Always return a company's template JSON format.
No new factual knowledge is required, and a short experiment is preferred.
Which method should be considered first?
For format and behavior changes, first try explicit instructions and examples; fine-tuning changes parameters and requires a training cycle.
Detailed explanation
A format change can often be tested quickly with prompt engineering.
This is excessive for a simple format requirement.
Temperature affects randomness and does not replace schema validation.
Generated output still needs schema and business-rule checks.
Check official documentation
Compare the following points with the official documentation.
Amazon Bedrock公式Prompt management・モデルカスタマイズの説明を確認する。
Official references
What to confirm
形式適応にプロンプト、パラメータ更新に微調整という使い分けを説明できる。
Key points
- Prompting
- Fine-tuning
- Format validation
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
- Compare the definition with the AIF-C01 objective and the service documentation before applying it.
- Keep the concept separate from a particular answer choice or implementation step.
Foundation review
Behavior versus parameters
Prompting changes the request context; fine-tuning changes the model and its release artifact.
Output contract
Sampling settings alone do not guarantee that a response follows a machine-readable format.