{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.05518763796909492,
  "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": 0.0009054738155100494,
      "clip_ratio/high_mean": 0.00035781565529759975,
      "clip_ratio/low_mean": 0.00016371723540942185,
      "clip_ratio/low_min": 1.144509224104695e-05,
      "clip_ratio/region_mean": 0.0005215328827034682,
      "completions/clipped_ratio": 0.4879464626312256,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1021.8,
      "completions/mean_length": 802.8379760742188,
      "completions/mean_terminated_length": 591.0183349609375,
      "completions/min_length": 16.6,
      "completions/min_terminated_length": 16.6,
      "entropy": 0.6993172228336334,
      "epoch": 0.005518763796909493,
      "frac_reward_zero_std": 0.648214316368103,
      "grad_norm": 0.123046875,
      "kl": 0.0009140914742602035,
      "learning_rate": 9.994100796397954e-06,
      "loss": 0.0096,
      "num_tokens": 2134981.0,
      "reward": 0.17656995058059693,
      "reward_std": 0.3300946533679962,
      "rewards/reward_fn/mean": 0.17656994760036468,
      "rewards/reward_fn/std": 0.3300946354866028,
      "step": 10,
      "step_time": 62.725463995942846
    },
    {
      "clip_ratio/high_max": 0.0009650078689446672,
      "clip_ratio/high_mean": 0.00040530616533942523,
      "clip_ratio/low_mean": 0.00019620073144324123,
      "clip_ratio/low_min": 2.9645057293237186e-05,
      "clip_ratio/region_mean": 0.0006015069084241986,
      "completions/clipped_ratio": 0.5308036029338836,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1019.4,
      "completions/mean_length": 814.8719116210938,
      "completions/mean_terminated_length": 578.806640625,
      "completions/min_length": 10.2,
      "completions/min_terminated_length": 10.2,
      "entropy": 0.7009260475635528,
      "epoch": 0.011037527593818985,
      "frac_reward_zero_std": 0.6464286088943482,
      "grad_norm": 0.115234375,
      "kl": 0.0012107100861612707,
      "learning_rate": 9.889612861977855e-06,
      "loss": 0.009,
      "num_tokens": 4322090.0,
      "reward": 0.16465030908584594,
      "reward_std": 0.31232678294181826,
      "rewards/reward_fn/mean": 0.16465030312538148,
      "rewards/reward_fn/std": 0.31232677698135375,
      "step": 20,
      "step_time": 59.907306081056596
    },
    {
      "clip_ratio/high_max": 0.0007639058487256989,
      "clip_ratio/high_mean": 0.00030917047697585074,
      "clip_ratio/low_mean": 0.00016517131589353085,
      "clip_ratio/low_min": 3.0255078672780656e-05,
      "clip_ratio/region_mean": 0.0004743418045109138,
      "completions/clipped_ratio": 0.49821431636810304,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1017.6,
      "completions/mean_length": 810.8062744140625,
      "completions/mean_terminated_length": 601.0194946289063,
      "completions/min_length": 18.2,
      "completions/min_terminated_length": 18.2,
      "entropy": 0.707157039642334,
      "epoch": 0.016556291390728478,
      "frac_reward_zero_std": 0.6964286088943481,
      "grad_norm": 0.08984375,
      "kl": 0.0015323497413191945,
      "learning_rate": 9.657180469054213e-06,
      "loss": 0.0056,
      "num_tokens": 6483648.0,
      "reward": 0.15746280252933503,
      "reward_std": 0.31524514555931094,
      "rewards/reward_fn/mean": 0.15746279954910278,
      "rewards/reward_fn/std": 0.315245121717453,
      "step": 30,
      "step_time": 69.35127711733803
    },
    {
      "clip_ratio/high_max": 0.0010893745173234492,
      "clip_ratio/high_mean": 0.0004552829515887424,
      "clip_ratio/low_mean": 0.00020492849944275804,
      "clip_ratio/low_min": 3.608180813898798e-05,
      "clip_ratio/region_mean": 0.0006602114357519895,
      "completions/clipped_ratio": 0.4924107313156128,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1022.4,
      "completions/mean_length": 801.3585205078125,
      "completions/mean_terminated_length": 584.6675537109375,
      "completions/min_length": 28.6,
      "completions/min_terminated_length": 28.6,
      "entropy": 0.7392709702253342,
      "epoch": 0.02207505518763797,
      "frac_reward_zero_std": 0.650000023841858,
      "grad_norm": 0.10205078125,
      "kl": 0.001778936560731381,
      "learning_rate": 9.302885579019626e-06,
      "loss": 0.0102,
      "num_tokens": 8664175.0,
      "reward": 0.1526116132736206,
      "reward_std": 0.3074562013149261,
      "rewards/reward_fn/mean": 0.15261160731315612,
      "rewards/reward_fn/std": 0.3074561834335327,
      "step": 40,
      "step_time": 66.20743731297553
    },
    {
      "clip_ratio/high_max": 0.0009445919014979154,
      "clip_ratio/high_mean": 0.00039524411113234237,
      "clip_ratio/low_mean": 0.00019868952222168446,
      "clip_ratio/low_min": 6.7668152041733265e-06,
      "clip_ratio/region_mean": 0.0005939336377196013,
      "completions/clipped_ratio": 0.4633928775787354,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1021.6,
      "completions/mean_length": 785.5151977539062,
      "completions/mean_terminated_length": 579.9617919921875,
      "completions/min_length": 17.2,
      "completions/min_terminated_length": 17.2,
      "entropy": 0.7299870848655701,
      "epoch": 0.02759381898454746,
      "frac_reward_zero_std": 0.6625000357627868,
      "grad_norm": 0.13671875,
      "kl": 0.0018997497216332703,
      "learning_rate": 8.83599887835493e-06,
      "loss": 0.0106,
      "num_tokens": 10787441.0,
      "reward": 0.17087055444717408,
      "reward_std": 0.3202264249324799,
      "rewards/reward_fn/mean": 0.17087054550647734,
      "rewards/reward_fn/std": 0.3202264130115509,
      "step": 50,
      "step_time": 61.80385860861279
    },
    {
      "clip_ratio/high_max": 0.0008307943062391132,
      "clip_ratio/high_mean": 0.00035787131600955034,
      "clip_ratio/low_mean": 0.00023181447832030245,
      "clip_ratio/low_min": 2.61040790064726e-05,
      "clip_ratio/region_mean": 0.0005896858012420126,
      "completions/clipped_ratio": 0.4741071581840515,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1011.2,
      "completions/mean_length": 792.1500244140625,
      "completions/mean_terminated_length": 583.1341064453125,
      "completions/min_length": 16.8,
      "completions/min_terminated_length": 16.8,
      "entropy": 0.8175096482038497,
      "epoch": 0.033112582781456956,
      "frac_reward_zero_std": 0.6428571581840515,
      "grad_norm": 0.1142578125,
      "kl": 0.0020997423096559944,
      "learning_rate": 8.268737196446264e-06,
      "loss": 0.01,
      "num_tokens": 12937373.0,
      "reward": 0.170543172955513,
      "reward_std": 0.30973091125488283,
      "rewards/reward_fn/mean": 0.17054316401481628,
      "rewards/reward_fn/std": 0.3097309172153473,
      "step": 60,
      "step_time": 81.64577775509096
    },
    {
      "clip_ratio/high_max": 0.0010142354993149638,
      "clip_ratio/high_mean": 0.0004649270667869132,
      "clip_ratio/low_mean": 0.0002056810655631125,
      "clip_ratio/low_min": 2.441328761051409e-05,
      "clip_ratio/region_mean": 0.0006706080996082165,
      "completions/clipped_ratio": 0.4513392984867096,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1021.2,
      "completions/mean_length": 780.0888671875,
      "completions/mean_terminated_length": 579.7443481445313,
      "completions/min_length": 17.6,
      "completions/min_terminated_length": 17.6,
      "entropy": 0.8121535569429398,
      "epoch": 0.03863134657836645,
      "frac_reward_zero_std": 0.6232143044471741,
      "grad_norm": 0.1298828125,
      "kl": 0.002086066803894937,
      "learning_rate": 7.61594383291065e-06,
      "loss": 0.0068,
      "num_tokens": 15043532.0,
      "reward": 0.1762128084897995,
      "reward_std": 0.32508295178413393,
      "rewards/reward_fn/mean": 0.17621279656887054,
      "rewards/reward_fn/std": 0.3250829458236694,
      "step": 70,
      "step_time": 67.39459452703595
    },
    {
      "clip_ratio/high_max": 0.0011546267720405012,
      "clip_ratio/high_mean": 0.0004448194318683818,
      "clip_ratio/low_mean": 0.00022613786859437823,
      "clip_ratio/low_min": 3.512381881591864e-05,
      "clip_ratio/region_mean": 0.0006709573091939092,
      "completions/clipped_ratio": 0.45714287757873534,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1017.6,
      "completions/mean_length": 780.5415405273437,
      "completions/mean_terminated_length": 574.924560546875,
      "completions/min_length": 13.4,
      "completions/min_terminated_length": 13.4,
      "entropy": 0.8120243638753891,
      "epoch": 0.04415011037527594,
      "frac_reward_zero_std": 0.6267857432365418,
      "grad_norm": 0.125,
      "kl": 0.0022045036195777355,
      "learning_rate": 6.894700159171535e-06,
      "loss": 0.0074,
      "num_tokens": 17150985.0,
      "reward": 0.18558780550956727,
      "reward_std": 0.3320887923240662,
      "rewards/reward_fn/mean": 0.1855877935886383,
      "rewards/reward_fn/std": 0.3320887744426727,
      "step": 80,
      "step_time": 69.55923435157165
    },
    {
      "clip_ratio/high_max": 0.0010666975751519204,
      "clip_ratio/high_mean": 0.0004059886821778491,
      "clip_ratio/low_mean": 0.00018196528690168635,
      "clip_ratio/low_min": 1.4905737771186978e-05,
      "clip_ratio/region_mean": 0.0005879539618035779,
      "completions/clipped_ratio": 0.4683035910129547,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1018.8,
      "completions/mean_length": 786.5406616210937,
      "completions/mean_terminated_length": 577.4362426757813,
      "completions/min_length": 18.8,
      "completions/min_terminated_length": 18.8,
      "entropy": 0.8206083297729492,
      "epoch": 0.04966887417218543,
      "frac_reward_zero_std": 0.6660714626312256,
      "grad_norm": 0.107421875,
      "kl": 0.0022262637387029825,
      "learning_rate": 6.123878657343648e-06,
      "loss": 0.0095,
      "num_tokens": 19265236.0,
      "reward": 0.16312500536441804,
      "reward_std": 0.3078223645687103,
      "rewards/reward_fn/mean": 0.16312500536441804,
      "rewards/reward_fn/std": 0.30782235860824586,
      "step": 90,
      "step_time": 63.29917922359891
    },
    {
      "clip_ratio/high_max": 0.001015995239140466,
      "clip_ratio/high_mean": 0.0005060771596617997,
      "clip_ratio/low_mean": 0.000221432710532099,
      "clip_ratio/low_min": 4.502671981754247e-05,
      "clip_ratio/region_mean": 0.0007275098643731326,
      "completions/clipped_ratio": 0.45044645071029665,
      "completions/max_length": 1024.0,
      "completions/max_terminated_length": 1021.4,
      "completions/mean_length": 771.9370849609375,
      "completions/mean_terminated_length": 565.00205078125,
      "completions/min_length": 17.0,
      "completions/min_terminated_length": 17.0,
      "entropy": 0.8362967848777771,
      "epoch": 0.05518763796909492,
      "frac_reward_zero_std": 0.6035714626312256,
      "grad_norm": 0.09912109375,
      "kl": 0.0022062973876018077,
      "learning_rate": 5.323649091872179e-06,
      "loss": 0.0076,
      "num_tokens": 21324495.0,
      "reward": 0.19067709147930145,
      "reward_std": 0.32638007402420044,
      "rewards/reward_fn/mean": 0.19067708551883697,
      "rewards/reward_fn/std": 0.32638007402420044,
      "step": 100,
      "step_time": 47.039065402187404
    }
  ],
  "logging_steps": 10,
  "max_steps": 200,
  "num_input_tokens_seen": 21324495,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}