@inproceedings{sun-etal-2020-distill,
title = "Distill and Replay for Continual Language Learning",
author = "Sun, Jingyuan and
Wang, Shaonan and
Zhang, Jiajun and
Zong, Chengqing",
editor = "Scott, Donia and
Bel, Nuria and
Zong, Chengqing",
booktitle = "Proceedings of the 28th International Conference on Computational Linguistics",
month = dec,
year = "2020",
address = "Barcelona, Spain (Online)",
publisher = "International Committee on Computational Linguistics",
url = "https://aclanthology.org/2020.coling-main.318/",
doi = "10.18653/v1/2020.coling-main.318",
pages = "3569--3579",
abstract = "Accumulating knowledge to tackle new tasks without necessarily forgetting the old ones is a hallmark of human-like intelligence. But the current dominant paradigm of machine learning is still to train a model that works well on static datasets. When learning tasks in a stream where data distribution may fluctuate, fitting on new tasks often leads to forgetting on the previous ones. We propose a simple yet effective framework that continually learns natural language understanding tasks with one model. Our framework distills knowledge and replays experience from previous tasks when fitting on a new task, thus named DnR (distill and replay). The framework is based on language models and can be smoothly built with different language model architectures. Experimental results demonstrate that DnR outperfoms previous state-of-the-art models in continually learning tasks of the same type but from different domains, as well as tasks of different types. With the distillation method, we further show that it`s possible for DnR to incrementally compress the model size while still outperforming most of the baselines. We hope that DnR could promote the empirical application of continual language learning, and contribute to building human-level language intelligence minimally bothered by catastrophic forgetting."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sun-etal-2020-distill">
<titleInfo>
<title>Distill and Replay for Continual Language Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jingyuan</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shaonan</namePart>
<namePart type="family">Wang</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>2020-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 28th International Conference on Computational Linguistics</title>
</titleInfo>
<name type="personal">
<namePart type="given">Donia</namePart>
<namePart type="family">Scott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nuria</namePart>
<namePart type="family">Bel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chengqing</namePart>
<namePart type="family">Zong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>International Committee on Computational Linguistics</publisher>
<place>
<placeTerm type="text">Barcelona, Spain (Online)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Accumulating knowledge to tackle new tasks without necessarily forgetting the old ones is a hallmark of human-like intelligence. But the current dominant paradigm of machine learning is still to train a model that works well on static datasets. When learning tasks in a stream where data distribution may fluctuate, fitting on new tasks often leads to forgetting on the previous ones. We propose a simple yet effective framework that continually learns natural language understanding tasks with one model. Our framework distills knowledge and replays experience from previous tasks when fitting on a new task, thus named DnR (distill and replay). The framework is based on language models and can be smoothly built with different language model architectures. Experimental results demonstrate that DnR outperfoms previous state-of-the-art models in continually learning tasks of the same type but from different domains, as well as tasks of different types. With the distillation method, we further show that it‘s possible for DnR to incrementally compress the model size while still outperforming most of the baselines. We hope that DnR could promote the empirical application of continual language learning, and contribute to building human-level language intelligence minimally bothered by catastrophic forgetting.</abstract>
<identifier type="citekey">sun-etal-2020-distill</identifier>
<identifier type="doi">10.18653/v1/2020.coling-main.318</identifier>
<location>
<url>https://aclanthology.org/2020.coling-main.318/</url>
</location>
<part>
<date>2020-12</date>
<extent unit="page">
<start>3569</start>
<end>3579</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Distill and Replay for Continual Language Learning
%A Sun, Jingyuan
%A Wang, Shaonan
%A Zhang, Jiajun
%A Zong, Chengqing
%Y Scott, Donia
%Y Bel, Nuria
%Y Zong, Chengqing
%S Proceedings of the 28th International Conference on Computational Linguistics
%D 2020
%8 December
%I International Committee on Computational Linguistics
%C Barcelona, Spain (Online)
%F sun-etal-2020-distill
%X Accumulating knowledge to tackle new tasks without necessarily forgetting the old ones is a hallmark of human-like intelligence. But the current dominant paradigm of machine learning is still to train a model that works well on static datasets. When learning tasks in a stream where data distribution may fluctuate, fitting on new tasks often leads to forgetting on the previous ones. We propose a simple yet effective framework that continually learns natural language understanding tasks with one model. Our framework distills knowledge and replays experience from previous tasks when fitting on a new task, thus named DnR (distill and replay). The framework is based on language models and can be smoothly built with different language model architectures. Experimental results demonstrate that DnR outperfoms previous state-of-the-art models in continually learning tasks of the same type but from different domains, as well as tasks of different types. With the distillation method, we further show that it‘s possible for DnR to incrementally compress the model size while still outperforming most of the baselines. We hope that DnR could promote the empirical application of continual language learning, and contribute to building human-level language intelligence minimally bothered by catastrophic forgetting.
%R 10.18653/v1/2020.coling-main.318
%U https://aclanthology.org/2020.coling-main.318/
%U https://doi.org/10.18653/v1/2020.coling-main.318
%P 3569-3579
Markdown (Informal)
[Distill and Replay for Continual Language Learning](https://aclanthology.org/2020.coling-main.318/) (Sun et al., COLING 2020)
ACL
- Jingyuan Sun, Shaonan Wang, Jiajun Zhang, and Chengqing Zong. 2020. Distill and Replay for Continual Language Learning. In Proceedings of the 28th International Conference on Computational Linguistics, pages 3569–3579, Barcelona, Spain (Online). International Committee on Computational Linguistics.