{"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/understanding-back-translation-at-scale","title":"Understanding Back-Translation at Scale","arxiv_id":"1808.09381","date":"2018-08-28","proceeding":"EMNLP 2018 10","authors":["Sergey Edunov","Myle Ott","Michael Auli","David Grangier"],"abstract":"An effective method to improve neural machine translation with monolingual\ndata is to augment the parallel training corpus with back-translations of\ntarget language sentences. This work broadens the understanding of\nback-translation and investigates a number of methods to generate synthetic\nsource sentences. We find that in all but resource poor settings\nback-translations obtained via sampling or noised beam outputs are most\neffective. Our analysis shows that sampling or noisy synthetic data gives a\nmuch stronger training signal than data generated by beam or greedy search. We\nalso compare how synthetic data compares to genuine bitext and study various\ndomain effects. Finally, we scale to hundreds of millions of monolingual\nsentences and achieve a new state of the art of 35 BLEU on the WMT'14\nEnglish-German test set.","url_abs":"http://arxiv.org/abs/1808.09381v2","url_pdf":"http://arxiv.org/pdf/1808.09381v2.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":"understanding-back-translation-at-scale","repo_url":"https://github.com/pytorch/fairseq","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"understanding-back-translation-at-scale","repo_url":"https://github.com/facebookresearch/fairseq","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"understanding-back-translation-at-scale","repo_url":"https://github.com/valentinmace/noisy-text","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/machine-translation-on-wmt2014-english-french","task":"Machine Translation","dataset":"WMT2014 English-French","model":"Noisy back-translation","rank_in_archive_order":2,"of":57,"metrics":{"BLEU score":"45.6","Hardware Burden":"180G","SacreBLEU":"43.8"},"uses_additional_data":true},{"leaderboard":"/sota/machine-translation-on-wmt2014-english-german","task":"Machine Translation","dataset":"WMT2014 English-German","model":"Noisy back-translation","rank_in_archive_order":2,"of":91,"metrics":{"BLEU score":"35.0","Hardware Burden":"146G","SacreBLEU":"33.8"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1808.09381","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1808.09381"}},"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/pytorch/fairseq","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/facebookresearch/fairseq","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/valentinmace/noisy-text","reach":null}],"summary":{"ran_draft_wrong":1},"by_repo_kind":{"listed":{"samples":1,"ran":1,"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":"e7cec6288ed860f1","entry":"random_token_permutation","repo":"valentinmace/noisy-text","repo_kind":"listed","path":"noise_functions.py","file_url":"https://github.com/valentinmace/noisy-text/blob/HEAD/noise_functions.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e7cec6288ed860f1"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}