Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
4a3a171
1 Parent(s): e9d7276

Add X-CSQA-pl data files

Browse files
README.md CHANGED
@@ -763,13 +763,13 @@ dataset_info:
763
  dtype: string
764
  splits:
765
  - name: test
766
- num_bytes: 231781
767
  num_examples: 1074
768
  - name: validation
769
- num_bytes: 220096
770
  num_examples: 1000
771
- download_size: 7519903
772
- dataset_size: 451877
773
  - config_name: X-CSQA-pt
774
  features:
775
  - name: id
@@ -975,6 +975,12 @@ configs:
975
  path: X-CSQA-nl/test-*
976
  - split: validation
977
  path: X-CSQA-nl/validation-*
 
 
 
 
 
 
978
  - config_name: X-CSQA-zh
979
  data_files:
980
  - split: test
 
763
  dtype: string
764
  splits:
765
  - name: test
766
+ num_bytes: 231479
767
  num_examples: 1074
768
  - name: validation
769
+ num_bytes: 219814
770
  num_examples: 1000
771
+ download_size: 245829
772
+ dataset_size: 451293
773
  - config_name: X-CSQA-pt
774
  features:
775
  - name: id
 
975
  path: X-CSQA-nl/test-*
976
  - split: validation
977
  path: X-CSQA-nl/validation-*
978
+ - config_name: X-CSQA-pl
979
+ data_files:
980
+ - split: test
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
X-CSQA-pl/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73b58a714885d6589dcb24eb04329cb1ff699565e067a26d9de56d3bebb6dc8f
3
+ size 127512
X-CSQA-pl/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2439aa3654e5e94f68596b2657fbb7cfdd861e35134c6f586cc46c951b31381
3
+ size 118317
dataset_infos.json CHANGED
@@ -535,48 +535,38 @@
535
  "features": {
536
  "id": {
537
  "dtype": "string",
538
- "id": null,
539
  "_type": "Value"
540
  },
541
  "lang": {
542
  "dtype": "string",
543
- "id": null,
544
  "_type": "Value"
545
  },
546
  "question": {
547
  "stem": {
548
  "dtype": "string",
549
- "id": null,
550
  "_type": "Value"
551
  },
552
  "choices": {
553
  "feature": {
554
  "label": {
555
  "dtype": "string",
556
- "id": null,
557
  "_type": "Value"
558
  },
559
  "text": {
560
  "dtype": "string",
561
- "id": null,
562
  "_type": "Value"
563
  }
564
  },
565
- "length": -1,
566
- "id": null,
567
  "_type": "Sequence"
568
  }
569
  },
570
  "answerKey": {
571
  "dtype": "string",
572
- "id": null,
573
  "_type": "Value"
574
  }
575
  },
576
- "post_processed": null,
577
- "supervised_keys": null,
578
- "task_templates": null,
579
  "builder_name": "xcsr",
 
580
  "config_name": "X-CSQA-pl",
581
  "version": {
582
  "version_str": "1.1.0",
@@ -588,27 +578,20 @@
588
  "splits": {
589
  "test": {
590
  "name": "test",
591
- "num_bytes": 231781,
592
  "num_examples": 1074,
593
- "dataset_name": "xcsr"
594
  },
595
  "validation": {
596
  "name": "validation",
597
- "num_bytes": 220096,
598
  "num_examples": 1000,
599
- "dataset_name": "xcsr"
600
- }
601
- },
602
- "download_checksums": {
603
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
604
- "num_bytes": 7519903,
605
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
606
  }
607
  },
608
- "download_size": 7519903,
609
- "post_processing_size": null,
610
- "dataset_size": 451877,
611
- "size_in_bytes": 7971780
612
  },
613
  "X-CSQA-pt": {
614
  "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",
 
535
  "features": {
536
  "id": {
537
  "dtype": "string",
 
538
  "_type": "Value"
539
  },
540
  "lang": {
541
  "dtype": "string",
 
542
  "_type": "Value"
543
  },
544
  "question": {
545
  "stem": {
546
  "dtype": "string",
 
547
  "_type": "Value"
548
  },
549
  "choices": {
550
  "feature": {
551
  "label": {
552
  "dtype": "string",
 
553
  "_type": "Value"
554
  },
555
  "text": {
556
  "dtype": "string",
 
557
  "_type": "Value"
558
  }
559
  },
 
 
560
  "_type": "Sequence"
561
  }
562
  },
563
  "answerKey": {
564
  "dtype": "string",
 
565
  "_type": "Value"
566
  }
567
  },
 
 
 
568
  "builder_name": "xcsr",
569
+ "dataset_name": "xcsr",
570
  "config_name": "X-CSQA-pl",
571
  "version": {
572
  "version_str": "1.1.0",
 
578
  "splits": {
579
  "test": {
580
  "name": "test",
581
+ "num_bytes": 231479,
582
  "num_examples": 1074,
583
+ "dataset_name": null
584
  },
585
  "validation": {
586
  "name": "validation",
587
+ "num_bytes": 219814,
588
  "num_examples": 1000,
589
+ "dataset_name": null
 
 
 
 
 
 
590
  }
591
  },
592
+ "download_size": 245829,
593
+ "dataset_size": 451293,
594
+ "size_in_bytes": 697122
 
595
  },
596
  "X-CSQA-pt": {
597
  "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",