54-type PII coverage
The publisher states the model classifies 54 sensitive-information types spanning personal, financial, medical, and contact categories.
Open Source Model Profile · OpenMed
OpenMed-PII-SuperClinical-Small-44M-v1
OpenMed-PII-SuperClinical-Small-44M-v1 is an OpenMed token-classification fine-tune for PII detection. Its card reports 54 sensitive-information types with Apache-2.0 licensing.
- Publisher
- OpenMed
- Task
- token-classification
- Model type
- deberta-v2
- License
- apache-2.0
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
OpenMed-PII-SuperClinical-Small-44M-v1 is published by OpenMed for token-classification PII detection. The captured configuration identifies DebertaV2ForTokenClassification with model type deberta-v2, and Safetensors metadata reports 141,385,834 parameters. According to the model card, it identifies 54 types of sensitive information such as names, addresses, SSNs, and medical record numbers.
Recorded capabilities
0.9539 micro-F1 report
According to the model card’s 2,000-sample Nemotron-PII test split, the model reports 0.9539 micro F1 with 0.9549 precision and 0.9529 recall.
DeBERTa small fine-tune
Hub and page context name microsoft/deberta-v3-small as base with nvidia/Nemotron-PII as training dataset and a 384-token setup.
Use cases in the source record
- Automated redaction of PII in clinical notes, medical records, and documents, as listed in the card’s intended de-identification use.
- Compliance-support, preprocessing, and audit workflows for identifying PII across document collections under HIPAA and GDPR handling.
Limitations and unknowns
- The 44M name and the captured 141,385,834-parameter record differ; both values are reported without reconciling them.
- Reported scores are publisher test results, not Ethen-measured benchmarks; challenging entities such as occupation, time, and sexuality score lower and may need post-processing.
- The card notes English-primary training, possible false negatives, and context sensitivity for critical applications.
- Provider state is historical snapshot data and should be refreshed before being presented as current.
Source and provenance
Source: OpenMed/OpenMed-PII-SuperClinical-Small-44M-v1
Captured: Unknown. Processed: 2026-09-07T19:35:50.841486+00:00.
OpenMed-PII-SuperClinical-Small-44M-v1 PII Detection Model | 44M Parameters | Open Source Model Description OpenMed-PII-SuperClinical-Small-44M-v1 is a transformer-based token classification model fine-tuned for Personally Identifiable Information (PII) detection in text. This model identifies and classifies 54 types of sensitive information including names, addresses, SSNs, medical record numbers, and more. Key Features High Accuracy : Achieves strong F1 scores across diverse PII categories Comprehensive Coverage : Detects 50+ entity types spanning personal, financial, medical, and contact information Privacy-Focused : Designed for…
F001F002F003F004F005F006F007F008F009F010F011F012F013F014F016F017F018