Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
json
Sub-tasks:
fact-checking
Languages:
English
Size:
100K - 1M
License:
File size: 256 Bytes
edf1e6c |
1 2 3 4 |
copenlu--fever_gold_evidence/json-train.parquet filter=lfs diff=lfs merge=lfs -text
copenlu--fever_gold_evidence/json-test.parquet filter=lfs diff=lfs merge=lfs -text
copenlu--fever_gold_evidence/json-validation.parquet filter=lfs diff=lfs merge=lfs -text
|