A team wants to prototype document summarization and question answering with a small amount of internal data.
It will not train a large model from scratch.
Choose two benefits of using a foundation model.
A pre-trained foundation model can shorten initial development and can be adapted with prompts, retrieval, or other methods; quality and confidentiality still require controls.
Detailed explanation
Confidentiality depends on access control, encryption, service settings, and policy.
Pre-training does not remove the need for evaluation and safeguards.
Prompts, retrieval, and fine-tuning can adapt a foundation model to a use case.
The team can use broad capabilities without building a base model from scratch.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01ガイドのDomain 2.1とAmazon Bedrock公式概要で基盤モデルの利用形態を確認する。
Official references
What to confirm
開発短縮と適応性を利点として挙げ、品質保証や機密性を自動保証と誤認しない。
Key points
- Pre-training
- Adaptation
- Benefits versus guarantees
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
Reuse
A pretrained model can provide a starting capability without collecting and training a task model from zero.
Capability boundary
Pretraining does not guarantee current knowledge, domain accuracy, safety, or suitability for a particular use.