Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
621ad19
1 Parent(s): 9150acc

Add X-CSQA-es data files

Browse files
README.md CHANGED
@@ -601,13 +601,13 @@ dataset_info:
601
  dtype: string
602
  splits:
603
  - name: test
604
- num_bytes: 237119
605
  num_examples: 1074
606
  - name: validation
607
- num_bytes: 224779
608
  num_examples: 1000
609
- download_size: 7519903
610
- dataset_size: 461898
611
  - config_name: X-CSQA-fr
612
  features:
613
  - name: id
@@ -945,6 +945,12 @@ configs:
945
  path: X-CSQA-en/test-*
946
  - split: validation
947
  path: X-CSQA-en/validation-*
 
 
 
 
 
 
948
  - config_name: X-CSQA-zh
949
  data_files:
950
  - split: test
 
601
  dtype: string
602
  splits:
603
  - name: test
604
+ num_bytes: 236817
605
  num_examples: 1074
606
  - name: validation
607
+ num_bytes: 224497
608
  num_examples: 1000
609
+ download_size: 238810
610
+ dataset_size: 461314
611
  - config_name: X-CSQA-fr
612
  features:
613
  - name: id
 
945
  path: X-CSQA-en/test-*
946
  - split: validation
947
  path: X-CSQA-en/validation-*
948
+ - config_name: X-CSQA-es
949
+ data_files:
950
+ - split: test
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
X-CSQA-es/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef789374918ad4d28c127d4f2786a63f8e9884c7d473d0c9ece3dbe2d4747c7a
3
+ size 123793
X-CSQA-es/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9bfe47edd6782b700aa5ae2c444cb6c1dbc75c7bab28c66a767ef3e1d98fa292
3
+ size 115017
dataset_infos.json CHANGED
@@ -205,48 +205,38 @@
205
  "features": {
206
  "id": {
207
  "dtype": "string",
208
- "id": null,
209
  "_type": "Value"
210
  },
211
  "lang": {
212
  "dtype": "string",
213
- "id": null,
214
  "_type": "Value"
215
  },
216
  "question": {
217
  "stem": {
218
  "dtype": "string",
219
- "id": null,
220
  "_type": "Value"
221
  },
222
  "choices": {
223
  "feature": {
224
  "label": {
225
  "dtype": "string",
226
- "id": null,
227
  "_type": "Value"
228
  },
229
  "text": {
230
  "dtype": "string",
231
- "id": null,
232
  "_type": "Value"
233
  }
234
  },
235
- "length": -1,
236
- "id": null,
237
  "_type": "Sequence"
238
  }
239
  },
240
  "answerKey": {
241
  "dtype": "string",
242
- "id": null,
243
  "_type": "Value"
244
  }
245
  },
246
- "post_processed": null,
247
- "supervised_keys": null,
248
- "task_templates": null,
249
  "builder_name": "xcsr",
 
250
  "config_name": "X-CSQA-es",
251
  "version": {
252
  "version_str": "1.1.0",
@@ -258,27 +248,20 @@
258
  "splits": {
259
  "test": {
260
  "name": "test",
261
- "num_bytes": 237119,
262
  "num_examples": 1074,
263
- "dataset_name": "xcsr"
264
  },
265
  "validation": {
266
  "name": "validation",
267
- "num_bytes": 224779,
268
  "num_examples": 1000,
269
- "dataset_name": "xcsr"
270
- }
271
- },
272
- "download_checksums": {
273
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
274
- "num_bytes": 7519903,
275
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
276
  }
277
  },
278
- "download_size": 7519903,
279
- "post_processing_size": null,
280
- "dataset_size": 461898,
281
- "size_in_bytes": 7981801
282
  },
283
  "X-CSQA-fr": {
284
  "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",
 
205
  "features": {
206
  "id": {
207
  "dtype": "string",
 
208
  "_type": "Value"
209
  },
210
  "lang": {
211
  "dtype": "string",
 
212
  "_type": "Value"
213
  },
214
  "question": {
215
  "stem": {
216
  "dtype": "string",
 
217
  "_type": "Value"
218
  },
219
  "choices": {
220
  "feature": {
221
  "label": {
222
  "dtype": "string",
 
223
  "_type": "Value"
224
  },
225
  "text": {
226
  "dtype": "string",
 
227
  "_type": "Value"
228
  }
229
  },
 
 
230
  "_type": "Sequence"
231
  }
232
  },
233
  "answerKey": {
234
  "dtype": "string",
 
235
  "_type": "Value"
236
  }
237
  },
 
 
 
238
  "builder_name": "xcsr",
239
+ "dataset_name": "xcsr",
240
  "config_name": "X-CSQA-es",
241
  "version": {
242
  "version_str": "1.1.0",
 
248
  "splits": {
249
  "test": {
250
  "name": "test",
251
+ "num_bytes": 236817,
252
  "num_examples": 1074,
253
+ "dataset_name": null
254
  },
255
  "validation": {
256
  "name": "validation",
257
+ "num_bytes": 224497,
258
  "num_examples": 1000,
259
+ "dataset_name": null
 
 
 
 
 
 
260
  }
261
  },
262
+ "download_size": 238810,
263
+ "dataset_size": 461314,
264
+ "size_in_bytes": 700124
 
265
  },
266
  "X-CSQA-fr": {
267
  "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",