ppo_v83/policy_update235.state.json

{
  "checkpoint_schema_version": 2,
  "hidden_reset_policy": "episode_done",
  "update": 236,
  "stage": 5,
  "ent_scale": 4.761904761904762,
  "lr_now": 0.00025,
  "total_env_steps": 519168,
  "recent_wins": [
    0.0,
    1.0,
    0.0,
    1.0,
    1.0,
    1.0,
    0.0,
    0.0,
    0.0,
    0.0,
    1.0,
    1.0,
    0.0,
    1.0,
    1.0,
    0.0,
    1.0,
    0.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    0.0,
    0.0
  ],
  "recent_conversions": [
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0,
    1.0
  ],
  "best_eval_win": null,
  "best_eval_score": null,
  "v81_curriculum": true,
  "curriculum_schedule": "v8.3",
  "reward_profile": "v8.3-closeout-v1",
  "return_stats": null,
  "stage_env_targets": [
    24,
    24,
    24,
    24,
    24,
    24,
    16,
    12,
    10,
    12,
    10,
    8,
    8,
    8,
    8
  ],
  "envs_per_shard": 24,
  "requested_env_target": null
}