evaluation/evaluation_metrics.json

{
  "overall": {
    "accuracy": 0.14,
    "precision_macro": 0.33,
    "recall_macro": 0.78,
    "f1_macro": 0.46
  },
  "per_class": {
    "precision": {
      "math": 0.42,
      "graphs": 0.4,
      "strings": 0.41,
      "number theory": 0.16,
      "trees": 0.45,
      "geometry": 0.25,
      "games": 0.27,
      "probabilities": 0.1,
      "other": 0.54
    },
    "recall": {
      "math": 0.74,
      "graphs": 0.75,
      "strings": 0.93,
      "number theory": 0.76,
      "trees": 0.79,
      "geometry": 0.84,
      "games": 0.8,
      "probabilities": 0.47,
      "other": 0.95
    },
    "f1": {
      "math": 0.53,
      "graphs": 0.52,
      "strings": 0.56,
      "number theory": 0.27,
      "trees": 0.57,
      "geometry": 0.39,
      "games": 0.4,
      "probabilities": 0.16,
      "other": 0.68
    }
  }
}