Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
5a483a2
1 Parent(s): 2ec1431

Add X-CSQA-ru data files

Browse files
README.md CHANGED
@@ -817,13 +817,13 @@ dataset_info:
817
  dtype: string
818
  splits:
819
  - name: test
820
- num_bytes: 342051
821
  num_examples: 1074
822
  - name: validation
823
- num_bytes: 324006
824
  num_examples: 1000
825
- download_size: 7519903
826
- dataset_size: 666057
827
  - config_name: X-CSQA-sw
828
  features:
829
  - name: id
@@ -987,6 +987,12 @@ configs:
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
 
817
  dtype: string
818
  splits:
819
  - name: test
820
+ num_bytes: 341749
821
  num_examples: 1074
822
  - name: validation
823
+ num_bytes: 323724
824
  num_examples: 1000
825
+ download_size: 296252
826
+ dataset_size: 665473
827
  - config_name: X-CSQA-sw
828
  features:
829
  - name: id
 
987
  path: X-CSQA-pt/test-*
988
  - split: validation
989
  path: X-CSQA-pt/validation-*
990
+ - config_name: X-CSQA-ru
991
+ data_files:
992
+ - split: test
993
+ path: X-CSQA-ru/test-*
994
+ - split: validation
995
+ path: X-CSQA-ru/validation-*
996
  - config_name: X-CSQA-zh
997
  data_files:
998
  - split: test
X-CSQA-ru/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ba031ff6919f86fd2cc408b29aab2b0598de07aed3c75bd3f6643305630cb72
3
+ size 153653
X-CSQA-ru/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff000ebc60c1428ae62c0ec469a8834a386265992fc2edc251b84005f21ba50a
3
+ size 142599
dataset_infos.json CHANGED
@@ -667,48 +667,38 @@
667
  "features": {
668
  "id": {
669
  "dtype": "string",
670
- "id": null,
671
  "_type": "Value"
672
  },
673
  "lang": {
674
  "dtype": "string",
675
- "id": null,
676
  "_type": "Value"
677
  },
678
  "question": {
679
  "stem": {
680
  "dtype": "string",
681
- "id": null,
682
  "_type": "Value"
683
  },
684
  "choices": {
685
  "feature": {
686
  "label": {
687
  "dtype": "string",
688
- "id": null,
689
  "_type": "Value"
690
  },
691
  "text": {
692
  "dtype": "string",
693
- "id": null,
694
  "_type": "Value"
695
  }
696
  },
697
- "length": -1,
698
- "id": null,
699
  "_type": "Sequence"
700
  }
701
  },
702
  "answerKey": {
703
  "dtype": "string",
704
- "id": null,
705
  "_type": "Value"
706
  }
707
  },
708
- "post_processed": null,
709
- "supervised_keys": null,
710
- "task_templates": null,
711
  "builder_name": "xcsr",
 
712
  "config_name": "X-CSQA-ru",
713
  "version": {
714
  "version_str": "1.1.0",
@@ -720,27 +710,20 @@
720
  "splits": {
721
  "test": {
722
  "name": "test",
723
- "num_bytes": 342051,
724
  "num_examples": 1074,
725
- "dataset_name": "xcsr"
726
  },
727
  "validation": {
728
  "name": "validation",
729
- "num_bytes": 324006,
730
  "num_examples": 1000,
731
- "dataset_name": "xcsr"
732
- }
733
- },
734
- "download_checksums": {
735
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
736
- "num_bytes": 7519903,
737
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
738
  }
739
  },
740
- "download_size": 7519903,
741
- "post_processing_size": null,
742
- "dataset_size": 666057,
743
- "size_in_bytes": 8185960
744
  },
745
  "X-CSQA-ar": {
746
  "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",
 
667
  "features": {
668
  "id": {
669
  "dtype": "string",
 
670
  "_type": "Value"
671
  },
672
  "lang": {
673
  "dtype": "string",
 
674
  "_type": "Value"
675
  },
676
  "question": {
677
  "stem": {
678
  "dtype": "string",
 
679
  "_type": "Value"
680
  },
681
  "choices": {
682
  "feature": {
683
  "label": {
684
  "dtype": "string",
 
685
  "_type": "Value"
686
  },
687
  "text": {
688
  "dtype": "string",
 
689
  "_type": "Value"
690
  }
691
  },
 
 
692
  "_type": "Sequence"
693
  }
694
  },
695
  "answerKey": {
696
  "dtype": "string",
 
697
  "_type": "Value"
698
  }
699
  },
 
 
 
700
  "builder_name": "xcsr",
701
+ "dataset_name": "xcsr",
702
  "config_name": "X-CSQA-ru",
703
  "version": {
704
  "version_str": "1.1.0",
 
710
  "splits": {
711
  "test": {
712
  "name": "test",
713
+ "num_bytes": 341749,
714
  "num_examples": 1074,
715
+ "dataset_name": null
716
  },
717
  "validation": {
718
  "name": "validation",
719
+ "num_bytes": 323724,
720
  "num_examples": 1000,
721
+ "dataset_name": null
 
 
 
 
 
 
722
  }
723
  },
724
+ "download_size": 296252,
725
+ "dataset_size": 665473,
726
+ "size_in_bytes": 961725
 
727
  },
728
  "X-CSQA-ar": {
729
  "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",