Datasets › XGLUE

XGLUE

Introduced by Yaobo Liang et al. in XGLUE: A New Benchmark Dataset for Cross-lingual Pre-training, Understanding and Generation3 Apr 2020 archive 2025-07-28

XGLUE is an evaluation benchmark XGLUE,which is composed of 11 tasks that span 19 languages. For each task, the training data is only available in English. This means that to succeed at XGLUE, a model must have a strong zero-shot cross-lingual transfer capability to learn from the English data of a specific task and transfer what it learned to other languages. Comparing to its concurrent work XTREME, XGLUE has two characteristics: First, it includes cross-lingual NLU and cross-lingual NLG tasks at the same time; Second, besides including 5 existing cross-lingual tasks (i.e. NER, POS, MLQA, PAWS-X and XNLI), XGLUE selects 6 new tasks from Bing scenarios as well, including News Classification (NC), Query-Ad Matching (QADSM), Web Page Ranking (WPR), QA Matching (QAM), Question Generation (QG) and News Title Generation (NTG). Such diversities of languages, tasks and task origin provide a comprehensive benchmark for quantifying the quality of a pre-trained model on cross-lingual natural language understanding and generation.

Homepage Benchmarks Papers JSON

Benchmarks archive 2025-07-28

All 2 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

		First row (archive order)	Paper	Code
Few-shot NER	XGLUE	mGPT Avg F1 0.85	mGPT: Few-Shot Learners Go Multilingual	ai-forever/mgpt	1	Compare
Part-Of-Speech Tagging	XGLUE	mGPT Avg. F1 0.56	mGPT: Few-Shot Learners Go Multilingual	ai-forever/mgpt	1	Compare

Papers archive 2025-07-28

1 shown of 1 paper with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 22. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

			Date	Samples run Syntology
mGPT: Few-Shot Learners Go Multilingual	1	2	15 Apr 2022	not harvested

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

Natural Language Inference Cross-Lingual NER Part-Of-Speech Tagging Natural Language Understanding Cross-Lingual Natural Language Inference Few-shot NER Cross-Lingual POS Tagging

License archive 2025-07-28

Custom (non-commercial)

Modalities archive 2025-07-28

Texts

Languages archive 2025-07-28

English French Spanish German Italian Chinese Japanese Russian Portuguese Arabic Bulgarian Danish Dutch Finnish Hebrew Hindi Hungarian Indonesian Korean Polish Swedish Thai Turkish Urdu Vietnamese Greek Swahili

Variants archive 2025-07-28

XGLUE

1 variant name, as the archive lists them.

At a glance archive 2025-07-28

Introduced: 3 Apr 2020
Papers: 22
Benchmarks: 2
With a result: 1
Licence: Custom (non-commercial)
Modalities: Texts
Languages: English, French, Spanish, German, Italian, Chinese, Japanese, Russian, Portuguese, Arabic, Bulgarian, Danish, Dutch, Finnish, Hebrew, Hindi, Hungarian, Indonesian, Korean, Polish, Swedish, Thai, Turkish, Urdu, Vietnamese, Greek, Swahili
Homepage: microsoft.github.io/XGLUE/

Samples run Syntology

Of the 1 paper with a result here, 0 have harvested code samples in Syntology's graph. Summed over those: Syntology ran 0 of 0 samples; 0 unverified. 0 of the 0 have no sample that ran. Each row in the Papers table states its own "ran N of M samples" with the unverified count; a sample that ran executed on a synthesized fixture, which is not a correctness claim about any result. Read from the graph 2026-09-24.

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