Universal Language Model Fine-tuning for Text Classification
J. Howard, and S. Ruder. (2018)cite arxiv:1801.06146Comment: ACL 2018, fixed denominator in Equation 3, line 3.
Abstract
Inductive transfer learning has greatly impacted computer vision, but
existing approaches in NLP still require task-specific modifications and
training from scratch. We propose Universal Language Model Fine-tuning
(ULMFiT), an effective transfer learning method that can be applied to any task
in NLP, and introduce techniques that are key for fine-tuning a language model.
Our method significantly outperforms the state-of-the-art on six text
classification tasks, reducing the error by 18-24% on the majority of datasets.
Furthermore, with only 100 labeled examples, it matches the performance of
training from scratch on 100x more data. We open-source our pretrained models
and code.
Description
Universal Language Model Fine-tuning for Text Classification
%0 Generic
%1 howard2018universal
%A Howard, Jeremy
%A Ruder, Sebastian
%D 2018
%K dfg-antrag-steckbriefe nlp transfer transferlearning ulmfit
%T Universal Language Model Fine-tuning for Text Classification
%U http://arxiv.org/abs/1801.06146
%X Inductive transfer learning has greatly impacted computer vision, but
existing approaches in NLP still require task-specific modifications and
training from scratch. We propose Universal Language Model Fine-tuning
(ULMFiT), an effective transfer learning method that can be applied to any task
in NLP, and introduce techniques that are key for fine-tuning a language model.
Our method significantly outperforms the state-of-the-art on six text
classification tasks, reducing the error by 18-24% on the majority of datasets.
Furthermore, with only 100 labeled examples, it matches the performance of
training from scratch on 100x more data. We open-source our pretrained models
and code.
@misc{howard2018universal,
abstract = {Inductive transfer learning has greatly impacted computer vision, but
existing approaches in NLP still require task-specific modifications and
training from scratch. We propose Universal Language Model Fine-tuning
(ULMFiT), an effective transfer learning method that can be applied to any task
in NLP, and introduce techniques that are key for fine-tuning a language model.
Our method significantly outperforms the state-of-the-art on six text
classification tasks, reducing the error by 18-24% on the majority of datasets.
Furthermore, with only 100 labeled examples, it matches the performance of
training from scratch on 100x more data. We open-source our pretrained models
and code.},
added-at = {2019-09-10T23:02:14.000+0200},
author = {Howard, Jeremy and Ruder, Sebastian},
biburl = {https://www.bibsonomy.org/bibtex/2d7ba313d42b8da06500bef2945b7b968/albinzehe},
description = {Universal Language Model Fine-tuning for Text Classification},
interhash = {4a4b12e7b6b8f26c4fc6cdcbecfbbdc3},
intrahash = {d7ba313d42b8da06500bef2945b7b968},
keywords = {dfg-antrag-steckbriefe nlp transfer transferlearning ulmfit},
note = {cite arxiv:1801.06146Comment: ACL 2018, fixed denominator in Equation 3, line 3},
timestamp = {2021-01-12T14:02:20.000+0100},
title = {Universal Language Model Fine-tuning for Text Classification},
url = {http://arxiv.org/abs/1801.06146},
year = 2018
}