comprehensive_results_binary.json

{
  "val": {
    "loss": 0.3197322487831116,
    "accuracy": 0.9238754325259516,
    "precision": 0.910958904109589,
    "recall": 0.9366197183098591,
    "f1": 0.9236111111111112,
    "macro_f1": 0.923874521072797,
    "roc_auc": 0.9701782121299224,
    "non_gas_precision": 0.9370629370629371,
    "non_gas_recall": 0.9115646258503401,
    "non_gas_f1": 0.9241379310344827,
    "runtime": 12.9943,
    "samples_per_second": 22.241,
    "steps_per_second": 1.462,
    "confusion_matrix": [
      [
        134,
        13
      ],
      [
        9,
        133
      ]
    ]
  },
  "test_id": {
    "loss": 0.336225688457489,
    "accuracy": 0.9065743944636678,
    "precision": 0.9020979020979021,
    "recall": 0.9084507042253521,
    "f1": 0.9052631578947369,
    "macro_f1": 0.9065564936231363,
    "roc_auc": 0.9741304972693302,
    "non_gas_precision": 0.910958904109589,
    "non_gas_recall": 0.9047619047619048,
    "non_gas_f1": 0.9078498293515358,
    "runtime": 13.0057,
    "samples_per_second": 22.221,
    "steps_per_second": 1.461,
    "confusion_matrix": [
      [
        133,
        14
      ],
      [
        13,
        129
      ]
    ]
  },
  "test_ood": {
    "loss": 0.5277629494667053,
    "accuracy": 0.8695652173913043,
    "precision": 0.994579945799458,
    "recall": 0.8398169336384439,
    "f1": 0.9106699751861043,
    "macro_f1": 0.8345296184655353,
    "roc_auc": 0.9881404835339767,
    "non_gas_precision": 0.6174863387978142,
    "non_gas_recall": 0.9826086956521739,
    "non_gas_f1": 0.7583892617449665,
    "runtime": 13.2146,
    "samples_per_second": 41.772,
    "steps_per_second": 2.649,
    "confusion_matrix": [
      [
        113,
        2
      ],
      [
        70,
        367
      ]
    ]
  },
  "delta_f1": -0.005406817291367383,
  "delta_f1_interpretation": "\u2705 Robust generalization (minimal domain sensitivity)"
}