dotnet/machinelearning · error · ArgumentException

{nameof(numberOfCVFolds)} must be at least 2

Error message

{nameof(numberOfCVFolds)} must be at least 2

What it means

UserInputValidationUtil.ValidateNumberOfCVFoldsArg rejects cross-validation fold counts below 2 with ArgumentException naming the parameter — cross-validation requires at least a train/test split, so 0 or 1 folds is meaningless.

Source

Thrown at src/Microsoft.ML.AutoML/Utils/UserInputValidationUtil.cs:57

            ValidatePath(path);
        }

        public static void ValidateInferColumnsArgs(string path, string labelColumn)
        {
            ValidateLabelColumn(labelColumn);
            ValidatePath(path);
        }

        public static void ValidateInferColumnsArgs(string path)
        {
            ValidatePath(path);
        }

        public static void ValidateNumberOfCVFoldsArg(uint numberOfCVFolds)
        {
            if (numberOfCVFolds <= 1)
            {
                throw new ArgumentException($"{nameof(numberOfCVFolds)} must be at least 2", nameof(numberOfCVFolds));
            }
        }

        public static void ValidateSamplingKey(string samplingKeyColumnName, string groupIdColumnName, TaskKind task)
        {
            if (task == TaskKind.Ranking && samplingKeyColumnName != null && samplingKeyColumnName != groupIdColumnName)
            {
                throw new ArgumentException($"If provided, {nameof(samplingKeyColumnName)} must be the same as {nameof(groupIdColumnName)} for Ranking Experiments", samplingKeyColumnName);
            }
        }

        private static void ValidateTrainData(IDataView trainData, ColumnInformation columnInformation)
        {
            if (trainData == null)
            {
                throw new ArgumentNullException(nameof(trainData), "Training data cannot be null");
            }

View on GitHub (pinned to 7b76e69cf9)

Solutions

1. Pass numberOfCVFolds >= 2 (call ValidateNumberOfCVFoldsArg early to fail fast).
2. Clamp user/config-supplied fold values to Math.Max(2, value) before calling the API.
3. Use non-cross-validation execution (train/validation split) if only one split is desired.

Example fix

// before
experimentSettings.NumberOfCVFolds = 1;
// after
experimentSettings.NumberOfCVFolds = Math.Max(2, userConfiguredFolds);

if (numberOfCVFolds < 2)
    throw new ArgumentException(nameof(numberOfCVFolds), "Must be at least 2");
UserInputValidationUtil.ValidateNumberOfCVFoldsArg(numberOfCVFolds);

bool IsValidCvFolds(uint folds) => folds >= 2;

try { UserInputValidationUtil.ValidateNumberOfCVFoldsArg(folds); }
catch (ArgumentException ex) { folds = 5; logger.LogWarning(ex.Message + "; defaulting to 5"); }

When it happens

Trigger: Calling an AutoML experiment API (e.g. CreateExperiment/Execute with cross-validation) passing numberOfCVFolds = 0 or 1, typically from user-supplied settings or config defaults.

Common situations: Users configuring fold counts from app settings that default to 0/1; off-by-one when interpreting 'number of validation runs'; UI inputs allowing 1.

Related errors

• Please provide a message with content.
• Expected either {0} or {1} to be provided
• {nameof(ChannelMessageKind)}.{e.Kind} is not yet implemented
• Training failed with the exception: {_history.Last().Excepti
• Metric {typeof(TMetrics)} not implemented

AI-assisted analysis of dotnet/machinelearning@7b76e69cf9 (2026-09-11). Data as JSON: /api/errors/0aa7e8d4e3891d77. Report an issue: GitHub.