<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="maimon-etal-2025-slamming">
    <titleInfo>
        <title>Slamming: Training a Speech Language Model on One GPU in a Day</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Gallil</namePart>
        <namePart type="family">Maimon</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Avishai</namePart>
        <namePart type="family">Elmakies</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yossi</namePart>
        <namePart type="family">Adi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: ACL 2025</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Wanxiang</namePart>
            <namePart type="family">Che</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Joyce</namePart>
            <namePart type="family">Nabende</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ekaterina</namePart>
            <namePart type="family">Shutova</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Mohammad</namePart>
            <namePart type="given">Taher</namePart>
            <namePart type="family">Pilehvar</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-256-5</identifier>
    </relatedItem>
    <abstract>We introduce *Slam*, a recipe for training high-quality Speech Language Models (SLMs) on a single academic GPU in 24 hours. We do so through empirical analysis of model initialisation and architecture, synthetic training data, preference optimisation with synthetic data and tweaking all other components. We empirically demonstrate that this training recipe also scales well with more compute getting results on par with leading SLMs in a fraction of the compute cost. We hope these insights will make SLM training and research more accessible. In the context of SLM scaling laws, our results far outperform predicted compute optimal performance, giving an optimistic view to SLM feasibility. See code, data, models, samples - https://pages.cs.huji.ac.il/adiyoss-lab/slamming .</abstract>
    <identifier type="citekey">maimon-etal-2025-slamming</identifier>
    <identifier type="doi">10.18653/v1/2025.findings-acl.631</identifier>
    <location>
        <url>https://aclanthology.org/2025.findings-acl.631/</url>
    </location>
    <part>
        <date>2025-07</date>
        <extent unit="page">
            <start>12201</start>
            <end>12216</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Slamming: Training a Speech Language Model on One GPU in a Day
%A Maimon, Gallil
%A Elmakies, Avishai
%A Adi, Yossi
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F maimon-etal-2025-slamming
%X We introduce *Slam*, a recipe for training high-quality Speech Language Models (SLMs) on a single academic GPU in 24 hours. We do so through empirical analysis of model initialisation and architecture, synthetic training data, preference optimisation with synthetic data and tweaking all other components. We empirically demonstrate that this training recipe also scales well with more compute getting results on par with leading SLMs in a fraction of the compute cost. We hope these insights will make SLM training and research more accessible. In the context of SLM scaling laws, our results far outperform predicted compute optimal performance, giving an optimistic view to SLM feasibility. See code, data, models, samples - https://pages.cs.huji.ac.il/adiyoss-lab/slamming .
%R 10.18653/v1/2025.findings-acl.631
%U https://aclanthology.org/2025.findings-acl.631/
%U https://doi.org/10.18653/v1/2025.findings-acl.631
%P 12201-12216

Markdown (Informal)

[Slamming: Training a Speech Language Model on One GPU in a Day](https://aclanthology.org/2025.findings-acl.631/) (Maimon et al., Findings 2025)

• Slamming: Training a Speech Language Model on One GPU in a Day (Maimon et al., Findings 2025)

ACL

• Gallil Maimon, Avishai Elmakies, and Yossi Adi. 2025. Slamming: Training a Speech Language Model on One GPU in a Day. In Findings of the Association for Computational Linguistics: ACL 2025, pages 12201–12216, Vienna, Austria. Association for Computational Linguistics.