Users submit text, images, and short audio.
Choose a model and evaluate its fit.
Choose two practices.
Confirm input and output modalities, language and quality requirements, safety behavior, latency, quotas, and cost with representative test cases.
Detailed explanation
A model must accept and produce the required data types.
A model must accept and produce the required data types.
Capability is broader than a model label.
Capability is broader than a model label.
Model capabilities differ.
Model capabilities differ.
Actual behavior and limits remain unknown.
Actual behavior and limits remain unknown.
Try it yourself
An example you can run in a temporary verification environment.
AWS公式AIF-C01 Domain 2.1の基盤モデル、マルチモーダル、評価を確認する。Expected result
モダリティ対応の有無と実運用品質を分けて評価できる。Key points
- Modality
- Evaluation
- Cost
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.