Datasets:

Multilinguality:
multilingual
Size Categories:
1K<n<10K
Annotations Creators:
crowdsourced
ArXiv:
Tags:
License:
albertvillanova HF staff commited on
Commit
14ca350
1 Parent(s): 9682793

Add X-CODAH-sw data files

Browse files
README.md CHANGED
@@ -406,13 +406,13 @@ dataset_info:
406
  dtype: string
407
  splits:
408
  - name: test
409
- num_bytes: 423707
410
  num_examples: 1000
411
  - name: validation
412
- num_bytes: 124882
413
  num_examples: 300
414
- download_size: 7519903
415
- dataset_size: 548589
416
  - config_name: X-CODAH-ur
417
  features:
418
  - name: id
@@ -1005,6 +1005,12 @@ configs:
1005
  path: X-CODAH-ru/test-*
1006
  - split: validation
1007
  path: X-CODAH-ru/validation-*
 
 
 
 
 
 
1008
  - config_name: X-CODAH-vi
1009
  data_files:
1010
  - split: test
 
406
  dtype: string
407
  splits:
408
  - name: test
409
+ num_bytes: 423421
410
  num_examples: 1000
411
  - name: validation
412
+ num_bytes: 124770
413
  num_examples: 300
414
+ download_size: 214100
415
+ dataset_size: 548191
416
  - config_name: X-CODAH-ur
417
  features:
418
  - name: id
 
1005
  path: X-CODAH-ru/test-*
1006
  - split: validation
1007
  path: X-CODAH-ru/validation-*
1008
+ - config_name: X-CODAH-sw
1009
+ data_files:
1010
+ - split: test
1011
+ path: X-CODAH-sw/test-*
1012
+ - split: validation
1013
+ path: X-CODAH-sw/validation-*
1014
  - config_name: X-CODAH-vi
1015
  data_files:
1016
  - split: test
X-CODAH-sw/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20a3bd3e08344bb50a5e4ad3847ceb831af678c3502d649603775a99bac55522
3
+ size 163235
X-CODAH-sw/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:564928508c05cbb5c52c56ffb1bf99ea1421e5ab4875ef00467e5d1c88731247
3
+ size 50865
dataset_infos.json CHANGED
@@ -2043,53 +2043,42 @@
2043
  "features": {
2044
  "id": {
2045
  "dtype": "string",
2046
- "id": null,
2047
  "_type": "Value"
2048
  },
2049
  "lang": {
2050
  "dtype": "string",
2051
- "id": null,
2052
  "_type": "Value"
2053
  },
2054
  "question_tag": {
2055
  "dtype": "string",
2056
- "id": null,
2057
  "_type": "Value"
2058
  },
2059
  "question": {
2060
  "stem": {
2061
  "dtype": "string",
2062
- "id": null,
2063
  "_type": "Value"
2064
  },
2065
  "choices": {
2066
  "feature": {
2067
  "label": {
2068
  "dtype": "string",
2069
- "id": null,
2070
  "_type": "Value"
2071
  },
2072
  "text": {
2073
  "dtype": "string",
2074
- "id": null,
2075
  "_type": "Value"
2076
  }
2077
  },
2078
- "length": -1,
2079
- "id": null,
2080
  "_type": "Sequence"
2081
  }
2082
  },
2083
  "answerKey": {
2084
  "dtype": "string",
2085
- "id": null,
2086
  "_type": "Value"
2087
  }
2088
  },
2089
- "post_processed": null,
2090
- "supervised_keys": null,
2091
- "task_templates": null,
2092
  "builder_name": "xcsr",
 
2093
  "config_name": "X-CODAH-sw",
2094
  "version": {
2095
  "version_str": "1.1.0",
@@ -2101,27 +2090,20 @@
2101
  "splits": {
2102
  "test": {
2103
  "name": "test",
2104
- "num_bytes": 423707,
2105
  "num_examples": 1000,
2106
- "dataset_name": "xcsr"
2107
  },
2108
  "validation": {
2109
  "name": "validation",
2110
- "num_bytes": 124882,
2111
  "num_examples": 300,
2112
- "dataset_name": "xcsr"
2113
- }
2114
- },
2115
- "download_checksums": {
2116
- "https://inklab.usc.edu/XCSR/xcsr_datasets.zip": {
2117
- "num_bytes": 7519903,
2118
- "checksum": "c45b29ece740643252d5402e76be1e33f96f9d6910053f79e80d39887f10c85e"
2119
  }
2120
  },
2121
- "download_size": 7519903,
2122
- "post_processing_size": null,
2123
- "dataset_size": 548589,
2124
- "size_in_bytes": 8068492
2125
  },
2126
  "X-CODAH-ur": {
2127
  "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",
 
2043
  "features": {
2044
  "id": {
2045
  "dtype": "string",
 
2046
  "_type": "Value"
2047
  },
2048
  "lang": {
2049
  "dtype": "string",
 
2050
  "_type": "Value"
2051
  },
2052
  "question_tag": {
2053
  "dtype": "string",
 
2054
  "_type": "Value"
2055
  },
2056
  "question": {
2057
  "stem": {
2058
  "dtype": "string",
 
2059
  "_type": "Value"
2060
  },
2061
  "choices": {
2062
  "feature": {
2063
  "label": {
2064
  "dtype": "string",
 
2065
  "_type": "Value"
2066
  },
2067
  "text": {
2068
  "dtype": "string",
 
2069
  "_type": "Value"
2070
  }
2071
  },
 
 
2072
  "_type": "Sequence"
2073
  }
2074
  },
2075
  "answerKey": {
2076
  "dtype": "string",
 
2077
  "_type": "Value"
2078
  }
2079
  },
 
 
 
2080
  "builder_name": "xcsr",
2081
+ "dataset_name": "xcsr",
2082
  "config_name": "X-CODAH-sw",
2083
  "version": {
2084
  "version_str": "1.1.0",
 
2090
  "splits": {
2091
  "test": {
2092
  "name": "test",
2093
+ "num_bytes": 423421,
2094
  "num_examples": 1000,
2095
+ "dataset_name": null
2096
  },
2097
  "validation": {
2098
  "name": "validation",
2099
+ "num_bytes": 124770,
2100
  "num_examples": 300,
2101
+ "dataset_name": null
 
 
 
 
 
 
2102
  }
2103
  },
2104
+ "download_size": 214100,
2105
+ "dataset_size": 548191,
2106
+ "size_in_bytes": 762291
 
2107
  },
2108
  "X-CODAH-ur": {
2109
  "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",