training_sources.json

{
  "base_model": "OpenMed/OpenMed-PII-mLiteClinical-Base-135M-v1",
  "current_public_reference": "temsa/OpenMed-mLiteClinical-IrishCorePII-135M-v1",
  "release_purpose": "Targeted weak-context PPSN recovery for the IrishCorePII release line.",
  "recommended_thresholds": {
    "ppsn_min_score": 0.5,
    "other_min_score": 0.4
  },
  "training_mix_summary": [
    {
      "component": "duplicated weak-context PPSN regression cases",
      "weight": 7.0
    },
    {
      "component": "Irish PPSN and phone edge-case replay",
      "weight": 3.0
    },
    {
      "component": "synthetic PPSN focus data with weak-context positives and hard negatives",
      "weight": 4.0
    },
    {
      "component": "broader Irish core PII replay mix",
      "weight": 5.0
    },
    {
      "component": "spec-driven Irish PPSN and Eircode synthetic data",
      "weight": 1.0
    }
  ],
  "upstream_attribution": [
    {
      "name": "joelniklaus/mapa",
      "license": "CC-BY-4.0"
    },
    {
      "name": "gretelai/synthetic_pii_finance_multilingual",
      "license": "Apache-2.0"
    }
  ]
}