{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 100.0,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 652.46875,
      "epoch": 0.2,
      "grad_norm": 0.052754957228899,
      "kl": 0.0,
      "learning_rate": 4.0000000000000003e-07,
      "loss": -0.0,
      "reward": 0.21875,
      "reward_std": 0.30173346400260925,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 1
    },
    {
      "completion_length": 749.75,
      "epoch": 0.4,
      "grad_norm": 0.03587888181209564,
      "kl": 0.0,
      "learning_rate": 8.000000000000001e-07,
      "loss": -0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 2
    },
    {
      "completion_length": 693.03125,
      "epoch": 0.6,
      "grad_norm": 0.03723589703440666,
      "kl": 0.00010526180267333984,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.0,
      "reward": 0.1875,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 3
    },
    {
      "completion_length": 811.0,
      "epoch": 0.8,
      "grad_norm": 0.03902453929185867,
      "kl": 0.0001246929168701172,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 4
    },
    {
      "completion_length": 659.625,
      "epoch": 1.0,
      "grad_norm": 0.07274740189313889,
      "kl": 0.00017940998077392578,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 5
    },
    {
      "completion_length": 675.5,
      "epoch": 1.2,
      "grad_norm": 0.043703626841306686,
      "kl": 0.00015532970428466797,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 6
    },
    {
      "completion_length": 744.5,
      "epoch": 1.4,
      "grad_norm": 0.04744574800133705,
      "kl": 7.432699203491211e-05,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 7
    },
    {
      "completion_length": 710.0625,
      "epoch": 1.6,
      "grad_norm": 0.049952100962400436,
      "kl": 0.00013005733489990234,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0,
      "reward": 0.28125,
      "reward_std": 0.3608423173427582,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.21875,
      "step": 8
    },
    {
      "completion_length": 808.3125,
      "epoch": 1.8,
      "grad_norm": 0.02467654086649418,
      "kl": 7.808208465576172e-05,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 9
    },
    {
      "completion_length": 637.9375,
      "epoch": 2.0,
      "grad_norm": 0.0368252769112587,
      "kl": 0.0001710653305053711,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 10
    },
    {
      "completion_length": 672.125,
      "epoch": 2.2,
      "grad_norm": 0.29078933596611023,
      "kl": 0.0002989768981933594,
      "learning_rate": 4.4e-06,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 11
    },
    {
      "completion_length": 812.3125,
      "epoch": 2.4,
      "grad_norm": 0.03565595671534538,
      "kl": 0.00011175870895385742,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 12
    },
    {
      "completion_length": 685.46875,
      "epoch": 2.6,
      "grad_norm": 0.055703457444906235,
      "kl": 0.00016438961029052734,
      "learning_rate": 5.2e-06,
      "loss": 0.0,
      "reward": 0.25,
      "reward_std": 0.3945523276925087,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.03125,
      "step": 13
    },
    {
      "completion_length": 728.9375,
      "epoch": 2.8,
      "grad_norm": 0.04565276578068733,
      "kl": 9.632110595703125e-05,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.0,
      "reward": 0.1875,
      "reward_std": 0.3104073107242584,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 14
    },
    {
      "completion_length": 670.6875,
      "epoch": 3.0,
      "grad_norm": 0.061348024755716324,
      "kl": 0.00018405914306640625,
      "learning_rate": 6e-06,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 15
    },
    {
      "completion_length": 719.3125,
      "epoch": 3.2,
      "grad_norm": 0.03140898421406746,
      "kl": 0.00011992454528808594,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 16
    },
    {
      "completion_length": 782.375,
      "epoch": 3.4,
      "grad_norm": 0.047895658761262894,
      "kl": 0.00013840198516845703,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 17
    },
    {
      "completion_length": 685.5,
      "epoch": 3.6,
      "grad_norm": 0.03884062170982361,
      "kl": 0.00011730194091796875,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 18
    },
    {
      "completion_length": 742.78125,
      "epoch": 3.8,
      "grad_norm": 0.0465020127594471,
      "kl": 0.0001150369644165039,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.0,
      "reward": 0.28125,
      "reward_std": 0.2630179077386856,
      "rewards/accuracy_reward": 0.28125,
      "rewards/format_reward": 0.0,
      "step": 19
    },
    {
      "completion_length": 666.625,
      "epoch": 4.0,
      "grad_norm": 0.04878535866737366,
      "kl": 0.00016605854034423828,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 20
    },
    {
      "completion_length": 713.96875,
      "epoch": 4.2,
      "grad_norm": 0.05213127285242081,
      "kl": 0.00013446807861328125,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.0,
      "reward": 0.25,
      "reward_std": 0.3745020925998688,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.1875,
      "step": 21
    },
    {
      "completion_length": 696.5625,
      "epoch": 4.4,
      "grad_norm": 0.0467567965388298,
      "kl": 0.00017702579498291016,
      "learning_rate": 8.8e-06,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 22
    },
    {
      "completion_length": 768.40625,
      "epoch": 4.6,
      "grad_norm": 0.059018660336732864,
      "kl": 0.00012123584747314453,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 23
    },
    {
      "completion_length": 762.59375,
      "epoch": 4.8,
      "grad_norm": 0.10580332577228546,
      "kl": 0.00010251998901367188,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 24
    },
    {
      "completion_length": 708.75,
      "epoch": 5.0,
      "grad_norm": 0.053475841879844666,
      "kl": 0.00011232495307922363,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 25
    },
    {
      "completion_length": 681.53125,
      "epoch": 5.2,
      "grad_norm": 0.06291075795888901,
      "kl": 9.40561294555664e-05,
      "learning_rate": 1.04e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 26
    },
    {
      "completion_length": 737.15625,
      "epoch": 5.4,
      "grad_norm": 0.0963231772184372,
      "kl": 0.00013685226440429688,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 27
    },
    {
      "completion_length": 807.21875,
      "epoch": 5.6,
      "grad_norm": 0.03735160082578659,
      "kl": 0.0001347064971923828,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 28
    },
    {
      "completion_length": 795.34375,
      "epoch": 5.8,
      "grad_norm": 0.024533439427614212,
      "kl": 9.292364120483398e-05,
      "learning_rate": 1.16e-05,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 29
    },
    {
      "completion_length": 638.125,
      "epoch": 6.0,
      "grad_norm": 0.03820430859923363,
      "kl": 0.00011861324310302734,
      "learning_rate": 1.2e-05,
      "loss": 0.0,
      "reward": 0.25,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 672.0,
      "epoch": 6.2,
      "grad_norm": 0.04252497851848602,
      "kl": 0.00012922286987304688,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.0,
      "reward": 0.21875,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.03125,
      "step": 31
    },
    {
      "completion_length": 752.53125,
      "epoch": 6.4,
      "grad_norm": 0.02208935283124447,
      "kl": 0.00011682510375976562,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 32
    },
    {
      "completion_length": 876.03125,
      "epoch": 6.6,
      "grad_norm": 0.03968479111790657,
      "kl": 0.00010192394256591797,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 33
    },
    {
      "completion_length": 685.3125,
      "epoch": 6.8,
      "grad_norm": 0.056483570486307144,
      "kl": 0.0001461505889892578,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 34
    },
    {
      "completion_length": 734.375,
      "epoch": 7.0,
      "grad_norm": 0.03762107715010643,
      "kl": 0.00012969970703125,
      "learning_rate": 1.4e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 35
    },
    {
      "completion_length": 822.375,
      "epoch": 7.2,
      "grad_norm": 0.021994464099407196,
      "kl": 0.0001533031463623047,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.0,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 36
    },
    {
      "completion_length": 682.875,
      "epoch": 7.4,
      "grad_norm": 0.02833479456603527,
      "kl": 9.298324584960938e-05,
      "learning_rate": 1.48e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 37
    },
    {
      "completion_length": 754.125,
      "epoch": 7.6,
      "grad_norm": 0.03455812484025955,
      "kl": 0.00015020370483398438,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 38
    },
    {
      "completion_length": 655.71875,
      "epoch": 7.8,
      "grad_norm": 0.016996435821056366,
      "kl": 0.00019550323486328125,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.0,
      "reward": 0.21875,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "completion_length": 728.5625,
      "epoch": 8.0,
      "grad_norm": 0.04107189178466797,
      "kl": 0.00013899803161621094,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 40
    },
    {
      "completion_length": 640.5625,
      "epoch": 8.2,
      "grad_norm": 0.05016947537660599,
      "kl": 0.00017881393432617188,
      "learning_rate": 1.64e-05,
      "loss": 0.0,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 41
    },
    {
      "completion_length": 678.28125,
      "epoch": 8.4,
      "grad_norm": 0.044850900769233704,
      "kl": 0.00013256072998046875,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.0,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 42
    },
    {
      "completion_length": 752.53125,
      "epoch": 8.6,
      "grad_norm": 0.04446345195174217,
      "kl": 0.0001544952392578125,
      "learning_rate": 1.72e-05,
      "loss": 0.0,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 43
    },
    {
      "completion_length": 799.4375,
      "epoch": 8.8,
      "grad_norm": 0.032816775143146515,
      "kl": 0.00015306472778320312,
      "learning_rate": 1.76e-05,
      "loss": 0.0,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 44
    },
    {
      "completion_length": 762.375,
      "epoch": 9.0,
      "grad_norm": 0.062291160225868225,
      "kl": 0.00016069412231445312,
      "learning_rate": 1.8e-05,
      "loss": 0.0,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 45
    },
    {
      "completion_length": 773.53125,
      "epoch": 9.2,
      "grad_norm": 0.05443451181054115,
      "kl": 0.0001690387725830078,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 46
    },
    {
      "completion_length": 718.65625,
      "epoch": 9.4,
      "grad_norm": 0.05950860306620598,
      "kl": 0.00017261505126953125,
      "learning_rate": 1.88e-05,
      "loss": 0.0,
      "reward": 0.25,
      "reward_std": 0.4261348247528076,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0625,
      "step": 47
    },
    {
      "completion_length": 704.1875,
      "epoch": 9.6,
      "grad_norm": 0.03597826883196831,
      "kl": 0.00015497207641601562,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.0,
      "reward": 0.21875,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.03125,
      "step": 48
    },
    {
      "completion_length": 653.78125,
      "epoch": 9.8,
      "grad_norm": 0.0006673262105323374,
      "kl": 0.00023221969604492188,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.0,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 49
    },
    {
      "completion_length": 818.8125,
      "epoch": 10.0,
      "grad_norm": 0.052374593913555145,
      "kl": 0.00018906593322753906,
      "learning_rate": 2e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 50
    },
    {
      "completion_length": 675.125,
      "epoch": 10.2,
      "grad_norm": 0.06491909921169281,
      "kl": 0.0004253387451171875,
      "learning_rate": 1.9999756307053947e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 51
    },
    {
      "completion_length": 634.96875,
      "epoch": 10.4,
      "grad_norm": 0.06437934935092926,
      "kl": 0.0003037452697753906,
      "learning_rate": 1.9999025240093045e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 52
    },
    {
      "completion_length": 641.90625,
      "epoch": 10.6,
      "grad_norm": 0.043609969317913055,
      "kl": 0.0001957416534423828,
      "learning_rate": 1.9997806834748455e-05,
      "loss": 0.0,
      "reward": 0.25,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.0,
      "step": 53
    },
    {
      "completion_length": 818.125,
      "epoch": 10.8,
      "grad_norm": 0.03535196930170059,
      "kl": 0.00021600723266601562,
      "learning_rate": 1.9996101150403543e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 54
    },
    {
      "completion_length": 794.3125,
      "epoch": 11.0,
      "grad_norm": 0.03906484693288803,
      "kl": 0.00020003318786621094,
      "learning_rate": 1.999390827019096e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 55
    },
    {
      "completion_length": 656.75,
      "epoch": 11.2,
      "grad_norm": 0.04632149264216423,
      "kl": 0.0003056526184082031,
      "learning_rate": 1.9991228300988586e-05,
      "loss": 0.0,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 56
    },
    {
      "completion_length": 673.40625,
      "epoch": 11.4,
      "grad_norm": 0.027432706207036972,
      "kl": 0.0002613067626953125,
      "learning_rate": 1.9988061373414342e-05,
      "loss": 0.0,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 57
    },
    {
      "completion_length": 688.46875,
      "epoch": 11.6,
      "grad_norm": 0.11954957991838455,
      "kl": 0.00041103363037109375,
      "learning_rate": 1.9984407641819812e-05,
      "loss": 0.0,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 58
    },
    {
      "completion_length": 724.15625,
      "epoch": 11.8,
      "grad_norm": 0.04400650039315224,
      "kl": 0.0005452632904052734,
      "learning_rate": 1.9980267284282718e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 59
    },
    {
      "completion_length": 906.3125,
      "epoch": 12.0,
      "grad_norm": 0.0012655943864956498,
      "kl": 0.0003103017807006836,
      "learning_rate": 1.9975640502598243e-05,
      "loss": 0.0,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "completion_length": 744.15625,
      "epoch": 12.2,
      "grad_norm": 0.05378064140677452,
      "kl": 0.0006260871887207031,
      "learning_rate": 1.9970527522269204e-05,
      "loss": 0.0,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.1875,
      "step": 61
    },
    {
      "completion_length": 710.0625,
      "epoch": 12.4,
      "grad_norm": 0.05093209445476532,
      "kl": 0.0006451606750488281,
      "learning_rate": 1.9964928592495046e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 62
    },
    {
      "completion_length": 623.5,
      "epoch": 12.6,
      "grad_norm": 0.038004543632268906,
      "kl": 0.0005998611450195312,
      "learning_rate": 1.9958843986159705e-05,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.13363061845302582,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 63
    },
    {
      "completion_length": 743.0625,
      "epoch": 12.8,
      "grad_norm": 0.1318265050649643,
      "kl": 0.0010037422180175781,
      "learning_rate": 1.9952273999818312e-05,
      "loss": 0.0,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 64
    },
    {
      "completion_length": 658.0,
      "epoch": 13.0,
      "grad_norm": 0.05642435699701309,
      "kl": 0.0009012222290039062,
      "learning_rate": 1.9945218953682736e-05,
      "loss": 0.0,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 65
    },
    {
      "completion_length": 736.90625,
      "epoch": 13.2,
      "grad_norm": 0.046298280358314514,
      "kl": 0.000713348388671875,
      "learning_rate": 1.9937679191605964e-05,
      "loss": 0.0,
      "reward": 0.1875,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 66
    },
    {
      "completion_length": 690.90625,
      "epoch": 13.4,
      "grad_norm": 0.06122737377882004,
      "kl": 0.0012760162353515625,
      "learning_rate": 1.992965508106537e-05,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0,
      "step": 67
    },
    {
      "completion_length": 746.96875,
      "epoch": 13.6,
      "grad_norm": 0.05444188043475151,
      "kl": 0.0017440319061279297,
      "learning_rate": 1.9921147013144782e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 68
    },
    {
      "completion_length": 746.21875,
      "epoch": 13.8,
      "grad_norm": 0.04397672787308693,
      "kl": 0.0015916824340820312,
      "learning_rate": 1.991215540251542e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 69
    },
    {
      "completion_length": 743.625,
      "epoch": 14.0,
      "grad_norm": 0.03757720813155174,
      "kl": 0.0013751983642578125,
      "learning_rate": 1.9902680687415704e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 70
    },
    {
      "completion_length": 724.8125,
      "epoch": 14.2,
      "grad_norm": 0.04262088984251022,
      "kl": 0.0010733604431152344,
      "learning_rate": 1.9892723329629885e-05,
      "loss": 0.0,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 71
    },
    {
      "completion_length": 768.65625,
      "epoch": 14.4,
      "grad_norm": 0.03520955517888069,
      "kl": 0.0016698837280273438,
      "learning_rate": 1.988228381446553e-05,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 72
    },
    {
      "completion_length": 635.8125,
      "epoch": 14.6,
      "grad_norm": 0.07774507254362106,
      "kl": 0.004955291748046875,
      "learning_rate": 1.987136265072988e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 73
    },
    {
      "completion_length": 743.875,
      "epoch": 14.8,
      "grad_norm": 0.08741147816181183,
      "kl": 0.002712249755859375,
      "learning_rate": 1.985996037070505e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3471825420856476,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.15625,
      "step": 74
    },
    {
      "completion_length": 564.5625,
      "epoch": 15.0,
      "grad_norm": 0.07323095947504044,
      "kl": 0.0051746368408203125,
      "learning_rate": 1.9848077530122083e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 75
    },
    {
      "completion_length": 824.25,
      "epoch": 15.2,
      "grad_norm": 0.04077565670013428,
      "kl": 0.0030851364135742188,
      "learning_rate": 1.983571470813386e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 76
    },
    {
      "completion_length": 847.9375,
      "epoch": 15.4,
      "grad_norm": 0.025483058765530586,
      "kl": 0.0022611618041992188,
      "learning_rate": 1.982287250728689e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 77
    },
    {
      "completion_length": 823.5625,
      "epoch": 15.6,
      "grad_norm": 0.06061682850122452,
      "kl": 0.005008697509765625,
      "learning_rate": 1.9809551553491918e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 78
    },
    {
      "completion_length": 580.59375,
      "epoch": 15.8,
      "grad_norm": 0.07759010046720505,
      "kl": 0.0047168731689453125,
      "learning_rate": 1.979575249599344e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3808925524353981,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 79
    },
    {
      "completion_length": 648.875,
      "epoch": 16.0,
      "grad_norm": 0.04703226685523987,
      "kl": 0.00841522216796875,
      "learning_rate": 1.9781476007338058e-05,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "completion_length": 727.96875,
      "epoch": 16.2,
      "grad_norm": 0.04409215599298477,
      "kl": 0.0054225921630859375,
      "learning_rate": 1.9766722783341682e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0,
      "step": 81
    },
    {
      "completion_length": 829.15625,
      "epoch": 16.4,
      "grad_norm": 0.05709024518728256,
      "kl": 0.006870269775390625,
      "learning_rate": 1.9751493543055634e-05,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 82
    },
    {
      "completion_length": 703.1875,
      "epoch": 16.6,
      "grad_norm": 0.06334344297647476,
      "kl": 0.0034942626953125,
      "learning_rate": 1.9735789028731603e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 83
    },
    {
      "completion_length": 697.4375,
      "epoch": 16.8,
      "grad_norm": 0.05215838551521301,
      "kl": 0.00803375244140625,
      "learning_rate": 1.9719610005785466e-05,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 84
    },
    {
      "completion_length": 782.8125,
      "epoch": 17.0,
      "grad_norm": 0.059087853878736496,
      "kl": 0.004886627197265625,
      "learning_rate": 1.9702957262759964e-05,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 85
    },
    {
      "completion_length": 801.5625,
      "epoch": 17.2,
      "grad_norm": 0.02566160261631012,
      "kl": 0.0037174224853515625,
      "learning_rate": 1.9685831611286312e-05,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 86
    },
    {
      "completion_length": 734.25,
      "epoch": 17.4,
      "grad_norm": 0.08946952223777771,
      "kl": 0.00384521484375,
      "learning_rate": 1.9668233886044597e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 87
    },
    {
      "completion_length": 677.03125,
      "epoch": 17.6,
      "grad_norm": 0.08532679080963135,
      "kl": 0.00876617431640625,
      "learning_rate": 1.9650164944723116e-05,
      "loss": 0.0004,
      "reward": 0.21875,
      "reward_std": 0.4218914955854416,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.21875,
      "step": 88
    },
    {
      "completion_length": 617.75,
      "epoch": 17.8,
      "grad_norm": 0.0515192411839962,
      "kl": 0.0021572113037109375,
      "learning_rate": 1.9631625667976584e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 89
    },
    {
      "completion_length": 755.0,
      "epoch": 18.0,
      "grad_norm": 0.04105350002646446,
      "kl": 0.0040836334228515625,
      "learning_rate": 1.961261695938319e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 90
    },
    {
      "completion_length": 789.90625,
      "epoch": 18.2,
      "grad_norm": 0.04714483767747879,
      "kl": 0.0016145706176757812,
      "learning_rate": 1.9593139745400575e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 91
    },
    {
      "completion_length": 710.78125,
      "epoch": 18.4,
      "grad_norm": 0.05726596340537071,
      "kl": 0.0026378631591796875,
      "learning_rate": 1.9573194975320672e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 92
    },
    {
      "completion_length": 646.375,
      "epoch": 18.6,
      "grad_norm": 0.03785916790366173,
      "kl": 0.0055751800537109375,
      "learning_rate": 1.9552783621223437e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 93
    },
    {
      "completion_length": 649.0,
      "epoch": 18.8,
      "grad_norm": 0.039185889065265656,
      "kl": 0.003993988037109375,
      "learning_rate": 1.9531906677929472e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 94
    },
    {
      "completion_length": 722.625,
      "epoch": 19.0,
      "grad_norm": 0.051298946142196655,
      "kl": 0.0039043426513671875,
      "learning_rate": 1.9510565162951538e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 95
    },
    {
      "completion_length": 701.03125,
      "epoch": 19.2,
      "grad_norm": 0.05333693325519562,
      "kl": 0.006542205810546875,
      "learning_rate": 1.9488760116444966e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 96
    },
    {
      "completion_length": 758.40625,
      "epoch": 19.4,
      "grad_norm": 0.059845130890607834,
      "kl": 0.00335693359375,
      "learning_rate": 1.9466492601156964e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3104073107242584,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 97
    },
    {
      "completion_length": 707.0,
      "epoch": 19.6,
      "grad_norm": 0.029148058965802193,
      "kl": 0.00183868408203125,
      "learning_rate": 1.944376370237481e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 98
    },
    {
      "completion_length": 637.3125,
      "epoch": 19.8,
      "grad_norm": 0.05930906534194946,
      "kl": 0.0045318603515625,
      "learning_rate": 1.942057452787297e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 99
    },
    {
      "completion_length": 890.5625,
      "epoch": 20.0,
      "grad_norm": 0.004534106235951185,
      "kl": 0.0011739730834960938,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.0,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "completion_length": 773.53125,
      "epoch": 20.2,
      "grad_norm": 0.06249362602829933,
      "kl": 0.0035648345947265625,
      "learning_rate": 1.937281989491892e-05,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.125,
      "step": 101
    },
    {
      "completion_length": 697.0625,
      "epoch": 20.4,
      "grad_norm": 0.039780255407094955,
      "kl": 0.0038700103759765625,
      "learning_rate": 1.9348256763960146e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 102
    },
    {
      "completion_length": 619.34375,
      "epoch": 20.6,
      "grad_norm": 0.07436488568782806,
      "kl": 0.0042095184326171875,
      "learning_rate": 1.9323238012155125e-05,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.4218914955854416,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 103
    },
    {
      "completion_length": 714.875,
      "epoch": 20.8,
      "grad_norm": 0.003782485844567418,
      "kl": 0.0012178421020507812,
      "learning_rate": 1.9297764858882516e-05,
      "loss": 0.0,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 104
    },
    {
      "completion_length": 819.0,
      "epoch": 21.0,
      "grad_norm": 0.03427855297923088,
      "kl": 0.00182342529296875,
      "learning_rate": 1.9271838545667876e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 105
    },
    {
      "completion_length": 836.125,
      "epoch": 21.2,
      "grad_norm": 0.06285113841295242,
      "kl": 0.0041866302490234375,
      "learning_rate": 1.9245460336123136e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 106
    },
    {
      "completion_length": 708.3125,
      "epoch": 21.4,
      "grad_norm": 0.04544105380773544,
      "kl": 0.0020160675048828125,
      "learning_rate": 1.9218631515885007e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.249358132481575,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 107
    },
    {
      "completion_length": 702.96875,
      "epoch": 21.6,
      "grad_norm": 0.031117625534534454,
      "kl": 0.0014514923095703125,
      "learning_rate": 1.9191353392552346e-05,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 108
    },
    {
      "completion_length": 726.0,
      "epoch": 21.8,
      "grad_norm": 0.02800888568162918,
      "kl": 0.0017895698547363281,
      "learning_rate": 1.9163627295622397e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.13363061845302582,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 109
    },
    {
      "completion_length": 734.0625,
      "epoch": 22.0,
      "grad_norm": 0.05484931170940399,
      "kl": 0.0052490234375,
      "learning_rate": 1.913545457642601e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 110
    },
    {
      "completion_length": 683.0,
      "epoch": 22.2,
      "grad_norm": 0.048629019409418106,
      "kl": 0.00447845458984375,
      "learning_rate": 1.910683660806177e-05,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 111
    },
    {
      "completion_length": 746.875,
      "epoch": 22.4,
      "grad_norm": 0.010094250552356243,
      "kl": 0.0025186538696289062,
      "learning_rate": 1.907777478532909e-05,
      "loss": 0.0001,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 112
    },
    {
      "completion_length": 723.84375,
      "epoch": 22.6,
      "grad_norm": 0.0752173513174057,
      "kl": 0.005557537078857422,
      "learning_rate": 1.9048270524660197e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 113
    },
    {
      "completion_length": 769.0625,
      "epoch": 22.8,
      "grad_norm": 0.0477801151573658,
      "kl": 0.0017805099487304688,
      "learning_rate": 1.901832526405114e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 114
    },
    {
      "completion_length": 699.6875,
      "epoch": 23.0,
      "grad_norm": 0.04839959368109703,
      "kl": 0.0012121200561523438,
      "learning_rate": 1.8987940462991673e-05,
      "loss": 0.0,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 115
    },
    {
      "completion_length": 707.34375,
      "epoch": 23.2,
      "grad_norm": 0.08084084093570709,
      "kl": 0.0027523040771484375,
      "learning_rate": 1.895711760239413e-05,
      "loss": 0.0001,
      "reward": 0.21875,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 116
    },
    {
      "completion_length": 739.59375,
      "epoch": 23.4,
      "grad_norm": 0.025769440457224846,
      "kl": 0.0030221939086914062,
      "learning_rate": 1.892585818452126e-05,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 117
    },
    {
      "completion_length": 723.46875,
      "epoch": 23.6,
      "grad_norm": 0.058995552361011505,
      "kl": 0.0026149749755859375,
      "learning_rate": 1.889416373291298e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3104073107242584,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 118
    },
    {
      "completion_length": 686.65625,
      "epoch": 23.8,
      "grad_norm": 0.05975084751844406,
      "kl": 0.0029544830322265625,
      "learning_rate": 1.8862035792312148e-05,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 119
    },
    {
      "completion_length": 659.9375,
      "epoch": 24.0,
      "grad_norm": 0.11447711288928986,
      "kl": 0.0073566436767578125,
      "learning_rate": 1.8829475928589272e-05,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 120
    },
    {
      "completion_length": 648.0,
      "epoch": 24.2,
      "grad_norm": 0.12400014698505402,
      "kl": 0.004389762878417969,
      "learning_rate": 1.879648572866617e-05,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 121
    },
    {
      "completion_length": 711.34375,
      "epoch": 24.4,
      "grad_norm": 0.0695832222700119,
      "kl": 0.006267547607421875,
      "learning_rate": 1.8763066800438638e-05,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 122
    },
    {
      "completion_length": 811.03125,
      "epoch": 24.6,
      "grad_norm": 0.03884181007742882,
      "kl": 0.0032482147216796875,
      "learning_rate": 1.8729220772698096e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 123
    },
    {
      "completion_length": 776.0,
      "epoch": 24.8,
      "grad_norm": 0.06304960697889328,
      "kl": 0.005489349365234375,
      "learning_rate": 1.869494929505219e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 124
    },
    {
      "completion_length": 766.8125,
      "epoch": 25.0,
      "grad_norm": 0.046755433082580566,
      "kl": 0.002269744873046875,
      "learning_rate": 1.866025403784439e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 125
    },
    {
      "completion_length": 737.90625,
      "epoch": 25.2,
      "grad_norm": 0.1073736622929573,
      "kl": 0.0057697296142578125,
      "learning_rate": 1.8625136692072577e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 126
    },
    {
      "completion_length": 575.90625,
      "epoch": 25.4,
      "grad_norm": 0.07604625076055527,
      "kl": 0.00994110107421875,
      "learning_rate": 1.8589598969306646e-05,
      "loss": 0.0004,
      "reward": 0.28125,
      "reward_std": 0.3787454217672348,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.09375,
      "step": 127
    },
    {
      "completion_length": 664.96875,
      "epoch": 25.6,
      "grad_norm": 0.048758264631032944,
      "kl": 0.005596160888671875,
      "learning_rate": 1.855364260160507e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 128
    },
    {
      "completion_length": 776.71875,
      "epoch": 25.8,
      "grad_norm": 0.06416219472885132,
      "kl": 0.01025390625,
      "learning_rate": 1.851726934143048e-05,
      "loss": 0.0004,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 129
    },
    {
      "completion_length": 810.0625,
      "epoch": 26.0,
      "grad_norm": 0.033319141715765,
      "kl": 0.0031871795654296875,
      "learning_rate": 1.848048096156426e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "completion_length": 564.875,
      "epoch": 26.2,
      "grad_norm": 0.09029232710599899,
      "kl": 0.01023101806640625,
      "learning_rate": 1.8443279255020153e-05,
      "loss": 0.0004,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 131
    },
    {
      "completion_length": 675.59375,
      "epoch": 26.4,
      "grad_norm": 0.02714545652270317,
      "kl": 0.004695892333984375,
      "learning_rate": 1.8405666034956842e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 132
    },
    {
      "completion_length": 761.46875,
      "epoch": 26.6,
      "grad_norm": 0.02261636219918728,
      "kl": 0.00455474853515625,
      "learning_rate": 1.836764313458962e-05,
      "loss": 0.0002,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 133
    },
    {
      "completion_length": 793.53125,
      "epoch": 26.8,
      "grad_norm": 0.076606884598732,
      "kl": 0.007883071899414062,
      "learning_rate": 1.8329212407100996e-05,
      "loss": 0.0003,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 134
    },
    {
      "completion_length": 641.375,
      "epoch": 27.0,
      "grad_norm": 0.06023378297686577,
      "kl": 0.008708953857421875,
      "learning_rate": 1.8290375725550417e-05,
      "loss": 0.0003,
      "reward": 0.25,
      "reward_std": 0.4492306634783745,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.125,
      "step": 135
    },
    {
      "completion_length": 720.78125,
      "epoch": 27.2,
      "grad_norm": 0.07514262944459915,
      "kl": 0.0074329376220703125,
      "learning_rate": 1.8251134982782952e-05,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 136
    },
    {
      "completion_length": 667.25,
      "epoch": 27.4,
      "grad_norm": 0.06626664847135544,
      "kl": 0.0064849853515625,
      "learning_rate": 1.821149209133704e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 137
    },
    {
      "completion_length": 691.46875,
      "epoch": 27.6,
      "grad_norm": 0.06274402886629105,
      "kl": 0.0063629150390625,
      "learning_rate": 1.8171448983351284e-05,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 138
    },
    {
      "completion_length": 670.0,
      "epoch": 27.8,
      "grad_norm": 0.05724909529089928,
      "kl": 0.0061397552490234375,
      "learning_rate": 1.8131007610470278e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 139
    },
    {
      "completion_length": 703.1875,
      "epoch": 28.0,
      "grad_norm": 0.09771377593278885,
      "kl": 0.0037441253662109375,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 140
    },
    {
      "completion_length": 614.125,
      "epoch": 28.2,
      "grad_norm": 0.07325705140829086,
      "kl": 0.009479522705078125,
      "learning_rate": 1.804893797355914e-05,
      "loss": 0.0004,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 141
    },
    {
      "completion_length": 746.75,
      "epoch": 28.4,
      "grad_norm": 0.050279486924409866,
      "kl": 0.0060100555419921875,
      "learning_rate": 1.8007313709487334e-05,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 142
    },
    {
      "completion_length": 752.0625,
      "epoch": 28.6,
      "grad_norm": 0.03074537217617035,
      "kl": 0.001331329345703125,
      "learning_rate": 1.7965299180241963e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 143
    },
    {
      "completion_length": 725.90625,
      "epoch": 28.8,
      "grad_norm": 0.0496852733194828,
      "kl": 0.0050640106201171875,
      "learning_rate": 1.792289643355191e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 144
    },
    {
      "completion_length": 633.0625,
      "epoch": 29.0,
      "grad_norm": 0.0302935391664505,
      "kl": 0.0042266845703125,
      "learning_rate": 1.788010753606722e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "completion_length": 753.09375,
      "epoch": 29.2,
      "grad_norm": 0.05465216562151909,
      "kl": 0.005916595458984375,
      "learning_rate": 1.78369345732584e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 146
    },
    {
      "completion_length": 798.53125,
      "epoch": 29.4,
      "grad_norm": 0.07724983990192413,
      "kl": 0.006725311279296875,
      "learning_rate": 1.7793379649314743e-05,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 147
    },
    {
      "completion_length": 670.96875,
      "epoch": 29.6,
      "grad_norm": 0.052045710384845734,
      "kl": 0.004421234130859375,
      "learning_rate": 1.7749444887041797e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 148
    },
    {
      "completion_length": 729.8125,
      "epoch": 29.8,
      "grad_norm": 0.04775654524564743,
      "kl": 0.002147674560546875,
      "learning_rate": 1.7705132427757895e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 149
    },
    {
      "completion_length": 677.6875,
      "epoch": 30.0,
      "grad_norm": 0.06813061982393265,
      "kl": 0.0017709732055664062,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.0001,
      "reward": 0.25,
      "reward_std": 0.3514062538743019,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "completion_length": 746.71875,
      "epoch": 30.2,
      "grad_norm": 0.059449851512908936,
      "kl": 0.0029287338256835938,
      "learning_rate": 1.761538307536737e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 151
    },
    {
      "completion_length": 686.21875,
      "epoch": 30.4,
      "grad_norm": 0.05846744030714035,
      "kl": 0.0047817230224609375,
      "learning_rate": 1.7569950556517566e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.13363061845302582,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 152
    },
    {
      "completion_length": 752.6875,
      "epoch": 30.6,
      "grad_norm": 0.06499540060758591,
      "kl": 0.0039234161376953125,
      "learning_rate": 1.7524149088957244e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 153
    },
    {
      "completion_length": 840.90625,
      "epoch": 30.8,
      "grad_norm": 0.01074282918125391,
      "kl": 0.002513885498046875,
      "learning_rate": 1.747798090498532e-05,
      "loss": 0.0001,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 154
    },
    {
      "completion_length": 624.0,
      "epoch": 31.0,
      "grad_norm": 0.0682206004858017,
      "kl": 0.0042247772216796875,
      "learning_rate": 1.7431448254773943e-05,
      "loss": 0.0002,
      "reward": 0.3125,
      "reward_std": 0.3745020925998688,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.15625,
      "step": 155
    },
    {
      "completion_length": 750.84375,
      "epoch": 31.2,
      "grad_norm": 0.028054354712367058,
      "kl": 0.002704620361328125,
      "learning_rate": 1.7384553406258842e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 156
    },
    {
      "completion_length": 797.875,
      "epoch": 31.4,
      "grad_norm": 0.05043506994843483,
      "kl": 0.0019140243530273438,
      "learning_rate": 1.7337298645028764e-05,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 157
    },
    {
      "completion_length": 752.4375,
      "epoch": 31.6,
      "grad_norm": 0.03465605154633522,
      "kl": 0.005340576171875,
      "learning_rate": 1.7289686274214116e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 158
    },
    {
      "completion_length": 691.3125,
      "epoch": 31.8,
      "grad_norm": 0.047109831124544144,
      "kl": 0.003963470458984375,
      "learning_rate": 1.7241718614374678e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 159
    },
    {
      "completion_length": 690.5,
      "epoch": 32.0,
      "grad_norm": 0.0693172961473465,
      "kl": 0.0039043426513671875,
      "learning_rate": 1.7193398003386514e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 160
    },
    {
      "completion_length": 709.65625,
      "epoch": 32.2,
      "grad_norm": 0.2563571333885193,
      "kl": 0.009204864501953125,
      "learning_rate": 1.7144726796328034e-05,
      "loss": 0.0004,
      "reward": 0.1875,
      "reward_std": 0.408231720328331,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.15625,
      "step": 161
    },
    {
      "completion_length": 702.03125,
      "epoch": 32.4,
      "grad_norm": 0.07080015540122986,
      "kl": 0.0035953521728515625,
      "learning_rate": 1.709570736536521e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 162
    },
    {
      "completion_length": 829.84375,
      "epoch": 32.6,
      "grad_norm": 0.04313138872385025,
      "kl": 0.0013427734375,
      "learning_rate": 1.7046342099635948e-05,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 163
    },
    {
      "completion_length": 624.0,
      "epoch": 32.8,
      "grad_norm": 0.11450140178203583,
      "kl": 0.0047283172607421875,
      "learning_rate": 1.6996633405133656e-05,
      "loss": 0.0002,
      "reward": 0.25,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.03125,
      "step": 164
    },
    {
      "completion_length": 763.4375,
      "epoch": 33.0,
      "grad_norm": 0.06312569975852966,
      "kl": 0.008212089538574219,
      "learning_rate": 1.6946583704589973e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 165
    },
    {
      "completion_length": 755.6875,
      "epoch": 33.2,
      "grad_norm": 0.07195674628019333,
      "kl": 0.0042572021484375,
      "learning_rate": 1.68961954373567e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 166
    },
    {
      "completion_length": 824.8125,
      "epoch": 33.4,
      "grad_norm": 0.05315883830189705,
      "kl": 0.0027074813842773438,
      "learning_rate": 1.684547105928689e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 167
    },
    {
      "completion_length": 751.78125,
      "epoch": 33.6,
      "grad_norm": 0.044143546372652054,
      "kl": 0.0056304931640625,
      "learning_rate": 1.6794413042615168e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 168
    },
    {
      "completion_length": 715.0,
      "epoch": 33.8,
      "grad_norm": 0.054993364959955215,
      "kl": 0.0035266876220703125,
      "learning_rate": 1.6743023875837233e-05,
      "loss": 0.0001,
      "reward": 0.28125,
      "reward_std": 0.3608423173427582,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0625,
      "step": 169
    },
    {
      "completion_length": 711.3125,
      "epoch": 34.0,
      "grad_norm": 0.04282531887292862,
      "kl": 0.00383758544921875,
      "learning_rate": 1.6691306063588583e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 170
    },
    {
      "completion_length": 710.8125,
      "epoch": 34.2,
      "grad_norm": 0.04921108856797218,
      "kl": 0.0023479461669921875,
      "learning_rate": 1.6639262126522417e-05,
      "loss": 0.0001,
      "reward": 0.34375,
      "reward_std": 0.3787454217672348,
      "rewards/accuracy_reward": 0.3125,
      "rewards/format_reward": 0.03125,
      "step": 171
    },
    {
      "completion_length": 752.34375,
      "epoch": 34.4,
      "grad_norm": 0.05801001563668251,
      "kl": 0.0030841827392578125,
      "learning_rate": 1.6586894601186804e-05,
      "loss": 0.0001,
      "reward": 0.25,
      "reward_std": 0.4261348247528076,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.09375,
      "step": 172
    },
    {
      "completion_length": 642.21875,
      "epoch": 34.6,
      "grad_norm": 0.07046660035848618,
      "kl": 0.007915496826171875,
      "learning_rate": 1.6534206039901057e-05,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 173
    },
    {
      "completion_length": 828.03125,
      "epoch": 34.8,
      "grad_norm": 0.06986822932958603,
      "kl": 0.0038127899169921875,
      "learning_rate": 1.6481199010631312e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 174
    },
    {
      "completion_length": 678.125,
      "epoch": 35.0,
      "grad_norm": 0.06373422592878342,
      "kl": 0.0052967071533203125,
      "learning_rate": 1.6427876096865394e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 175
    },
    {
      "completion_length": 740.34375,
      "epoch": 35.2,
      "grad_norm": 0.33381345868110657,
      "kl": 0.044974327087402344,
      "learning_rate": 1.63742398974869e-05,
      "loss": 0.0018,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 176
    },
    {
      "completion_length": 682.5,
      "epoch": 35.4,
      "grad_norm": 0.07477016001939774,
      "kl": 0.005031585693359375,
      "learning_rate": 1.632029302664851e-05,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.408231720328331,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.125,
      "step": 177
    },
    {
      "completion_length": 734.8125,
      "epoch": 35.6,
      "grad_norm": 0.06294216960668564,
      "kl": 0.0035686492919921875,
      "learning_rate": 1.6266038113644605e-05,
      "loss": 0.0001,
      "reward": 0.25,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.09375,
      "step": 178
    },
    {
      "completion_length": 633.65625,
      "epoch": 35.8,
      "grad_norm": 0.059178903698921204,
      "kl": 0.0047435760498046875,
      "learning_rate": 1.6211477802783105e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 179
    },
    {
      "completion_length": 709.8125,
      "epoch": 36.0,
      "grad_norm": 0.015170346945524216,
      "kl": 0.004337310791015625,
      "learning_rate": 1.6156614753256583e-05,
      "loss": 0.0002,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "completion_length": 736.25,
      "epoch": 36.2,
      "grad_norm": 0.062141355127096176,
      "kl": 0.0058135986328125,
      "learning_rate": 1.610145163901268e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 181
    },
    {
      "completion_length": 832.59375,
      "epoch": 36.4,
      "grad_norm": 0.0728365033864975,
      "kl": 0.002017974853515625,
      "learning_rate": 1.6045991148623752e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.408231720328331,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.125,
      "step": 182
    },
    {
      "completion_length": 679.78125,
      "epoch": 36.6,
      "grad_norm": 0.03516562283039093,
      "kl": 0.004940032958984375,
      "learning_rate": 1.599023598515586e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 183
    },
    {
      "completion_length": 664.0,
      "epoch": 36.8,
      "grad_norm": 0.05773026868700981,
      "kl": 0.00171661376953125,
      "learning_rate": 1.5934188866037017e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 184
    },
    {
      "completion_length": 674.625,
      "epoch": 37.0,
      "grad_norm": 0.06871882826089859,
      "kl": 0.002918243408203125,
      "learning_rate": 1.5877852522924733e-05,
      "loss": 0.0001,
      "reward": 0.28125,
      "reward_std": 0.3471629247069359,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.125,
      "step": 185
    },
    {
      "completion_length": 749.28125,
      "epoch": 37.2,
      "grad_norm": 0.12351617217063904,
      "kl": 0.0049076080322265625,
      "learning_rate": 1.5821229701572897e-05,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.408231720328331,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.125,
      "step": 186
    },
    {
      "completion_length": 737.09375,
      "epoch": 37.4,
      "grad_norm": 0.05196770280599594,
      "kl": 0.0021238327026367188,
      "learning_rate": 1.5764323161697933e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.3535533845424652,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 187
    },
    {
      "completion_length": 746.6875,
      "epoch": 37.6,
      "grad_norm": 0.044122546911239624,
      "kl": 0.0019297599792480469,
      "learning_rate": 1.570713567684432e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 188
    },
    {
      "completion_length": 736.28125,
      "epoch": 37.8,
      "grad_norm": 0.007503041066229343,
      "kl": 0.0021915435791015625,
      "learning_rate": 1.564967003424938e-05,
      "loss": 0.0001,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 189
    },
    {
      "completion_length": 586.0,
      "epoch": 38.0,
      "grad_norm": 0.16468028724193573,
      "kl": 0.0054779052734375,
      "learning_rate": 1.5591929034707468e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 190
    },
    {
      "completion_length": 658.4375,
      "epoch": 38.2,
      "grad_norm": 0.08969616889953613,
      "kl": 0.0038738250732421875,
      "learning_rate": 1.553391549243344e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 191
    },
    {
      "completion_length": 608.90625,
      "epoch": 38.4,
      "grad_norm": 0.19583958387374878,
      "kl": 0.0025424957275390625,
      "learning_rate": 1.5475632234925505e-05,
      "loss": 0.0001,
      "reward": 0.25,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0625,
      "step": 192
    },
    {
      "completion_length": 585.4375,
      "epoch": 38.6,
      "grad_norm": 0.07371523231267929,
      "kl": 0.004715919494628906,
      "learning_rate": 1.54170821028274e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 193
    },
    {
      "completion_length": 789.09375,
      "epoch": 38.8,
      "grad_norm": 0.05235208570957184,
      "kl": 0.003101348876953125,
      "learning_rate": 1.5358267949789968e-05,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 194
    },
    {
      "completion_length": 794.875,
      "epoch": 39.0,
      "grad_norm": 0.042146533727645874,
      "kl": 0.00682830810546875,
      "learning_rate": 1.529919264233205e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 195
    },
    {
      "completion_length": 628.90625,
      "epoch": 39.2,
      "grad_norm": 0.04325362667441368,
      "kl": 0.003551483154296875,
      "learning_rate": 1.5239859059700794e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 196
    },
    {
      "completion_length": 705.9375,
      "epoch": 39.4,
      "grad_norm": 0.05797632038593292,
      "kl": 0.003326416015625,
      "learning_rate": 1.5180270093731305e-05,
      "loss": 0.0001,
      "reward": 0.25,
      "reward_std": 0.3514062538743019,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.09375,
      "step": 197
    },
    {
      "completion_length": 785.375,
      "epoch": 39.6,
      "grad_norm": 0.045176297426223755,
      "kl": 0.0056095123291015625,
      "learning_rate": 1.5120428648705716e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 198
    },
    {
      "completion_length": 761.125,
      "epoch": 39.8,
      "grad_norm": 0.0534488782286644,
      "kl": 0.001953125,
      "learning_rate": 1.5060337641211637e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 199
    },
    {
      "completion_length": 606.0,
      "epoch": 40.0,
      "grad_norm": 0.06155118718743324,
      "kl": 0.007213592529296875,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 200
    },
    {
      "completion_length": 648.1875,
      "epoch": 40.2,
      "grad_norm": 0.06065668910741806,
      "kl": 0.005786895751953125,
      "learning_rate": 1.493941866584231e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 201
    },
    {
      "completion_length": 713.9375,
      "epoch": 40.4,
      "grad_norm": 0.04357614740729332,
      "kl": 0.005336761474609375,
      "learning_rate": 1.4878596591387329e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 202
    },
    {
      "completion_length": 721.46875,
      "epoch": 40.6,
      "grad_norm": 0.03795577213168144,
      "kl": 0.004741668701171875,
      "learning_rate": 1.4817536741017153e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 203
    },
    {
      "completion_length": 761.4375,
      "epoch": 40.8,
      "grad_norm": 0.03983302414417267,
      "kl": 0.003192901611328125,
      "learning_rate": 1.4756242090702756e-05,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 204
    },
    {
      "completion_length": 662.625,
      "epoch": 41.0,
      "grad_norm": 0.05180246755480766,
      "kl": 0.004192352294921875,
      "learning_rate": 1.469471562785891e-05,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.3471825420856476,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.125,
      "step": 205
    },
    {
      "completion_length": 687.65625,
      "epoch": 41.2,
      "grad_norm": 0.03887093439698219,
      "kl": 0.003993988037109375,
      "learning_rate": 1.463296035119862e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 206
    },
    {
      "completion_length": 799.8125,
      "epoch": 41.4,
      "grad_norm": 0.0740731805562973,
      "kl": 0.0067596435546875,
      "learning_rate": 1.4570979270586944e-05,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 207
    },
    {
      "completion_length": 725.75,
      "epoch": 41.6,
      "grad_norm": 0.04128188639879227,
      "kl": 0.0065822601318359375,
      "learning_rate": 1.4508775406894308e-05,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 208
    },
    {
      "completion_length": 652.8125,
      "epoch": 41.8,
      "grad_norm": 0.06086068972945213,
      "kl": 0.00821685791015625,
      "learning_rate": 1.4446351791849276e-05,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 209
    },
    {
      "completion_length": 662.6875,
      "epoch": 42.0,
      "grad_norm": 0.01791488565504551,
      "kl": 0.005237579345703125,
      "learning_rate": 1.4383711467890776e-05,
      "loss": 0.0002,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 210
    },
    {
      "completion_length": 735.96875,
      "epoch": 42.2,
      "grad_norm": 0.028158580884337425,
      "kl": 0.0033512115478515625,
      "learning_rate": 1.4320857488019826e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 211
    },
    {
      "completion_length": 727.5625,
      "epoch": 42.4,
      "grad_norm": 0.057374995201826096,
      "kl": 0.008304595947265625,
      "learning_rate": 1.4257792915650728e-05,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 212
    },
    {
      "completion_length": 610.75,
      "epoch": 42.6,
      "grad_norm": 0.24425353109836578,
      "kl": 0.03440093994140625,
      "learning_rate": 1.4194520824461773e-05,
      "loss": 0.0014,
      "reward": 0.21875,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 213
    },
    {
      "completion_length": 718.1875,
      "epoch": 42.8,
      "grad_norm": 0.042232200503349304,
      "kl": 0.0050792694091796875,
      "learning_rate": 1.413104429824542e-05,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 214
    },
    {
      "completion_length": 798.0625,
      "epoch": 43.0,
      "grad_norm": 0.03367134928703308,
      "kl": 0.00554656982421875,
      "learning_rate": 1.4067366430758004e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 215
    },
    {
      "completion_length": 809.03125,
      "epoch": 43.2,
      "grad_norm": 0.08703216165304184,
      "kl": 0.0069732666015625,
      "learning_rate": 1.4003490325568953e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 216
    },
    {
      "completion_length": 718.71875,
      "epoch": 43.4,
      "grad_norm": 0.059716470539569855,
      "kl": 0.004016876220703125,
      "learning_rate": 1.3939419095909513e-05,
      "loss": 0.0002,
      "reward": 0.28125,
      "reward_std": 0.35564958304166794,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0625,
      "step": 217
    },
    {
      "completion_length": 762.03125,
      "epoch": 43.6,
      "grad_norm": 0.03960856795310974,
      "kl": 0.008747100830078125,
      "learning_rate": 1.3875155864521031e-05,
      "loss": 0.0004,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 218
    },
    {
      "completion_length": 668.28125,
      "epoch": 43.8,
      "grad_norm": 0.02855999954044819,
      "kl": 0.0038890838623046875,
      "learning_rate": 1.3810703763502744e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 219
    },
    {
      "completion_length": 712.375,
      "epoch": 44.0,
      "grad_norm": 0.07352375984191895,
      "kl": 0.013210296630859375,
      "learning_rate": 1.3746065934159123e-05,
      "loss": 0.0005,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 220
    },
    {
      "completion_length": 648.0625,
      "epoch": 44.2,
      "grad_norm": 0.06787135452032089,
      "kl": 0.010593414306640625,
      "learning_rate": 1.3681245526846782e-05,
      "loss": 0.0004,
      "reward": 0.21875,
      "reward_std": 0.4218914955854416,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.125,
      "step": 221
    },
    {
      "completion_length": 692.90625,
      "epoch": 44.4,
      "grad_norm": 0.05384477600455284,
      "kl": 0.004001617431640625,
      "learning_rate": 1.3616245700820922e-05,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3377464786171913,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0,
      "step": 222
    },
    {
      "completion_length": 801.59375,
      "epoch": 44.6,
      "grad_norm": 0.05627313256263733,
      "kl": 0.0066280364990234375,
      "learning_rate": 1.3551069624081372e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 223
    },
    {
      "completion_length": 713.0625,
      "epoch": 44.8,
      "grad_norm": 0.03934725746512413,
      "kl": 0.00460052490234375,
      "learning_rate": 1.3485720473218153e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 224
    },
    {
      "completion_length": 774.75,
      "epoch": 45.0,
      "grad_norm": 0.043650925159454346,
      "kl": 0.006531715393066406,
      "learning_rate": 1.342020143325669e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 225
    },
    {
      "completion_length": 652.8125,
      "epoch": 45.2,
      "grad_norm": 0.0472080297768116,
      "kl": 0.0072612762451171875,
      "learning_rate": 1.3354515697502552e-05,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 226
    },
    {
      "completion_length": 741.59375,
      "epoch": 45.4,
      "grad_norm": 0.0659644827246666,
      "kl": 0.0036468505859375,
      "learning_rate": 1.3288666467385834e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 227
    },
    {
      "completion_length": 763.625,
      "epoch": 45.6,
      "grad_norm": 0.0662030428647995,
      "kl": 0.0071868896484375,
      "learning_rate": 1.3222656952305113e-05,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 228
    },
    {
      "completion_length": 768.84375,
      "epoch": 45.8,
      "grad_norm": 0.05112076178193092,
      "kl": 0.00908660888671875,
      "learning_rate": 1.3156490369471026e-05,
      "loss": 0.0004,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 229
    },
    {
      "completion_length": 638.125,
      "epoch": 46.0,
      "grad_norm": 0.06220560520887375,
      "kl": 0.00731658935546875,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.0003,
      "reward": 0.21875,
      "reward_std": 0.3608423173427582,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 230
    },
    {
      "completion_length": 725.09375,
      "epoch": 46.2,
      "grad_norm": 0.12178517132997513,
      "kl": 0.00751495361328125,
      "learning_rate": 1.3023698907504447e-05,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 231
    },
    {
      "completion_length": 782.40625,
      "epoch": 46.4,
      "grad_norm": 0.04633370414376259,
      "kl": 0.0020923614501953125,
      "learning_rate": 1.2957080500440469e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 232
    },
    {
      "completion_length": 774.6875,
      "epoch": 46.6,
      "grad_norm": 0.06594963371753693,
      "kl": 0.005680084228515625,
      "learning_rate": 1.2890317969444716e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 233
    },
    {
      "completion_length": 691.125,
      "epoch": 46.8,
      "grad_norm": 0.03636356443166733,
      "kl": 0.003635406494140625,
      "learning_rate": 1.2823414568428767e-05,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 234
    },
    {
      "completion_length": 593.4375,
      "epoch": 47.0,
      "grad_norm": 0.08689063787460327,
      "kl": 0.010068893432617188,
      "learning_rate": 1.2756373558169992e-05,
      "loss": 0.0004,
      "reward": 0.3125,
      "reward_std": 0.38298875093460083,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.1875,
      "step": 235
    },
    {
      "completion_length": 659.875,
      "epoch": 47.2,
      "grad_norm": 0.0875919908285141,
      "kl": 0.008754730224609375,
      "learning_rate": 1.2689198206152657e-05,
      "loss": 0.0004,
      "reward": 0.1875,
      "reward_std": 0.3945523276925087,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.1875,
      "step": 236
    },
    {
      "completion_length": 829.03125,
      "epoch": 47.4,
      "grad_norm": 0.03692341968417168,
      "kl": 0.0024633407592773438,
      "learning_rate": 1.2621891786408648e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 237
    },
    {
      "completion_length": 734.40625,
      "epoch": 47.6,
      "grad_norm": 0.04285416752099991,
      "kl": 0.0069580078125,
      "learning_rate": 1.2554457579357906e-05,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 238
    },
    {
      "completion_length": 700.125,
      "epoch": 47.8,
      "grad_norm": 0.06877611577510834,
      "kl": 0.008152008056640625,
      "learning_rate": 1.2486898871648552e-05,
      "loss": 0.0003,
      "reward": 0.28125,
      "reward_std": 0.35564958304166794,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.15625,
      "step": 239
    },
    {
      "completion_length": 654.0625,
      "epoch": 48.0,
      "grad_norm": 0.057643674314022064,
      "kl": 0.00344085693359375,
      "learning_rate": 1.2419218955996677e-05,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 240
    },
    {
      "completion_length": 788.0,
      "epoch": 48.2,
      "grad_norm": 0.06129004433751106,
      "kl": 0.008901596069335938,
      "learning_rate": 1.23514211310259e-05,
      "loss": 0.0004,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 241
    },
    {
      "completion_length": 709.0625,
      "epoch": 48.4,
      "grad_norm": 0.04215136915445328,
      "kl": 0.00811004638671875,
      "learning_rate": 1.2283508701106559e-05,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 242
    },
    {
      "completion_length": 768.75,
      "epoch": 48.6,
      "grad_norm": 0.08925128728151321,
      "kl": 0.002838134765625,
      "learning_rate": 1.2215484976194675e-05,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 243
    },
    {
      "completion_length": 743.59375,
      "epoch": 48.8,
      "grad_norm": 0.040025435388088226,
      "kl": 0.0040283203125,
      "learning_rate": 1.2147353271670634e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 244
    },
    {
      "completion_length": 562.875,
      "epoch": 49.0,
      "grad_norm": 0.072207011282444,
      "kl": 0.0041351318359375,
      "learning_rate": 1.2079116908177592e-05,
      "loss": 0.0002,
      "reward": 0.3125,
      "reward_std": 0.3745020925998688,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.1875,
      "step": 245
    },
    {
      "completion_length": 757.75,
      "epoch": 49.2,
      "grad_norm": 0.05225052312016487,
      "kl": 0.0013132095336914062,
      "learning_rate": 1.2010779211459649e-05,
      "loss": 0.0001,
      "reward": 0.25,
      "reward_std": 0.3514062538743019,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.0,
      "step": 246
    },
    {
      "completion_length": 717.96875,
      "epoch": 49.4,
      "grad_norm": 0.031644195318222046,
      "kl": 0.0064697265625,
      "learning_rate": 1.194234351219972e-05,
      "loss": 0.0003,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 247
    },
    {
      "completion_length": 689.21875,
      "epoch": 49.6,
      "grad_norm": 0.06784337013959885,
      "kl": 0.00656890869140625,
      "learning_rate": 1.187381314585725e-05,
      "loss": 0.0003,
      "reward": 0.375,
      "reward_std": 0.4629100561141968,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.15625,
      "step": 248
    },
    {
      "completion_length": 747.375,
      "epoch": 49.8,
      "grad_norm": 0.033385973423719406,
      "kl": 0.006641387939453125,
      "learning_rate": 1.1805191452505602e-05,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 249
    },
    {
      "completion_length": 778.5,
      "epoch": 50.0,
      "grad_norm": 0.039894282817840576,
      "kl": 0.00530242919921875,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 250
    },
    {
      "completion_length": 664.9375,
      "epoch": 50.2,
      "grad_norm": 0.01323878113180399,
      "kl": 0.0035419464111328125,
      "learning_rate": 1.1667687467161025e-05,
      "loss": 0.0001,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 251
    },
    {
      "completion_length": 784.5,
      "epoch": 50.4,
      "grad_norm": 0.04773200675845146,
      "kl": 0.004486083984375,
      "learning_rate": 1.159881187691835e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 252
    },
    {
      "completion_length": 681.25,
      "epoch": 50.6,
      "grad_norm": 0.0516289658844471,
      "kl": 0.0051116943359375,
      "learning_rate": 1.1529858362840383e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 253
    },
    {
      "completion_length": 719.25,
      "epoch": 50.8,
      "grad_norm": 0.0592305026948452,
      "kl": 0.006103515625,
      "learning_rate": 1.1460830285624119e-05,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.125,
      "step": 254
    },
    {
      "completion_length": 713.375,
      "epoch": 51.0,
      "grad_norm": 0.060044627636671066,
      "kl": 0.01081085205078125,
      "learning_rate": 1.1391731009600655e-05,
      "loss": 0.0004,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 255
    },
    {
      "completion_length": 727.59375,
      "epoch": 51.2,
      "grad_norm": 0.0568021722137928,
      "kl": 0.0041046142578125,
      "learning_rate": 1.1322563902571227e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 256
    },
    {
      "completion_length": 747.21875,
      "epoch": 51.4,
      "grad_norm": 0.05668547376990318,
      "kl": 0.007038116455078125,
      "learning_rate": 1.1253332335643043e-05,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 257
    },
    {
      "completion_length": 710.59375,
      "epoch": 51.6,
      "grad_norm": 0.06153063476085663,
      "kl": 0.0038509368896484375,
      "learning_rate": 1.1184039683065014e-05,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3377464786171913,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0,
      "step": 258
    },
    {
      "completion_length": 716.8125,
      "epoch": 51.8,
      "grad_norm": 0.05457403510808945,
      "kl": 0.005252838134765625,
      "learning_rate": 1.1114689322063255e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 259
    },
    {
      "completion_length": 797.9375,
      "epoch": 52.0,
      "grad_norm": 0.046778466552495956,
      "kl": 0.0037317276000976562,
      "learning_rate": 1.1045284632676535e-05,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 260
    },
    {
      "completion_length": 745.5,
      "epoch": 52.2,
      "grad_norm": 0.042898040264844894,
      "kl": 0.004665374755859375,
      "learning_rate": 1.0975828997591496e-05,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.03125,
      "step": 261
    },
    {
      "completion_length": 659.28125,
      "epoch": 52.4,
      "grad_norm": 0.08597607165575027,
      "kl": 0.0042476654052734375,
      "learning_rate": 1.0906325801977804e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 262
    },
    {
      "completion_length": 754.375,
      "epoch": 52.6,
      "grad_norm": 0.03376508876681328,
      "kl": 0.006850242614746094,
      "learning_rate": 1.083677843332316e-05,
      "loss": 0.0003,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 263
    },
    {
      "completion_length": 835.71875,
      "epoch": 52.8,
      "grad_norm": 0.04447947070002556,
      "kl": 0.004009246826171875,
      "learning_rate": 1.0767190281268187e-05,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 264
    },
    {
      "completion_length": 605.0625,
      "epoch": 53.0,
      "grad_norm": 0.011051284149289131,
      "kl": 0.003787994384765625,
      "learning_rate": 1.0697564737441254e-05,
      "loss": 0.0002,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 265
    },
    {
      "completion_length": 741.125,
      "epoch": 53.2,
      "grad_norm": 0.06511937826871872,
      "kl": 0.00412750244140625,
      "learning_rate": 1.0627905195293135e-05,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3377464786171913,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 266
    },
    {
      "completion_length": 680.6875,
      "epoch": 53.4,
      "grad_norm": 0.0632292777299881,
      "kl": 0.0049114227294921875,
      "learning_rate": 1.055821504993164e-05,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 267
    },
    {
      "completion_length": 737.59375,
      "epoch": 53.6,
      "grad_norm": 0.06620203703641891,
      "kl": 0.0019378662109375,
      "learning_rate": 1.0488497697956134e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 268
    },
    {
      "completion_length": 784.1875,
      "epoch": 53.8,
      "grad_norm": 0.032115936279296875,
      "kl": 0.0035762786865234375,
      "learning_rate": 1.0418756537291996e-05,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 269
    },
    {
      "completion_length": 699.0,
      "epoch": 54.0,
      "grad_norm": 0.05448479577898979,
      "kl": 0.0056209564208984375,
      "learning_rate": 1.0348994967025012e-05,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 270
    },
    {
      "completion_length": 714.5,
      "epoch": 54.2,
      "grad_norm": 0.011306805536150932,
      "kl": 0.0034999847412109375,
      "learning_rate": 1.0279216387235691e-05,
      "loss": 0.0001,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 271
    },
    {
      "completion_length": 628.3125,
      "epoch": 54.4,
      "grad_norm": 0.0956725999712944,
      "kl": 0.004085540771484375,
      "learning_rate": 1.0209424198833571e-05,
      "loss": 0.0002,
      "reward": 0.25,
      "reward_std": 0.4492306634783745,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.03125,
      "step": 272
    },
    {
      "completion_length": 685.78125,
      "epoch": 54.6,
      "grad_norm": 0.047145962715148926,
      "kl": 0.0092620849609375,
      "learning_rate": 1.0139621803391454e-05,
      "loss": 0.0004,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 273
    },
    {
      "completion_length": 674.84375,
      "epoch": 54.8,
      "grad_norm": 0.05443866550922394,
      "kl": 0.00234222412109375,
      "learning_rate": 1.0069812602979617e-05,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 274
    },
    {
      "completion_length": 757.0625,
      "epoch": 55.0,
      "grad_norm": 0.0682177022099495,
      "kl": 0.0035009384155273438,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3471825420856476,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.1875,
      "step": 275
    },
    {
      "completion_length": 818.875,
      "epoch": 55.2,
      "grad_norm": 0.05054046958684921,
      "kl": 0.002574920654296875,
      "learning_rate": 9.930187397020385e-06,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3945523276925087,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 276
    },
    {
      "completion_length": 679.1875,
      "epoch": 55.4,
      "grad_norm": 0.036176227033138275,
      "kl": 0.0063190460205078125,
      "learning_rate": 9.860378196608549e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 277
    },
    {
      "completion_length": 638.375,
      "epoch": 55.6,
      "grad_norm": 0.044918473809957504,
      "kl": 0.004294395446777344,
      "learning_rate": 9.790575801166432e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 278
    },
    {
      "completion_length": 730.65625,
      "epoch": 55.8,
      "grad_norm": 0.03417769446969032,
      "kl": 0.005695343017578125,
      "learning_rate": 9.720783612764314e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 279
    },
    {
      "completion_length": 729.4375,
      "epoch": 56.0,
      "grad_norm": 0.03874381631612778,
      "kl": 0.004076480865478516,
      "learning_rate": 9.651005032974994e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 280
    },
    {
      "completion_length": 735.1875,
      "epoch": 56.2,
      "grad_norm": 0.04026850312948227,
      "kl": 0.005176544189453125,
      "learning_rate": 9.581243462708007e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 281
    },
    {
      "completion_length": 741.5625,
      "epoch": 56.4,
      "grad_norm": 0.08385323733091354,
      "kl": 0.00628662109375,
      "learning_rate": 9.511502302043867e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 282
    },
    {
      "completion_length": 620.4375,
      "epoch": 56.6,
      "grad_norm": 0.0641460195183754,
      "kl": 0.00861358642578125,
      "learning_rate": 9.441784950068362e-06,
      "loss": 0.0003,
      "reward": 0.28125,
      "reward_std": 0.3608423173427582,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0625,
      "step": 283
    },
    {
      "completion_length": 735.75,
      "epoch": 56.8,
      "grad_norm": 0.06547683477401733,
      "kl": 0.0036602020263671875,
      "learning_rate": 9.372094804706867e-06,
      "loss": 0.0001,
      "reward": 0.28125,
      "reward_std": 0.3608423173427582,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.09375,
      "step": 284
    },
    {
      "completion_length": 807.6875,
      "epoch": 57.0,
      "grad_norm": 0.038492776453495026,
      "kl": 0.0035247802734375,
      "learning_rate": 9.302435262558748e-06,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 285
    },
    {
      "completion_length": 740.21875,
      "epoch": 57.2,
      "grad_norm": 0.05591354891657829,
      "kl": 0.003871917724609375,
      "learning_rate": 9.232809718731815e-06,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3377464786171913,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 286
    },
    {
      "completion_length": 731.875,
      "epoch": 57.4,
      "grad_norm": 0.09159969538450241,
      "kl": 0.0052947998046875,
      "learning_rate": 9.163221566676847e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 287
    },
    {
      "completion_length": 676.84375,
      "epoch": 57.6,
      "grad_norm": 0.0690065249800682,
      "kl": 0.0053958892822265625,
      "learning_rate": 9.093674198022201e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 288
    },
    {
      "completion_length": 684.1875,
      "epoch": 57.8,
      "grad_norm": 0.04371067136526108,
      "kl": 0.0026721954345703125,
      "learning_rate": 9.024171002408507e-06,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 289
    },
    {
      "completion_length": 622.375,
      "epoch": 58.0,
      "grad_norm": 0.04673806577920914,
      "kl": 0.00656890869140625,
      "learning_rate": 8.954715367323468e-06,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 290
    },
    {
      "completion_length": 761.25,
      "epoch": 58.2,
      "grad_norm": 0.06499247997999191,
      "kl": 0.004596710205078125,
      "learning_rate": 8.885310677936746e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 291
    },
    {
      "completion_length": 731.25,
      "epoch": 58.4,
      "grad_norm": 0.045903149992227554,
      "kl": 0.006435394287109375,
      "learning_rate": 8.815960316934991e-06,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 292
    },
    {
      "completion_length": 724.4375,
      "epoch": 58.6,
      "grad_norm": 0.04144304245710373,
      "kl": 0.0032796859741210938,
      "learning_rate": 8.746667664356957e-06,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 293
    },
    {
      "completion_length": 638.6875,
      "epoch": 58.8,
      "grad_norm": 0.0686144009232521,
      "kl": 0.007534027099609375,
      "learning_rate": 8.677436097428775e-06,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 294
    },
    {
      "completion_length": 770.125,
      "epoch": 59.0,
      "grad_norm": 0.04020367190241814,
      "kl": 0.00449371337890625,
      "learning_rate": 8.60826899039935e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 295
    },
    {
      "completion_length": 550.59375,
      "epoch": 59.2,
      "grad_norm": 0.07181453704833984,
      "kl": 0.014190673828125,
      "learning_rate": 8.539169714375885e-06,
      "loss": 0.0006,
      "reward": 0.1875,
      "reward_std": 0.249358132481575,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 296
    },
    {
      "completion_length": 785.3125,
      "epoch": 59.4,
      "grad_norm": 0.043180983513593674,
      "kl": 0.0039882659912109375,
      "learning_rate": 8.47014163715962e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.13363061845302582,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 297
    },
    {
      "completion_length": 756.9375,
      "epoch": 59.6,
      "grad_norm": 0.041922781616449356,
      "kl": 0.0078277587890625,
      "learning_rate": 8.401188123081653e-06,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 298
    },
    {
      "completion_length": 720.21875,
      "epoch": 59.8,
      "grad_norm": 0.05826785787940025,
      "kl": 0.0022363662719726562,
      "learning_rate": 8.332312532838978e-06,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 299
    },
    {
      "completion_length": 755.5,
      "epoch": 60.0,
      "grad_norm": 0.0637492761015892,
      "kl": 0.005036354064941406,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3808925524353981,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 300
    },
    {
      "completion_length": 615.53125,
      "epoch": 60.2,
      "grad_norm": 0.054740484803915024,
      "kl": 0.01085662841796875,
      "learning_rate": 8.194808547494401e-06,
      "loss": 0.0004,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 301
    },
    {
      "completion_length": 678.96875,
      "epoch": 60.4,
      "grad_norm": 0.03264128044247627,
      "kl": 0.00649261474609375,
      "learning_rate": 8.126186854142752e-06,
      "loss": 0.0003,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 302
    },
    {
      "completion_length": 731.40625,
      "epoch": 60.6,
      "grad_norm": 0.07293147593736649,
      "kl": 0.0033702850341796875,
      "learning_rate": 8.057656487800283e-06,
      "loss": 0.0001,
      "reward": 0.3125,
      "reward_std": 0.3514062538743019,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.1875,
      "step": 303
    },
    {
      "completion_length": 719.84375,
      "epoch": 60.8,
      "grad_norm": 0.04588692635297775,
      "kl": 0.003902435302734375,
      "learning_rate": 7.989220788540356e-06,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 304
    },
    {
      "completion_length": 795.375,
      "epoch": 61.0,
      "grad_norm": 0.042266327887773514,
      "kl": 0.00453948974609375,
      "learning_rate": 7.92088309182241e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 305
    },
    {
      "completion_length": 609.5,
      "epoch": 61.2,
      "grad_norm": 0.054133933037519455,
      "kl": 0.009552001953125,
      "learning_rate": 7.852646728329368e-06,
      "loss": 0.0004,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0,
      "step": 306
    },
    {
      "completion_length": 747.625,
      "epoch": 61.4,
      "grad_norm": 0.03229310363531113,
      "kl": 0.003002166748046875,
      "learning_rate": 7.784515023805328e-06,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0,
      "step": 307
    },
    {
      "completion_length": 826.34375,
      "epoch": 61.6,
      "grad_norm": 0.04594132676720619,
      "kl": 0.005923271179199219,
      "learning_rate": 7.716491298893443e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.125,
      "step": 308
    },
    {
      "completion_length": 705.78125,
      "epoch": 61.8,
      "grad_norm": 0.037182100117206573,
      "kl": 0.0042591094970703125,
      "learning_rate": 7.6485788689741e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 309
    },
    {
      "completion_length": 776.0,
      "epoch": 62.0,
      "grad_norm": 0.04344649985432625,
      "kl": 0.0043048858642578125,
      "learning_rate": 7.580781044003324e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 310
    },
    {
      "completion_length": 673.53125,
      "epoch": 62.2,
      "grad_norm": 0.04054214432835579,
      "kl": 0.00882720947265625,
      "learning_rate": 7.513101128351454e-06,
      "loss": 0.0004,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 311
    },
    {
      "completion_length": 765.5625,
      "epoch": 62.4,
      "grad_norm": 0.057973332703113556,
      "kl": 0.002285003662109375,
      "learning_rate": 7.445542420642097e-06,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 312
    },
    {
      "completion_length": 553.28125,
      "epoch": 62.6,
      "grad_norm": 0.14677725732326508,
      "kl": 0.009618759155273438,
      "learning_rate": 7.378108213591355e-06,
      "loss": 0.0004,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.125,
      "step": 313
    },
    {
      "completion_length": 815.0,
      "epoch": 62.8,
      "grad_norm": 0.06965812295675278,
      "kl": 0.007441520690917969,
      "learning_rate": 7.310801793847344e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 314
    },
    {
      "completion_length": 696.75,
      "epoch": 63.0,
      "grad_norm": 0.07354584336280823,
      "kl": 0.004512786865234375,
      "learning_rate": 7.243626441830009e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 315
    },
    {
      "completion_length": 800.84375,
      "epoch": 63.2,
      "grad_norm": 0.025326721370220184,
      "kl": 0.00466156005859375,
      "learning_rate": 7.176585431571235e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 316
    },
    {
      "completion_length": 620.875,
      "epoch": 63.4,
      "grad_norm": 0.09194564074277878,
      "kl": 0.008543014526367188,
      "learning_rate": 7.109682030555283e-06,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 317
    },
    {
      "completion_length": 775.125,
      "epoch": 63.6,
      "grad_norm": 0.049211449921131134,
      "kl": 0.0024623870849609375,
      "learning_rate": 7.042919499559538e-06,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 318
    },
    {
      "completion_length": 611.71875,
      "epoch": 63.8,
      "grad_norm": 0.0728549137711525,
      "kl": 0.0084991455078125,
      "learning_rate": 6.976301092495556e-06,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 319
    },
    {
      "completion_length": 763.5625,
      "epoch": 64.0,
      "grad_norm": 0.05590234696865082,
      "kl": 0.0051479339599609375,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3471629247069359,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.125,
      "step": 320
    },
    {
      "completion_length": 720.03125,
      "epoch": 64.2,
      "grad_norm": 0.06959515064954758,
      "kl": 0.004337310791015625,
      "learning_rate": 6.843509630528977e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 321
    },
    {
      "completion_length": 721.0625,
      "epoch": 64.4,
      "grad_norm": 0.03896469622850418,
      "kl": 0.0036754608154296875,
      "learning_rate": 6.777343047694891e-06,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 322
    },
    {
      "completion_length": 554.03125,
      "epoch": 64.6,
      "grad_norm": 0.056396856904029846,
      "kl": 0.00882720947265625,
      "learning_rate": 6.711333532614168e-06,
      "loss": 0.0004,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 323
    },
    {
      "completion_length": 808.28125,
      "epoch": 64.8,
      "grad_norm": 0.14124415814876556,
      "kl": 0.0091400146484375,
      "learning_rate": 6.645484302497452e-06,
      "loss": 0.0004,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 324
    },
    {
      "completion_length": 799.125,
      "epoch": 65.0,
      "grad_norm": 0.060791462659835815,
      "kl": 0.0049304962158203125,
      "learning_rate": 6.579798566743314e-06,
      "loss": 0.0002,
      "reward": 0.25,
      "reward_std": 0.3650856465101242,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.09375,
      "step": 325
    },
    {
      "completion_length": 793.5,
      "epoch": 65.2,
      "grad_norm": 0.04707815870642662,
      "kl": 0.004053592681884766,
      "learning_rate": 6.5142795267818505e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 326
    },
    {
      "completion_length": 672.71875,
      "epoch": 65.4,
      "grad_norm": 0.06408388912677765,
      "kl": 0.005802154541015625,
      "learning_rate": 6.448930375918632e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 327
    },
    {
      "completion_length": 639.4375,
      "epoch": 65.6,
      "grad_norm": 0.08114010840654373,
      "kl": 0.007213592529296875,
      "learning_rate": 6.383754299179079e-06,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 328
    },
    {
      "completion_length": 701.84375,
      "epoch": 65.8,
      "grad_norm": 0.06492974609136581,
      "kl": 0.006916999816894531,
      "learning_rate": 6.318754473153221e-06,
      "loss": 0.0003,
      "reward": 0.375,
      "reward_std": 0.4355512708425522,
      "rewards/accuracy_reward": 0.34375,
      "rewards/format_reward": 0.03125,
      "step": 329
    },
    {
      "completion_length": 848.9375,
      "epoch": 66.0,
      "grad_norm": 0.05233583599328995,
      "kl": 0.00559234619140625,
      "learning_rate": 6.25393406584088e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 330
    },
    {
      "completion_length": 700.375,
      "epoch": 66.2,
      "grad_norm": 0.029833296313881874,
      "kl": 0.0030221939086914062,
      "learning_rate": 6.18929623649726e-06,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 331
    },
    {
      "completion_length": 638.84375,
      "epoch": 66.4,
      "grad_norm": 0.07965300232172012,
      "kl": 0.005283355712890625,
      "learning_rate": 6.124844135478971e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 332
    },
    {
      "completion_length": 695.90625,
      "epoch": 66.6,
      "grad_norm": 0.13027136027812958,
      "kl": 0.00688934326171875,
      "learning_rate": 6.06058090409049e-06,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 333
    },
    {
      "completion_length": 760.15625,
      "epoch": 66.8,
      "grad_norm": 0.04611360654234886,
      "kl": 0.003757476806640625,
      "learning_rate": 5.996509674431053e-06,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 334
    },
    {
      "completion_length": 764.375,
      "epoch": 67.0,
      "grad_norm": 0.07783949375152588,
      "kl": 0.006290435791015625,
      "learning_rate": 5.932633569242e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 335
    },
    {
      "completion_length": 708.5,
      "epoch": 67.2,
      "grad_norm": 0.044424816966056824,
      "kl": 0.0040302276611328125,
      "learning_rate": 5.868955701754584e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 336
    },
    {
      "completion_length": 784.625,
      "epoch": 67.4,
      "grad_norm": 0.06684647500514984,
      "kl": 0.00331878662109375,
      "learning_rate": 5.8054791755382286e-06,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 337
    },
    {
      "completion_length": 722.09375,
      "epoch": 67.6,
      "grad_norm": 0.04156530275940895,
      "kl": 0.00507354736328125,
      "learning_rate": 5.742207084349274e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 338
    },
    {
      "completion_length": 870.40625,
      "epoch": 67.8,
      "grad_norm": 0.030239231884479523,
      "kl": 0.004021644592285156,
      "learning_rate": 5.679142511980176e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.13363061845302582,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 339
    },
    {
      "completion_length": 620.625,
      "epoch": 68.0,
      "grad_norm": 0.07272662222385406,
      "kl": 0.00936126708984375,
      "learning_rate": 5.616288532109225e-06,
      "loss": 0.0004,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 340
    },
    {
      "completion_length": 786.21875,
      "epoch": 68.2,
      "grad_norm": 0.04447874799370766,
      "kl": 0.00745391845703125,
      "learning_rate": 5.553648208150728e-06,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 341
    },
    {
      "completion_length": 609.71875,
      "epoch": 68.4,
      "grad_norm": 0.04590880125761032,
      "kl": 0.01007843017578125,
      "learning_rate": 5.491224593105695e-06,
      "loss": 0.0004,
      "reward": 0.21875,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0,
      "step": 342
    },
    {
      "completion_length": 858.71875,
      "epoch": 68.6,
      "grad_norm": 0.04522442817687988,
      "kl": 0.003849029541015625,
      "learning_rate": 5.429020729413062e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 343
    },
    {
      "completion_length": 684.34375,
      "epoch": 68.8,
      "grad_norm": 0.06102772057056427,
      "kl": 0.00911712646484375,
      "learning_rate": 5.367039648801386e-06,
      "loss": 0.0004,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 344
    },
    {
      "completion_length": 678.5,
      "epoch": 69.0,
      "grad_norm": 0.046152278780937195,
      "kl": 0.0076751708984375,
      "learning_rate": 5.305284372141095e-06,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.13363061845302582,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 345
    },
    {
      "completion_length": 766.46875,
      "epoch": 69.2,
      "grad_norm": 0.040787436068058014,
      "kl": 0.004444122314453125,
      "learning_rate": 5.243757909297247e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 346
    },
    {
      "completion_length": 700.53125,
      "epoch": 69.4,
      "grad_norm": 0.06111907958984375,
      "kl": 0.005175590515136719,
      "learning_rate": 5.1824632589828465e-06,
      "loss": 0.0002,
      "reward": 0.28125,
      "reward_std": 0.3471629247069359,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.03125,
      "step": 347
    },
    {
      "completion_length": 632.34375,
      "epoch": 69.6,
      "grad_norm": 0.05885278806090355,
      "kl": 0.005603790283203125,
      "learning_rate": 5.121403408612672e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 348
    },
    {
      "completion_length": 705.09375,
      "epoch": 69.8,
      "grad_norm": 0.033596571534872055,
      "kl": 0.00606536865234375,
      "learning_rate": 5.060581334157693e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 349
    },
    {
      "completion_length": 665.5,
      "epoch": 70.0,
      "grad_norm": 0.060441531240940094,
      "kl": 0.00395965576171875,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 350
    },
    {
      "completion_length": 699.8125,
      "epoch": 70.2,
      "grad_norm": 0.19340521097183228,
      "kl": 0.004101753234863281,
      "learning_rate": 4.939662358788364e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 351
    },
    {
      "completion_length": 673.28125,
      "epoch": 70.4,
      "grad_norm": 0.09181207418441772,
      "kl": 0.009540557861328125,
      "learning_rate": 4.879571351294287e-06,
      "loss": 0.0004,
      "reward": 0.25,
      "reward_std": 0.3514062538743019,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.03125,
      "step": 352
    },
    {
      "completion_length": 740.53125,
      "epoch": 70.6,
      "grad_norm": 0.04517640545964241,
      "kl": 0.0024280548095703125,
      "learning_rate": 4.8197299062687e-06,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 353
    },
    {
      "completion_length": 610.71875,
      "epoch": 70.8,
      "grad_norm": 0.06916794180870056,
      "kl": 0.007266998291015625,
      "learning_rate": 4.76014094029921e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 354
    },
    {
      "completion_length": 751.9375,
      "epoch": 71.0,
      "grad_norm": 0.05887198820710182,
      "kl": 0.00473785400390625,
      "learning_rate": 4.700807357667953e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.125,
      "step": 355
    },
    {
      "completion_length": 707.3125,
      "epoch": 71.2,
      "grad_norm": 0.1130240187048912,
      "kl": 0.007659912109375,
      "learning_rate": 4.641732050210032e-06,
      "loss": 0.0003,
      "reward": 0.3125,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.125,
      "step": 356
    },
    {
      "completion_length": 682.5625,
      "epoch": 71.4,
      "grad_norm": 0.050379782915115356,
      "kl": 0.00505828857421875,
      "learning_rate": 4.582917897172603e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 357
    },
    {
      "completion_length": 696.1875,
      "epoch": 71.6,
      "grad_norm": 0.056728295981884,
      "kl": 0.002349853515625,
      "learning_rate": 4.524367765074499e-06,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 358
    },
    {
      "completion_length": 741.21875,
      "epoch": 71.8,
      "grad_norm": 0.05675540864467621,
      "kl": 0.0052700042724609375,
      "learning_rate": 4.46608450756656e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 359
    },
    {
      "completion_length": 714.0625,
      "epoch": 72.0,
      "grad_norm": 0.039005063474178314,
      "kl": 0.005115509033203125,
      "learning_rate": 4.408070965292534e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 360
    },
    {
      "completion_length": 730.5625,
      "epoch": 72.2,
      "grad_norm": 0.05122857168316841,
      "kl": 0.007793426513671875,
      "learning_rate": 4.350329965750622e-06,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 361
    },
    {
      "completion_length": 740.375,
      "epoch": 72.4,
      "grad_norm": 0.06879931688308716,
      "kl": 0.003704547882080078,
      "learning_rate": 4.292864323155684e-06,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 362
    },
    {
      "completion_length": 695.0,
      "epoch": 72.6,
      "grad_norm": 0.049194369465112686,
      "kl": 0.004791259765625,
      "learning_rate": 4.235676838302069e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 363
    },
    {
      "completion_length": 748.375,
      "epoch": 72.8,
      "grad_norm": 0.04388902708888054,
      "kl": 0.00705718994140625,
      "learning_rate": 4.178770298427107e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 364
    },
    {
      "completion_length": 724.375,
      "epoch": 73.0,
      "grad_norm": 0.0331445075571537,
      "kl": 0.004261016845703125,
      "learning_rate": 4.12214747707527e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 365
    },
    {
      "completion_length": 775.90625,
      "epoch": 73.2,
      "grad_norm": 0.07193587720394135,
      "kl": 0.006439208984375,
      "learning_rate": 4.065811133962987e-06,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.408231720328331,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.125,
      "step": 366
    },
    {
      "completion_length": 698.34375,
      "epoch": 73.4,
      "grad_norm": 0.05865119770169258,
      "kl": 0.0106048583984375,
      "learning_rate": 4.009764014844143e-06,
      "loss": 0.0004,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 367
    },
    {
      "completion_length": 744.4375,
      "epoch": 73.6,
      "grad_norm": 0.05057159438729286,
      "kl": 0.0068912506103515625,
      "learning_rate": 3.954008851376252e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 368
    },
    {
      "completion_length": 759.40625,
      "epoch": 73.8,
      "grad_norm": 0.06823986768722534,
      "kl": 0.004940032958984375,
      "learning_rate": 3.898548360987325e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 369
    },
    {
      "completion_length": 724.8125,
      "epoch": 74.0,
      "grad_norm": 0.00893895048648119,
      "kl": 0.00299835205078125,
      "learning_rate": 3.8433852467434175e-06,
      "loss": 0.0001,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 370
    },
    {
      "completion_length": 781.625,
      "epoch": 74.2,
      "grad_norm": 0.044992513954639435,
      "kl": 0.0033054351806640625,
      "learning_rate": 3.7885221972168974e-06,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 371
    },
    {
      "completion_length": 725.5,
      "epoch": 74.4,
      "grad_norm": 0.10380952805280685,
      "kl": 0.00518035888671875,
      "learning_rate": 3.7339618863553983e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 372
    },
    {
      "completion_length": 699.15625,
      "epoch": 74.6,
      "grad_norm": 0.07051531225442886,
      "kl": 0.009708404541015625,
      "learning_rate": 3.679706973351491e-06,
      "loss": 0.0004,
      "reward": 0.21875,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 373
    },
    {
      "completion_length": 729.96875,
      "epoch": 74.8,
      "grad_norm": 0.09161454439163208,
      "kl": 0.00778961181640625,
      "learning_rate": 3.625760102513103e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 374
    },
    {
      "completion_length": 682.6875,
      "epoch": 75.0,
      "grad_norm": 0.03867654129862785,
      "kl": 0.00751495361328125,
      "learning_rate": 3.5721239031346067e-06,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 375
    },
    {
      "completion_length": 734.5,
      "epoch": 75.2,
      "grad_norm": 0.07278414070606232,
      "kl": 0.00582122802734375,
      "learning_rate": 3.5188009893686916e-06,
      "loss": 0.0002,
      "reward": 0.25,
      "reward_std": 0.4355512708425522,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.15625,
      "step": 376
    },
    {
      "completion_length": 716.5625,
      "epoch": 75.4,
      "grad_norm": 0.08173063397407532,
      "kl": 0.006717681884765625,
      "learning_rate": 3.4657939600989453e-06,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 377
    },
    {
      "completion_length": 672.71875,
      "epoch": 75.6,
      "grad_norm": 0.05544436722993851,
      "kl": 0.010814666748046875,
      "learning_rate": 3.4131053988131947e-06,
      "loss": 0.0004,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 378
    },
    {
      "completion_length": 679.375,
      "epoch": 75.8,
      "grad_norm": 0.04979521036148071,
      "kl": 0.013836860656738281,
      "learning_rate": 3.360737873477584e-06,
      "loss": 0.0006,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 379
    },
    {
      "completion_length": 833.0625,
      "epoch": 76.0,
      "grad_norm": 0.07625052332878113,
      "kl": 0.00481414794921875,
      "learning_rate": 3.308693936411421e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 380
    },
    {
      "completion_length": 692.78125,
      "epoch": 76.2,
      "grad_norm": 0.04101628437638283,
      "kl": 0.010272979736328125,
      "learning_rate": 3.2569761241627694e-06,
      "loss": 0.0004,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 381
    },
    {
      "completion_length": 740.8125,
      "epoch": 76.4,
      "grad_norm": 0.04744882509112358,
      "kl": 0.003627777099609375,
      "learning_rate": 3.2055869573848374e-06,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 382
    },
    {
      "completion_length": 705.40625,
      "epoch": 76.6,
      "grad_norm": 0.0408770926296711,
      "kl": 0.0054168701171875,
      "learning_rate": 3.1545289407131128e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 383
    },
    {
      "completion_length": 764.25,
      "epoch": 76.8,
      "grad_norm": 0.053316760808229446,
      "kl": 0.004515647888183594,
      "learning_rate": 3.103804562643302e-06,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 384
    },
    {
      "completion_length": 755.8125,
      "epoch": 77.0,
      "grad_norm": 0.1009504497051239,
      "kl": 0.015666961669921875,
      "learning_rate": 3.0534162954100264e-06,
      "loss": 0.0006,
      "reward": 0.25,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0625,
      "step": 385
    },
    {
      "completion_length": 747.53125,
      "epoch": 77.2,
      "grad_norm": 0.06267711520195007,
      "kl": 0.0065822601318359375,
      "learning_rate": 3.003366594866345e-06,
      "loss": 0.0003,
      "reward": 0.21875,
      "reward_std": 0.3377464786171913,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 386
    },
    {
      "completion_length": 754.53125,
      "epoch": 77.4,
      "grad_norm": 0.03025928884744644,
      "kl": 0.00618743896484375,
      "learning_rate": 2.953657900364053e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 387
    },
    {
      "completion_length": 648.21875,
      "epoch": 77.6,
      "grad_norm": 0.18330256640911102,
      "kl": 0.025907516479492188,
      "learning_rate": 2.9042926346347932e-06,
      "loss": 0.001,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 388
    },
    {
      "completion_length": 588.59375,
      "epoch": 77.8,
      "grad_norm": 0.03385114669799805,
      "kl": 0.010234832763671875,
      "learning_rate": 2.855273203671969e-06,
      "loss": 0.0004,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0,
      "step": 389
    },
    {
      "completion_length": 733.75,
      "epoch": 78.0,
      "grad_norm": 0.03537575900554657,
      "kl": 0.002895355224609375,
      "learning_rate": 2.8066019966134907e-06,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 390
    },
    {
      "completion_length": 701.375,
      "epoch": 78.2,
      "grad_norm": 0.03745219111442566,
      "kl": 0.0038595199584960938,
      "learning_rate": 2.7582813856253276e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 391
    },
    {
      "completion_length": 736.21875,
      "epoch": 78.4,
      "grad_norm": 0.05290169268846512,
      "kl": 0.004032135009765625,
      "learning_rate": 2.7103137257858867e-06,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3377464786171913,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 392
    },
    {
      "completion_length": 672.46875,
      "epoch": 78.6,
      "grad_norm": 0.06984294950962067,
      "kl": 0.008174896240234375,
      "learning_rate": 2.6627013549712355e-06,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 393
    },
    {
      "completion_length": 685.625,
      "epoch": 78.8,
      "grad_norm": 0.12042836844921112,
      "kl": 0.00705718994140625,
      "learning_rate": 2.615446593741161e-06,
      "loss": 0.0003,
      "reward": 0.21875,
      "reward_std": 0.4218914955854416,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.125,
      "step": 394
    },
    {
      "completion_length": 872.375,
      "epoch": 79.0,
      "grad_norm": 0.047877345234155655,
      "kl": 0.005802154541015625,
      "learning_rate": 2.5685517452260566e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 395
    },
    {
      "completion_length": 723.78125,
      "epoch": 79.2,
      "grad_norm": 0.08731101453304291,
      "kl": 0.0076141357421875,
      "learning_rate": 2.522019095014683e-06,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 396
    },
    {
      "completion_length": 682.75,
      "epoch": 79.4,
      "grad_norm": 0.0575302317738533,
      "kl": 0.007720470428466797,
      "learning_rate": 2.4758509110427576e-06,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 397
    },
    {
      "completion_length": 788.3125,
      "epoch": 79.6,
      "grad_norm": 0.04634345695376396,
      "kl": 0.005237579345703125,
      "learning_rate": 2.4300494434824373e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 398
    },
    {
      "completion_length": 715.0625,
      "epoch": 79.8,
      "grad_norm": 0.06102108210325241,
      "kl": 0.00475311279296875,
      "learning_rate": 2.3846169246326345e-06,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3608423173427582,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 399
    },
    {
      "completion_length": 703.0625,
      "epoch": 80.0,
      "grad_norm": 0.042385708540678024,
      "kl": 0.008289337158203125,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.0003,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 400
    },
    {
      "completion_length": 704.59375,
      "epoch": 80.2,
      "grad_norm": 0.03629639372229576,
      "kl": 0.004337310791015625,
      "learning_rate": 2.2948675722421086e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 401
    },
    {
      "completion_length": 652.84375,
      "epoch": 80.4,
      "grad_norm": 0.06909455358982086,
      "kl": 0.007781982421875,
      "learning_rate": 2.2505551129582047e-06,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 402
    },
    {
      "completion_length": 583.03125,
      "epoch": 80.6,
      "grad_norm": 0.0329122431576252,
      "kl": 0.010089874267578125,
      "learning_rate": 2.206620350685257e-06,
      "loss": 0.0004,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 403
    },
    {
      "completion_length": 904.125,
      "epoch": 80.8,
      "grad_norm": 0.027277404442429543,
      "kl": 0.0045719146728515625,
      "learning_rate": 2.163065426741603e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 404
    },
    {
      "completion_length": 714.0625,
      "epoch": 81.0,
      "grad_norm": 0.05145667865872383,
      "kl": 0.006084442138671875,
      "learning_rate": 2.119892463932781e-06,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 405
    },
    {
      "completion_length": 705.71875,
      "epoch": 81.2,
      "grad_norm": 0.06335460394620895,
      "kl": 0.007598876953125,
      "learning_rate": 2.0771035664480944e-06,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 406
    },
    {
      "completion_length": 675.40625,
      "epoch": 81.4,
      "grad_norm": 0.06961791962385178,
      "kl": 0.00555419921875,
      "learning_rate": 2.0347008197580376e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 407
    },
    {
      "completion_length": 738.1875,
      "epoch": 81.6,
      "grad_norm": 0.04456457123160362,
      "kl": 0.0035190582275390625,
      "learning_rate": 1.9926862905126663e-06,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 408
    },
    {
      "completion_length": 718.5625,
      "epoch": 81.8,
      "grad_norm": 0.05315074697136879,
      "kl": 0.005313873291015625,
      "learning_rate": 1.95106202644086e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 409
    },
    {
      "completion_length": 738.9375,
      "epoch": 82.0,
      "grad_norm": 0.04968252405524254,
      "kl": 0.0063343048095703125,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 410
    },
    {
      "completion_length": 643.0,
      "epoch": 82.2,
      "grad_norm": 0.06702003628015518,
      "kl": 0.0022411346435546875,
      "learning_rate": 1.8689923895297247e-06,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 411
    },
    {
      "completion_length": 818.625,
      "epoch": 82.4,
      "grad_norm": 0.031119490042328835,
      "kl": 0.003907203674316406,
      "learning_rate": 1.8285510166487154e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 412
    },
    {
      "completion_length": 712.1875,
      "epoch": 82.6,
      "grad_norm": 0.10696296393871307,
      "kl": 0.01055145263671875,
      "learning_rate": 1.7885079086629598e-06,
      "loss": 0.0004,
      "reward": 0.1875,
      "reward_std": 0.3104073107242584,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.1875,
      "step": 413
    },
    {
      "completion_length": 683.3125,
      "epoch": 82.8,
      "grad_norm": 0.05855996534228325,
      "kl": 0.004909515380859375,
      "learning_rate": 1.7488650172170496e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 414
    },
    {
      "completion_length": 700.3125,
      "epoch": 83.0,
      "grad_norm": 0.054597459733486176,
      "kl": 0.00568389892578125,
      "learning_rate": 1.709624274449584e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 415
    },
    {
      "completion_length": 739.03125,
      "epoch": 83.2,
      "grad_norm": 0.05254088714718819,
      "kl": 0.004650115966796875,
      "learning_rate": 1.6707875928990059e-06,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.2630179077386856,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0,
      "step": 416
    },
    {
      "completion_length": 725.96875,
      "epoch": 83.4,
      "grad_norm": 0.06761158257722855,
      "kl": 0.0069293975830078125,
      "learning_rate": 1.6323568654103838e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 417
    },
    {
      "completion_length": 596.34375,
      "epoch": 83.6,
      "grad_norm": 0.09070757031440735,
      "kl": 0.0081939697265625,
      "learning_rate": 1.5943339650431578e-06,
      "loss": 0.0003,
      "reward": 0.21875,
      "reward_std": 0.4355708882212639,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 418
    },
    {
      "completion_length": 725.90625,
      "epoch": 83.8,
      "grad_norm": 0.06117472052574158,
      "kl": 0.013702392578125,
      "learning_rate": 1.5567207449798517e-06,
      "loss": 0.0005,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 419
    },
    {
      "completion_length": 580.25,
      "epoch": 84.0,
      "grad_norm": 0.02816312201321125,
      "kl": 0.004856109619140625,
      "learning_rate": 1.5195190384357405e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 420
    },
    {
      "completion_length": 763.875,
      "epoch": 84.2,
      "grad_norm": 0.03096804767847061,
      "kl": 0.003997802734375,
      "learning_rate": 1.4827306585695234e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 421
    },
    {
      "completion_length": 747.84375,
      "epoch": 84.4,
      "grad_norm": 0.03320998325943947,
      "kl": 0.00650787353515625,
      "learning_rate": 1.446357398394934e-06,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 422
    },
    {
      "completion_length": 659.09375,
      "epoch": 84.6,
      "grad_norm": 0.051016248762607574,
      "kl": 0.0054378509521484375,
      "learning_rate": 1.4104010306933558e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 423
    },
    {
      "completion_length": 801.59375,
      "epoch": 84.8,
      "grad_norm": 0.06237417459487915,
      "kl": 0.0053157806396484375,
      "learning_rate": 1.3748633079274254e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 424
    },
    {
      "completion_length": 756.8125,
      "epoch": 85.0,
      "grad_norm": 0.07189148664474487,
      "kl": 0.005405426025390625,
      "learning_rate": 1.339745962155613e-06,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.27439429610967636,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0625,
      "step": 425
    },
    {
      "completion_length": 577.5,
      "epoch": 85.2,
      "grad_norm": 0.10670667886734009,
      "kl": 0.01203155517578125,
      "learning_rate": 1.30505070494781e-06,
      "loss": 0.0005,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.1875,
      "step": 426
    },
    {
      "completion_length": 713.0,
      "epoch": 85.4,
      "grad_norm": 0.044564273208379745,
      "kl": 0.0067901611328125,
      "learning_rate": 1.2707792273019049e-06,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 427
    },
    {
      "completion_length": 783.53125,
      "epoch": 85.6,
      "grad_norm": 0.04508211091160774,
      "kl": 0.0023241043090820312,
      "learning_rate": 1.2369331995613664e-06,
      "loss": 0.0001,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0625,
      "step": 428
    },
    {
      "completion_length": 721.625,
      "epoch": 85.8,
      "grad_norm": 0.05431554093956947,
      "kl": 0.0087738037109375,
      "learning_rate": 1.2035142713338366e-06,
      "loss": 0.0004,
      "reward": 0.21875,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.03125,
      "step": 429
    },
    {
      "completion_length": 759.1875,
      "epoch": 86.0,
      "grad_norm": 0.058978427201509476,
      "kl": 0.005031585693359375,
      "learning_rate": 1.1705240714107301e-06,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 430
    },
    {
      "completion_length": 701.28125,
      "epoch": 86.2,
      "grad_norm": 0.0821586400270462,
      "kl": 0.004241943359375,
      "learning_rate": 1.1379642076878528e-06,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.408231720328331,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 431
    },
    {
      "completion_length": 796.5625,
      "epoch": 86.4,
      "grad_norm": 0.036505650728940964,
      "kl": 0.00543975830078125,
      "learning_rate": 1.1058362670870248e-06,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 432
    },
    {
      "completion_length": 724.375,
      "epoch": 86.6,
      "grad_norm": 0.03882104530930519,
      "kl": 0.00560760498046875,
      "learning_rate": 1.0741418154787443e-06,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 433
    },
    {
      "completion_length": 678.34375,
      "epoch": 86.8,
      "grad_norm": 0.084920734167099,
      "kl": 0.0077667236328125,
      "learning_rate": 1.042882397605871e-06,
      "loss": 0.0003,
      "reward": 0.15625,
      "reward_std": 0.24511480331420898,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.09375,
      "step": 434
    },
    {
      "completion_length": 689.375,
      "epoch": 87.0,
      "grad_norm": 0.08935703337192535,
      "kl": 0.008365631103515625,
      "learning_rate": 1.012059537008332e-06,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.249358132481575,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0625,
      "step": 435
    },
    {
      "completion_length": 583.40625,
      "epoch": 87.2,
      "grad_norm": 0.09143051505088806,
      "kl": 0.009525299072265625,
      "learning_rate": 9.816747359488632e-07,
      "loss": 0.0004,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 436
    },
    {
      "completion_length": 686.09375,
      "epoch": 87.4,
      "grad_norm": 0.07910123467445374,
      "kl": 0.0091705322265625,
      "learning_rate": 9.517294753398066e-07,
      "loss": 0.0004,
      "reward": 0.375,
      "reward_std": 0.3104073107242584,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.15625,
      "step": 437
    },
    {
      "completion_length": 749.65625,
      "epoch": 87.6,
      "grad_norm": 0.049836572259664536,
      "kl": 0.0059814453125,
      "learning_rate": 9.222252146709143e-07,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 438
    },
    {
      "completion_length": 702.03125,
      "epoch": 87.8,
      "grad_norm": 0.03987400233745575,
      "kl": 0.0032215118408203125,
      "learning_rate": 8.931633919382299e-07,
      "loss": 0.0001,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 439
    },
    {
      "completion_length": 788.125,
      "epoch": 88.0,
      "grad_norm": 0.014637740328907967,
      "kl": 0.0043621063232421875,
      "learning_rate": 8.645454235739903e-07,
      "loss": 0.0002,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 440
    },
    {
      "completion_length": 646.28125,
      "epoch": 88.2,
      "grad_norm": 0.040815550833940506,
      "kl": 0.00289154052734375,
      "learning_rate": 8.363727043776037e-07,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 441
    },
    {
      "completion_length": 759.625,
      "epoch": 88.4,
      "grad_norm": 0.055915869772434235,
      "kl": 0.0036458969116210938,
      "learning_rate": 8.086466074476562e-07,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.03125,
      "step": 442
    },
    {
      "completion_length": 658.78125,
      "epoch": 88.6,
      "grad_norm": 0.10622189939022064,
      "kl": 0.011785507202148438,
      "learning_rate": 7.81368484114996e-07,
      "loss": 0.0005,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 443
    },
    {
      "completion_length": 796.6875,
      "epoch": 88.8,
      "grad_norm": 0.04346534237265587,
      "kl": 0.003444671630859375,
      "learning_rate": 7.545396638768698e-07,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 444
    },
    {
      "completion_length": 767.25,
      "epoch": 89.0,
      "grad_norm": 0.07092897593975067,
      "kl": 0.0074462890625,
      "learning_rate": 7.281614543321269e-07,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 445
    },
    {
      "completion_length": 700.53125,
      "epoch": 89.2,
      "grad_norm": 0.1183854192495346,
      "kl": 0.0064239501953125,
      "learning_rate": 7.022351411174866e-07,
      "loss": 0.0003,
      "reward": 0.25,
      "reward_std": 0.4261348247528076,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.15625,
      "step": 446
    },
    {
      "completion_length": 706.8125,
      "epoch": 89.4,
      "grad_norm": 0.05067533254623413,
      "kl": 0.0027251243591308594,
      "learning_rate": 6.767619878448783e-07,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 447
    },
    {
      "completion_length": 683.9375,
      "epoch": 89.6,
      "grad_norm": 0.06551023572683334,
      "kl": 0.007556915283203125,
      "learning_rate": 6.517432360398556e-07,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 448
    },
    {
      "completion_length": 803.625,
      "epoch": 89.8,
      "grad_norm": 0.02727729082107544,
      "kl": 0.0027456283569335938,
      "learning_rate": 6.271801050810856e-07,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 449
    },
    {
      "completion_length": 690.875,
      "epoch": 90.0,
      "grad_norm": 0.0609399788081646,
      "kl": 0.006710052490234375,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 450
    },
    {
      "completion_length": 765.28125,
      "epoch": 90.2,
      "grad_norm": 0.060198985040187836,
      "kl": 0.006627082824707031,
      "learning_rate": 5.794254721270331e-07,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.03125,
      "step": 451
    },
    {
      "completion_length": 777.875,
      "epoch": 90.4,
      "grad_norm": 0.05776260048151016,
      "kl": 0.005610466003417969,
      "learning_rate": 5.562362976251901e-07,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 452
    },
    {
      "completion_length": 602.21875,
      "epoch": 90.6,
      "grad_norm": 0.07143893837928772,
      "kl": 0.0063629150390625,
      "learning_rate": 5.335073988430373e-07,
      "loss": 0.0003,
      "reward": 0.1875,
      "reward_std": 0.3945523276925087,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 453
    },
    {
      "completion_length": 734.90625,
      "epoch": 90.8,
      "grad_norm": 0.08987980335950851,
      "kl": 0.00484466552734375,
      "learning_rate": 5.112398835550348e-07,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.3061639815568924,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.0625,
      "step": 454
    },
    {
      "completion_length": 775.875,
      "epoch": 91.0,
      "grad_norm": 0.07045432925224304,
      "kl": 0.007686614990234375,
      "learning_rate": 4.894348370484648e-07,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 455
    },
    {
      "completion_length": 742.59375,
      "epoch": 91.2,
      "grad_norm": 0.1062343493103981,
      "kl": 0.00803375244140625,
      "learning_rate": 4.6809332207053083e-07,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 456
    },
    {
      "completion_length": 709.25,
      "epoch": 91.4,
      "grad_norm": 0.03916856646537781,
      "kl": 0.0048675537109375,
      "learning_rate": 4.4721637877656377e-07,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 457
    },
    {
      "completion_length": 641.375,
      "epoch": 91.6,
      "grad_norm": 0.052550166845321655,
      "kl": 0.01186370849609375,
      "learning_rate": 4.268050246793276e-07,
      "loss": 0.0005,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 458
    },
    {
      "completion_length": 828.25,
      "epoch": 91.8,
      "grad_norm": 0.0332893468439579,
      "kl": 0.0023899078369140625,
      "learning_rate": 4.068602545994249e-07,
      "loss": 0.0001,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 459
    },
    {
      "completion_length": 713.5625,
      "epoch": 92.0,
      "grad_norm": 0.029494035989046097,
      "kl": 0.0042266845703125,
      "learning_rate": 3.8738304061681107e-07,
      "loss": 0.0002,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 460
    },
    {
      "completion_length": 670.34375,
      "epoch": 92.2,
      "grad_norm": 0.11828664690256119,
      "kl": 0.0059661865234375,
      "learning_rate": 3.68374332023419e-07,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0,
      "step": 461
    },
    {
      "completion_length": 502.15625,
      "epoch": 92.4,
      "grad_norm": 0.19351941347122192,
      "kl": 0.0099029541015625,
      "learning_rate": 3.498350552768859e-07,
      "loss": 0.0004,
      "reward": 0.125,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 462
    },
    {
      "completion_length": 840.84375,
      "epoch": 92.6,
      "grad_norm": 0.035693634301424026,
      "kl": 0.003574371337890625,
      "learning_rate": 3.3176611395540625e-07,
      "loss": 0.0001,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 463
    },
    {
      "completion_length": 765.71875,
      "epoch": 92.8,
      "grad_norm": 0.05469142273068428,
      "kl": 0.00876617431640625,
      "learning_rate": 3.1416838871368925e-07,
      "loss": 0.0004,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 464
    },
    {
      "completion_length": 780.125,
      "epoch": 93.0,
      "grad_norm": 0.03061514161527157,
      "kl": 0.0030832290649414062,
      "learning_rate": 2.970427372400353e-07,
      "loss": 0.0001,
      "reward": 0.125,
      "reward_std": 0.13363061845302582,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 465
    },
    {
      "completion_length": 581.3125,
      "epoch": 93.2,
      "grad_norm": 0.1440468281507492,
      "kl": 0.00862884521484375,
      "learning_rate": 2.8038999421453827e-07,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 466
    },
    {
      "completion_length": 684.09375,
      "epoch": 93.4,
      "grad_norm": 0.0781330093741417,
      "kl": 0.008890151977539062,
      "learning_rate": 2.6421097126839714e-07,
      "loss": 0.0004,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 467
    },
    {
      "completion_length": 828.75,
      "epoch": 93.6,
      "grad_norm": 0.03663882240653038,
      "kl": 0.0069370269775390625,
      "learning_rate": 2.4850645694436736e-07,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.1293872892856598,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.0,
      "step": 468
    },
    {
      "completion_length": 683.15625,
      "epoch": 93.8,
      "grad_norm": 0.08346743881702423,
      "kl": 0.006175994873046875,
      "learning_rate": 2.332772166583208e-07,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.4218914955854416,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.09375,
      "step": 469
    },
    {
      "completion_length": 633.1875,
      "epoch": 94.0,
      "grad_norm": 0.11486148834228516,
      "kl": 0.006305694580078125,
      "learning_rate": 2.1852399266194312e-07,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 470
    },
    {
      "completion_length": 747.28125,
      "epoch": 94.2,
      "grad_norm": 0.0489036962389946,
      "kl": 0.0030517578125,
      "learning_rate": 2.0424750400655947e-07,
      "loss": 0.0001,
      "reward": 0.1875,
      "reward_std": 0.249358132481575,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 471
    },
    {
      "completion_length": 736.71875,
      "epoch": 94.4,
      "grad_norm": 0.053127698600292206,
      "kl": 0.0052032470703125,
      "learning_rate": 1.9044844650808468e-07,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.03125,
      "step": 472
    },
    {
      "completion_length": 737.71875,
      "epoch": 94.6,
      "grad_norm": 0.035883210599422455,
      "kl": 0.008480072021484375,
      "learning_rate": 1.7712749271311392e-07,
      "loss": 0.0003,
      "reward": 0.03125,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.03125,
      "step": 473
    },
    {
      "completion_length": 727.875,
      "epoch": 94.8,
      "grad_norm": 0.05309724435210228,
      "kl": 0.007781982421875,
      "learning_rate": 1.6428529186614195e-07,
      "loss": 0.0003,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 474
    },
    {
      "completion_length": 729.625,
      "epoch": 95.0,
      "grad_norm": 0.12343677878379822,
      "kl": 0.006969451904296875,
      "learning_rate": 1.519224698779198e-07,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 475
    },
    {
      "completion_length": 705.96875,
      "epoch": 95.2,
      "grad_norm": 0.09638237953186035,
      "kl": 0.00696563720703125,
      "learning_rate": 1.400396292949513e-07,
      "loss": 0.0003,
      "reward": 0.125,
      "reward_std": 0.2314550280570984,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.125,
      "step": 476
    },
    {
      "completion_length": 730.15625,
      "epoch": 95.4,
      "grad_norm": 0.09068649262189865,
      "kl": 0.0046634674072265625,
      "learning_rate": 1.2863734927012094e-07,
      "loss": 0.0002,
      "reward": 0.28125,
      "reward_std": 0.4534739926457405,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.0625,
      "step": 477
    },
    {
      "completion_length": 757.65625,
      "epoch": 95.6,
      "grad_norm": 0.03434085100889206,
      "kl": 0.0056858062744140625,
      "learning_rate": 1.1771618553447217e-07,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 478
    },
    {
      "completion_length": 729.53125,
      "epoch": 95.8,
      "grad_norm": 0.07220398634672165,
      "kl": 0.005321025848388672,
      "learning_rate": 1.0727667037011668e-07,
      "loss": 0.0002,
      "reward": 0.25,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.09375,
      "step": 479
    },
    {
      "completion_length": 654.8125,
      "epoch": 96.0,
      "grad_norm": 0.056842703372240067,
      "kl": 0.005825042724609375,
      "learning_rate": 9.731931258429638e-08,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 480
    },
    {
      "completion_length": 799.78125,
      "epoch": 96.2,
      "grad_norm": 0.05823909491300583,
      "kl": 0.01227569580078125,
      "learning_rate": 8.784459748458318e-08,
      "loss": 0.0005,
      "reward": 0.1875,
      "reward_std": 0.249358132481575,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 481
    },
    {
      "completion_length": 693.40625,
      "epoch": 96.4,
      "grad_norm": 0.042686577886343,
      "kl": 0.0037841796875,
      "learning_rate": 7.885298685522235e-08,
      "loss": 0.0002,
      "reward": 0.1875,
      "reward_std": 0.2587745785713196,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 482
    },
    {
      "completion_length": 720.8125,
      "epoch": 96.6,
      "grad_norm": 0.057090874761343,
      "kl": 0.005649089813232422,
      "learning_rate": 7.034491893463059e-08,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.09375,
      "step": 483
    },
    {
      "completion_length": 713.9375,
      "epoch": 96.8,
      "grad_norm": 0.10989940166473389,
      "kl": 0.004550933837890625,
      "learning_rate": 6.232080839403631e-08,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 484
    },
    {
      "completion_length": 689.125,
      "epoch": 97.0,
      "grad_norm": 0.0833081528544426,
      "kl": 0.010522842407226562,
      "learning_rate": 5.4781046317267103e-08,
      "loss": 0.0004,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.03125,
      "step": 485
    },
    {
      "completion_length": 623.125,
      "epoch": 97.2,
      "grad_norm": 0.06911883503198624,
      "kl": 0.00821685791015625,
      "learning_rate": 4.772600018168816e-08,
      "loss": 0.0003,
      "reward": 0.34375,
      "reward_std": 0.38816186785697937,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.125,
      "step": 486
    },
    {
      "completion_length": 737.75,
      "epoch": 97.4,
      "grad_norm": 0.05464695766568184,
      "kl": 0.0027599334716796875,
      "learning_rate": 4.115601384029666e-08,
      "loss": 0.0001,
      "reward": 0.25,
      "reward_std": 0.3650856465101242,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0625,
      "step": 487
    },
    {
      "completion_length": 767.03125,
      "epoch": 97.6,
      "grad_norm": 0.05080058425664902,
      "kl": 0.004467010498046875,
      "learning_rate": 3.50714075049563e-08,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.03125,
      "step": 488
    },
    {
      "completion_length": 708.21875,
      "epoch": 97.8,
      "grad_norm": 0.06618014723062515,
      "kl": 0.0099945068359375,
      "learning_rate": 2.947247773079753e-08,
      "loss": 0.0004,
      "reward": 0.09375,
      "reward_std": 0.2651650384068489,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 489
    },
    {
      "completion_length": 797.6875,
      "epoch": 98.0,
      "grad_norm": 0.05233558639883995,
      "kl": 0.00853729248046875,
      "learning_rate": 2.4359497401758026e-08,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 490
    },
    {
      "completion_length": 698.40625,
      "epoch": 98.2,
      "grad_norm": 0.0935780480504036,
      "kl": 0.00585174560546875,
      "learning_rate": 1.973271571728441e-08,
      "loss": 0.0002,
      "reward": 0.21875,
      "reward_std": 0.4355708882212639,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.15625,
      "step": 491
    },
    {
      "completion_length": 749.0,
      "epoch": 98.4,
      "grad_norm": 0.07994402945041656,
      "kl": 0.004262447357177734,
      "learning_rate": 1.5592358180189782e-08,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.22201896458864212,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.15625,
      "step": 492
    },
    {
      "completion_length": 702.625,
      "epoch": 98.6,
      "grad_norm": 0.06925726681947708,
      "kl": 0.005274772644042969,
      "learning_rate": 1.1938626585660252e-08,
      "loss": 0.0002,
      "reward": 0.25,
      "reward_std": 0.3514062538743019,
      "rewards/accuracy_reward": 0.21875,
      "rewards/format_reward": 0.03125,
      "step": 493
    },
    {
      "completion_length": 611.65625,
      "epoch": 98.8,
      "grad_norm": 0.03248089924454689,
      "kl": 0.00778961181640625,
      "learning_rate": 8.771699011416169e-09,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 494
    },
    {
      "completion_length": 742.3125,
      "epoch": 99.0,
      "grad_norm": 0.07433319091796875,
      "kl": 0.0054531097412109375,
      "learning_rate": 6.091729809042379e-09,
      "loss": 0.0002,
      "reward": 0.15625,
      "reward_std": 0.3198433741927147,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.125,
      "step": 495
    },
    {
      "completion_length": 701.59375,
      "epoch": 99.2,
      "grad_norm": 0.08704384416341782,
      "kl": 0.006954193115234375,
      "learning_rate": 3.898849596456477e-09,
      "loss": 0.0003,
      "reward": 0.25,
      "reward_std": 0.2925042062997818,
      "rewards/accuracy_reward": 0.15625,
      "rewards/format_reward": 0.09375,
      "step": 496
    },
    {
      "completion_length": 791.5,
      "epoch": 99.4,
      "grad_norm": 0.027208929881453514,
      "kl": 0.0045757293701171875,
      "learning_rate": 2.193165251545004e-09,
      "loss": 0.0002,
      "reward": 0.0625,
      "reward_std": 0.1157275140285492,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0625,
      "step": 497
    },
    {
      "completion_length": 686.59375,
      "epoch": 99.6,
      "grad_norm": 0.043633949011564255,
      "kl": 0.00383758544921875,
      "learning_rate": 9.74759906957612e-10,
      "loss": 0.0002,
      "reward": 0.125,
      "reward_std": 0.2177756354212761,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.09375,
      "step": 498
    },
    {
      "completion_length": 763.5625,
      "epoch": 99.8,
      "grad_norm": 0.05561517924070358,
      "kl": 0.0047931671142578125,
      "learning_rate": 2.436929460525317e-10,
      "loss": 0.0002,
      "reward": 0.09375,
      "reward_std": 0.2041158601641655,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.0625,
      "step": 499
    },
    {
      "completion_length": 607.375,
      "epoch": 100.0,
      "grad_norm": 0.06856601685285568,
      "kl": 0.00896453857421875,
      "learning_rate": 0.0,
      "loss": 0.0004,
      "reward": 0.1875,
      "reward_std": 0.3335031494498253,
      "rewards/accuracy_reward": 0.09375,
      "rewards/format_reward": 0.09375,
      "step": 500
    },
    {
      "epoch": 100.0,
      "step": 500,
      "total_flos": 0.0,
      "train_loss": 0.00019750852083552672,
      "train_runtime": 15991.8306,
      "train_samples_per_second": 0.125,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "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": 4,
  "trial_name": null,
  "trial_params": null
}