Datasets:
Tasks:
Token Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
named-entity-recognition
Languages:
English
Size:
100K - 1M

davidberenstein1957
HF staff
Upload data/train-00000-of-00001-1cc4d2bfaa7112a4.parquet with huggingface_hub
31d45c7