IOT: Instance-wise Layer Reordering for Transformer Structures

Related tags

Deep LearningIOT
Overview

Introduction

This repository contains the code for Instance-wise Ordered Transformer (IOT), which is introduced in the ICLR2021 paper IOT: Instance-wise Layer Reordering for Transformer Structures.

If you find this work helpful in your research, please cite as:

@inproceedings{
zhu2021iot,
title={{\{}IOT{\}}: Instance-wise Layer Reordering for Transformer Structures},
author={Jinhua Zhu and Lijun Wu and Yingce Xia and Shufang Xie and Tao Qin and Wengang Zhou and Houqiang Li and Tie-Yan Liu},
booktitle={International Conference on Learning Representations},
year={2021},
url={https://openreview.net/forum?id=ipUPfYxWZvM}
}

Requirements and Installation

  • PyTorch version == 1.0.0
  • Python version >= 3.5

To install IOT:

git clone https://github.com/instance-wise-ordered-transformer/IOT
cd IOT
pip install --editable .

Getting Started

Take IWSLT14 De-En translation as an example.

Data Preprocessing

cd examples/translation/
bash prepare-iwslt14.sh
cd ../..

TEXT=examples/translation/iwslt14.tokenized.de-en
python preprocess.py --source-lang de --target-lang en \
    --trainpref $TEXT/train --validpref $TEXT/valid --testpref $TEXT/test \
    --destdir data-bin/iwslt14.tokenized.de-en --joined-dictionary

Training

Encoder order is set to be the default one without reordering (ENCODER_MAX_ORDER=1), since the paper finds that both reordering encoder and decoder is not good as reordering decoder only.

#!/bin/bash
export CUDA_VISIBLE_DEVICES=${1:-0}
nvidia-smi

ENCODER_MAX_ORDER=1
DECODER_MAX_ORDER=3
DECODER_ORDER="0 3 5"
DIVERSITY=0.1
GS_MAX=20
GS_MIN=2
GS_R=0
GS_UF=5000
KL=0.01
CLAMPVAL=0.05

DECODER_ORDER_NAME=`echo $DECODER_ORDER | sed 's/ //g'`
SAVE_DIR=checkpoints/dec_${DECODER_MAX_ORDER}_order_${DECODER_ORDER_NAME}_div_${DIVERSITY}_gsmax_${GS_MAX}_gsmin_${GS_MIN}_gsr_${GS_R}_gsuf_${GS_UF}_kl_${KL}_clampval_${CLAMPVAL}
mkdir -p ${SAVE_DIR}

python -u train.py data-bin/iwslt14.tokenized.de-en -a transformer_iwslt_de_en \
--optimizer adam --lr 0.0005 -s de -t en --label-smoothing 0.1 --dropout 0.3 --max-tokens 4000 \
--min-lr 1e-09 --lr-scheduler inverse_sqrt --weight-decay 0.0001 --criterion label_smoothed_cross_entropy \
--max-update 100000 --warmup-updates 4000 --warmup-init-lr 1e-07 --adam-betas '(0.9,0.98)' \
--save-dir $SAVE_DIR --share-all-embeddings  --gs-clamp --decoder-orders $DECODER_ORDER  \
--encoder-max-order $ENCODER_MAX_ORDER  --decoder-max-order $DECODER_MAX_ORDER  --diversity $DIVERSITY \
--gumbel-softmax-max $GS_MAX  --gumbel-softmax-min $GS_MIN --gumbel-softmax-tau-r $GS_R  --gumbel-softmax-update-freq $GS_UF \
--kl $KL --clamp-value $CLAMPVAL | tee -a ${SAVE_DIR}/train.log

Evaluation

#!/bin/bash
set -x
set -e

pip install -e . --user
export CUDA_VISIBLE_DEVICES=${1:-0}
nvidia-smi

ENCODER_MAX_ORDER=1
DECODER_MAX_ORDER=3
DECODER_ORDER="0 3 5"
DIVERSITY=0.1
GS_MAX=20
GS_MIN=2
GS_R=0
GS_UF=5000
KL=0.01
CLAMPVAL=0.05

DECODER_ORDER_NAME=`echo $DECODER_ORDER | sed 's/ //g'`
SAVE_DIR=checkpoints/dec_${DECODER_MAX_ORDER}_order_${DECODER_ORDER_NAME}_div_${DIVERSITY}_gsmax_${GS_MAX}_gsmin_${GS_MIN}_gsr_${GS_R}_gsuf_${GS_UF}_kl_${KL}_clampval_${CLAMPVAL}

python generate.py data-bin/iwslt14.tokenized.de-en \
  --path $SAVE_DIR/checkpint_best.pt \
  --batch-size 128 --beam 5 --remove-bpe --quiet --num-ckts $DECODER_MAX_ORDER 
Streamlit App For Product Analysis - Streamlit App For Product Analysis

Streamlit_App_For_Product_Analysis Здравствуйте! Перед вами дашборд, позволяющий

Grigory Sirotkin 1 Jan 10, 2022
A dataset for online Arabic calligraphy

Calliar Calliar is a dataset for Arabic calligraphy. The dataset consists of 2500 json files that contain strokes manually annotated for Arabic callig

ARBML 114 Dec 28, 2022
WSDM2022 Challenge - Large scale temporal graph link prediction

WSDM 2022 Large-scale Temporal Graph Link Prediction - Baseline and Initial Test Set WSDM Cup Website link Link to this challenge This branch offers A

Deep Graph Library 34 Dec 29, 2022
Roger Labbe 13k Dec 29, 2022
Distributed Deep learning with Keras & Spark

Elephas: Distributed Deep Learning with Keras & Spark Elephas is an extension of Keras, which allows you to run distributed deep learning models at sc

Max Pumperla 1.6k Jan 05, 2023
This Jupyter notebook shows one way to implement a simple first-order low-pass filter on sampled data in discrete time.

How to Implement a First-Order Low-Pass Filter in Discrete Time We often teach or learn about filters in continuous time, but then need to implement t

Joshua Marshall 4 Aug 24, 2022
This is the official PyTorch implementation of the paper "TransFG: A Transformer Architecture for Fine-grained Recognition" (Ju He, Jie-Neng Chen, Shuai Liu, Adam Kortylewski, Cheng Yang, Yutong Bai, Changhu Wang, Alan Yuille).

TransFG: A Transformer Architecture for Fine-grained Recognition Official PyTorch code for the paper: TransFG: A Transformer Architecture for Fine-gra

Ju He 307 Jan 03, 2023
Interactive Image Generation via Generative Adversarial Networks

iGAN: Interactive Image Generation via Generative Adversarial Networks Project | Youtube | Paper Recent projects: [pix2pix]: Torch implementation for

Jun-Yan Zhu 3.9k Dec 23, 2022
PyTorch implementation of SMODICE: Versatile Offline Imitation Learning via State Occupancy Matching

SMODICE: Versatile Offline Imitation Learning via State Occupancy Matching This is the official PyTorch implementation of SMODICE: Versatile Offline I

Jason Ma 14 Aug 30, 2022
Attention mechanism with MNIST dataset

[TensorFlow] Attention mechanism with MNIST dataset Usage $ python run.py Result Training Loss graph. Test Each figure shows input digit, attention ma

YeongHyeon Park 12 Jun 10, 2022
Backdoor Attack through Frequency Domain

Backdoor Attack through Frequency Domain DEPENDENCIES python==3.8.3 numpy==1.19.4 tensorflow==2.4.0 opencv==4.5.1 idx2numpy==1.2.3 pytorch==1.7.0 Data

5 Jun 18, 2022
DaReCzech is a dataset for text relevance ranking in Czech

Dataset DaReCzech is a dataset for text relevance ranking in Czech. The dataset consists of more than 1.6M annotated query-documents pairs,

Seznam.cz a.s. 8 Jul 26, 2022
SeqAttack: a framework for adversarial attacks on token classification models

A framework for adversarial attacks against token classification models

Walter 23 Nov 25, 2022
Official PaddlePaddle implementation of Paint Transformer

Paint Transformer: Feed Forward Neural Painting with Stroke Prediction [Paper] [Paddle Implementation] Update We have optimized the serial inference p

TianweiLin 284 Dec 31, 2022
Reproducible research and reusable acyclic workflows in Python. Execute code on HPC systems as if you executed them on your personal computer!

Reproducible research and reusable acyclic workflows in Python. Execute code on HPC systems as if you executed them on your machine! Motivation Would

Joeri Hermans 15 Sep 11, 2022
2020 CCF大数据与计算智能大赛-非结构化商业文本信息中隐私信息识别-第7名方案

2020CCF-NER 2020 CCF大数据与计算智能大赛-非结构化商业文本信息中隐私信息识别-第7名方案 bert base + flat + crf + fgm + swa + pu learning策略 + clue数据集 = test1单模0.906 词向量

67 Oct 19, 2022
Create UIs for prototyping your machine learning model in 3 minutes

Note: We just launched Hosted, where anyone can upload their interface for permanent hosting. Check it out! Welcome to Gradio Quickly create customiza

Gradio 11.7k Jan 07, 2023
project page for VinVL

VinVL: Revisiting Visual Representations in Vision-Language Models Updates 02/28/2021: Project page built. Introduction This repository is the project

308 Jan 09, 2023
A repository for benchmarking neural vocoders by their quality and speed.

License The majority of VocBench is licensed under CC-BY-NC, however portions of the project are available under separate license terms: Wavenet, Para

Meta Research 177 Dec 12, 2022
Pytorch Lightning code guideline for conferences

Deep learning project seed Use this seed to start new deep learning / ML projects. Built in setup.py Built in requirements Examples with MNIST Badges

Pytorch Lightning 1k Jan 02, 2023