Entropy Aware Training for Fast and Accurate Distributed GNN

Several distributed frameworks have been developed to scale Graph Neural Networks (GNNs) on billion-size graphs. On several benchmarks, we observe that the graph partitions generated by these frameworks have heterogeneous data distributions and class imbalance, affecting convergence, and resulting i...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	arXiv.org 2023-11
Hauptverfasser:	Deshmukh, Dhruv, Gupta, Gagan Raj, Chawla, Manisha, Jatala, Vishwesh, Haldar, Anirban
Format:	Artikel
Sprache:	eng
Schlagworte:	Algorithms Benchmarks Convergence Entropy Graph neural networks Partitions (mathematics) Training
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!