comprehensive_results_binary.json

{
  "val": {
    "loss": 0.32649385929107666,
    "accuracy": 0.9307958477508651,
    "precision": 0.9485294117647058,
    "recall": 0.9084507042253521,
    "f1": 0.9280575539568345,
    "macro_f1": 0.9306954436450839,
    "roc_auc": 0.9782504551116221,
    "non_gas_precision": 0.9150326797385621,
    "non_gas_recall": 0.9523809523809523,
    "non_gas_f1": 0.9333333333333333,
    "runtime": 12.4882,
    "samples_per_second": 23.142,
    "steps_per_second": 1.521,
    "confusion_matrix": [
      [
        140,
        7
      ],
      [
        13,
        129
      ]
    ]
  },
  "test_id": {
    "loss": 0.25096753239631653,
    "accuracy": 0.9307958477508651,
    "precision": 0.9357142857142857,
    "recall": 0.9225352112676056,
    "f1": 0.9290780141843972,
    "macro_f1": 0.9307552233084149,
    "roc_auc": 0.9802146210596915,
    "non_gas_precision": 0.9261744966442953,
    "non_gas_recall": 0.9387755102040817,
    "non_gas_f1": 0.9324324324324325,
    "runtime": 12.5304,
    "samples_per_second": 23.064,
    "steps_per_second": 1.516,
    "confusion_matrix": [
      [
        138,
        9
      ],
      [
        11,
        131
      ]
    ]
  },
  "test_ood": {
    "loss": 0.6652491688728333,
    "accuracy": 0.8442028985507246,
    "precision": 0.9915966386554622,
    "recall": 0.8100686498855835,
    "f1": 0.8916876574307305,
    "macro_f1": 0.8071341512960104,
    "roc_auc": 0.9828275793453388,
    "non_gas_precision": 0.5743589743589743,
    "non_gas_recall": 0.9739130434782609,
    "non_gas_f1": 0.7225806451612903,
    "runtime": 12.7439,
    "samples_per_second": 43.315,
    "steps_per_second": 2.746,
    "confusion_matrix": [
      [
        112,
        3
      ],
      [
        83,
        354
      ]
    ]
  },
  "delta_f1": 0.03739035675366664,
  "delta_f1_interpretation": "\u2705 Robust generalization (minimal domain sensitivity)"
}