Skip to content

Latest commit

 

History

History
15 lines (8 loc) · 429 Bytes

File metadata and controls

15 lines (8 loc) · 429 Bytes

Evaluation Utilities

Standalone evaluation utilities for question generation and LLM judging.

These utilities can be used independently of the knowledge base. For KB-integrated evaluation workflows, see the kb.eval module.

Question Generation

::: kb_mcp.eval_utils.generate_qa_pairs_keypoint

::: kb_mcp.eval_utils.generate_qa_pairs_persona

Answer Judging

::: kb_mcp.eval_utils.llm_judge_answer