eval/eval_summary.json

{
  "checkpoint": "./ppo_checkpoints/policy_update_00976.pt",
  "trained_at_update": 976,
  "n_episodes": 40,
  "controllers": [
    "ppo",
    "scripted",
    "random"
  ],
  "summaries": [
    {
      "controller": "ppo",
      "n_episodes": 40,
      "mean_collisions": 0.075,
      "mean_intent_rate": 0.983,
      "mean_noop_rate": 0.0,
      "mean_stop_rate": 0.0,
      "mean_min_depth": 0.831,
      "mean_reward": 151.649,
      "total_collisions": 3,
      "collision_free_episodes": 37
    },
    {
      "controller": "scripted",
      "n_episodes": 40,
      "mean_collisions": 1.375,
      "mean_intent_rate": 0.498,
      "mean_noop_rate": 0.469,
      "mean_stop_rate": 0.0,
      "mean_min_depth": 0.546,
      "mean_reward": 38.206,
      "total_collisions": 55,
      "collision_free_episodes": 32
    },
    {
      "controller": "random",
      "n_episodes": 40,
      "mean_collisions": 0.175,
      "mean_intent_rate": 0.144,
      "mean_noop_rate": 0.132,
      "mean_stop_rate": 0.51,
      "mean_min_depth": 0.761,
      "mean_reward": -1.275,
      "total_collisions": 7,
      "collision_free_episodes": 35
    }
  ],
  "elapsed_s": 7.6
}