@inproceedings{jin-etal-2023-parameter,
title = "Parameter-efficient Tuning for Large Language Model without Calculating Its Gradients",
author = "Jin, Feihu and
Zhang, Jiajun and
Zong, Chengqing",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.22",
doi = "10.18653/v1/2023.emnlp-main.22",
pages = "321--330",
abstract = "Fine-tuning all parameters of large language models (LLMs) requires significant computational resources and is time-consuming. Recent parameter-efficient tuning methods such as Adapter tuning, Prefix tuning, and LoRA allow for updating a small subset of parameters in large language models. However, they can only save approximately 30{\%} of the training memory requirements, due to the problem that gradient computation and backpropagation are still necessary for these methods. This paper proposes a novel parameter-efficient tuning method for LLMs without calculating their gradients. Leveraging the discernible similarities between the parameter-efficient modules of the same task learned by both large and small language models, we put forward a strategy for transferring the parameter-efficient modules, originally derived from small language models to much larger ones. To ensure a smooth and effective adaptation process, we further introduce a Bridge model to guarantee dimensional consistency while also stimulating a dynamic interaction between the models. We demonstrate the effectiveness of our method using the T5 and GPT-2 series of language models on the SuperGLUE benchmark. Our method achieves comparable performance to both fine-tuning and parameter-efficient tuning on large language models without needing gradient-based optimization. Additionally, our method achieves up to 5.7x memory reduction compared to parameter-efficient tuning.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="jin-etal-2023-parameter">
<titleInfo>
<title>Parameter-efficient Tuning for Large Language Model without Calculating Its Gradients</title>
</titleInfo>
<name type="personal">
<namePart type="given">Feihu</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chengqing</namePart>
<namePart type="family">Zong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Juan</namePart>
<namePart type="family">Pino</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kalika</namePart>
<namePart type="family">Bali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Singapore</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Fine-tuning all parameters of large language models (LLMs) requires significant computational resources and is time-consuming. Recent parameter-efficient tuning methods such as Adapter tuning, Prefix tuning, and LoRA allow for updating a small subset of parameters in large language models. However, they can only save approximately 30% of the training memory requirements, due to the problem that gradient computation and backpropagation are still necessary for these methods. This paper proposes a novel parameter-efficient tuning method for LLMs without calculating their gradients. Leveraging the discernible similarities between the parameter-efficient modules of the same task learned by both large and small language models, we put forward a strategy for transferring the parameter-efficient modules, originally derived from small language models to much larger ones. To ensure a smooth and effective adaptation process, we further introduce a Bridge model to guarantee dimensional consistency while also stimulating a dynamic interaction between the models. We demonstrate the effectiveness of our method using the T5 and GPT-2 series of language models on the SuperGLUE benchmark. Our method achieves comparable performance to both fine-tuning and parameter-efficient tuning on large language models without needing gradient-based optimization. Additionally, our method achieves up to 5.7x memory reduction compared to parameter-efficient tuning.</abstract>
<identifier type="citekey">jin-etal-2023-parameter</identifier>
<identifier type="doi">10.18653/v1/2023.emnlp-main.22</identifier>
<location>
<url>https://aclanthology.org/2023.emnlp-main.22</url>
</location>
<part>
<date>2023-12</date>
<extent unit="page">
<start>321</start>
<end>330</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Parameter-efficient Tuning for Large Language Model without Calculating Its Gradients
%A Jin, Feihu
%A Zhang, Jiajun
%A Zong, Chengqing
%Y Bouamor, Houda
%Y Pino, Juan
%Y Bali, Kalika
%S Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
%D 2023
%8 December
%I Association for Computational Linguistics
%C Singapore
%F jin-etal-2023-parameter
%X Fine-tuning all parameters of large language models (LLMs) requires significant computational resources and is time-consuming. Recent parameter-efficient tuning methods such as Adapter tuning, Prefix tuning, and LoRA allow for updating a small subset of parameters in large language models. However, they can only save approximately 30% of the training memory requirements, due to the problem that gradient computation and backpropagation are still necessary for these methods. This paper proposes a novel parameter-efficient tuning method for LLMs without calculating their gradients. Leveraging the discernible similarities between the parameter-efficient modules of the same task learned by both large and small language models, we put forward a strategy for transferring the parameter-efficient modules, originally derived from small language models to much larger ones. To ensure a smooth and effective adaptation process, we further introduce a Bridge model to guarantee dimensional consistency while also stimulating a dynamic interaction between the models. We demonstrate the effectiveness of our method using the T5 and GPT-2 series of language models on the SuperGLUE benchmark. Our method achieves comparable performance to both fine-tuning and parameter-efficient tuning on large language models without needing gradient-based optimization. Additionally, our method achieves up to 5.7x memory reduction compared to parameter-efficient tuning.
%R 10.18653/v1/2023.emnlp-main.22
%U https://aclanthology.org/2023.emnlp-main.22
%U https://doi.org/10.18653/v1/2023.emnlp-main.22
%P 321-330
Markdown (Informal)
[Parameter-efficient Tuning for Large Language Model without Calculating Its Gradients](https://aclanthology.org/2023.emnlp-main.22) (Jin et al., EMNLP 2023)
ACL