Applying "Load What You Need: Smaller Versions of Multilingual BERT" to LaBSE

Last update: Sep 02, 2022

Overview

smaller-LaBSE

LaBSE(Language-agnostic BERT Sentence Embedding) is a very good method to get sentence embeddings across languages. But it is hard to fine-tune due to the parameter size(~=471M) of this model. For instance, if I fine-tune this model with Adam optimizer, I need the GPU that has VRAM at least 7.5GB = 471M * (parameters 4 bytes + gradients 4 bytes + momentums 4 bytes + variances 4 bytes). So I applied "Load What You Need: Smaller Multilingual Transformers" method to LaBSE to reduce parameter size since most of this model's parameter is the word embedding table(~=385M).

The smaller version of LaBSE is evaluated for 14 languages using tatoeba dataset. It shows we can reduce LaBSE's parameters to 47% without a big performance drop.

If you need the PyTorch version, see https://github.com/Geotrend-research/smaller-transformers. I followed most of the steps in the paper.

Model	#param(transformer)	#param(word embedding)	#param(model)	vocab size
tfhub_LaBSE	85.1M	384.9M	470.9M	501,153
15lang_LaBSE	85.1M	133.1M	219.2M	173,347

Used Languages

English (en or eng)
French (fr or fra)
Spanish (es or spa)
German (de or deu)
Chinese (zh, zh_classical or cmn)
Arabic (ar or ara)
Italian (it or ita)
Japanese (ja or jpn)
Korean (ko or kor)
Dutch (nl or nld)
Polish (pl or pol)
Portuguese (pt or por)
Thai (th or tha)
Turkish (tr or tur)
Russian (ru or rus)

I selected the languages multilingual-USE supports.

Scripts

A smaller version of the vocab was constructed based on the frequency of tokens using Wikipedia dump data. I followed most of the algorithms in the paper to extract proper vocab for each language and rewrite it for TensorFlow.

Convert weight

mkdir -p downloads/labse-2
curl -L https://tfhub.dev/google/LaBSE/2?tf-hub-format=compressed -o downloads/labse-2.tar.gz
tar -xf downloads/labse-2.tar.gz -C downloads/labse-2/
python save_as_weight_from_saved_model.py

Select vocabs

./download_dataset.sh
python select_vocab.py

Make smaller LaBSE

./make_smaller_labse.py

Evaluate tatoeba

./download_tatoeba_dataset.sh
# evaluate TFHub LaBSE
./evaluate_tatoeba.sh
# evaluate the smaller LaBSE
./evaluate_tatoeba.sh \
    --model models/LaBSE_en-fr-es-de-zh-ar-zh_classical-it-ja-ko-nl-pl-pt-th-tr-ru/1/ \
    --preprocess models/LaBSE_en-fr-es-de-zh-ar-zh_classical-it-ja-ko-nl-pl-pt-th-tr-ru_preprocess/1/

Results

Tatoeba

Model	fr	es	de	zh	ar	it	ja	ko	nl	pl	pt	th	tr	ru	avg
tfHub_LaBSE(en→xx)	95.90	98.10	99.30	96.10	90.70	95.30	96.40	94.10	97.50	97.90	95.70	82.85	98.30	95.30	95.25
tfHub_LaBSE(xx→en)	96.00	98.80	99.40	96.30	91.20	94.00	96.50	92.90	97.00	97.80	95.40	83.58	98.50	95.30	95.19
15lang_LaBSE(en→xx)	95.20	98.00	99.20	96.10	90.50	95.20	96.30	93.50	97.50	97.90	95.80	82.85	98.30	95.40	95.13
15lang_LaBSE(xx→en)	95.40	98.70	99.40	96.30	91.10	94.00	96.30	92.70	96.70	97.80	95.40	83.58	98.50	95.20	95.08

Accuracy(%) of the Tatoeba datasets.
If the strategy to select vocabs is changed or the corpus used in the selection step is changed to the corpus similar to the evaluation dataset, it is expected to reduce the performance drop.

References

Load What You Need: Smaller Versions of Multilingual BERT (Paper: https://arxiv.org/abs/2010.05609, GitHub: https://github.com/Geotrend-research/smaller-transformers)
Language-agnostic BERT Sentence Embedding: https://arxiv.org/abs/2007.01852
TFHub - LaBSE: https://tfhub.dev/google/LaBSE/2
LaBSE blog post: https://ai.googleblog.com/2020/08/language-agnostic-bert-sentence.html
Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond: https://arxiv.org/abs/1812.10464

Comments

Training time and Machine configuration

Hi, thanks for your sharing model. I want to make a smaller model, just contains two languages(en, zh). And I want to know the kind of machine GPU and how long does it need to cost?

opened by QzzIsCoding 2
Publish model to HuggingFace Model Hub?

I migrated the full LaBSE model from TF to PyTorch and uploaded them to the HuggingFace model hub. I saw this model on the TF hub and started migrating it for uploading to the HF Hub. I realized then that this wasn't published by Google but by @jeongukjae, so wanted to check with you before uploading it.

I have exported the model locally. I'm happy to check the changes in and upload the exported model if that's fine for you :).

opened by setu4993 2

Applying "Load What You Need: Smaller Versions of Multilingual BERT" to LaBSE

Related tags

Overview

smaller-LaBSE

Used Languages

Scripts

Convert weight

Select vocabs

Make smaller LaBSE

Evaluate tatoeba

Results

Tatoeba

References

You might also like...

Comments

Training time and Machine configuration

Publish model to HuggingFace Model Hub?

Releases(15lang-1)

15lang-1(Sep 19, 2021)

Owner

Jeong Ukjae

Pytorch version of BERT-whitening

Official PyTorch implementation of SegFormer

Data and code to support "Applied Natural Language Processing" (INFO 256, Fall 2021, UC Berkeley)

VMD Audio/Text control with natural language

NLPIR tutorial: pretrain for IR. pre-train on raw textual corpus, fine-tune on MS MARCO Document Ranking

Crie tokens de autenticação íntegros e seguros com UToken.

AI-Broad-casting - AI Broad casting with python

Code for EMNLP 2021 main conference paper "Text AutoAugment: Learning Compositional Augmentation Policy for Text Classification"

2021 AI CUP Competition on Traditional Chinese Scene Text Recognition - Intermediate Contest

A python framework to transform natural language questions to queries in a database query language.

This script just scrapes the most recent Nepali news from Kathmandu Post and notifies the user about current events at regular intervals.It sends out the most recent news at random!

[AAAI 21] Curriculum Labeling: Revisiting Pseudo-Labeling for Semi-Supervised Learning

PyTorch implementation of Microsoft's text-to-speech system FastSpeech 2: Fast and High-Quality End-to-End Text to Speech.

A unified tokenization tool for Images, Chinese and English.

Official code repository of the paper Linear Transformers Are Secretly Fast Weight Programmers.

Scikit-learn style model finetuning for NLP

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

A sentence aligner for comparable corpora

The implementation of Parameter Differentiation based Multilingual Neural Machine Translation

Kestrel Threat Hunting Language