知识卡片

普通读书笔记卡 · 8787.b · 生成模型:GPT-6

小批量由质量门禁支撑

理解等级:中级

小批量把一次变更控制在容易理解和验证的范围,沿途用数据测试、模型评价、预生产部署和部署后检查持续反馈。绿色构建只代表通过已定义的门禁;发布信心还取决于门禁是否覆盖相关风险,而不是颜色本身。

这张卡片看不懂?

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

    • · 查看原书依据

      Along the path to production, we instrument multiple quality gates to give us rapid and comprehensive feedback on the quality of our changes. In the ML context, quality gates include automated tests (for software, data, and model), deployments to preproduction environments, post-deployment tests, and other fitness functions that we described in Chapter 6. When the quality gates are comprehensive, we can confidently deploy any green

    参考来源

    - 位置:《Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)》第09章《MLOps与持续交付》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)/_epub-src/OEBPS/ch09.html) - 结论依据:摘录中关于小批量由质量门禁支撑的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:Along the path to production, we instrument multiple quality gates to give us rapid and comprehensive feedback on the quality of our changes. In the ML context, quality gates include automated tests (for software, data, and model), deployments to preproduction environments, post-deployment tests, and other fitness functions that we described in Chapter 6. When the quality gates are comprehensive, we can confidently deploy any green