查看更多>>摘要：From Washington, D.C., NewsRx journali sts report that a patent application by the inventors Ahn, Junwhan (San Jose, CA , US); Baeuml, Martin (Wollerau, CH); Bailey, Alexander (Wollerau, CH); Beirami, Ahmad (New York, NY, US); Chen, Zhifeng (Sunnyvale, CA, US); Huang, Yanping (Mo untain View, CA, US); Jia, Wenhao (Saratoga, CA, US); Lan, Chang (Kirkland, WA, US); Mudgal, Sidharth (Mountain View, CA, US); Schelin, Leif (Zurich, CH); Stroh man, Trevor (Sunnyvale, CA, US); Taropa, Emanuel (Los Altos, CA, US); Xu, Yuanzh ong (Mountain View, CA, US); Zheng, Yanyan (Palo Alto, CA, US), filed on April 1 9, 2023, was made available online on September 19, 2024. No assignee for this patent application has been made. News editors obtained the following quote from the background information suppli ed by the inventors: "Large language models (LLMs) are particular types of machi ne learning models that can perform various natural language processing (NLP) ta sks, such as language generation, machine translation, and questionanswering. T hese LLMs are typically trained on enormous amounts of diverse data including da ta from, but not limited to, webpages, electronic books, software code, electron ic news articles, and machine translation data. Accordingly, these LLMs leverage the underlying data on which they were trained in performing these various NLP tasks. For instance, in performing a language generation task, these LLMs can pr ocess a natural language (NL) based input that is received from a client device, and generate a NL based output that is responsive to the NL based input and tha t is to be rendered at the client device. However, in generating the NL based ou tput utilizing these LLMs, additional latency is introduced that may not be pres ent absent utilizing these LLMs. This additional latency can prolong user intera ctions with these LLMs and detract from a user experience with these LLMs. Accor dingly, there is a need in the art for reducing latency in utilizing these LLMs. "
NETL
NSTL