LV-BERT: Exploiting Layer Variety for BERT (Findings of ACL 2021)

Overview

LV-BERT

Introduction

In this repo, we introduce LV-BERT by exploiting layer variety for BERT. For detailed description and experimental results, please refer to our paper LV-BERT: Exploiting Layer Variety for BERT (Findings of ACL 2021).

Requirements

  • Python 3.6
  • TensorFlow 1.15
  • numpy
  • scikit-learn

Experiments

Firstly, set your data dir (absolute) to place datasets and models by

DATA_DIR=/path/to/data/dir

Fine-tining

We give the instruction to fine-tune a pre-trained LV-BERT-small (13M parameters) on GLUE. You can refer to this Google Colab notebook for a quick example. All models of different are provided this Google Drive folder. The models are pre-trained 1M steps with sequence length 128 to save compute. *_seq512 named models are trained for more 100K steps with sequence length 512 whichs are used for long-sequence tasks like SQuAD. See our paper for more details on model performance.

  1. Create your data directory.
mkdir -p $DATA_DIR/models && cp vocab.txt $DATA_DIR/

Put the pre-trained model in the corresponding directory

mv lv-bert_small $DATA_DIR/models/
  1. Download the GLUE data by running
python3 download_glue_data.py
  1. Set up the data by running
cd glue_data && mv CoLA cola && mv MNLI mnli && mv MRPC mrpc && mv QNLI qnli && mv QQP qqp && mv RTE rte && mv SST-2 sst && mv STS-B sts && mv diagnostic/diagnostic.tsv mnli && mkdir -p $DATA_DIR/finetuning_data && mv * $DATA_DIR/finetuning_data && cd ..
  1. Fine-tune the model by running
bash finetune.sh $DATA_DIR

PS: (a) You can test different tasks by changing configs in finetune.sh. (b) Some of the datasets on GLUE are small, causing that the results may vary substantially for different random seeds. The same as ELECTRA, we report the median of 10 fine-tuning runs from the same pre-trained model for each result.

Pre-training

We give the instruction to pre-train LV-BERT-small (13M parameters) using the OpenWebText corpus.

  1. First download the OpenWebText pre-traing corpus (12G).

  2. After downloading the pre-training corpus, build the pre-training dataset tf-record by running

bash build_data.sh $DATA_DIR
  1. Then, pre-train the model by running
bash pretrain.sh $DATA_DIR

Bibtex

@inproceedings{yu2021lv-bert,
        author = {Yu, Weihao and Jiang, Zihang and Chen, Fei, Hou, Qibin and Feng, Jiashi},
        title = {LV-BERT: Exploiting Layer Variety for BERT},
        booktitle = {Findings of ACL},
        month = {August},
        year = {2021}
}

Reference

This repo is based on the repo ELECTRA.

You might also like...
Pre-trained BERT Models for Ancient and Medieval Greek, and associated code for LaTeCH 2021 paper titled -
Pre-trained BERT Models for Ancient and Medieval Greek, and associated code for LaTeCH 2021 paper titled - "A Pilot Study for BERT Language Modelling and Morphological Analysis for Ancient and Medieval Greek"

Ancient Greek BERT The first and only available Ancient Greek sub-word BERT model! State-of-the-art post fine-tuning on Part-of-Speech Tagging and Mor

Pytorch implementation for the EMNLP 2020 (Findings) paper: Connecting the Dots: A Knowledgeable Path Generator for Commonsense Question Answering

Path-Generator-QA This is a Pytorch implementation for the EMNLP 2020 (Findings) paper: Connecting the Dots: A Knowledgeable Path Generator for Common

🌈 PyTorch Implementation for EMNLP'21 Findings
🌈 PyTorch Implementation for EMNLP'21 Findings "Reasoning Visual Dialog with Sparse Graph Learning and Knowledge Transfer"

SGLKT-VisDial Pytorch Implementation for the paper: Reasoning Visual Dialog with Sparse Graph Learning and Knowledge Transfer Gi-Cheon Kang, Junseok P

Deep generative modeling for time-stamped heterogeneous data, enabling high-fidelity models for a large variety of spatio-temporal domains.
Deep generative modeling for time-stamped heterogeneous data, enabling high-fidelity models for a large variety of spatio-temporal domains.

Neural Spatio-Temporal Point Processes [arxiv] Ricky T. Q. Chen, Brandon Amos, Maximilian Nickel Abstract. We propose a new class of parameterizations

LaneDet is an open source lane detection toolbox based on PyTorch that aims to pull together a wide variety of state-of-the-art lane detection models
LaneDet is an open source lane detection toolbox based on PyTorch that aims to pull together a wide variety of state-of-the-art lane detection models

LaneDet is an open source lane detection toolbox based on PyTorch that aims to pull together a wide variety of state-of-the-art lane detection models. Developers can reproduce these SOTA methods and build their own methods.

Code and dataset for ACL2018 paper "Exploiting Document Knowledge for Aspect-level Sentiment Classification"

Aspect-level Sentiment Classification Code and dataset for ACL2018 [paper] ‘‘Exploiting Document Knowledge for Aspect-level Sentiment Classification’’

Moving Object Segmentation in 3D LiDAR Data: A Learning-based Approach Exploiting Sequential Data
Moving Object Segmentation in 3D LiDAR Data: A Learning-based Approach Exploiting Sequential Data

LiDAR-MOS: Moving Object Segmentation in 3D LiDAR Data This repo contains the code for our paper: Moving Object Segmentation in 3D LiDAR Data: A Learn

This repository contains codes of ICCV2021 paper: SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation

SO-Pose This repository contains codes of ICCV2021 paper: SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation This paper is basically an

Comments
  • CVE-2007-4559 Patch

    CVE-2007-4559 Patch

    Patching CVE-2007-4559

    Hi, we are security researchers from the Advanced Research Center at Trellix. We have began a campaign to patch a widespread bug named CVE-2007-4559. CVE-2007-4559 is a 15 year old bug in the Python tarfile package. By using extract() or extractall() on a tarfile object without sanitizing input, a maliciously crafted .tar file could perform a directory path traversal attack. We found at least one unsantized extractall() in your codebase and are providing a patch for you via pull request. The patch essentially checks to see if all tarfile members will be extracted safely and throws an exception otherwise. We encourage you to use this patch or your own solution to secure against CVE-2007-4559. Further technical information about the vulnerability can be found in this blog.

    If you have further questions you may contact us through this projects lead researcher Kasimir Schulz.

    opened by TrellixVulnTeam 0
Owner
Weihao Yu
PhD student at NUS
Weihao Yu
null 190 Jan 3, 2023
The source codes for ACL 2021 paper 'BoB: BERT Over BERT for Training Persona-based Dialogue Models from Limited Personalized Data'

BoB: BERT Over BERT for Training Persona-based Dialogue Models from Limited Personalized Data This repository provides the implementation details for

null 124 Dec 27, 2022
Official repository for "Exploiting Session Information in BERT-based Session-aware Sequential Recommendation", SIGIR 2022 short.

Session-aware BERT4Rec Official repository for "Exploiting Session Information in BERT-based Session-aware Sequential Recommendation", SIGIR 2022 shor

Jamie J. Seol 22 Dec 13, 2022
[NAACL & ACL 2021] SapBERT: Self-alignment pretraining for BERT.

SapBERT: Self-alignment pretraining for BERT This repo holds code for the SapBERT model presented in our NAACL 2021 paper: Self-Alignment Pretraining

Cambridge Language Technology Lab 104 Dec 7, 2022
git git《Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual Tracking》(CVPR 2021) GitHub:git2] 《Masksembles for Uncertainty Estimation》(CVPR 2021) GitHub:git3]

Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual Tracking Ning Wang, Wengang Zhou, Jie Wang, and Houqiang Li Accepted by CVPR

NingWang 236 Dec 22, 2022
Data augmentation for NLP, accepted at EMNLP 2021 Findings

AEDA: An Easier Data Augmentation Technique for Text Classification This is the code for the EMNLP 2021 paper AEDA: An Easier Data Augmentation Techni

Akbar Karimi 81 Dec 9, 2022
Implementation for "Exploiting Aliasing for Manga Restoration" (CVPR 2021)

[CVPR Paper](To appear) | [Project Website](To appear) | BibTex Introduction As a popular entertainment art form, manga enriches the line drawings det

null 133 Dec 15, 2022
An official implementation of "Exploiting a Joint Embedding Space for Generalized Zero-Shot Semantic Segmentation" (ICCV 2021) in PyTorch.

Exploiting a Joint Embedding Space for Generalized Zero-Shot Semantic Segmentation This is an official implementation of the paper "Exploiting a Joint

CV Lab @ Yonsei University 35 Oct 26, 2022
Code for our NeurIPS 2021 paper 'Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation'

Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation (NeurIPS 2021) Code for our NeurIPS 2021 paper 'Exploiting the Intri

Shiqi Yang 53 Dec 25, 2022
Source code for NAACL 2021 paper "TR-BERT: Dynamic Token Reduction for Accelerating BERT Inference"

TR-BERT Source code and dataset for "TR-BERT: Dynamic Token Reduction for Accelerating BERT Inference". The code is based on huggaface's transformers.

THUNLP 37 Oct 30, 2022