{
  "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": 395.39176177978516,
      "epoch": 0.017057569296375266,
      "grad_norm": 1.4015110731124878,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0686,
      "reward": 0.7399553880095482,
      "reward_std": 0.35724927112460136,
      "rewards/accuracy_reward": 0.7399553880095482,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 417.85381507873535,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.43299397826194763,
      "kl": 0.00021469593048095703,
      "learning_rate": 2.5e-06,
      "loss": 0.0718,
      "reward": 0.7338169980794191,
      "reward_std": 0.3763962681405246,
      "rewards/accuracy_reward": 0.7335379812866449,
      "rewards/format_reward": 0.00027901786961592734,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 490.7464515686035,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.2890031933784485,
      "kl": 1.2583311080932618,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.2281,
      "reward": 0.8033482536673546,
      "reward_std": 0.2846699129790068,
      "rewards/accuracy_reward": 0.8033482536673546,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 485.9149772644043,
      "epoch": 0.255863539445629,
      "grad_norm": 0.14673297107219696,
      "kl": 0.0069026947021484375,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0851,
      "reward": 0.8765625387430191,
      "reward_std": 0.18012245232239366,
      "rewards/accuracy_reward": 0.8765625387430191,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 562.7558280944825,
      "epoch": 0.3411513859275053,
      "grad_norm": 1.1095651388168335,
      "kl": 0.0181610107421875,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0494,
      "reward": 0.8392857491970063,
      "reward_std": 0.18913990762084723,
      "rewards/accuracy_reward": 0.8392857491970063,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 552.2873062133789,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.21138454973697662,
      "kl": 0.006878089904785156,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0371,
      "reward": 0.843080398440361,
      "reward_std": 0.1689525332301855,
      "rewards/accuracy_reward": 0.843080398440361,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 576.7803855895996,
      "epoch": 0.511727078891258,
      "grad_norm": 0.3545551896095276,
      "kl": 0.00529937744140625,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0392,
      "reward": 0.8198661088943482,
      "reward_std": 0.16675762198865413,
      "rewards/accuracy_reward": 0.8198661088943482,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 623.479264831543,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.11543061584234238,
      "kl": 0.0050445556640625,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0352,
      "reward": 0.7696428924798966,
      "reward_std": 0.20560304205864668,
      "rewards/accuracy_reward": 0.7696428924798966,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 636.8741333007813,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.12837538123130798,
      "kl": 0.004553985595703125,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0437,
      "reward": 0.7522321760654449,
      "reward_std": 0.21456908099353314,
      "rewards/accuracy_reward": 0.7522321760654449,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 710.5674453735352,
      "epoch": 0.767590618336887,
      "grad_norm": 0.11472984403371811,
      "kl": 174.40458183288575,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 14.2489,
      "reward": 0.5919643111526967,
      "reward_std": 0.2533439327031374,
      "rewards/accuracy_reward": 0.5919643111526967,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 737.0203475952148,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.13668760657310486,
      "kl": 0.003937911987304687,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0285,
      "reward": 0.5479910984635353,
      "reward_std": 0.24692099243402482,
      "rewards/accuracy_reward": 0.5479910984635353,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 727.8326248168945,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.13492952287197113,
      "kl": 0.004103469848632813,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0362,
      "reward": 0.5658482372760772,
      "reward_std": 0.24713426027446986,
      "rewards/accuracy_reward": 0.5658482372760772,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 732.3923924763998,
      "epoch": 0.9893390191897654,
      "kl": 0.0041942596435546875,
      "reward": 0.5450149066746235,
      "reward_std": 0.287875646725297,
      "rewards/accuracy_reward": 0.5450149066746235,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 1.2875895330618168,
      "train_runtime": 16023.0623,
      "train_samples_per_second": 0.468,
      "train_steps_per_second": 0.004
    }
  ],
  "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
}