@inproceedings{bunzeck-etal-2024-graphemes,
title = "Graphemes vs. phonemes: battling it out in character-based language models",
author = "Bunzeck, Bastian and
Duran, Daniel and
Schade, Leonie and
Zarrie{\ss}, Sina",
editor = "Hu, Michael Y. and
Mueller, Aaron and
Ross, Candace and
Williams, Adina and
Linzen, Tal and
Zhuang, Chengxu and
Choshen, Leshem and
Cotterell, Ryan and
Warstadt, Alex and
Wilcox, Ethan Gotlieb",
booktitle = "The 2nd BabyLM Challenge at the 28th Conference on Computational Natural Language Learning",
month = nov,
year = "2024",
address = "Miami, FL, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.conll-babylm.5/",
pages = "54--64",
abstract = "We present grapheme-llama and phoneme-llama, character-based language models trained for the 2024 BabyLM challenge. Through these models, we explore an under-researched approach to downsizing: replacing subword-based tokenization with character-level tokenization, drastically reducing the vocabulary size. The grapheme model is trained on a standard BabyLM dataset, while the phoneme model uses a phoneme-converted version of this dataset. Results show that grapheme-based models perform better overall, achieving scores comparable to subword-based models on grammatical benchmarks. Despite lower performance, phoneme models also demonstrate promising grammatical learning. We argue that our results challenge conventional wisdom on language modeling techniques and open up novel research questions with character- and phoneme-based models as objects of inquiry."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bunzeck-etal-2024-graphemes">
<titleInfo>
<title>Graphemes vs. phonemes: battling it out in character-based language models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Bastian</namePart>
<namePart type="family">Bunzeck</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daniel</namePart>
<namePart type="family">Duran</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Leonie</namePart>
<namePart type="family">Schade</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sina</namePart>
<namePart type="family">Zarrieß</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>The 2nd BabyLM Challenge at the 28th Conference on Computational Natural Language Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Michael</namePart>
<namePart type="given">Y</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aaron</namePart>
<namePart type="family">Mueller</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Candace</namePart>
<namePart type="family">Ross</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Adina</namePart>
<namePart type="family">Williams</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tal</namePart>
<namePart type="family">Linzen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chengxu</namePart>
<namePart type="family">Zhuang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Leshem</namePart>
<namePart type="family">Choshen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ryan</namePart>
<namePart type="family">Cotterell</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alex</namePart>
<namePart type="family">Warstadt</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ethan</namePart>
<namePart type="given">Gotlieb</namePart>
<namePart type="family">Wilcox</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, FL, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We present grapheme-llama and phoneme-llama, character-based language models trained for the 2024 BabyLM challenge. Through these models, we explore an under-researched approach to downsizing: replacing subword-based tokenization with character-level tokenization, drastically reducing the vocabulary size. The grapheme model is trained on a standard BabyLM dataset, while the phoneme model uses a phoneme-converted version of this dataset. Results show that grapheme-based models perform better overall, achieving scores comparable to subword-based models on grammatical benchmarks. Despite lower performance, phoneme models also demonstrate promising grammatical learning. We argue that our results challenge conventional wisdom on language modeling techniques and open up novel research questions with character- and phoneme-based models as objects of inquiry.</abstract>
<identifier type="citekey">bunzeck-etal-2024-graphemes</identifier>
<location>
<url>https://aclanthology.org/2024.conll-babylm.5/</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>54</start>
<end>64</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Graphemes vs. phonemes: battling it out in character-based language models
%A Bunzeck, Bastian
%A Duran, Daniel
%A Schade, Leonie
%A Zarrieß, Sina
%Y Hu, Michael Y.
%Y Mueller, Aaron
%Y Ross, Candace
%Y Williams, Adina
%Y Linzen, Tal
%Y Zhuang, Chengxu
%Y Choshen, Leshem
%Y Cotterell, Ryan
%Y Warstadt, Alex
%Y Wilcox, Ethan Gotlieb
%S The 2nd BabyLM Challenge at the 28th Conference on Computational Natural Language Learning
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, FL, USA
%F bunzeck-etal-2024-graphemes
%X We present grapheme-llama and phoneme-llama, character-based language models trained for the 2024 BabyLM challenge. Through these models, we explore an under-researched approach to downsizing: replacing subword-based tokenization with character-level tokenization, drastically reducing the vocabulary size. The grapheme model is trained on a standard BabyLM dataset, while the phoneme model uses a phoneme-converted version of this dataset. Results show that grapheme-based models perform better overall, achieving scores comparable to subword-based models on grammatical benchmarks. Despite lower performance, phoneme models also demonstrate promising grammatical learning. We argue that our results challenge conventional wisdom on language modeling techniques and open up novel research questions with character- and phoneme-based models as objects of inquiry.
%U https://aclanthology.org/2024.conll-babylm.5/
%P 54-64
Markdown (Informal)
[Graphemes vs. phonemes: battling it out in character-based language models](https://aclanthology.org/2024.conll-babylm.5/) (Bunzeck et al., CoNLL-BabyLM 2024)
ACL