|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/devops-to-mlops-bootcamp/
课程评论：没有评论
课程名称：终极DevOps到MLOps培训营 - 构建ML CI/CD管道 课程概述： 本培训营是一门实践导向的课程，旨在帮助DevOps工程师和基础设施专业人士过渡到日益增长的MLOps领域。随着AI/ML迅速成为现代应用程序的一部分，MLOps成为了机器学习模型与生产系统之间的重要桥梁。在本课程中，学员将围绕一个真实世界的回归案例——预测房价，完成从数据处理到在Kubernetes上生产部署的整个过程。 课程内容包括： - 使用Docker和MLFlow设置实验跟踪环境； - 理解机器学习生命周期，掌握数据工程、特征工程和模型实验的实操技能； - 使用FastAPI打包模型，并与基于Streamlit的用户界面一起部署； - 编写GitHub Actions工作流，自动化ML管道以实现持续集成，并使用DockerHub推送模型容器； - 建立可扩展的推理基础设施，利用Kubernetes暴露服务，并通过服务发现连接前端和后端； - 深入了解生产级模型服务（使用Seldon Core）并使用Prometheus和Grafana仪表板监控部署； - 探索基于GitOps的持续交付，通过ArgoCD管理和自动化Kubernetes集群的变更。 完成课程后，学员将掌握运作和自动化机器学习工作流的知识与实践经验，为MLOps和AI平台工程师角色的工作做好准备。
This hands-on bootcamp is designed to help DevOps Engineers and infrastructure professionals transition into the growing field of MLOps. With AI/ML rapidly becoming an integral part of modern applications, MLOps has emerged as the critical bridge between machine learning models and production systems.In this course, you will work on a real-world regression use case - predicting house prices - and take it all the way from data processing to production deployment on Kubernetes. You'll start by setting up your environment using Docker and MLFlow for tracking experiments. You'll understand the machine learning lifecycle and get hands-on experience with data engineering, feature engineering, and model experimentation using Jupyter notebooks.Next, you'll package the model with FastAPI and deploy it alongside a Streamlit-based UI. You'll write GitHub Actions workflows to automate your ML pipeline for CI and use DockerHub to push your model containers.In the later stages, you'll build a scalable inference infrastructure using Kubernetes, expose services, and connect frontends and backends using service discovery. You'll explore production-grade model serving with Seldon Core and monitor your deployments with Prometheus and Grafana dashboards.Finally, you'll explore GitOps-based continuous delivery using ArgoCD to manage and deploy changes to your Kubernetes cluster in a clean and automated way.By the end of this course, you'll be equipped with the knowledge and hands-on experience to operate and automate machine learning workflows using DevOps practices - making you job-ready for MLOps and AI Platform Engineering roles.
