Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
bf9faae
1 Parent(s): 3c45489

Add X-CODAH-pt data files

Browse files
README.md CHANGED
@@ -348,13 +348,13 @@ dataset_info:
348
  dtype: string
349
  splits:
350
  - name: test
351
- num_bytes: 455869
352
  num_examples: 1000
353
  - name: validation
354
- num_bytes: 132045
355
  num_examples: 300
356
- download_size: 7519903
357
- dataset_size: 587914
358
  - config_name: X-CODAH-ru
359
  features:
360
  - name: id
@@ -981,6 +981,12 @@ configs:
981
  path: X-CODAH-pl/test-*
982
  - split: validation
983
  path: X-CODAH-pl/validation-*
 
 
 
 
 
 
984
  - config_name: X-CODAH-zh
985
  data_files:
986
  - split: test
 
348
  dtype: string
349
  splits:
350
  - name: test
351
+ num_bytes: 455583
352
  num_examples: 1000
353
  - name: validation
354
+ num_bytes: 131933
355
  num_examples: 300
356
+ download_size: 238858
357
+ dataset_size: 587516
358
  - config_name: X-CODAH-ru
359
  features:
360
  - name: id
 
981
  path: X-CODAH-pl/test-*
982
  - split: validation
983
  path: X-CODAH-pl/validation-*
984
+ - config_name: X-CODAH-pt
985
+ data_files:
986
+ - split: test
987
+ path: X-CODAH-pt/test-*
988
+ - split: validation
989
+ path: X-CODAH-pt/validation-*
990
  - config_name: X-CODAH-zh
991
  data_files:
992
  - split: test
X-CODAH-pt/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e34f45c2401b1462f14fea6cd2d157be674d4dbe98867086934fd26059b8a639
3
+ size 182516
X-CODAH-pt/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0e26f8f1356a9df6d8c2a99720d33af519306db7fe38ae52f8a3cfc8ddcfebd
3
+ size 56342
dataset_infos.json CHANGED
@@ -1693,53 +1693,42 @@
1693
  "features": {
1694
  "id": {
1695
  "dtype": "string",
1696
- "id": null,
1697
  "_type": "Value"
1698
  },
1699
  "lang": {
1700
  "dtype": "string",
1701
- "id": null,
1702
  "_type": "Value"
1703
  },
1704
  "question_tag": {
1705
  "dtype": "string",
1706
- "id": null,
1707
  "_type": "Value"
1708
  },
1709
  "question": {
1710
  "stem": {
1711
  "dtype": "string",
1712
- "id": null,
1713
  "_type": "Value"
1714
  },
1715
  "choices": {
1716
  "feature": {
1717
  "label": {
1718
  "dtype": "string",
1719
- "id": null,
1720
  "_type": "Value"
1721
  },
1722
  "text": {
1723
  "dtype": "string",
1724
- "id": null,
1725
  "_type": "Value"
1726
  }
1727
  },
1728
- "length": -1,
1729
- "id": null,
1730
  "_type": "Sequence"
1731
  }
1732
  },
1733
  "answerKey": {
1734
  "dtype": "string",
1735
- "id": null,
1736
  "_type": "Value"
1737
  }
1738
  },
1739
- "post_processed": null,
1740
- "supervised_keys": null,
1741
- "task_templates": null,
1742
  "builder_name": "xcsr",
 
1743
  "config_name": "X-CODAH-pt",
1744
  "version": {
1745
  "version_str": "1.1.0",
@@ -1751,27 +1740,20 @@
1751
  "splits": {
1752
  "test": {
1753
  "name": "test",
1754
- "num_bytes": 455869,
1755
  "num_examples": 1000,
1756
- "dataset_name": "xcsr"
1757
  },
1758
  "validation": {
1759
  "name": "validation",
1760
- "num_bytes": 132045,
1761
  "num_examples": 300,
1762
- "dataset_name": "xcsr"
1763
- }
1764
- },
1765
- "download_checksums": {
1766
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
1767
- "num_bytes": 7519903,
1768
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
1769
  }
1770
  },
1771
- "download_size": 7519903,
1772
- "post_processing_size": null,
1773
- "dataset_size": 587914,
1774
- "size_in_bytes": 8107817
1775
  },
1776
  "X-CODAH-ru": {
1777
  "description": "To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.\n",
 
1693
  "features": {
1694
  "id": {
1695
  "dtype": "string",
 
1696
  "_type": "Value"
1697
  },
1698
  "lang": {
1699
  "dtype": "string",
 
1700
  "_type": "Value"
1701
  },
1702
  "question_tag": {
1703
  "dtype": "string",
 
1704
  "_type": "Value"
1705
  },
1706
  "question": {
1707
  "stem": {
1708
  "dtype": "string",
 
1709
  "_type": "Value"
1710
  },
1711
  "choices": {
1712
  "feature": {
1713
  "label": {
1714
  "dtype": "string",
 
1715
  "_type": "Value"
1716
  },
1717
  "text": {
1718
  "dtype": "string",
 
1719
  "_type": "Value"
1720
  }
1721
  },
 
 
1722
  "_type": "Sequence"
1723
  }
1724
  },
1725
  "answerKey": {
1726
  "dtype": "string",
 
1727
  "_type": "Value"
1728
  }
1729
  },
 
 
 
1730
  "builder_name": "xcsr",
1731
+ "dataset_name": "xcsr",
1732
  "config_name": "X-CODAH-pt",
1733
  "version": {
1734
  "version_str": "1.1.0",
 
1740
  "splits": {
1741
  "test": {
1742
  "name": "test",
1743
+ "num_bytes": 455583,
1744
  "num_examples": 1000,
1745
+ "dataset_name": null
1746
  },
1747
  "validation": {
1748
  "name": "validation",
1749
+ "num_bytes": 131933,
1750
  "num_examples": 300,
1751
+ "dataset_name": null
 
 
 
 
 
 
1752
  }
1753
  },
1754
+ "download_size": 238858,
1755
+ "dataset_size": 587516,
1756
+ "size_in_bytes": 826374
 
1757
  },
1758
  "X-CODAH-ru": {
1759
  "description": "To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.\n",