Namespace: Moderation
A scorer that uses OpenAI's moderation API to determine if AI response contains ANY flagged content.
Param
Param
Optional. Threshold to use to determine whether content has exceeded threshold. By
default, it uses OpenAI's default. (Using flagged from the response payload.)
Param
Optional. Specific categories to look for. If not set, all categories will be considered.
Functions
partial
▸ partial<T>(args): Scorer<string, Omit<{ threshold?: number } & OpenAIAuth, T> & Partial<Pick<{ threshold?: number } & OpenAIAuth, T>>>
Type parameters
| Name | Type |
|---|---|
T | extends keyof OpenAIAuth | "threshold" |
Parameters
| Name | Type |
|---|---|
args | { [K in keyof OpenAIAuth | "threshold"]: (Object & OpenAIAuth)[K] } |
Returns
Scorer<string, Omit<{ threshold?: number } & OpenAIAuth, T> & Partial<Pick<{ threshold?: number } & OpenAIAuth, T>>>