借助 Amazon SageMaker AI 简化 AI 模型开发生命周期.pdf

《借助 Amazon SageMaker AI 简化 AI 模型开发生命周期.pdf》由会员分享，可在线阅读，更多相关《借助 Amazon SageMaker AI 简化 AI 模型开发生命周期.pdf（39页珍藏版）》请在三个皮匠报告上搜索。

1、A I M 3 6 4Streamline AI model development lifecycle with Amazon SageMaker AIKhushboo SrivastavaSr.Product Manager TechnicalAmazon Web ServicesBruno PistoneSr.WW Specialist SAGenAIAmazon Web ServicesManikandan ParamasivanSenior Staff Architect-Data,ML&AIKOHO 2025,Amazon Web Services,Inc.or its affil

2、iates.All rights reserved.$202BGenAI spending by 20281$7TIncrease in global GDP(7%)2STATE OF GENERATIVE AI IN THE ENTERPRISES32%overall AI spending29%CAGRProductivity increases 1.5 percentage points over next 10 years1.IDC 2.Goldman Sachs 2025,Amazon Web Services,Inc.or its affiliates.All rights res

3、erved.of enterprises are actively advancing their generative AI initiatives in 2025,with 92%planning to increase investment by 2027https:/aloa.co/ai/resources/industry-insights/ai-statsof organizations choose AI models that are 13B parameters or smaller,suggesting preference for customizable,cost-ef

4、fective models rather than larger generic oneshttps:/ organizations now use AI in at least one business functionhttps:/ AI adoption momentum is real 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.Disparate and disconnected ML tools increases time to marketIsolation between team m

5、embers reduces productivity and collaborationChallenging to govern AI and ML projects efficientlyTraining infrastructure provision and management To build,train,and deploy AI models at scale is challenging 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.C L A S S I C M LG E N A ID

6、ata prepPrepare data assets and pipelines,manage data quality and biasBuildExperiment and automate the execution of build pipelinesTrainTrain ML models at scale with automation DeployAutomate the execution of deploy pipelines into productionA U T O M A T E W I T H A I O P S A N D G O V E R N A N C E