{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9893390191897654,
  "eval_steps": 500,
  "global_step": 58,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 395.3694381713867,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.25610095262527466,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.1043,
      "reward": 1.9107143729925156,
      "reward_std": 0.3902299851179123,
      "rewards/accuracy_reward": 0.6462053805589676,
      "rewards/compress_reward": 0.3247768059372902,
      "rewards/format_reward": 0.9397321790456772,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 398.57200622558594,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.3123304545879364,
      "kl": 0.00016838312149047852,
      "learning_rate": 2.5e-06,
      "loss": 0.1161,
      "reward": 1.8558873608708382,
      "reward_std": 0.4281642036512494,
      "rewards/accuracy_reward": 0.5892857434228063,
      "rewards/compress_reward": 0.3148716646246612,
      "rewards/format_reward": 0.9517299514263868,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 367.61095199584963,
      "epoch": 0.17057569296375266,
      "grad_norm": 2.265721321105957,
      "kl": 0.015511131286621094,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.1141,
      "reward": 1.9002233117818832,
      "reward_std": 0.40832622945308683,
      "rewards/accuracy_reward": 0.5988839618861675,
      "rewards/compress_reward": 0.3475446566939354,
      "rewards/format_reward": 0.9537946879863739,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 313.22568435668944,
      "epoch": 0.255863539445629,
      "grad_norm": 0.40780770778656006,
      "kl": 0.0651092529296875,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0857,
      "reward": 1.9601563423871995,
      "reward_std": 0.39528256431221964,
      "rewards/accuracy_reward": 0.5776785969734192,
      "rewards/compress_reward": 0.4090401969850063,
      "rewards/format_reward": 0.9734375372529029,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 286.3823780059814,
      "epoch": 0.3411513859275053,
      "grad_norm": 4.390214443206787,
      "kl": 0.06104736328125,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0709,
      "reward": 2.0002232909202577,
      "reward_std": 0.36102210320532324,
      "rewards/accuracy_reward": 0.5935268148779869,
      "rewards/compress_reward": 0.4294643089175224,
      "rewards/format_reward": 0.9772321805357933,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 293.0426471710205,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.9113842844963074,
      "kl": 0.069866943359375,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0508,
      "reward": 2.032812610268593,
      "reward_std": 0.3311908446252346,
      "rewards/accuracy_reward": 0.5979910969734192,
      "rewards/compress_reward": 0.4497768051922321,
      "rewards/format_reward": 0.985044677555561,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 315.613631439209,
      "epoch": 0.511727078891258,
      "grad_norm": 1.5531269311904907,
      "kl": 0.064349365234375,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0379,
      "reward": 2.0685268968343733,
      "reward_std": 0.2967603411525488,
      "rewards/accuracy_reward": 0.6350446686148643,
      "rewards/compress_reward": 0.442857164144516,
      "rewards/format_reward": 0.990625025331974,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 313.02188873291016,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.4903877377510071,
      "kl": 0.056243896484375,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0389,
      "reward": 2.054687598347664,
      "reward_std": 0.31466612182557585,
      "rewards/accuracy_reward": 0.6426339596509933,
      "rewards/compress_reward": 0.428348234295845,
      "rewards/format_reward": 0.983705396950245,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 295.896439743042,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.32720160484313965,
      "kl": 0.061297607421875,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0381,
      "reward": 2.0698661684989927,
      "reward_std": 0.29936849512159824,
      "rewards/accuracy_reward": 0.6511161006987095,
      "rewards/compress_reward": 0.43348216339945794,
      "rewards/format_reward": 0.9852678909897804,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 294.11586112976073,
      "epoch": 0.767590618336887,
      "grad_norm": 0.4028303623199463,
      "kl": 0.06295166015625,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0326,
      "reward": 2.073660808801651,
      "reward_std": 0.3071202915161848,
      "rewards/accuracy_reward": 0.6421875312924386,
      "rewards/compress_reward": 0.4457589529454708,
      "rewards/format_reward": 0.9857143208384513,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 289.0868427276611,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.41271042823791504,
      "kl": 0.06861572265625,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0452,
      "reward": 2.044196516275406,
      "reward_std": 0.32505202740430833,
      "rewards/accuracy_reward": 0.6229910977184773,
      "rewards/compress_reward": 0.44084823578596116,
      "rewards/format_reward": 0.9803571760654449,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 286.33059158325193,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.5375810861587524,
      "kl": 0.07474365234375,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0347,
      "reward": 2.0625000923871992,
      "reward_std": 0.34456842839717866,
      "rewards/accuracy_reward": 0.6479911044239998,
      "rewards/compress_reward": 0.4363839469850063,
      "rewards/format_reward": 0.9781250387430191,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 274.45362917582196,
      "epoch": 0.9893390191897654,
      "kl": 0.074554443359375,
      "reward": 2.0604539712270102,
      "reward_std": 0.31418706725041073,
      "rewards/accuracy_reward": 0.6287202710906664,
      "rewards/compress_reward": 0.4499628196159999,
      "rewards/format_reward": 0.98177087555329,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.05939787125279163,
      "train_runtime": 12451.444,
      "train_samples_per_second": 0.602,
      "train_steps_per_second": 0.005
    }
  ],
  "logging_steps": 5,
  "max_steps": 58,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}