Interface PreferenceOptimizationDataStats

Statistics computed for datasets used for preference optimization. This data type is not supported in Gemini API.

interface PreferenceOptimizationDataStats {
    scoresDistribution?: DatasetDistribution;
    scoreVariancePerExampleDistribution?: DatasetDistribution;
    totalBillableTokenCount?: string;
    tuningDatasetExampleCount?: string;
    tuningStepCount?: string;
    userDatasetExamples?: GeminiPreferenceExample[];
    userInputTokenDistribution?: DatasetDistribution;
    userOutputTokenDistribution?: DatasetDistribution;
}

Properties

scoresDistribution? scoreVariancePerExampleDistribution? totalBillableTokenCount? tuningDatasetExampleCount? tuningStepCount? userDatasetExamples? userInputTokenDistribution? userOutputTokenDistribution?

Properties

OptionalscoresDistribution

scoresDistribution?: DatasetDistribution

Output only. Dataset distributions for scores.

OptionalscoreVariancePerExampleDistribution

scoreVariancePerExampleDistribution?: DatasetDistribution

Output only. Dataset distributions for scores variance per example.

OptionaltotalBillableTokenCount

totalBillableTokenCount?: string

Output only. Number of billable tokens in the tuning dataset.

OptionaltuningDatasetExampleCount

tuningDatasetExampleCount?: string

Output only. Number of examples in the tuning dataset.

OptionaltuningStepCount

tuningStepCount?: string

Output only. Number of tuning steps for this Tuning Job.

OptionaluserDatasetExamples

userDatasetExamples?: GeminiPreferenceExample[]

Output only. Sample user examples in the training dataset.

OptionaluserInputTokenDistribution

userInputTokenDistribution?: DatasetDistribution

Output only. Dataset distributions for the user input tokens.

OptionaluserOutputTokenDistribution

userOutputTokenDistribution?: DatasetDistribution

Output only. Dataset distributions for the user output tokens.

Settings

On This Page

Properties