{"url":"/dataset/xtreme","name":"XTREME","full_name":"Cross-Lingual Transfer Evaluation of Multilingual Encoders","description_markdown":"The **Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME)** benchmark was introduced to encourage more research on multilingual transfer learning,. XTREME covers 40 typologically diverse languages spanning 12 language families and includes 9 tasks that require reasoning about different levels of syntax or semantics.\r\n\r\nThe languages in XTREME are selected to maximize language diversity, coverage in existing tasks, and availability of training data. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks, and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil (spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the Niger-Congo languages Swahili and Yoruba, spoken in Africa.","description_withheld":null,"homepage":"https://sites.research.google/xtreme","introduced_date":"2020-01-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/xtreme-a-massively-multilingual-multi-task-1","title":"XTREME: A Massively Multilingual Multi-task Benchmark for Evaluating Cross-lingual Generalisation","first_author":"Junjie Hu","url":null},"license":null,"modalities":[{"name":"Texts","url":"/datasets/modality/texts"}],"tasks":[{"name":"Token Classification","url":"/task/token-classification","datasets_with_task":"/datasets/task/token-classification"},{"name":"Zero-Shot Cross-Lingual Transfer","url":"/task/zero-shot-cross-lingual-transfer","datasets_with_task":"/datasets/task/zero-shot-cross-lingual-transfer"}],"languages":[{"name":"Telugu","url":"/datasets/language/telugu"},{"name":"Swahili","url":"/datasets/language/swahili"}],"variants":["XTREME"],"data_loaders":[{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/google/xtreme","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/xtreme","frameworks":["tf","pytorch","jax"]}],"num_papers_in_archive":55,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/zero-shot-cross-lingual-transfer-on-xtreme","task":"Zero-Shot Cross-Lingual Transfer","dataset_variant":"XTREME","rows":25,"metrics":["Avg","AVG","Sentence-pair Classification","Structured Prediction","Question Answering","Sentence Retrieval"],"first_row_in_archive_order":{"model":"Turing ULR v6","paper":"/paper/xlm-e-cross-lingual-language-model-pre","metrics":{"Avg":"85.5","Question Answering":"77.1","Sentence Retrieval":"94.4","Sentence-pair Classification":"91.0","Structured Prediction":"83.8"},"code_links":[{"title":"microsoft/unilm","url":"https://github.com/microsoft/unilm"},{"title":"CZWin32768/xnlg","url":"https://github.com/CZWin32768/xnlg"},{"title":"Mind23-2/MindCode-151","url":"https://github.com/Mind23-2/MindCode-151"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/token-classification-on-xtreme","task":"Token Classification","dataset_variant":"XTREME","rows":0,"metrics":["F1"],"first_row_in_archive_order":null,"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/xlm-e-cross-lingual-language-model-pre","title":"XLM-E: Cross-lingual Language Model Pre-training via ELECTRA","date":"2021-06-30","rows_on_this_dataset":2,"code_links":3,"syntology":null},{"paper":"/paper/ernie-m-enhanced-multilingual-representation","title":"ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora","date":"2020-12-31","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":0,"samples_unverified":1,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/mt5-a-massively-multilingual-pre-trained-text","title":"mT5: A massively multilingual pre-trained text-to-text transformer","date":"2020-10-22","rows_on_this_dataset":1,"code_links":8,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":13,"samples_ran":0,"samples_unverified":13,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/filter-an-enhanced-fusion-method-for-cross","title":"FILTER: An Enhanced Fusion Method for Cross-lingual Language Understanding","date":"2020-09-10","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":2,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/infoxlm-an-information-theoretic-framework","title":"InfoXLM: An Information-Theoretic Framework for Cross-Lingual Language Model Pre-Training","date":"2020-07-15","rows_on_this_dataset":1,"code_links":4,"syntology":null},{"paper":"/paper/english-intermediate-task-training-improves","title":"English Intermediate-Task Training Improves Zero-Shot Cross-Lingual Transfer Too","date":"2020-05-26","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/xtreme-a-massively-multilingual-multi-task-1","title":"XTREME: A Massively Multilingual Multi-task Benchmark for Evaluating Cross-lingual Generalisation","date":"2020-01-01","rows_on_this_dataset":1,"code_links":2,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":3,"samples_harvested":16,"samples_ran":2,"samples_unverified":14,"pointer_only_for_licence":1,"papers_with_no_sample_that_ran":2,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}