Datasets:
Tasks:
Token Classification
Sub-tasks:
named-entity-recognition
Languages:
German
Size:
100K<n<1M
License:

albertvillanova
HF staff
Convert dataset sizes from base 2 to base 10 in the dataset card (#4)
a42b589