Projecte Aina’s Basque-Catalan machine translation model
Model description
This model was trained from scratch using the Fairseq toolkit on a combination of Basque-Catalan datasets
totalling approximately 75 million sentence pairs.Parallel Basque-Catalan data was collected from Opus and additional synthetic data was created from
the Projecte Aina ES-CA corpus by translating the Spanish side using the ES-EU translator of HiTZ. The model was evaluated on the Flores and NTREX evaluation datasets.
Intended uses and limitations
You can use this model for machine translation from Basque to Catalan.
At the time of submission, no measures have been taken to estimate the bias and toxicity embedded in the model.
However, we are well aware that our models may be biased. We intend to conduct research in these areas in the future, and if completed, this model card will be updated.
Training
Training data
The Basque-Catalan data is a combination of publicly available bilingual datasets collected from Opus.
Additional synthetic parallel data were created from the Projecte Aina ES-CA corpus.
Training procedure
Data preparation
All datasets are filtered for language alignment, deduplicated and filtered to remove any sentence pairs with a cosine similarity of less than 0.75.
This is done using sentence embeddings calculated using LaBSE.
The filtered datasets are then concatenated to form the final training corpus and 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 a 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 hyperparameters 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
48.000
Optimizer
adam
Adam betas
(0.9, 0.980)
Clip norm
0.0
Learning rate
5e-4
Lr. schedurer
inverse sqrt
Warmup updates
8000
Dropout
0.1
Label smoothing
0.1
The model was trained for 19.000 updates on the parallel data collected from the web.
This data was then concatenated with the synthetic parallel data and training continued for a total of 30.000 updates.
Weights were saved every 1000 updates and reported results are the average of the last 4 checkpoints.
Evaluation
Variable and metrics
We use the BLEU score for evaluation on test sets: Flores-200 and
NTREX.
This work is funded by the Ministerio para la Transformación Digital y de la Función Pública - Funded by EU – NextGenerationEU
within the framework of the project ILENIA
with reference 2022/TL22/00215337.
Disclaimer
Click to expand
The model published in this repository is intended for a generalist purpose and is available to third parties under a permissive Apache License, Version 2.0.
Be aware that the model may have biases and/or any other undesirable distortions.
When third parties deploy or provide systems and/or services to other parties using this model (or any system based on it)
or become users of the model, they should note that it is their responsibility to mitigate the risks arising from its 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 model (Barcelona Supercomputing Center)
be liable for any results arising from the use made by third parties.
