Datasets:
Tasks:
Token Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
named-entity-recognition
Size:
< 1K
License:
Download default/test_rm/0000.parquet from ZurichNLP/swissner: direct link, hf CLI and curl.
- Browser
- Download file 54.7 kB
-
https://huggingface.co/datasets/ZurichNLP/swissner/resolve/refs%2Fconvert%2Fparquet/default/test_rm/0000.parquet
- Command line
-
hf download hf://datasets/ZurichNLP/swissner@refs/convert/parquet/default/test_rm/0000.parquet
-
curl -L -o 0000.parquet https://huggingface.co/datasets/ZurichNLP/swissner/resolve/refs%2Fconvert%2Fparquet/default/test_rm/0000.parquet
54.7 kB
- Xet hash:
- af628ff6cf75bd73edf4e71e164a95515f76bd66f1492895812625b1269a1f68
- Size of remote file:
- 54.7 kB
- SHA256:
- 3eb1c086cdfb3bbf3de4a569bbf9821a6b79b14a36b75c6ae13889ba8f904877
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.