{
  "base_adapter": "/tmp/atomicvision_publish_runner/output/base_adapter",
  "candidates": {
    "base": {
      "hard": {
        "done": 1.0,
        "f1": 0.81621578125,
        "fail": 0.0,
        "mae": 0.02591959375,
        "normalized": 1.0,
        "reward": 4.69171921875,
        "strict": 1.0,
        "submit": 1.0
      },
      "hard_f1_delta_vs_base": 0.0,
      "hard_reward_delta_vs_base": 0.0,
      "label": "base",
      "medium": {
        "done": 1.0,
        "f1": 0.789137,
        "fail": 0.0,
        "mae": 0.027124218749999998,
        "normalized": 1.0,
        "reward": 4.50648265625,
        "strict": 1.0,
        "submit": 1.0
      },
      "medium_f1_delta_vs_base": 0.0,
      "medium_reward_delta_vs_base": 0.0
    },
    "checkpoint-1": {
      "hard": {
        "done": 1.0,
        "f1": 0.8206800937500001,
        "fail": 0.0,
        "mae": 0.02552296875,
        "normalized": 1.0,
        "reward": 4.714775875,
        "strict": 1.0,
        "submit": 1.0
      },
      "hard_f1_delta_vs_base": 0.004464312500000012,
      "hard_reward_delta_vs_base": 0.02305665624999964,
      "label": "checkpoint-1",
      "medium": {
        "done": 1.0,
        "f1": 0.789137,
        "fail": 0.0,
        "mae": 0.027124218749999998,
        "normalized": 1.0,
        "reward": 4.50648265625,
        "strict": 1.0,
        "submit": 1.0
      },
      "medium_f1_delta_vs_base": 0.0,
      "medium_reward_delta_vs_base": 0.0
    }
  },
  "checkpoint_steps": [
    1
  ],
  "dataset_counts": {
    "submit_after_reference": 12,
    "submit_prior": 4
  },
  "episodes_per_difficulty": 16,
  "eval_difficulties": [
    "medium",
    "hard"
  ],
  "eval_episodes": 32,
  "eval_seed_start": 10000,
  "learning_rate": 1e-06,
  "max_updates": 1,
  "profile": "hard_recall_micro_repair",
  "promotion_candidate": "checkpoint-1",
  "seed_start": 3600,
  "train_difficulties": [
    "hard"
  ]
}