Papers › Cross-lingual Language Model Pretraining

Cross-lingual Language Model Pretraining

22 Jan 2019NeurIPS 2019 12arXiv:1901.07291archive 2025-07-28

Guillaume Lample, Alexis Conneau

Recent studies have demonstrated the efficiency of generative pretraining for English natural language understanding. In this work, we extend this approach to multiple languages and show the effectiveness of cross-lingual pretraining. We propose two methods to learn cross-lingual language models (XLMs): one unsupervised that only relies on monolingual data, and one supervised that leverages parallel data with a new cross-lingual language model objective. We obtain state-of-the-art results on cross-lingual classification, unsupervised and supervised machine translation. On XNLI, our approach pushes the state of the art by an absolute gain of 4.9% accuracy. On unsupervised machine translation, we obtain 34.3 BLEU on WMT'16 German-English, improving the previous state of the art by more than 9 BLEU. On supervised machine translation, we obtain a new state of the art of 38.5 BLEU on WMT'16 Romanian-English, outperforming the previous best approach by more than 4 BLEU. Our code and pretrained models will be made publicly available.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 7 code samples harvested from 3 repositories linked to this paper: 5 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 1 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code; here there are none.

By repository: community (archive-listed): 7 samples from 3 repositories, 5 ran (5 with no instrument failure; 0 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 5 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

17 repositories listed; official and paper-mentioned ones first.

1-punchMan/CLTS mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Unsupervised Cross-lingual Representation Learning at Scale (arXiv:1911.02116; that paper's own run record: 40 ran (of which 13 constructed an object rather than computing a result; 35 with no instrument failure: 4 honoured, 1 violated, 30 with no contract checked; 5 where Syntology's instrument failed) · 19 unverified); nothing checks that they implement this paper's method. report
JunnYu/xlm_paddle mentioned on GitHubpaddleApache-2.0Syntology: 3 samples, 2 ran (0 instrument-failed) · 1 unverified report
Somefive/XNLI mentioned on GitHubpytorchSyntology: not harvested report
Tikquuss/meta_XLM mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
deterministic-algorithms-lab/Large-XLM mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Unsupervised Cross-lingual Representation Learning at Scale (arXiv:1911.02116; that paper's own run record: 40 ran (of which 13 constructed an object rather than computing a result; 35 with no instrument failure: 4 honoured, 1 violated, 30 with no contract checked; 5 where Syntology's instrument failed) · 19 unverified); nothing checks that they implement this paper's method. report
facebookresearch/MLQA mentioned on GitHubSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to MLQA: Evaluating Cross-lingual Extractive Question Answering (arXiv:1910.07475; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
facebookresearch/UnsupervisedMT mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
facebookresearch/XLM mentioned on GitHubpytorchSyntology: not harvested report
feyzaakyurek/XLM-LwLL mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
fshdnc/enfi-XLM mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
huggingface/transformers mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Attention Is All You Need (arXiv:1706.03762; that paper's own run record: 610 ran (of which 293 constructed an object rather than computing a result; 529 with no instrument failure: 45 honoured, 15 violated, 469 with no contract checked; 81 where Syntology's instrument failed) · 336 unverified), HuggingFace's Transformers: State-of-the-art Natural Language Processing (arXiv:1910.03771; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified), DeBERTa: Decoding-enhanced BERT with Disentangled Attention (arXiv:2006.03654; that paper's own run record: 4 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 2 honoured, 1 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 9 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
kheeong/XLM_OWN mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
samwisegamjeee/pytorch-transformers mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to XLNet: Generalized Autoregressive Pretraining for Language Understanding (arXiv:1906.08237; that paper's own run record: 15 ran (of which 2 constructed an object rather than computing a result; 10 with no instrument failure: 2 honoured, 0 violated, 8 with no contract checked; 5 where Syntology's instrument failed) · 9 unverified); nothing checks that they implement this paper's method. report
2023-MindSpore-1/ms-code-169 mindsporeSyntology: not harvested report
MS-P3/code7 mindsporeSyntology: not harvested report
MS-P3/code7 mindsporeSyntology: not harvested report
MiuGod0126/ConvS2S_Paddle paddleApache-2.0Syntology: 3 samples, 2 ran (0 instrument-failed) · 1 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

7 samples harvested: 5 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 1 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · violated contract
4ran
2unverified

Licence: 1 of the 7 samples is pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 3 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: 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. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

JunnYu/xlm_paddle community (archive-listed) 3 samples, 2 ran (0 instrument-failed) · 1 unverified
get_pairs in 260 papers xlm_paddle/tokenizer.py ran Apache-2.0 (permissive) · 9372a18a70742351 · report fixture: signature word:tuple/1/str · returned set · level recorded 2026-09-24
lowercase_and_remove_accent xlm_paddle/tokenizer.py ran Apache-2.0 (permissive) · 9b4a5d4b20c0f99a · report fixture: signature text:list/1/str · returned list · level recorded 2026-09-24
replace_unicode_punct xlm_paddle/tokenizer.py unverified Apache-2.0 (permissive) · 22db157c2db3501f · report
Tikquuss/meta_XLM community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
couple XLM/meta_train.py ran · violated contract licence not identified · pointer only · a205cce260e31d6f · report fixture: signature array:list/1/str, sep:str/0/str · returned list · level recorded 2026-09-22
MiuGod0126/ConvS2S_Paddle community (archive-listed) 3 samples, 2 ran (0 instrument-failed) · 1 unverified
check_different align.py ran fingerprinted Apache-2.0 (permissive) · 67fe142c7fcc1801 · report fixture: signature ls1:list/1/str, ls2:list/1/str · returned tuple · level recorded 2026-09-24
extend_conv_spec models/convs2s.py ran Apache-2.0 (permissive) · 92126365537a14c7 · report fixture: signature convolutions:list/1/tuple · returned tuple · level recorded 2026-09-24
load_embedding in 15 papers models/utils.py unverified Apache-2.0 (permissive) · 112594e8e25c52c6 · report

Tasks

Language ModelingLanguage ModellingMachine TranslationNatural Language UnderstandingTranslationUnsupervised Machine Translationmodel

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Machine Translation WMT2016 Romanian-English MLM pretraining BLEU score 35.3 #4 of 21 Archive leaderboard report
Natural Language Inference XNLI French XLM (MLM+TLM) Accuracy 80.2 #5 of 6 Archive leaderboard report
Unsupervised Machine Translation WMT2014 English-French MLM pretraining for encoder and decoder BLEU 33.4 #4 of 7 Archive leaderboard report
Unsupervised Machine Translation WMT2014 French-English MLM pretraining for encoder and decoder BLEU 33.3 #4 of 7 Archive leaderboard report
Unsupervised Machine Translation WMT2016 English--Romanian MLM pretraining for encoder and decoder BLEU 33.3 #2 of 2 Archive leaderboard report
Unsupervised Machine Translation WMT2016 English-German MLM pretraining for encoder and decoder BLEU 26.4 #4 of 7 Archive leaderboard report
Unsupervised Machine Translation WMT2016 English-Romanian MLM pretraining for encoder and decoder BLEU 33.3 #2 of 3 Archive leaderboard report
Unsupervised Machine Translation WMT2016 German-English MLM pretraining for encoder and decoder BLEU 34.3 #4 of 7 Archive leaderboard report
Unsupervised Machine Translation WMT2016 Romanian-English MLM pretraining for encoder and decoder BLEU 31.8 #3 of 3 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Introduced by this paper: XLM

AdamAttentionAttention DropoutBPEDense ConnectionsDropoutLayer NormalizationLinear LayerMulti-Head AttentionResidual ConnectionSoftmaxXLM

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections