{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 228,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 313.40625,
      "epoch": 0.02631578947368421,
      "grad_norm": 1.734404490317045,
      "kl": 0.0,
      "learning_rate": 9.973684210526315e-07,
      "loss": -0.0,
      "reward": 0.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.03125,
      "step": 1
    },
    {
      "completion_length": 223.125,
      "epoch": 0.05263157894736842,
      "grad_norm": 2.7739676691945543,
      "kl": 0.000339508056640625,
      "learning_rate": 9.947368421052631e-07,
      "loss": 0.0,
      "reward": 0.78125,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.65625,
      "rewards/format_reward": 0.125,
      "step": 2
    },
    {
      "completion_length": 225.125,
      "epoch": 0.07894736842105263,
      "grad_norm": 2.751260848291789,
      "kl": 0.000408172607421875,
      "learning_rate": 9.921052631578947e-07,
      "loss": 0.0,
      "reward": 0.84375,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.09375,
      "step": 3
    },
    {
      "completion_length": 218.96875,
      "epoch": 0.10526315789473684,
      "grad_norm": 2.815805743579428,
      "kl": 0.0003662109375,
      "learning_rate": 9.894736842105263e-07,
      "loss": 0.0,
      "reward": 0.71875,
      "reward_std": 0.33183756470680237,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.15625,
      "step": 4
    },
    {
      "completion_length": 232.375,
      "epoch": 0.13157894736842105,
      "grad_norm": 3.4873903451415527,
      "kl": 0.00054168701171875,
      "learning_rate": 9.868421052631579e-07,
      "loss": 0.0,
      "reward": 0.96875,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.09375,
      "step": 5
    },
    {
      "completion_length": 209.28125,
      "epoch": 0.15789473684210525,
      "grad_norm": 11.352612794141292,
      "kl": 0.0010986328125,
      "learning_rate": 9.842105263157894e-07,
      "loss": 0.0,
      "reward": 0.75,
      "reward_std": 0.2992308735847473,
      "rewards/accuracy_reward": 0.6875,
      "rewards/format_reward": 0.0625,
      "step": 6
    },
    {
      "completion_length": 239.09375,
      "epoch": 0.18421052631578946,
      "grad_norm": 2.429253829069759,
      "kl": 0.001617431640625,
      "learning_rate": 9.81578947368421e-07,
      "loss": 0.0001,
      "reward": 0.9375,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.09375,
      "step": 7
    },
    {
      "completion_length": 199.34375,
      "epoch": 0.21052631578947367,
      "grad_norm": 2.6508949173291008,
      "kl": 0.003173828125,
      "learning_rate": 9.789473684210526e-07,
      "loss": 0.0001,
      "reward": 1.0625,
      "reward_std": 0.26933756470680237,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.15625,
      "step": 8
    },
    {
      "completion_length": 217.15625,
      "epoch": 0.23684210526315788,
      "grad_norm": 3.643602680405586,
      "kl": 0.003082275390625,
      "learning_rate": 9.763157894736842e-07,
      "loss": 0.0001,
      "reward": 1.09375,
      "reward_std": 0.3125,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.25,
      "step": 9
    },
    {
      "completion_length": 184.5625,
      "epoch": 0.2631578947368421,
      "grad_norm": 4.633586265396643,
      "kl": 0.006134033203125,
      "learning_rate": 9.736842105263158e-07,
      "loss": 0.0002,
      "reward": 1.15625,
      "reward_std": 0.5290063619613647,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.40625,
      "step": 10
    },
    {
      "completion_length": 227.28125,
      "epoch": 0.2894736842105263,
      "grad_norm": 4.047589204398526,
      "kl": 0.0068359375,
      "learning_rate": 9.710526315789474e-07,
      "loss": 0.0003,
      "reward": 1.1875,
      "reward_std": 0.5834712982177734,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.375,
      "step": 11
    },
    {
      "completion_length": 178.09375,
      "epoch": 0.3157894736842105,
      "grad_norm": 4.471112815496676,
      "kl": 0.010498046875,
      "learning_rate": 9.68421052631579e-07,
      "loss": 0.0004,
      "reward": 1.4375,
      "reward_std": 0.5492308735847473,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.46875,
      "step": 12
    },
    {
      "completion_length": 195.96875,
      "epoch": 0.34210526315789475,
      "grad_norm": 5.344105755138263,
      "kl": 0.0128173828125,
      "learning_rate": 9.657894736842105e-07,
      "loss": 0.0005,
      "reward": 1.1875,
      "reward_std": 0.32216876745224,
      "rewards/accuracy_reward": 0.6875,
      "rewards/format_reward": 0.5,
      "step": 13
    },
    {
      "completion_length": 216.59375,
      "epoch": 0.3684210526315789,
      "grad_norm": 5.518182292437182,
      "kl": 0.01080322265625,
      "learning_rate": 9.63157894736842e-07,
      "loss": 0.0004,
      "reward": 1.46875,
      "reward_std": 0.3696783781051636,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.53125,
      "step": 14
    },
    {
      "completion_length": 182.5625,
      "epoch": 0.39473684210526316,
      "grad_norm": 4.4023382034222625,
      "kl": 0.01953125,
      "learning_rate": 9.605263157894737e-07,
      "loss": 0.0008,
      "reward": 1.4375,
      "reward_std": 0.4242308735847473,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.6875,
      "step": 15
    },
    {
      "completion_length": 202.09375,
      "epoch": 0.42105263157894735,
      "grad_norm": 3.344194998556422,
      "kl": 0.017333984375,
      "learning_rate": 9.578947368421053e-07,
      "loss": 0.0007,
      "reward": 1.40625,
      "reward_std": 0.40400636196136475,
      "rewards/accuracy_reward": 0.71875,
      "rewards/format_reward": 0.6875,
      "step": 16
    },
    {
      "completion_length": 208.375,
      "epoch": 0.4473684210526316,
      "grad_norm": 3.4905855548266636,
      "kl": 0.018310546875,
      "learning_rate": 9.552631578947368e-07,
      "loss": 0.0007,
      "reward": 1.25,
      "reward_std": 0.5615255832672119,
      "rewards/accuracy_reward": 0.59375,
      "rewards/format_reward": 0.65625,
      "step": 17
    },
    {
      "completion_length": 185.1875,
      "epoch": 0.47368421052631576,
      "grad_norm": 3.4648471201200115,
      "kl": 0.018798828125,
      "learning_rate": 9.526315789473683e-07,
      "loss": 0.0008,
      "reward": 1.59375,
      "reward_std": 0.3125,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.71875,
      "step": 18
    },
    {
      "completion_length": 170.1875,
      "epoch": 0.5,
      "grad_norm": 3.516316903937348,
      "kl": 0.027099609375,
      "learning_rate": 9.499999999999999e-07,
      "loss": 0.0011,
      "reward": 1.8125,
      "reward_std": 0.32216876745224,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.8125,
      "step": 19
    },
    {
      "completion_length": 143.6875,
      "epoch": 0.5263157894736842,
      "grad_norm": 4.856219822054919,
      "kl": 0.034423828125,
      "learning_rate": 9.473684210526315e-07,
      "loss": 0.0014,
      "reward": 1.5,
      "reward_std": 0.4963996410369873,
      "rewards/accuracy_reward": 0.6875,
      "rewards/format_reward": 0.8125,
      "step": 20
    },
    {
      "completion_length": 158.8125,
      "epoch": 0.5526315789473685,
      "grad_norm": 1.6022523114697775,
      "kl": 0.0341796875,
      "learning_rate": 9.447368421052632e-07,
      "loss": 0.0014,
      "reward": 1.59375,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.59375,
      "rewards/format_reward": 1.0,
      "step": 21
    },
    {
      "completion_length": 160.1875,
      "epoch": 0.5789473684210527,
      "grad_norm": 2.545246246786829,
      "kl": 0.037109375,
      "learning_rate": 9.421052631578948e-07,
      "loss": 0.0015,
      "reward": 1.375,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.46875,
      "rewards/format_reward": 0.90625,
      "step": 22
    },
    {
      "completion_length": 172.40625,
      "epoch": 0.6052631578947368,
      "grad_norm": 4.312576092147522,
      "kl": 0.0279541015625,
      "learning_rate": 9.394736842105263e-07,
      "loss": 0.0011,
      "reward": 1.5625,
      "reward_std": 0.375,
      "rewards/accuracy_reward": 0.6875,
      "rewards/format_reward": 0.875,
      "step": 23
    },
    {
      "completion_length": 135.625,
      "epoch": 0.631578947368421,
      "grad_norm": 3.07360466610783,
      "kl": 0.03466796875,
      "learning_rate": 9.368421052631579e-07,
      "loss": 0.0014,
      "reward": 1.78125,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.90625,
      "step": 24
    },
    {
      "completion_length": 147.625,
      "epoch": 0.6578947368421053,
      "grad_norm": 1.799001126451805,
      "kl": 0.0322265625,
      "learning_rate": 9.342105263157895e-07,
      "loss": 0.0013,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 25
    },
    {
      "completion_length": 143.59375,
      "epoch": 0.6842105263157895,
      "grad_norm": 2.968199508863703,
      "kl": 0.03857421875,
      "learning_rate": 9.31578947368421e-07,
      "loss": 0.0015,
      "reward": 1.6875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.71875,
      "rewards/format_reward": 0.96875,
      "step": 26
    },
    {
      "completion_length": 128.71875,
      "epoch": 0.7105263157894737,
      "grad_norm": 2.0469322588476735,
      "kl": 0.03466796875,
      "learning_rate": 9.289473684210526e-07,
      "loss": 0.0014,
      "reward": 1.78125,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.96875,
      "step": 27
    },
    {
      "completion_length": 142.03125,
      "epoch": 0.7368421052631579,
      "grad_norm": 10.08604224085363,
      "kl": 0.030517578125,
      "learning_rate": 9.263157894736841e-07,
      "loss": 0.0012,
      "reward": 1.78125,
      "reward_std": 0.33183756470680237,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.90625,
      "step": 28
    },
    {
      "completion_length": 155.125,
      "epoch": 0.7631578947368421,
      "grad_norm": 3.1120163936500065,
      "kl": 0.030517578125,
      "learning_rate": 9.236842105263157e-07,
      "loss": 0.0012,
      "reward": 1.78125,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.9375,
      "step": 29
    },
    {
      "completion_length": 147.625,
      "epoch": 0.7894736842105263,
      "grad_norm": 4.576665670394314,
      "kl": 0.04443359375,
      "learning_rate": 9.210526315789473e-07,
      "loss": 0.0018,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 30
    },
    {
      "completion_length": 151.53125,
      "epoch": 0.8157894736842105,
      "grad_norm": 1.2450324331157339,
      "kl": 0.031494140625,
      "learning_rate": 9.184210526315789e-07,
      "loss": 0.0013,
      "reward": 1.84375,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 1.0,
      "step": 31
    },
    {
      "completion_length": 140.53125,
      "epoch": 0.8421052631578947,
      "grad_norm": 1.4646561367312099,
      "kl": 0.026611328125,
      "learning_rate": 9.157894736842105e-07,
      "loss": 0.0011,
      "reward": 1.40625,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.40625,
      "rewards/format_reward": 1.0,
      "step": 32
    },
    {
      "completion_length": 145.5,
      "epoch": 0.868421052631579,
      "grad_norm": 2.7991930077966813,
      "kl": 0.02978515625,
      "learning_rate": 9.13157894736842e-07,
      "loss": 0.0012,
      "reward": 1.84375,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 33
    },
    {
      "completion_length": 159.25,
      "epoch": 0.8947368421052632,
      "grad_norm": 2.578749059913484,
      "kl": 0.0238037109375,
      "learning_rate": 9.105263157894737e-07,
      "loss": 0.001,
      "reward": 1.84375,
      "reward_std": 0.20683756470680237,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 34
    },
    {
      "completion_length": 170.15625,
      "epoch": 0.9210526315789473,
      "grad_norm": 1.4408060069389652,
      "kl": 0.023681640625,
      "learning_rate": 9.078947368421053e-07,
      "loss": 0.0009,
      "reward": 1.9375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 35
    },
    {
      "completion_length": 161.40625,
      "epoch": 0.9473684210526315,
      "grad_norm": 4.227155952460155,
      "kl": 0.03271484375,
      "learning_rate": 9.052631578947368e-07,
      "loss": 0.0013,
      "reward": 1.53125,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.59375,
      "rewards/format_reward": 0.9375,
      "step": 36
    },
    {
      "completion_length": 190.625,
      "epoch": 0.9736842105263158,
      "grad_norm": 7.245608776040312,
      "kl": 0.022705078125,
      "learning_rate": 9.026315789473684e-07,
      "loss": 0.0009,
      "reward": 1.65625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.65625,
      "rewards/format_reward": 1.0,
      "step": 37
    },
    {
      "completion_length": 189.1999969482422,
      "epoch": 1.0,
      "grad_norm": 1.9920127298691528,
      "kl": 0.024658203125,
      "learning_rate": 9e-07,
      "loss": 0.0009,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 38
    },
    {
      "completion_length": 185.3125,
      "epoch": 1.0263157894736843,
      "grad_norm": 2.789439958145857,
      "kl": 0.044921875,
      "learning_rate": 8.973684210526315e-07,
      "loss": 0.0018,
      "reward": 1.6875,
      "reward_std": 0.26933756470680237,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.875,
      "step": 39
    },
    {
      "completion_length": 179.8125,
      "epoch": 1.0526315789473684,
      "grad_norm": 1.2098048720944379,
      "kl": 0.0218505859375,
      "learning_rate": 8.947368421052631e-07,
      "loss": 0.0009,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 40
    },
    {
      "completion_length": 194.90625,
      "epoch": 1.0789473684210527,
      "grad_norm": 3.242857448772222,
      "kl": 0.08837890625,
      "learning_rate": 8.921052631578947e-07,
      "loss": 0.0035,
      "reward": 1.65625,
      "reward_std": 0.28956207633018494,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.90625,
      "step": 41
    },
    {
      "completion_length": 196.0,
      "epoch": 1.1052631578947367,
      "grad_norm": 0.9529596375150978,
      "kl": 0.02685546875,
      "learning_rate": 8.894736842105263e-07,
      "loss": 0.0011,
      "reward": 1.71875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.96875,
      "step": 42
    },
    {
      "completion_length": 177.375,
      "epoch": 1.131578947368421,
      "grad_norm": 0.17243337346927284,
      "kl": 0.023681640625,
      "learning_rate": 8.868421052631579e-07,
      "loss": 0.0009,
      "reward": 1.5,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 43
    },
    {
      "completion_length": 176.9375,
      "epoch": 1.1578947368421053,
      "grad_norm": 3.9901952590413767,
      "kl": 0.04345703125,
      "learning_rate": 8.842105263157895e-07,
      "loss": 0.0017,
      "reward": 1.78125,
      "reward_std": 0.38466876745224,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.90625,
      "step": 44
    },
    {
      "completion_length": 183.125,
      "epoch": 1.1842105263157894,
      "grad_norm": 2.3843351470911363,
      "kl": 0.02880859375,
      "learning_rate": 8.815789473684209e-07,
      "loss": 0.0012,
      "reward": 1.625,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.65625,
      "rewards/format_reward": 0.96875,
      "step": 45
    },
    {
      "completion_length": 166.96875,
      "epoch": 1.2105263157894737,
      "grad_norm": 4.071274547532942,
      "kl": 0.025146484375,
      "learning_rate": 8.789473684210525e-07,
      "loss": 0.001,
      "reward": 1.84375,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 1.0,
      "step": 46
    },
    {
      "completion_length": 197.125,
      "epoch": 1.236842105263158,
      "grad_norm": 2.008714180363393,
      "kl": 0.02392578125,
      "learning_rate": 8.763157894736841e-07,
      "loss": 0.001,
      "reward": 1.8125,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 47
    },
    {
      "completion_length": 206.0,
      "epoch": 1.263157894736842,
      "grad_norm": 3.0635549935930406,
      "kl": 0.02490234375,
      "learning_rate": 8.736842105263158e-07,
      "loss": 0.001,
      "reward": 1.5625,
      "reward_std": 0.32216876745224,
      "rewards/accuracy_reward": 0.65625,
      "rewards/format_reward": 0.90625,
      "step": 48
    },
    {
      "completion_length": 204.65625,
      "epoch": 1.2894736842105263,
      "grad_norm": 2.552435438840004,
      "kl": 0.057373046875,
      "learning_rate": 8.710526315789474e-07,
      "loss": 0.0023,
      "reward": 1.65625,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.71875,
      "rewards/format_reward": 0.9375,
      "step": 49
    },
    {
      "completion_length": 199.65625,
      "epoch": 1.3157894736842106,
      "grad_norm": 2.449737423390204,
      "kl": 0.0291748046875,
      "learning_rate": 8.684210526315789e-07,
      "loss": 0.0012,
      "reward": 1.75,
      "reward_std": 0.26933756470680237,
      "rewards/accuracy_reward": 0.78125,
      "rewards/format_reward": 0.96875,
      "step": 50
    },
    {
      "completion_length": 205.3125,
      "epoch": 1.3421052631578947,
      "grad_norm": 2.0269064836147876,
      "kl": 0.024169921875,
      "learning_rate": 8.657894736842105e-07,
      "loss": 0.001,
      "reward": 1.71875,
      "reward_std": 0.23673085868358612,
      "rewards/accuracy_reward": 0.78125,
      "rewards/format_reward": 0.9375,
      "step": 51
    },
    {
      "completion_length": 201.4375,
      "epoch": 1.368421052631579,
      "grad_norm": 7.212764333373468,
      "kl": 0.2314453125,
      "learning_rate": 8.631578947368421e-07,
      "loss": 0.0092,
      "reward": 1.84375,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.90625,
      "step": 52
    },
    {
      "completion_length": 205.625,
      "epoch": 1.3947368421052633,
      "grad_norm": 1.932007216297956,
      "kl": 0.028076171875,
      "learning_rate": 8.605263157894737e-07,
      "loss": 0.0011,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 53
    },
    {
      "completion_length": 213.59375,
      "epoch": 1.4210526315789473,
      "grad_norm": 1.1006621969472858,
      "kl": 0.0216064453125,
      "learning_rate": 8.578947368421053e-07,
      "loss": 0.0009,
      "reward": 1.71875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.96875,
      "step": 54
    },
    {
      "completion_length": 227.78125,
      "epoch": 1.4473684210526316,
      "grad_norm": 3.5124295614824095,
      "kl": 0.0260009765625,
      "learning_rate": 8.552631578947367e-07,
      "loss": 0.001,
      "reward": 1.75,
      "reward_std": 0.32216876745224,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.90625,
      "step": 55
    },
    {
      "completion_length": 209.03125,
      "epoch": 1.4736842105263157,
      "grad_norm": 2.110914394169814,
      "kl": 0.0274658203125,
      "learning_rate": 8.526315789473683e-07,
      "loss": 0.0011,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.9375,
      "step": 56
    },
    {
      "completion_length": 221.59375,
      "epoch": 1.5,
      "grad_norm": 2.6295758757346395,
      "kl": 0.0244140625,
      "learning_rate": 8.499999999999999e-07,
      "loss": 0.001,
      "reward": 1.6875,
      "reward_std": 0.3986847400665283,
      "rewards/accuracy_reward": 0.78125,
      "rewards/format_reward": 0.90625,
      "step": 57
    },
    {
      "completion_length": 213.34375,
      "epoch": 1.526315789473684,
      "grad_norm": 2.2321151195761013,
      "kl": 0.034912109375,
      "learning_rate": 8.473684210526315e-07,
      "loss": 0.0014,
      "reward": 1.53125,
      "reward_std": 0.3660780191421509,
      "rewards/accuracy_reward": 0.65625,
      "rewards/format_reward": 0.875,
      "step": 58
    },
    {
      "completion_length": 210.375,
      "epoch": 1.5526315789473686,
      "grad_norm": 5.701493957758128,
      "kl": 0.023681640625,
      "learning_rate": 8.447368421052631e-07,
      "loss": 0.0009,
      "reward": 1.84375,
      "reward_std": 0.20683756470680237,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 59
    },
    {
      "completion_length": 218.34375,
      "epoch": 1.5789473684210527,
      "grad_norm": 2.948270031439115,
      "kl": 0.034912109375,
      "learning_rate": 8.421052631578947e-07,
      "loss": 0.0014,
      "reward": 1.84375,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.9375,
      "step": 60
    },
    {
      "completion_length": 206.0625,
      "epoch": 1.6052631578947367,
      "grad_norm": 2.3435977803629857,
      "kl": 0.029296875,
      "learning_rate": 8.394736842105262e-07,
      "loss": 0.0012,
      "reward": 1.84375,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.90625,
      "step": 61
    },
    {
      "completion_length": 212.625,
      "epoch": 1.631578947368421,
      "grad_norm": 2.238463297830657,
      "kl": 0.033935546875,
      "learning_rate": 8.368421052631579e-07,
      "loss": 0.0014,
      "reward": 1.8125,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.90625,
      "step": 62
    },
    {
      "completion_length": 201.625,
      "epoch": 1.6578947368421053,
      "grad_norm": 1.1227656281112715,
      "kl": 0.0311279296875,
      "learning_rate": 8.342105263157895e-07,
      "loss": 0.0012,
      "reward": 1.84375,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 63
    },
    {
      "completion_length": 209.28125,
      "epoch": 1.6842105263157894,
      "grad_norm": 1.9439659057227383,
      "kl": 0.024658203125,
      "learning_rate": 8.315789473684211e-07,
      "loss": 0.001,
      "reward": 1.8125,
      "reward_std": 0.17423085868358612,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 64
    },
    {
      "completion_length": 180.625,
      "epoch": 1.7105263157894737,
      "grad_norm": 3.463850548121871,
      "kl": 0.0322265625,
      "learning_rate": 8.289473684210527e-07,
      "loss": 0.0013,
      "reward": 1.75,
      "reward_std": 0.18217839300632477,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.9375,
      "step": 65
    },
    {
      "completion_length": 197.125,
      "epoch": 1.736842105263158,
      "grad_norm": 3.8258176830789394,
      "kl": 0.03955078125,
      "learning_rate": 8.263157894736841e-07,
      "loss": 0.0016,
      "reward": 1.78125,
      "reward_std": 0.3125,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.875,
      "step": 66
    },
    {
      "completion_length": 189.9375,
      "epoch": 1.763157894736842,
      "grad_norm": 2.491889740867337,
      "kl": 0.0289306640625,
      "learning_rate": 8.236842105263157e-07,
      "loss": 0.0012,
      "reward": 1.625,
      "reward_std": 0.32216876745224,
      "rewards/accuracy_reward": 0.6875,
      "rewards/format_reward": 0.9375,
      "step": 67
    },
    {
      "completion_length": 207.90625,
      "epoch": 1.7894736842105263,
      "grad_norm": 1.9661579400854214,
      "kl": 0.0228271484375,
      "learning_rate": 8.210526315789473e-07,
      "loss": 0.0009,
      "reward": 1.8125,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 68
    },
    {
      "completion_length": 191.46875,
      "epoch": 1.8157894736842106,
      "grad_norm": 2.0969574556289436,
      "kl": 0.0238037109375,
      "learning_rate": 8.184210526315789e-07,
      "loss": 0.0009,
      "reward": 1.78125,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.9375,
      "step": 69
    },
    {
      "completion_length": 195.21875,
      "epoch": 1.8421052631578947,
      "grad_norm": 2.7728294805183435,
      "kl": 0.033203125,
      "learning_rate": 8.157894736842105e-07,
      "loss": 0.0013,
      "reward": 1.75,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.78125,
      "rewards/format_reward": 0.96875,
      "step": 70
    },
    {
      "completion_length": 186.5,
      "epoch": 1.868421052631579,
      "grad_norm": 2.7390859558965324,
      "kl": 0.0296630859375,
      "learning_rate": 8.131578947368421e-07,
      "loss": 0.0012,
      "reward": 1.75,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.9375,
      "step": 71
    },
    {
      "completion_length": 180.3125,
      "epoch": 1.8947368421052633,
      "grad_norm": 1.1232297875868242,
      "kl": 0.0223388671875,
      "learning_rate": 8.105263157894736e-07,
      "loss": 0.0009,
      "reward": 1.84375,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 1.0,
      "step": 72
    },
    {
      "completion_length": 170.5,
      "epoch": 1.9210526315789473,
      "grad_norm": 2.513759360761248,
      "kl": 0.035888671875,
      "learning_rate": 8.078947368421052e-07,
      "loss": 0.0014,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 1.0,
      "step": 73
    },
    {
      "completion_length": 179.8125,
      "epoch": 1.9473684210526314,
      "grad_norm": 1.5397907763128271,
      "kl": 0.03076171875,
      "learning_rate": 8.052631578947368e-07,
      "loss": 0.0012,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 74
    },
    {
      "completion_length": 192.90625,
      "epoch": 1.973684210526316,
      "grad_norm": 2.3945832564462486,
      "kl": 0.030517578125,
      "learning_rate": 8.026315789473685e-07,
      "loss": 0.0012,
      "reward": 1.8125,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.90625,
      "step": 75
    },
    {
      "completion_length": 173.1999969482422,
      "epoch": 2.0,
      "grad_norm": 0.21034421259552658,
      "kl": 0.0283203125,
      "learning_rate": 8e-07,
      "loss": 0.0011,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 76
    },
    {
      "completion_length": 167.34375,
      "epoch": 2.026315789473684,
      "grad_norm": 0.16707349379321665,
      "kl": 0.0234375,
      "learning_rate": 7.973684210526315e-07,
      "loss": 0.0009,
      "reward": 1.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 77
    },
    {
      "completion_length": 180.21875,
      "epoch": 2.0526315789473686,
      "grad_norm": 2.2651586426224957,
      "kl": 0.033935546875,
      "learning_rate": 7.947368421052631e-07,
      "loss": 0.0014,
      "reward": 1.71875,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.78125,
      "rewards/format_reward": 0.9375,
      "step": 78
    },
    {
      "completion_length": 169.59375,
      "epoch": 2.0789473684210527,
      "grad_norm": 3.313614338083391,
      "kl": 0.0250244140625,
      "learning_rate": 7.921052631578947e-07,
      "loss": 0.001,
      "reward": 1.84375,
      "reward_std": 0.20683756470680237,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 79
    },
    {
      "completion_length": 191.0625,
      "epoch": 2.1052631578947367,
      "grad_norm": 2.47725789712835,
      "kl": 0.026123046875,
      "learning_rate": 7.894736842105263e-07,
      "loss": 0.001,
      "reward": 1.71875,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.96875,
      "step": 80
    },
    {
      "completion_length": 166.5625,
      "epoch": 2.1315789473684212,
      "grad_norm": 2.5401593096604174,
      "kl": 0.025146484375,
      "learning_rate": 7.868421052631579e-07,
      "loss": 0.001,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 81
    },
    {
      "completion_length": 167.125,
      "epoch": 2.1578947368421053,
      "grad_norm": 4.201187568205836,
      "kl": 0.029052734375,
      "learning_rate": 7.842105263157895e-07,
      "loss": 0.0012,
      "reward": 1.84375,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 82
    },
    {
      "completion_length": 176.0,
      "epoch": 2.1842105263157894,
      "grad_norm": 1.4343300785247641,
      "kl": 0.033203125,
      "learning_rate": 7.81578947368421e-07,
      "loss": 0.0013,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 83
    },
    {
      "completion_length": 168.3125,
      "epoch": 2.2105263157894735,
      "grad_norm": 0.16761476129799044,
      "kl": 0.02197265625,
      "learning_rate": 7.789473684210526e-07,
      "loss": 0.0009,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 84
    },
    {
      "completion_length": 181.0625,
      "epoch": 2.236842105263158,
      "grad_norm": 4.446198173885463,
      "kl": 0.02978515625,
      "learning_rate": 7.763157894736841e-07,
      "loss": 0.0012,
      "reward": 1.71875,
      "reward_std": 0.20683756470680237,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.96875,
      "step": 85
    },
    {
      "completion_length": 170.3125,
      "epoch": 2.263157894736842,
      "grad_norm": 1.8263263739181415,
      "kl": 0.03857421875,
      "learning_rate": 7.736842105263157e-07,
      "loss": 0.0015,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 86
    },
    {
      "completion_length": 179.96875,
      "epoch": 2.2894736842105265,
      "grad_norm": 1.381330554307507,
      "kl": 0.033935546875,
      "learning_rate": 7.710526315789473e-07,
      "loss": 0.0014,
      "reward": 1.9375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 87
    },
    {
      "completion_length": 159.34375,
      "epoch": 2.3157894736842106,
      "grad_norm": 0.2946471541942504,
      "kl": 0.0252685546875,
      "learning_rate": 7.684210526315788e-07,
      "loss": 0.001,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 88
    },
    {
      "completion_length": 171.75,
      "epoch": 2.3421052631578947,
      "grad_norm": 1.584493461989643,
      "kl": 0.0296630859375,
      "learning_rate": 7.657894736842105e-07,
      "loss": 0.0012,
      "reward": 1.875,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 89
    },
    {
      "completion_length": 190.65625,
      "epoch": 2.3684210526315788,
      "grad_norm": 1.7628444339170648,
      "kl": 0.0238037109375,
      "learning_rate": 7.631578947368421e-07,
      "loss": 0.001,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 90
    },
    {
      "completion_length": 185.15625,
      "epoch": 2.3947368421052633,
      "grad_norm": 3.6845999878329385,
      "kl": 0.02587890625,
      "learning_rate": 7.605263157894737e-07,
      "loss": 0.001,
      "reward": 1.78125,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.96875,
      "step": 91
    },
    {
      "completion_length": 183.84375,
      "epoch": 2.4210526315789473,
      "grad_norm": 2.3369306813630963,
      "kl": 0.023193359375,
      "learning_rate": 7.578947368421053e-07,
      "loss": 0.0009,
      "reward": 1.8125,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 1.0,
      "step": 92
    },
    {
      "completion_length": 202.0,
      "epoch": 2.4473684210526314,
      "grad_norm": 0.12160989183498855,
      "kl": 0.0206298828125,
      "learning_rate": 7.552631578947369e-07,
      "loss": 0.0008,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 93
    },
    {
      "completion_length": 191.375,
      "epoch": 2.473684210526316,
      "grad_norm": 2.0039936393070352,
      "kl": 0.038818359375,
      "learning_rate": 7.526315789473684e-07,
      "loss": 0.0016,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.90625,
      "step": 94
    },
    {
      "completion_length": 195.4375,
      "epoch": 2.5,
      "grad_norm": 0.9198562015276629,
      "kl": 0.019287109375,
      "learning_rate": 7.5e-07,
      "loss": 0.0008,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 95
    },
    {
      "completion_length": 198.375,
      "epoch": 2.526315789473684,
      "grad_norm": 1.905361197078109,
      "kl": 0.0234375,
      "learning_rate": 7.473684210526315e-07,
      "loss": 0.0009,
      "reward": 1.875,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 96
    },
    {
      "completion_length": 211.96875,
      "epoch": 2.5526315789473686,
      "grad_norm": 2.688187014358546,
      "kl": 0.033447265625,
      "learning_rate": 7.447368421052631e-07,
      "loss": 0.0013,
      "reward": 1.71875,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.90625,
      "step": 97
    },
    {
      "completion_length": 202.71875,
      "epoch": 2.5789473684210527,
      "grad_norm": 0.10763886108168262,
      "kl": 0.020751953125,
      "learning_rate": 7.421052631578947e-07,
      "loss": 0.0008,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 98
    },
    {
      "completion_length": 212.84375,
      "epoch": 2.6052631578947367,
      "grad_norm": 4.567098501948653,
      "kl": 0.026611328125,
      "learning_rate": 7.394736842105262e-07,
      "loss": 0.0011,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 99
    },
    {
      "completion_length": 204.46875,
      "epoch": 2.6315789473684212,
      "grad_norm": 2.22062431618725,
      "kl": 0.021484375,
      "learning_rate": 7.368421052631578e-07,
      "loss": 0.0009,
      "reward": 1.8125,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9375,
      "step": 100
    },
    {
      "completion_length": 203.9375,
      "epoch": 2.6578947368421053,
      "grad_norm": 1.9855490970882193,
      "kl": 0.024169921875,
      "learning_rate": 7.342105263157894e-07,
      "loss": 0.001,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 101
    },
    {
      "completion_length": 195.1875,
      "epoch": 2.6842105263157894,
      "grad_norm": 15.487188747210284,
      "kl": 0.0267333984375,
      "learning_rate": 7.315789473684211e-07,
      "loss": 0.0011,
      "reward": 1.78125,
      "reward_std": 0.38466876745224,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.96875,
      "step": 102
    },
    {
      "completion_length": 207.6875,
      "epoch": 2.7105263157894735,
      "grad_norm": 1.7691677725960633,
      "kl": 0.021728515625,
      "learning_rate": 7.289473684210527e-07,
      "loss": 0.0009,
      "reward": 1.78125,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.96875,
      "step": 103
    },
    {
      "completion_length": 205.1875,
      "epoch": 2.736842105263158,
      "grad_norm": 1.5203098901138603,
      "kl": 0.039306640625,
      "learning_rate": 7.263157894736843e-07,
      "loss": 0.0016,
      "reward": 1.84375,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.9375,
      "step": 104
    },
    {
      "completion_length": 221.75,
      "epoch": 2.763157894736842,
      "grad_norm": 2.3064006637283727,
      "kl": 0.022705078125,
      "learning_rate": 7.236842105263158e-07,
      "loss": 0.0009,
      "reward": 1.84375,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 105
    },
    {
      "completion_length": 220.4375,
      "epoch": 2.7894736842105265,
      "grad_norm": 0.1378506542128109,
      "kl": 0.0194091796875,
      "learning_rate": 7.210526315789473e-07,
      "loss": 0.0008,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 106
    },
    {
      "completion_length": 228.1875,
      "epoch": 2.8157894736842106,
      "grad_norm": 2.092529341782894,
      "kl": 0.0390625,
      "learning_rate": 7.184210526315789e-07,
      "loss": 0.0016,
      "reward": 1.75,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.875,
      "step": 107
    },
    {
      "completion_length": 228.34375,
      "epoch": 2.8421052631578947,
      "grad_norm": 2.211022858437151,
      "kl": 0.0380859375,
      "learning_rate": 7.157894736842105e-07,
      "loss": 0.0015,
      "reward": 1.78125,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.875,
      "step": 108
    },
    {
      "completion_length": 233.84375,
      "epoch": 2.8684210526315788,
      "grad_norm": 2.6608361097986206,
      "kl": 0.037109375,
      "learning_rate": 7.131578947368421e-07,
      "loss": 0.0015,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 109
    },
    {
      "completion_length": 221.90625,
      "epoch": 2.8947368421052633,
      "grad_norm": 1.4299660876064793,
      "kl": 0.0294189453125,
      "learning_rate": 7.105263157894736e-07,
      "loss": 0.0012,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 110
    },
    {
      "completion_length": 215.9375,
      "epoch": 2.9210526315789473,
      "grad_norm": 2.947584639988725,
      "kl": 0.0400390625,
      "learning_rate": 7.078947368421052e-07,
      "loss": 0.0016,
      "reward": 1.8125,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9375,
      "step": 111
    },
    {
      "completion_length": 226.40625,
      "epoch": 2.9473684210526314,
      "grad_norm": 2.398240956616203,
      "kl": 0.0286865234375,
      "learning_rate": 7.052631578947368e-07,
      "loss": 0.0011,
      "reward": 1.8125,
      "reward_std": 0.26933756470680237,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 112
    },
    {
      "completion_length": 215.84375,
      "epoch": 2.973684210526316,
      "grad_norm": 0.8798051326161257,
      "kl": 0.0238037109375,
      "learning_rate": 7.026315789473684e-07,
      "loss": 0.001,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 113
    },
    {
      "completion_length": 190.60000610351562,
      "epoch": 3.0,
      "grad_norm": 2.0124186084740154,
      "kl": 0.0260009765625,
      "learning_rate": 7e-07,
      "loss": 0.0011,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 114
    },
    {
      "completion_length": 238.21875,
      "epoch": 3.026315789473684,
      "grad_norm": 1.5457245532265884,
      "kl": 0.0262451171875,
      "learning_rate": 6.973684210526314e-07,
      "loss": 0.0011,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 115
    },
    {
      "completion_length": 227.03125,
      "epoch": 3.0526315789473686,
      "grad_norm": 1.359045427027539,
      "kl": 0.0238037109375,
      "learning_rate": 6.947368421052631e-07,
      "loss": 0.001,
      "reward": 1.8125,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9375,
      "step": 116
    },
    {
      "completion_length": 237.25,
      "epoch": 3.0789473684210527,
      "grad_norm": 1.428818791833506,
      "kl": 0.028076171875,
      "learning_rate": 6.921052631578947e-07,
      "loss": 0.0011,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.9375,
      "step": 117
    },
    {
      "completion_length": 205.0,
      "epoch": 3.1052631578947367,
      "grad_norm": 1.693130406970002,
      "kl": 0.0245361328125,
      "learning_rate": 6.894736842105263e-07,
      "loss": 0.001,
      "reward": 1.875,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 118
    },
    {
      "completion_length": 236.96875,
      "epoch": 3.1315789473684212,
      "grad_norm": 1.634819729321223,
      "kl": 0.02294921875,
      "learning_rate": 6.868421052631579e-07,
      "loss": 0.0009,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.9375,
      "step": 119
    },
    {
      "completion_length": 220.0625,
      "epoch": 3.1578947368421053,
      "grad_norm": 3.14637521801096,
      "kl": 0.039306640625,
      "learning_rate": 6.842105263157895e-07,
      "loss": 0.0016,
      "reward": 1.8125,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9375,
      "step": 120
    },
    {
      "completion_length": 201.53125,
      "epoch": 3.1842105263157894,
      "grad_norm": 2.3778574866457967,
      "kl": 0.0289306640625,
      "learning_rate": 6.81578947368421e-07,
      "loss": 0.0012,
      "reward": 1.75,
      "reward_std": 0.22706207633018494,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.9375,
      "step": 121
    },
    {
      "completion_length": 229.78125,
      "epoch": 3.2105263157894735,
      "grad_norm": 2.5277288485260327,
      "kl": 0.0294189453125,
      "learning_rate": 6.789473684210526e-07,
      "loss": 0.0012,
      "reward": 1.78125,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.875,
      "step": 122
    },
    {
      "completion_length": 240.5,
      "epoch": 3.236842105263158,
      "grad_norm": 2.042837126991465,
      "kl": 0.0478515625,
      "learning_rate": 6.763157894736842e-07,
      "loss": 0.0019,
      "reward": 1.8125,
      "reward_std": 0.18217839300632477,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.90625,
      "step": 123
    },
    {
      "completion_length": 224.3125,
      "epoch": 3.263157894736842,
      "grad_norm": 2.1974563034268653,
      "kl": 0.040771484375,
      "learning_rate": 6.736842105263158e-07,
      "loss": 0.0016,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 124
    },
    {
      "completion_length": 224.90625,
      "epoch": 3.2894736842105265,
      "grad_norm": 1.4722641563956271,
      "kl": 0.031982421875,
      "learning_rate": 6.710526315789473e-07,
      "loss": 0.0013,
      "reward": 1.8125,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9375,
      "step": 125
    },
    {
      "completion_length": 211.84375,
      "epoch": 3.3157894736842106,
      "grad_norm": 2.093454619766443,
      "kl": 0.033203125,
      "learning_rate": 6.684210526315788e-07,
      "loss": 0.0013,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.9375,
      "step": 126
    },
    {
      "completion_length": 192.46875,
      "epoch": 3.3421052631578947,
      "grad_norm": 1.1372861693420884,
      "kl": 0.0201416015625,
      "learning_rate": 6.657894736842104e-07,
      "loss": 0.0008,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 127
    },
    {
      "completion_length": 214.375,
      "epoch": 3.3684210526315788,
      "grad_norm": 1.9444796335102086,
      "kl": 0.0269775390625,
      "learning_rate": 6.63157894736842e-07,
      "loss": 0.0011,
      "reward": 1.75,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.9375,
      "step": 128
    },
    {
      "completion_length": 209.65625,
      "epoch": 3.3947368421052633,
      "grad_norm": 0.0851305493593319,
      "kl": 0.0203857421875,
      "learning_rate": 6.605263157894737e-07,
      "loss": 0.0008,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 129
    },
    {
      "completion_length": 232.0625,
      "epoch": 3.4210526315789473,
      "grad_norm": 0.12864507491944657,
      "kl": 0.0216064453125,
      "learning_rate": 6.578947368421053e-07,
      "loss": 0.0009,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 130
    },
    {
      "completion_length": 208.875,
      "epoch": 3.4473684210526314,
      "grad_norm": 3.4231669534201856,
      "kl": 0.02001953125,
      "learning_rate": 6.552631578947369e-07,
      "loss": 0.0008,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 131
    },
    {
      "completion_length": 204.5,
      "epoch": 3.473684210526316,
      "grad_norm": 2.21570182693823,
      "kl": 0.048828125,
      "learning_rate": 6.526315789473684e-07,
      "loss": 0.002,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.875,
      "step": 132
    },
    {
      "completion_length": 206.25,
      "epoch": 3.5,
      "grad_norm": 1.517302820797273,
      "kl": 0.032958984375,
      "learning_rate": 6.5e-07,
      "loss": 0.0013,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 133
    },
    {
      "completion_length": 217.1875,
      "epoch": 3.526315789473684,
      "grad_norm": 2.2833303387756936,
      "kl": 0.0301513671875,
      "learning_rate": 6.473684210526316e-07,
      "loss": 0.0012,
      "reward": 1.75,
      "reward_std": 0.3071783781051636,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.90625,
      "step": 134
    },
    {
      "completion_length": 230.21875,
      "epoch": 3.5526315789473686,
      "grad_norm": 2.6381647603485794,
      "kl": 0.024658203125,
      "learning_rate": 6.447368421052632e-07,
      "loss": 0.001,
      "reward": 1.875,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.90625,
      "step": 135
    },
    {
      "completion_length": 207.21875,
      "epoch": 3.5789473684210527,
      "grad_norm": 2.3085400870612287,
      "kl": 0.0302734375,
      "learning_rate": 6.421052631578947e-07,
      "loss": 0.0012,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 136
    },
    {
      "completion_length": 209.1875,
      "epoch": 3.6052631578947367,
      "grad_norm": 2.4121255629772715,
      "kl": 0.02685546875,
      "learning_rate": 6.394736842105262e-07,
      "loss": 0.0011,
      "reward": 1.8125,
      "reward_std": 0.32216876745224,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 137
    },
    {
      "completion_length": 205.46875,
      "epoch": 3.6315789473684212,
      "grad_norm": 2.0949973089334737,
      "kl": 0.023193359375,
      "learning_rate": 6.368421052631578e-07,
      "loss": 0.0009,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.90625,
      "step": 138
    },
    {
      "completion_length": 194.59375,
      "epoch": 3.6578947368421053,
      "grad_norm": 1.7090110132102732,
      "kl": 0.0260009765625,
      "learning_rate": 6.342105263157894e-07,
      "loss": 0.001,
      "reward": 1.8125,
      "reward_std": 0.17423085868358612,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 139
    },
    {
      "completion_length": 213.90625,
      "epoch": 3.6842105263157894,
      "grad_norm": 0.8764355994796962,
      "kl": 0.0245361328125,
      "learning_rate": 6.31578947368421e-07,
      "loss": 0.001,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 1.0,
      "step": 140
    },
    {
      "completion_length": 224.875,
      "epoch": 3.7105263157894735,
      "grad_norm": 2.2541970263043556,
      "kl": 0.0264892578125,
      "learning_rate": 6.289473684210526e-07,
      "loss": 0.0011,
      "reward": 1.875,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.9375,
      "step": 141
    },
    {
      "completion_length": 209.8125,
      "epoch": 3.736842105263158,
      "grad_norm": 3.877348291575896,
      "kl": 0.029296875,
      "learning_rate": 6.263157894736842e-07,
      "loss": 0.0012,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 142
    },
    {
      "completion_length": 196.8125,
      "epoch": 3.763157894736842,
      "grad_norm": 1.9426605971843807,
      "kl": 0.02734375,
      "learning_rate": 6.236842105263158e-07,
      "loss": 0.0011,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 1.0,
      "step": 143
    },
    {
      "completion_length": 214.78125,
      "epoch": 3.7894736842105265,
      "grad_norm": 1.1912038149884154,
      "kl": 0.0257568359375,
      "learning_rate": 6.210526315789474e-07,
      "loss": 0.001,
      "reward": 1.9375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 144
    },
    {
      "completion_length": 194.71875,
      "epoch": 3.8157894736842106,
      "grad_norm": 1.0148600812787658,
      "kl": 0.0284423828125,
      "learning_rate": 6.18421052631579e-07,
      "loss": 0.0011,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 145
    },
    {
      "completion_length": 199.75,
      "epoch": 3.8421052631578947,
      "grad_norm": 2.2871586093721183,
      "kl": 0.023681640625,
      "learning_rate": 6.157894736842105e-07,
      "loss": 0.0009,
      "reward": 1.78125,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.9375,
      "step": 146
    },
    {
      "completion_length": 207.40625,
      "epoch": 3.8684210526315788,
      "grad_norm": 3.206835111803621,
      "kl": 0.0286865234375,
      "learning_rate": 6.131578947368421e-07,
      "loss": 0.0011,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 147
    },
    {
      "completion_length": 187.96875,
      "epoch": 3.8947368421052633,
      "grad_norm": 1.53853550529444,
      "kl": 0.030517578125,
      "learning_rate": 6.105263157894736e-07,
      "loss": 0.0012,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 148
    },
    {
      "completion_length": 186.9375,
      "epoch": 3.9210526315789473,
      "grad_norm": 6.971515446773084,
      "kl": 0.0277099609375,
      "learning_rate": 6.078947368421052e-07,
      "loss": 0.0011,
      "reward": 1.78125,
      "reward_std": 0.16456207633018494,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.96875,
      "step": 149
    },
    {
      "completion_length": 206.53125,
      "epoch": 3.9473684210526314,
      "grad_norm": 0.13450382946284248,
      "kl": 0.0255126953125,
      "learning_rate": 6.052631578947368e-07,
      "loss": 0.001,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 150
    },
    {
      "completion_length": 194.53125,
      "epoch": 3.973684210526316,
      "grad_norm": 2.2091993045896023,
      "kl": 0.032470703125,
      "learning_rate": 6.026315789473684e-07,
      "loss": 0.0013,
      "reward": 1.875,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 151
    },
    {
      "completion_length": 196.40000915527344,
      "epoch": 4.0,
      "grad_norm": 1.3188375542411617,
      "kl": 0.0269775390625,
      "learning_rate": 6e-07,
      "loss": 0.001,
      "reward": 1.8000000715255737,
      "reward_std": 0.10000000149011612,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 1.0,
      "step": 152
    },
    {
      "completion_length": 219.96875,
      "epoch": 4.026315789473684,
      "grad_norm": 1.450378087434608,
      "kl": 0.031982421875,
      "learning_rate": 5.973684210526316e-07,
      "loss": 0.0013,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 153
    },
    {
      "completion_length": 208.5,
      "epoch": 4.052631578947368,
      "grad_norm": 1.360466997392594,
      "kl": 0.02294921875,
      "learning_rate": 5.947368421052631e-07,
      "loss": 0.0009,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 1.0,
      "step": 154
    },
    {
      "completion_length": 211.3125,
      "epoch": 4.078947368421052,
      "grad_norm": 2.0037027871213335,
      "kl": 0.0250244140625,
      "learning_rate": 5.921052631578946e-07,
      "loss": 0.001,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 155
    },
    {
      "completion_length": 216.6875,
      "epoch": 4.105263157894737,
      "grad_norm": 2.313443563639608,
      "kl": 0.025634765625,
      "learning_rate": 5.894736842105262e-07,
      "loss": 0.001,
      "reward": 1.78125,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.9375,
      "step": 156
    },
    {
      "completion_length": 220.5625,
      "epoch": 4.131578947368421,
      "grad_norm": 2.0199190895534085,
      "kl": 0.0240478515625,
      "learning_rate": 5.868421052631579e-07,
      "loss": 0.001,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 157
    },
    {
      "completion_length": 220.96875,
      "epoch": 4.157894736842105,
      "grad_norm": 1.6701959561617372,
      "kl": 0.033447265625,
      "learning_rate": 5.842105263157895e-07,
      "loss": 0.0013,
      "reward": 1.84375,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 1.0,
      "step": 158
    },
    {
      "completion_length": 196.90625,
      "epoch": 4.184210526315789,
      "grad_norm": 1.7086265895226307,
      "kl": 0.0252685546875,
      "learning_rate": 5.81578947368421e-07,
      "loss": 0.001,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 1.0,
      "step": 159
    },
    {
      "completion_length": 222.125,
      "epoch": 4.2105263157894735,
      "grad_norm": 1.0143061245565415,
      "kl": 0.0264892578125,
      "learning_rate": 5.789473684210526e-07,
      "loss": 0.0011,
      "reward": 1.9375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 160
    },
    {
      "completion_length": 219.375,
      "epoch": 4.2368421052631575,
      "grad_norm": 1.3926076080810952,
      "kl": 0.0244140625,
      "learning_rate": 5.763157894736842e-07,
      "loss": 0.001,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 1.0,
      "step": 161
    },
    {
      "completion_length": 223.4375,
      "epoch": 4.2631578947368425,
      "grad_norm": 2.697570795451099,
      "kl": 0.02783203125,
      "learning_rate": 5.736842105263158e-07,
      "loss": 0.0011,
      "reward": 1.78125,
      "reward_std": 0.3125,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.96875,
      "step": 162
    },
    {
      "completion_length": 198.84375,
      "epoch": 4.2894736842105265,
      "grad_norm": 2.723312360174212,
      "kl": 0.051025390625,
      "learning_rate": 5.710526315789474e-07,
      "loss": 0.002,
      "reward": 1.875,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.90625,
      "step": 163
    },
    {
      "completion_length": 209.5625,
      "epoch": 4.315789473684211,
      "grad_norm": 1.2500212363626535,
      "kl": 0.0306396484375,
      "learning_rate": 5.68421052631579e-07,
      "loss": 0.0012,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 1.0,
      "step": 164
    },
    {
      "completion_length": 182.96875,
      "epoch": 4.342105263157895,
      "grad_norm": 4.275280121593393,
      "kl": 0.060791015625,
      "learning_rate": 5.657894736842104e-07,
      "loss": 0.0024,
      "reward": 1.875,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.9375,
      "step": 165
    },
    {
      "completion_length": 213.75,
      "epoch": 4.368421052631579,
      "grad_norm": 1.1686035487358153,
      "kl": 0.020263671875,
      "learning_rate": 5.63157894736842e-07,
      "loss": 0.0008,
      "reward": 1.90625,
      "reward_std": 0.11967839300632477,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 166
    },
    {
      "completion_length": 211.9375,
      "epoch": 4.394736842105263,
      "grad_norm": 2.0545085657135003,
      "kl": 0.03271484375,
      "learning_rate": 5.605263157894736e-07,
      "loss": 0.0013,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 167
    },
    {
      "completion_length": 214.25,
      "epoch": 4.421052631578947,
      "grad_norm": 2.7092593699581244,
      "kl": 0.033935546875,
      "learning_rate": 5.578947368421052e-07,
      "loss": 0.0014,
      "reward": 1.84375,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.90625,
      "step": 168
    },
    {
      "completion_length": 207.375,
      "epoch": 4.447368421052632,
      "grad_norm": 3.818309778210316,
      "kl": 0.034423828125,
      "learning_rate": 5.552631578947368e-07,
      "loss": 0.0014,
      "reward": 1.875,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.875,
      "step": 169
    },
    {
      "completion_length": 205.9375,
      "epoch": 4.473684210526316,
      "grad_norm": 2.368829180909962,
      "kl": 0.048828125,
      "learning_rate": 5.526315789473684e-07,
      "loss": 0.002,
      "reward": 1.8125,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 170
    },
    {
      "completion_length": 207.5,
      "epoch": 4.5,
      "grad_norm": 1.6236162370126592,
      "kl": 0.037841796875,
      "learning_rate": 5.5e-07,
      "loss": 0.0015,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 171
    },
    {
      "completion_length": 207.34375,
      "epoch": 4.526315789473684,
      "grad_norm": 2.3953458691444154,
      "kl": 0.0211181640625,
      "learning_rate": 5.473684210526316e-07,
      "loss": 0.0008,
      "reward": 1.875,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 172
    },
    {
      "completion_length": 210.8125,
      "epoch": 4.552631578947368,
      "grad_norm": 1.486062353837139,
      "kl": 0.0673828125,
      "learning_rate": 5.447368421052632e-07,
      "loss": 0.0027,
      "reward": 1.9375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.9375,
      "step": 173
    },
    {
      "completion_length": 174.53125,
      "epoch": 4.578947368421053,
      "grad_norm": 1.7278209085262042,
      "kl": 0.0225830078125,
      "learning_rate": 5.421052631578948e-07,
      "loss": 0.0009,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 174
    },
    {
      "completion_length": 226.0,
      "epoch": 4.605263157894737,
      "grad_norm": 1.6511067115151974,
      "kl": 0.0242919921875,
      "learning_rate": 5.394736842105264e-07,
      "loss": 0.001,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.9375,
      "step": 175
    },
    {
      "completion_length": 200.28125,
      "epoch": 4.631578947368421,
      "grad_norm": 1.334373475158102,
      "kl": 0.026611328125,
      "learning_rate": 5.368421052631578e-07,
      "loss": 0.0011,
      "reward": 1.875,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 176
    },
    {
      "completion_length": 182.65625,
      "epoch": 4.657894736842105,
      "grad_norm": 2.4814923989942677,
      "kl": 0.04248046875,
      "learning_rate": 5.342105263157894e-07,
      "loss": 0.0017,
      "reward": 1.875,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.9375,
      "step": 177
    },
    {
      "completion_length": 200.03125,
      "epoch": 4.684210526315789,
      "grad_norm": 3.041738594857993,
      "kl": 0.11669921875,
      "learning_rate": 5.31578947368421e-07,
      "loss": 0.0047,
      "reward": 1.84375,
      "reward_std": 0.25966876745224,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.9375,
      "step": 178
    },
    {
      "completion_length": 195.59375,
      "epoch": 4.7105263157894735,
      "grad_norm": 1.899182559287601,
      "kl": 0.03076171875,
      "learning_rate": 5.289473684210526e-07,
      "loss": 0.0012,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 179
    },
    {
      "completion_length": 181.09375,
      "epoch": 4.7368421052631575,
      "grad_norm": 2.708628929951145,
      "kl": 0.032470703125,
      "learning_rate": 5.263157894736842e-07,
      "loss": 0.0013,
      "reward": 1.84375,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.9375,
      "step": 180
    },
    {
      "completion_length": 169.375,
      "epoch": 4.7631578947368425,
      "grad_norm": 2.4496400189287133,
      "kl": 0.02685546875,
      "learning_rate": 5.236842105263157e-07,
      "loss": 0.0011,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 181
    },
    {
      "completion_length": 176.96875,
      "epoch": 4.7894736842105265,
      "grad_norm": 2.9005276777481366,
      "kl": 0.07080078125,
      "learning_rate": 5.210526315789473e-07,
      "loss": 0.0028,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 182
    },
    {
      "completion_length": 189.5625,
      "epoch": 4.815789473684211,
      "grad_norm": 1.7466406933098233,
      "kl": 0.0361328125,
      "learning_rate": 5.184210526315789e-07,
      "loss": 0.0014,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 183
    },
    {
      "completion_length": 182.75,
      "epoch": 4.842105263157895,
      "grad_norm": 3.0915190811109423,
      "kl": 0.03662109375,
      "learning_rate": 5.157894736842106e-07,
      "loss": 0.0015,
      "reward": 1.84375,
      "reward_std": 0.20683756470680237,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.9375,
      "step": 184
    },
    {
      "completion_length": 179.65625,
      "epoch": 4.868421052631579,
      "grad_norm": 1.5329175378095479,
      "kl": 0.03173828125,
      "learning_rate": 5.131578947368422e-07,
      "loss": 0.0013,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 185
    },
    {
      "completion_length": 177.5,
      "epoch": 4.894736842105263,
      "grad_norm": 3.142723176522207,
      "kl": 0.053466796875,
      "learning_rate": 5.105263157894736e-07,
      "loss": 0.0021,
      "reward": 1.8125,
      "reward_std": 0.32216876745224,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.90625,
      "step": 186
    },
    {
      "completion_length": 174.90625,
      "epoch": 4.921052631578947,
      "grad_norm": 1.196148914730756,
      "kl": 0.0264892578125,
      "learning_rate": 5.078947368421052e-07,
      "loss": 0.0011,
      "reward": 1.84375,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.96875,
      "step": 187
    },
    {
      "completion_length": 178.3125,
      "epoch": 4.947368421052632,
      "grad_norm": 2.2072736380699345,
      "kl": 0.03759765625,
      "learning_rate": 5.052631578947368e-07,
      "loss": 0.0015,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.90625,
      "step": 188
    },
    {
      "completion_length": 182.09375,
      "epoch": 4.973684210526316,
      "grad_norm": 2.7998315573284165,
      "kl": 0.0341796875,
      "learning_rate": 5.026315789473684e-07,
      "loss": 0.0014,
      "reward": 1.875,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 189
    },
    {
      "completion_length": 189.1999969482422,
      "epoch": 5.0,
      "grad_norm": 1.893706828876122,
      "kl": 0.028076171875,
      "learning_rate": 5e-07,
      "loss": 0.0013,
      "reward": 2.0,
      "reward_std": 0.10000000149011612,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 190
    },
    {
      "completion_length": 164.15625,
      "epoch": 5.026315789473684,
      "grad_norm": 2.9669522352070454,
      "kl": 0.0262451171875,
      "learning_rate": 4.973684210526316e-07,
      "loss": 0.001,
      "reward": 1.8125,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 1.0,
      "step": 191
    },
    {
      "completion_length": 170.8125,
      "epoch": 5.052631578947368,
      "grad_norm": 12.901112580540412,
      "kl": 0.03466796875,
      "learning_rate": 4.947368421052631e-07,
      "loss": 0.0014,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 192
    },
    {
      "completion_length": 166.125,
      "epoch": 5.078947368421052,
      "grad_norm": 1.3248581834542839,
      "kl": 0.028076171875,
      "learning_rate": 4.921052631578947e-07,
      "loss": 0.0011,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 1.0,
      "step": 193
    },
    {
      "completion_length": 193.8125,
      "epoch": 5.105263157894737,
      "grad_norm": 0.8884292803910787,
      "kl": 0.03857421875,
      "learning_rate": 4.894736842105263e-07,
      "loss": 0.0015,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 194
    },
    {
      "completion_length": 182.40625,
      "epoch": 5.131578947368421,
      "grad_norm": 2.757137196326431,
      "kl": 0.039794921875,
      "learning_rate": 4.868421052631579e-07,
      "loss": 0.0016,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 195
    },
    {
      "completion_length": 178.0,
      "epoch": 5.157894736842105,
      "grad_norm": 1.9816664424042243,
      "kl": 0.033935546875,
      "learning_rate": 4.842105263157895e-07,
      "loss": 0.0014,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 196
    },
    {
      "completion_length": 183.0,
      "epoch": 5.184210526315789,
      "grad_norm": 1.9196290684503896,
      "kl": 0.033935546875,
      "learning_rate": 4.81578947368421e-07,
      "loss": 0.0014,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 197
    },
    {
      "completion_length": 170.9375,
      "epoch": 5.2105263157894735,
      "grad_norm": 2.7310336290366672,
      "kl": 0.039306640625,
      "learning_rate": 4.789473684210526e-07,
      "loss": 0.0016,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.9375,
      "step": 198
    },
    {
      "completion_length": 199.9375,
      "epoch": 5.2368421052631575,
      "grad_norm": 4.563786198068683,
      "kl": 0.032470703125,
      "learning_rate": 4.7631578947368416e-07,
      "loss": 0.0013,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 1.0,
      "step": 199
    },
    {
      "completion_length": 178.875,
      "epoch": 5.2631578947368425,
      "grad_norm": 2.806953879352004,
      "kl": 0.025146484375,
      "learning_rate": 4.7368421052631574e-07,
      "loss": 0.001,
      "reward": 1.875,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.90625,
      "step": 200
    },
    {
      "completion_length": 176.375,
      "epoch": 5.2894736842105265,
      "grad_norm": 7.829372436365277,
      "kl": 0.045166015625,
      "learning_rate": 4.710526315789474e-07,
      "loss": 0.0018,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.90625,
      "step": 201
    },
    {
      "completion_length": 196.78125,
      "epoch": 5.315789473684211,
      "grad_norm": 1.877921158869899,
      "kl": 0.036376953125,
      "learning_rate": 4.6842105263157896e-07,
      "loss": 0.0015,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 202
    },
    {
      "completion_length": 197.40625,
      "epoch": 5.342105263157895,
      "grad_norm": 1.6222324732681357,
      "kl": 0.03369140625,
      "learning_rate": 4.657894736842105e-07,
      "loss": 0.0014,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 203
    },
    {
      "completion_length": 213.15625,
      "epoch": 5.368421052631579,
      "grad_norm": 6.619886502818049,
      "kl": 0.042236328125,
      "learning_rate": 4.6315789473684207e-07,
      "loss": 0.0017,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.90625,
      "step": 204
    },
    {
      "completion_length": 194.75,
      "epoch": 5.394736842105263,
      "grad_norm": 0.9797938214561028,
      "kl": 0.031005859375,
      "learning_rate": 4.6052631578947365e-07,
      "loss": 0.0012,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 205
    },
    {
      "completion_length": 190.40625,
      "epoch": 5.421052631578947,
      "grad_norm": 1.53389371808685,
      "kl": 0.029296875,
      "learning_rate": 4.5789473684210523e-07,
      "loss": 0.0012,
      "reward": 1.875,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 206
    },
    {
      "completion_length": 192.46875,
      "epoch": 5.447368421052632,
      "grad_norm": 2.136939765137761,
      "kl": 0.0308837890625,
      "learning_rate": 4.5526315789473687e-07,
      "loss": 0.0012,
      "reward": 1.8125,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.84375,
      "rewards/format_reward": 0.96875,
      "step": 207
    },
    {
      "completion_length": 212.125,
      "epoch": 5.473684210526316,
      "grad_norm": 1.0475214147359295,
      "kl": 0.03857421875,
      "learning_rate": 4.526315789473684e-07,
      "loss": 0.0015,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 208
    },
    {
      "completion_length": 177.5625,
      "epoch": 5.5,
      "grad_norm": 1.8307066237928853,
      "kl": 0.031494140625,
      "learning_rate": 4.5e-07,
      "loss": 0.0013,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 209
    },
    {
      "completion_length": 201.09375,
      "epoch": 5.526315789473684,
      "grad_norm": 1.3031513844942098,
      "kl": 0.027587890625,
      "learning_rate": 4.4736842105263156e-07,
      "loss": 0.0011,
      "reward": 1.9375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 210
    },
    {
      "completion_length": 196.84375,
      "epoch": 5.552631578947368,
      "grad_norm": 2.760088682963021,
      "kl": 0.03369140625,
      "learning_rate": 4.4473684210526314e-07,
      "loss": 0.0013,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 211
    },
    {
      "completion_length": 193.75,
      "epoch": 5.578947368421053,
      "grad_norm": 1.6739291437505646,
      "kl": 0.02490234375,
      "learning_rate": 4.421052631578947e-07,
      "loss": 0.001,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 212
    },
    {
      "completion_length": 205.4375,
      "epoch": 5.605263157894737,
      "grad_norm": 1.5800944814376778,
      "kl": 0.0272216796875,
      "learning_rate": 4.3947368421052625e-07,
      "loss": 0.0011,
      "reward": 1.96875,
      "reward_std": 0.0625,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.96875,
      "step": 213
    },
    {
      "completion_length": 193.5625,
      "epoch": 5.631578947368421,
      "grad_norm": 1.8111835312620033,
      "kl": 0.036376953125,
      "learning_rate": 4.368421052631579e-07,
      "loss": 0.0015,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.96875,
      "step": 214
    },
    {
      "completion_length": 190.125,
      "epoch": 5.657894736842105,
      "grad_norm": 0.1974692551272531,
      "kl": 0.03173828125,
      "learning_rate": 4.3421052631578947e-07,
      "loss": 0.0013,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 215
    },
    {
      "completion_length": 230.0625,
      "epoch": 5.684210526315789,
      "grad_norm": 1.6873208428311255,
      "kl": 0.030517578125,
      "learning_rate": 4.3157894736842105e-07,
      "loss": 0.0012,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 1.0,
      "step": 216
    },
    {
      "completion_length": 202.03125,
      "epoch": 5.7105263157894735,
      "grad_norm": 3.116825134691819,
      "kl": 0.027099609375,
      "learning_rate": 4.2894736842105263e-07,
      "loss": 0.0011,
      "reward": 1.78125,
      "reward_std": 0.38466876745224,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.84375,
      "step": 217
    },
    {
      "completion_length": 204.5,
      "epoch": 5.7368421052631575,
      "grad_norm": 2.874175288864497,
      "kl": 0.049560546875,
      "learning_rate": 4.2631578947368416e-07,
      "loss": 0.002,
      "reward": 1.78125,
      "reward_std": 0.33183756470680237,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.875,
      "step": 218
    },
    {
      "completion_length": 206.0,
      "epoch": 5.7631578947368425,
      "grad_norm": 3.6065549701065147,
      "kl": 0.03369140625,
      "learning_rate": 4.2368421052631575e-07,
      "loss": 0.0013,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 1.0,
      "step": 219
    },
    {
      "completion_length": 229.4375,
      "epoch": 5.7894736842105265,
      "grad_norm": 0.12087159180956061,
      "kl": 0.02978515625,
      "learning_rate": 4.2105263157894733e-07,
      "loss": 0.0012,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 220
    },
    {
      "completion_length": 209.125,
      "epoch": 5.815789473684211,
      "grad_norm": 1.8675902479008377,
      "kl": 0.0303955078125,
      "learning_rate": 4.1842105263157896e-07,
      "loss": 0.0012,
      "reward": 1.90625,
      "reward_std": 0.1875,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.9375,
      "step": 221
    },
    {
      "completion_length": 209.28125,
      "epoch": 5.842105263157895,
      "grad_norm": 1.5117991062273646,
      "kl": 0.026611328125,
      "learning_rate": 4.1578947368421054e-07,
      "loss": 0.0011,
      "reward": 1.90625,
      "reward_std": 0.13466878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 1.0,
      "step": 222
    },
    {
      "completion_length": 182.4375,
      "epoch": 5.868421052631579,
      "grad_norm": 2.212351078142714,
      "kl": 0.029052734375,
      "learning_rate": 4.1315789473684207e-07,
      "loss": 0.0012,
      "reward": 1.9375,
      "reward_std": 0.125,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.96875,
      "step": 223
    },
    {
      "completion_length": 215.34375,
      "epoch": 5.894736842105263,
      "grad_norm": 0.15800252702932996,
      "kl": 0.0274658203125,
      "learning_rate": 4.1052631578947365e-07,
      "loss": 0.0011,
      "reward": 2.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 1.0,
      "step": 224
    },
    {
      "completion_length": 230.53125,
      "epoch": 5.921052631578947,
      "grad_norm": 2.4908383388400943,
      "kl": 0.041015625,
      "learning_rate": 4.0789473684210524e-07,
      "loss": 0.0016,
      "reward": 1.8125,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.90625,
      "step": 225
    },
    {
      "completion_length": 195.21875,
      "epoch": 5.947368421052632,
      "grad_norm": 3.022857722705474,
      "kl": 0.034912109375,
      "learning_rate": 4.052631578947368e-07,
      "loss": 0.0014,
      "reward": 1.875,
      "reward_std": 0.25,
      "rewards/accuracy_reward": 0.96875,
      "rewards/format_reward": 0.90625,
      "step": 226
    },
    {
      "completion_length": 214.15625,
      "epoch": 5.973684210526316,
      "grad_norm": 9.73752516040991,
      "kl": 0.039306640625,
      "learning_rate": 4.026315789473684e-07,
      "loss": 0.0016,
      "reward": 1.875,
      "reward_std": 0.19716878235340118,
      "rewards/accuracy_reward": 0.90625,
      "rewards/format_reward": 0.96875,
      "step": 227
    },
    {
      "completion_length": 196.8000030517578,
      "epoch": 6.0,
      "grad_norm": 2.2614617007968665,
      "kl": 0.0281982421875,
      "learning_rate": 4e-07,
      "loss": 0.0011,
      "reward": 1.600000023841858,
      "reward_std": 0.515470027923584,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.800000011920929,
      "step": 228
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 380,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 38,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}