Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
Datasets:
Sprakbanken
/
Norwegian_idioms
like
2
Follow
Nasjonalbiblioteket Språkbanken
7
Tasks:
Text Generation
Fill-Mask
Modalities:
Text
Formats:
json
Languages:
Norwegian
Norwegian Bokmål
Norwegian Nynorsk
Size:
1K - 10K
Libraries:
Datasets
pandas
Croissant
+ 1
License:
cc0-1.0
Dataset card
Viewer
Files
Files and versions
Community
1
refs/convert/parquet
Norwegian_idioms
/
default
/
test
1 contributor
History:
1 commit
parquet-converter
Update parquet files
8b1bc67
verified
2 months ago
0000.parquet
65 kB
LFS
Update parquet files
2 months ago