public InferredPipeline(IEnumerable <SuggestedTransform> transforms,
SuggestedTrainer trainer,
MLContext context = null)
{
Transforms = transforms.Select(t => t.Clone()).ToList();
Trainer = trainer.Clone();
_context = context ?? new MLContext();
AddNormalizationTransforms();
}
/// <summary>
/// Given a predictor type & target max num of iterations, return a set of all permissible trainers (with their sweeper params, if defined).
/// </summary>
/// <returns>Array of viable learners.</returns>
public static IEnumerable <SuggestedTrainer> AllowedTrainers(MLContext mlContext, TaskKind task,
int maxIterations)
{
var trainerExtensions = TrainerExtensionCatalog.GetTrainers(task, maxIterations);
var trainers = new List <SuggestedTrainer>();
foreach (var trainerExtension in trainerExtensions)
{
var learner = new SuggestedTrainer(mlContext, trainerExtension);
trainers.Add(learner);
}
return(trainers.ToArray());
}
private static void SampleHyperparameters(SuggestedTrainer trainer, IEnumerable <PipelineRunResult> history, bool isMaximizingMetric)
{
var sps = ConvertToValueGenerators(trainer.SweepParams);
var sweeper = new SmacSweeper(
new SmacSweeper.Arguments
{
SweptParameters = sps
});
IEnumerable <PipelineRunResult> historyToUse = history
.Where(r => r.RunSucceded && r.Pipeline.Trainer.TrainerName == trainer.TrainerName && r.Pipeline.Trainer.HyperParamSet != null);
// get new set of hyperparameter values
var proposedParamSet = sweeper.ProposeSweeps(1, historyToUse.Select(h => h.ToRunResult(isMaximizingMetric))).First();
// associate proposed param set with trainer, so that smart hyperparam
// sweepers (like KDO) can map them back.
trainer.SetHyperparamValues(proposedParamSet);
}