|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/mlops-exhaustive-guide-aws-gcp-apple-cases/
课程评论：没有评论
课程名称：MLOps：机器学习部署（AWS、GCP与Apple） 课程概述： 该课程专注于机器学习的最佳实践，包括自动化和模型部署。无论你是理论学习者还是实践动手者，课程都将帮助你建立扎实的理论基础，并通过实践运用新学的概念来加强技能。MLOps在过去六年中帮助我在IT、食品和旅游行业成功实现了多个机器学习项目，通过现代云计算和开源软件，我解决了复杂的商业挑战，并推动了职业发展。 我们将覆盖多种现代MLOps技术，如AWS Sagemaker、Kubeflow、Azure机器学习、MLflow、GCP Vertex AI等，并探讨它们如何协同工作。课程中还将介绍数据漂移这一常见问题，学习如何及时发现和减轻其对模型性能的影响，包括使用简单的可视化方法和EvidentlyAI工具。 此外，课程还将教授如何使用MLFlow来跟踪机器学习实验，讲解如何在本地和通过云服务dagshub进行操作。最后，学员将学习如何将基于ML的Web应用（Flask微服务和UI）部署到AWS云。 更新内容： - 2023年8月：第3章和第7章新增额外材料 - 2023年10月：新增“ML Ops市场概述”章节，了解市场统计数据、趋势、薪资和职位预期 - 2023年11月：新增“数据漂移”部分（增加1小时内容），学习如何发现和处理数据漂移 - 2023年12月：课程结构更新，新增“你将学到什么”视频 - 2024年2月：新增“MLFlow”部分及“基于ML的Web应用部署到AWS”章节 学习目标： 完成课程后，你将能够： - 设置CI/CD管道 - 将ML模型打包为Docker容器 - 在本地和云端运行AutoML - 训练适用于Apple设备的ML模型 - 使用MLflow框架监控和记录ML实验 - 在AWS SageMaker中设置和管理MLOps管道 - 在GCP VertexAI中操作模型注册表和端点 - 使用EvidentlyAI发现数据漂移并进行高级模型性能分析 - 结合MLFlow使用EvidentlyAI跟踪ML实验 - 使用dagshub这一云端ML跟踪环境 - 部署基于ML的Web应用到AWS云 欢迎加入这个有趣且与行业密切相关的课程，提升你的MLOps和云计算能力！课程根据学生反馈定期更新，添加新讲座和实践案例。
xxxxxxxxxxxx[Course Updates]:- 08.2023: + New Extra materials in Chapters #3 & #7- 10.2023: + "MLOps Market Overview" chapter. Learn key Market stats, trends + Salaries & Role Expectations- 11.2023: + "Data Drifts" section (+1 hour of content). Learn to discover data drifts & deal with them.EvidentlyAI + MLFlow integration- 12.2023: Course Structure Updates + new "What you will learn" video- 02.2024: + "MLFLow" Section + "ML-powered Web App Deployment to AWS" SectionxxxxxxxxxxxWould you like to learn best practices of Automation & ML models Deployment?Maybe you would also like to practice doing it?You've come to the right place!There's no better way to achieve that than by creating a strong theoretical foundation and getting hands dirty by applying newly learnt concepts in practice straight away!MLOps has been helping me automate & roll out robust, easily maintainable and state-of-the-art ML in IT, Food and Travel industries over the last 6 years.With the help of modern Cloud Computing and open source software I've brought live dozens of ML research projects, successfully solved very complex Business challenges and even changed the country where I live & work!There are many different technologies powering modern ml ops. Some of them are: AWS Sagemaker, Kubeflow, Azure machine learning, mlflow, GCP Vertex AI, dvc etc. We will cover many of them and see how they work together.The course will also teach you about Data Drifts: a common issue arising in the world of Machine Learning models. We will learn what these are, how to discover them in a timely manner and what actions to take to mitigate their effect on model's performance.We will use a variety techniques for that: from simple visual analysis using histograms & box plots all the way to learning EvidentlyAI.Additionally, we will look into using MLFlow to track your ML experiments. There're 2 ways to do that covered in the Course: using MLFlow locally on your machine and using SaaS service called dagshub.Finally, we will deploy an ML-powered Web App (Flask micro-service + UI) to the internet using AWS Cloud.Join me in this fun and Industry-shaped course to get new skills and improve your MLOps & Cloud acumen!By the end of this course you will be able to:Set up CI & CD pipelinesPackage ML models into DockerRun AutoML locally & in the CloudTrain ML models for Apple devicesMonitor and Log ML experiments with mlflow frameworkSet up and manage MLOps pipelines in AWS SageMakerOperate Model Registry & Endpoints in GCP VertexAIUse EvidentlyAI to discover Data Drifts and conduct advanced analysis of ML model performanceUse EvidentlyAI together with MLFlow to track ML experimentsUse MLFLow to track your ML ExperimentsLearn dagshub - a Cloud-based ML tracking environment building on top of MLFlowDeploy ML-powered Web App with UI to AWS CloudBoost your Career and MLOps studying efficiencyThe course isn't static! I collect students' feedback and periodically update the materials: add new lectures and practice cases!
