Must-read papers on improving efficiency for pre-trained language models.

Overview

Awesome Efficient PLM Papers

Must-read papers on improving efficiency for pre-trained language models.

The paper list is mainly mantained by Lei Li and Shuhuai Ren.

Knowledge Distillation

  1. DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter NeurIPS workshop

    Victor Sanh, Lysandre Debut, Julien Chaumond, Thomas Wolf [pdf] [project]

  2. Patient Knowledge Distillation for BERT Model Compression EMNLP 2019

    Siqi Sun, Yu Cheng, Zhe Gan, Jingjing Liu [pdf] [project]

  3. Well-Read Students Learn Better: On the Importance of Pre-training Compact Models Preprint

    Iulia Turc, Ming-Wei Chang, Kenton Lee, Kristina Toutanova [pdf] [project]

  4. TinyBERT: Distilling BERT for Natural Language Understanding Findings of EMNLP 2020

    Xiaoqi Jiao, Yichun Yin, Lifeng Shang, Xin Jiang, Xiao Chen, Linlin Li, Fang Wang, Qun Liu [pdf] [project]

  5. BERT-of-Theseus: Compressing BERT by Progressive Module Replacing EMNLP 2020

    Canwen Xu, Wangchunshu Zhou, Tao Ge, Furu Wei, Ming Zhou [pdf] [project]

  6. MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers NeurIPS 2020

    Wenhui Wang, Furu Wei, Li Dong, Hangbo Bao, Nan Yang, Ming Zhou [pdf] [project]

  7. BERT-EMD: Many-to-Many Layer Mapping for BERT Compression with Earth Mover's Distance EMNLP 2020

    Jianquan Li, Xiaokang Liu, Honghong Zhao, Ruifeng Xu, Min Yang, Yaohong Jin [pdf] [project]

  8. MixKD: Towards Efficient Distillation of Large-scale Language Models ICLR 2021

    Kevin J Liang, Weituo Hao, Dinghan Shen, Yufan Zhou, Weizhu Chen, Changyou Chen, Lawrence Carin [pdf]

  9. Meta-KD: A Meta Knowledge Distillation Framework for Language Model Compression across Domains ACL-IJCNLP 2021

    Haojie Pan, Chengyu Wang, Minghui Qiu, Yichang Zhang, Yaliang Li, Jun Huang [pdf]

  10. MATE-KD: Masked Adversarial TExt, a Companion to Knowledge Distillation ACL-IJCNLP 2021

    Ahmad Rashid, Vasileios Lioutas, Mehdi Rezagholizadeh [pdf]

  11. Structural Knowledge Distillation: Tractably Distilling Information for Structured Predictor ACL-IJCNLP 2021

    Xinyu Wang, Yong Jiang, Zhaohui Yan, Zixia Jia, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu [pdf] [project]

  12. Weight Distillation: Transferring the Knowledge in Neural Network Parameters ACL-IJCNLP 2021

    Ye Lin, Yanyang Li, Ziyang Wang, Bei Li, Quan Du, Tong Xiao, Jingbo Zhu [pdf]

  13. Marginal Utility Diminishes: Exploring the Minimum Knowledge for BERT Knowledge Distillation ACL-IJCNLP 2021

    Yuanxin Liu, Fandong Meng, Zheng Lin, Weiping Wang, Jie Zhou [pdf]

  14. MiniLMv2: Multi-Head Self-Attention Relation Distillation for Compressing Pretrained Transformers Findings of ACL-IJCNLP 2021

    Wenhui Wang, Hangbo Bao, Shaohan Huang, Li Dong, Furu Wei [pdf] [project]

  15. One Teacher is Enough? Pre-trained Language Model Distillation from Multiple Teachers Findings of ACL-IJCNLP 2021

    Chuhan Wu, Fangzhao Wu, Yongfeng Huang [pdf]

Dynamic Early Exiting

  1. DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference ACL 2020

    Ji Xin, Raphael Tang, Jaejun Lee, Yaoliang Yu, Jimmy Lin [pdf] [project]

  2. FastBERT: a Self-distilling BERT with Adaptive Inference Time ACL 2020

    Weijie Liu, Peng Zhou, Zhe Zhao, Zhiruo Wang, Haotang Deng, Qi Ju [pdf] [project]

  3. The Right Tool for the Job: Matching Model and Instance Complexities ACL 2020

    Roy Schwartz, Gabriel Stanovsky, Swabha Swayamdipta, Jesse Dodge, Noah A. Smith [pdf] [project]

  4. A Global Past-Future Early Exit Method for Accelerating Inference of Pre-trained Language Models NAACL 2021

    Kaiyuan Liao, Yi Zhang, Xuancheng Ren, Qi Su, Xu Sun, Bin He [pdf] [project]

  5. CascadeBERT: Accelerating Inference of Pre-trained Language Models via Calibrated Complete Models Cascade Preprint

    Lei Li, Yankai Lin, Deli Chen, Shuhuai Ren, Peng Li, Jie Zhou, Xu Sun [pdf] [project]

  6. Early Exiting BERT for Efficient Document Ranking SustaiNLP 2020

    Ji Xin, Rodrigo Nogueira, Yaoliang Yu, and Jimmy Lin [pdf] [project]

  7. BERxiT: Early Exiting for BERT with Better Fine-Tuning and Extension to Regression EACL 2021

    Ji Xin, Raphael Tang, Yaoliang Yu, and Jimmy Lin [pdf] [project]

  8. Accelerating BERT Inference for Sequence Labeling via Early-Exit ACL 2021

    Xiaonan Li, Yunfan Shao, Tianxiang Sun, Hang Yan, Xipeng Qiu, Xuanjing Huang [pdf] [project]

  9. BERT Loses Patience: Fast and Robust Inference with Early Exit NeurIPS 2020

    Wangchunshu Zhou, Canwen Xu, Tao Ge, Julian McAuley, Ke Xu, Furu Wei [pdf] [project]

  10. Early Exiting with Ensemble Internal Classifiers Preprint

    Tianxiang Sun, Yunhua Zhou, Xiangyang Liu, Xinyu Zhang, Hao Jiang, Zhao Cao, Xuanjing Huang, Xipeng Qiu [pdf]

Quantization

  1. Q-BERT: Hessian Based Ultra Low Precision Quantization of BERT AAAI 2020

    Sheng Shen, Zhen Dong, Jiayu Ye, Linjian Ma, Zhewei Yao, Amir Gholami, Michael W. Mahoney, Kurt Keutzer [pdf] [project]

  2. TernaryBERT: Distillation-aware Ultra-low Bit BERT EMNLP 2020

    Wei Zhang, Lu Hou, Yichun Yin, Lifeng Shang, Xiao Chen, Xin Jiang, Qun Liu [pdf] [project]

  3. Q8BERT: Quantized 8Bit BERT NeurIPS 2019 Workshop

    Ofir Zafrir, Guy Boudoukh, Peter Izsak, Moshe Wasserblat [pdf] [project]

  4. BinaryBERT: Pushing the Limit of BERT Quantization EMNLP 2020

    Haoli Bai, Wei Zhang, Lu Hou, Lifeng Shang, Jing Jin, Xin Jiang, Qun Liu, Michael Lyu, Irwin King [pdf] [project]

  5. I-BERT: Integer-only BERT Quantization ICML 2021

    Sehoon Kim, Amir Gholami, Zhewei Yao, Michael W. Mahoney, Kurt Keutzer [pdf] [project]

Pruning

  1. Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned ACL 2019

    Elena Voita, David Talbot, Fedor Moiseev, Rico Sennrich, Ivan Titov [pdf] [project]

  2. Are Sixteen Heads Really Better than One? NeurIPS 2019

    Paul Michel, Omer Levy, Graham Neubig [pdf] [project]

  3. The Lottery Ticket Hypothesis for Pre-trained BERT Networks NeurIPS 2020

    Tianlong Chen, Jonathan Frankle, Shiyu Chang, Sijia Liu, Yang Zhang, Zhangyang Wang, Michael Carbin [pdf] [project]

  4. Movement Pruning: Adaptive Sparsity by Fine-Tuning NeurIPS 2020

    Victor Sanh, Thomas Wolf, Alexander M. Rush [pdf] [project]

  5. Reducing Transformer Depth on Demand with Structured Dropout Preprint

    Angela Fan, Edouard Grave, Armand Joulin [pdf]

  6. When BERT Plays the Lottery, All Tickets Are Winning EMNLP 2020

    Sai Prasanna, Anna Rogers, Anna Rumshisky [pdf] [project]

  7. Structured Pruning of a BERT-based Question Answering Model Preprint

    J.S. McCarley, Rishav Chakravarti, Avirup Sil [pdf]

  8. Structured Pruning of Large Language Models EMNLP 2020

    Ziheng Wang, Jeremy Wohlwend, Tao Lei [pdf] [project]

  9. Rethinking Network Pruning -- under the Pre-train and Fine-tune Paradigm NAACL 2021

    Dongkuan Xu, Ian E.H. Yen, Jinxi Zhao, Zhibin Xiao [pdf]

  10. Super Tickets in Pre-Trained Language Models: From Model Compression to Improving Generalization ACL 2021

    Chen Liang, Simiao Zuo, Minshuo Chen, Haoming Jiang, Xiaodong Liu, Pengcheng He, Tuo Zhao, Weizhu Chen [pdf] [project]

Contribution

If you find any related work not included in the list, do not hesitate to raise a PR to help us complete the list.

Owner
Tobias Lee
On the way becoming an NLPer.
Tobias Lee
Every Google, Azure & IBM text to speech voice for free

TTS-Grabber Quick thing i made about a year ago to download any text with any tts voice, over 630 voices to choose from currently. It will split the i

16 Dec 07, 2022
Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis (SV2TTS)

This repository is an implementation of Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis (SV2TTS) with a vocoder that works in real-time. Feel free to check my the

Corentin Jemine 38.5k Jan 03, 2023
This repository collects together basic linguistic processing data for using dataset dumps from the Common Voice project

Common Voice Utils This repository collects together basic linguistic processing data for using dataset dumps from the Common Voice project. It aims t

Francis Tyers 40 Dec 20, 2022
Simple multilingual lemmatizer for Python, especially useful for speed and efficiency

Simplemma: a simple multilingual lemmatizer for Python Purpose Lemmatization is the process of grouping together the inflected forms of a word so they

Adrien Barbaresi 70 Dec 29, 2022
✨Fast Coreference Resolution in spaCy with Neural Networks

✨ NeuralCoref 4.0: Coreference Resolution in spaCy with Neural Networks. NeuralCoref is a pipeline extension for spaCy 2.1+ which annotates and resolv

Hugging Face 2.6k Jan 04, 2023
Words_And_Phrases - Just a repo for useful words and phrases that might come handy in some scenarios. Feel free to add yours

Words_And_Phrases Just a repo for useful words and phrases that might come handy in some scenarios. Feel free to add yours Abbreviations Abbreviation

Subhadeep Mandal 1 Feb 01, 2022
Persian-lexicon - A lexicon of 70K unique Persian (Farsi) words

Persian Lexicon This repo uses Uppsala Persian Corpus (UPC) to construct a lexic

Saman Vaisipour 7 Apr 01, 2022
IndoBERTweet is the first large-scale pretrained model for Indonesian Twitter. Published at EMNLP 2021 (main conference)

IndoBERTweet 🐦 🇮🇩 1. Paper Fajri Koto, Jey Han Lau, and Timothy Baldwin. IndoBERTweet: A Pretrained Language Model for Indonesian Twitter with Effe

IndoLEM 40 Nov 30, 2022
NewsMTSC: (Multi-)Target-dependent Sentiment Classification in News Articles

NewsMTSC: (Multi-)Target-dependent Sentiment Classification in News Articles NewsMTSC is a dataset for target-dependent sentiment classification (TSC)

Felix Hamborg 79 Dec 30, 2022
Crowd sourced training data for Rasa NLU models

NLU Training Data Crowd-sourced training data for the development and testing of Rasa NLU models. If you're interested in grabbing some data feel free

Rasa 169 Dec 26, 2022
This repository is home to the Optimus data transformation plugins for various data processing needs.

Transformers Optimus's transformation plugins are implementations of Task and Hook interfaces that allows execution of arbitrary jobs in optimus. To i

Open Data Platform 37 Dec 14, 2022
中文生成式预训练模型

T5 PEGASUS 中文生成式预训练模型,以mT5为基础架构和初始权重,通过类似PEGASUS的方式进行预训练。 详情可见:https://kexue.fm/archives/8209 Tokenizer 我们将T5 PEGASUS的Tokenizer换成了BERT的Tokenizer,它对中文更

410 Jan 03, 2023
A python gui program to generate reddit text to speech videos from the id of any post.

Reddit text to speech generator A python gui program to generate reddit text to speech videos from the id of any post. Current functionality Generate

Aadvik 17 Dec 19, 2022
Various Algorithms for Short Text Mining

Short Text Mining in Python Introduction This package shorttext is a Python package that facilitates supervised and unsupervised learning for short te

Kwan-Yuet 466 Dec 06, 2022
Segmenter - Transformer for Semantic Segmentation

Segmenter - Transformer for Semantic Segmentation

592 Dec 27, 2022
Amazon Multilingual Counterfactual Dataset (AMCD)

Amazon Multilingual Counterfactual Dataset (AMCD)

35 Sep 20, 2022
American Sign Language (ASL) to Text Converter

Signterpreter American Sign Language (ASL) to Text Converter Recommendations Although there is grayscale and gaussian blur, we recommend that you use

0 Feb 20, 2022
Generate vector graphics from a textual caption

VectorAscent: Generate vector graphics from a textual description Example "a painting of an evergreen tree" python text_to_painting.py --prompt "a pai

Ajay Jain 97 Dec 15, 2022
Coreference resolution for English, French, German and Polish, optimised for limited training data and easily extensible for further languages

Coreferee Author: Richard Paul Hudson, Explosion AI 1. Introduction 1.1 The basic idea 1.2 Getting started 1.2.1 English 1.2.2 French 1.2.3 German 1.2

Explosion 70 Dec 12, 2022
RoNER is a Named Entity Recognition model based on a pre-trained BERT transformer model trained on RONECv2

RoNER RoNER is a Named Entity Recognition model based on a pre-trained BERT transformer model trained on RONECv2. It is meant to be an easy to use, hi

Stefan Dumitrescu 9 Nov 07, 2022