試験
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 20 / 65
Which operational policy is most appropriate for ensuring responsiveness while minimizing energy consumption in a traffic prediction service?
①Keep the model running at full capacity at all times to maintain low latency.
②Only perform batch processing at night and do not respond during the day.
③Dynamically schedule inference, for example by invoking lightweight models on demand.
④Always use the full-size model for all requests.
⑤Save all logs and re-run inference on them later.