reports/baseline_test_hard_metrics.json

{
  "split": "test_hard",
  "num_samples": 150,
  "evaluation_mode": "pretrained_segface",
  "model_label": "baseline_segface:swin_base",
  "threshold": 0.5,
  "image_size": 512,
  "checkpoint": "",
  "prepared_root": "/workspace/runpod_upload_ready/data/aihub_hairmask_hq_budget_50k_test_only",
  "raw_root": "",
  "model_name": "swin_base",
  "iou": 0.18030471801757814,
  "dice": 0.30552232023678627,
  "precision": 0.18030471801757814,
  "recall": 1.0,
  "mean_iou": 0.18030471801757814,
  "mean_dice": 0.29506855229536694,
  "mean_precision": 0.18030471801757814,
  "mean_recall": 1.0
}