{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.25,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 1.6,
      "completions/max_terminated_length": 1.6,
      "completions/mean_length": 1.15,
      "completions/mean_terminated_length": 1.15,
      "completions/min_length": 1.0,
      "completions/min_terminated_length": 1.0,
      "entropy": 0.44410082772374154,
      "epoch": 0.025,
      "frac_reward_zero_std": 1.0,
      "grad_norm": 0.0,
      "learning_rate": 9.1e-07,
      "loss": 0.0,
      "num_tokens": 3586.0,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/reward_helper/mean": 0.0,
      "rewards/reward_helper/std": 0.0,
      "sampling/importance_sampling_ratio/max": 1.0006482005119324,
      "sampling/importance_sampling_ratio/mean": 0.9993368148803711,
      "sampling/importance_sampling_ratio/min": 0.9954026460647583,
      "sampling/sampling_logp_difference/max": 0.00986330844461918,
      "sampling/sampling_logp_difference/mean": 0.004519483260810375,
      "step": 10,
      "step_time": 5.278360854899991
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.075,
      "completions/max_length": 52.0,
      "completions/max_terminated_length": 1.0,
      "completions/mean_length": 20.125,
      "completions/mean_terminated_length": 1.0,
      "completions/min_length": 1.0,
      "completions/min_terminated_length": 1.0,
      "entropy": 0.5618356347084046,
      "epoch": 0.05,
      "frac_reward_zero_std": 0.8,
      "grad_norm": 0.0,
      "learning_rate": 8.1e-07,
      "loss": -0.0875,
      "num_tokens": 8279.0,
      "reward": 0.015000000596046448,
      "reward_std": 0.030000001192092896,
      "rewards/reward_helper/mean": 0.015000000596046448,
      "rewards/reward_helper/std": 0.030000004172325134,
      "sampling/importance_sampling_ratio/max": 1.0033257365226746,
      "sampling/importance_sampling_ratio/mean": 0.9744328677654266,
      "sampling/importance_sampling_ratio/min": 0.8951155811548233,
      "sampling/sampling_logp_difference/max": 0.039345417730510236,
      "sampling/sampling_logp_difference/mean": 0.004816015157848597,
      "step": 20,
      "step_time": 1.230997591900001
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.025,
      "completions/max_length": 46.5,
      "completions/max_terminated_length": 21.0,
      "completions/mean_length": 12.525,
      "completions/mean_terminated_length": 6.15,
      "completions/min_length": 1.0,
      "completions/min_terminated_length": 1.0,
      "entropy": 0.5129512194544077,
      "epoch": 0.075,
      "frac_reward_zero_std": 0.9,
      "grad_norm": 0.0,
      "learning_rate": 7.1e-07,
      "loss": -0.0936,
      "num_tokens": 12500.0,
      "reward": 0.007500000298023224,
      "reward_std": 0.015000000596046448,
      "rewards/reward_helper/mean": 0.007500000298023224,
      "rewards/reward_helper/std": 0.015000002086162567,
      "sampling/importance_sampling_ratio/max": 1.0458860635757445,
      "sampling/importance_sampling_ratio/mean": 1.0027949154376983,
      "sampling/importance_sampling_ratio/min": 0.9623779714107513,
      "sampling/sampling_logp_difference/max": 0.040722915157675746,
      "sampling/sampling_logp_difference/mean": 0.006484306254424155,
      "step": 30,
      "step_time": 1.29722429359999
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 80.8,
      "completions/max_terminated_length": 4.3,
      "completions/mean_length": 33.725,
      "completions/mean_terminated_length": 1.85,
      "completions/min_length": 1.0,
      "completions/min_terminated_length": 1.0,
      "entropy": 0.8639558412134647,
      "epoch": 0.1,
      "frac_reward_zero_std": 0.7,
      "grad_norm": 0.0,
      "learning_rate": 6.1e-07,
      "loss": -0.3408,
      "num_tokens": 17421.0,
      "reward": 0.017500000633299352,
      "reward_std": 0.035000001266598704,
      "rewards/reward_helper/mean": 0.017500000633299352,
      "rewards/reward_helper/std": 0.03500000424683094,
      "sampling/importance_sampling_ratio/max": 1.1232336461544037,
      "sampling/importance_sampling_ratio/mean": 1.0079659342765808,
      "sampling/importance_sampling_ratio/min": 0.9237572312355041,
      "sampling/sampling_logp_difference/max": 0.08927474431693554,
      "sampling/sampling_logp_difference/mean": 0.009856639616191386,
      "step": 40,
      "step_time": 1.325083582900001
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.175,
      "completions/max_length": 143.6,
      "completions/max_terminated_length": 55.9,
      "completions/mean_length": 59.35,
      "completions/mean_terminated_length": 15.916666793823243,
      "completions/min_length": 1.0,
      "completions/min_terminated_length": 1.0,
      "entropy": 1.408535435795784,
      "epoch": 0.125,
      "frac_reward_zero_std": 0.6,
      "grad_norm": 0.0,
      "learning_rate": 5.1e-07,
      "loss": -0.5045,
      "num_tokens": 22655.0,
      "reward": 0.025000000931322576,
      "reward_std": 0.05000000186264515,
      "rewards/reward_helper/mean": 0.025000000931322576,
      "rewards/reward_helper/std": 0.05000000633299351,
      "sampling/importance_sampling_ratio/max": 1.2991383910179137,
      "sampling/importance_sampling_ratio/mean": 1.045933473110199,
      "sampling/importance_sampling_ratio/min": 0.9206243932247162,
      "sampling/sampling_logp_difference/max": 0.10195044428110123,
      "sampling/sampling_logp_difference/mean": 0.011015815567225218,
      "step": 50,
      "step_time": 1.4795691685999999
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.15,
      "completions/max_length": 119.2,
      "completions/max_terminated_length": 68.1,
      "completions/mean_length": 57.775,
      "completions/mean_terminated_length": 19.425,
      "completions/min_length": 26.5,
      "completions/min_terminated_length": 0.9,
      "entropy": 1.288734181970358,
      "epoch": 0.15,
      "frac_reward_zero_std": 0.6,
      "grad_norm": 6.278634548187256,
      "learning_rate": 4.0999999999999994e-07,
      "loss": -0.2518,
      "num_tokens": 28242.0,
      "reward": 0.0325000012293458,
      "reward_std": 0.05232050903141498,
      "rewards/reward_helper/mean": 0.0325000012293458,
      "rewards/reward_helper/std": 0.052320512011647224,
      "sampling/importance_sampling_ratio/max": 1.1803278803825379,
      "sampling/importance_sampling_ratio/mean": 1.020902580022812,
      "sampling/importance_sampling_ratio/min": 0.8570977985858917,
      "sampling/sampling_logp_difference/max": 0.08151691369712352,
      "sampling/sampling_logp_difference/mean": 0.010703828372061252,
      "step": 60,
      "step_time": 1.4942365265000093
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.2,
      "completions/max_length": 187.8,
      "completions/max_terminated_length": 80.4,
      "completions/mean_length": 76.725,
      "completions/mean_terminated_length": 33.31666717529297,
      "completions/min_length": 1.0,
      "completions/min_terminated_length": 1.0,
      "entropy": 2.0541629776358605,
      "epoch": 0.175,
      "frac_reward_zero_std": 0.4,
      "grad_norm": 10.47471809387207,
      "learning_rate": 3.1e-07,
      "loss": -0.4418,
      "num_tokens": 35115.0,
      "reward": 0.05500000212341547,
      "reward_std": 0.08464101739227772,
      "rewards/reward_helper/mean": 0.05500000212341547,
      "rewards/reward_helper/std": 0.08464102186262608,
      "sampling/importance_sampling_ratio/max": 1.1099666595458983,
      "sampling/importance_sampling_ratio/mean": 0.9488288760185242,
      "sampling/importance_sampling_ratio/min": 0.7294621288776397,
      "sampling/sampling_logp_difference/max": 0.14803548753261567,
      "sampling/sampling_logp_difference/mean": 0.014868053188547493,
      "step": 70,
      "step_time": 1.6123995632000174
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.5,
      "completions/max_length": 230.5,
      "completions/max_terminated_length": 85.9,
      "completions/mean_length": 149.775,
      "completions/mean_terminated_length": 52.43333396911621,
      "completions/min_length": 83.3,
      "completions/min_terminated_length": 32.1,
      "entropy": 2.11170057579875,
      "epoch": 0.2,
      "frac_reward_zero_std": 0.4,
      "grad_norm": 5.259692668914795,
      "learning_rate": 2.0999999999999997e-07,
      "loss": -0.2111,
      "num_tokens": 43822.0,
      "reward": 0.08250000271946192,
      "reward_std": 0.10280726440250873,
      "rewards/reward_helper/mean": 0.08250000271946192,
      "rewards/reward_helper/std": 0.10280726738274097,
      "sampling/importance_sampling_ratio/max": 1.539601093530655,
      "sampling/importance_sampling_ratio/mean": 1.1134967446327209,
      "sampling/importance_sampling_ratio/min": 0.7393792182207107,
      "sampling/sampling_logp_difference/max": 0.1634996145963669,
      "sampling/sampling_logp_difference/mean": 0.014190209750086068,
      "step": 80,
      "step_time": 1.6749743500999899
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.65,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 76.9,
      "completions/mean_length": 196.8,
      "completions/mean_terminated_length": 67.38333435058594,
      "completions/min_length": 83.5,
      "completions/min_terminated_length": 57.9,
      "entropy": 1.6522414579987525,
      "epoch": 0.225,
      "frac_reward_zero_std": 0.0,
      "grad_norm": 6.119882583618164,
      "learning_rate": 1.0999999999999999e-07,
      "loss": -0.2799,
      "num_tokens": 55182.0,
      "reward": 0.19750000536441803,
      "reward_std": 0.22683281898498536,
      "rewards/reward_helper/mean": 0.19750000536441803,
      "rewards/reward_helper/std": 0.22683282494544982,
      "sampling/importance_sampling_ratio/max": 1.3945529162883759,
      "sampling/importance_sampling_ratio/mean": 0.9089209794998169,
      "sampling/importance_sampling_ratio/min": 0.5628022223711013,
      "sampling/sampling_logp_difference/max": 0.1890378475189209,
      "sampling/sampling_logp_difference/mean": 0.0153791518881917,
      "step": 90,
      "step_time": 1.7072683768000103
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.6,
      "completions/max_length": 255.0,
      "completions/max_terminated_length": 120.7,
      "completions/mean_length": 197.3,
      "completions/mean_terminated_length": 87.23333358764648,
      "completions/min_length": 112.7,
      "completions/min_terminated_length": 61.5,
      "entropy": 1.7242027707397938,
      "epoch": 0.25,
      "frac_reward_zero_std": 0.1,
      "grad_norm": 10.829450607299805,
      "learning_rate": 1e-08,
      "loss": -0.1902,
      "num_tokens": 65950.0,
      "reward": 0.11250000465661288,
      "reward_std": 0.13696152605116368,
      "rewards/reward_helper/mean": 0.11250000465661288,
      "rewards/reward_helper/std": 0.13696153201162814,
      "sampling/importance_sampling_ratio/max": 1.7122498869895935,
      "sampling/importance_sampling_ratio/mean": 1.096574491262436,
      "sampling/importance_sampling_ratio/min": 0.6619945377111435,
      "sampling/sampling_logp_difference/max": 0.15592530369758606,
      "sampling/sampling_logp_difference/mean": 0.015373438503593207,
      "step": 100,
      "step_time": 1.7581579990000022
    }
  ],
  "logging_steps": 10,
  "max_steps": 100,
  "num_input_tokens_seen": 65950,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}