Datasets:
Tasks:
Token Classification
Formats:
parquet
Sub-tasks:
named-entity-recognition
Languages:
Arabic
Size:
10K - 100K
License:
Download validation.parquet from HebArabNlpProject/ShamNER: direct link, hf CLI and curl.
- Browser
- Download file 166 kB
-
https://huggingface.co/datasets/HebArabNlpProject/ShamNER/resolve/main/validation.parquet
- Command line
-
hf download hf://datasets/HebArabNlpProject/ShamNER/validation.parquet
-
curl -L -o validation.parquet https://huggingface.co/datasets/HebArabNlpProject/ShamNER/resolve/main/validation.parquet
166 kB
- Xet hash:
- a3e48c4e669545e78c822d9714916b8105d8a7a1d83820c7386b03fcefa8b934
- Size of remote file:
- 166 kB
- SHA256:
- 23973340c3dffdb0b2196072141a80a177d0e2e15954814427a5e5eb59dcaa9c
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.