Documentation
¶
Index ¶
- type RagasLLMJudge
- func (j *RagasLLMJudge) EvaluateAnswerRelevance(ctx context.Context, query string, answer string) (float32, string, error)
- func (j *RagasLLMJudge) EvaluateContextPrecision(ctx context.Context, query string, chunks []*entity.Chunk) (float32, string, error)
- func (j *RagasLLMJudge) EvaluateFaithfulness(ctx context.Context, query string, chunks []*entity.Chunk, answer string) (float32, string, error)
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type RagasLLMJudge ¶
type RagasLLMJudge struct {
// contains filtered or unexported fields
}
RagasLLMJudge implements the LLMJudge interface using standard RAGAS-style prompts. It leverages a strong LLM (like GPT-4) to grade the pipeline's output.
func NewRagasLLMJudge ¶
func NewRagasLLMJudge(judgeLLM chat.Client) *RagasLLMJudge
func (*RagasLLMJudge) EvaluateAnswerRelevance ¶
func (j *RagasLLMJudge) EvaluateAnswerRelevance(ctx context.Context, query string, answer string) (float32, string, error)
EvaluateAnswerRelevance checks if the answer actually answers the user's question.
Source Files
Click to show internal directories.
Click to hide internal directories.