Datasets:
Tasks:
Token Classification
Modalities:
Text
Formats:
parquet
Languages:
Slovenian
Size:
10K - 100K
DOI:
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -8,6 +8,8 @@ size_categories:
|
|
| 8 |
- 10K<n<100K
|
| 9 |
source_datasets:
|
| 10 |
- extended
|
|
|
|
|
|
|
| 11 |
tags:
|
| 12 |
- text
|
| 13 |
- pii
|
|
@@ -314,11 +316,11 @@ The canonical representation is model-agnostic and can be converted for architec
|
|
| 314 |
|
| 315 |
## License
|
| 316 |
|
| 317 |
-
MultiPII-X
|
| 318 |
|
| 319 |
-
|
| 320 |
|
| 321 |
-
|
| 322 |
|
| 323 |
## Citation
|
| 324 |
|
|
|
|
| 8 |
- 10K<n<100K
|
| 9 |
source_datasets:
|
| 10 |
- extended
|
| 11 |
+
license:
|
| 12 |
+
- cc-by-4.0
|
| 13 |
tags:
|
| 14 |
- text
|
| 15 |
- pii
|
|
|
|
| 316 |
|
| 317 |
## License
|
| 318 |
|
| 319 |
+
MultiPII-X is released under the **Creative Commons Attribution 4.0 International (CC BY 4.0)** license.
|
| 320 |
|
| 321 |
+
Users may share and adapt the dataset, including for commercial use, provided appropriate attribution is given and changes are indicated.
|
| 322 |
|
| 323 |
+
Upstream source datasets and corpora retain their own attribution requirements. See `CITATION.cff` for provenance and source citations.
|
| 324 |
|
| 325 |
## Citation
|
| 326 |
|