Salvato in:
Dettagli Bibliografici
Autori principali: Lu, Taixi, Wang, Haoyu, Shao, Huajie, Gao, Jing, Yao, Huaxiu
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2402.15991
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911783882588160
author Lu, Taixi
Wang, Haoyu
Shao, Huajie
Gao, Jing
Yao, Huaxiu
author_facet Lu, Taixi
Wang, Haoyu
Shao, Huajie
Gao, Jing
Yao, Huaxiu
contents Cross-lingual natural language understanding (NLU) is a critical task in natural language processing (NLP). Recent advancements have seen multilingual pre-trained language models (mPLMs) significantly enhance the performance of these tasks. However, mPLMs necessitate substantial resources and incur high computational costs during inference, posing challenges for deployment in real-world and real-time systems. Existing model cascade methods seek to enhance inference efficiency by greedily selecting the lightest model capable of processing the current input from a variety of models, based on model confidence scores. Nonetheless, deep models tend to exhibit overconfidence, and confidence distributions vary across languages. This leads to the emission of confident but incorrect predictions by smaller models, hindering their ability to generalize effectively across test languages. In this study, we introduce a confidence calibration model cascade ($C^3$) method. This approach, simple yet effective, involves calibration prior to cascade inference, thereby enhancing cascade accuracy through more reliable predictions. Extensive experiments conducted on three cross-lingual benchmarks demonstrate that $C^3$ significantly outperforms all state-of-the-art baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2402_15991
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle $C^3$: Confidence Calibration Model Cascade for Inference-Efficient Cross-Lingual Natural Language Understanding
Lu, Taixi
Wang, Haoyu
Shao, Huajie
Gao, Jing
Yao, Huaxiu
Computation and Language
Cross-lingual natural language understanding (NLU) is a critical task in natural language processing (NLP). Recent advancements have seen multilingual pre-trained language models (mPLMs) significantly enhance the performance of these tasks. However, mPLMs necessitate substantial resources and incur high computational costs during inference, posing challenges for deployment in real-world and real-time systems. Existing model cascade methods seek to enhance inference efficiency by greedily selecting the lightest model capable of processing the current input from a variety of models, based on model confidence scores. Nonetheless, deep models tend to exhibit overconfidence, and confidence distributions vary across languages. This leads to the emission of confident but incorrect predictions by smaller models, hindering their ability to generalize effectively across test languages. In this study, we introduce a confidence calibration model cascade ($C^3$) method. This approach, simple yet effective, involves calibration prior to cascade inference, thereby enhancing cascade accuracy through more reliable predictions. Extensive experiments conducted on three cross-lingual benchmarks demonstrate that $C^3$ significantly outperforms all state-of-the-art baselines.
title $C^3$: Confidence Calibration Model Cascade for Inference-Efficient Cross-Lingual Natural Language Understanding
topic Computation and Language
url https://arxiv.org/abs/2402.15991