{"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/multilingual-translation-with-extensible","title":"Multilingual Translation with Extensible Multilingual Pretraining and Finetuning","arxiv_id":"2008.00401","date":"2020-08-02","proceeding":null,"authors":["Yuqing Tang","Chau Tran","Xi-An Li","Peng-Jen Chen","Naman Goyal","Vishrav Chaudhary","Jiatao Gu","Angela Fan"],"abstract":"Recent work demonstrates the potential of multilingual pretraining of creating one model that can be used for various tasks in different languages. Previous work in multilingual pretraining has demonstrated that machine translation systems can be created by finetuning on bitext. In this work, we show that multilingual translation models can be created through multilingual finetuning. Instead of finetuning on one direction, a pretrained model is finetuned on many directions at the same time. Compared to multilingual models trained from scratch, starting from pretrained models incorporates the benefits of large quantities of unlabeled monolingual data, which is particularly important for low resource languages where bitext is not available. We demonstrate that pretrained models can be extended to incorporate additional languages without loss of performance. We double the number of languages in mBART to support multilingual machine translation models of 50 languages. Finally, we create the ML50 benchmark, covering low, mid, and high resource languages, to facilitate reproducible research by standardizing training and evaluation data. On ML50, we demonstrate that multilingual finetuning improves on average 1 BLEU over the strongest baselines (being either multilingual from scratch or bilingual finetuning) while improving 9.3 BLEU on average over bilingual baselines from scratch.","url_abs":"https://arxiv.org/abs/2008.00401v1","url_pdf":"https://arxiv.org/pdf/2008.00401v1.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":"multilingual-translation-with-extensible","repo_url":"https://github.com/UKPLab/EasyNMT","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"multilingual-translation-with-extensible","repo_url":"https://github.com/huggingface/transformers","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"multilingual-translation-with-extensible","repo_url":"https://github.com/kenji-imamura/sentpiece_mimic","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"multilingual-translation-with-extensible","repo_url":"https://github.com/russiannlp/rucola","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"multilingual-translation-with-extensible","repo_url":"https://github.com/xhluca/dl-translate","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multilingual-translation-with-extensible","repo_url":"https://github.com/xhlulu/dl-translate","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"translation","task_name":"Translation"}],"methods":[{"method_slug":"mbart","method_name":"mBART"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2008.00401","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2008.00401"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/huggingface/transformers","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/russiannlp/rucola","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kenji-imamura/sentpiece_mimic","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/UKPLab/EasyNMT","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/xhluca/dl-translate","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/xhlulu/dl-translate","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":2},"by_repo_kind":{"listed":{"samples":2,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"6f6b1d2434485fd7","entry":"logsumexp","repo":"kenji-imamura/sentpiece_mimic","repo_kind":"listed","path":"src/SentPiece.py","file_url":"https://github.com/kenji-imamura/sentpiece_mimic/blob/HEAD/src/SentPiece.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6f6b1d2434485fd7"}},{"code_sha256_prefix":"18c68ce4f301ae56","entry":"read_corpus","repo":"kenji-imamura/sentpiece_mimic","repo_kind":"listed","path":"src/SentPiece.py","file_url":"https://github.com/kenji-imamura/sentpiece_mimic/blob/HEAD/src/SentPiece.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"18c68ce4f301ae56"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}