{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 280.6875,
      "epoch": 0.1,
      "grad_norm": 53.05036163330078,
      "kl": 0.0,
      "learning_rate": 4.965903258506806e-07,
      "loss": -0.0,
      "reward": 3.6584795396775007,
      "reward_std": 0.5753003612917382,
      "rewards/concensus_correctness_reward_func": 1.011624988168478,
      "rewards/consensus_reward_func": 1.0625,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.0,
      "rewards/question_recreation_reward_func": 0.656604582327418,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.171875,
      "rewards/xmlcount_reward_func": 0.7558750007301569,
      "step": 2
    },
    {
      "completion_length": 171.03125,
      "epoch": 0.2,
      "grad_norm": 169.1420135498047,
      "kl": 10.232354334380943,
      "learning_rate": 4.698684378016222e-07,
      "loss": 0.0102,
      "reward": 6.453376889228821,
      "reward_std": 0.14471963373944163,
      "rewards/concensus_correctness_reward_func": 1.8348749801516533,
      "rewards/consensus_reward_func": 2.0,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.125,
      "rewards/question_recreation_reward_func": 0.9452207088470459,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.34375,
      "rewards/xmlcount_reward_func": 1.2045312523841858,
      "step": 4
    },
    {
      "completion_length": 167.6875,
      "epoch": 0.3,
      "grad_norm": 273064.90625,
      "kl": 468753.35245687794,
      "learning_rate": 4.193203929064353e-07,
      "loss": 468.7533,
      "reward": 6.525232225656509,
      "reward_std": 0.3649999493500218,
      "rewards/concensus_correctness_reward_func": 2.006375003606081,
      "rewards/consensus_reward_func": 1.8125,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.25,
      "rewards/question_recreation_reward_func": 0.8971072547137737,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.359375,
      "rewards/xmlcount_reward_func": 1.1998750008642673,
      "step": 6
    },
    {
      "completion_length": 160.8125,
      "epoch": 0.4,
      "grad_norm": 167036304.0,
      "kl": 3727724.481420556,
      "learning_rate": 3.5042385616324236e-07,
      "loss": 3727.7244,
      "reward": 7.081181287765503,
      "reward_std": 0.10486791450239252,
      "rewards/concensus_correctness_reward_func": 2.170249997638166,
      "rewards/consensus_reward_func": 2.0,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.375,
      "rewards/question_recreation_reward_func": 0.9093687888234854,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.40625,
      "rewards/xmlcount_reward_func": 1.2203124985098839,
      "step": 8
    },
    {
      "completion_length": 147.0625,
      "epoch": 0.5,
      "grad_norm": 4871.97412109375,
      "kl": 539.2818314363249,
      "learning_rate": 2.706448363680831e-07,
      "loss": 0.5393,
      "reward": 6.877287954092026,
      "reward_std": 0.44364456451148726,
      "rewards/concensus_correctness_reward_func": 2.1652499809861183,
      "rewards/consensus_reward_func": 1.9375,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.375,
      "rewards/question_recreation_reward_func": 0.8499442553147674,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.328125,
      "rewards/xmlcount_reward_func": 1.2214687392115593,
      "step": 10
    },
    {
      "completion_length": 175.53125,
      "epoch": 0.6,
      "grad_norm": 50525.3671875,
      "kl": 2825.989122052677,
      "learning_rate": 1.886286282148002e-07,
      "loss": 2.826,
      "reward": 6.0079357624053955,
      "reward_std": 0.43696475913748145,
      "rewards/concensus_correctness_reward_func": 1.7586249820888042,
      "rewards/consensus_reward_func": 1.9375,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.0,
      "rewards/question_recreation_reward_func": 0.7420295961201191,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.359375,
      "rewards/xmlcount_reward_func": 1.210406243801117,
      "step": 12
    },
    {
      "completion_length": 177.8125,
      "epoch": 0.7,
      "grad_norm": 505.6527404785156,
      "kl": 1061.56522446312,
      "learning_rate": 1.1326296046939333e-07,
      "loss": 1.0616,
      "reward": 6.597704261541367,
      "reward_std": 0.4343329582188744,
      "rewards/concensus_correctness_reward_func": 2.0284374840557575,
      "rewards/consensus_reward_func": 1.8125,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.25,
      "rewards/question_recreation_reward_func": 0.8877355121076107,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.375,
      "rewards/xmlcount_reward_func": 1.2440312504768372,
      "step": 14
    },
    {
      "completion_length": 144.71875,
      "epoch": 0.8,
      "grad_norm": 49.680084228515625,
      "kl": 12.081052286550403,
      "learning_rate": 5.271487265090163e-08,
      "loss": 0.0121,
      "reward": 6.665501236915588,
      "reward_std": 0.5225161880371161,
      "rewards/concensus_correctness_reward_func": 2.029374983161688,
      "rewards/consensus_reward_func": 1.9375,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.1875,
      "rewards/question_recreation_reward_func": 0.8862824384123087,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.390625,
      "rewards/xmlcount_reward_func": 1.234218753874302,
      "step": 16
    },
    {
      "completion_length": 147.0625,
      "epoch": 0.9,
      "grad_norm": 730.200927734375,
      "kl": 86.71727948682383,
      "learning_rate": 1.3545689574841341e-08,
      "loss": 0.0867,
      "reward": 6.578728884458542,
      "reward_std": 0.26941974822693737,
      "rewards/concensus_correctness_reward_func": 2.021124981343746,
      "rewards/consensus_reward_func": 1.75,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.3125,
      "rewards/question_recreation_reward_func": 0.9322289191186428,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.359375,
      "rewards/xmlcount_reward_func": 1.2034999914467335,
      "step": 18
    },
    {
      "completion_length": 142.25,
      "epoch": 1.0,
      "grad_norm": 13.476627349853516,
      "kl": 7.980941329384223,
      "learning_rate": 0.0,
      "loss": 0.008,
      "reward": 6.966006487607956,
      "reward_std": 0.08144115762843285,
      "rewards/concensus_correctness_reward_func": 2.1567499935626984,
      "rewards/consensus_reward_func": 2.0,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.25,
      "rewards/question_recreation_reward_func": 0.9232877704780549,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.421875,
      "rewards/xmlcount_reward_func": 1.2140937522053719,
      "step": 20
    },
    {
      "epoch": 1.0,
      "step": 20,
      "total_flos": 0.0,
      "train_loss": 420.1021536417305,
      "train_runtime": 134.2834,
      "train_samples_per_second": 2.383,
      "train_steps_per_second": 0.149
    }
  ],
  "logging_steps": 2,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 25,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}