message/training_manifest.json

{
  "task": "message",
  "created_at": "20260923T175009Z",
  "python": "3.12.13",
  "platform": "Linux-7.2.6-1-cachyos-x86_64-with-glibc2.44",
  "config": {
    "seed": 42,
    "output_dir": "ml/artifacts",
    "data": {
      "url_dataset": "PhiUSIIL_Phishing_URL_Dataset.csv",
      "phishtank_dataset": "verified_online.csv",
      "benign_domains": "top-1m.csv",
      "message_dataset": "Dataset_5971.csv",
      "sms_collection": "SMSSpamCollection",
      "transaction_n_sample": 1000000,
      "transaction_cache": "ml/.txn-cache"
    },
    "split": {
      "train": 0.7,
      "validation": 0.15,
      "test": 0.15
    },
    "model": {
      "n_estimators": 300,
      "max_depth": 5,
      "learning_rate": 0.04,
      "min_child_weight": 3,
      "subsample": 0.85,
      "colsample_bytree": 0.85,
      "reg_alpha": 0.2,
      "reg_lambda": 2.0,
      "tree_method": "hist"
    },
    "transaction_model": {
      "n_estimators": 200,
      "max_depth": 6,
      "learning_rate": 0.1,
      "min_child_weight": 3,
      "subsample": 0.9,
      "colsample_bytree": 0.9,
      "reg_alpha": 0.2,
      "reg_lambda": 2.0,
      "tree_method": "hist"
    }
  }
}