{
  "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": 594.6719055175781,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.38917914032936096,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": -0.0015,
      "reward": 0.5937500223517418,
      "reward_std": 0.36258383840322495,
      "rewards/accuracy_reward": 0.5937500223517418,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.426923751831,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.6033169031143188,
      "kl": 0.00022223591804504395,
      "learning_rate": 2.5e-06,
      "loss": 0.0208,
      "reward": 0.5965402042493224,
      "reward_std": 0.35134637588635087,
      "rewards/accuracy_reward": 0.5965402042493224,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.2098503112793,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.3742579519748688,
      "kl": 0.012548351287841797,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0495,
      "reward": 0.6843750298023223,
      "reward_std": 0.3002105034887791,
      "rewards/accuracy_reward": 0.6843750298023223,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.2727981567383,
      "epoch": 0.255863539445629,
      "grad_norm": 0.1519230604171753,
      "kl": 0.0043125152587890625,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0656,
      "reward": 0.7558036029338837,
      "reward_std": 0.2108498839661479,
      "rewards/accuracy_reward": 0.7558036029338837,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 585.0951133728028,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.16383488476276398,
      "kl": 0.005002593994140625,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0501,
      "reward": 0.7734375342726707,
      "reward_std": 0.18315920066088437,
      "rewards/accuracy_reward": 0.7734375342726707,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 592.6384147644043,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.2767745852470398,
      "kl": 0.005355453491210938,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0484,
      "reward": 0.7578125357627868,
      "reward_std": 0.18339989371597767,
      "rewards/accuracy_reward": 0.7578125357627868,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.6895370483398,
      "epoch": 0.511727078891258,
      "grad_norm": 0.3095102608203888,
      "kl": 0.007829666137695312,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0344,
      "reward": 0.750669676065445,
      "reward_std": 0.18129065297544003,
      "rewards/accuracy_reward": 0.750669676065445,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 594.9719017028808,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.13182398676872253,
      "kl": 0.007412338256835937,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.032,
      "reward": 0.7395089618861675,
      "reward_std": 0.16979889776557683,
      "rewards/accuracy_reward": 0.7395089618861675,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.8926574707032,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.7764875888824463,
      "kl": 0.009193038940429688,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.039,
      "reward": 0.7620536044239998,
      "reward_std": 0.18032782282680274,
      "rewards/accuracy_reward": 0.7620536044239998,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 600.9457855224609,
      "epoch": 0.767590618336887,
      "grad_norm": 0.2261986881494522,
      "kl": 0.009584808349609375,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0393,
      "reward": 0.7482143178582191,
      "reward_std": 0.1897371619939804,
      "rewards/accuracy_reward": 0.7482143178582191,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.5201164245606,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.10736213624477386,
      "kl": 0.006787872314453125,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0339,
      "reward": 0.737723246216774,
      "reward_std": 0.1853496261872351,
      "rewards/accuracy_reward": 0.737723246216774,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.9147575378418,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.8183253407478333,
      "kl": 0.011393356323242187,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0353,
      "reward": 0.7736607447266579,
      "reward_std": 0.19916837718337774,
      "rewards/accuracy_reward": 0.7736607447266579,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.7449696858724,
      "epoch": 0.9893390191897654,
      "kl": 0.0070037841796875,
      "reward": 0.7593006392319998,
      "reward_std": 0.19209627310434976,
      "rewards/accuracy_reward": 0.7593006392319998,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.03989953156721232,
      "train_runtime": 12322.7154,
      "train_samples_per_second": 0.609,
      "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
}