aggregate_metrics.json

{
  "dataset": "ViFN",
  "dataset_rows": 1406,
  "split_counts": {
    "train": 1124,
    "validation": 141,
    "test": 141
  },
  "split_seed": 42,
  "source_sha256": "80ca7eab9383bb0487e64c857b81956f3abd5147aaa19888d4ca8cdff7fd3a50",
  "canonical_data_sha256": "dcf674944d5a4c323f1817ff33c088ff41e427b26b9892e31cae70c077f6e5f0",
  "base_model": "bert-base-multilingual-cased",
  "base_revision": "3f076fdb1ab68d5b2880cb87a0886f315b8146f8",
  "fine_tuning_seeds": [
    42,
    22,
    202
  ],
  "representative_seed": 42,
  "selection_rule": "highest development Macro-F1; test metrics are not used for selection",
  "aggregate_metrics": {
    "dev_accuracy": {
      "mean": 0.8486997635933807,
      "std": 0.014763588648695997,
      "values": [
        0.8368794326241135,
        0.8652482269503546,
        0.8439716312056738
      ]
    },
    "dev_macro_precision": {
      "mean": 0.8529317119960395,
      "std": 0.013506402205727543,
      "values": [
        0.8403540903540904,
        0.8672064777327935,
        0.8512345679012345
      ]
    },
    "dev_macro_recall": {
      "mean": 0.8483232729711604,
      "std": 0.014845199869479062,
      "values": [
        0.8365191146881288,
        0.864989939637827,
        0.8434607645875252
      ]
    },
    "dev_macro_f1": {
      "mean": 0.8481242000514174,
      "std": 0.014993029393505908,
      "values": [
        0.8363526265327749,
        0.8650037792894936,
        0.8430161943319838
      ]
    },
    "dev_weighted_f1": {
      "mean": 0.8481886833633626,
      "std": 0.014975323748613392,
      "values": [
        0.8364184772941922,
        0.8650445205663038,
        0.843103052229592
      ]
    },
    "test_accuracy": {
      "mean": 0.8250591016548463,
      "std": 0.008189365520420235,
      "values": [
        0.8156028368794326,
        0.8297872340425532,
        0.8297872340425532
      ]
    },
    "test_macro_precision": {
      "mean": 0.829765076745669,
      "std": 0.010976776319693734,
      "values": [
        0.8178137651821862,
        0.8320850202429151,
        0.8393964448119058
      ]
    },
    "test_macro_recall": {
      "mean": 0.8254527162977867,
      "std": 0.008278278300672315,
      "values": [
        0.8158953722334004,
        0.8300804828973842,
        0.8303822937625754
      ]
    },
    "test_macro_f1": {
      "mean": 0.8245628443823872,
      "std": 0.007971415110036854,
      "values": [
        0.8153706688154714,
        0.8295729250604351,
        0.828744939271255
      ]
    },
    "test_weighted_f1": {
      "mean": 0.8245014945559426,
      "std": 0.007959906612465843,
      "values": [
        0.8153242352026792,
        0.8295300632640115,
        0.8286501852011371
      ]
    }
  },
  "training": {
    "epochs": 3,
    "learning_rate": 2e-05,
    "weight_decay": 0.01,
    "warmup_ratio": 0.1,
    "effective_batch_size": 8,
    "max_length": 256,
    "text_mode": "raw",
    "class_weighting": false,
    "resampling": false
  },
  "environment": {
    "python": "3.12.13",
    "execution_mode": "full"
  },
  "created_at_utc": "2026-07-18T18:01:08.422406+00:00"
}