search

Helsinki-NLP / Opus-MT

Open neural machine translation models and web services
MIT License
592 stars 71 forks source link

Is it possible to finetune a OPUS model that has once been finetuned? #50

Open Smilauer opened 2 years ago

Smilauer commented 2 years ago

Is it possible to finetune a model that has once been finetuned? Thus, incrementally improving the model gradually. If not, I will have several new finetuned models. If I have several already finetuned models is it possible to use them at the same time or do I have to compose 2 new .tmx into one .tmx file before the fine-tuning??

jorgtied commented 2 years ago

Is that a question about finetuning in OPUS-CAT or using the procedures in OPUS-MT-train?

Smilauer commented 2 years ago

Hello, my question is concerning the fine-tuning in the OPUS- CAT as described in the article OPUS-CAT: "Desktop NMT with CAT integration and local fine-tuning." Tommi Nieminen The University of Helsinki, Yliopistonkatu 3, 00014 University of Helsinki, Finland tommi.nieminen@helsinki.fi I have tried to use this excellent interactive software to make the translator English -> Interlingua (see www.interlingua.com) by fine-tuning English -> Catalan, see my paper in http://www.interlingva.cz/Le_experimentos_con_le_machina_OPUS_CAT.pdf which is written in Interlingua (= simplified Latin, similar to Catalan ). The translation showed surprisingly good results. I repeatedly wanted to fine-tune this model again, with additional phrase pairs EN->IA, but the process was started and never has finished. I tried it with more combinations CZ (my mother language), IA, EN, and I have never experienced a successful ending when the source model was fine-tuned before. I would appreciate any advice, on how to subsequently improve the translators with Interlingua. See also the other papers on the page http://www.interlingva.cz/Le_Experimentos_con_traduction_automatic.pdf . Many thanks

jorgtied commented 2 years ago

Could we move this issue to https://github.com/Helsinki-NLP/OPUS-CAT?