Training Command-Line Options

Command-line options shared by runnable training programs. These records describe process-level choices such as step counts, log destinations, and CUDA allocator sampling. Model definitions and the Trainer.train API do not depend on them.

structure TorchLean.CLI.Training.RunOptions : Type

Step, batching, logging, and allocator-reporting options for a training command.

• steps : Nat

  Number of optimizer updates.

• batchSize : Nat

  Number of in-memory samples consumed by one optimizer update.

• logDestination : Training.LogDestination

  Destination for the JSON training log.

• cudaMemorySampleEvery : Nat

  Number of completed steps between CUDA allocator samples; 0 selects the default policy.

def TorchLean.CLI.Training.instReprRunOptions.repr : RunOptions → Nat → Std.Format

@[instance_reducible]

instance TorchLean.CLI.Training.instReprRunOptions : Repr RunOptions

def TorchLean.CLI.Training.RunOptions.parse (exeName : String) (arguments : List String) (defaultLogPath : System.FilePath) (defaultSteps defaultBatchSize : Nat := 1) (allowZeroSteps : Bool := false) : Except String (RunOptions × List String)

Parse the common options accepted by runnable training commands.

structure TorchLean.CLI.Training.OptimizerOptions : Type

Training command options that also select a learning rate.

• steps : Nat

  Number of optimizer updates.

• batchSize : Nat

  Number of in-memory samples consumed by one optimizer update.

• logDestination : Training.LogDestination

  Destination for the JSON training log.

• cudaMemorySampleEvery : Nat

  Number of completed steps between CUDA allocator samples; 0 selects the default policy.

• learningRate : Float

  Learning rate passed to the command's optimizer constructor.

@[instance_reducible]

instance TorchLean.CLI.Training.instReprOptimizerOptions : Repr OptimizerOptions

def TorchLean.CLI.Training.instReprOptimizerOptions.repr : OptimizerOptions → Nat → Std.Format

def TorchLean.CLI.Training.OptimizerOptions.toRunOptions (options : OptimizerOptions) : RunOptions

Step, batching, and logging settings without the optimizer learning rate.

def TorchLean.CLI.Training.OptimizerOptions.parse (exeName : String) (arguments : List String) (defaultLogPath : System.FilePath) (defaultSteps : Nat := 1) (defaultLearningRate : Float := 1e-3) (defaultBatchSize : Nat := 1) (allowZeroSteps : Bool := false) : Except String (OptimizerOptions × List String)

Parse run options followed by a positive --lr value.