eval_metrics.json

{
  "val": {
    "seqeval_f1": 0.9025009061254077,
    "exact_match": {
      "precision": 0.8688524590163934,
      "recall": 0.9149198520345253,
      "f1": 0.8912912912912914,
      "tp": 742,
      "fp": 112,
      "fn": 69
    },
    "partial_match": {
      "precision": 0.9379391100702577,
      "recall": 0.9876695437731196,
      "f1": 0.9621621621621621,
      "partial_tp": 59
    }
  },
  "test": {
    "seqeval_f1": 0.9124124124124123,
    "exact_match": {
      "precision": 0.8814872192099148,
      "recall": 0.9222042139384117,
      "f1": 0.9013861386138614,
      "tp": 1138,
      "fp": 153,
      "fn": 96
    },
    "partial_match": {
      "precision": 0.9550735863671572,
      "recall": 0.9991896272285251,
      "f1": 0.9766336633663366,
      "partial_tp": 95
    }
  },
  "ood": {
    "seqeval_f1": 0.6672992977794648,
    "exact_match": {
      "precision": 0.658359621451104,
      "recall": 0.7204004142216086,
      "f1": 0.6879841766935884,
      "tp": 2087,
      "fp": 1083,
      "fn": 810
    },
    "partial_match": {
      "precision": 0.8290220820189275,
      "recall": 0.90714532274767,
      "f1": 0.8663260260425252,
      "partial_tp": 541
    }
  }
}