Jittor: a novel deep learning framework with meta-operators and unified graph execution

This paper introduces Jittor, a fully just-in-time (JIT) compiled deep learning framework. With JIT compilation, we can achieve higher performance while making systems highly customizable. Jittor provides classes of Numpy-like operators, which we call meta-operators. A deep learning model built upon...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	Science China. Information sciences 2020-12, Vol.63 (12), p.222103, Article 222103
Hauptverfasser:	Hu, Shi-Min, Liang, Dun, Yang, Guo-Ye, Yang, Guo-Wei, Zhou, Wen-Yang
Format:	Artikel
Sprache:	eng
Schlagworte:	Algorithms Back propagation Computer Science Deep learning Graphs Information Systems and Communication Service Iterative methods Libraries Operators Optimization techniques Propagation Research Paper Science
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!