AWS Certification Exam AI Practitioner (Practice Questions) set13

問題 50 / 65

When labeled data are very scarce and onboard computing resources are limited, what is the best way to efficiently adapt a foundation model to a marine classification task?

Full fine-tuning that completely retrains the model.
Parameter-efficient fine-tuning methods such as LoRA and adapters.
Increase the number of labels through instruction tuning.
Conduct large-scale continuous pre-training.
Use a simple rule-based filter.

当サイトでは、ユーザー体験の向上を目的としてCookieを使用しています。サイトの利用を継続することで、Cookieの使用に同意したものとみなされます。