nemotron-pii-ready/evaluations/span-based_deberta/eval_summary.json

{
  "model_name": "Span-Based DeBERTa",
  "dataset": "nemotron-pii-ready",
  "seeds": [
    42,
    123,
    456,
    789,
    999
  ],
  "max_samples_per_seed": 1000,
  "avg_precision": 0.8795,
  "avg_recall": 0.95,
  "avg_f1": 0.8994,
  "per_seed": {
    "42": {
      "precision": 0.8789,
      "recall": 0.9533,
      "f1": 0.9006
    },
    "123": {
      "precision": 0.8768,
      "recall": 0.9449,
      "f1": 0.895
    },
    "456": {
      "precision": 0.8736,
      "recall": 0.9405,
      "f1": 0.8915
    },
    "789": {
      "precision": 0.8883,
      "recall": 0.9628,
      "f1": 0.9104
    },
    "999": {
      "precision": 0.8799,
      "recall": 0.9483,
      "f1": 0.8995
    }
  },
  "timestamp": "2026-04-05T04:56:05.629306"
}