知识卡片
自主团队仍需管理重复建设
端到端团队减少交接,却可能各自重建训练基础设施等公共能力,增加首发和长期维护成本。让重复需求可见,再决定是否建立共同平台或赋能机制;自主性与共享方向可以兼容,不能用全部集中化作为唯一答案。
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
- · 高效机器学习组织
原书出处 · 1 处
- · 查看原书依据
both teams have independently spent significant effort in solving similar problems. For example, to solve the problem of provisioning scalable ML training infrastructure, Eduardo’s team wrote some intricate DevOps scripts to provision and tear down GPU instances. Unknown to Eduardo, Ellie’s team has integrated a GPU-as-a-service offering.
参考来源
- 位置:《Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)》第11章《高效机器学习组织》(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/ch11.html)
- 结论依据:摘录中关于自主团队仍需管理重复建设的定义、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:both teams have independently spent significant effort in solving similar problems. For example, to solve the problem of provisioning scalable ML training infrastructure, Eduardo’s team wrote some intricate DevOps scripts to provision and tear down GPU instances. Unknown to Eduardo, Ellie’s team has integrated a GPU-as-a-service offering.