@inproceedings{nozaki-etal-2025-efficient,
title = "Efficient Vocabulary Reduction for Small Language Models",
author = "Nozaki, Yuta and
Nakashima, Dai and
Sato, Ryo and
Asaba, Naoki and
Kawamura, Shintaro",
editor = "Rambow, Owen and
Wanner, Leo and
Apidianaki, Marianna and
Al-Khalifa, Hend and
Eugenio, Barbara Di and
Schockaert, Steven and
Darwish, Kareem and
Agarwal, Apoorv",
booktitle = "Proceedings of the 31st International Conference on Computational Linguistics: Industry Track",
month = jan,
year = "2025",
address = "Abu Dhabi, UAE",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.coling-industry.64/",
pages = "771--783",
abstract = "The increasing size of large language models (LLMs) poses significant challenges due to their high computational costs and energy consumption, making their deployment in industrial settings difficult. Small language models (SLMs) have been introduced to mitigate these challenges by reducing model size while preserving performance. However, the embedding layer, which occupies a significant portion of the model, remains a bottleneck in model compression efforts. In this paper, we valdated vocabulary reduction as a solution to compress the embedding layer and reduce model size without significant loss of performance. We conduct a series of experiments to investigate how vocabulary reduction affects GPU memory footprint, inference speed, and task performance. Our results show that while performance generally declines with vocabulary reduction, fine-tuning can recover much of the lost performance. Moreover, in some tasks, such as truthfulness and summarization, the vocabulary-reduced models outperform the baseline. Finally, we demonstrate that vocabulary reduction can be effectively applied in domain adaptation, particularly in the medical domain, and in multilingual adaptation, improving task efficiency and cross-lingual robustness."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="nozaki-etal-2025-efficient">
<titleInfo>
<title>Efficient Vocabulary Reduction for Small Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yuta</namePart>
<namePart type="family">Nozaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dai</namePart>
<namePart type="family">Nakashima</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ryo</namePart>
<namePart type="family">Sato</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naoki</namePart>
<namePart type="family">Asaba</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shintaro</namePart>
<namePart type="family">Kawamura</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-01</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 31st International Conference on Computational Linguistics: Industry Track</title>
</titleInfo>
<name type="personal">
<namePart type="given">Owen</namePart>
<namePart type="family">Rambow</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Leo</namePart>
<namePart type="family">Wanner</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marianna</namePart>
<namePart type="family">Apidianaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hend</namePart>
<namePart type="family">Al-Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Barbara</namePart>
<namePart type="given">Di</namePart>
<namePart type="family">Eugenio</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steven</namePart>
<namePart type="family">Schockaert</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kareem</namePart>
<namePart type="family">Darwish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Apoorv</namePart>
<namePart type="family">Agarwal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, UAE</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The increasing size of large language models (LLMs) poses significant challenges due to their high computational costs and energy consumption, making their deployment in industrial settings difficult. Small language models (SLMs) have been introduced to mitigate these challenges by reducing model size while preserving performance. However, the embedding layer, which occupies a significant portion of the model, remains a bottleneck in model compression efforts. In this paper, we valdated vocabulary reduction as a solution to compress the embedding layer and reduce model size without significant loss of performance. We conduct a series of experiments to investigate how vocabulary reduction affects GPU memory footprint, inference speed, and task performance. Our results show that while performance generally declines with vocabulary reduction, fine-tuning can recover much of the lost performance. Moreover, in some tasks, such as truthfulness and summarization, the vocabulary-reduced models outperform the baseline. Finally, we demonstrate that vocabulary reduction can be effectively applied in domain adaptation, particularly in the medical domain, and in multilingual adaptation, improving task efficiency and cross-lingual robustness.</abstract>
<identifier type="citekey">nozaki-etal-2025-efficient</identifier>
<location>
<url>https://aclanthology.org/2025.coling-industry.64/</url>
</location>
<part>
<date>2025-01</date>
<extent unit="page">
<start>771</start>
<end>783</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Efficient Vocabulary Reduction for Small Language Models
%A Nozaki, Yuta
%A Nakashima, Dai
%A Sato, Ryo
%A Asaba, Naoki
%A Kawamura, Shintaro
%Y Rambow, Owen
%Y Wanner, Leo
%Y Apidianaki, Marianna
%Y Al-Khalifa, Hend
%Y Eugenio, Barbara Di
%Y Schockaert, Steven
%Y Darwish, Kareem
%Y Agarwal, Apoorv
%S Proceedings of the 31st International Conference on Computational Linguistics: Industry Track
%D 2025
%8 January
%I Association for Computational Linguistics
%C Abu Dhabi, UAE
%F nozaki-etal-2025-efficient
%X The increasing size of large language models (LLMs) poses significant challenges due to their high computational costs and energy consumption, making their deployment in industrial settings difficult. Small language models (SLMs) have been introduced to mitigate these challenges by reducing model size while preserving performance. However, the embedding layer, which occupies a significant portion of the model, remains a bottleneck in model compression efforts. In this paper, we valdated vocabulary reduction as a solution to compress the embedding layer and reduce model size without significant loss of performance. We conduct a series of experiments to investigate how vocabulary reduction affects GPU memory footprint, inference speed, and task performance. Our results show that while performance generally declines with vocabulary reduction, fine-tuning can recover much of the lost performance. Moreover, in some tasks, such as truthfulness and summarization, the vocabulary-reduced models outperform the baseline. Finally, we demonstrate that vocabulary reduction can be effectively applied in domain adaptation, particularly in the medical domain, and in multilingual adaptation, improving task efficiency and cross-lingual robustness.
%U https://aclanthology.org/2025.coling-industry.64/
%P 771-783
Markdown (Informal)
[Efficient Vocabulary Reduction for Small Language Models](https://aclanthology.org/2025.coling-industry.64/) (Nozaki et al., COLING 2025)
ACL
- Yuta Nozaki, Dai Nakashima, Ryo Sato, Naoki Asaba, and Shintaro Kawamura. 2025. Efficient Vocabulary Reduction for Small Language Models. In Proceedings of the 31st International Conference on Computational Linguistics: Industry Track, pages 771–783, Abu Dhabi, UAE. Association for Computational Linguistics.