dotnet/machinelearning · error · ArgumentException

The runner metric manager is of type {_metricManager.GetType

Error message

The runner metric manager is of type {_metricManager.GetType()} which expected to be of type {typeof(ITrainValidateDatasetManager)} or {typeof(ICrossValidateDatasetManager)}

What it means

RegressionExperiment.RunAsync's evaluation loop dispatches on the type of _metricManager and only handles ITrainValidateDatasetManager and ICrossValidateDatasetManager. When the manager is another type, neither branch produces a TrialResult and control reaches the trailing throw of ArgumentException, indicating a misconfigured runner.

Source

Thrown at src/Microsoft.ML.AutoML/API/RegressionExperiment.cs:450

                            var loss = metricManager.IsMaximize ? -metric : metric;

                            stopWatch.Stop();


                            return Task.FromResult(new TrialResult<RegressionMetrics>()
                            {
                                Loss = loss,
                                Metric = metric,
                                Metrics = metrics,
                                Model = model,
                                TrialSettings = settings,
                                DurationInMilliseconds = stopWatch.ElapsedMilliseconds,
                                Pipeline = refitPipeline,
                            } as TrialResult);
                        }
                    }

                    throw new ArgumentException($"The runner metric manager is of type {_metricManager.GetType()} which expected to be of type {typeof(ITrainValidateDatasetManager)} or {typeof(ICrossValidateDatasetManager)}");
                }
            }
            catch (Exception ex) when (ct.IsCancellationRequested)
            {
                throw new OperationCanceledException(ex.Message, ex.InnerException);
            }
            catch (Exception)
            {
                throw;
            }
        }

        public void Dispose()
        {
            _context.CancelExecution();
            _context = null;
        }

View on GitHub (pinned to 7b76e69cf9)

Solutions

Construct the experiment through the standard builders so a TrainTestSplit (ITrainValidateDatasetManager) or CrossValidation (ICrossValidateDatasetManager) manager is used.
Make custom dataset managers implement one of the two expected interfaces.
Validate the manager type at experiment setup and throw early with a descriptive message.
Align all Microsoft.ML.AutoML package references to the same version.

Example fix

// before
experiment.SetDataset(new MyDatasetManager(data)); // unsupported type
await experiment.RunAsync(ct);
// after
experiment.SetDataset(new CrossValidationDatasetManager(data, numFolds: 5));
await experiment.RunAsync(ct);

Defensive patterns

Strategy: type-guard

Validate before calling

// before RunAsync:
if (_metricManager is not ITrainValidateDatasetManager and not ICrossValidateDatasetManager)
    throw new InvalidOperationException("Use TrainTestSplit or CrossValidation dataset manager");

Type guard

bool IsValidManager(object m) => m is ITrainValidateDatasetManager or ICrossValidateDatasetManager;

Try / catch

try { var best = await experiment.RunAsync(ct); }
catch (ArgumentException ex) when (ex.Message.Contains("metric manager")) { // rebuild experiment with standard dataset manager
    throw new InvalidOperationException("Regression experiment dataset manager misconfigured", ex); }

Prevention

Build regression experiments through mlContext.Auto().CreateRegressionExperiment/StandardPipelines
Avoid custom IDatasetManager implementations unless they implement the expected interfaces
Add a construction-time assertion for the manager type

When it happens

Trigger: Running a regression AutoML experiment whose metric/dataset manager object is neither ITrainValidateDatasetManager nor ICrossValidateDatasetManager (custom implementation or wrong object injected), so both switch branches are skipped and the terminal ArgumentException is hit.

Common situations: Hand-rolled AutoML runners; dependency-injection misconfiguration supplying a custom IDatasetManager; package version mismatch where runner interfaces were refactored.

Understand the failure class

Background: Type mismatch errors: IllegalArgumentException, TypeError and type guards across 150 open-source libraries — this error's family across 150 libraries.

Related errors

AI-assisted analysis of dotnet/machinelearning@7b76e69cf9 (2026-09-11). Data as JSON: /api/errors/2d8d0f74ec86c34e. Report an issue: GitHub.