Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
f958747
1 Parent(s): 621ad19

Add X-CSQA-fr data files

Browse files
README.md CHANGED
@@ -628,13 +628,13 @@ dataset_info:
628
  dtype: string
629
  splits:
630
  - name: test
631
- num_bytes: 244254
632
  num_examples: 1074
633
  - name: validation
634
- num_bytes: 231678
635
  num_examples: 1000
636
- download_size: 7519903
637
- dataset_size: 475932
638
  - config_name: X-CSQA-hi
639
  features:
640
  - name: id
@@ -951,6 +951,12 @@ configs:
951
  path: X-CSQA-es/test-*
952
  - split: validation
953
  path: X-CSQA-es/validation-*
 
 
 
 
 
 
954
  - config_name: X-CSQA-zh
955
  data_files:
956
  - split: test
 
628
  dtype: string
629
  splits:
630
  - name: test
631
+ num_bytes: 243952
632
  num_examples: 1074
633
  - name: validation
634
+ num_bytes: 231396
635
  num_examples: 1000
636
+ download_size: 244676
637
+ dataset_size: 475348
638
  - config_name: X-CSQA-hi
639
  features:
640
  - name: id
 
951
  path: X-CSQA-es/test-*
952
  - split: validation
953
  path: X-CSQA-es/validation-*
954
+ - config_name: X-CSQA-fr
955
+ data_files:
956
+ - split: test
957
+ path: X-CSQA-fr/test-*
958
+ - split: validation
959
+ path: X-CSQA-fr/validation-*
960
  - config_name: X-CSQA-zh
961
  data_files:
962
  - split: test
X-CSQA-fr/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ff82bc034d7362625318ff7803e947d56abe080604dc37bda660141b8bac94f
3
+ size 127005
X-CSQA-fr/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3d94097dfda92373d1f678f567b31c6943265b0146d1e4bec0ce4c0378fe4b1
3
+ size 117671
dataset_infos.json CHANGED
@@ -271,48 +271,38 @@
271
  "features": {
272
  "id": {
273
  "dtype": "string",
274
- "id": null,
275
  "_type": "Value"
276
  },
277
  "lang": {
278
  "dtype": "string",
279
- "id": null,
280
  "_type": "Value"
281
  },
282
  "question": {
283
  "stem": {
284
  "dtype": "string",
285
- "id": null,
286
  "_type": "Value"
287
  },
288
  "choices": {
289
  "feature": {
290
  "label": {
291
  "dtype": "string",
292
- "id": null,
293
  "_type": "Value"
294
  },
295
  "text": {
296
  "dtype": "string",
297
- "id": null,
298
  "_type": "Value"
299
  }
300
  },
301
- "length": -1,
302
- "id": null,
303
  "_type": "Sequence"
304
  }
305
  },
306
  "answerKey": {
307
  "dtype": "string",
308
- "id": null,
309
  "_type": "Value"
310
  }
311
  },
312
- "post_processed": null,
313
- "supervised_keys": null,
314
- "task_templates": null,
315
  "builder_name": "xcsr",
 
316
  "config_name": "X-CSQA-fr",
317
  "version": {
318
  "version_str": "1.1.0",
@@ -324,27 +314,20 @@
324
  "splits": {
325
  "test": {
326
  "name": "test",
327
- "num_bytes": 244254,
328
  "num_examples": 1074,
329
- "dataset_name": "xcsr"
330
  },
331
  "validation": {
332
  "name": "validation",
333
- "num_bytes": 231678,
334
  "num_examples": 1000,
335
- "dataset_name": "xcsr"
336
- }
337
- },
338
- "download_checksums": {
339
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
340
- "num_bytes": 7519903,
341
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
342
  }
343
  },
344
- "download_size": 7519903,
345
- "post_processing_size": null,
346
- "dataset_size": 475932,
347
- "size_in_bytes": 7995835
348
  },
349
  "X-CSQA-it": {
350
  "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",
 
271
  "features": {
272
  "id": {
273
  "dtype": "string",
 
274
  "_type": "Value"
275
  },
276
  "lang": {
277
  "dtype": "string",
 
278
  "_type": "Value"
279
  },
280
  "question": {
281
  "stem": {
282
  "dtype": "string",
 
283
  "_type": "Value"
284
  },
285
  "choices": {
286
  "feature": {
287
  "label": {
288
  "dtype": "string",
 
289
  "_type": "Value"
290
  },
291
  "text": {
292
  "dtype": "string",
 
293
  "_type": "Value"
294
  }
295
  },
 
 
296
  "_type": "Sequence"
297
  }
298
  },
299
  "answerKey": {
300
  "dtype": "string",
 
301
  "_type": "Value"
302
  }
303
  },
 
 
 
304
  "builder_name": "xcsr",
305
+ "dataset_name": "xcsr",
306
  "config_name": "X-CSQA-fr",
307
  "version": {
308
  "version_str": "1.1.0",
 
314
  "splits": {
315
  "test": {
316
  "name": "test",
317
+ "num_bytes": 243952,
318
  "num_examples": 1074,
319
+ "dataset_name": null
320
  },
321
  "validation": {
322
  "name": "validation",
323
+ "num_bytes": 231396,
324
  "num_examples": 1000,
325
+ "dataset_name": null
 
 
 
 
 
 
326
  }
327
  },
328
+ "download_size": 244676,
329
+ "dataset_size": 475348,
330
+ "size_in_bytes": 720024
 
331
  },
332
  "X-CSQA-it": {
333
  "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",