课程定位
CS 329S 是一门项目制的 ML 系统课,讲的是“模型之外”的工程:数据怎么管、模型怎么选、训练怎么扩、上线怎么监控、团队怎么协作。它也会讨论隐私、公平、安全等真实系统必须面对的问题。
主要内容
- 数据工程:数据管理、标注、版本、质量
- 模型选择:baseline、评估、误差分析
- 训练与扩展:分布式、成本、实验管理
- 部署:服务化、批/流、延迟与吞吐
- 监控与迭代:drift、回滚、A/B
- 人与社会:隐私、公平、安全、团队协作
- 项目:设计并实现一个 ML 系统
课程大纲 / 课件
以 Winter 2022 公开 syllabus 整理;slides / notes 在 syllabus 页。
先修
CS 229/230/231N/224N 至少一门;熟悉至少一个框架(PyTorch/TF/JAX)。
考核与工作量
项目为主 + 作业/阅读。具体以当季官网为准。
适合谁
想从“调模型”走向“做系统”的 ML engineer;想补工程化能力的研究者;技术产品经理。
学习建议
配合《Designing Machine Learning Systems》。项目要包含监控和回滚方案,不要只交模型。
相关课程
官方课程描述(英文原文)
This project-based course covers the iterative process for designing, developing, and deploying machine learning systems. It focuses on systems that require massive datasets and compute resources, such as large neural networks. Students will learn about data management, data engineering, approaches to model selection, training, scaling, how to continually monitor and deploy changes to ML systems, as well as the human side of ML projects. In the process, students will learn about important issues including privacy, fairness, and security. Pre-requisites: At least one of the following; CS229, CS230, CS231N, CS224N or equivalent. Students should have a good understanding of machine learning algorithms and should be familiar with at least one framework such as TensorFlow, PyTorch, JAX.