Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
b1e305c
1 Parent(s): 4595814

Add X-CODAH-es data files

Browse files
README.md CHANGED
@@ -145,13 +145,13 @@ dataset_info:
145
  dtype: string
146
  splits:
147
  - name: test
148
- num_bytes: 451240
149
  num_examples: 1000
150
  - name: validation
151
- num_bytes: 130790
152
  num_examples: 300
153
- download_size: 7519903
154
- dataset_size: 582030
155
  - config_name: X-CODAH-fr
156
  features:
157
  - name: id
@@ -945,6 +945,12 @@ configs:
945
  path: X-CODAH-en/test-*
946
  - split: validation
947
  path: X-CODAH-en/validation-*
 
 
 
 
 
 
948
  - config_name: X-CODAH-zh
949
  data_files:
950
  - split: test
 
145
  dtype: string
146
  splits:
147
  - name: test
148
+ num_bytes: 450954
149
  num_examples: 1000
150
  - name: validation
151
+ num_bytes: 130678
152
  num_examples: 300
153
+ download_size: 242647
154
+ dataset_size: 581632
155
  - config_name: X-CODAH-fr
156
  features:
157
  - name: id
 
945
  path: X-CODAH-en/test-*
946
  - split: validation
947
  path: X-CODAH-en/validation-*
948
+ - config_name: X-CODAH-es
949
+ data_files:
950
+ - split: test
951
+ path: X-CODAH-es/test-*
952
+ - split: validation
953
+ path: X-CODAH-es/validation-*
954
  - config_name: X-CODAH-zh
955
  data_files:
956
  - split: test
X-CODAH-es/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87609acbd0051832bf24020a91c2da6d6ebdcded7027d59707a852e0d04c2bd3
3
+ size 185567
X-CODAH-es/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8900f2b24f9d051ad8e71830b7e733340d74ea9cc4f5a0ef5b0231ad65efb94e
3
+ size 57080
dataset_infos.json CHANGED
@@ -1273,53 +1273,42 @@
1273
  "features": {
1274
  "id": {
1275
  "dtype": "string",
1276
- "id": null,
1277
  "_type": "Value"
1278
  },
1279
  "lang": {
1280
  "dtype": "string",
1281
- "id": null,
1282
  "_type": "Value"
1283
  },
1284
  "question_tag": {
1285
  "dtype": "string",
1286
- "id": null,
1287
  "_type": "Value"
1288
  },
1289
  "question": {
1290
  "stem": {
1291
  "dtype": "string",
1292
- "id": null,
1293
  "_type": "Value"
1294
  },
1295
  "choices": {
1296
  "feature": {
1297
  "label": {
1298
  "dtype": "string",
1299
- "id": null,
1300
  "_type": "Value"
1301
  },
1302
  "text": {
1303
  "dtype": "string",
1304
- "id": null,
1305
  "_type": "Value"
1306
  }
1307
  },
1308
- "length": -1,
1309
- "id": null,
1310
  "_type": "Sequence"
1311
  }
1312
  },
1313
  "answerKey": {
1314
  "dtype": "string",
1315
- "id": null,
1316
  "_type": "Value"
1317
  }
1318
  },
1319
- "post_processed": null,
1320
- "supervised_keys": null,
1321
- "task_templates": null,
1322
  "builder_name": "xcsr",
 
1323
  "config_name": "X-CODAH-es",
1324
  "version": {
1325
  "version_str": "1.1.0",
@@ -1331,27 +1320,20 @@
1331
  "splits": {
1332
  "test": {
1333
  "name": "test",
1334
- "num_bytes": 451240,
1335
  "num_examples": 1000,
1336
- "dataset_name": "xcsr"
1337
  },
1338
  "validation": {
1339
  "name": "validation",
1340
- "num_bytes": 130790,
1341
  "num_examples": 300,
1342
- "dataset_name": "xcsr"
1343
- }
1344
- },
1345
- "download_checksums": {
1346
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
1347
- "num_bytes": 7519903,
1348
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
1349
  }
1350
  },
1351
- "download_size": 7519903,
1352
- "post_processing_size": null,
1353
- "dataset_size": 582030,
1354
- "size_in_bytes": 8101933
1355
  },
1356
  "X-CODAH-fr": {
1357
  "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",
 
1273
  "features": {
1274
  "id": {
1275
  "dtype": "string",
 
1276
  "_type": "Value"
1277
  },
1278
  "lang": {
1279
  "dtype": "string",
 
1280
  "_type": "Value"
1281
  },
1282
  "question_tag": {
1283
  "dtype": "string",
 
1284
  "_type": "Value"
1285
  },
1286
  "question": {
1287
  "stem": {
1288
  "dtype": "string",
 
1289
  "_type": "Value"
1290
  },
1291
  "choices": {
1292
  "feature": {
1293
  "label": {
1294
  "dtype": "string",
 
1295
  "_type": "Value"
1296
  },
1297
  "text": {
1298
  "dtype": "string",
 
1299
  "_type": "Value"
1300
  }
1301
  },
 
 
1302
  "_type": "Sequence"
1303
  }
1304
  },
1305
  "answerKey": {
1306
  "dtype": "string",
 
1307
  "_type": "Value"
1308
  }
1309
  },
 
 
 
1310
  "builder_name": "xcsr",
1311
+ "dataset_name": "xcsr",
1312
  "config_name": "X-CODAH-es",
1313
  "version": {
1314
  "version_str": "1.1.0",
 
1320
  "splits": {
1321
  "test": {
1322
  "name": "test",
1323
+ "num_bytes": 450954,
1324
  "num_examples": 1000,
1325
+ "dataset_name": null
1326
  },
1327
  "validation": {
1328
  "name": "validation",
1329
+ "num_bytes": 130678,
1330
  "num_examples": 300,
1331
+ "dataset_name": null
 
 
 
 
 
 
1332
  }
1333
  },
1334
+ "download_size": 242647,
1335
+ "dataset_size": 581632,
1336
+ "size_in_bytes": 824279
 
1337
  },
1338
  "X-CODAH-fr": {
1339
  "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",