Add ai4privacy/pii-masking-300k to model card metadata (evaluation benchmark)

README.md

@@ -1,6 +1,8 @@
 ---
 license: apache-2.0
 pipeline_tag: token-classification
+datasets:
+- ai4privacy/pii-masking-300k
 library_name: transformers
 tags:
 - transformers.js
@@ -182,19 +184,3 @@ Additional caution is warranted in high-sensitivity settings such as medical, le
 - Evaluate in-domain with local policy references before production.
 - Use task-specific fine-tuning when policy differs from base boundaries.
 - Keep human review paths for high-sensitivity workflows.
-
-## Evaluation Benchmarks and Data Attribution
-
-As described in the [release announcement](https://openai.com/index/introducing-openai-privacy-filter/), Privacy Filter was evaluated on the [PII-Masking-300k](https://huggingface.co/datasets/ai4privacy/pii-masking-300k) benchmark by [Ai4Privacy](https://huggingface.co/ai4privacy) (a project affiliated with [Ai Suisse SA](https://www.aisuisse.com/)), where it achieves an F1 score of 96% (94.04% precision, 98.04% recall), and 97.43% F1 on a corrected version of the benchmark.
-
-If you use or build on this evaluation, please cite the benchmark:
-
-```bibtex
-@dataset{ai4privacy_pii_masking_300k,
-  author = {Ai4Privacy},
-  title = {PII-Masking-300k},
-  year = {2024},
-  publisher = {Hugging Face},
-  url = {https://huggingface.co/datasets/ai4privacy/pii-masking-300k}
-}
-```