An implementation of the Pay Attention when Required transformer

Overview

Pay Attention when Required (PAR) Transformer-XL

An implementation of the Pay Attention when Required transformer from the paper: https://arxiv.org/pdf/2009.04534.pdf

alt text [source: Jonathan Kernes]

Quick overview

The Pay Attention when Required Transformer (Mandava, et. al. 2020) is just a regular transformer-XL (Dai et. al. 2019)[https://arxiv.org/pdf/1901.02860.pdf] , but the ratio of attention and dense layers has been optimized. This optimization is performed by allowing the network to choose which types of layer it prefers in each block of the network. The present implementation is not an exact replica of the author's efforts. Instead, we perform a simultaneous optimization procedure on both the model architecture and model parameters. The search is performed using a SuperNet, which is a sequential neural network composed of stochastic blocks, as shown in the figure below (taken from the paper. Please don't sue me!)

alt text [Source: Mandava et. al. 2020]

The key component is a Gumbel-Softmax layer [(Jang et al., 2016) and (Maddison et al., 2016). jang link: https://arxiv.org/pdf/1611.01144.pdf]. This layer is a continuous representation of a discrete sampling from a Categorical distribution, thereby allowing us to use gradients to learn parameters of a discrete distribution. (Recall a categorical is a distrbution over K states with kth state having probability pi_k, and we must have the normalization condition \sum_{i=1}^K pi_i = 1)

As the model learns, it is free to adjust both the usual model parameters, as well as its architecture search parameters pi, indicating the probability of choosing either

  1. Attention

  2. Dense

  3. Identity

for any given stochastic block. We perform simulated annealing: since the categorical distribution is approximated by a continuous representation, we get some scores like (0.02, 0.98, 0.02) for the probability of say sampling that state 2 is picked. The sharpness of this is set by a parameter \tau (the temperature), with a categorical distribution the limit tau-->0. Simulated annealing means we begin with tau=1 to let the model figure out what it wants, then slowly decrease tau so the distribution approaches a categorical.

All of this is implemented on the freely available wiki-text2 dataset.

Explanation of the main GIF: The main gif is the result of our experiments. It shows the pi distribution for each stochastic block of a 6 block SuperNet, as a function of training iterations. The number indicates the probability of the most likely layer type (darker means more probable). As you can see, the model learns to put attention in the beginning, and dense layers at the end.

Requirements

Usual ML stuff, if you have a conda environment, python 3+, TensorFlow 2+ you should be ok. You will need TensorFlow Text as well to handle the SentencePiece Tokenization

If you choose to run your own tokenizer (a flag option in data_utils for handling new text data), you will also need to download the SentencePiece package: https://github.com/google/sentencepiece

Data

The dataset used is Wiki-text2. We have provided a copy of this in the data folder, along with some preprocessed data for training. In order to reproduce this from scratch, run the shell script

./create_tfrecords.sh

This will download the wiki-text2 dataset from its source, then proceed to clean, batch, and write the data to a tfrecords file. The shell script calls build_data.py which offers more control over what type of data to generate. The general parameters you will want to tune are:

*batch_size *seq_len.

You can also supply your own dataset instead of the one provided. The underlying tokenizer uses sentencepiece (Kudo): https://github.com/google/sentencepiece, which works at the byte level and can handle any kind of input. Simply change the --input_text flag to your file, and set the desired --vocab_size.

Why do we need to specify the batch size? Transformer XL uses memory states to form a recurrent, long range network. After analyzing a particular sequence say [A,B] of the sequence [A,B,C,D], the results of [A,B] are fed into the [C,D] calculation with a stop gradient. Therefore, we must be sure that each datapoint follows chronologically from the previous one.

This is achieved by context batching (see data_utils.py function) where we break the entire dataset into batch_size segments, then pull in order one sequence from each batch at a time to form the dataset. Because of this, note that adding more shards to the data could result in a large loss (order of batch_size*seq_len*shards), as each shard will drop the remaining datapoint of size (batch_size*seq_len) to keep the tensor shapes.

Addtional technical details

Per the original Transformer-XL, we also implement an adaptive softmax layer (Grave et. al. 2017, https://arxiv.org/abs/1609.04309) to deal with a potentially large number of outputs in the final dense layer. This implemenation is inspired by the TF 1.0 example at https://github.com/yangsaiyong/tf-adaptive-softmax-lstm-lm. To use the adaptive softmax, set the --cutoffs= flag in train.py. The cutoffs are the max values of each bin, and should NOT include the vocab size (i.e. the max cutoff of the final bin). If no cutoffs are specified, the model defaults to normal softmax.

For completeness, we have also provided a script optimal_cuts.py that determines the optimal cutoffs given a return space separated file of unigram probabilities (based on the assumptions of Grave et. al. regarding GPU computation complexity -- see the paper for details). The algorithm uses dynamic programming, but is quite slow at O(KN^2), for K cutoffs and N vocab words. In principle it's a one time cost to determine the cutoffs, but we are impatient and recommend to just play around with the cutoffs instead. See the script for flag details

Training and Benchmarks

The default model we use has memory length 16, feed-forward dimension 1024, attention dimension 128, and 6 stochastic blocks, with an adaptive softmax layer and 2 clusters. We trained on a colab GPU for 20 epochs, taking a total of 37 minutes. We use an Adam optimzer with cosine rate decay: an initial warmup of 4000 steps and a maximum learning rate of 1e-4, decaying to zero at the end of training. Our training benchmarks are:

Iteration (thousands) Train_perplexity Validation_perplexity Time
2.7k 163.9 114.4 1m 58s
8.5k 78.56 62.33 5m 37s
14.1k 65.71 51.88 9m 28s
28.3k 48.52 42.61 18m 40s
48.1k 41.85 39.57 31m 51s
56.5k 42.12 39.41 37m 14s

To train, simply run the shell script

./base_model.sh

adjusting the parameters as you see fit. The above model is the default configuration. To train in colab, simply open up the notebook "colab.ipynb" and follow the instructions. This is most easily done by going to [google.colab.com] and searching this repository in github. The benefit of colab, is it's easier to play around with the model after training.

While training, we have provided two ways to monitor the output

  1. A tensorboard log. The colab notebook takes care of running this for you. In the terminal, first create a 'logs' directory, then run the command tensorboard --logdir logs in a separate tab. This will open a port where you can view live plots of the learning rate, tau annealing, train/valid loss and perplexity.

  2. An output log saved to training_log.log. This will log the model summary, parameters, etc. as well as print out loss updates every 100 steps and save it to the log file.

Thanks for reading this far!

Enjoy! And thank you to the wonderful researchers that inspired this project.

If you would like to contribute, or have any comments questions concerns please open a pull request or email me directly.

Source code and dataset for ACL 2019 paper "ERNIE: Enhanced Language Representation with Informative Entities"

ERNIE Source code and dataset for "ERNIE: Enhanced Language Representation with Informative Entities" Reqirements: Pytorch=0.4.1 Python3 tqdm boto3 r

THUNLP 1.3k Dec 30, 2022
ZUNIT - Toward Zero-Shot Unsupervised Image-to-Image Translation

ZUNIT Dependencies you can install all the dependencies by pip install -r requirements.txt Datasets Download CUB dataset. Unzip the birds.zip at ./da

Chen Yuanqi 9 Jun 24, 2022
An extension for asreview implements a version of the tf-idf feature extractor that saves the matrix and the vocabulary.

Extension - matrix and vocabulary extractor for TF-IDF and Doc2Vec An extension for ASReview that adds a tf-idf extractor that saves the matrix and th

ASReview 4 Jun 17, 2022
We have built a Voice based Personal Assistant for people to access files hands free in their device using natural language processing.

Voice Based Personal Assistant We have built a Voice based Personal Assistant for people to access files hands free in their device using natural lang

Rushabh 2 Nov 13, 2021
Python bot created with Selenium that can guess the daily Wordle word correct 96.8% of the time.

Wordle_Bot Python bot created with Selenium that can guess the daily Wordle word correct 96.8% of the time. It will log onto the wordle website and en

Lucas Polidori 15 Dec 11, 2022
This repo contains simple to use, pretrained/training-less models for speaker diarization.

PyDiar This repo contains simple to use, pretrained/training-less models for speaker diarization. Supported Models Binary Key Speaker Modeling Based o

12 Jan 20, 2022
PyTorch Implementation of VAENAR-TTS: Variational Auto-Encoder based Non-AutoRegressive Text-to-Speech Synthesis.

VAENAR-TTS - PyTorch Implementation PyTorch Implementation of VAENAR-TTS: Variational Auto-Encoder based Non-AutoRegressive Text-to-Speech Synthesis.

Keon Lee 67 Nov 14, 2022
Conversational text Analysis using various NLP techniques

Conversational text Analysis using various NLP techniques

Rita Anjana 159 Jan 06, 2023
Repo for Enhanced Seq2Seq Autoencoder via Contrastive Learning for Abstractive Text Summarization

ESACL: Enhanced Seq2Seq Autoencoder via Contrastive Learning for AbstractiveText Summarization This repo is for our paper "Enhanced Seq2Seq Autoencode

Rachel Zheng 14 Nov 01, 2022
一个基于Nonebot2和go-cqhttp的娱乐性qq机器人

Takker - 一个普通的QQ机器人 此项目为基于 Nonebot2 和 go-cqhttp 开发,以 Sqlite 作为数据库的QQ群娱乐机器人 关于 纯兴趣开发,部分功能借鉴了大佬们的代码,作为Q群的娱乐+功能性Bot 声明 此项目仅用于学习交流,请勿用于非法用途 这是开发者的第一个Pytho

风屿 79 Dec 29, 2022
Dual languaged (rus+eng) tool for packing and unpacking archives of Silky Engine.

SilkyArcTool English Dual languaged (rus+eng) GUI tool for packing and unpacking archives of Silky Engine. It is not the same arc as used in Ai6WIN. I

Tester 5 Sep 15, 2022
Sploitus - Command line search tool for sploitus.com. Think searchsploit, but with more POCs

Sploitus Command line search tool for sploitus.com. Think searchsploit, but with

watchdog2000 5 Mar 07, 2022
This is a general repo that helps you develop fast/effective NLP classifiers using Huggingface

NLP Classifier Introduction This project trains a bert model on any NLP classifcation model. And uses the model in make predictions on new data using

Abdullah Tarek 3 Mar 11, 2022
Converts python code into c++ by using OpenAI CODEX.

🦾 codex_py2cpp 🤖 OpenAI Codex Python to C++ Code Generator Your Python Code is too slow? 🐌 You want to speed it up but forgot how to code in C++? ⌨

Alexander 423 Jan 01, 2023
结巴中文分词

jieba “结巴”中文分词:做最好的 Python 中文分词组件 "Jieba" (Chinese for "to stutter") Chinese text segmentation: built to be the best Python Chinese word segmentation

Sun Junyi 29.8k Jan 02, 2023
LightSeq: A High-Performance Inference Library for Sequence Processing and Generation

LightSeq is a high performance inference library for sequence processing and generation implemented in CUDA. It enables highly efficient computation of modern NLP models such as BERT, GPT2, Transform

Bytedance Inc. 2.5k Jan 03, 2023
FB ID CLONER WUTHOT CHECKPOINT, FACEBOOK ID CLONE FROM FILE

* MY SOCIAL MEDIA : Programming And Memes Want to contact Mr. Error ? CONTACT : [ema

Mr. Error 9 Jun 17, 2021
Codes to pre-train Japanese T5 models

t5-japanese Codes to pre-train a T5 (Text-to-Text Transfer Transformer) model pre-trained on Japanese web texts. The model is available at https://hug

Megagon Labs 37 Dec 25, 2022
🚀 RocketQA, dense retrieval for information retrieval and question answering, including both Chinese and English state-of-the-art models.

In recent years, the dense retrievers based on pre-trained language models have achieved remarkable progress. To facilitate more developers using cutt

475 Jan 04, 2023
IEEEXtreme15.0 Questions And Answers

IEEEXtreme15.0 Questions And Answers IEEEXtreme is a global challenge in which teams of IEEE Student members – advised and proctored by an IEEE member

Dilan Perera 15 Oct 24, 2022