{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9893390191897654,
  "eval_steps": 100,
  "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": 608.1997985839844,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.4826625768130922,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0276,
      "reward": 0.6250000298023224,
      "reward_std": 0.3469353225082159,
      "rewards/accuracy_reward": 0.6250000298023224,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.1880884170532,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.7824728423768934,
      "kl": 0.00032770633697509766,
      "learning_rate": 2.5e-06,
      "loss": 0.0222,
      "reward": 0.6166294906288385,
      "reward_std": 0.35452946182340384,
      "rewards/accuracy_reward": 0.6166294906288385,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.6335121154785,
      "epoch": 0.17057569296375266,
      "grad_norm": 4602.5686953405275,
      "kl": 0.2378152847290039,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0859,
      "reward": 0.6770089596509934,
      "reward_std": 0.31036885008215903,
      "rewards/accuracy_reward": 0.6770089596509934,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.3538200378418,
      "epoch": 0.255863539445629,
      "grad_norm": 0.30519508536223,
      "kl": 0.055579757690429686,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0695,
      "reward": 0.7611607506871223,
      "reward_std": 0.22096406035125254,
      "rewards/accuracy_reward": 0.7611607506871223,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.3939994812011,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.15882451560463448,
      "kl": 0.0035945892333984373,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0507,
      "reward": 0.7741071805357933,
      "reward_std": 0.2031545776873827,
      "rewards/accuracy_reward": 0.7741071805357933,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.604940032959,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.16805967209081119,
      "kl": 0.004913711547851562,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0412,
      "reward": 0.7656250342726707,
      "reward_std": 0.1831065765582025,
      "rewards/accuracy_reward": 0.7656250342726707,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 598.2044891357422,
      "epoch": 0.511727078891258,
      "grad_norm": 0.14888598857678317,
      "kl": 0.00452728271484375,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0398,
      "reward": 0.7508928909897804,
      "reward_std": 0.17886008992791175,
      "rewards/accuracy_reward": 0.7508928909897804,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 588.5198966979981,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.1519392828021144,
      "kl": 0.005135726928710937,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0301,
      "reward": 0.7504464626312256,
      "reward_std": 0.16700280252844096,
      "rewards/accuracy_reward": 0.7504464626312256,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 575.4761436462402,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.10504938622311973,
      "kl": 0.004594039916992187,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0386,
      "reward": 0.7741071790456772,
      "reward_std": 0.1698190384544432,
      "rewards/accuracy_reward": 0.7741071790456772,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.2636466979981,
      "epoch": 0.767590618336887,
      "grad_norm": 0.10062535580724605,
      "kl": 0.00447540283203125,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0291,
      "reward": 0.7636161088943482,
      "reward_std": 0.17512967977672816,
      "rewards/accuracy_reward": 0.7636161088943482,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 594.1404296875,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.13619152463583725,
      "kl": 0.004338836669921875,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0318,
      "reward": 0.7573661044239998,
      "reward_std": 0.1743252847343683,
      "rewards/accuracy_reward": 0.7573661044239998,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 588.1810516357422,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.15370397597420127,
      "kl": 0.00505828857421875,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0362,
      "reward": 0.7852678924798966,
      "reward_std": 0.17964433841407298,
      "rewards/accuracy_reward": 0.7852678924798966,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 574.9235763549805,
      "epoch": 0.9893390191897654,
      "kl": 0.006348927815755208,
      "reward": 0.7786458656191826,
      "reward_std": 0.15960007222990194,
      "rewards/accuracy_reward": 0.7786458656191826,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.04329907550508606,
      "train_runtime": 7330.3865,
      "train_samples_per_second": 1.023,
      "train_steps_per_second": 0.008
    }
  ],
  "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
}