ModelZoo.instReprForecastWindowModelTrainFlags.repr : ForecastWindowModelTrainFlags → ℕ → Std.Format

def NN.Examples.ModelZoo.ForecastWindowModelTrainFlags.parse (exeName : String) (args : List String) (defaultLogPath : System.FilePath) (defaultSteps : ℕ := 100) (defaultLr : Float := 1e-2) (parseData : List String → Except String (ForecastWindowDataFlags × List String)) : Except String (ForecastWindowModelTrainFlags × List String)

Parse forecasting data and the standard model-training flags.

structure NN.Examples.ModelZoo.CsvTrainFlagsextends TorchLean.CLI.Training.OptimizerOptions : Type

Optimizer/training flags for a model command that reads one supervised CSV.

• steps : ℕ
• batchSize : ℕ
• log : Training.LogDestination
• logPath : System.FilePath
• cudaMemWatch : ℕ
• lr : Float
• csvPath : System.FilePath

  CSV file containing model inputs and targets.

• seed : ℕ

  Seed used for model initialization and data shuffling.

@[implicit_reducible]

instance NN.Examples.ModelZoo.instReprCsvTrainFlags : Repr CsvTrainFlags

def NN.Examples.ModelZoo.instReprCsvTrainFlags.repr : CsvTrainFlags → ℕ → Std.Format

def NN.Examples.ModelZoo.parseCsvTrainFlags (exeName : String) (args : List String) (defaultCsv defaultLogPath : System.FilePath) (defaultSteps : ℕ := 1) (defaultLr : Float := 1e-3) (allowZeroSteps : Bool := false) : Except String (CsvTrainFlags × List String)

Parse a CSV path, seed, and the standard model-training flags.

def TorchLean.CLI.Training.RunOptions.toTrainerOptions (opts : RunOptions) (enableLog : Bool := true) (logEvery : ℕ := 0) (title : String := "Training") (notes : Array String := #[]) : Trainer.TrainOptions

Convert parsed command options to the options consumed by Trainer.train.

def TorchLean.CLI.Training.RunOptions.toTrainerOptionsWhenRequested (opts : RunOptions) (args : List String) (logEvery : ℕ := 0) (title : String := "Training") (notes : Array String := #[]) : Trainer.TrainOptions

Enable logging only when the command line explicitly contains --log.

def TorchLean.CLI.Training.OptimizerOptions.toTrainerOptions (opts : OptimizerOptions) (logEvery : ℕ := 0) (title : String := "Training") (notes : Array String := #[]) : Trainer.TrainOptions

Convert parsed optimizer and command options to the options consumed by Trainer.train.