Edit model card

Aina Project's Catalan-Spanish machine translation model for the administrative domain.

Table of Contents

Model description

This model is a fine-tuned version of projecte-aina/mt-aina-ca-es for the administrative domain. Additionally, the model is evaluated on several public datasets comprising 5 different domains (general, adminstrative, technology, biomedical, and news).

Intended uses and limitations

You can use this model for machine translation of administrative texts from Catalan to Spanish.

How to use

Usage

Required libraries:

pip install ctranslate2 pyonmttok

Translate a sentence using python

import ctranslate2
import pyonmttok
from huggingface_hub import snapshot_download
model_dir = snapshot_download(repo_id="projecte-aina/mt-aina-ca-es-adm", revision="main")

tokenizer=pyonmttok.Tokenizer(mode="none", sp_model_path = model_dir + "/spm.model")
tokenized=tokenizer.tokenize("Benvingut al projecte Aina!")

translator = ctranslate2.Translator(model_dir)
translated = translator.translate_batch([tokenized[0]])
print(tokenizer.detokenize(translated[0][0]['tokens']))

Limitations and Bias

At the time of submission, no measures have been taken to estimate the bias and toxicity embedded in the model. However, we are aware that our models may be biased since the corpora have been collected using crawling techniques on multiple web sources. We intend to conduct research in these areas in the future, and if completed, this model card will be updated.

Training

Training data

The original model was trained on a combination of the following datasets:

Dataset Sentences
DOCG v2 8.472.786
El Periodico 6.483.106
EuroParl 1.876.669
WikiMatrix 1.421.077
Wikimedia 335.955
QED 71.867
TED2020 v1 52.177
CCMatrix v1 56.103.820
MultiCCAligned v1 2.433.418
ParaCrawl 15.327.808
Total 92.578.683

This fine-tuned model is further trained using:

Dataset Sentences
AINA AAPP 62.773

Training procedure

Data preparation

All datasets are concatenated and filtered using the mBERT Gencata parallel filter and cleaned using the clean-corpus-n.pl script from moses, allowing sentences between 5 and 150 words.

Before training, the punctuation is normalized using a modified version of the join-single-file.py script from SoftCatalà

Tokenization

All data is tokenized using sentencepiece, with 50 thousand token sentencepiece model learned from the combination of all filtered training data. This model is included.

Hyperparameters

The model is based on the Transformer-XLarge proposed by Subramanian et al. The following hyperparamenters were set on the Fairseq toolkit:

Hyperparameter Value
Architecture transformer_vaswani_wmt_en_de_big
Embedding size 1024
Feedforward size 4096
Number of heads 16
Encoder layers 24
Decoder layers 6
Normalize before attention True
--share-decoder-input-output-embed True
--share-all-embeddings True
Effective batch size 96.000
Optimizer adam
Adam betas (0.9, 0.980)
Clip norm 0.0
Learning rate 1e-3
Lr. schedurer inverse sqrt
Warmup updates 4000
Dropout 0.1
Label smoothing 0.1

The original model was trained using shards of 10 million sentences, for a total of 13.000 updates. Weights were saved every 1000 updates and reported results are the average of the last 6 checkpoints.

For the fine-tuning, the model continued training for 9000 updates with reduced maximum learning rate of 5e-5.

Evaluation

Variable and metrics

We use the BLEU score for evaluation on test sets: Flores-101, TaCon, United Nations, Cybersecurity, wmt19 biomedical test set, wmt13 news test set, aina aapp

Evaluation results

Below are the evaluation results on the machine translation from Catalan to Spanish compared to Softcatalà and Google Translate:

Test set SoftCatalà Google Translate mt-aina-ca-es mt-aina-ca-es-adm
Spanish Constitution 70,7 77,1 75,5 67,2
United Nations 78,1 84,3 86,3 78,0
Flores 101 dev 23,5 24 24,1 22,4
Flores 101 devtest 24,1 24,2 24,4 22,9
Cybersecurity 67,3 76,9 75,1 65,7
wmt 19 biomedical 60,4 62,7 63,0 59,2
wmt 13 news 22,5 23,1 23,4 20,4
aina_aapp 80,9 81,4 82,8 85,5
Average 53,4 56,7 56,7 52,6

Additional information

Author

Text Mining Unit (TeMU) at the Barcelona Supercomputing Center (bsc-temu@bsc.es)

Contact information

For further information, send an email to aina@bsc.es

Copyright

Copyright (c) 2022 Text Mining Unit at Barcelona Supercomputing Center

Licensing Information

Apache License, Version 2.0

Funding

This work was funded by the Departament de la Vicepresidència i de Polítiques Digitals i Territori de la Generalitat de Catalunya within the framework of Projecte AINA.

Disclaimer

Click to expand

The models published in this repository are intended for a generalist purpose and are available to third parties. These models may have bias and/or any other undesirable distortions.

When third parties, deploy or provide systems and/or services to other parties using any of these models (or using systems based on these models) or become users of the models, they should note that it is their responsibility to mitigate the risks arising from their use and, in any event, to comply with applicable regulations, including regulations regarding the use of Artificial Intelligence.

In no event shall the owner and creator of the models (BSC – Barcelona Supercomputing Center) be liable for any results arising from the use made by third parties of these models.

Downloads last month
0
Unable to determine this model's library. Check the docs .

Collection including projecte-aina/mt-aina-ca-es-adm