sharanharsoor/pii-ner-model-gretel

PII NER Model

This model is trained to detect Personal Identifiable Information (PII) in text. It can identify various types of PII including:

• Names (first and last)
• Email addresses
• Phone numbers
• Social Security Numbers (SSN)
• Credit card numbers
• Medical record numbers
• Employee IDs
• Addresses
• And more

Usage

from transformers import AutoModelForTokenClassification, AutoTokenizer

model = AutoModelForTokenClassification.from_pretrained("sharanharsoor/pii-ner-model")
tokenizer = AutoTokenizer.from_pretrained("sharanharsoor/pii-ner-model")

Training Data

The model was trained on the Gretel PII dataset with custom modifications.

Performance

The model achieves:

• F1 Score: 0.9451
• Precision: 0.9401
• Recall: 0.9502
• Accuracy: 0.9952