ByT5: Towards a token-free future with pre-trained byte-to-byte models

Related tags

Text Data & NLPbyt5
Overview

ByT5: Towards a token-free future with pre-trained byte-to-byte models

ByT5 is a tokenizer-free extension of the mT5 model. Instead of using a subword vocabulary like most other pretrained language models (BERT, XLM-R, T5, GPT-3), our ByT5 model operates directly on UTF-8 bytes, removing the need for any text preprocessing. Beyond the reduction in system complexity, we find that parameter-matched ByT5 models are competitive with mT5 across a range of tasks, and outperform mT5 on tasks that involve noisy text or are sensitive to spelling and pronunciation. This repo can be used to reproduce the experiments in the ByT5 paper.

Usage

Training

To run this code, you need to install the t5 library. General instructions for training, fine-tuning, evaluation, and exporting models for inference can be found in the t5 repo. In order to use the additional ByT5 tasks provided in this library with the t5_mesh_transformer command, run from this directory and add the flag --module_import="byt5.tasks".

To train a ByT5-Large model on the mc4 task from scratch as described in the paper:

export PROJECT=yourproject
export ZONE=yourzone
export BUCKET=yourbucket
export TPU=yourtpu

ctpu up --name=$TPU --project=$PROJECT --zone=$ZONE --tpu-size=v3-256 --tpu-only --noconf

TASK=byt5_mc4
MODEL_DIR="${BUCKET}${TASK}"

python -m t5.models.mesh_transformer_main \
  --tpu="${TPU}" \
  --gcp_project="${PROJECT}" \
  --tpu_zone="${ZONE}" \
  --model_dir="${MODEL_DIR}" \
  --gin_file="models/byt5.large.gin" \
  --gin_param="MIXTURE_NAME = '${TASK}'" \
  --gin_param="utils.run.sequence_length = {'inputs': 1024, 'targets': 189}" \
  --gin_param="utils.run.batch_size = ('tokens_per_batch', 1048576)" \
  --gin_param="[email protected]_rate_schedules.rsqrt_no_ramp_down" \
  --gin_param="run.train_steps = 1000000" \
  --gin_param="utils.tpu_mesh_shape.model_parallelism = 1" \
  --gin_param="utils.tpu_mesh_shape.tpu_topology = 'v3-256'" \
  --eval_mode="perplexity_eval" \
  --eval_gin_param="mesh_eval_dataset_fn.num_eval_examples = 10000" \
  --t5_tfds_data_dir="${BUCKET}/t5-tfds" \
  --module_import="byt5.tasks"

Fine-Tuning

The example below shows how to finetune the ByT5-Large model on the XNLI zeroshot task.

export PROJECT=yourproject
export ZONE=yourzone
export BUCKET=yourbucket
export TPU=yourtpu

ctpu up --name=$TPU --project=$PROJECT --zone=$ZONE --tpu-size=v3-256 --tpu-only --noconf

TASK=byt5_xnli_zeroshot
PRETRAINED_DIR=gs://t5-data/pretrained_models/byt5/large
PRETRAINED_STEPS=1000000
FINETUNE_STEPS=262144
MODEL_DIR="${BUCKET}${TASK}"

# Run fine-tuning
python -m t5.models.mesh_transformer_main \
  --tpu="${TPU}" \
  --gcp_project="${PROJECT}" \
  --tpu_zone="${ZONE}" \
  --model_dir="${MODEL_DIR}" \
  --gin_file="${PRETRAINED_DIR}/operative_config.gin" \
  --gin_param="utils.tpu_mesh_shape.tpu_topology = 'v3-256'" \
  --gin_param="MIXTURE_NAME = '${TASK}'" \
  --gin_param="utils.run.train_steps=$((PRETRAINED_STEPS+FINETUNE_STEPS))" \
  --gin_param="utils.run.init_checkpoint='${PRETRAINED_DIR}/model.ckpt-${PRETRAINED_STEPS}'" \
  --t5_tfds_data_dir="${BUCKET}/t5-tfds" \
  --module_import="byt5.tasks"
  --gin_param="utils.run.batch_size = ('tokens_per_batch', 1048576)" \
  --gin_param="utils.run.sequence_length = {'inputs': 2048, 'targets': 56}"
  --eval_gin_param="Bitransformer.decode.max_decode_length = 56" \

The remaining experiments are shown in the tasks.py file.

Released Model Checkpoints

We have released the following checkpoints for pre-trained models described in our paper:

How to Cite

If you extend or use this work, please cite the paper where it was introduced:

@misc{xue2021byt5,
    title={ByT5: Towards a token-free future with pre-trained byte-to-byte models},
    author={Linting Xue and Aditya Barua and Noah Constant and Rami Al-Rfou and Sharan Narang and Mihir Kale and Adam Roberts and Colin Raffel},
    year={2021},
    eprint={2105.13626},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

This is not an officially supported Google product.

Owner
Google Research
Google Research
GitHub Repository
A curated list of efficient attention modules

awesome-fast-attention A curated list of efficient attention modules

Sepehr Sameni 891 Dec 22, 2022
Implementing SimCSE(paper, official repository) using TensorFlow 2 and KR-BERT.

KR-BERT-SimCSE Implementing SimCSE(paper, official repository) using TensorFlow 2 and KR-BERT. Training Unsupervised python train_unsupervised.py --mi

Jeong Ukjae 27 Dec 12, 2022
TensorFlow code and pre-trained models for BERT

BERT ***** New March 11th, 2020: Smaller BERT Models ***** This is a release of 24 smaller BERT models (English only, uncased, trained with WordPiece

Google Research 32.9k Jan 08, 2023
A paper list of pre-trained language models (PLMs).

Large-scale pre-trained language models (PLMs) such as BERT and GPT have achieved great success and become a milestone in NLP.

RUCAIBox 124 Jan 02, 2023
My implementation of Safaricom Machine Learning Codility test. The code has bugs, logical I guess I made errors and any correction will be appreciated.

Safaricom_Codility Machine Learning 2022 The test entails two questions. Question 1 was on Machine Learning. Question 2 was on SQL I ran out of time.

Lawrence M. 1 Mar 03, 2022
Ongoing research training transformer language models at scale, including: BERT & GPT-2

What is this fork of Megatron-LM and Megatron-DeepSpeed This is a detached fork of https://github.com/microsoft/Megatron-DeepSpeed, which in itself is

BigScience Workshop 316 Jan 03, 2023
Wind Speed Prediction using LSTMs in PyTorch

Implementation of Deep-Forecast using PyTorch Deep Forecast: Deep Learning-based Spatio-Temporal Forecasting Adapted from original implementation Setu

Onur Kaplan 151 Dec 14, 2022
Large-scale pretraining for dialogue

A State-of-the-Art Large-scale Pretrained Response Generation Model (DialoGPT) This repository contains the source code and trained model for a large-

Microsoft 1.8k Jan 07, 2023
A method to generate speech across multiple speakers

VoiceLoop PyTorch implementation of the method described in the paper VoiceLoop: Voice Fitting and Synthesis via a Phonological Loop. VoiceLoop is a n

Facebook Archive 873 Dec 15, 2022
Library for Russian imprecise rhymes generation

TOM RHYMER Library for Russian imprecise rhymes generation. Quick Start Generate rhymes by any given rhyme scheme (aabb, abab, aaccbb, etc ...): from

Alexey Karnachev 6 Oct 18, 2022
Code for hyperboloid embeddings for knowledge graph entities

Implementation for the papers: Self-Supervised Hyperboloid Representations from Logical Queries over Knowledge Graphs, Nurendra Choudhary, Nikhil Rao,

30 Dec 10, 2022
Accurately generate all possible forms of an English word e.g "election" --> "elect", "electoral", "electorate" etc.

Accurately generate all possible forms of an English word Word forms can accurately generate all possible forms of an English word. It can conjugate v

Dibya Chakravorty 570 Dec 31, 2022
A program that uses real statistics to choose the best times to bet on BloxFlip's crash gamemode

Bloxflip Smart Bet A program that uses real statistics to choose the best times to bet on BloxFlip's crash gamemode. https://bloxflip.com/crash. THIS

43 Jan 05, 2023
PhoNLP: A BERT-based multi-task learning toolkit for part-of-speech tagging, named entity recognition and dependency parsing

PhoNLP is a multi-task learning model for joint part-of-speech (POS) tagging, named entity recognition (NER) and dependency parsing. Experiments on Vietnamese benchmark datasets show that PhoNLP prod

VinAI Research 109 Dec 02, 2022
DeepAmandine is an artificial intelligence that allows you to talk to it for hours, you won't know the difference.

DeepAmandine This is an artificial intelligence based on GPT-3 that you can chat with, it is very nice and makes a lot of jokes. We wish you a good ex

BuyWithCrypto 3 Apr 19, 2022
⛵️The official PyTorch implementation for "BERT-of-Theseus: Compressing BERT by Progressive Module Replacing" (EMNLP 2020).

BERT-of-Theseus Code for paper "BERT-of-Theseus: Compressing BERT by Progressive Module Replacing". BERT-of-Theseus is a new compressed BERT by progre

Kevin Canwen Xu 284 Nov 25, 2022
Materials (slides, code, assignments) for the NYU class I teach on NLP and ML Systems (Master of Engineering).

FREE_7773 Repo containing material for the NYU class (Master of Engineering) I teach on NLP, ML Sys etc. For context on what the class is trying to ac

Jacopo Tagliabue 90 Dec 19, 2022
:P Some basic stuff I'm gonna use for my upcoming Agile Software Development and Devops

reverse-image-search-py bash script.sh img_name.jpg Requirements pip install requests pip install pyshorteners Dry run [ Sudhanva M 3 Dec 18, 2021

BERN2: an advanced neural biomedical namedentity recognition and normalization tool

BERN2 We present BERN2 (Advanced Biomedical Entity Recognition and Normalization), a tool that improves the previous neural network-based NER tool by

DMIS Laboratory - Korea University 99 Jan 06, 2023
Korean Sentence Embedding Repository

Korean-Sentence-Embedding 🍭 Korean sentence embedding repository. You can download the pre-trained models and inference right away, also it provides

80 Jan 02, 2023