GhostRNN: Reducing State Redundancy in RNN with Cheap Operations

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhou, Hang, Zheng, Xiaoxu, Wang, Yunhe, Mi, Michael Bi, Xiong, Deyi, Han, Kai
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916491940593664
author Zhou, Hang
Zheng, Xiaoxu
Wang, Yunhe
Mi, Michael Bi
Xiong, Deyi
Han, Kai
author_facet Zhou, Hang
Zheng, Xiaoxu
Wang, Yunhe
Mi, Michael Bi
Xiong, Deyi
Han, Kai
contents Recurrent neural network (RNNs) that are capable of modeling long-distance dependencies are widely used in various speech tasks, eg., keyword spotting (KWS) and speech enhancement (SE). Due to the limitation of power and memory in low-resource devices, efficient RNN models are urgently required for real-world applications. In this paper, we propose an efficient RNN architecture, GhostRNN, which reduces hidden state redundancy with cheap operations. In particular, we observe that partial dimensions of hidden states are similar to the others in trained RNN models, suggesting that redundancy exists in specific RNNs. To reduce the redundancy and hence computational cost, we propose to first generate a few intrinsic states, and then apply cheap operations to produce ghost states based on the intrinsic states. Experiments on KWS and SE tasks demonstrate that the proposed GhostRNN significantly reduces the memory usage (~40%) and computation cost while keeping performance similar.
format Preprint
id arxiv_https___arxiv_org_abs_2411_14489
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle GhostRNN: Reducing State Redundancy in RNN with Cheap Operations
Zhou, Hang
Zheng, Xiaoxu
Wang, Yunhe
Mi, Michael Bi
Xiong, Deyi
Han, Kai
Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
Recurrent neural network (RNNs) that are capable of modeling long-distance dependencies are widely used in various speech tasks, eg., keyword spotting (KWS) and speech enhancement (SE). Due to the limitation of power and memory in low-resource devices, efficient RNN models are urgently required for real-world applications. In this paper, we propose an efficient RNN architecture, GhostRNN, which reduces hidden state redundancy with cheap operations. In particular, we observe that partial dimensions of hidden states are similar to the others in trained RNN models, suggesting that redundancy exists in specific RNNs. To reduce the redundancy and hence computational cost, we propose to first generate a few intrinsic states, and then apply cheap operations to produce ghost states based on the intrinsic states. Experiments on KWS and SE tasks demonstrate that the proposed GhostRNN significantly reduces the memory usage (~40%) and computation cost while keeping performance similar.
title GhostRNN: Reducing State Redundancy in RNN with Cheap Operations
topic Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2411.14489