modernbert-pii

ModernBERT fine-tuned for token classification (NER) to detect PII in Australian clinical notes. Labels use the BIO scheme over these entity types:

  • ACD
  • ADDRESS
  • AGE
  • ATTENDING_NURSE
  • COUNTRY_OF_BIRTH
  • DATE_OF_BIRTH
  • DVA_NUMBER
  • EMAIL
  • FULL_NAME
  • GUARDIANSHIP
  • IHI
  • MEDICARE_NUMBER
  • MRN
  • NEXT_OF_KIN
  • OCCUPATION
  • OTHER_CLINICAL_STAFF
  • PHONE_NUMBER
  • PREFERRED_NAME
  • PRIVATE_HEALTH_NUMBER
  • REFERRING_GP
  • SEX_GENDER
  • TREATING_CLINICIAN
  • WORKCOVER

Usage

from transformers import pipeline

ner = pipeline("token-classification", model="Cbezz/modernbert-pii", aggregation_strategy="simple")
ner("Patient John Smith, DOB 12/03/1980, MRN 4456219.")
Downloads last month
42
Safetensors
Model size
0.1B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Cbezz/modernbert-pii

Finetuned
(1385)
this model