Dynamic ASR Pathways: An Adaptive Masking Approach Towards Efficient Pruning of A Multilingual ASR Model

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xie, Jiamin, Li, Ke, Guo, Jinxi, Tjandra, Andros, Shangguan, Yuan, Sari, Leda, Wu, Chunyang, Jia, Junteng, Mahadeokar, Jay, Kalinli, Ozlem
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909651638943744
author Xie, Jiamin
Li, Ke
Guo, Jinxi
Tjandra, Andros
Shangguan, Yuan
Sari, Leda
Wu, Chunyang
Jia, Junteng
Mahadeokar, Jay
Kalinli, Ozlem
author_facet Xie, Jiamin
Li, Ke
Guo, Jinxi
Tjandra, Andros
Shangguan, Yuan
Sari, Leda
Wu, Chunyang
Jia, Junteng
Mahadeokar, Jay
Kalinli, Ozlem
contents Neural network pruning offers an effective method for compressing a multilingual automatic speech recognition (ASR) model with minimal performance loss. However, it entails several rounds of pruning and re-training needed to be run for each language. In this work, we propose the use of an adaptive masking approach in two scenarios for pruning a multilingual ASR model efficiently, each resulting in sparse monolingual models or a sparse multilingual model (named as Dynamic ASR Pathways). Our approach dynamically adapts the sub-network, avoiding premature decisions about a fixed sub-network structure. We show that our approach outperforms existing pruning methods when targeting sparse monolingual models. Further, we illustrate that Dynamic ASR Pathways jointly discovers and trains better sub-networks (pathways) of a single multilingual model by adapting from different sub-network initializations, thereby reducing the need for language-specific pruning.
format Preprint
id arxiv_https___arxiv_org_abs_2309_13018
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Dynamic ASR Pathways: An Adaptive Masking Approach Towards Efficient Pruning of A Multilingual ASR Model
Xie, Jiamin
Li, Ke
Guo, Jinxi
Tjandra, Andros
Shangguan, Yuan
Sari, Leda
Wu, Chunyang
Jia, Junteng
Mahadeokar, Jay
Kalinli, Ozlem
Audio and Speech Processing
Artificial Intelligence
Computation and Language
Machine Learning
Sound
Neural network pruning offers an effective method for compressing a multilingual automatic speech recognition (ASR) model with minimal performance loss. However, it entails several rounds of pruning and re-training needed to be run for each language. In this work, we propose the use of an adaptive masking approach in two scenarios for pruning a multilingual ASR model efficiently, each resulting in sparse monolingual models or a sparse multilingual model (named as Dynamic ASR Pathways). Our approach dynamically adapts the sub-network, avoiding premature decisions about a fixed sub-network structure. We show that our approach outperforms existing pruning methods when targeting sparse monolingual models. Further, we illustrate that Dynamic ASR Pathways jointly discovers and trains better sub-networks (pathways) of a single multilingual model by adapting from different sub-network initializations, thereby reducing the need for language-specific pruning.
title Dynamic ASR Pathways: An Adaptive Masking Approach Towards Efficient Pruning of A Multilingual ASR Model
topic Audio and Speech Processing
Artificial Intelligence
Computation and Language
Machine Learning
Sound
url https://arxiv.org/abs/2309.13018