NAACL 2022findings4 citations
Speeding Up Entmax
Maxat Tezekbayev, Vassilina Nikoulina, Matthias Gallé, Zhenisbek Assylbekov
Abstract
Softmax is the de facto standard for normalizing logits in modern neural networks for language processing. However, by producing a dense probability distribution each token in the vocabulary has a nonzero chance of being selected at each generation step, leading to a variety of reported problems in text generation. 𝛼-entmax of Peters et al. (2019) solves this problem, but is unfortunately slower than softmax. In this paper, we propose an alternative to 𝛼-entmax, which keeps its virtuous characteristics, but is as fast as optimized softmax and achieves on par or better performance in machine translation task.
BibTeX
@inproceedings{tezekbayev-etal-2022-speeding,
title = "Speeding Up Entmax",
author = "Tezekbayev, Maxat and
Nikoulina, Vassilina and
Gall{\'e}, Matthias and
Assylbekov, Zhenisbek",
editor = "Carpuat, Marine and
de Marneffe, Marie-Catherine and
Meza Ruiz, Ivan Vladimir",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2022",
month = jul,
year = "2022",
address = "Seattle, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.findings-naacl.86/",
doi = "10.18653/v1/2022.findings-naacl.86",
pages = "1142--1158"
}