← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8743.a · 生成模型:GPT-6

镜像体积和攻击面要看实际内容

理解等级:中级

包或环境含不需要组件,会增加下载、存储及潜在漏洞暴露。应按实际依赖裁剪并检查,不以工具名字断言Conda环境总是数GB或任何Poetry环境都更安全;开发与生产差异也需有意管理。

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

      unnecessarily large attack surface area. If we were to use conda for development and another dependency manager for production, it would introduce: (i) training-serving asymmetry, and (ii) a fork in your dependency management solution, both of which need to be maintained.

    参考来源

    - 位置:《Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)》第03章《依赖管理的原则与工具》(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/ch03.html) - 结论依据:摘录中关于镜像体积和攻击面要看实际内容的定义、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:unnecessarily large attack surface area. If we were to use conda for development and another dependency manager for production, it would introduce: (i) training-serving asymmetry, and (ii) a fork in your dependency management solution, both of which need to be maintained.