知识卡片
ML 的 SLO 与 SLI
SLO 是对具体度量(SLI)的预设阈值,定义系统是否达标。ML 系统的 SLO 更难定,因为数据或世界微小变化就能悄然降级模型。可从关注点切分入手:service/training/应用三分;传统 golden signals 与 ML 内部信号二分;以及 ML 增强应用自身的业务目的 SLO(如点击率、可归因收入),并按客户分片(地域/类型)度量。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- Introduction · 查看原书依据
Defining and Measuring SLOs Service-level objectives (SLOs) are predefined thresholds for specific measurements, often known as service-level indicators (SLIs), that define whether the system is performing according to requirements.
参考来源
- 位置:《Reliable Machine Learning》第01章《Introduction》(源文件:extracted/Reliable Machine Learning/_epub-src/OEBPS/ch01.xhtml)
- 结论依据:原文给出 SLO/SLI 定义,并说明 ML 下分 serving/training/应用分别设 SLO 的思路。
- 原始内容:Defining and Measuring SLOs Service-level objectives (SLOs) are predefined thresholds for specific measurements, often known as service-level indicators (SLIs), that define whether the system is performing according to requirements.