temsa/OpenMed-mLiteClinical-IrishCorePII-135M-v1

Full transformers checkpoint derived from OpenMed/OpenMed-PII-mLiteClinical-Base-135M-v1 and tuned for Irish core PII:

• PPSN
• account_number
• bank_routing_number
• credit_debit_card
• PASSPORT_NUMBER
• postcode
• phone_number
• email
• first_name
• last_name
• swift_bic

The main focus is English + Irish Gaelic (ga) handling for Irish administrative, citizen-support, and HSE-style text.

Included Artifacts

• Full transformers model files in the repo root
• Dynamic int8 ONNX export in onnx/model_quantized.onnx
• inference_mask.py for the full model
• inference_mask_onnx.py for the ONNX int8 artifact
• clean benchmark summaries in eval/

Recommended Inference

Highest accuracy:

python3 inference_mask.py \
  --text "My PPSN is 1234567TW and call me on 087 123 4567." \
  --json

Fast CPU path:

python3 inference_mask_onnx.py \
  --text "My PPSN is 1234567TW and call me on 087 123 4567." \
  --json

Dedicated Eircode example:

python3 inference_mask.py \
  --text "My Eircode is D02 X285." \
  --json

Benchmarks

Reference comparison on the manual Irish core suite and PPSN regression suites:

Label	Base OpenMed	Previous Public Model	This Release	ONNX Q8
PPSN	0.0000	0.0800	0.8000	0.7273
account_number	0.3333	0.3333	1.0000	1.0000
bank_routing_number	0.0000	0.0000	1.0000	1.0000
credit_debit_card	0.1538	0.1818	1.0000	0.3333
PASSPORT_NUMBER	0.0000	0.0000	1.0000	1.0000
postcode	0.0000	0.0000	1.0000	1.0000
phone_number	0.0000	0.0000	0.8571	0.8571
email	0.7059	1.0000	1.0000	1.0000
first_name	0.8947	0.8947	1.0000	1.0000
last_name	0.8889	0.8889	1.0000	1.0000
swift_bic	0.0000	0.0000	1.0000	1.0000

Edge and multilingual PPSN checks:

Suite	Base OpenMed	Previous Public Model	This Release	ONNX Q8
edge_ppsn	0.0000	0.4211	0.5000	0.4000
edge_phone_number	0.1429	0.1429	0.6316	0.5000
multilingual_ppsn	0.0000	0.9704	0.9940	0.9882

Multilingual PPSN throughput on CPU (eval/multilingual_ppsn_v1_all.jsonl):

• Base OpenMed: 42.30 examples/s
• Previous public PPSN model: 42.63 examples/s
• This release: 41.18 examples/s
• ONNX Q8: 81.99 examples/s

Practical Reading Of The Benchmarks

• This release is materially better than the previous public PPSN-only model on Irish phones, Eircodes, account details, passport numbers, and names.
• The bundled ONNX int8 export is useful for CPU speed, but it is not accuracy-identical to the full checkpoint.
• The largest ONNX drops are on credit_debit_card and some PPSN edge cases. Use the full model when those matter.

License And Attribution

• Model weights in this repo are distributed under Apache-2.0.
• Base model: OpenMed/OpenMed-PII-mLiteClinical-Base-135M-v1
• Training data included synthetic Irish data plus attributed upstream data from:
  • joelniklaus/mapa (cc-by-4.0)
  • gretelai/synthetic_pii_finance_multilingual (apache-2.0)
• See NOTICE for attribution details.