Investigating techniques for a deeper understanding of Neural Machine Translation (NMT) systems through data filtering and fine-tuning strategies

Lichao Zhu (1) , Maria Zimina-Poirot (1) , Maud Bénard (1) , Behnoosh Namdarzadeh (1) , Nicolas Ballier (1, 2) , Guillaume Wisniewski (2) , Jean-Baptiste Yunès

1 CLILLAC-ARP (URP_3967) - Centre de Linguistique Inter-langues, de Lexicologie, de Linguistique Anglaise et de Corpus
2 LLF - UMR7110 - Laboratoire de Linguistique Formelle

In the context of this biomedical shared task, we have implemented data filters to enhance the selection of relevant training data for finetuning from the available training data sources. Specifically, we have employed textometric analysis to detect repetitive segments within the test set, which we have then used for refining the training data used to fine-tune the mBart-50 baseline model. Through this approach, we aim to achieve several objectives: developing a practical fine-tuning strategy for training biomedical in-domain fr<>en models, defining criteria for filtering in-domain training data, and comparing model predictions, finetuning data in accordance with the test set to gain a deeper insight into the functioning of Neural Machine Translation (NMT) systems.

Keywords: NMT Neural Machine Translation; Corpora filtering; In-domain data; Fine-tuning; French - English specialised translation

Domains: Computation and Language [cs.CL]; Document and Text Processing

Lichao Zhu, Maria Zimina-Poirot, Maud Bénard, Behnoosh Namdarzadeh, Nicolas Ballier, et al.. Investigating techniques for a deeper understanding of Neural Machine Translation (NMT) systems through data filtering and fine-tuning strategies. 8th World Machine Translation, Association for Computational Linguistics, Dec 2023, Singapore (SG), Singapore. pp.275-281. ⟨hal-04306041⟩