用实测数据替代猜测：SageMaker AI 上 TTS 自动扩缩目标设定
原标题：Stop guessing auto scaling targets for TTS on Amazon SageMaker AI
AI 摘要
AWS 机器学习博客发布技术文章，介绍在 Amazon SageMaker AI 上部署 Kokoro-82M 文本转语音（TTS）模型的方法。文章指出，TTS 工作负载不适合用默认自动扩缩指标，因此改用 SageMaker 原生发布的 ConcurrentRequestsPerModel 指标，并通过实测两个参数——单实例在 SLO 内可承受的最大并发 Q_max 和一次扩容端到端耗时 T_total——用算术推导出扩容阈值、缩容阈值及冷却时间，从而替代凭猜测设定的扩缩目标。
正文节选
Stop guessing auto scaling targets for TTS on Amazon SageMaker AI A traffic spike hits your self-hosted text-to-speech (TTS) endpoint, and requests start queueing. Latency climbs past your service level objective (SLO) right when the most callers are listening. The auto scaling policy behind that endpoint was supposed to prevent this, but the target value it scales on was a guess, and guesses like that fail silently until real traffic proves them wrong. Get that number wrong, and it’s paid for i