Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
2ec1431
1 Parent(s): 4a3a171

Add X-CSQA-pt data files

Browse files
README.md CHANGED
@@ -790,13 +790,13 @@ dataset_info:
790
  dtype: string
791
  splits:
792
  - name: test
793
- num_bytes: 235771
794
  num_examples: 1074
795
  - name: validation
796
- num_bytes: 223067
797
  num_examples: 1000
798
- download_size: 7519903
799
- dataset_size: 458838
800
  - config_name: X-CSQA-ru
801
  features:
802
  - name: id
@@ -981,6 +981,12 @@ configs:
981
  path: X-CSQA-pl/test-*
982
  - split: validation
983
  path: X-CSQA-pl/validation-*
 
 
 
 
 
 
984
  - config_name: X-CSQA-zh
985
  data_files:
986
  - split: test
 
790
  dtype: string
791
  splits:
792
  - name: test
793
+ num_bytes: 235469
794
  num_examples: 1074
795
  - name: validation
796
+ num_bytes: 222785
797
  num_examples: 1000
798
+ download_size: 238902
799
+ dataset_size: 458254
800
  - config_name: X-CSQA-ru
801
  features:
802
  - name: id
 
981
  path: X-CSQA-pl/test-*
982
  - split: validation
983
  path: X-CSQA-pl/validation-*
984
+ - config_name: X-CSQA-pt
985
+ data_files:
986
+ - split: test
987
+ path: X-CSQA-pt/test-*
988
+ - split: validation
989
+ path: X-CSQA-pt/validation-*
990
  - config_name: X-CSQA-zh
991
  data_files:
992
  - split: test
X-CSQA-pt/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:338eaee42c8c26cd3b343f1cb7b1ec366cf2ccbd234fba18633e3adb75e8f6e6
3
+ size 123812
X-CSQA-pt/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1356f434852bedba7fc3f8b121364f4efb20a6d85f0e21f6bd5d678495632ad3
3
+ size 115090
dataset_infos.json CHANGED
@@ -601,48 +601,38 @@
601
  "features": {
602
  "id": {
603
  "dtype": "string",
604
- "id": null,
605
  "_type": "Value"
606
  },
607
  "lang": {
608
  "dtype": "string",
609
- "id": null,
610
  "_type": "Value"
611
  },
612
  "question": {
613
  "stem": {
614
  "dtype": "string",
615
- "id": null,
616
  "_type": "Value"
617
  },
618
  "choices": {
619
  "feature": {
620
  "label": {
621
  "dtype": "string",
622
- "id": null,
623
  "_type": "Value"
624
  },
625
  "text": {
626
  "dtype": "string",
627
- "id": null,
628
  "_type": "Value"
629
  }
630
  },
631
- "length": -1,
632
- "id": null,
633
  "_type": "Sequence"
634
  }
635
  },
636
  "answerKey": {
637
  "dtype": "string",
638
- "id": null,
639
  "_type": "Value"
640
  }
641
  },
642
- "post_processed": null,
643
- "supervised_keys": null,
644
- "task_templates": null,
645
  "builder_name": "xcsr",
 
646
  "config_name": "X-CSQA-pt",
647
  "version": {
648
  "version_str": "1.1.0",
@@ -654,27 +644,20 @@
654
  "splits": {
655
  "test": {
656
  "name": "test",
657
- "num_bytes": 235771,
658
  "num_examples": 1074,
659
- "dataset_name": "xcsr"
660
  },
661
  "validation": {
662
  "name": "validation",
663
- "num_bytes": 223067,
664
  "num_examples": 1000,
665
- "dataset_name": "xcsr"
666
- }
667
- },
668
- "download_checksums": {
669
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
670
- "num_bytes": 7519903,
671
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
672
  }
673
  },
674
- "download_size": 7519903,
675
- "post_processing_size": null,
676
- "dataset_size": 458838,
677
- "size_in_bytes": 7978741
678
  },
679
  "X-CSQA-ru": {
680
  "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",
 
601
  "features": {
602
  "id": {
603
  "dtype": "string",
 
604
  "_type": "Value"
605
  },
606
  "lang": {
607
  "dtype": "string",
 
608
  "_type": "Value"
609
  },
610
  "question": {
611
  "stem": {
612
  "dtype": "string",
 
613
  "_type": "Value"
614
  },
615
  "choices": {
616
  "feature": {
617
  "label": {
618
  "dtype": "string",
 
619
  "_type": "Value"
620
  },
621
  "text": {
622
  "dtype": "string",
 
623
  "_type": "Value"
624
  }
625
  },
 
 
626
  "_type": "Sequence"
627
  }
628
  },
629
  "answerKey": {
630
  "dtype": "string",
 
631
  "_type": "Value"
632
  }
633
  },
 
 
 
634
  "builder_name": "xcsr",
635
+ "dataset_name": "xcsr",
636
  "config_name": "X-CSQA-pt",
637
  "version": {
638
  "version_str": "1.1.0",
 
644
  "splits": {
645
  "test": {
646
  "name": "test",
647
+ "num_bytes": 235469,
648
  "num_examples": 1074,
649
+ "dataset_name": null
650
  },
651
  "validation": {
652
  "name": "validation",
653
+ "num_bytes": 222785,
654
  "num_examples": 1000,
655
+ "dataset_name": null
 
 
 
 
 
 
656
  }
657
  },
658
+ "download_size": 238902,
659
+ "dataset_size": 458254,
660
+ "size_in_bytes": 697156
 
661
  },
662
  "X-CSQA-ru": {
663
  "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",