AWS Certification Exam AI Practitioner (Practice Questions) set15
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 21 / 65
Which of the following is the appropriate cost trade-off when using parameter-efficient methods such as LoRA or Adapters for fine-tuning?
①These methods always improve model performance without limit.
②While it can reduce the resources required for training and storage, there are limits to achieving extreme performance improvements.
③Parameter efficiency automatically solves data privacy problems.
④These methods inevitably worsen inference latency.
⑤Parameter efficiency eliminates the upfront cost of fine-tuning.