Robotics & Machine Learning Daily News2024,Issue(Nov.25) :99-102.

Researchers Submit Patent Application, 'Compilation System And Method', for Appr oval (USPTO 20240370239)

研究人员为APR OVAL(USPTO 20240370239)提交专利申请“汇编系统和方法”

Robotics & Machine Learning Daily News2024,Issue(Nov.25) :99-102.

Researchers Submit Patent Application, 'Compilation System And Method', for Appr oval (USPTO 20240370239)

研究人员为APR OVAL(USPTO 20240370239)提交专利申请“汇编系统和方法”

扫码查看

摘要

新闻编辑从发明人提供的背景信息中获得了以下引文:“机器学习中常用的编程框架有两种:(ML):a)老派基于图形的框架,如TensorFlow 1和B)渴望执行的框架,如PyTorch或TensorFlow2. 急切地按顺序执行源程序的步骤,并将每个操作从源程序中分派给源程序将立即执行,当他们被计数红色。一种基于图的方法另一方面,根据一个表示关系的图形,对整个程序进行动态编译在源程序中的各种变量和操作之间,然后只编译一次执行编译程序。Eager和基于图形的编程框架也可以扩展而不仅仅是机器学习。

Abstract

News editors obtained the following quote from the background information suppli ed by the inventors:“There are two kinds of programming framework commonly used in machine learning (ML): a) old-schoolgraph-based frameworks like TensorFlow 1, and b) eager execution frameworks like PyTorch or TensorFlow2. Eager executi on steps through the source program sequentially and dispatches each operation f rom thesource program to be executed immediately, as-and-when they are encounte red. A graph-based approachon the other hand compiles the entire program in adv ance based on a graph representing relationshipsbetween the various variables a nd operations within the source program, and then only once compiledexecutes th e compiled program. Eager and graph-based programming frameworks can also be ext endedto other applications, not just machine learning.

Key words

Cyborgs/Emerging Technologies/Machine Learning/Patent Application

引用本文复制引用

出版年

2024
Robotics & Machine Learning Daily News

Robotics & Machine Learning Daily News

ISSN:
段落导航相关论文