{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.992,
  "eval_steps": 10,
  "global_step": 279,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 410.9166774749756,
      "epoch": 0.010666666666666666,
      "grad_norm": 0.5299306511878967,
      "kl": 0.0,
      "learning_rate": 1.0714285714285714e-07,
      "loss": 0.0126,
      "reward": 0.7291666772216558,
      "reward_std": 0.1530931107699871,
      "rewards/accuracy_reward": 0.7291666772216558,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 422.85417556762695,
      "epoch": 0.021333333333333333,
      "grad_norm": 0.4661875367164612,
      "kl": 0.0,
      "learning_rate": 2.1428571428571428e-07,
      "loss": 0.0183,
      "reward": 0.7916666753590107,
      "reward_std": 0.24859581887722015,
      "rewards/accuracy_reward": 0.7916666753590107,
      "rewards/format_reward": 0.0,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 438.2708549499512,
      "epoch": 0.032,
      "grad_norm": 0.8872589468955994,
      "kl": -8.471310138702393e-06,
      "learning_rate": 3.214285714285714e-07,
      "loss": 0.0187,
      "reward": 0.8958333432674408,
      "reward_std": 0.1705273948609829,
      "rewards/accuracy_reward": 0.8958333432674408,
      "rewards/format_reward": 0.0,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 436.6666736602783,
      "epoch": 0.042666666666666665,
      "grad_norm": 0.09845240414142609,
      "kl": -7.5623393058776855e-06,
      "learning_rate": 4.2857142857142857e-07,
      "loss": 0.0108,
      "reward": 0.7916666734963655,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.7916666734963655,
      "rewards/format_reward": 0.0,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 416.70834732055664,
      "epoch": 0.05333333333333334,
      "grad_norm": 0.565118134021759,
      "kl": -5.11854887008667e-06,
      "learning_rate": 5.357142857142858e-07,
      "loss": 0.0196,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 430.27084732055664,
      "epoch": 0.064,
      "grad_norm": 0.9171705842018127,
      "kl": 1.4778226613998413e-05,
      "learning_rate": 6.428571428571428e-07,
      "loss": 0.0091,
      "reward": 0.8750000111758709,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.8750000111758709,
      "rewards/format_reward": 0.0,
      "step": 6
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 400.00000762939453,
      "epoch": 0.07466666666666667,
      "grad_norm": 0.5704061388969421,
      "kl": -8.910894393920898e-06,
      "learning_rate": 7.5e-07,
      "loss": -0.0,
      "reward": 0.9166666679084301,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9166666679084301,
      "rewards/format_reward": 0.0,
      "step": 7
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 342.5833396911621,
      "epoch": 0.08533333333333333,
      "grad_norm": 0.4634876251220703,
      "kl": 3.3229589462280273e-06,
      "learning_rate": 8.571428571428571e-07,
      "loss": 0.0097,
      "reward": 0.9166666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 8
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 446.02084159851074,
      "epoch": 0.096,
      "grad_norm": 0.4667946994304657,
      "kl": 5.4001808166503906e-05,
      "learning_rate": 9.642857142857145e-07,
      "loss": 0.005,
      "reward": 0.7708333358168602,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.7708333358168602,
      "rewards/format_reward": 0.0,
      "step": 9
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 410.5416831970215,
      "epoch": 0.10666666666666667,
      "grad_norm": 0.0004501362855080515,
      "kl": 1.0106712579727173e-05,
      "learning_rate": 1.0714285714285716e-06,
      "loss": 0.0,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 337.3333377838135,
      "epoch": 0.11733333333333333,
      "grad_norm": 0.0002833263133652508,
      "kl": 1.4334917068481445e-05,
      "learning_rate": 1.1785714285714285e-06,
      "loss": 0.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 11
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 459.43750762939453,
      "epoch": 0.128,
      "grad_norm": 1.180604338645935,
      "kl": 0.0019370466470718384,
      "learning_rate": 1.2857142857142856e-06,
      "loss": 0.0508,
      "reward": 0.7291666697710752,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.7291666697710752,
      "rewards/format_reward": 0.0,
      "step": 12
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 333.41667556762695,
      "epoch": 0.13866666666666666,
      "grad_norm": 0.06026441603899002,
      "kl": -1.620757393538952e-05,
      "learning_rate": 1.392857142857143e-06,
      "loss": -0.0034,
      "reward": 0.895833333954215,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.895833333954215,
      "rewards/format_reward": 0.0,
      "step": 13
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 326.9166774749756,
      "epoch": 0.14933333333333335,
      "grad_norm": 0.03230924531817436,
      "kl": -2.1424144506454468e-05,
      "learning_rate": 1.5e-06,
      "loss": -0.0172,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 14
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 446.2500114440918,
      "epoch": 0.16,
      "grad_norm": 0.8342331051826477,
      "kl": 0.005100443959236145,
      "learning_rate": 1.6071428571428572e-06,
      "loss": 0.0341,
      "reward": 0.7708333395421505,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.7708333395421505,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 426.06251525878906,
      "epoch": 0.17066666666666666,
      "grad_norm": 0.014114960096776485,
      "kl": 0.0043880753219127655,
      "learning_rate": 1.7142857142857143e-06,
      "loss": 0.0003,
      "reward": 0.625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 16
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 471.20834732055664,
      "epoch": 0.18133333333333335,
      "grad_norm": 0.015604740008711815,
      "kl": 0.007741432636976242,
      "learning_rate": 1.8214285714285714e-06,
      "loss": 0.0003,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 17
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 458.2708511352539,
      "epoch": 0.192,
      "grad_norm": 0.45530879497528076,
      "kl": 0.021405041217803955,
      "learning_rate": 1.928571428571429e-06,
      "loss": 0.018,
      "reward": 0.7708333358168602,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.7708333358168602,
      "rewards/format_reward": 0.0,
      "step": 18
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 307.43750762939453,
      "epoch": 0.20266666666666666,
      "grad_norm": 0.0004259841807652265,
      "kl": 2.8014183044433594e-05,
      "learning_rate": 2.035714285714286e-06,
      "loss": 0.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 19
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 529.1458492279053,
      "epoch": 0.21333333333333335,
      "grad_norm": 1.0350282192230225,
      "kl": 0.023723304271697998,
      "learning_rate": 2.142857142857143e-06,
      "loss": 0.0356,
      "reward": 0.5625000055879354,
      "reward_std": 0.1705273911356926,
      "rewards/accuracy_reward": 0.5625000055879354,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 365.87500953674316,
      "epoch": 0.224,
      "grad_norm": 0.01775968447327614,
      "kl": 0.00790809583850205,
      "learning_rate": 2.25e-06,
      "loss": 0.0005,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 21
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 454.8958377838135,
      "epoch": 0.23466666666666666,
      "grad_norm": 0.05057643726468086,
      "kl": 0.026109017431735992,
      "learning_rate": 2.357142857142857e-06,
      "loss": 0.0149,
      "reward": 0.7708333376795053,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.7708333376795053,
      "rewards/format_reward": 0.0,
      "step": 22
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 450.50000762939453,
      "epoch": 0.24533333333333332,
      "grad_norm": 0.1461334526538849,
      "kl": 0.0741237998008728,
      "learning_rate": 2.4642857142857143e-06,
      "loss": -0.0041,
      "reward": 0.7500000055879354,
      "reward_std": 0.23899271711707115,
      "rewards/accuracy_reward": 0.7500000055879354,
      "rewards/format_reward": 0.0,
      "step": 23
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 387.18750762939453,
      "epoch": 0.256,
      "grad_norm": 0.5979704260826111,
      "kl": 0.022205054759979248,
      "learning_rate": 2.571428571428571e-06,
      "loss": 0.0201,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 24
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 500.43750762939453,
      "epoch": 0.26666666666666666,
      "grad_norm": 0.7714133262634277,
      "kl": 0.6279669404029846,
      "learning_rate": 2.678571428571429e-06,
      "loss": 0.1006,
      "reward": 0.5000000055879354,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.5000000055879354,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 489.0833435058594,
      "epoch": 0.2773333333333333,
      "grad_norm": 0.1741541475057602,
      "kl": 0.5411486104130745,
      "learning_rate": 2.785714285714286e-06,
      "loss": 0.0444,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 26
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 483.9375190734863,
      "epoch": 0.288,
      "grad_norm": 0.11233610659837723,
      "kl": 0.18515753000974655,
      "learning_rate": 2.892857142857143e-06,
      "loss": 0.0476,
      "reward": 0.8125,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.0,
      "step": 27
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 411.14584732055664,
      "epoch": 0.2986666666666667,
      "grad_norm": 0.0639011412858963,
      "kl": 0.11302819848060608,
      "learning_rate": 3e-06,
      "loss": 0.0116,
      "reward": 0.8333333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.8333333358168602,
      "rewards/format_reward": 0.0,
      "step": 28
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 398.83334732055664,
      "epoch": 0.30933333333333335,
      "grad_norm": 0.020812369883060455,
      "kl": 0.03338313102722168,
      "learning_rate": 2.999882508108334e-06,
      "loss": 0.0026,
      "reward": 0.9166666679084301,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9166666679084301,
      "rewards/format_reward": 0.0,
      "step": 29
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 409.3125114440918,
      "epoch": 0.32,
      "grad_norm": 0.5567188262939453,
      "kl": 0.6645529866218567,
      "learning_rate": 2.9995300508391286e-06,
      "loss": 0.0308,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 435.50001525878906,
      "epoch": 0.33066666666666666,
      "grad_norm": 0.3817005455493927,
      "kl": 0.3913140296936035,
      "learning_rate": 2.998942683406879e-06,
      "loss": 0.0262,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 31
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 345.9375114440918,
      "epoch": 0.3413333333333333,
      "grad_norm": 0.01121461670845747,
      "kl": 0.0008133649826049805,
      "learning_rate": 2.998120497826133e-06,
      "loss": 0.0072,
      "reward": 0.895833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.895833333954215,
      "rewards/format_reward": 0.0,
      "step": 32
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 313.35417556762695,
      "epoch": 0.352,
      "grad_norm": 0.05279409885406494,
      "kl": 0.002323627471923828,
      "learning_rate": 2.9970636228970757e-06,
      "loss": -0.0081,
      "reward": 0.7916666753590107,
      "reward_std": 0.24859581515192986,
      "rewards/accuracy_reward": 0.7916666753590107,
      "rewards/format_reward": 0.0,
      "step": 33
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 269.2291736602783,
      "epoch": 0.3626666666666667,
      "grad_norm": 0.054334308952093124,
      "kl": 0.002706080675125122,
      "learning_rate": 2.995772224185354e-06,
      "loss": 0.0001,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 34
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 453.14584732055664,
      "epoch": 0.37333333333333335,
      "grad_norm": 0.48555248975753784,
      "kl": 0.8578243255615234,
      "learning_rate": 2.9942465039961373e-06,
      "loss": 0.1078,
      "reward": 0.687500013038516,
      "reward_std": 0.21764283254742622,
      "rewards/accuracy_reward": 0.687500013038516,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 378.81250953674316,
      "epoch": 0.384,
      "grad_norm": 0.047343045473098755,
      "kl": 0.0028142929077148438,
      "learning_rate": 2.992486701342427e-06,
      "loss": 0.0103,
      "reward": 0.7916666734963655,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.7916666734963655,
      "rewards/format_reward": 0.0,
      "step": 36
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 407.458345413208,
      "epoch": 0.39466666666666667,
      "grad_norm": 0.37469470500946045,
      "kl": 0.9185905456542969,
      "learning_rate": 2.9904930919076145e-06,
      "loss": 0.0625,
      "reward": 0.7291666679084301,
      "reward_std": 0.1705273911356926,
      "rewards/accuracy_reward": 0.7291666679084301,
      "rewards/format_reward": 0.0,
      "step": 37
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 441.22918128967285,
      "epoch": 0.4053333333333333,
      "grad_norm": 0.34893208742141724,
      "kl": 1.2932720184326172,
      "learning_rate": 2.9882659880022905e-06,
      "loss": 0.1287,
      "reward": 0.7083333376795053,
      "reward_std": 0.24859581887722015,
      "rewards/accuracy_reward": 0.7083333376795053,
      "rewards/format_reward": 0.0,
      "step": 38
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 406.0208396911621,
      "epoch": 0.416,
      "grad_norm": 0.1943722516298294,
      "kl": 0.6224098205566406,
      "learning_rate": 2.985805738515323e-06,
      "loss": 0.0537,
      "reward": 0.8541666697710752,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8541666697710752,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 370.4166774749756,
      "epoch": 0.4266666666666667,
      "grad_norm": 0.4532138705253601,
      "kl": 0.5034923553466797,
      "learning_rate": 2.9831127288592004e-06,
      "loss": 0.0644,
      "reward": 0.8333333358168602,
      "reward_std": 0.20148037374019623,
      "rewards/accuracy_reward": 0.8333333358168602,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 506.4166851043701,
      "epoch": 0.43733333333333335,
      "grad_norm": 0.2463473230600357,
      "kl": 1.607778549194336,
      "learning_rate": 2.9801873809096543e-06,
      "loss": 0.1184,
      "reward": 0.7291666753590107,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.7291666753590107,
      "rewards/format_reward": 0.0,
      "step": 41
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 393.9166793823242,
      "epoch": 0.448,
      "grad_norm": 0.7764784097671509,
      "kl": 0.9653873443603516,
      "learning_rate": 2.97703015293957e-06,
      "loss": 0.1261,
      "reward": 0.8125000074505806,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 42
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 422.12501525878906,
      "epoch": 0.45866666666666667,
      "grad_norm": 1.4099422693252563,
      "kl": 1.4124107360839844,
      "learning_rate": 2.9736415395471975e-06,
      "loss": 0.1981,
      "reward": 0.7708333414047956,
      "reward_std": 0.2350771203637123,
      "rewards/accuracy_reward": 0.7708333414047956,
      "rewards/format_reward": 0.0,
      "step": 43
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 312.47917556762695,
      "epoch": 0.4693333333333333,
      "grad_norm": 0.00318885431624949,
      "kl": 0.0023512840270996094,
      "learning_rate": 2.970022071578666e-06,
      "loss": 0.0001,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 44
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 417.3125057220459,
      "epoch": 0.48,
      "grad_norm": 0.4240525960922241,
      "kl": 1.9218940734863281,
      "learning_rate": 2.9661723160448267e-06,
      "loss": 0.1428,
      "reward": 0.8750000055879354,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.8750000055879354,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 384.7083396911621,
      "epoch": 0.49066666666666664,
      "grad_norm": 0.033430058509111404,
      "kl": 0.0019054412841796875,
      "learning_rate": 2.962092876032427e-06,
      "loss": 0.0056,
      "reward": 0.8333333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.8333333358168602,
      "rewards/format_reward": 0.0,
      "step": 46
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 459.4166793823242,
      "epoch": 0.5013333333333333,
      "grad_norm": 1.4057543277740479,
      "kl": 2.4302802085876465,
      "learning_rate": 2.957784390609632e-06,
      "loss": 0.1647,
      "reward": 0.7083333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.7083333358168602,
      "rewards/format_reward": 0.0,
      "step": 47
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 348.0625114440918,
      "epoch": 0.512,
      "grad_norm": 0.002613195450976491,
      "kl": 0.0018477439880371094,
      "learning_rate": 2.9532475347259113e-06,
      "loss": 0.0001,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 48
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 416.31250762939453,
      "epoch": 0.5226666666666666,
      "grad_norm": 0.7081112861633301,
      "kl": 1.2155981063842773,
      "learning_rate": 2.9484830191063057e-06,
      "loss": 0.0487,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 49
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 479.18751525878906,
      "epoch": 0.5333333333333333,
      "grad_norm": 0.11860348284244537,
      "kl": 0.8176593780517578,
      "learning_rate": 2.9434915901400857e-06,
      "loss": 0.0756,
      "reward": 0.7708333414047956,
      "reward_std": 0.2446802221238613,
      "rewards/accuracy_reward": 0.7708333414047956,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 336.14584159851074,
      "epoch": 0.544,
      "grad_norm": 0.05025426298379898,
      "kl": 0.0013027191162109375,
      "learning_rate": 2.938274029763826e-06,
      "loss": 0.0322,
      "reward": 0.8125000074505806,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 51
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 414.06250953674316,
      "epoch": 0.5546666666666666,
      "grad_norm": 0.5244683623313904,
      "kl": 1.343430519104004,
      "learning_rate": 2.9328311553389116e-06,
      "loss": 0.091,
      "reward": 0.791666679084301,
      "reward_std": 0.16661180555820465,
      "rewards/accuracy_reward": 0.791666679084301,
      "rewards/format_reward": 0.0,
      "step": 52
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 418.7083435058594,
      "epoch": 0.5653333333333334,
      "grad_norm": 0.29313525557518005,
      "kl": 0.7863144874572754,
      "learning_rate": 2.927163819523492e-06,
      "loss": 0.0602,
      "reward": 0.6041666679084301,
      "reward_std": 0.1705273911356926,
      "rewards/accuracy_reward": 0.6041666679084301,
      "rewards/format_reward": 0.0,
      "step": 53
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 369.54168128967285,
      "epoch": 0.576,
      "grad_norm": 0.413883239030838,
      "kl": 0.3720831871032715,
      "learning_rate": 2.9212729101389084e-06,
      "loss": 0.0502,
      "reward": 0.7916666734963655,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.7916666734963655,
      "rewards/format_reward": 0.0,
      "step": 54
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 555.5625190734863,
      "epoch": 0.5866666666666667,
      "grad_norm": 0.8179274797439575,
      "kl": 1.2964344024658203,
      "learning_rate": 2.9151593500306103e-06,
      "loss": 0.0543,
      "reward": 0.5625,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 374.7291736602783,
      "epoch": 0.5973333333333334,
      "grad_norm": 0.002118342788890004,
      "kl": 0.0013017654418945312,
      "learning_rate": 2.9088240969235863e-06,
      "loss": 0.0001,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 56
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 328.9166793823242,
      "epoch": 0.608,
      "grad_norm": 0.0338607020676136,
      "kl": 0.0011858940124511719,
      "learning_rate": 2.902268143272333e-06,
      "loss": -0.0089,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 57
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 351.5416793823242,
      "epoch": 0.6186666666666667,
      "grad_norm": 0.5058808922767639,
      "kl": 0.20941162109375,
      "learning_rate": 2.895492516105378e-06,
      "loss": 0.0457,
      "reward": 0.8125000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 58
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 574.2708396911621,
      "epoch": 0.6293333333333333,
      "grad_norm": 0.6014701128005981,
      "kl": 0.8367283344268799,
      "learning_rate": 2.888498276864393e-06,
      "loss": 0.0236,
      "reward": 0.6458333358168602,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.6458333358168602,
      "rewards/format_reward": 0.0,
      "step": 59
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 444.9166774749756,
      "epoch": 0.64,
      "grad_norm": 0.07577009499073029,
      "kl": 0.07219600677490234,
      "learning_rate": 2.8812865212379087e-06,
      "loss": 0.0071,
      "reward": 0.8125,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 378.2291736602783,
      "epoch": 0.6506666666666666,
      "grad_norm": 0.0443667508661747,
      "kl": 0.0021257400512695312,
      "learning_rate": 2.8738583789896743e-06,
      "loss": 0.0073,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206207633018494,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 61
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 514.6458396911621,
      "epoch": 0.6613333333333333,
      "grad_norm": 0.8298602104187012,
      "kl": 0.23528003692626953,
      "learning_rate": 2.866215013781668e-06,
      "loss": 0.0501,
      "reward": 0.7291666753590107,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.7291666753590107,
      "rewards/format_reward": 0.0,
      "step": 62
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 423.06250762939453,
      "epoch": 0.672,
      "grad_norm": 0.5626036524772644,
      "kl": 0.08101940155029297,
      "learning_rate": 2.858357622991806e-06,
      "loss": 0.0213,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 63
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 396.2916774749756,
      "epoch": 0.6826666666666666,
      "grad_norm": 0.4362681210041046,
      "kl": 0.0890655517578125,
      "learning_rate": 2.8502874375263646e-06,
      "loss": 0.0404,
      "reward": 0.8958333432674408,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.8958333432674408,
      "rewards/format_reward": 0.0,
      "step": 64
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 349.8958435058594,
      "epoch": 0.6933333333333334,
      "grad_norm": 0.42613545060157776,
      "kl": 0.058725833892822266,
      "learning_rate": 2.8420057216271533e-06,
      "loss": 0.0178,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 425.60417556762695,
      "epoch": 0.704,
      "grad_norm": 0.44066762924194336,
      "kl": 0.1816251277923584,
      "learning_rate": 2.8335137726734607e-06,
      "loss": 0.0316,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 66
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 369.0833435058594,
      "epoch": 0.7146666666666667,
      "grad_norm": 0.12847420573234558,
      "kl": 0.4885525703430176,
      "learning_rate": 2.8248129209788164e-06,
      "loss": 0.061,
      "reward": 0.8541666753590107,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.8541666753590107,
      "rewards/format_reward": 0.0,
      "step": 67
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 384.7708435058594,
      "epoch": 0.7253333333333334,
      "grad_norm": 0.29157620668411255,
      "kl": 0.36469125747680664,
      "learning_rate": 2.8159045295825857e-06,
      "loss": 0.0067,
      "reward": 0.7916666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 68
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 307.75000953674316,
      "epoch": 0.736,
      "grad_norm": 0.02736111357808113,
      "kl": 0.001550436019897461,
      "learning_rate": 2.806789994036445e-06,
      "loss": -0.0033,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 69
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 382.1250057220459,
      "epoch": 0.7466666666666667,
      "grad_norm": 0.057907938957214355,
      "kl": 0.44846928119659424,
      "learning_rate": 2.7974707421857584e-06,
      "loss": 0.0653,
      "reward": 0.833333333954215,
      "reward_std": 0.11949636042118073,
      "rewards/accuracy_reward": 0.833333333954215,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 505.02084732055664,
      "epoch": 0.7573333333333333,
      "grad_norm": 0.9753615856170654,
      "kl": 1.7791624069213867,
      "learning_rate": 2.7879482339458973e-06,
      "loss": 0.1034,
      "reward": 0.708333333954215,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.708333333954215,
      "rewards/format_reward": 0.0,
      "step": 71
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 432.9791831970215,
      "epoch": 0.768,
      "grad_norm": 0.3005034625530243,
      "kl": 0.9472742080688477,
      "learning_rate": 2.778223961073537e-06,
      "loss": 0.0849,
      "reward": 0.6666666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.6666666716337204,
      "rewards/format_reward": 0.0,
      "step": 72
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 353.5416793823242,
      "epoch": 0.7786666666666666,
      "grad_norm": 0.2883008122444153,
      "kl": 0.44875526428222656,
      "learning_rate": 2.7682994469329646e-06,
      "loss": 0.0592,
      "reward": 0.8750000074505806,
      "reward_std": 0.18404608592391014,
      "rewards/accuracy_reward": 0.8750000074505806,
      "rewards/format_reward": 0.0,
      "step": 73
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 426.81251525878906,
      "epoch": 0.7893333333333333,
      "grad_norm": 0.286614328622818,
      "kl": 0.5456581115722656,
      "learning_rate": 2.7581762462574334e-06,
      "loss": 0.0588,
      "reward": 0.8125000074505806,
      "reward_std": 0.2621145099401474,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 74
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.6875152587891,
      "epoch": 0.8,
      "grad_norm": 0.09037178009748459,
      "kl": 1.1991617679595947,
      "learning_rate": 2.747855944905606e-06,
      "loss": 0.0889,
      "reward": 0.6666666772216558,
      "reward_std": 0.22155842930078506,
      "rewards/accuracy_reward": 0.6666666772216558,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 396.75001525878906,
      "epoch": 0.8106666666666666,
      "grad_norm": 0.04595573991537094,
      "kl": 0.0017056465148925781,
      "learning_rate": 2.73734015961312e-06,
      "loss": 0.0054,
      "reward": 0.8958333432674408,
      "reward_std": 0.1705273911356926,
      "rewards/accuracy_reward": 0.8958333432674408,
      "rewards/format_reward": 0.0,
      "step": 76
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 421.25001525878906,
      "epoch": 0.8213333333333334,
      "grad_norm": 0.6178137063980103,
      "kl": 0.7442665100097656,
      "learning_rate": 2.7266305377393183e-06,
      "loss": 0.1024,
      "reward": 0.8541666716337204,
      "reward_std": 0.18796168267726898,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 77
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 423.0625114440918,
      "epoch": 0.832,
      "grad_norm": 0.040635429322719574,
      "kl": 0.18666815757751465,
      "learning_rate": 2.7157287570091765e-06,
      "loss": 0.0134,
      "reward": 0.8541666679084301,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.8541666679084301,
      "rewards/format_reward": 0.0,
      "step": 78
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 460.87500762939453,
      "epoch": 0.8426666666666667,
      "grad_norm": 0.2019713968038559,
      "kl": 0.32639598846435547,
      "learning_rate": 2.7046365252504835e-06,
      "loss": 0.029,
      "reward": 0.8333333432674408,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.8333333432674408,
      "rewards/format_reward": 0.0,
      "step": 79
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 368.5000114440918,
      "epoch": 0.8533333333333334,
      "grad_norm": 0.07597187906503677,
      "kl": 1.0420165061950684,
      "learning_rate": 2.693355580126295e-06,
      "loss": 0.0706,
      "reward": 0.8541666679084301,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.8541666679084301,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 421.4375057220459,
      "epoch": 0.864,
      "grad_norm": 0.33045217394828796,
      "kl": 0.7187583446502686,
      "learning_rate": 2.6818876888627203e-06,
      "loss": 0.0402,
      "reward": 0.7500000074505806,
      "reward_std": 0.18404608592391014,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.0,
      "step": 81
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 438.87500762939453,
      "epoch": 0.8746666666666667,
      "grad_norm": 1.0251787900924683,
      "kl": 2.417191505432129,
      "learning_rate": 2.6702346479720777e-06,
      "loss": 0.132,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 82
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 351.25000762939453,
      "epoch": 0.8853333333333333,
      "grad_norm": 0.25971361994743347,
      "kl": 0.2927316427230835,
      "learning_rate": 2.6583982829714576e-06,
      "loss": 0.0172,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 83
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 382.8541736602783,
      "epoch": 0.896,
      "grad_norm": 0.3502584993839264,
      "kl": 0.24036216735839844,
      "learning_rate": 2.6463804480967447e-06,
      "loss": 0.0407,
      "reward": 0.9375000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.9375000074505806,
      "rewards/format_reward": 0.0,
      "step": 84
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 392.93751335144043,
      "epoch": 0.9066666666666666,
      "grad_norm": 0.3297070562839508,
      "kl": 0.43697142601013184,
      "learning_rate": 2.6341830260121445e-06,
      "loss": 0.0187,
      "reward": 0.7083333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.7083333358168602,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 438.6875114440918,
      "epoch": 0.9173333333333333,
      "grad_norm": 0.11217549443244934,
      "kl": 0.9480280876159668,
      "learning_rate": 2.6218079275152485e-06,
      "loss": 0.0422,
      "reward": 0.6666666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.6666666716337204,
      "rewards/format_reward": 0.0,
      "step": 86
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 421.08334732055664,
      "epoch": 0.928,
      "grad_norm": 0.12261732667684555,
      "kl": 0.28173160552978516,
      "learning_rate": 2.609257091237699e-06,
      "loss": 0.069,
      "reward": 0.8125000111758709,
      "reward_std": 0.2446802258491516,
      "rewards/accuracy_reward": 0.8125000111758709,
      "rewards/format_reward": 0.0,
      "step": 87
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 501.3958396911621,
      "epoch": 0.9386666666666666,
      "grad_norm": 0.6337574124336243,
      "kl": 1.5301032066345215,
      "learning_rate": 2.5965324833414926e-06,
      "loss": 0.1127,
      "reward": 0.5833333358168602,
      "reward_std": 0.16661180555820465,
      "rewards/accuracy_reward": 0.5833333358168602,
      "rewards/format_reward": 0.0,
      "step": 88
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 602.4166851043701,
      "epoch": 0.9493333333333334,
      "grad_norm": 0.9542731642723083,
      "kl": 1.2789087295532227,
      "learning_rate": 2.5836360972109642e-06,
      "loss": 0.1305,
      "reward": 0.5416666772216558,
      "reward_std": 0.3680921718478203,
      "rewards/accuracy_reward": 0.5416666772216558,
      "rewards/format_reward": 0.0,
      "step": 89
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 352.4166774749756,
      "epoch": 0.96,
      "grad_norm": 0.8385551571846008,
      "kl": 0.3083925247192383,
      "learning_rate": 2.5705699531405174e-06,
      "loss": 0.0655,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 396.2291793823242,
      "epoch": 0.9706666666666667,
      "grad_norm": 0.10351618379354477,
      "kl": 0.11728858947753906,
      "learning_rate": 2.55733609801813e-06,
      "loss": 0.0038,
      "reward": 0.7708333395421505,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.7708333395421505,
      "rewards/format_reward": 0.0,
      "step": 91
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 361.708345413208,
      "epoch": 0.9813333333333333,
      "grad_norm": 0.31291908025741577,
      "kl": 0.4343738555908203,
      "learning_rate": 2.5439366050046985e-06,
      "loss": 0.0274,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 92
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 518.5625190734863,
      "epoch": 0.992,
      "grad_norm": 0.18174269795417786,
      "kl": 1.6261186599731445,
      "learning_rate": 2.5303735732092655e-06,
      "loss": 0.1018,
      "reward": 0.5416666734963655,
      "reward_std": 0.30354244634509087,
      "rewards/accuracy_reward": 0.5416666734963655,
      "rewards/format_reward": 0.0,
      "step": 93
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 425.37501525878906,
      "epoch": 1.0106666666666666,
      "grad_norm": 0.16784830391407013,
      "kl": 0.8756067752838135,
      "learning_rate": 2.51664912736018e-06,
      "loss": 0.0937,
      "reward": 0.7708333358168602,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.7708333358168602,
      "rewards/format_reward": 0.0,
      "step": 94
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 466.3541717529297,
      "epoch": 1.0213333333333334,
      "grad_norm": 0.2611272633075714,
      "kl": 1.1532807350158691,
      "learning_rate": 2.5027654174722495e-06,
      "loss": 0.0989,
      "reward": 0.791666679084301,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.791666679084301,
      "rewards/format_reward": 0.0,
      "step": 95
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 377.7708435058594,
      "epoch": 1.032,
      "grad_norm": 0.008225666359066963,
      "kl": 0.005245685577392578,
      "learning_rate": 2.4887246185099237e-06,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 96
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 407.2916774749756,
      "epoch": 1.0426666666666666,
      "grad_norm": 0.38802120089530945,
      "kl": 0.638700008392334,
      "learning_rate": 2.474528930046578e-06,
      "loss": 0.0255,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 97
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 357.0208396911621,
      "epoch": 1.0533333333333332,
      "grad_norm": 0.1373687982559204,
      "kl": 0.00467681884765625,
      "learning_rate": 2.4601805759199317e-06,
      "loss": -0.0094,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 98
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.1875076293945,
      "epoch": 1.064,
      "grad_norm": 0.5215266346931458,
      "kl": 1.7781472206115723,
      "learning_rate": 2.445681803883678e-06,
      "loss": 0.1212,
      "reward": 0.6666666734963655,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.6666666734963655,
      "rewards/format_reward": 0.0,
      "step": 99
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 403.22917556762695,
      "epoch": 1.0746666666666667,
      "grad_norm": 0.6477581262588501,
      "kl": 0.8606910705566406,
      "learning_rate": 2.4310348852553554e-06,
      "loss": 0.1011,
      "reward": 0.7500000093132257,
      "reward_std": 0.23116152733564377,
      "rewards/accuracy_reward": 0.7500000093132257,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 442.1041717529297,
      "epoch": 1.0853333333333333,
      "grad_norm": 0.5244895815849304,
      "kl": 0.815155029296875,
      "learning_rate": 2.4162421145605308e-06,
      "loss": 0.0327,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 101
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 475.5208396911621,
      "epoch": 1.096,
      "grad_norm": 0.3228438198566437,
      "kl": 0.40970706939697266,
      "learning_rate": 2.4013058091733546e-06,
      "loss": 0.0419,
      "reward": 0.854166679084301,
      "reward_std": 0.2350771203637123,
      "rewards/accuracy_reward": 0.854166679084301,
      "rewards/format_reward": 0.0,
      "step": 102
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 384.1666774749756,
      "epoch": 1.1066666666666667,
      "grad_norm": 0.5393307209014893,
      "kl": 0.36811351776123047,
      "learning_rate": 2.3862283089535265e-06,
      "loss": 0.0679,
      "reward": 0.9375000074505806,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.9375000074505806,
      "rewards/format_reward": 0.0,
      "step": 103
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 398.56250953674316,
      "epoch": 1.1173333333333333,
      "grad_norm": 0.5991567373275757,
      "kl": 0.8719997406005859,
      "learning_rate": 2.371011975879742e-06,
      "loss": 0.0415,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 104
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 336.1875057220459,
      "epoch": 1.1280000000000001,
      "grad_norm": 0.33208516240119934,
      "kl": 0.25864124298095703,
      "learning_rate": 2.3556591936796777e-06,
      "loss": 0.0648,
      "reward": 0.9166666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 105
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 339.81250762939453,
      "epoch": 1.1386666666666667,
      "grad_norm": 0.4861837923526764,
      "kl": 0.20621585845947266,
      "learning_rate": 2.340172367456564e-06,
      "loss": 0.0464,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 106
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 474.6666736602783,
      "epoch": 1.1493333333333333,
      "grad_norm": 0.7674474120140076,
      "kl": 1.2895288467407227,
      "learning_rate": 2.3245539233124133e-06,
      "loss": 0.1183,
      "reward": 0.7500000093132257,
      "reward_std": 0.23116153478622437,
      "rewards/accuracy_reward": 0.7500000093132257,
      "rewards/format_reward": 0.0,
      "step": 107
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 340.708345413208,
      "epoch": 1.16,
      "grad_norm": 0.07937922328710556,
      "kl": 0.3579130172729492,
      "learning_rate": 2.308806307967955e-06,
      "loss": 0.0211,
      "reward": 0.9166666679084301,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9166666679084301,
      "rewards/format_reward": 0.0,
      "step": 108
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 403.0208511352539,
      "epoch": 1.1706666666666667,
      "grad_norm": 0.41769587993621826,
      "kl": 0.6602458953857422,
      "learning_rate": 2.2929319883793442e-06,
      "loss": 0.0381,
      "reward": 0.8125000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 109
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 457.3958377838135,
      "epoch": 1.1813333333333333,
      "grad_norm": 0.8391091823577881,
      "kl": 1.37548828125,
      "learning_rate": 2.276933451351696e-06,
      "loss": 0.057,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 481.6041793823242,
      "epoch": 1.192,
      "grad_norm": 1.0701651573181152,
      "kl": 1.6614551544189453,
      "learning_rate": 2.2608132031495184e-06,
      "loss": 0.0747,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 111
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 376.54167556762695,
      "epoch": 1.2026666666666666,
      "grad_norm": 0.09668245911598206,
      "kl": 0.32042503356933594,
      "learning_rate": 2.244573769104084e-06,
      "loss": 0.0267,
      "reward": 0.7916666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 112
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 383.02084732055664,
      "epoch": 1.2133333333333334,
      "grad_norm": 0.4020495116710663,
      "kl": 0.5682315826416016,
      "learning_rate": 2.228217693217826e-06,
      "loss": 0.0309,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 113
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 537.7708549499512,
      "epoch": 1.224,
      "grad_norm": 0.3334648907184601,
      "kl": 2.139446258544922,
      "learning_rate": 2.211747537765808e-06,
      "loss": 0.2109,
      "reward": 0.7083333469927311,
      "reward_std": 0.23116152733564377,
      "rewards/accuracy_reward": 0.7083333469927311,
      "rewards/format_reward": 0.0,
      "step": 114
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 331.33334159851074,
      "epoch": 1.2346666666666666,
      "grad_norm": 0.06051883101463318,
      "kl": 0.015924453735351562,
      "learning_rate": 2.195165882894322e-06,
      "loss": 0.0011,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 115
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.1458473205566,
      "epoch": 1.2453333333333334,
      "grad_norm": 0.6789928674697876,
      "kl": 2.4610023498535156,
      "learning_rate": 2.1784753262166986e-06,
      "loss": 0.1618,
      "reward": 0.541666679084301,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.541666679084301,
      "rewards/format_reward": 0.0,
      "step": 116
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 489.72917556762695,
      "epoch": 1.256,
      "grad_norm": 0.6535123586654663,
      "kl": 0.9884786605834961,
      "learning_rate": 2.161678482406374e-06,
      "loss": 0.0395,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 117
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 625.6250133514404,
      "epoch": 1.2666666666666666,
      "grad_norm": 0.41955748200416565,
      "kl": 1.4302806854248047,
      "learning_rate": 2.144777982787286e-06,
      "loss": 0.1144,
      "reward": 0.6458333432674408,
      "reward_std": 0.1705273911356926,
      "rewards/accuracy_reward": 0.6458333432674408,
      "rewards/format_reward": 0.0,
      "step": 118
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 332.87500762939453,
      "epoch": 1.2773333333333334,
      "grad_norm": 0.7490342855453491,
      "kl": 0.2764606475830078,
      "learning_rate": 2.127776474921661e-06,
      "loss": 0.0512,
      "reward": 0.9166666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 119
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 547.7916831970215,
      "epoch": 1.288,
      "grad_norm": 0.06379815191030502,
      "kl": 0.8844208717346191,
      "learning_rate": 2.1106766221952607e-06,
      "loss": 0.072,
      "reward": 0.5625000018626451,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.5625000018626451,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 388.1875171661377,
      "epoch": 1.2986666666666666,
      "grad_norm": 0.8159778714179993,
      "kl": 0.38848018646240234,
      "learning_rate": 2.0934811034001427e-06,
      "loss": 0.0851,
      "reward": 0.8958333395421505,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.8958333395421505,
      "rewards/format_reward": 0.0,
      "step": 121
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 481.2500171661377,
      "epoch": 1.3093333333333335,
      "grad_norm": 0.12465062737464905,
      "kl": 0.829371452331543,
      "learning_rate": 2.0761926123150166e-06,
      "loss": 0.0618,
      "reward": 0.7083333432674408,
      "reward_std": 0.10206207633018494,
      "rewards/accuracy_reward": 0.7083333432674408,
      "rewards/format_reward": 0.0,
      "step": 122
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 513.5000152587891,
      "epoch": 1.32,
      "grad_norm": 0.6524884104728699,
      "kl": 0.4481773376464844,
      "learning_rate": 2.058813857283244e-06,
      "loss": 0.083,
      "reward": 0.7708333376795053,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.7708333376795053,
      "rewards/format_reward": 0.0,
      "step": 123
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 420.89584732055664,
      "epoch": 1.3306666666666667,
      "grad_norm": 0.379934161901474,
      "kl": 0.38955020904541016,
      "learning_rate": 2.041347560788562e-06,
      "loss": 0.0625,
      "reward": 0.7708333358168602,
      "reward_std": 0.19756478071212769,
      "rewards/accuracy_reward": 0.7708333358168602,
      "rewards/format_reward": 0.0,
      "step": 124
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 457.2083435058594,
      "epoch": 1.3413333333333333,
      "grad_norm": 1.2472268342971802,
      "kl": 0.42907142639160156,
      "learning_rate": 2.0237964590285924e-06,
      "loss": 0.081,
      "reward": 0.8333333432674408,
      "reward_std": 0.24859581515192986,
      "rewards/accuracy_reward": 0.8333333432674408,
      "rewards/format_reward": 0.0,
      "step": 125
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 323.04167556762695,
      "epoch": 1.3519999999999999,
      "grad_norm": 0.0032101995311677456,
      "kl": 0.0042514801025390625,
      "learning_rate": 2.0061633014861975e-06,
      "loss": 0.0002,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 126
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 407.9166774749756,
      "epoch": 1.3626666666666667,
      "grad_norm": 0.02511999011039734,
      "kl": 0.005228996276855469,
      "learning_rate": 1.988450850498758e-06,
      "loss": 0.0025,
      "reward": 0.770833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.770833333954215,
      "rewards/format_reward": 0.0,
      "step": 127
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 428.04167556762695,
      "epoch": 1.3733333333333333,
      "grad_norm": 1.0854262113571167,
      "kl": 0.6306896209716797,
      "learning_rate": 1.970661880825437e-06,
      "loss": 0.1133,
      "reward": 0.5625000149011612,
      "reward_std": 0.1530931033194065,
      "rewards/accuracy_reward": 0.5625000149011612,
      "rewards/format_reward": 0.0,
      "step": 128
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 418.27084159851074,
      "epoch": 1.384,
      "grad_norm": 0.5975901484489441,
      "kl": 0.6131801605224609,
      "learning_rate": 1.952799179212498e-06,
      "loss": 0.0908,
      "reward": 0.9166666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 129
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 289.5833377838135,
      "epoch": 1.3946666666666667,
      "grad_norm": 0.005492673255503178,
      "kl": 0.007350921630859375,
      "learning_rate": 1.934865543956745e-06,
      "loss": 0.0003,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 386.58333587646484,
      "epoch": 1.4053333333333333,
      "grad_norm": 0.7348778247833252,
      "kl": 1.185323715209961,
      "learning_rate": 1.916863784467152e-06,
      "loss": 0.0511,
      "reward": 0.8333333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.8333333432674408,
      "rewards/format_reward": 0.0,
      "step": 131
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 492.58334159851074,
      "epoch": 1.416,
      "grad_norm": 1.0415639877319336,
      "kl": 2.898162841796875,
      "learning_rate": 1.8987967208247542e-06,
      "loss": 0.14,
      "reward": 0.7916666679084301,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.7916666679084301,
      "rewards/format_reward": 0.0,
      "step": 132
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 326.75000953674316,
      "epoch": 1.4266666666666667,
      "grad_norm": 0.03604130446910858,
      "kl": 0.004380226135253906,
      "learning_rate": 1.8806671833408633e-06,
      "loss": 0.0115,
      "reward": 0.9166666679084301,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9166666679084301,
      "rewards/format_reward": 0.0,
      "step": 133
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 377.62501335144043,
      "epoch": 1.4373333333333334,
      "grad_norm": 0.4862348139286041,
      "kl": 0.6692104339599609,
      "learning_rate": 1.8624780121136837e-06,
      "loss": 0.0363,
      "reward": 0.8333333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.8333333358168602,
      "rewards/format_reward": 0.0,
      "step": 134
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 422.0000114440918,
      "epoch": 1.448,
      "grad_norm": 0.44507503509521484,
      "kl": 0.6348180770874023,
      "learning_rate": 1.8442320565833972e-06,
      "loss": 0.0274,
      "reward": 0.7916666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 135
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 451.2500114440918,
      "epoch": 1.4586666666666668,
      "grad_norm": 1.021637201309204,
      "kl": 2.4706764221191406,
      "learning_rate": 1.8259321750857772e-06,
      "loss": 0.1354,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 136
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 399.66667556762695,
      "epoch": 1.4693333333333334,
      "grad_norm": 0.9214147925376892,
      "kl": 1.3475456237792969,
      "learning_rate": 1.8075812344044163e-06,
      "loss": 0.0669,
      "reward": 0.8333333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.8333333358168602,
      "rewards/format_reward": 0.0,
      "step": 137
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 329.75000381469727,
      "epoch": 1.48,
      "grad_norm": 0.5980573892593384,
      "kl": 0.4876108169555664,
      "learning_rate": 1.7891821093216275e-06,
      "loss": 0.0794,
      "reward": 0.8541666772216558,
      "reward_std": 0.1530931070446968,
      "rewards/accuracy_reward": 0.8541666772216558,
      "rewards/format_reward": 0.0,
      "step": 138
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 505.9791793823242,
      "epoch": 1.4906666666666666,
      "grad_norm": 0.9238595366477966,
      "kl": 2.1415843963623047,
      "learning_rate": 1.770737682168091e-06,
      "loss": 0.1121,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 139
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 273.5416774749756,
      "epoch": 1.5013333333333332,
      "grad_norm": 0.05083390697836876,
      "kl": 0.00457763671875,
      "learning_rate": 1.7522508423713243e-06,
      "loss": -0.0013,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 476.83334732055664,
      "epoch": 1.512,
      "grad_norm": 0.4643917381763458,
      "kl": 0.5749650001525879,
      "learning_rate": 1.7337244860030312e-06,
      "loss": 0.0285,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 141
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 416.8333435058594,
      "epoch": 1.5226666666666666,
      "grad_norm": 0.31729748845100403,
      "kl": 1.1347312927246094,
      "learning_rate": 1.7151615153254192e-06,
      "loss": 0.0634,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 142
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 341.6458435058594,
      "epoch": 1.5333333333333332,
      "grad_norm": 0.3522985875606537,
      "kl": 0.23196029663085938,
      "learning_rate": 1.6965648383365416e-06,
      "loss": 0.0507,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 143
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 382.3541793823242,
      "epoch": 1.544,
      "grad_norm": 0.3495376408100128,
      "kl": 0.11490726470947266,
      "learning_rate": 1.6779373683147423e-06,
      "loss": 0.0229,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 144
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 388.39584732055664,
      "epoch": 1.5546666666666666,
      "grad_norm": 0.15049666166305542,
      "kl": 0.20890235900878906,
      "learning_rate": 1.6592820233622743e-06,
      "loss": 0.0084,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 335.06250762939453,
      "epoch": 1.5653333333333332,
      "grad_norm": 0.5136704444885254,
      "kl": 0.2004833221435547,
      "learning_rate": 1.6406017259481624e-06,
      "loss": 0.0476,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 146
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 366.9583435058594,
      "epoch": 1.576,
      "grad_norm": 0.0030216865707188845,
      "kl": 0.00506901741027832,
      "learning_rate": 1.621899402450379e-06,
      "loss": 0.0002,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 147
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 452.7083339691162,
      "epoch": 1.5866666666666667,
      "grad_norm": 0.14610090851783752,
      "kl": 0.6161727905273438,
      "learning_rate": 1.6031779826974138e-06,
      "loss": 0.0747,
      "reward": 0.8125,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.0,
      "step": 148
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 442.9791774749756,
      "epoch": 1.5973333333333333,
      "grad_norm": 0.2566995918750763,
      "kl": 0.46244239807128906,
      "learning_rate": 1.5844403995092944e-06,
      "loss": 0.0504,
      "reward": 0.8125000074505806,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 149
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 466.14584732055664,
      "epoch": 1.608,
      "grad_norm": 0.06670026481151581,
      "kl": 0.7022981643676758,
      "learning_rate": 1.5656895882381465e-06,
      "loss": 0.0494,
      "reward": 0.6666666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.6666666716337204,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 482.3541717529297,
      "epoch": 1.6186666666666667,
      "grad_norm": 0.3101685643196106,
      "kl": 0.8194246292114258,
      "learning_rate": 1.5469284863083523e-06,
      "loss": 0.0407,
      "reward": 0.7708333395421505,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.7708333395421505,
      "rewards/format_reward": 0.0,
      "step": 151
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 385.39584159851074,
      "epoch": 1.6293333333333333,
      "grad_norm": 0.2913825511932373,
      "kl": 0.36629152297973633,
      "learning_rate": 1.5281600327563853e-06,
      "loss": 0.0093,
      "reward": 0.8125,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.0,
      "step": 152
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 512.3333473205566,
      "epoch": 1.6400000000000001,
      "grad_norm": 0.46484774351119995,
      "kl": 0.29232025146484375,
      "learning_rate": 1.5093871677703947e-06,
      "loss": 0.0114,
      "reward": 0.8541666716337204,
      "reward_std": 0.19756478071212769,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 153
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.0000171661377,
      "epoch": 1.6506666666666665,
      "grad_norm": 0.28327611088752747,
      "kl": 1.0828266143798828,
      "learning_rate": 1.4906128322296056e-06,
      "loss": 0.0889,
      "reward": 0.562500013038516,
      "reward_std": 0.21764283999800682,
      "rewards/accuracy_reward": 0.562500013038516,
      "rewards/format_reward": 0.0,
      "step": 154
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 475.395845413208,
      "epoch": 1.6613333333333333,
      "grad_norm": 0.740557849407196,
      "kl": 0.5760955810546875,
      "learning_rate": 1.471839967243615e-06,
      "loss": 0.0731,
      "reward": 0.6250000149011612,
      "reward_std": 0.23116153106093407,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.0,
      "step": 155
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 688.854175567627,
      "epoch": 1.6720000000000002,
      "grad_norm": 0.5783205628395081,
      "kl": 1.8222312927246094,
      "learning_rate": 1.453071513691648e-06,
      "loss": 0.1042,
      "reward": 0.45833333395421505,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.45833333395421505,
      "rewards/format_reward": 0.0,
      "step": 156
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 435.1666831970215,
      "epoch": 1.6826666666666665,
      "grad_norm": 0.4338362216949463,
      "kl": 0.7707328796386719,
      "learning_rate": 1.4343104117618536e-06,
      "loss": 0.0664,
      "reward": 0.7291666772216558,
      "reward_std": 0.1530931070446968,
      "rewards/accuracy_reward": 0.7291666772216558,
      "rewards/format_reward": 0.0,
      "step": 157
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 404.00000953674316,
      "epoch": 1.6933333333333334,
      "grad_norm": 0.9795002937316895,
      "kl": 0.4046802520751953,
      "learning_rate": 1.415559600490706e-06,
      "loss": 0.07,
      "reward": 0.8958333358168602,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.8958333358168602,
      "rewards/format_reward": 0.0,
      "step": 158
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 396.31250762939453,
      "epoch": 1.704,
      "grad_norm": 0.49897342920303345,
      "kl": 0.7032661437988281,
      "learning_rate": 1.3968220173025865e-06,
      "loss": 0.0281,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 159
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 595.3541870117188,
      "epoch": 1.7146666666666666,
      "grad_norm": 0.21808795630931854,
      "kl": 1.0136184692382812,
      "learning_rate": 1.3781005975496212e-06,
      "loss": 0.0788,
      "reward": 0.6458333395421505,
      "reward_std": 0.1530931033194065,
      "rewards/accuracy_reward": 0.6458333395421505,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 516.604175567627,
      "epoch": 1.7253333333333334,
      "grad_norm": 1.431018590927124,
      "kl": 1.447868824005127,
      "learning_rate": 1.359398274051838e-06,
      "loss": 0.1777,
      "reward": 0.666666692122817,
      "reward_std": 0.3061862178146839,
      "rewards/accuracy_reward": 0.666666692122817,
      "rewards/format_reward": 0.0,
      "step": 161
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 471.5000114440918,
      "epoch": 1.736,
      "grad_norm": 0.04113710671663284,
      "kl": 0.004864692687988281,
      "learning_rate": 1.340717976637726e-06,
      "loss": -0.0195,
      "reward": 0.8541666772216558,
      "reward_std": 0.1530931070446968,
      "rewards/accuracy_reward": 0.8541666772216558,
      "rewards/format_reward": 0.0,
      "step": 162
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 390.00001335144043,
      "epoch": 1.7466666666666666,
      "grad_norm": 1.3767911195755005,
      "kl": 0.3090934753417969,
      "learning_rate": 1.3220626316852584e-06,
      "loss": 0.0901,
      "reward": 0.8958333432674408,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.8958333432674408,
      "rewards/format_reward": 0.0,
      "step": 163
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 521.0000152587891,
      "epoch": 1.7573333333333334,
      "grad_norm": 0.575884997844696,
      "kl": 1.0921440124511719,
      "learning_rate": 1.303435161663459e-06,
      "loss": 0.0502,
      "reward": 0.645833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.645833333954215,
      "rewards/format_reward": 0.0,
      "step": 164
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 381.9791774749756,
      "epoch": 1.768,
      "grad_norm": 0.03651285544037819,
      "kl": 0.5413856506347656,
      "learning_rate": 1.2848384846745813e-06,
      "loss": 0.0345,
      "reward": 0.895833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.895833333954215,
      "rewards/format_reward": 0.0,
      "step": 165
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 553.6666851043701,
      "epoch": 1.7786666666666666,
      "grad_norm": 0.6268121600151062,
      "kl": 0.81475830078125,
      "learning_rate": 1.2662755139969693e-06,
      "loss": 0.0824,
      "reward": 0.6666666753590107,
      "reward_std": 0.2861081585288048,
      "rewards/accuracy_reward": 0.6666666753590107,
      "rewards/format_reward": 0.0,
      "step": 166
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 396.8958511352539,
      "epoch": 1.7893333333333334,
      "grad_norm": 0.4399033486843109,
      "kl": 0.28545475006103516,
      "learning_rate": 1.2477491576286764e-06,
      "loss": 0.0433,
      "reward": 0.8333333432674408,
      "reward_std": 0.10206207633018494,
      "rewards/accuracy_reward": 0.8333333432674408,
      "rewards/format_reward": 0.0,
      "step": 167
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 452.0208435058594,
      "epoch": 1.8,
      "grad_norm": 0.207831010222435,
      "kl": 1.3086605072021484,
      "learning_rate": 1.2292623178319094e-06,
      "loss": 0.1024,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 168
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 431.87500762939453,
      "epoch": 1.8106666666666666,
      "grad_norm": 0.3532598912715912,
      "kl": 0.8703994750976562,
      "learning_rate": 1.2108178906783732e-06,
      "loss": 0.0434,
      "reward": 0.8125,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.0,
      "step": 169
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 354.2083444595337,
      "epoch": 1.8213333333333335,
      "grad_norm": 0.49789828062057495,
      "kl": 0.22895240783691406,
      "learning_rate": 1.1924187655955838e-06,
      "loss": 0.0548,
      "reward": 0.7916666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 452.8958492279053,
      "epoch": 1.8319999999999999,
      "grad_norm": 0.4020465612411499,
      "kl": 1.2119274139404297,
      "learning_rate": 1.1740678249142229e-06,
      "loss": 0.1281,
      "reward": 0.6666666772216558,
      "reward_std": 0.22155842557549477,
      "rewards/accuracy_reward": 0.6666666772216558,
      "rewards/format_reward": 0.0,
      "step": 171
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 377.2291793823242,
      "epoch": 1.8426666666666667,
      "grad_norm": 0.29985761642456055,
      "kl": 0.6539144515991211,
      "learning_rate": 1.1557679434166027e-06,
      "loss": 0.0712,
      "reward": 0.7708333395421505,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.7708333395421505,
      "rewards/format_reward": 0.0,
      "step": 172
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 324.75000762939453,
      "epoch": 1.8533333333333335,
      "grad_norm": 0.23136773705482483,
      "kl": 0.8882865905761719,
      "learning_rate": 1.1375219878863162e-06,
      "loss": 0.0518,
      "reward": 0.895833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.895833333954215,
      "rewards/format_reward": 0.0,
      "step": 173
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 401.5000114440918,
      "epoch": 1.8639999999999999,
      "grad_norm": 0.8711624145507812,
      "kl": 1.3646907806396484,
      "learning_rate": 1.1193328166591372e-06,
      "loss": 0.1095,
      "reward": 0.770833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.770833333954215,
      "rewards/format_reward": 0.0,
      "step": 174
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 431.9166774749756,
      "epoch": 1.8746666666666667,
      "grad_norm": 0.585869550704956,
      "kl": 0.7460689544677734,
      "learning_rate": 1.1012032791752457e-06,
      "loss": 0.0468,
      "reward": 0.6875000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000074505806,
      "rewards/format_reward": 0.0,
      "step": 175
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 448.4166774749756,
      "epoch": 1.8853333333333333,
      "grad_norm": 0.3197134733200073,
      "kl": 1.1435546875,
      "learning_rate": 1.0831362155328477e-06,
      "loss": 0.1067,
      "reward": 0.7500000037252903,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.7500000037252903,
      "rewards/format_reward": 0.0,
      "step": 176
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 477.83334732055664,
      "epoch": 1.896,
      "grad_norm": 0.07629247009754181,
      "kl": 0.042983055114746094,
      "learning_rate": 1.0651344560432548e-06,
      "loss": 0.0104,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 177
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 386.8333492279053,
      "epoch": 1.9066666666666667,
      "grad_norm": 0.1934867650270462,
      "kl": 0.40998268127441406,
      "learning_rate": 1.0472008207875023e-06,
      "loss": 0.0402,
      "reward": 0.833333333954215,
      "reward_std": 0.11949636042118073,
      "rewards/accuracy_reward": 0.833333333954215,
      "rewards/format_reward": 0.0,
      "step": 178
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 411.6041793823242,
      "epoch": 1.9173333333333333,
      "grad_norm": 0.7401374578475952,
      "kl": 0.5113000869750977,
      "learning_rate": 1.0293381191745637e-06,
      "loss": 0.0632,
      "reward": 0.8541666772216558,
      "reward_std": 0.1530931033194065,
      "rewards/accuracy_reward": 0.8541666772216558,
      "rewards/format_reward": 0.0,
      "step": 179
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 509.08334159851074,
      "epoch": 1.928,
      "grad_norm": 0.3369535803794861,
      "kl": 0.47974681854248047,
      "learning_rate": 1.0115491495012424e-06,
      "loss": 0.0261,
      "reward": 0.8125000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 491.93751335144043,
      "epoch": 1.9386666666666668,
      "grad_norm": 1.289297103881836,
      "kl": 1.9831523895263672,
      "learning_rate": 9.938366985138028e-07,
      "loss": 0.0808,
      "reward": 0.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.0,
      "step": 181
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 346.0208435058594,
      "epoch": 1.9493333333333334,
      "grad_norm": 0.11761274188756943,
      "kl": 0.36895179748535156,
      "learning_rate": 9.762035409714077e-07,
      "loss": 0.0403,
      "reward": 0.8958333432674408,
      "reward_std": 0.1705273948609829,
      "rewards/accuracy_reward": 0.8958333432674408,
      "rewards/format_reward": 0.0,
      "step": 182
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 270.33334159851074,
      "epoch": 1.96,
      "grad_norm": 0.044683996587991714,
      "kl": 0.0055255889892578125,
      "learning_rate": 9.586524392114385e-07,
      "loss": -0.0001,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 183
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 360.00000953674316,
      "epoch": 1.9706666666666668,
      "grad_norm": 0.05016176402568817,
      "kl": 0.24141979217529297,
      "learning_rate": 9.411861427167567e-07,
      "loss": 0.0187,
      "reward": 0.895833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.895833333954215,
      "rewards/format_reward": 0.0,
      "step": 184
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 497.2708435058594,
      "epoch": 1.9813333333333332,
      "grad_norm": 0.6449931859970093,
      "kl": 1.508819580078125,
      "learning_rate": 9.238073876849837e-07,
      "loss": 0.1108,
      "reward": 0.6250000037252903,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.6250000037252903,
      "rewards/format_reward": 0.0,
      "step": 185
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 464.22917556762695,
      "epoch": 1.992,
      "grad_norm": 0.14938627183437347,
      "kl": 0.8099231719970703,
      "learning_rate": 9.065188965998574e-07,
      "loss": 0.0968,
      "reward": 0.7708333358168602,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.7708333358168602,
      "rewards/format_reward": 0.0,
      "step": 186
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 446.5416736602783,
      "epoch": 2.010666666666667,
      "grad_norm": 0.1458408087491989,
      "kl": 0.4970722198486328,
      "learning_rate": 8.893233778047394e-07,
      "loss": 0.0488,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 187
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 314.37500953674316,
      "epoch": 2.021333333333333,
      "grad_norm": 0.005513062234967947,
      "kl": 0.00611114501953125,
      "learning_rate": 8.722235250783389e-07,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 188
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 304.8541793823242,
      "epoch": 2.032,
      "grad_norm": 0.005480589345097542,
      "kl": 0.0071258544921875,
      "learning_rate": 8.552220172127148e-07,
      "loss": 0.0003,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 189
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 457.7916774749756,
      "epoch": 2.042666666666667,
      "grad_norm": 0.545896053314209,
      "kl": 0.8441686630249023,
      "learning_rate": 8.383215175936263e-07,
      "loss": 0.0277,
      "reward": 0.7916666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 190
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 393.2708435058594,
      "epoch": 2.0533333333333332,
      "grad_norm": 0.12136304378509521,
      "kl": 0.7232379913330078,
      "learning_rate": 8.215246737833017e-07,
      "loss": 0.049,
      "reward": 0.895833333954215,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.895833333954215,
      "rewards/format_reward": 0.0,
      "step": 191
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 377.54167556762695,
      "epoch": 2.064,
      "grad_norm": 0.19407136738300323,
      "kl": 0.533238410949707,
      "learning_rate": 8.048341171056786e-07,
      "loss": 0.076,
      "reward": 0.8125,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.0,
      "step": 192
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 335.2916679382324,
      "epoch": 2.074666666666667,
      "grad_norm": 0.03280942142009735,
      "kl": 0.007334709167480469,
      "learning_rate": 7.882524622341923e-07,
      "loss": 0.0047,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 193
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 537.7083473205566,
      "epoch": 2.0853333333333333,
      "grad_norm": 0.1604704111814499,
      "kl": 1.1008033752441406,
      "learning_rate": 7.717823067821738e-07,
      "loss": 0.0828,
      "reward": 0.5000000055879354,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.5000000055879354,
      "rewards/format_reward": 0.0,
      "step": 194
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 460.43750762939453,
      "epoch": 2.096,
      "grad_norm": 0.19094224274158478,
      "kl": 0.7224607467651367,
      "learning_rate": 7.554262308959165e-07,
      "loss": 0.064,
      "reward": 0.7083333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.7083333432674408,
      "rewards/format_reward": 0.0,
      "step": 195
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 435.10418701171875,
      "epoch": 2.1066666666666665,
      "grad_norm": 0.3023138642311096,
      "kl": 0.37677001953125,
      "learning_rate": 7.39186796850482e-07,
      "loss": 0.0505,
      "reward": 0.791666679084301,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.791666679084301,
      "rewards/format_reward": 0.0,
      "step": 196
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 484.3333492279053,
      "epoch": 2.1173333333333333,
      "grad_norm": 0.6410330533981323,
      "kl": 1.1864428520202637,
      "learning_rate": 7.23066548648304e-07,
      "loss": 0.1031,
      "reward": 0.7916666697710752,
      "reward_std": 0.18404608964920044,
      "rewards/accuracy_reward": 0.7916666697710752,
      "rewards/format_reward": 0.0,
      "step": 197
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 567.4791870117188,
      "epoch": 2.128,
      "grad_norm": 0.27650487422943115,
      "kl": 0.9266481399536133,
      "learning_rate": 7.070680116206564e-07,
      "loss": 0.0619,
      "reward": 0.7500000055879354,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.7500000055879354,
      "rewards/format_reward": 0.0,
      "step": 198
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 388.14584732055664,
      "epoch": 2.1386666666666665,
      "grad_norm": 0.05082610622048378,
      "kl": 0.008310317993164062,
      "learning_rate": 6.911936920320457e-07,
      "loss": -0.0025,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 199
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 363.8125114440918,
      "epoch": 2.1493333333333333,
      "grad_norm": 0.48688116669654846,
      "kl": 0.13292503356933594,
      "learning_rate": 6.754460766875871e-07,
      "loss": 0.0419,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 200
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 316.06250381469727,
      "epoch": 2.16,
      "grad_norm": 0.02585996314883232,
      "kl": 0.0049285888671875,
      "learning_rate": 6.598276325434357e-07,
      "loss": 0.0002,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 201
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 381.1875114440918,
      "epoch": 2.1706666666666665,
      "grad_norm": 0.4411306083202362,
      "kl": 0.1497488021850586,
      "learning_rate": 6.443408063203226e-07,
      "loss": 0.0297,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 202
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 510.8125114440918,
      "epoch": 2.1813333333333333,
      "grad_norm": 0.290724515914917,
      "kl": 0.42789649963378906,
      "learning_rate": 6.289880241202581e-07,
      "loss": 0.0153,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 203
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 362.4791774749756,
      "epoch": 2.192,
      "grad_norm": 0.003817596472799778,
      "kl": 0.005370616912841797,
      "learning_rate": 6.137716910464735e-07,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 204
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 317.6875114440918,
      "epoch": 2.2026666666666666,
      "grad_norm": 0.002196200890466571,
      "kl": 0.0045452117919921875,
      "learning_rate": 5.986941908266454e-07,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 205
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 431.1666831970215,
      "epoch": 2.2133333333333334,
      "grad_norm": 0.6445562839508057,
      "kl": 0.5915031433105469,
      "learning_rate": 5.837578854394692e-07,
      "loss": 0.0613,
      "reward": 0.8333333414047956,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.8333333414047956,
      "rewards/format_reward": 0.0,
      "step": 206
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 339.18750762939453,
      "epoch": 2.224,
      "grad_norm": 0.5580638647079468,
      "kl": 0.17171144485473633,
      "learning_rate": 5.689651147446448e-07,
      "loss": 0.0382,
      "reward": 0.9583333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.9583333432674408,
      "rewards/format_reward": 0.0,
      "step": 207
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 460.56250762939453,
      "epoch": 2.2346666666666666,
      "grad_norm": 0.6676971316337585,
      "kl": 1.0391263961791992,
      "learning_rate": 5.543181961163219e-07,
      "loss": 0.0394,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 208
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 361.5416774749756,
      "epoch": 2.2453333333333334,
      "grad_norm": 0.38774293661117554,
      "kl": 0.4868354797363281,
      "learning_rate": 5.398194240800684e-07,
      "loss": 0.1008,
      "reward": 0.9375,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.0,
      "step": 209
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 437.0000057220459,
      "epoch": 2.2560000000000002,
      "grad_norm": 0.33576539158821106,
      "kl": 0.5966892242431641,
      "learning_rate": 5.254710699534226e-07,
      "loss": 0.058,
      "reward": 0.7916666716337204,
      "reward_std": 0.11949636042118073,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 210
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 508.60416984558105,
      "epoch": 2.2666666666666666,
      "grad_norm": 0.7082143425941467,
      "kl": 1.8151893615722656,
      "learning_rate": 5.112753814900761e-07,
      "loss": 0.0719,
      "reward": 0.6250000037252903,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.6250000037252903,
      "rewards/format_reward": 0.0,
      "step": 211
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 388.4375114440918,
      "epoch": 2.2773333333333334,
      "grad_norm": 0.28254204988479614,
      "kl": 0.4129981994628906,
      "learning_rate": 4.97234582527751e-07,
      "loss": 0.0403,
      "reward": 0.9166666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 212
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 411.25001525878906,
      "epoch": 2.288,
      "grad_norm": 0.08238724619150162,
      "kl": 0.5282511711120605,
      "learning_rate": 4.8335087263982e-07,
      "loss": 0.008,
      "reward": 0.7916666679084301,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.7916666679084301,
      "rewards/format_reward": 0.0,
      "step": 213
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 454.0208511352539,
      "epoch": 2.2986666666666666,
      "grad_norm": 0.18948504328727722,
      "kl": 0.6926717758178711,
      "learning_rate": 4.6962642679073465e-07,
      "loss": 0.0757,
      "reward": 0.9375,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.0,
      "step": 214
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 503.2916851043701,
      "epoch": 2.3093333333333335,
      "grad_norm": 1.367248296737671,
      "kl": 1.4507637023925781,
      "learning_rate": 4.560633949953017e-07,
      "loss": 0.1942,
      "reward": 0.8333333469927311,
      "reward_std": 0.23116153106093407,
      "rewards/accuracy_reward": 0.8333333469927311,
      "rewards/format_reward": 0.0,
      "step": 215
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 318.1250057220459,
      "epoch": 2.32,
      "grad_norm": 0.10031232982873917,
      "kl": 0.13898754119873047,
      "learning_rate": 4.426639019818704e-07,
      "loss": 0.0057,
      "reward": 0.9166666716337204,
      "reward_std": 0.11949636042118073,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 216
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 404.50000762939453,
      "epoch": 2.3306666666666667,
      "grad_norm": 0.1769079864025116,
      "kl": 0.1834850311279297,
      "learning_rate": 4.2943004685948286e-07,
      "loss": 0.0405,
      "reward": 0.7708333432674408,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.7708333432674408,
      "rewards/format_reward": 0.0,
      "step": 217
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 474.0416831970215,
      "epoch": 2.3413333333333335,
      "grad_norm": 0.1727655977010727,
      "kl": 1.332137107849121,
      "learning_rate": 4.163639027890363e-07,
      "loss": 0.1002,
      "reward": 0.7291666753590107,
      "reward_std": 0.1801304928958416,
      "rewards/accuracy_reward": 0.7291666753590107,
      "rewards/format_reward": 0.0,
      "step": 218
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 409.56250953674316,
      "epoch": 2.352,
      "grad_norm": 0.43854233622550964,
      "kl": 0.5045089721679688,
      "learning_rate": 4.034675166585079e-07,
      "loss": 0.0126,
      "reward": 0.7291666772216558,
      "reward_std": 0.1530931070446968,
      "rewards/accuracy_reward": 0.7291666772216558,
      "rewards/format_reward": 0.0,
      "step": 219
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 460.62500953674316,
      "epoch": 2.3626666666666667,
      "grad_norm": 0.16729138791561127,
      "kl": 1.1355094909667969,
      "learning_rate": 3.907429087623007e-07,
      "loss": 0.1109,
      "reward": 0.5833333414047956,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.5833333414047956,
      "rewards/format_reward": 0.0,
      "step": 220
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 334.5208435058594,
      "epoch": 2.3733333333333335,
      "grad_norm": 0.18300044536590576,
      "kl": 0.4584054946899414,
      "learning_rate": 3.7819207248475174e-07,
      "loss": 0.0409,
      "reward": 0.9375,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.0,
      "step": 221
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 372.2916717529297,
      "epoch": 2.384,
      "grad_norm": 0.019170992076396942,
      "kl": 0.0048065185546875,
      "learning_rate": 3.6581697398785615e-07,
      "loss": 0.0016,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 222
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 314.7708396911621,
      "epoch": 2.3946666666666667,
      "grad_norm": 0.3603329062461853,
      "kl": 0.2915935516357422,
      "learning_rate": 3.5361955190325563e-07,
      "loss": 0.04,
      "reward": 0.9375000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.9375000074505806,
      "rewards/format_reward": 0.0,
      "step": 223
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 330.10417556762695,
      "epoch": 2.405333333333333,
      "grad_norm": 0.02583317644894123,
      "kl": 0.007556915283203125,
      "learning_rate": 3.4160171702854286e-07,
      "loss": -0.0041,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 224
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 356.208345413208,
      "epoch": 2.416,
      "grad_norm": 0.4580487608909607,
      "kl": 0.6083364486694336,
      "learning_rate": 3.297653520279226e-07,
      "loss": 0.028,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 225
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 319.85416984558105,
      "epoch": 2.4266666666666667,
      "grad_norm": 0.003119911067187786,
      "kl": 0.0054950714111328125,
      "learning_rate": 3.1811231113727974e-07,
      "loss": 0.0002,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 226
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 331.4166736602783,
      "epoch": 2.437333333333333,
      "grad_norm": 0.14595112204551697,
      "kl": 0.3737983703613281,
      "learning_rate": 3.066444198737053e-07,
      "loss": 0.0356,
      "reward": 0.8958333358168602,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.8958333358168602,
      "rewards/format_reward": 0.0,
      "step": 227
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 371.7708435058594,
      "epoch": 2.448,
      "grad_norm": 0.3323851227760315,
      "kl": 0.1330718994140625,
      "learning_rate": 2.9536347474951645e-07,
      "loss": 0.0337,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 228
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 497.87501525878906,
      "epoch": 2.458666666666667,
      "grad_norm": 0.6714590191841125,
      "kl": 0.9981689453125,
      "learning_rate": 2.842712429908237e-07,
      "loss": 0.0643,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 229
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 513.4583377838135,
      "epoch": 2.469333333333333,
      "grad_norm": 0.55454421043396,
      "kl": 0.8866720199584961,
      "learning_rate": 2.733694622606824e-07,
      "loss": 0.0355,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 230
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 370.70834159851074,
      "epoch": 2.48,
      "grad_norm": 0.4770406186580658,
      "kl": 0.3101310729980469,
      "learning_rate": 2.626598403868801e-07,
      "loss": 0.0699,
      "reward": 0.8958333395421505,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8958333395421505,
      "rewards/format_reward": 0.0,
      "step": 231
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 318.6666717529297,
      "epoch": 2.490666666666667,
      "grad_norm": 0.5790646076202393,
      "kl": 0.1632075309753418,
      "learning_rate": 2.521440550943943e-07,
      "loss": 0.0443,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 232
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 346.87500953674316,
      "epoch": 2.501333333333333,
      "grad_norm": 0.004121846053749323,
      "kl": 0.0057392120361328125,
      "learning_rate": 2.418237537425666e-07,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 233
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 322.7916736602783,
      "epoch": 2.512,
      "grad_norm": 0.0034579967614263296,
      "kl": 0.0049915313720703125,
      "learning_rate": 2.3170055306703512e-07,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 234
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 489.39584732055664,
      "epoch": 2.522666666666667,
      "grad_norm": 0.17027877271175385,
      "kl": 0.9452896118164062,
      "learning_rate": 2.2177603892646277e-07,
      "loss": 0.0827,
      "reward": 0.8125000037252903,
      "reward_std": 0.19756478071212769,
      "rewards/accuracy_reward": 0.8125000037252903,
      "rewards/format_reward": 0.0,
      "step": 235
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 363.7916717529297,
      "epoch": 2.533333333333333,
      "grad_norm": 0.5640117526054382,
      "kl": 0.5935440063476562,
      "learning_rate": 2.1205176605410259e-07,
      "loss": 0.0887,
      "reward": 0.9375000074505806,
      "reward_std": 0.11558076739311218,
      "rewards/accuracy_reward": 0.9375000074505806,
      "rewards/format_reward": 0.0,
      "step": 236
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 416.31251335144043,
      "epoch": 2.544,
      "grad_norm": 0.24723418056964874,
      "kl": 0.49085044860839844,
      "learning_rate": 2.0252925781424153e-07,
      "loss": 0.0418,
      "reward": 0.8541666716337204,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.8541666716337204,
      "rewards/format_reward": 0.0,
      "step": 237
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 424.7500114440918,
      "epoch": 2.554666666666667,
      "grad_norm": 0.24722225964069366,
      "kl": 0.3246746063232422,
      "learning_rate": 1.932100059635551e-07,
      "loss": 0.0162,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 238
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 480.9583435058594,
      "epoch": 2.5653333333333332,
      "grad_norm": 0.18559230864048004,
      "kl": 1.0236377716064453,
      "learning_rate": 1.8409547041741436e-07,
      "loss": 0.1079,
      "reward": 0.6458333432674408,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.6458333432674408,
      "rewards/format_reward": 0.0,
      "step": 239
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 452.0833435058594,
      "epoch": 2.576,
      "grad_norm": 0.8793236017227173,
      "kl": 1.7606868743896484,
      "learning_rate": 1.751870790211838e-07,
      "loss": 0.0704,
      "reward": 0.7708333395421505,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.7708333395421505,
      "rewards/format_reward": 0.0,
      "step": 240
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 479.8125190734863,
      "epoch": 2.586666666666667,
      "grad_norm": 0.611665666103363,
      "kl": 0.9804134368896484,
      "learning_rate": 1.6648622732653946e-07,
      "loss": 0.0375,
      "reward": 0.7291666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.7291666716337204,
      "rewards/format_reward": 0.0,
      "step": 241
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 406.4375057220459,
      "epoch": 2.5973333333333333,
      "grad_norm": 0.6699697971343994,
      "kl": 1.0984725952148438,
      "learning_rate": 1.5799427837284686e-07,
      "loss": 0.0453,
      "reward": 0.770833333954215,
      "reward_std": 0.05103103443980217,
      "rewards/accuracy_reward": 0.770833333954215,
      "rewards/format_reward": 0.0,
      "step": 242
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 313.6250114440918,
      "epoch": 2.608,
      "grad_norm": 0.0033394964411854744,
      "kl": 0.0048961639404296875,
      "learning_rate": 1.4971256247363534e-07,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 243
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 367.7083377838135,
      "epoch": 2.618666666666667,
      "grad_norm": 0.2537068724632263,
      "kl": 0.1256580352783203,
      "learning_rate": 1.416423770081941e-07,
      "loss": 0.0114,
      "reward": 0.8750000055879354,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.8750000055879354,
      "rewards/format_reward": 0.0,
      "step": 244
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 477.12501335144043,
      "epoch": 2.6293333333333333,
      "grad_norm": 0.13006748259067535,
      "kl": 1.0239896774291992,
      "learning_rate": 1.3378498621833236e-07,
      "loss": 0.1092,
      "reward": 0.8125000018626451,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8125000018626451,
      "rewards/format_reward": 0.0,
      "step": 245
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 442.1041774749756,
      "epoch": 2.64,
      "grad_norm": 0.17547762393951416,
      "kl": 0.6624622344970703,
      "learning_rate": 1.2614162101032601e-07,
      "loss": 0.0534,
      "reward": 0.8333333414047956,
      "reward_std": 0.16661180183291435,
      "rewards/accuracy_reward": 0.8333333414047956,
      "rewards/format_reward": 0.0,
      "step": 246
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 318.12500762939453,
      "epoch": 2.6506666666666665,
      "grad_norm": 0.0035871269647032022,
      "kl": 0.005970954895019531,
      "learning_rate": 1.1871347876209121e-07,
      "loss": 0.0002,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 1.0,
      "rewards/format_reward": 0.0,
      "step": 247
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 492.2500114440918,
      "epoch": 2.6613333333333333,
      "grad_norm": 0.5926199555397034,
      "kl": 1.2046470642089844,
      "learning_rate": 1.1150172313560741e-07,
      "loss": 0.149,
      "reward": 0.7708333376795053,
      "reward_std": 0.1801304966211319,
      "rewards/accuracy_reward": 0.7708333376795053,
      "rewards/format_reward": 0.0,
      "step": 248
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 458.6250190734863,
      "epoch": 2.672,
      "grad_norm": 0.17869782447814941,
      "kl": 0.8741664886474609,
      "learning_rate": 1.0450748389462184e-07,
      "loss": 0.0953,
      "reward": 0.6875000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.6875000074505806,
      "rewards/format_reward": 0.0,
      "step": 249
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 473.43751335144043,
      "epoch": 2.6826666666666665,
      "grad_norm": 0.2680971026420593,
      "kl": 1.3310651779174805,
      "learning_rate": 9.773185672766705e-08,
      "loss": 0.0763,
      "reward": 0.8125,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.0,
      "step": 250
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 662.4166793823242,
      "epoch": 2.6933333333333334,
      "grad_norm": 0.7904369235038757,
      "kl": 1.6705427169799805,
      "learning_rate": 9.117590307641361e-08,
      "loss": 0.0995,
      "reward": 0.5833333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.5833333358168602,
      "rewards/format_reward": 0.0,
      "step": 251
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 412.395845413208,
      "epoch": 2.7039999999999997,
      "grad_norm": 0.226091206073761,
      "kl": 0.5224666595458984,
      "learning_rate": 8.484064996939023e-08,
      "loss": 0.0417,
      "reward": 0.8125000055879354,
      "reward_std": 0.1705273911356926,
      "rewards/accuracy_reward": 0.8125000055879354,
      "rewards/format_reward": 0.0,
      "step": 252
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 430.7291717529297,
      "epoch": 2.7146666666666666,
      "grad_norm": 0.31987854838371277,
      "kl": 1.5644493103027344,
      "learning_rate": 7.872708986109183e-08,
      "loss": 0.098,
      "reward": 0.7083333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.7083333432674408,
      "rewards/format_reward": 0.0,
      "step": 253
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 453.6666831970215,
      "epoch": 2.7253333333333334,
      "grad_norm": 0.8331900835037231,
      "kl": 0.355438232421875,
      "learning_rate": 7.283618047650808e-08,
      "loss": 0.0767,
      "reward": 0.8125000149011612,
      "reward_std": 0.1530931070446968,
      "rewards/accuracy_reward": 0.8125000149011612,
      "rewards/format_reward": 0.0,
      "step": 254
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 331.43750762939453,
      "epoch": 2.7359999999999998,
      "grad_norm": 0.05422873795032501,
      "kl": 0.006491661071777344,
      "learning_rate": 6.716884466108869e-08,
      "loss": -0.0096,
      "reward": 0.8750000055879354,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.8750000055879354,
      "rewards/format_reward": 0.0,
      "step": 255
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 438.8125114440918,
      "epoch": 2.7466666666666666,
      "grad_norm": 0.29196617007255554,
      "kl": 0.3569297790527344,
      "learning_rate": 6.17259702361741e-08,
      "loss": 0.065,
      "reward": 0.9166666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 256
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 483.833345413208,
      "epoch": 2.7573333333333334,
      "grad_norm": 0.6647077202796936,
      "kl": 0.5238819122314453,
      "learning_rate": 5.65084098599144e-08,
      "loss": 0.0487,
      "reward": 0.916666679084301,
      "reward_std": 0.16661179810762405,
      "rewards/accuracy_reward": 0.916666679084301,
      "rewards/format_reward": 0.0,
      "step": 257
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 415.60417556762695,
      "epoch": 2.768,
      "grad_norm": 0.5700039267539978,
      "kl": 0.6588878631591797,
      "learning_rate": 5.151698089369433e-08,
      "loss": 0.0699,
      "reward": 0.9166666716337204,
      "reward_std": 0.11949635669589043,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 258
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 368.1458435058594,
      "epoch": 2.7786666666666666,
      "grad_norm": 0.08835466206073761,
      "kl": 0.25104331970214844,
      "learning_rate": 4.675246527408872e-08,
      "loss": 0.0241,
      "reward": 0.9166666679084301,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9166666679084301,
      "rewards/format_reward": 0.0,
      "step": 259
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 425.5000057220459,
      "epoch": 2.7893333333333334,
      "grad_norm": 0.4357761740684509,
      "kl": 1.2623291015625,
      "learning_rate": 4.221560939036828e-08,
      "loss": 0.0757,
      "reward": 0.7916666716337204,
      "reward_std": 0.11949636042118073,
      "rewards/accuracy_reward": 0.7916666716337204,
      "rewards/format_reward": 0.0,
      "step": 260
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 544.2708473205566,
      "epoch": 2.8,
      "grad_norm": 0.46760129928588867,
      "kl": 1.0151500701904297,
      "learning_rate": 3.790712396757317e-08,
      "loss": 0.0416,
      "reward": 0.8333333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.8333333358168602,
      "rewards/format_reward": 0.0,
      "step": 261
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 368.8125114440918,
      "epoch": 2.8106666666666666,
      "grad_norm": 0.78629070520401,
      "kl": 0.3953113555908203,
      "learning_rate": 3.382768395517355e-08,
      "loss": 0.0723,
      "reward": 0.8333333432674408,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.8333333432674408,
      "rewards/format_reward": 0.0,
      "step": 262
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 464.1041793823242,
      "epoch": 2.8213333333333335,
      "grad_norm": 0.040255844593048096,
      "kl": 0.7203226089477539,
      "learning_rate": 2.997792842133445e-08,
      "loss": 0.063,
      "reward": 0.8541666679084301,
      "reward_std": 0.13301505148410797,
      "rewards/accuracy_reward": 0.8541666679084301,
      "rewards/format_reward": 0.0,
      "step": 263
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 322.7500057220459,
      "epoch": 2.832,
      "grad_norm": 0.03899736702442169,
      "kl": 0.004932403564453125,
      "learning_rate": 2.6358460452802768e-08,
      "loss": -0.0013,
      "reward": 0.9375,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.9375,
      "rewards/format_reward": 0.0,
      "step": 264
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 364.3125114440918,
      "epoch": 2.8426666666666667,
      "grad_norm": 0.32864269614219666,
      "kl": 0.29061126708984375,
      "learning_rate": 2.296984706042993e-08,
      "loss": 0.0524,
      "reward": 0.8958333395421505,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8958333395421505,
      "rewards/format_reward": 0.0,
      "step": 265
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 338.2083377838135,
      "epoch": 2.8533333333333335,
      "grad_norm": 0.03260638937354088,
      "kl": 0.00400543212890625,
      "learning_rate": 1.9812619090345984e-08,
      "loss": -0.0003,
      "reward": 0.9583333358168602,
      "reward_std": 0.06454972922801971,
      "rewards/accuracy_reward": 0.9583333358168602,
      "rewards/format_reward": 0.0,
      "step": 266
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.6666870117188,
      "epoch": 2.864,
      "grad_norm": 0.4221813976764679,
      "kl": 2.3057403564453125,
      "learning_rate": 1.688727114079963e-08,
      "loss": 0.1908,
      "reward": 0.7291666772216558,
      "reward_std": 0.290023747831583,
      "rewards/accuracy_reward": 0.7291666772216558,
      "rewards/format_reward": 0.0,
      "step": 267
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 440.7708511352539,
      "epoch": 2.8746666666666667,
      "grad_norm": 0.2516721189022064,
      "kl": 1.0313682556152344,
      "learning_rate": 1.4194261484677007e-08,
      "loss": 0.0911,
      "reward": 0.8750000055879354,
      "reward_std": 0.10206206887960434,
      "rewards/accuracy_reward": 0.8750000055879354,
      "rewards/format_reward": 0.0,
      "step": 268
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 297.8958435058594,
      "epoch": 2.8853333333333335,
      "grad_norm": 0.19661647081375122,
      "kl": 0.2102203369140625,
      "learning_rate": 1.1734011997709626e-08,
      "loss": 0.0125,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 269
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 423.9791793823242,
      "epoch": 2.896,
      "grad_norm": 0.604397714138031,
      "kl": 0.8948936462402344,
      "learning_rate": 9.50690809238558e-09,
      "loss": 0.0391,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 270
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 411.77084159851074,
      "epoch": 2.9066666666666667,
      "grad_norm": 0.29994437098503113,
      "kl": 0.6658239364624023,
      "learning_rate": 7.513298657572831e-09,
      "loss": 0.0989,
      "reward": 0.875,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 271
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 357.5833396911621,
      "epoch": 2.9173333333333336,
      "grad_norm": 0.49588629603385925,
      "kl": 0.7907133102416992,
      "learning_rate": 5.753496003862935e-09,
      "loss": 0.0417,
      "reward": 0.8125000074505806,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.8125000074505806,
      "rewards/format_reward": 0.0,
      "step": 272
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 495.18750953674316,
      "epoch": 2.928,
      "grad_norm": 0.4009920060634613,
      "kl": 0.8979015350341797,
      "learning_rate": 4.227775814646173e-09,
      "loss": 0.0522,
      "reward": 0.5625,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 273
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 372.1666774749756,
      "epoch": 2.9386666666666668,
      "grad_norm": 0.31020382046699524,
      "kl": 0.2064361572265625,
      "learning_rate": 2.9363771029242705e-09,
      "loss": 0.0285,
      "reward": 0.9791666716337204,
      "reward_std": 0.05103103816509247,
      "rewards/accuracy_reward": 0.9791666716337204,
      "rewards/format_reward": 0.0,
      "step": 274
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 432.27084732055664,
      "epoch": 2.9493333333333336,
      "grad_norm": 0.5533910393714905,
      "kl": 0.7583274841308594,
      "learning_rate": 1.8795021738670093e-09,
      "loss": 0.0362,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 275
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 389.7916774749756,
      "epoch": 2.96,
      "grad_norm": 0.3034490942955017,
      "kl": 0.48420238494873047,
      "learning_rate": 1.0573165931209162e-09,
      "loss": 0.0194,
      "reward": 0.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.0,
      "step": 276
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 427.93751335144043,
      "epoch": 2.970666666666667,
      "grad_norm": 0.19339117407798767,
      "kl": 0.1923542022705078,
      "learning_rate": 4.699491608713413e-10,
      "loss": 0.0268,
      "reward": 0.9166666716337204,
      "reward_std": 0.12909945845603943,
      "rewards/accuracy_reward": 0.9166666716337204,
      "rewards/format_reward": 0.0,
      "step": 277
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.9791831970215,
      "epoch": 2.981333333333333,
      "grad_norm": 1.0409412384033203,
      "kl": 2.48989200592041,
      "learning_rate": 1.1749189166609943e-10,
      "loss": 0.1356,
      "reward": 0.6666666679084301,
      "reward_std": 0.10206207260489464,
      "rewards/accuracy_reward": 0.6666666679084301,
      "rewards/format_reward": 0.0,
      "step": 278
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 389.6250114440918,
      "epoch": 2.992,
      "grad_norm": 0.8131476044654846,
      "kl": 0.6751537322998047,
      "learning_rate": 0.0,
      "loss": 0.0895,
      "reward": 0.7291666828095913,
      "reward_std": 0.21764283627271652,
      "rewards/accuracy_reward": 0.7291666828095913,
      "rewards/format_reward": 0.0,
      "step": 279
    },
    {
      "epoch": 2.992,
      "step": 279,
      "total_flos": 0.0,
      "train_loss": 0.050804489867134,
      "train_runtime": 11595.9773,
      "train_samples_per_second": 0.194,
      "train_steps_per_second": 0.024
    }
  ],
  "logging_steps": 1,
  "max_steps": 279,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2,
  "trial_name": null,
  "trial_params": null
}