Configuration options for Content Moderation training.

interface CustomizationFeatureContentModerationConfig {
    ConfidenceThreshold?: number;
}

Properties

ConfidenceThreshold?: number

The confidence level you plan to use to identify if unsafe content is present during inference.