Soft Alignment Objectives for Robust Adaptation of Language Generation

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14330%2F23%3A00130912" target="_blank" >RIV/00216224:14330/23:00130912 - isvavai.cz</a>

Výsledek na webu

<a href="https://aclanthology.org/2023.acl-long.492" target="_blank" >https://aclanthology.org/2023.acl-long.492</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/2023.acl-long.492" target="_blank" >10.18653/v1/2023.acl-long.492</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Soft Alignment Objectives for Robust Adaptation of Language Generation

Popis výsledku v původním jazyce

Domain adaptation allows generative language models to address specific flaws caused by the domain shift of their application. However, the traditional adaptation by further training on in-domain data rapidly weakens the model's ability to generalize to other domains, making the open-ended deployments of the adapted models prone to errors. This work introduces novel training objectives built upon a semantic similarity of the predicted tokens to the reference. Our results show that (1) avoiding the common assumption of a single correct prediction by constructing the training target from tokens' semantic similarity can largely mitigate catastrophic forgetting of adaptation, while (2) preserving the adaptation in-domain quality, (3) with negligible additions to compute costs. In the broader context, the objectives grounded in a continuous token similarity pioneer the exploration of the middle ground between the efficient but na"{i}ve exact-match token-level objectives and expressive but computationally- and resource-intensive sequential objectives.

Název v anglickém jazyce

Soft Alignment Objectives for Robust Adaptation of Language Generation

Popis výsledku anglicky

Domain adaptation allows generative language models to address specific flaws caused by the domain shift of their application. However, the traditional adaptation by further training on in-domain data rapidly weakens the model's ability to generalize to other domains, making the open-ended deployments of the adapted models prone to errors. This work introduces novel training objectives built upon a semantic similarity of the predicted tokens to the reference. Our results show that (1) avoiding the common assumption of a single correct prediction by constructing the training target from tokens' semantic similarity can largely mitigate catastrophic forgetting of adaptation, while (2) preserving the adaptation in-domain quality, (3) with negligible additions to compute costs. In the broader context, the objectives grounded in a continuous token similarity pioneer the exploration of the middle ground between the efficient but na"{i}ve exact-match token-level objectives and expressive but computationally- and resource-intensive sequential objectives.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

ISBN

9781959429722

ISSN

0736-587X

e-ISSN

—

Počet stran výsledku

17

Strana od-do

8837-8853

Název nakladatele

Association for Computational Linguistics

Místo vydání

Toronto, Canada

Místo konání akce

Toronto, Canada

Datum konání akce

9. 7. 2023

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—