{
  "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": 424.46875,
      "epoch": 0.1,
      "grad_norm": 4.621828079223633,
      "kl": 0.0,
      "learning_rate": 4.965903258506806e-07,
      "loss": -0.0,
      "reward": 1.155371225439012,
      "reward_std": 1.157450430560857,
      "rewards/concensus_correctness_reward_func": 0.263374999165535,
      "rewards/consensus_reward_func": 0.1875,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.25,
      "rewards/question_recreation_reward_func": 0.2525274936342612,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.20196874719113111,
      "step": 2
    },
    {
      "completion_length": 423.6875,
      "epoch": 0.2,
      "grad_norm": 7.296565532684326,
      "kl": 0.0013047958818788175,
      "learning_rate": 4.698684378016222e-07,
      "loss": 0.0,
      "reward": 0.8222260437905788,
      "reward_std": 0.636947439139476,
      "rewards/concensus_correctness_reward_func": 0.05025000125169754,
      "rewards/consensus_reward_func": 0.0,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.125,
      "rewards/question_recreation_reward_func": 0.315944813657552,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.3310312391258776,
      "step": 4
    },
    {
      "completion_length": 550.5625,
      "epoch": 0.3,
      "grad_norm": 10.931665420532227,
      "kl": 0.001646797751163831,
      "learning_rate": 4.193203929064353e-07,
      "loss": 0.0,
      "reward": 0.8547968231141567,
      "reward_std": 1.5359444813802838,
      "rewards/concensus_correctness_reward_func": 0.23118750005960464,
      "rewards/consensus_reward_func": 0.0625,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.125,
      "rewards/question_recreation_reward_func": 0.3315468009095639,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.015625,
      "rewards/xmlcount_reward_func": 0.08893750491552055,
      "step": 6
    },
    {
      "completion_length": 527.5625,
      "epoch": 0.4,
      "grad_norm": 4.5462517738342285,
      "kl": 0.0033753589050320443,
      "learning_rate": 3.5042385616324236e-07,
      "loss": 0.0,
      "reward": 1.172942945966497,
      "reward_std": 1.0442042673239484,
      "rewards/concensus_correctness_reward_func": 0.28293750062584877,
      "rewards/consensus_reward_func": 0.1875,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.0625,
      "rewards/question_recreation_reward_func": 0.39450543955899775,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.24549999507144094,
      "step": 8
    },
    {
      "completion_length": 596.84375,
      "epoch": 0.5,
      "grad_norm": 5.691103458404541,
      "kl": 0.002918783080531284,
      "learning_rate": 2.706448363680831e-07,
      "loss": 0.0,
      "reward": 1.119660965166986,
      "reward_std": 1.2069853162392974,
      "rewards/concensus_correctness_reward_func": 0.1850624978542328,
      "rewards/consensus_reward_func": 0.125,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.1875,
      "rewards/question_recreation_reward_func": 0.3388797022635117,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.2832187442108989,
      "step": 10
    },
    {
      "completion_length": 592.96875,
      "epoch": 0.6,
      "grad_norm": 4.646827220916748,
      "kl": 0.0031351612269645557,
      "learning_rate": 1.886286282148002e-07,
      "loss": 0.0,
      "reward": 0.9737958908081055,
      "reward_std": 0.9545698054134846,
      "rewards/concensus_correctness_reward_func": 0.2938749995082617,
      "rewards/consensus_reward_func": 0.0625,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.0,
      "rewards/question_recreation_reward_func": 0.40088965790346265,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.21653124503791332,
      "step": 12
    },
    {
      "completion_length": 575.0,
      "epoch": 0.7,
      "grad_norm": 5.13407564163208,
      "kl": 0.004790089908055961,
      "learning_rate": 1.1326296046939333e-07,
      "loss": 0.0,
      "reward": 2.0287008844316006,
      "reward_std": 1.4835871619870886,
      "rewards/concensus_correctness_reward_func": 0.545687498524785,
      "rewards/consensus_reward_func": 0.25,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.3125,
      "rewards/question_recreation_reward_func": 0.5870445817708969,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.3334687491878867,
      "step": 14
    },
    {
      "completion_length": 541.90625,
      "epoch": 0.8,
      "grad_norm": 4.688849449157715,
      "kl": 0.004843814662308432,
      "learning_rate": 5.271487265090163e-08,
      "loss": 0.0,
      "reward": 1.3065089867450297,
      "reward_std": 0.554467682595714,
      "rewards/concensus_correctness_reward_func": 0.1198749989271164,
      "rewards/consensus_reward_func": 0.0625,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.375,
      "rewards/question_recreation_reward_func": 0.4176964627113193,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.3314374880865216,
      "step": 16
    },
    {
      "completion_length": 787.59375,
      "epoch": 0.9,
      "grad_norm": 2.6661102771759033,
      "kl": 0.0034712271644821158,
      "learning_rate": 1.3545689574841341e-08,
      "loss": 0.0,
      "reward": 1.0092257296200842,
      "reward_std": 1.405159011046635,
      "rewards/concensus_correctness_reward_func": 0.20868750102818012,
      "rewards/consensus_reward_func": 0.125,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.25,
      "rewards/question_recreation_reward_func": 0.315475704104756,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.11006249859929085,
      "step": 18
    },
    {
      "completion_length": 573.4375,
      "epoch": 1.0,
      "grad_norm": 3.8552772998809814,
      "kl": 0.004824902229302097,
      "learning_rate": 0.0,
      "loss": 0.0,
      "reward": 2.2213319819420576,
      "reward_std": 2.8470122441649437,
      "rewards/concensus_correctness_reward_func": 1.1250000074505806,
      "rewards/consensus_reward_func": 0.25,
      "rewards/cumulative_reward_2": 0.0,
      "rewards/final_correctness_reward_func": 0.25,
      "rewards/question_recreation_reward_func": 0.5113319638185203,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.03125,
      "rewards/xmlcount_reward_func": 0.05375000648200512,
      "step": 20
    },
    {
      "epoch": 1.0,
      "step": 20,
      "total_flos": 0.0,
      "train_loss": 3.0882656574249268e-06,
      "train_runtime": 382.8125,
      "train_samples_per_second": 0.836,
      "train_steps_per_second": 0.052
    }
  ],
  "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
}