Type Alias EmbeddingDistance
EmbeddingDistance: { evaluatorType: "embedding_distance"; feedbackKey: undefined | string; formatEvaluatorInputs: EvaluatorInputFormatter; agentTools?: undefined | StructuredToolInterface[]; chainOptions?: undefined | Partial<Omit<LLMEvalChainInput<EvalOutputType, BaseLanguageModelInterface>, "llm">>; criteria?: undefined | CriteriaLike; distanceMetric?: undefined | EmbeddingDistanceType; embedding?: any; llm?: any; } Type declaration
evaluatorType: "embedding_distance"
feedbackKey: undefined | string
Optional
agentTools?: undefined | StructuredToolInterface[]
Optional
chainOptions?: undefined | Partial<Omit<LLMEvalChainInput<EvalOutputType, BaseLanguageModelInterface>, "llm">>
Optional
criteria?: undefined | CriteriaLike
Optional
embedding?: any
Optional
llm?: any
Configuration to load a "EmbeddingDistanceEvalChain" evaluator, which embeds distances to score semantic difference between a prediction and reference.