通过模型路由AI代理优化成本和用户价值.pdf

《通过模型路由AI代理优化成本和用户价值.pdf》由会员分享，可在线阅读，更多相关《通过模型路由AI代理优化成本和用户价值.pdf（50页珍藏版）》请在三个皮匠报告上搜索。

1、Optimize Cost and User Value Through Model Routing AI AgentAditya GautamForward-looking StatementThis presentation has been prepared for informational purposes only.The information set forth herein does not purport to be complete or contain all relevant information.Statements contained herein are ma

2、de as of the date of this presentation unless stated otherwise.This presentation and the accompanying oral commentary may contain forward-looking statements.In some cases,forward-looking statements can be identified by terms such as“may”,“will”,“should”,“expects”,“plans”,“anticipates”,“could”,“inten

3、ds”,“projects”,“believes”,“estimates”,“predicts”,or“continue”,or the negative of these words or other similar terms or expressions that concern Databricks expectations,strategy,plans,or intentions.Forward-looking statements are based on information available at the time those statements are made and

4、 are inherently subject to risks and uncertainties that could cause actual results to differ materially from those expressed in or suggested by the forward-looking statements.Forward-looking statements should not be read as a guarantee of future performance or outcomes.Except as required by law,Data

5、bricks does not undertake any obligation to publicly update or revise any forward-looking statement,whether as a result of new information,future developments or otherwise.3AgendaLLM components and cost variabilityLlama and open source overviewLLM components and cost variabilityLLM Routing:Goal,Adva

6、ntages,CategoriesLLM Routing Technical DetailsCategories-Rule,Rewards and Classifier based methods.Features,Data,Loss function and trainingOffline/Online evaluation,measuring cost savings and performance LLM Routing through DatabricksMosaic AI Agent and Gateway,Model Serving,MLFlow,Unity Catalog etc