THE RISE OF THE LOTTERY HEROES: WHY ZERO-SHOT PRUNING IS HARD - Laboratoire Traitement et Communication de l'Information Access content directly
Conference Papers Year : 2022

THE RISE OF THE LOTTERY HEROES: WHY ZERO-SHOT PRUNING IS HARD

Abstract

Recent advances in deep learning optimization showed that just a subset of parameters are really necessary to successfully train a model. Potentially, such a discovery has broad impact from the theory to application; however, it is known that finding these trainable sub-network is a typically costly process. This inhibits practical applications: can the learned sub-graph structures in deep learning models be found at training time? In this work we explore such a possibility, observing and motivating why common approaches typically fail in the extreme scenarios of interest, and proposing an approach which potentially enables training with reduced computational effort. The experiments on either challenging architectures and datasets suggest the algorithmic accessibility over such a computational gain, and in particular a trade-off between accuracy achieved and training complexity deployed emerges.
Fichier principal
Vignette du fichier
ICIP_22_LotteryHeroes-1.pdf (813.83 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03766740 , version 1 (01-09-2022)

Identifiers

Cite

Enzo Tartaglione. THE RISE OF THE LOTTERY HEROES: WHY ZERO-SHOT PRUNING IS HARD. IEEE International Conference on Image Processing (ICIP 22), Oct 2022, Bordeaux, France. ⟨hal-03766740⟩
139 View
33 Download

Altmetric

Share

Gmail Facebook X LinkedIn More