ettin-68m-nemotron-pii model

Light Weight PII Detection Model | Open Source | 68M Parameters | 96.27 F1 Score | Blog Post

Overview

Ettin-68m-nemotron-pii is based on the ettin-encoder-68M model and fine-tuned over the Nemotron PII dataset. This model can detect 50+ PII entities in both structured and unstructured texts across various domains like healthcare, finance, legal, cybersecurity etc. With just 68M parameters, the model achieves a strong F1-score of 96.27.

Key Features

• Achieves strong F1-score of 96.27 with just 68M parameters.
• Outperforms popular LLMs like DeepSeek-V4-Flash (84.89) and GPT-4o-Mini (78.69).
• Detects 50+ PII entities in both structured and unstructured texts.
• Handles text in various domains like healthcare, finance, legal etc.

Model Visualization

Supported PII Entity Types

This model can detect the following 55 PII entity types

PII entity types with description

Entity	Description
account_number	Account Number
age	Age
api_key	API Key
bank_routing_number	Bank Routing Number
biometric_identifier	Biometric Identifier
blood_type	Blood Type
certificate_license_number	Certificate or License Number
city	City
company_name	Company Name
coordinate	Geographic Coordinate
country	Country
county	County
credit_debit_card	Credit or Debit Card Number
customer_id	Customer ID
cvv	Card Verification Value (CVV)
date	Date
date_of_birth	Date of Birth
date_time	Date and Time
device_identifier	Device Identifier
education_level	Education Level
email	Email Address
employee_id	Employee ID
employment_status	Employment Status
fax_number	Fax Number
first_name	First Name
gender	Gender
health_plan_beneficiary_number	Health Plan Beneficiary Number
http_cookie	HTTP Cookie
ipv4	IPv4 Address
ipv6	IPv6 Address
language	Language
last_name	Last Name
license_plate	Vehicle License Plate
mac_address	MAC Address
medical_record_number	Medical Record Number
national_id	National Identification Number
occupation	Occupation
password	Password
phone_number	Phone Number
pin	Personal Identification Number (PIN)
political_view	Political View
postcode	Postcode / Zip Code
race_ethnicity	Race or Ethnicity
religious_belief	Religious Belief
sexuality	Sexuality / Sexual Orientation
ssn	Social Security Number
state	State
street_address	Street Address
swift_bic	SWIFT / BIC Code
tax_id	Tax Identification Number
time	Time
unique_id	Unique Identifier
url	URL / Web Address
user_name	Username
vehicle_identifier	Vehicle Identification Number (VIN)

Usage

# First install Hugging Face transformers library
!pip install transformers

# Initialize and run the PII detection pipeline to extract PII entities
from transformers import pipeline

## Initialize the PII detection pipeline
ner = pipeline("ner", model="kalyan-ks/ettin-68m-nemotron-pii", aggregation_strategy="simple")

input_text = "Kalyan KS is from India. His email id is [email protected]"

## Run the PII detection to extract PII entities
pii_entities = ner(input_text)

## Process the extracted PII entities 
def format_pii_entities(entities, original_text):
    if not entities:
        return []

    merged_entities = []

    entities = sorted(entities, key=lambda x: x['start'])

    current_entity = {
        'start': entities[0]['start'],
        'end': entities[0]['end'],
        'label': entities[0]['entity_group'],
        'text': entities[0]['word']
    }

    for next_ent in entities[1:]:
        is_same_label = next_ent['entity_group'] == current_entity['label']
        is_adjacent = next_ent['start'] <= current_entity['end'] + 1

        if is_same_label and is_adjacent:
            current_entity['end'] = max(current_entity['end'], next_ent['end'])
            current_entity['text'] = original_text[current_entity['start']:current_entity['end']]
        else:
            merged_entities.append(clean_entity(current_entity))
            current_entity = {
                'start': next_ent['start'],
                'end': next_ent['end'],
                'label': next_ent['entity_group'],
                'text': next_ent['word']
            }

    merged_entities.append(clean_entity(current_entity))
    return merged_entities

def clean_entity(ent):

    raw_text = ent['text']
    stripped_text = raw_text.strip()
    leading_spaces = len(raw_text) - len(raw_text.lstrip())

    return {
        'start': ent['start'] + leading_spaces,
        'end': ent['start'] + leading_spaces + len(stripped_text),
        'text': stripped_text,
        'label': ent['label']
    }

# Display the extracted PII entities
formatted_entities = format_pii_entities(pii_entities, input_text)
print(formatted_entities)

# Output
[{'start': 0, 'end': 9, 'text': 'Kalyan KS', 'label': 'first_name'}, {'start': 18, 'end': 23, 'text': 'India', 'label': 'country'}, {'start': 41, 'end': 60, 'text': '[email protected]', 'label': 'email'}]

Evaluation

This model is evaluated on a 10k sample test set from Neomotron PII dataset and achieved the following results

Metric	Score
F1	96.27
Precision	96.35
Recall	96.19
Accuracy	99.26

Top Performing PII Entity Types

Entity	Precision	Recall	F1
biometric_identifier	0.9963	0.9966	0.9964
date_of_birth	0.9952	0.9963	0.9957
api_key	0.9932	0.9978	0.9955
mac_address	0.9929	0.9965	0.9947
email	0.9942	0.9942	0.9942
ipv4	0.9950	0.9933	0.9941
medical_record_number	0.9952	0.9904	0.9928
health_plan_beneficiary_number	0.9924	0.9925	0.9924
vehicle_identifier	0.9867	0.9977	0.9922
bank_routing_number	0.9967	0.9862	0.9914

Challenging PII Entity Types

Entity	Precision	Recall	F1
occupation	0.7200	0.5493	0.6232
time	0.8094	0.7781	0.7934
age	0.8333	0.9273	0.8778
political_view	0.8533	0.9247	0.8876
state	0.9077	0.8792	0.8932
fax_number	0.9047	0.9013	0.9030
company_name	0.9048	0.9072	0.9060
national_id	0.8995	0.9224	0.9108
education_level	0.9269	0.8973	0.9118
race_ethnicity	0.9027	0.9388	0.9204

Limitations

• Language: This model works well only for English language texts.
• Challenging PII Entity Types: Some of the entity types like occupation has low F1 score.

Citation

@misc{ettin-68m-pii-2026,
  title = {ettin-68m-nemotron-pii-2026: PII Detection Model},
  author = {Kalyan KS},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/kalyan-ks/ettin-17m-nemotron-pii}
}