English
Related papers

Related papers: Token Dropping for Efficient BERT Pretraining

200 papers

Subword tokenization is an essential part of modern large language models (LLMs), yet its specific contributions to training efficiency and model performance remain poorly understood. In this work, we decouple the effects of subword…

Computation and Language · Computer Science 2026-05-15 Théo Gigant , Bowen Peng , Jeffrey Quesnelle

Applying the Transformer architecture on the character level usually requires very deep architectures that are difficult and slow to train. These problems can be partially overcome by incorporating a segmentation into tokens in the model.…

Computation and Language · Computer Science 2020-09-30 Jindřich Libovický , Alexander Fraser

Transformers have become a predominant machine learning workload, they are not only the de-facto standard for natural language processing tasks, but they are also being deployed in other domains such as vision and speech recognition. Many…

Machine Learning · Computer Science 2022-06-23 Ibrahim Ahmed , Sahil Parmar , Matthew Boyd , Michael Beidler , Kris Kang , Bill Liu , Kyle Roach , John Kim , Dennis Abts

Many applications of large language models (LLMs) require long-context understanding, but models continue to struggle with such tasks. We hypothesize that conventional next-token prediction training could contribute to this, because each…

Computation and Language · Computer Science 2025-03-13 Falko Helm , Nico Daheim , Iryna Gurevych

In the low-data regime, it is difficult to train good supervised models from scratch. Instead practitioners turn to pre-trained models, leveraging transfer learning. Ensembling is an empirically and theoretically appealing way to construct…

Machine Learning · Computer Science 2020-10-20 Basil Mustafa , Carlos Riquelme , Joan Puigcerver , André Susano Pinto , Daniel Keysers , Neil Houlsby

We enhance auto-regressive language models by conditioning on document chunks retrieved from a large corpus, based on local similarity with preceding tokens. With a $2$ trillion token database, our Retrieval-Enhanced Transformer (RETRO)…

Language model pre-training, such as BERT, has significantly improved the performances of many natural language processing tasks. However, pre-trained language models are usually computationally expensive, so it is difficult to efficiently…

Computation and Language · Computer Science 2020-10-19 Xiaoqi Jiao , Yichun Yin , Lifeng Shang , Xin Jiang , Xiao Chen , Linlin Li , Fang Wang , Qun Liu

Long-context inference enhances the reasoning capability of Large Language Models (LLMs), but incurs significant computational overhead. Token-oriented methods, such as pruning and skipping, have shown great promise in reducing inference…

Computation and Language · Computer Science 2026-02-03 Zimeng Wu , Donghao Wang , Chaozhe Jin , Jiaxin Chen , Yunhong Wang

Masked language modeling (MLM) is one of the key sub-tasks in vision-language pretraining. In the cross-modal setting, tokens in the sentence are masked at random, and the model predicts the masked tokens given the image and the text. In…

Computation and Language · Computer Science 2021-09-07 Yonatan Bitton , Gabriel Stanovsky , Michael Elhadad , Roy Schwartz

With the success of large-scale pre-training in language tasks, there is an increasing trend of applying it to the domain of life sciences. In particular, pre-training methods based on DNA sequences have received increasing attention…

Artificial Intelligence · Computer Science 2024-09-10 Chaoqi Liang , Lifeng Qiao , Peng Ye , Nanqing Dong , Jianle Sun , Weiqiang Bai , Yuchen Ren , Xinzhu Ma , Hongliang Yan , Chunfeng Song , Wanli Ouyang , Wangmeng Zuo

Few-shot algorithms aim at learning new tasks provided only a handful of training examples. In this work we investigate few-shot learning in the setting where the data points are sequences of tokens and propose an efficient learning…

Machine Learning · Computer Science 2020-12-18 Lajanugen Logeswaran , Ann Lee , Myle Ott , Honglak Lee , Marc'Aurelio Ranzato , Arthur Szlam

Transferring a deep neural network trained on one problem to another requires only a small amount of data and little additional computation time. The same behaviour holds for ensembles of deep learning models typically superior to a single…

Machine Learning · Computer Science 2022-06-28 Ilya Shashkov , Nikita Balabin , Evgeny Burnaev , Alexey Zaytsev

Transfer learning techniques are particularly useful in NLP tasks where a sizable amount of high-quality annotated data is difficult to obtain. Current approaches directly adapt a pre-trained language model (LM) on in-domain text before…

The pre-training of large language models usually requires massive amounts of resources, both in terms of computation and data. Frequently used web sources such as Common Crawl might contain enough noise to make this pre-training…

Computation and Language · Computer Science 2022-07-15 Javier de la Rosa , Eduardo G. Ponferrada , Paulo Villegas , Pablo Gonzalez de Prado Salas , Manu Romero , Marıa Grandury

Pretraining NLP models with variants of Masked Language Model (MLM) objectives has recently led to a significant improvements on many tasks. This paper examines the benefits of pretrained models as a function of the number of training…

Computation and Language · Computer Science 2020-06-17 Sinong Wang , Madian Khabsa , Hao Ma

Large-scale pre-trained language models such as BERT have contributed significantly to the development of NLP. However, those models require large computational resources, making it difficult to be applied to mobile devices where computing…

Computation and Language · Computer Science 2023-08-02 Weixin Wu , Hankz Hankui Zhuo

Language model pre-training has proven to be useful in learning universal language representations. As a state-of-the-art language model pre-training model, BERT (Bidirectional Encoder Representations from Transformers) has achieved amazing…

Computation and Language · Computer Science 2020-02-06 Chi Sun , Xipeng Qiu , Yige Xu , Xuanjing Huang

Parameter-efficient tuning aims at updating only a small subset of parameters when adapting a pretrained model to downstream tasks. In this work, we introduce PASTA, in which we only modify the special token representations (e.g., [SEP] and…

Computation and Language · Computer Science 2023-02-15 Xiaocong Yang , James Y. Huang , Wenxuan Zhou , Muhao Chen

While behavior learning has made impressive progress in recent times, it lags behind computer vision and natural language processing due to its inability to leverage large, human-generated datasets. Human behaviors have wide variance,…

Machine Learning · Computer Science 2022-10-13 Nur Muhammad Mahi Shafiullah , Zichen Jeff Cui , Ariuntuya Altanzaya , Lerrel Pinto

Transformer models have achieved promising results on natural language processing (NLP) tasks including extractive question answering (QA). Common Transformer encoders used in NLP tasks process the hidden states of all input tokens in the…

Computation and Language · Computer Science 2022-05-17 Yue Guan , Zhengyi Li , Jingwen Leng , Zhouhan Lin , Minyi Guo , Yuhao Zhu