bert-tiny-training-mid-tuned-12260542-tuned-12260554-tuned-12260643

This model is a fine-tuned version of broadfield-dev/bert-tiny-training-mid-tuned-12260542-tuned-12260554 on the ai4privacy/pii-masking-400k dataset.

Training Details

• Task: TOKEN_CLS
• Columns: Input: source_text Output: privacy_mask
• Rows: Start: 0 End: 320000
• Epochs: 1
• Learning Rate: 0.0001
• Gradient Accumulation Steps: 4

Entity Labels

['ACCOUNTNUM', 'BUILDINGNUM', 'CITY', 'CREDITCARDNUMBER', 'DATEOFBIRTH', 'DRIVERLICENSENUM', 'EMAIL', 'GIVENNAME', 'IDCARDNUM', 'O', 'PASSWORD', 'SOCIALNUM', 'STREET', 'SURNAME', 'TAXNUM', 'TELEPHONENUM', 'USERNAME', 'ZIPCODE']