{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/multifit-efficient-multi-lingual-language","title":"MultiFiT: Efficient Multi-lingual Language Model Fine-tuning","arxiv_id":"1909.04761","date":"2019-09-10","proceeding":"IJCNLP 2019 11","authors":["Julian Martin Eisenschlos","Sebastian Ruder","Piotr Czapla","Marcin Kardas","Sylvain Gugger","Jeremy Howard"],"abstract":"Pretrained language models are promising particularly for low-resource languages as they only require unlabelled data. However, training existing models requires huge amounts of compute, while pretrained cross-lingual models often underperform on low-resource languages. We propose Multi-lingual language model Fine-Tuning (MultiFiT) to enable practitioners to train and fine-tune language models efficiently in their own language. In addition, we propose a zero-shot method using an existing pretrained cross-lingual model. We evaluate our methods on two widely used cross-lingual classification datasets where they outperform models pretrained on orders of magnitude more data and compute. We release all models and code.","url_abs":"https://arxiv.org/abs/1909.04761v2","url_pdf":"https://arxiv.org/pdf/1909.04761v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"multifit-efficient-multi-lingual-language","repo_url":"https://github.com/TheophileBlard/french-sentiment-analysis-with-bert","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multifit-efficient-multi-lingual-language","repo_url":"https://github.com/lukexyz/Language-Models","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"multifit-efficient-multi-lingual-language","repo_url":"https://github.com/n-waves/multifit","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multifit-efficient-multi-lingual-language","repo_url":"https://github.com/piegu/language-models","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"cross-lingual-document-classification","task_name":"Cross-Lingual Document Classification"},{"task_slug":"document-classification","task_name":"Document Classification"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":null,"task_name":"Zero-shot Cross-Lingual Document Classification"},{"task_slug":"model","task_name":"model"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/cross-lingual-document-classification-on-8","task":"Cross-Lingual Document Classification","dataset":"MLDoc Zero-Shot English-to-Chinese","model":"MultiFiT, pseudo","rank_in_archive_order":2,"of":5,"metrics":{"Accuracy":"82.48"},"uses_additional_data":false},{"leaderboard":"/sota/cross-lingual-document-classification-on-2","task":"Cross-Lingual Document Classification","dataset":"MLDoc Zero-Shot English-to-French","model":"MultiFiT, pseudo","rank_in_archive_order":2,"of":6,"metrics":{"Accuracy":"89.42"},"uses_additional_data":false},{"leaderboard":"/sota/cross-lingual-document-classification-on","task":"Cross-Lingual Document Classification","dataset":"MLDoc Zero-Shot English-to-German","model":"MultiFiT, pseudo","rank_in_archive_order":2,"of":5,"metrics":{"Accuracy":"91.62%"},"uses_additional_data":false},{"leaderboard":"/sota/cross-lingual-document-classification-on-10","task":"Cross-Lingual Document Classification","dataset":"MLDoc Zero-Shot English-to-Italian","model":"MultiFiT, pseudo","rank_in_archive_order":1,"of":4,"metrics":{"Accuracy":"76.02"},"uses_additional_data":false},{"leaderboard":"/sota/cross-lingual-document-classification-on-11","task":"Cross-Lingual Document Classification","dataset":"MLDoc Zero-Shot English-to-Japanese","model":"MultiFiT, pseudo","rank_in_archive_order":1,"of":3,"metrics":{"Accuracy":"69.57"},"uses_additional_data":false},{"leaderboard":"/sota/cross-lingual-document-classification-on-9","task":"Cross-Lingual Document Classification","dataset":"MLDoc Zero-Shot English-to-Russian","model":"MultiFiT, pseudo","rank_in_archive_order":2,"of":5,"metrics":{"Accuracy":"67.83"},"uses_additional_data":false},{"leaderboard":"/sota/cross-lingual-document-classification-on-1","task":"Cross-Lingual Document Classification","dataset":"MLDoc Zero-Shot English-to-Spanish","model":"MultiFiT, pseudo","rank_in_archive_order":2,"of":6,"metrics":{"Accuracy":"79.1"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1909.04761","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}