2.1Basic
Generate fixed classifications and creative drafts.
The reproducibility requirements differ.
Choose two settings.
× Incorrect
Use low randomness and reproducibility controls for fixed tasks, and evaluate quality and safety when increasing diversity for creative tasks.
Detailed explanation
CorrectEvaluate low randomness and fixed-seed controls for deterministic tasks
Output variation can be reduced.
Output variation can be reduced.
CorrectEvaluate quality and safety when increasing temperature or candidate range
Diversity is balanced with risk.
Diversity is balanced with risk.
IncorrectHigher temperature always improves factuality
Diversity and accuracy are different.
Diversity and accuracy are different.
IncorrectDifferent random settings always produce identical output
Generation can vary.
Generation can vary.
Try it yourself
An example you can run in a temporary verification environment.
AWS公式AIF-C01 Domain 2.1の生成パラメータ、ランダム性、モデル評価を確認する。Expected result
定型処理と創作処理で生成の揺らぎを要件に合わせて調整できる。Key points
- Temperature
- Diversity
- Reproducibility
Notes
- Environment: AWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
- Command output formatting can vary slightly by distribution or tool version.
- Run the example in a temporary directory or process when possible.
Foundation review
Read the scope first
Check whether the command acts on the current shell, a new process, an existing process, or a file.
Verify the observable result
Use the supplied command and compare the output with the expected result.
Question IDAWS-AIF-337
EnvironmentAWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
Last technical review2026-08-20
Report an error or rights issue