Adapt a model to legal terminology and a required output format.
Decide whether examples are enough or adaptation is needed.
Choose two appropriate considerations.
First measure a baseline with instructions, examples, and a schema; consider adaptation data and reevaluation only when that is insufficient.
Detailed explanation
Unnecessary cost and data governance work are added.
The choice is based on need and additional cost.
Improvement, regression, and safety cannot be compared.
The benefit available without new training data becomes clear.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 3.1のプロンプト、モデル適応、評価を確認する。
Official references
What to confirm
プロンプト改善と追加適応を費用・品質・データ要件で使い分けられる。
Key points
- Baseline
- Examples
- Adaptation
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
Baseline
A reference result measured before a change under the same evaluation conditions.
Model adaptation
Adjusting prompts, parameters, or model components for a task or domain.