test_result.json

{
  "baseline_scores": {
    "task_easy": 0.85,
    "task_medium": 0.85,
    "task_hard": 0.70,
    "overall_avg": 0.80
  },
  "mode": "llm",
  "model": "meta-llama/Llama-3.3-70B-Instruct",
  "runtime_seconds": 4.36,
  "detail": [
    {
      "task_id": "task_easy",
      "reward": 0.85,
      "mode": "llm",
      "done": false
    },
    {
      "task_id": "task_medium",
      "reward": 0.85,
      "mode": "llm",
      "done": false
    },
    {
      "task_id": "task_hard",
      "reward": 0.70,
      "mode": "llm",
      "done": false
    }
  ]
}