quiz bu lëkkale ak tegtal · Diggante Niveau

Jawriñu jàngat RAG ak quiz RAGAS

Saytu li nga xam ci jàngat RAG: njub, njariñu tontu, njubte ak fàttaliku, ak ni àttekati LLM yi di leen joxee poñ.

Yooni njiit yu jëm ciRAG ak Ragas
Laaj 1 de 8

Ban metrik mooy saytu ndax bépp laaj bu nekk ci tontu bi amna lu muy tekki ci muy tekki?

Xoolaatal laaj bu nekk8 laaj

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

  1. Ban metrik mooy saytu ndax bépp laaj bu nekk ci tontu bi amna lu muy tekki ci muy tekki?
    • Answer relevance
    • Context recall
    • Faithfulness
    • Context precision
  2. Ci metrics yii, yan ci ñoom moo soxla tontub royuwaay bu dëggu ci laaj bu nekk ci test bi?
    • Context recall
    • Faithfulness
    • Answer relevance
    • Response latency
  3. Lan mooy natt njubte ci contexte?
    • How quickly chunks are retrieved
    • Whether the answer is short
    • How many documents are in the index
    • Whether the relevant retrieved chunks are ranked near the top
  4. naka la RAGAS di xayma njariñu tontu?
    • By counting words shared by answer and context
    • By generating questions the answer would respond to and comparing them to the original question with embeddings
    • By asking the user to rate each answer
    • By checking the answer against a reference with exact string match
  5. Sistem bi dafay wane muy fàttaliku contexte bu bari waaye muy wane njubte ci contexte bu néew. Lan mooy li gëna mëna saafara?
    • Improve ranking, for example by adding a reranker
    • Switch to a smaller generation model
    • Remove the reference answers
    • Lower the number of test questions
  6. Why can a perfectly faithful answer still be wrong?
    • Faithfulness only checks grammar
    • Faithfulness ignores the retrieved context
    • It can faithfully repeat an outdated or incorrect document
    • Faithful answers are always too short
  7. Why should teams pin the judge model and settings when comparing RAG systems?
    • Pinning makes the judge free to run
    • Judges only work with one vector database
    • Pinning removes the need for test questions
    • LLM-judge scores vary with the judge model and prompt, so comparisons need identical settings
  8. Retrieval returns the right passages, but the answer adds unsupported details. Where should the team look first?
    • The chunking strategy
    • The prompt or the generation model
    • The embedding model
    • The number of documents indexed