गाइड से जुड़ी प्रश्नोत्तरी · मध्यम स्तर
शीघ्र संस्करण और प्रतिगमन परीक्षण प्रश्नोत्तरी
परीक्षण करता है कि संस्करणित प्रॉम्प्ट कॉन्फ़िगरेशन में क्या शामिल है, प्रतिगमन सुइट्स गैर-नियतिवाद को कैसे संभालते हैं, और मॉडल स्नैपशॉट और एलएलएम न्यायाधीशों को पिनिंग की आवश्यकता क्यों है।
प्रश्न 1 का 8
Which set of items should be versioned together as one prompt configuration?
प्रत्येक प्रश्न का पूर्वावलोकन करें8 प्रश्न
जैसे ही आप प्रश्नोत्तरी लेते हैं उत्तर और स्पष्टीकरण दिखाई देते हैं।
- Which set of items should be versioned together as one prompt configuration?
- Only the prompt wording
- Only the model name
- Prompt template, model identifier, sampling parameters and tool schemas
- Only the user inputs from production
- Why pin a dated model snapshot instead of a generic latest alias?
- So model changes happen only when you deliberately switch and re-test
- Because aliases are always slower
- Because snapshots are free
- Because aliases do not support system prompts
- What does the guide say about temperature 0?
- It guarantees identical outputs every time
- It makes regression testing unnecessary
- It only affects the length of outputs
- It reduces randomness but does not guarantee identical outputs across runs
- How should a regression suite handle nondeterministic outputs?
- Require exact string matches
- Use thresholds and repeated samples instead of exact equality
- Ignore any failing cases
- Only test with a single example
- What is a strong source of new test cases for a golden set?
- Past production failures turned into test cases
- Random text from the web
- Only cases the current prompt already passes
- Outputs copied from a competitor
- If an LLM judge grades outputs, what should you do?
- Let the judge model change freely
- Trust every verdict without review
- Pin the judge's model and prompt and check it against human labels
- Use the same prompt under test as the judge
- Why log the prompt version ID with each production request?
- To reduce token costs
- To improve model accuracy
- To satisfy the model provider's terms
- To tie problems to a specific deploy and roll back quickly
- Why list individual cases that flipped from pass to fail rather than only the average score?
- Averages are always wrong
- An unchanged average can hide new breakage offset by gains elsewhere
- Flipped cases reduce API costs
- Only flipped cases are graded