{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 15,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 324.875,
      "epoch": 0.7272727272727273,
      "grad_norm": 20.83338165283203,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0,
      "reward": 2.082763216458261,
      "reward_std": 1.3784077446907759,
      "rewards/concensus_correctness_reward_func": 0.31875000102445483,
      "rewards/consensus_reward_func": 0.375,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.3125,
      "rewards/question_recreation_reward_func": 0.554200713057071,
      "rewards/soft_format_reward_func": 0.015625,
      "rewards/strict_format_reward_func": 0.015625,
      "rewards/xmlcount_reward_func": 0.4910624995827675,
      "step": 2
    },
    {
      "completion_length": 302.17857142857144,
      "epoch": 1.3636363636363638,
      "grad_norm": 10.466278076171875,
      "kl": 0.004970156163575926,
      "learning_rate": 4.752422169756047e-07,
      "loss": 0.0,
      "reward": 3.386433482170105,
      "reward_std": 1.9209142122417688,
      "rewards/concensus_correctness_reward_func": 0.7719999892371041,
      "rewards/consensus_reward_func": 0.7857142857142857,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.2857142857142857,
      "rewards/question_recreation_reward_func": 0.7027192754404885,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.017857142857142856,
      "rewards/xmlcount_reward_func": 0.822428571326392,
      "step": 4
    },
    {
      "completion_length": 292.25,
      "epoch": 2.0,
      "grad_norm": 7.234833240509033,
      "kl": 0.013450991345702537,
      "learning_rate": 4.058724504646834e-07,
      "loss": 0.0,
      "reward": 4.393697313964367,
      "reward_std": 2.3138607857482776,
      "rewards/concensus_correctness_reward_func": 1.5589285743023669,
      "rewards/consensus_reward_func": 0.9285714285714286,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.42857142857142855,
      "rewards/question_recreation_reward_func": 0.674875912389585,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.017857142857142856,
      "rewards/xmlcount_reward_func": 0.7848928613322121,
      "step": 6
    },
    {
      "completion_length": 266.8125,
      "epoch": 2.7272727272727275,
      "grad_norm": 9.109390258789062,
      "kl": 0.035667407981236465,
      "learning_rate": 3.056302334890786e-07,
      "loss": 0.0,
      "reward": 4.11042432487011,
      "reward_std": 1.9496262441243744,
      "rewards/concensus_correctness_reward_func": 1.3073750026524067,
      "rewards/consensus_reward_func": 0.75,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.5,
      "rewards/question_recreation_reward_func": 0.6049242871813476,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.948124997317791,
      "step": 8
    },
    {
      "completion_length": 299.7142857142857,
      "epoch": 3.3636363636363638,
      "grad_norm": 33.44534683227539,
      "kl": 3.1398622704769616,
      "learning_rate": 1.9436976651092142e-07,
      "loss": 0.0027,
      "reward": 3.5893632641860416,
      "reward_std": 1.3633836537067379,
      "rewards/concensus_correctness_reward_func": 0.8663571423717907,
      "rewards/consensus_reward_func": 0.9285714285714286,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.14285714285714285,
      "rewards/question_recreation_reward_func": 0.7662561748708997,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.017857142857142856,
      "rewards/xmlcount_reward_func": 0.8674642784254891,
      "step": 10
    },
    {
      "completion_length": 319.42857142857144,
      "epoch": 4.0,
      "grad_norm": 8.600255966186523,
      "kl": 0.023560081858054867,
      "learning_rate": 9.412754953531663e-08,
      "loss": 0.0,
      "reward": 3.0599673305238997,
      "reward_std": 1.0951891348231584,
      "rewards/concensus_correctness_reward_func": 0.5134285640503679,
      "rewards/consensus_reward_func": 0.6428571428571429,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.35714285714285715,
      "rewards/question_recreation_reward_func": 0.6900031226021903,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.05357142857142857,
      "rewards/xmlcount_reward_func": 0.8029642807585853,
      "step": 12
    },
    {
      "completion_length": 288.03125,
      "epoch": 4.7272727272727275,
      "grad_norm": 22.622495651245117,
      "kl": 0.1138462919043377,
      "learning_rate": 2.475778302439524e-08,
      "loss": 0.0001,
      "reward": 3.153999723494053,
      "reward_std": 1.2494984324439429,
      "rewards/concensus_correctness_reward_func": 0.6776249930262566,
      "rewards/consensus_reward_func": 0.6875,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.375,
      "rewards/question_recreation_reward_func": 0.6243123151361942,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.7895624944940209,
      "step": 14
    },
    {
      "completion_length": 294.0833333333333,
      "epoch": 5.0,
      "kl": 1.2046936283198495,
      "reward": 3.455769936243693,
      "reward_std": 1.1367127535243828,
      "rewards/concensus_correctness_reward_func": 0.4769999956091245,
      "rewards/consensus_reward_func": 1.0,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.3333333333333333,
      "rewards/question_recreation_reward_func": 0.6737699757019678,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.041666666666666664,
      "rewards/xmlcount_reward_func": 0.9300000071525574,
      "step": 15,
      "total_flos": 0.0,
      "train_loss": 0.00045103654265403746,
      "train_runtime": 169.526,
      "train_samples_per_second": 1.416,
      "train_steps_per_second": 0.088
    }
  ],
  "logging_steps": 2,
  "max_steps": 15,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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
}