Datasets:
Formats:
csv
Sub-tasks:
semantic-similarity-classification
Languages:
Swedish
Size:
100K - 1M
License:
File too large to display, you can
check the raw version
instead.