Litcius/Paper detail

Towards End-to-End Optimization of LLM-based Applications with Ayo

Xin Tan, Yimin Jiang, Yitao Yang, Hong Xu

20256 citationsDOIOpen Access PDF

Abstract

Large language model (LLM)-based applications consist of both LLM and non-LLM components, each contributing to the end-to-end latency. Despite great efforts to optimize LLM inference, end-to-end workflow optimization has been overlooked. Existing frameworks employ coarse-grained orchestration with task modules, which confines optimizations to within each module and yields suboptimal scheduling decisions.

Topics & Concepts

End-to-end principleComputer scienceDead endMathematicsArtificial intelligenceFlow (mathematics)GeometryAdvanced Data Storage TechnologiesParallel Computing and Optimization TechniquesSoftware Testing and Debugging Techniques