Saved in:
Bibliographic Details
Main Authors: Anugraha, David, Winata, Genta Indra, Li, Chenyue, Irawan, Patrick Amadeus, Lee, En-Shiun Annie
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.09334
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916524257705984
author Anugraha, David
Winata, Genta Indra
Li, Chenyue
Irawan, Patrick Amadeus
Lee, En-Shiun Annie
author_facet Anugraha, David
Winata, Genta Indra
Li, Chenyue
Irawan, Patrick Amadeus
Lee, En-Shiun Annie
contents Performance prediction is a method to estimate the performance of Language Models (LMs) on various Natural Language Processing (NLP) tasks, mitigating computational costs associated with model capacity and data for fine-tuning. Our paper presents ProxyLM, a scalable task- and language-agnostic framework designed to predict the performance of LMs using proxy models. These proxy models act as surrogates, approximating the performance of the LM of interest. By leveraging these proxy models, ProxyLM significantly reduces computational overhead in task evaluations, achieving up to a 37.08x speedup over traditional methods, even with our smallest proxy models. Our results across multiple multilingual NLP tasks and various robustness tests demonstrate that ProxyLM not only adapts well to previously unseen languages in pre-trained LMs, but also generalizes effectively across different datasets, outperforming the state-of-the-art by at least 1.78x in terms of root-mean-square error (RMSE).
format Preprint
id arxiv_https___arxiv_org_abs_2406_09334
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle ProxyLM: Predicting Language Model Performance on Multilingual Tasks via Proxy Models
Anugraha, David
Winata, Genta Indra
Li, Chenyue
Irawan, Patrick Amadeus
Lee, En-Shiun Annie
Computation and Language
Performance prediction is a method to estimate the performance of Language Models (LMs) on various Natural Language Processing (NLP) tasks, mitigating computational costs associated with model capacity and data for fine-tuning. Our paper presents ProxyLM, a scalable task- and language-agnostic framework designed to predict the performance of LMs using proxy models. These proxy models act as surrogates, approximating the performance of the LM of interest. By leveraging these proxy models, ProxyLM significantly reduces computational overhead in task evaluations, achieving up to a 37.08x speedup over traditional methods, even with our smallest proxy models. Our results across multiple multilingual NLP tasks and various robustness tests demonstrate that ProxyLM not only adapts well to previously unseen languages in pre-trained LMs, but also generalizes effectively across different datasets, outperforming the state-of-the-art by at least 1.78x in terms of root-mean-square error (RMSE).
title ProxyLM: Predicting Language Model Performance on Multilingual Tasks via Proxy Models
topic Computation and Language
url https://arxiv.org/abs/2406.09334