atutej/m_lama

Name: atutej/m_lama
Creator: atutej
Published: 2024-03-14 19:36:35
License: 暂无描述

Hugging Face2024-03-14 更新2024-03-04 收录

下载链接：

https://hf-mirror.com/datasets/atutej/m_lama

下载链接

链接失效反馈

官方服务：

资源简介：

--- dataset_info: - config_name: af features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1364986 num_examples: 7331 download_size: 544481 dataset_size: 1364986 - config_name: ar features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4564504 num_examples: 19354 download_size: 1580143 dataset_size: 4564504 - config_name: az features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1467465 num_examples: 7653 download_size: 578396 dataset_size: 1467465 - config_name: be features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2285464 num_examples: 8853 download_size: 714406 dataset_size: 2285464 - config_name: bg features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3109085 num_examples: 12461 download_size: 1013009 dataset_size: 3109085 - config_name: bn features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2969863 num_examples: 8975 download_size: 748274 dataset_size: 2969863 - config_name: ca features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4620850 num_examples: 24287 download_size: 1940588 dataset_size: 4620850 - config_name: ceb features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1433194 num_examples: 6769 download_size: 524854 dataset_size: 1433194 - config_name: cs features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2997353 num_examples: 15848 download_size: 1246743 dataset_size: 2997353 - config_name: cy features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1901684 num_examples: 9915 download_size: 769225 dataset_size: 1901684 - config_name: da features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3672623 num_examples: 19636 download_size: 1535250 dataset_size: 3672623 - config_name: de features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 6348506 num_examples: 32548 download_size: 2613173 dataset_size: 6348506 - config_name: el features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3416098 num_examples: 12854 download_size: 1074167 dataset_size: 3416098 - config_name: en features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 7031572 num_examples: 37498 download_size: 3023574 dataset_size: 7031572 - config_name: es features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 6000790 num_examples: 31578 download_size: 2542929 dataset_size: 6000790 - config_name: et features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1847160 num_examples: 9880 download_size: 748222 dataset_size: 1847160 - config_name: eu features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2260887 num_examples: 11910 download_size: 921424 dataset_size: 2260887 - config_name: fa features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4482869 num_examples: 18481 download_size: 1497801 dataset_size: 4482869 - config_name: fi features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3575879 num_examples: 19017 download_size: 1477166 dataset_size: 3575879 - config_name: fr features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 6553643 num_examples: 33872 download_size: 2716208 dataset_size: 6553643 - config_name: ga features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2809813 num_examples: 13937 download_size: 1076939 dataset_size: 2809813 - config_name: gl features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2062413 num_examples: 10567 download_size: 817987 dataset_size: 2062413 - config_name: he features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3273282 num_examples: 14769 download_size: 1165490 dataset_size: 3273282 - config_name: hi features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2750247 num_examples: 8570 download_size: 707213 dataset_size: 2750247 - config_name: hr features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1766612 num_examples: 9322 download_size: 714362 dataset_size: 1766612 - config_name: hu features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3629786 num_examples: 18850 download_size: 1485748 dataset_size: 3629786 - config_name: hy features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2580835 num_examples: 10030 download_size: 809063 dataset_size: 2580835 - config_name: id features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2693872 num_examples: 14183 download_size: 1103155 dataset_size: 2693872 - config_name: it features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 5287655 num_examples: 27648 download_size: 2198936 dataset_size: 5287655 - config_name: ja features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 6105411 num_examples: 25356 download_size: 2091964 dataset_size: 6105411 - config_name: ka features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2649721 num_examples: 8099 download_size: 647390 dataset_size: 2649721 - config_name: ko features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3526211 num_examples: 16327 download_size: 1309593 dataset_size: 3526211 - config_name: la features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1581833 num_examples: 8061 download_size: 612760 dataset_size: 1581833 - config_name: lt features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1835683 num_examples: 9560 download_size: 736354 dataset_size: 1835683 - config_name: lv features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1649860 num_examples: 8474 download_size: 643807 dataset_size: 1649860 - config_name: ms features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1768627 num_examples: 9146 download_size: 702211 dataset_size: 1768627 - config_name: nl features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 6221612 num_examples: 32423 download_size: 2597145 dataset_size: 6221612 - config_name: pl features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4013247 num_examples: 20727 download_size: 1644648 dataset_size: 4013247 - config_name: pt features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4044269 num_examples: 21023 download_size: 1653658 dataset_size: 4044269 - config_name: ro features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2523121 num_examples: 12886 download_size: 1007651 dataset_size: 2523121 - config_name: ru features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 6405438 num_examples: 25335 download_size: 2129105 dataset_size: 6405438 - config_name: sk features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1942547 num_examples: 10205 download_size: 788723 dataset_size: 1942547 - config_name: sl features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3455705 num_examples: 18091 download_size: 1406987 dataset_size: 3455705 - config_name: sq features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2404246 num_examples: 12586 download_size: 956395 dataset_size: 2404246 - config_name: sr features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3104514 num_examples: 12477 download_size: 1027773 dataset_size: 3104514 - config_name: sv features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4536924 num_examples: 24240 download_size: 1905031 dataset_size: 4536924 - config_name: ta features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2546658 num_examples: 7223 download_size: 599177 dataset_size: 2546658 - config_name: th features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 3451558 num_examples: 9786 download_size: 851558 dataset_size: 3451558 - config_name: tr features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2701219 num_examples: 14209 download_size: 1101256 dataset_size: 2701219 - config_name: transliterated-hi features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1619992 num_examples: 8570 download_size: 646087 dataset_size: 1619992 - config_name: uk features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4528716 num_examples: 18035 download_size: 1523846 dataset_size: 4528716 - config_name: ur features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 1774430 num_examples: 7279 download_size: 576108 dataset_size: 1774430 - config_name: vi features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 2331103 num_examples: 11350 download_size: 893519 dataset_size: 2331103 - config_name: zh features: - name: uuid dtype: string - name: lineid dtype: uint32 - name: obj_uri dtype: string - name: obj_label dtype: string - name: sub_uri dtype: string - name: sub_label dtype: string - name: template dtype: string - name: language dtype: string - name: predicate_id dtype: string - name: options sequence: string splits: - name: test num_bytes: 4178875 num_examples: 21449 download_size: 1747217 dataset_size: 4178875 configs: - config_name: af data_files: - split: test path: af/test-* - config_name: ar data_files: - split: test path: ar/test-* - config_name: az data_files: - split: test path: az/test-* - config_name: be data_files: - split: test path: be/test-* - config_name: bg data_files: - split: test path: bg/test-* - config_name: bn data_files: - split: test path: bn/test-* - config_name: ca data_files: - split: test path: ca/test-* - config_name: ceb data_files: - split: test path: ceb/test-* - config_name: cs data_files: - split: test path: cs/test-* - config_name: cy data_files: - split: test path: cy/test-* - config_name: da data_files: - split: test path: da/test-* - config_name: de data_files: - split: test path: de/test-* - config_name: el data_files: - split: test path: el/test-* - config_name: en data_files: - split: test path: en/test-* - config_name: es data_files: - split: test path: es/test-* - config_name: et data_files: - split: test path: et/test-* - config_name: eu data_files: - split: test path: eu/test-* - config_name: fa data_files: - split: test path: fa/test-* - config_name: fi data_files: - split: test path: fi/test-* - config_name: fr data_files: - split: test path: fr/test-* - config_name: ga data_files: - split: test path: ga/test-* - config_name: gl data_files: - split: test path: gl/test-* - config_name: he data_files: - split: test path: he/test-* - config_name: hi data_files: - split: test path: hi/test-* - config_name: hr data_files: - split: test path: hr/test-* - config_name: hu data_files: - split: test path: hu/test-* - config_name: hy data_files: - split: test path: hy/test-* - config_name: id data_files: - split: test path: id/test-* - config_name: it data_files: - split: test path: it/test-* - config_name: ja data_files: - split: test path: ja/test-* - config_name: ka data_files: - split: test path: ka/test-* - config_name: ko data_files: - split: test path: ko/test-* - config_name: la data_files: - split: test path: la/test-* - config_name: lt data_files: - split: test path: lt/test-* - config_name: lv data_files: - split: test path: lv/test-* - config_name: ms data_files: - split: test path: ms/test-* - config_name: nl data_files: - split: test path: nl/test-* - config_name: pl data_files: - split: test path: pl/test-* - config_name: pt data_files: - split: test path: pt/test-* - config_name: ro data_files: - split: test path: ro/test-* - config_name: ru data_files: - split: test path: ru/test-* - config_name: sk data_files: - split: test path: sk/test-* - config_name: sl data_files: - split: test path: sl/test-* - config_name: sq data_files: - split: test path: sq/test-* - config_name: sr data_files: - split: test path: sr/test-* - config_name: sv data_files: - split: test path: sv/test-* - config_name: ta data_files: - split: test path: ta/test-* - config_name: th data_files: - split: test path: th/test-* - config_name: tr data_files: - split: test path: tr/test-* - config_name: transliterated-hi data_files: - split: test path: transliterated-hi/test-* - config_name: uk data_files: - split: test path: uk/test-* - config_name: ur data_files: - split: test path: ur/test-* - config_name: vi data_files: - split: test path: vi/test-* - config_name: zh data_files: - split: test path: zh/test-* --- Extension/Modification of the original m_lama dataset

本数据集为原始`m_lama`数据集的扩展修改版本。 ### 数据集配置详情本数据集包含多个语言专属配置，每个配置对应一种自然语言，所有配置的特征字段结构完全一致，仅数据集规模与文件大小存在差异。各配置的通用特征字段包括： 1. `uuid`：通用唯一识别码（UUID），数据类型为字符串 2. `lineid`：行编号，数据类型为无符号32位整数（uint32） 3. `obj_uri`：对象统一资源标识符（URI），数据类型为字符串 4. `obj_label`：对象标签，数据类型为字符串 5. `sub_uri`：主体统一资源标识符（URI），数据类型为字符串 6. `sub_label`：主体标签，数据类型为字符串 7. `template`：提示模板，数据类型为字符串 8. `language`：语言标识，数据类型为字符串 9. `predicate_id`：谓词ID，数据类型为字符串 10. `options`：候选选项序列，数据类型为字符串序列所有配置仅包含测试集划分，各配置的测试集参数如下： - 南非语（af）：测试集字节大小1364986，样本量7331；下载大小544481，数据集总大小1364986 - 阿拉伯语（ar）：测试集字节大小4564504，样本量19354；下载大小1580143，数据集总大小4564504 - 阿塞拜疆语（az）：测试集字节大小1467465，样本量7653；下载大小578396，数据集总大小1467465 - 白俄罗斯语（be）：测试集字节大小2285464，样本量8853；下载大小714406，数据集总大小2285464 - 保加利亚语（bg）：测试集字节大小3109085，样本量12461；下载大小1013009，数据集总大小3109085 - 孟加拉语（bn）：测试集字节大小2969863，样本量8975；下载大小748274，数据集总大小2969863 - 加泰罗尼亚语（ca）：测试集字节大小4620850，样本量24287；下载大小1940588，数据集总大小4620850 - 宿务语（ceb）：测试集字节大小1433194，样本量6769；下载大小524854，数据集总大小1433194 - 捷克语（cs）：测试集字节大小2997353，样本量15848；下载大小1246743，数据集总大小2997353 - 威尔士语（cy）：测试集字节大小1901684，样本量9915；下载大小769225，数据集总大小1901684 - 丹麦语（da）：测试集字节大小3672623，样本量19636；下载大小1535250，数据集总大小3672623 - 德语（de）：测试集字节大小6348506，样本量32548；下载大小2613173，数据集总大小6348506 - 希腊语（el）：测试集字节大小3416098，样本量12854；下载大小1074167，数据集总大小3416098 - 英语（en）：测试集字节大小7031572，样本量37498；下载大小3023574，数据集总大小7031572 - 西班牙语（es）：测试集字节大小6000790，样本量31578；下载大小2542929，数据集总大小6000790 - 爱沙尼亚语（et）：测试集字节大小1847160，样本量9880；下载大小748222，数据集总大小1847160 - 巴斯克语（eu）：测试集字节大小2260887，样本量11910；下载大小921424，数据集总大小2260887 - 波斯语（fa）：测试集字节大小4482869，样本量18481；下载大小1497801，数据集总大小4482869 - 芬兰语（fi）：测试集字节大小3575879，样本量19017；下载大小1477166，数据集总大小3575879 - 法语（fr）：测试集字节大小6553643，样本量33872；下载大小2716208，数据集总大小6553643 - 爱尔兰语（ga）：测试集字节大小2809813，样本量13937；下载大小1076939，数据集总大小2809813 - 加利西亚语（gl）：测试集字节大小2062413，样本量10567；下载大小817987，数据集总大小2062413 - 希伯来语（he）：测试集字节大小3273282，样本量14769；下载大小1165490，数据集总大小3273282 - 印地语（hi）：测试集字节大小2750247，样本量8570；下载大小707213，数据集总大小2750247 - 克罗地亚语（hr）：测试集字节大小1766612，样本量9322；下载大小714362，数据集总大小1766612 - 匈牙利语（hu）：测试集字节大小3629786，样本量18850；下载大小1485748，数据集总大小3629786 - 亚美尼亚语（hy）：测试集字节大小2580835，样本量10030；下载大小809063，数据集总大小2580835 - 印尼语（id）：测试集字节大小2693872，样本量14183；下载大小1103155，数据集总大小2693872 - 意大利语（it）：测试集字节大小5287655，样本量27648；下载大小2198936，数据集总大小5287655 - 日语（ja）：测试集字节大小6105411，样本量25356；下载大小2091964，数据集总大小6105411 - 格鲁吉亚语（ka）：测试集字节大小2649721，样本量8099；下载大小647390，数据集总大小2649721 - 韩语（ko）：测试集字节大小3526211，样本量16327；下载大小1309593，数据集总大小3526211 - 拉丁语（la）：测试集字节大小1581833，样本量8061；下载大小612760，数据集总大小1581833 - 立陶宛语（lt）：测试集字节大小1835683，样本量9560；下载大小736354，数据集总大小1835683 - 拉脱维亚语（lv）：测试集字节大小1649860，样本量8474；下载大小643807，数据集总大小1649860 - 马来语（ms）：测试集字节大小1768627，样本量9146；下载大小702211，数据集总大小1768627 - 荷兰语（nl）：测试集字节大小6221612，样本量32423；下载大小2597145，数据集总大小6221612 - 波兰语（pl）：测试集字节大小4013247，样本量20727；下载大小1644648，数据集总大小4013247 - 葡萄牙语（pt）：测试集字节大小4044269，样本量21023；下载大小1653658，数据集总大小4044269 - 罗马尼亚语（ro）：测试集字节大小2523121，样本量12886；下载大小1007651，数据集总大小2523121 - 俄语（ru）：测试集字节大小6405438，样本量25335；下载大小2129105，数据集总大小6405438 - 斯洛伐克语（sk）：测试集字节大小1942547，样本量10205；下载大小788723，数据集总大小1942547 - 斯洛文尼亚语（sl）：测试集字节大小3455705，样本量18091；下载大小1406987，数据集总大小3455705 - 阿尔巴尼亚语（sq）：测试集字节大小2404246，样本量12586；下载大小956395，数据集总大小2404246 - 塞尔维亚语（sr）：测试集字节大小3104514，样本量12477；下载大小1027773，数据集总大小3104514 - 瑞典语（sv）：测试集字节大小4536924，样本量24240；下载大小1905031，数据集总大小4536924 - 泰米尔语（ta）：测试集字节大小2546658，样本量7223；下载大小599177，数据集总大小2546658 - 泰语（th）：测试集字节大小3451558，样本量9786；下载大小851558，数据集总大小3451558 - 土耳其语（tr）：测试集字节大小2701219，样本量14209；下载大小1101256，数据集总大小2701219 - 转写版印地语（transliterated-hi）：测试集字节大小1619992，样本量8570；下载大小646087，数据集总大小1619992 - 乌克兰语（uk）：测试集字节大小4528716，样本量18035；下载大小1523846，数据集总大小4528716 - 乌尔都语（ur）：测试集字节大小1774430，样本量7279；下载大小576108，数据集总大小1774430 - 越南语（vi）：测试集字节大小2331103，样本量11350；下载大小893519，数据集总大小2331103 - 中文（zh）：测试集字节大小4178875，样本量21449；下载大小1747217，数据集总大小4178875 ### 数据集文件配置所有语言配置的测试集数据文件均存储在以对应语言代码命名的目录下，文件匹配模式为`[语言代码]/test-*`，例如南非语配置的数据文件路径为`af/test-*`，中文配置的路径为`zh/test-*`等。

提供机构：

atutej

原始信息汇总

数据集概述

该数据集包含多种语言的配置，每种语言配置下包含相同的特征和测试数据分割。以下是各语言配置的详细信息：

语言配置列表

af (南非荷兰语)
ar (阿拉伯语)
az (阿塞拜疆语)
be (白俄罗斯语)
bg (保加利亚语)
bn (孟加拉语)
ca (加泰罗尼亚语)
ceb (宿务语)
cs (捷克语)
cy (威尔士语)
da (丹麦语)
de (德语)
el (希腊语)
en (英语)
es (西班牙语)
et (爱沙尼亚语)
eu (巴斯克语)
fa (波斯语)
fi (芬兰语)
fr (法语)
ga (爱尔兰语)
gl (加利西亚语)
he (希伯来语)
hi (印地语)
hr (克罗地亚语)
hu (匈牙利语)
hy (亚美尼亚语)
id (印度尼西亚语)
it (意大利语)
ja (日语)
ka (格鲁吉亚语)
ko (韩语)
la (拉丁语)
lt (立陶宛语)
lv (拉脱维亚语)
ms (马来语)
nl (荷兰语)
pl (波兰语)
pt (葡萄牙语)

共同特征

每种语言配置包含以下特征：

uuid: 字符串类型
lineid: 32位无符号整数类型
obj_uri: 字符串类型
obj_label: 字符串类型
sub_uri: 字符串类型
sub_label: 字符串类型
template: 字符串类型
language: 字符串类型
predicate_id: 字符串类型
options: 字符串序列类型

数据分割

每种语言配置包含一个测试数据分割：

test: 包含数据字节数和示例数量

数据大小

每种语言配置的下载大小和数据集大小如下：

af: 下载大小 544481 字节，数据集大小 1364986 字节
ar: 下载大小 1580143 字节，数据集大小 4564504 字节
az: 下载大小 578396 字节，数据集大小 1467465 字节
be: 下载大小 714406 字节，数据集大小 2285464 字节
bg: 下载大小 1013009 字节，数据集大小 3109085 字节
bn: 下载大小 748274 字节，数据集大小 2969863 字节
ca: 下载大小 1940588 字节，数据集大小 4620850 字节
ceb: 下载大小 524854 字节，数据集大小 1433194 字节
cs: 下载大小 1246743 字节，数据集大小 2997353 字节
cy: 下载大小 769225 字节，数据集大小 1901684 字节
da: 下载大小 1535250 字节，数据集大小 3672623 字节
de: 下载大小 2613173 字节，数据集大小 6348506 字节
el: 下载大小 1074167 字节，数据集大小 3416098 字节
en: 下载大小 3023574 字节，数据集大小 7031572 字节
es: 下载大小 2542929 字节，数据集大小 6000790 字节
et: 下载大小 748222 字节，数据集大小 1847160 字节
eu: 下载大小 921424 字节，数据集大小 2260887 字节
fa: 下载大小 1497801 字节，数据集大小 4482869 字节
fi: 下载大小 1477166 字节，数据集大小 3575879 字节
fr: 下载大小 2716208 字节，数据集大小 6553643 字节
ga: 下载大小 1076939 字节，数据集大小 2809813 字节
gl: 下载大小 817987 字节，数据集大小 2062413 字节
he: 下载大小 1165490 字节，数据集大小 3273282 字节
hi: 下载大小 707213 字节，数据集大小 2750247 字节
hr: 下载大小 714362 字节，数据集大小 1766612 字节
hu: 下载大小 1485748 字节，数据集大小 3629786 字节
hy: 下载大小 809063 字节，数据集大小 2580835 字节
id: 下载大小 1103155 字节，数据集大小 2693872 字节
it: 下载大小 2198936 字节，数据集大小 5287655 字节
ja: 下载大小 2091964 字节，数据集大小 6105411 字节
ka: 下载大小 647390 字节，数据集大小 2649721 字节
ko: 下载大小 1309593 字节，数据集大小 3526211 字节
la: 下载大小 612760 字节，数据集大小 1581833 字节
lt: 下载大小 736354 字节，数据集大小 1835683 字节
lv: 下载大小 643807 字节，数据集大小 1649860 字节
ms: 下载大小 702211 字节，数据集大小 1768627 字节
nl: 下载大小 2597145 字节，数据集大小 6221612 字节
pl: 下载大小 1644648 字节，数据集大小 4013247 字节
pt: 下载大小 1580143 字节，数据集大小 4564504 字节

搜集汇总

数据集介绍

构建方式

在自然语言处理与知识图谱的交叉领域中，多语言知识探测任务日益受到关注。atutej/m_lama数据集正是为评估多语言预训练语言模型对结构化世界知识的掌握程度而构建。该数据集源自原始LAMA（LAnguage Model Analysis）基准，通过将维基百科知识库中的事实三元组（主语、关系、宾语）转化为自然语言完形填空模板，从而形成测试样本。每个样本包含唯一标识符、行号、主语与宾语的URI及标签、自然语言模板、语言标签、关系谓词ID以及一组候选宾语选项。数据集的构建过程确保了知识事实的跨语言对齐，覆盖了从南非荷兰语到中文等数十种语言，每种语言独立配置，便于针对性评估。

特点

atutej/m_lama数据集最显著的特点在于其广泛的多语言覆盖范围，囊括了超过50种不同语系的语言，从印欧语系到闪含语系乃至汉藏语系均有涉及，为跨语言知识推理研究提供了坚实的数据基础。每个语言配置均包含独立的测试集，样本量从数千到数万不等，反映了各语言维基百科的知识密度差异。数据集以结构化形式存储，每条记录详尽记录了知识事实的语义成分和元信息，特别是为每个样本预先生成了候选答案列表，这为评估语言模型的排他性知识选择能力提供了便利。此外，所有样本均保持统一的字段格式，极大简化了多语言实验的预处理流程。

使用方法

研究者可通过HuggingFace Datasets库便捷地加载atutej/m_lama数据集，指定所需语言配置即可获取对应测试集。使用方法上，通常将数据中的自然语言模板作为输入提示，要求语言模型预测被遮蔽的宾语标签。由于数据集提供了预定义的候选选项列表，用户既可以采用基于困惑度的排序方法进行零样本评估，也可以构建分类任务来微调模型。值得注意的是，数据集中包含的URI和谓词ID信息可用于与外部知识图谱进行关联分析，从而深入探究模型的知识边界。对于多语言对比实验，建议按语言配置分别评估，以揭示不同语言环境下知识表征的异同。

背景与挑战

背景概述

在自然语言处理领域，评估预训练语言模型对世界知识的掌握程度一直是研究焦点。mLAMA数据集（Multilingual LAnguage Model Analysis）由研究者于2020年左右创建，旨在系统性地衡量多语言模型在事实性知识方面的表现。该数据集基于Google Research的LAMA（LAnguage Model Analysis）框架进行多语言扩展，覆盖包括英语、法语、德语、中文等在内的数十种语言，每个语言配置均包含从维基百科知识库中提取的（主语、关系、宾语）三元组。核心研究问题在于探究跨语言模型是否能够准确回忆事实性知识，以及不同语言间知识表征的差异。该数据集的出现对多语言预训练模型的评估体系产生了深远影响，推动了语言模型知识记忆能力的标准化评测。

当前挑战

mLAMA数据集面临的核心挑战之一在于多语言事实知识的覆盖不均衡。低资源语言（如阿塞拜疆语、白俄罗斯语）的样本数量显著少于高资源语言（如英语、德语），导致模型性能评估存在偏差。构建过程中，从维基百科抽取三元组时面临关系模板多样性和语言特异性表达的难题，例如不同语言的句法结构差异要求为每种语言设计适配的模板。此外，数据集的静态性难以捕捉知识随时间演化的特性，使得基于固定时间点抽取的事实知识可能逐渐过时。这些挑战共同制约了mLAMA作为多语言知识记忆基准的全面性和时效性。

常用场景

经典使用场景

在自然语言处理与知识图谱的交叉领域中，m_lama数据集常被用于评估多语言预训练语言模型的事实知识存储能力。该数据集通过将维基百科中的实体关系三元组转化为完形填空形式的模板，例如填充“巴黎是_的首都”中的空白，从而精准检验模型对结构化知识的掌握程度。其覆盖53种语言、包含超过百万条测试样本的庞大规模，使其成为跨语言知识探测任务中的标杆性基准。研究者利用此数据集可系统性地对比不同语言模型在统一知识图谱上的表现差异，揭示模型在多语言环境下的知识迁移与记忆特性。

衍生相关工作

基于m_lama数据集，学术界衍生出一系列具有影响力的经典工作。其中，X-FACTR作为其多语言扩展版本，进一步增加了事实模板的多样性，用于更细致地分析模型在不同关系类型上的知识表现。另一项重要工作是mLAMA的持续训练策略研究，探索通过跨语言知识蒸馏来提升低资源语言的事实预测能力。此外，该数据集还被用于开发知识编辑（Knowledge Editing）方法，如修改模型中对特定事实的记忆，并评估编辑操作在不同语言上的泛化效果。这些衍生工作共同推动了多语言知识图谱与语言模型融合领域的理论进展与实践创新。

数据集最近研究