{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 820,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 767.1458587646484,
      "epoch": 0.0012195121951219512,
      "grad_norm": 0.3478255569934845,
      "kl": 0.0,
      "learning_rate": 3.658536585365854e-08,
      "loss": 0.0081,
      "reward": 0.14583333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "completion_length": 681.0208435058594,
      "epoch": 0.0024390243902439024,
      "grad_norm": 0.2987315058708191,
      "kl": 0.0,
      "learning_rate": 7.317073170731708e-08,
      "loss": 0.0155,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 2
    },
    {
      "completion_length": 757.9166870117188,
      "epoch": 0.003658536585365854,
      "grad_norm": 0.307852566242218,
      "kl": 0.00012946128845214844,
      "learning_rate": 1.097560975609756e-07,
      "loss": 0.0121,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 3
    },
    {
      "completion_length": 736.6875152587891,
      "epoch": 0.004878048780487805,
      "grad_norm": 0.38060635328292847,
      "kl": 0.0002968311309814453,
      "learning_rate": 1.4634146341463415e-07,
      "loss": -0.02,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 4
    },
    {
      "completion_length": 665.0208587646484,
      "epoch": 0.006097560975609756,
      "grad_norm": 0.2956765592098236,
      "kl": 0.00019156932830810547,
      "learning_rate": 1.8292682926829268e-07,
      "loss": 0.022,
      "reward": 0.18750000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "completion_length": 636.2291793823242,
      "epoch": 0.007317073170731708,
      "grad_norm": 0.42891716957092285,
      "kl": 0.0001494884490966797,
      "learning_rate": 2.195121951219512e-07,
      "loss": -0.0025,
      "reward": 0.20833334513008595,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.20833334513008595,
      "rewards/format_reward": 0.0,
      "step": 6
    },
    {
      "completion_length": 764.7708435058594,
      "epoch": 0.00853658536585366,
      "grad_norm": 0.3593457341194153,
      "kl": 0.0001971721649169922,
      "learning_rate": 2.5609756097560976e-07,
      "loss": 0.0092,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 7
    },
    {
      "completion_length": 687.5625152587891,
      "epoch": 0.00975609756097561,
      "grad_norm": 0.37344610691070557,
      "kl": 0.0001500844955444336,
      "learning_rate": 2.926829268292683e-07,
      "loss": 0.0041,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 8
    },
    {
      "completion_length": 639.5625152587891,
      "epoch": 0.01097560975609756,
      "grad_norm": 0.3451310694217682,
      "kl": 0.0001842975616455078,
      "learning_rate": 3.2926829268292686e-07,
      "loss": 0.0097,
      "reward": 0.1875000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 9
    },
    {
      "completion_length": 659.6875305175781,
      "epoch": 0.012195121951219513,
      "grad_norm": 0.3643101155757904,
      "kl": 0.00029087066650390625,
      "learning_rate": 3.6585365853658536e-07,
      "loss": -0.0028,
      "reward": 0.20833334513008595,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833334513008595,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "completion_length": 716.4166870117188,
      "epoch": 0.013414634146341463,
      "grad_norm": 0.2868013381958008,
      "kl": 0.0001595020294189453,
      "learning_rate": 4.0243902439024396e-07,
      "loss": -0.0071,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 11
    },
    {
      "completion_length": 636.5208587646484,
      "epoch": 0.014634146341463415,
      "grad_norm": 0.4205193519592285,
      "kl": 0.0001919269561767578,
      "learning_rate": 4.390243902439024e-07,
      "loss": 0.0003,
      "reward": 0.3958333469927311,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3958333469927311,
      "rewards/format_reward": 0.0,
      "step": 12
    },
    {
      "completion_length": 632.6875152587891,
      "epoch": 0.015853658536585366,
      "grad_norm": 0.18752971291542053,
      "kl": 0.00022792816162109375,
      "learning_rate": 4.75609756097561e-07,
      "loss": -0.004,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 13
    },
    {
      "completion_length": 547.1875228881836,
      "epoch": 0.01707317073170732,
      "grad_norm": 0.41111159324645996,
      "kl": 0.0004444122314453125,
      "learning_rate": 5.121951219512195e-07,
      "loss": -0.0099,
      "reward": 0.20833333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 14
    },
    {
      "completion_length": 711.9583587646484,
      "epoch": 0.018292682926829267,
      "grad_norm": 0.28387248516082764,
      "kl": 0.0002880096435546875,
      "learning_rate": 5.48780487804878e-07,
      "loss": 0.0152,
      "reward": 0.1666666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "completion_length": 707.4583587646484,
      "epoch": 0.01951219512195122,
      "grad_norm": 0.19555410742759705,
      "kl": 0.00035881996154785156,
      "learning_rate": 5.853658536585366e-07,
      "loss": 0.001,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 16
    },
    {
      "completion_length": 692.5833587646484,
      "epoch": 0.020731707317073172,
      "grad_norm": 0.5268789529800415,
      "kl": 0.00031948089599609375,
      "learning_rate": 6.219512195121951e-07,
      "loss": -0.0073,
      "reward": 0.291666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.291666679084301,
      "rewards/format_reward": 0.0,
      "step": 17
    },
    {
      "completion_length": 607.5000305175781,
      "epoch": 0.02195121951219512,
      "grad_norm": 0.5240080952644348,
      "kl": 0.00070953369140625,
      "learning_rate": 6.585365853658537e-07,
      "loss": 0.0463,
      "reward": 0.1458333358168602,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 18
    },
    {
      "completion_length": 670.8750305175781,
      "epoch": 0.023170731707317073,
      "grad_norm": 0.30199307203292847,
      "kl": 0.00074005126953125,
      "learning_rate": 6.951219512195122e-07,
      "loss": 0.0105,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 19
    },
    {
      "completion_length": 648.3958587646484,
      "epoch": 0.024390243902439025,
      "grad_norm": 0.31794947385787964,
      "kl": 0.0008096694946289062,
      "learning_rate": 7.317073170731707e-07,
      "loss": -0.0065,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "completion_length": 631.3125,
      "epoch": 0.025609756097560974,
      "grad_norm": 0.17207571864128113,
      "kl": 0.001129150390625,
      "learning_rate": 7.682926829268293e-07,
      "loss": -0.0036,
      "reward": 0.4166666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.4166666716337204,
      "rewards/format_reward": 0.0,
      "step": 21
    },
    {
      "completion_length": 741.9583435058594,
      "epoch": 0.026829268292682926,
      "grad_norm": 0.1936834156513214,
      "kl": 0.0007290840148925781,
      "learning_rate": 8.048780487804879e-07,
      "loss": 0.0063,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 22
    },
    {
      "completion_length": 714.7500305175781,
      "epoch": 0.02804878048780488,
      "grad_norm": 0.035597462207078934,
      "kl": 0.000858306884765625,
      "learning_rate": 8.414634146341464e-07,
      "loss": 0.0,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 23
    },
    {
      "completion_length": 678.1250305175781,
      "epoch": 0.02926829268292683,
      "grad_norm": 0.5932698249816895,
      "kl": 0.001789093017578125,
      "learning_rate": 8.780487804878048e-07,
      "loss": 0.0265,
      "reward": 0.16666666977107525,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 24
    },
    {
      "completion_length": 708.4583587646484,
      "epoch": 0.03048780487804878,
      "grad_norm": 0.39370182156562805,
      "kl": 0.0009379386901855469,
      "learning_rate": 9.146341463414634e-07,
      "loss": 0.0175,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "completion_length": 682.9375228881836,
      "epoch": 0.03170731707317073,
      "grad_norm": 0.3882644772529602,
      "kl": 0.0013561248779296875,
      "learning_rate": 9.51219512195122e-07,
      "loss": -0.0004,
      "reward": 0.22916666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 26
    },
    {
      "completion_length": 645.7291717529297,
      "epoch": 0.032926829268292684,
      "grad_norm": 0.22587056457996368,
      "kl": 0.001148223876953125,
      "learning_rate": 9.878048780487806e-07,
      "loss": 0.0154,
      "reward": 0.2916666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2916666716337204,
      "rewards/format_reward": 0.0,
      "step": 27
    },
    {
      "completion_length": 722.1250305175781,
      "epoch": 0.03414634146341464,
      "grad_norm": 0.240513876080513,
      "kl": 0.0011005401611328125,
      "learning_rate": 1.024390243902439e-06,
      "loss": 0.011,
      "reward": 0.27083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 28
    },
    {
      "completion_length": 739.7083740234375,
      "epoch": 0.03536585365853658,
      "grad_norm": 0.4110949635505676,
      "kl": 0.002593994140625,
      "learning_rate": 1.0609756097560976e-06,
      "loss": 0.0122,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 29
    },
    {
      "completion_length": 723.2916870117188,
      "epoch": 0.036585365853658534,
      "grad_norm": 0.29386106133461,
      "kl": 0.010051727294921875,
      "learning_rate": 1.097560975609756e-06,
      "loss": 0.0079,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 823.7916870117188,
      "epoch": 0.03780487804878049,
      "grad_norm": 0.3077484369277954,
      "kl": 0.001495361328125,
      "learning_rate": 1.1341463414634146e-06,
      "loss": 0.0054,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 31
    },
    {
      "completion_length": 715.5000152587891,
      "epoch": 0.03902439024390244,
      "grad_norm": 0.4433470368385315,
      "kl": 0.003887176513671875,
      "learning_rate": 1.1707317073170732e-06,
      "loss": 0.0097,
      "reward": 0.16666666977107525,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 32
    },
    {
      "completion_length": 779.6458435058594,
      "epoch": 0.04024390243902439,
      "grad_norm": 0.25064408779144287,
      "kl": 0.00261688232421875,
      "learning_rate": 1.2073170731707318e-06,
      "loss": 0.0028,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 33
    },
    {
      "completion_length": 720.6666717529297,
      "epoch": 0.041463414634146344,
      "grad_norm": 0.3388715386390686,
      "kl": 0.0028781890869140625,
      "learning_rate": 1.2439024390243902e-06,
      "loss": 0.0087,
      "reward": 0.0625,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 34
    },
    {
      "completion_length": 722.5833435058594,
      "epoch": 0.042682926829268296,
      "grad_norm": 0.46248888969421387,
      "kl": 0.0028228759765625,
      "learning_rate": 1.2804878048780488e-06,
      "loss": 0.0157,
      "reward": 0.2291666679084301,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "completion_length": 618.3958587646484,
      "epoch": 0.04390243902439024,
      "grad_norm": 0.5465686917304993,
      "kl": 0.00763702392578125,
      "learning_rate": 1.3170731707317074e-06,
      "loss": 0.0127,
      "reward": 0.1666666679084301,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 36
    },
    {
      "completion_length": 609.1458587646484,
      "epoch": 0.045121951219512194,
      "grad_norm": 0.49000048637390137,
      "kl": 0.008819580078125,
      "learning_rate": 1.3536585365853658e-06,
      "loss": -0.0059,
      "reward": 0.31250000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.31250000558793545,
      "rewards/format_reward": 0.0,
      "step": 37
    },
    {
      "completion_length": 670.7708435058594,
      "epoch": 0.046341463414634146,
      "grad_norm": 0.40482643246650696,
      "kl": 0.00677490234375,
      "learning_rate": 1.3902439024390244e-06,
      "loss": -0.0098,
      "reward": 0.2916666753590107,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2916666753590107,
      "rewards/format_reward": 0.0,
      "step": 38
    },
    {
      "completion_length": 747.5833587646484,
      "epoch": 0.0475609756097561,
      "grad_norm": 0.31426116824150085,
      "kl": 0.00624847412109375,
      "learning_rate": 1.4268292682926828e-06,
      "loss": 0.0092,
      "reward": 0.2083333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "completion_length": 783.8750305175781,
      "epoch": 0.04878048780487805,
      "grad_norm": 0.6139870882034302,
      "kl": 0.011974334716796875,
      "learning_rate": 1.4634146341463414e-06,
      "loss": -0.0117,
      "reward": 0.1666666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "completion_length": 696.1666870117188,
      "epoch": 0.05,
      "grad_norm": 0.3642057180404663,
      "kl": 0.0079345703125,
      "learning_rate": 1.5e-06,
      "loss": 0.0253,
      "reward": 0.2708333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 41
    },
    {
      "completion_length": 775.6666870117188,
      "epoch": 0.05121951219512195,
      "grad_norm": 1.5008519887924194,
      "kl": 0.014446258544921875,
      "learning_rate": 1.5365853658536586e-06,
      "loss": 0.0006,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 42
    },
    {
      "completion_length": 697.3750152587891,
      "epoch": 0.0524390243902439,
      "grad_norm": 3.9117236137390137,
      "kl": 0.04592132568359375,
      "learning_rate": 1.5731707317073172e-06,
      "loss": 0.0082,
      "reward": 0.20833333767950535,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.20833333767950535,
      "rewards/format_reward": 0.0,
      "step": 43
    },
    {
      "completion_length": 683.7500152587891,
      "epoch": 0.05365853658536585,
      "grad_norm": 0.2843371629714966,
      "kl": 0.015590667724609375,
      "learning_rate": 1.6097560975609759e-06,
      "loss": -0.0044,
      "reward": 0.3125000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.3125000037252903,
      "rewards/format_reward": 0.0,
      "step": 44
    },
    {
      "completion_length": 734.1875152587891,
      "epoch": 0.054878048780487805,
      "grad_norm": 0.4464169144630432,
      "kl": 0.0031719207763671875,
      "learning_rate": 1.6463414634146342e-06,
      "loss": 0.0145,
      "reward": 0.2500000149011612,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "completion_length": 669.2500152587891,
      "epoch": 0.05609756097560976,
      "grad_norm": 0.4611983001232147,
      "kl": 0.007305145263671875,
      "learning_rate": 1.6829268292682928e-06,
      "loss": 0.0402,
      "reward": 0.2708333432674408,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 46
    },
    {
      "completion_length": 690.3125305175781,
      "epoch": 0.05731707317073171,
      "grad_norm": 0.636078417301178,
      "kl": 0.00937652587890625,
      "learning_rate": 1.719512195121951e-06,
      "loss": 0.001,
      "reward": 0.2291666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 47
    },
    {
      "completion_length": 717.4583587646484,
      "epoch": 0.05853658536585366,
      "grad_norm": 0.29191645979881287,
      "kl": 0.00638580322265625,
      "learning_rate": 1.7560975609756096e-06,
      "loss": -0.013,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 48
    },
    {
      "completion_length": 709.6666870117188,
      "epoch": 0.05975609756097561,
      "grad_norm": 0.290912002325058,
      "kl": 0.01609039306640625,
      "learning_rate": 1.7926829268292682e-06,
      "loss": 0.0004,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 49
    },
    {
      "completion_length": 753.6041870117188,
      "epoch": 0.06097560975609756,
      "grad_norm": 0.6713863015174866,
      "kl": 0.02864837646484375,
      "learning_rate": 1.8292682926829268e-06,
      "loss": 0.0128,
      "reward": 0.16666666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "completion_length": 664.3958587646484,
      "epoch": 0.06219512195121951,
      "grad_norm": 2.1005263328552246,
      "kl": 0.026031494140625,
      "learning_rate": 1.8658536585365854e-06,
      "loss": -0.0023,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 51
    },
    {
      "completion_length": 727.8125305175781,
      "epoch": 0.06341463414634146,
      "grad_norm": 0.3386363387107849,
      "kl": 0.00882720947265625,
      "learning_rate": 1.902439024390244e-06,
      "loss": 0.0006,
      "reward": 0.37500000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.37500000558793545,
      "rewards/format_reward": 0.0,
      "step": 52
    },
    {
      "completion_length": 640.5625152587891,
      "epoch": 0.06463414634146342,
      "grad_norm": 0.5505751371383667,
      "kl": 0.01110076904296875,
      "learning_rate": 1.9390243902439024e-06,
      "loss": 0.0315,
      "reward": 0.3333333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3333333395421505,
      "rewards/format_reward": 0.0,
      "step": 53
    },
    {
      "completion_length": 691.6041870117188,
      "epoch": 0.06585365853658537,
      "grad_norm": 0.4202381372451782,
      "kl": 0.00687408447265625,
      "learning_rate": 1.9756097560975613e-06,
      "loss": -0.0032,
      "reward": 0.25000000931322575,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.25000000931322575,
      "rewards/format_reward": 0.0,
      "step": 54
    },
    {
      "completion_length": 665.6041717529297,
      "epoch": 0.06707317073170732,
      "grad_norm": 0.45118287205696106,
      "kl": 0.00640106201171875,
      "learning_rate": 2.0121951219512197e-06,
      "loss": -0.0008,
      "reward": 0.22916667349636555,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.22916667349636555,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "completion_length": 676.1458435058594,
      "epoch": 0.06829268292682927,
      "grad_norm": 0.4253048896789551,
      "kl": 0.0121307373046875,
      "learning_rate": 2.048780487804878e-06,
      "loss": 0.0127,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 56
    },
    {
      "completion_length": 643.2083435058594,
      "epoch": 0.06951219512195123,
      "grad_norm": 0.2859029769897461,
      "kl": 0.01929473876953125,
      "learning_rate": 2.0853658536585364e-06,
      "loss": 0.0249,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 57
    },
    {
      "completion_length": 682.6042022705078,
      "epoch": 0.07073170731707316,
      "grad_norm": 0.05482205003499985,
      "kl": 0.008087158203125,
      "learning_rate": 2.1219512195121953e-06,
      "loss": 0.0003,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 58
    },
    {
      "completion_length": 681.2083587646484,
      "epoch": 0.07195121951219512,
      "grad_norm": 0.21090681850910187,
      "kl": 0.008453369140625,
      "learning_rate": 2.1585365853658537e-06,
      "loss": -0.0079,
      "reward": 0.1666666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 59
    },
    {
      "completion_length": 650.0208435058594,
      "epoch": 0.07317073170731707,
      "grad_norm": 0.3090149760246277,
      "kl": 0.02374267578125,
      "learning_rate": 2.195121951219512e-06,
      "loss": 0.0123,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "completion_length": 619.6875152587891,
      "epoch": 0.07439024390243902,
      "grad_norm": 0.32917913794517517,
      "kl": 0.00824737548828125,
      "learning_rate": 2.231707317073171e-06,
      "loss": 0.0126,
      "reward": 0.20833333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833333767950535,
      "rewards/format_reward": 0.0,
      "step": 61
    },
    {
      "completion_length": 588.9375152587891,
      "epoch": 0.07560975609756097,
      "grad_norm": 1.3756786584854126,
      "kl": 0.030914306640625,
      "learning_rate": 2.2682926829268293e-06,
      "loss": -0.0259,
      "reward": 0.2500000037252903,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 62
    },
    {
      "completion_length": 656.0000305175781,
      "epoch": 0.07682926829268293,
      "grad_norm": 0.17273320257663727,
      "kl": 0.01983642578125,
      "learning_rate": 2.304878048780488e-06,
      "loss": 0.0008,
      "reward": 0.2916666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 63
    },
    {
      "completion_length": 612.5000305175781,
      "epoch": 0.07804878048780488,
      "grad_norm": 0.29986968636512756,
      "kl": 0.030426025390625,
      "learning_rate": 2.3414634146341465e-06,
      "loss": 0.0016,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 64
    },
    {
      "completion_length": 747.0417022705078,
      "epoch": 0.07926829268292683,
      "grad_norm": 0.3340769410133362,
      "kl": 0.060577392578125,
      "learning_rate": 2.378048780487805e-06,
      "loss": 0.0325,
      "reward": 0.08333333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "completion_length": 635.2708587646484,
      "epoch": 0.08048780487804878,
      "grad_norm": 0.5467635989189148,
      "kl": 0.033050537109375,
      "learning_rate": 2.4146341463414637e-06,
      "loss": -0.0231,
      "reward": 0.2083333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 66
    },
    {
      "completion_length": 609.7500152587891,
      "epoch": 0.08170731707317073,
      "grad_norm": 0.5021399855613708,
      "kl": 0.01324462890625,
      "learning_rate": 2.451219512195122e-06,
      "loss": -0.017,
      "reward": 0.20833333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 67
    },
    {
      "completion_length": 658.9375152587891,
      "epoch": 0.08292682926829269,
      "grad_norm": 0.49131351709365845,
      "kl": 0.0220947265625,
      "learning_rate": 2.4878048780487805e-06,
      "loss": 0.0005,
      "reward": 0.1250000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 68
    },
    {
      "completion_length": 677.2916870117188,
      "epoch": 0.08414634146341464,
      "grad_norm": 0.24393926560878754,
      "kl": 0.0197906494140625,
      "learning_rate": 2.524390243902439e-06,
      "loss": 0.0176,
      "reward": 0.31250000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.31250000558793545,
      "rewards/format_reward": 0.0,
      "step": 69
    },
    {
      "completion_length": 644.2291946411133,
      "epoch": 0.08536585365853659,
      "grad_norm": 0.25599101185798645,
      "kl": 0.01117706298828125,
      "learning_rate": 2.5609756097560977e-06,
      "loss": 0.0093,
      "reward": 0.2083333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "completion_length": 659.5625152587891,
      "epoch": 0.08658536585365853,
      "grad_norm": 0.5464791655540466,
      "kl": 0.0323333740234375,
      "learning_rate": 2.597560975609756e-06,
      "loss": -0.0021,
      "reward": 0.3125000111758709,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.3125000111758709,
      "rewards/format_reward": 0.0,
      "step": 71
    },
    {
      "completion_length": 677.1875152587891,
      "epoch": 0.08780487804878048,
      "grad_norm": 0.35093262791633606,
      "kl": 0.012786865234375,
      "learning_rate": 2.634146341463415e-06,
      "loss": -0.0113,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 72
    },
    {
      "completion_length": 680.4375152587891,
      "epoch": 0.08902439024390243,
      "grad_norm": 0.4046553671360016,
      "kl": 0.016693115234375,
      "learning_rate": 2.6707317073170733e-06,
      "loss": 0.0157,
      "reward": 0.2708333432674408,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 73
    },
    {
      "completion_length": 664.2708587646484,
      "epoch": 0.09024390243902439,
      "grad_norm": 0.3006143569946289,
      "kl": 0.011749267578125,
      "learning_rate": 2.7073170731707317e-06,
      "loss": 0.012,
      "reward": 0.31250000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.31250000558793545,
      "rewards/format_reward": 0.0,
      "step": 74
    },
    {
      "completion_length": 614.4166946411133,
      "epoch": 0.09146341463414634,
      "grad_norm": 0.3110095262527466,
      "kl": 0.015899658203125,
      "learning_rate": 2.7439024390243905e-06,
      "loss": -0.0002,
      "reward": 0.3333333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.3333333395421505,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "completion_length": 756.4791717529297,
      "epoch": 0.09268292682926829,
      "grad_norm": 0.24867770075798035,
      "kl": 0.013641357421875,
      "learning_rate": 2.780487804878049e-06,
      "loss": 0.0112,
      "reward": 0.3333333432674408,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.3333333432674408,
      "rewards/format_reward": 0.0,
      "step": 76
    },
    {
      "completion_length": 695.1666870117188,
      "epoch": 0.09390243902439024,
      "grad_norm": 0.37928274273872375,
      "kl": 0.0152130126953125,
      "learning_rate": 2.8170731707317073e-06,
      "loss": 0.0468,
      "reward": 0.16666666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 77
    },
    {
      "completion_length": 654.5416717529297,
      "epoch": 0.0951219512195122,
      "grad_norm": 0.5844159126281738,
      "kl": 0.0774383544921875,
      "learning_rate": 2.8536585365853657e-06,
      "loss": -0.0113,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 78
    },
    {
      "completion_length": 604.6875152587891,
      "epoch": 0.09634146341463415,
      "grad_norm": 0.6108522415161133,
      "kl": 0.0158538818359375,
      "learning_rate": 2.8902439024390245e-06,
      "loss": 0.0116,
      "reward": 0.3125000074505806,
      "reward_std": 0.25259074568748474,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 79
    },
    {
      "completion_length": 677.2916870117188,
      "epoch": 0.0975609756097561,
      "grad_norm": 0.44601452350616455,
      "kl": 0.012359619140625,
      "learning_rate": 2.926829268292683e-06,
      "loss": 0.0185,
      "reward": 0.29166666977107525,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.29166666977107525,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "completion_length": 705.3125228881836,
      "epoch": 0.09878048780487805,
      "grad_norm": 0.29111340641975403,
      "kl": 0.0164947509765625,
      "learning_rate": 2.9634146341463417e-06,
      "loss": 0.0134,
      "reward": 0.3125000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.3125000037252903,
      "rewards/format_reward": 0.0,
      "step": 81
    },
    {
      "completion_length": 664.2916870117188,
      "epoch": 0.1,
      "grad_norm": 0.28576967120170593,
      "kl": 0.01296234130859375,
      "learning_rate": 3e-06,
      "loss": 0.0119,
      "reward": 0.1458333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 82
    },
    {
      "completion_length": 677.0208435058594,
      "epoch": 0.10121951219512196,
      "grad_norm": 0.3842480480670929,
      "kl": 0.019012451171875,
      "learning_rate": 2.9999864091183917e-06,
      "loss": 0.0061,
      "reward": 0.22916666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 83
    },
    {
      "completion_length": 640.3541870117188,
      "epoch": 0.1024390243902439,
      "grad_norm": 0.437970906496048,
      "kl": 0.0176239013671875,
      "learning_rate": 2.999945636719849e-06,
      "loss": 0.0307,
      "reward": 0.37500000558793545,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.37500000558793545,
      "rewards/format_reward": 0.0,
      "step": 84
    },
    {
      "completion_length": 675.7083740234375,
      "epoch": 0.10365853658536585,
      "grad_norm": 0.30653998255729675,
      "kl": 0.021392822265625,
      "learning_rate": 2.999877683543216e-06,
      "loss": 0.027,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "completion_length": 681.7291946411133,
      "epoch": 0.1048780487804878,
      "grad_norm": 0.2532087564468384,
      "kl": 0.013885498046875,
      "learning_rate": 2.999782550819884e-06,
      "loss": 0.0228,
      "reward": 0.375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.0,
      "step": 86
    },
    {
      "completion_length": 771.5208587646484,
      "epoch": 0.10609756097560975,
      "grad_norm": 0.2872747480869293,
      "kl": 0.0281829833984375,
      "learning_rate": 2.99966024027377e-06,
      "loss": -0.0045,
      "reward": 0.18750000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 87
    },
    {
      "completion_length": 757.2500152587891,
      "epoch": 0.1073170731707317,
      "grad_norm": 0.34172654151916504,
      "kl": 0.01568603515625,
      "learning_rate": 2.9995107541212846e-06,
      "loss": 0.0176,
      "reward": 0.1041666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 88
    },
    {
      "completion_length": 709.6458435058594,
      "epoch": 0.10853658536585366,
      "grad_norm": 0.309572696685791,
      "kl": 0.021453857421875,
      "learning_rate": 2.999334095071293e-06,
      "loss": -0.0184,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 89
    },
    {
      "completion_length": 676.6042022705078,
      "epoch": 0.10975609756097561,
      "grad_norm": 0.4205746352672577,
      "kl": 0.02178955078125,
      "learning_rate": 2.9991302663250642e-06,
      "loss": -0.0122,
      "reward": 0.35416667722165585,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.35416667722165585,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "completion_length": 651.5000305175781,
      "epoch": 0.11097560975609756,
      "grad_norm": 0.2582562267780304,
      "kl": 0.022979736328125,
      "learning_rate": 2.9988992715762147e-06,
      "loss": 0.002,
      "reward": 0.0833333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 91
    },
    {
      "completion_length": 714.7083587646484,
      "epoch": 0.11219512195121951,
      "grad_norm": 0.4848339557647705,
      "kl": 0.037109375,
      "learning_rate": 2.9986411150106423e-06,
      "loss": 0.018,
      "reward": 0.29166666977107525,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.29166666977107525,
      "rewards/format_reward": 0.0,
      "step": 92
    },
    {
      "completion_length": 646.5625152587891,
      "epoch": 0.11341463414634147,
      "grad_norm": 0.2367382049560547,
      "kl": 0.02313232421875,
      "learning_rate": 2.9983558013064455e-06,
      "loss": 0.0004,
      "reward": 0.27083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 93
    },
    {
      "completion_length": 610.5833435058594,
      "epoch": 0.11463414634146342,
      "grad_norm": 0.2747872471809387,
      "kl": 0.018035888671875,
      "learning_rate": 2.998043335633845e-06,
      "loss": 0.027,
      "reward": 0.2291666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 94
    },
    {
      "completion_length": 656.7291717529297,
      "epoch": 0.11585365853658537,
      "grad_norm": 0.41957971453666687,
      "kl": 0.0257568359375,
      "learning_rate": 2.997703723655086e-06,
      "loss": 0.0319,
      "reward": 0.2500000074505806,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 95
    },
    {
      "completion_length": 671.5625152587891,
      "epoch": 0.11707317073170732,
      "grad_norm": 0.2613596022129059,
      "kl": 0.019031524658203125,
      "learning_rate": 2.9973369715243363e-06,
      "loss": 0.0148,
      "reward": 0.29166666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.29166666977107525,
      "rewards/format_reward": 0.0,
      "step": 96
    },
    {
      "completion_length": 653.4583435058594,
      "epoch": 0.11829268292682926,
      "grad_norm": 0.398412823677063,
      "kl": 0.0196075439453125,
      "learning_rate": 2.996943085887577e-06,
      "loss": 0.0413,
      "reward": 0.354166679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.354166679084301,
      "rewards/format_reward": 0.0,
      "step": 97
    },
    {
      "completion_length": 679.0416870117188,
      "epoch": 0.11951219512195121,
      "grad_norm": 0.5563205480575562,
      "kl": 0.031585693359375,
      "learning_rate": 2.996522073882477e-06,
      "loss": 0.0332,
      "reward": 0.2708333358168602,
      "reward_std": 0.25259073823690414,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 98
    },
    {
      "completion_length": 739.4791870117188,
      "epoch": 0.12073170731707317,
      "grad_norm": 0.44225770235061646,
      "kl": 0.023193359375,
      "learning_rate": 2.9960739431382697e-06,
      "loss": 0.0418,
      "reward": 0.2708333395421505,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 99
    },
    {
      "completion_length": 702.7500152587891,
      "epoch": 0.12195121951219512,
      "grad_norm": 0.4029344916343689,
      "kl": 0.03021240234375,
      "learning_rate": 2.9955987017756107e-06,
      "loss": 0.0081,
      "reward": 0.1875000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "completion_length": 691.0416870117188,
      "epoch": 0.12317073170731707,
      "grad_norm": 0.5099979043006897,
      "kl": 0.027618408203125,
      "learning_rate": 2.9950963584064327e-06,
      "loss": 0.0864,
      "reward": 0.458333358168602,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.458333358168602,
      "rewards/format_reward": 0.0,
      "step": 101
    },
    {
      "completion_length": 697.2500305175781,
      "epoch": 0.12439024390243902,
      "grad_norm": 0.3505668640136719,
      "kl": 0.030548095703125,
      "learning_rate": 2.9945669221337873e-06,
      "loss": 0.0143,
      "reward": 0.3125000074505806,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 102
    },
    {
      "completion_length": 624.0208587646484,
      "epoch": 0.12560975609756098,
      "grad_norm": 0.3838660717010498,
      "kl": 0.0420684814453125,
      "learning_rate": 2.994010402551682e-06,
      "loss": 0.0268,
      "reward": 0.25000000931322575,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.25000000931322575,
      "rewards/format_reward": 0.0,
      "step": 103
    },
    {
      "completion_length": 678.0000152587891,
      "epoch": 0.12682926829268293,
      "grad_norm": 1.0560173988342285,
      "kl": 0.076019287109375,
      "learning_rate": 2.9934268097449068e-06,
      "loss": -0.0293,
      "reward": 0.1875000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 104
    },
    {
      "completion_length": 707.5833587646484,
      "epoch": 0.12804878048780488,
      "grad_norm": 0.47438010573387146,
      "kl": 0.034027099609375,
      "learning_rate": 2.9928161542888487e-06,
      "loss": 0.0642,
      "reward": 0.2083333395421505,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 105
    },
    {
      "completion_length": 707.7708511352539,
      "epoch": 0.12926829268292683,
      "grad_norm": 0.4831258952617645,
      "kl": 0.037384033203125,
      "learning_rate": 2.9921784472493023e-06,
      "loss": 0.0701,
      "reward": 0.33333334513008595,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.33333334513008595,
      "rewards/format_reward": 0.0,
      "step": 106
    },
    {
      "completion_length": 649.8125152587891,
      "epoch": 0.13048780487804879,
      "grad_norm": 0.32380735874176025,
      "kl": 0.029388427734375,
      "learning_rate": 2.9915137001822686e-06,
      "loss": 0.0391,
      "reward": 0.3125,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3125,
      "rewards/format_reward": 0.0,
      "step": 107
    },
    {
      "completion_length": 747.3125152587891,
      "epoch": 0.13170731707317074,
      "grad_norm": 0.07502617686986923,
      "kl": 0.037628173828125,
      "learning_rate": 2.9908219251337465e-06,
      "loss": 0.0016,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 108
    },
    {
      "completion_length": 638.4583587646484,
      "epoch": 0.1329268292682927,
      "grad_norm": 0.3937641978263855,
      "kl": 0.0341949462890625,
      "learning_rate": 2.9901031346395125e-06,
      "loss": 0.058,
      "reward": 0.25000000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 109
    },
    {
      "completion_length": 687.3125152587891,
      "epoch": 0.13414634146341464,
      "grad_norm": 0.3746364414691925,
      "kl": 0.042083740234375,
      "learning_rate": 2.9893573417248957e-06,
      "loss": -0.0005,
      "reward": 0.2708333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "completion_length": 732.3541870117188,
      "epoch": 0.1353658536585366,
      "grad_norm": 0.3835819363594055,
      "kl": 0.04608154296875,
      "learning_rate": 2.98858455990454e-06,
      "loss": 0.0732,
      "reward": 0.31250001303851604,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.31250001303851604,
      "rewards/format_reward": 0.0,
      "step": 111
    },
    {
      "completion_length": 796.2708587646484,
      "epoch": 0.13658536585365855,
      "grad_norm": 0.35788479447364807,
      "kl": 0.052490234375,
      "learning_rate": 2.987784803182161e-06,
      "loss": 0.0287,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 112
    },
    {
      "completion_length": 729.2708435058594,
      "epoch": 0.1378048780487805,
      "grad_norm": 0.26465335488319397,
      "kl": 0.041168212890625,
      "learning_rate": 2.9869580860502894e-06,
      "loss": 0.0404,
      "reward": 0.1875000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 113
    },
    {
      "completion_length": 701.9791870117188,
      "epoch": 0.13902439024390245,
      "grad_norm": 0.32841095328330994,
      "kl": 0.048614501953125,
      "learning_rate": 2.9861044234900125e-06,
      "loss": 0.0224,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 114
    },
    {
      "completion_length": 853.8125305175781,
      "epoch": 0.1402439024390244,
      "grad_norm": 0.17240795493125916,
      "kl": 0.23248291015625,
      "learning_rate": 2.985223830970699e-06,
      "loss": -0.0091,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 115
    },
    {
      "completion_length": 732.9791870117188,
      "epoch": 0.14146341463414633,
      "grad_norm": 0.43660447001457214,
      "kl": 0.062744140625,
      "learning_rate": 2.98431632444972e-06,
      "loss": 0.0416,
      "reward": 0.29166667722165585,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.29166667722165585,
      "rewards/format_reward": 0.0,
      "step": 116
    },
    {
      "completion_length": 618.5208587646484,
      "epoch": 0.14268292682926828,
      "grad_norm": 0.32270222902297974,
      "kl": 0.047393798828125,
      "learning_rate": 2.9833819203721614e-06,
      "loss": 0.0548,
      "reward": 0.3958333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.3958333395421505,
      "rewards/format_reward": 0.0,
      "step": 117
    },
    {
      "completion_length": 697.3958587646484,
      "epoch": 0.14390243902439023,
      "grad_norm": 0.320021390914917,
      "kl": 0.07568359375,
      "learning_rate": 2.982420635670523e-06,
      "loss": 0.0067,
      "reward": 0.1875000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 118
    },
    {
      "completion_length": 601.6041870117188,
      "epoch": 0.14512195121951219,
      "grad_norm": 0.6032148003578186,
      "kl": 0.0601806640625,
      "learning_rate": 2.981432487764413e-06,
      "loss": -0.0196,
      "reward": 0.39583334513008595,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.39583334513008595,
      "rewards/format_reward": 0.0,
      "step": 119
    },
    {
      "completion_length": 648.2500152587891,
      "epoch": 0.14634146341463414,
      "grad_norm": 0.5923802852630615,
      "kl": 0.0625,
      "learning_rate": 2.980417494560234e-06,
      "loss": 0.0499,
      "reward": 0.2500000111758709,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2500000111758709,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "completion_length": 740.7500152587891,
      "epoch": 0.1475609756097561,
      "grad_norm": 0.39325180649757385,
      "kl": 0.0552978515625,
      "learning_rate": 2.979375674450855e-06,
      "loss": 0.031,
      "reward": 0.2291666679084301,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 121
    },
    {
      "completion_length": 757.0000152587891,
      "epoch": 0.14878048780487804,
      "grad_norm": 0.3565278947353363,
      "kl": 0.0526123046875,
      "learning_rate": 2.9783070463152816e-06,
      "loss": 0.0422,
      "reward": 0.10416666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 122
    },
    {
      "completion_length": 730.2916717529297,
      "epoch": 0.15,
      "grad_norm": 0.33752405643463135,
      "kl": 0.1187744140625,
      "learning_rate": 2.9772116295183124e-06,
      "loss": 0.0531,
      "reward": 0.29166667349636555,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.29166667349636555,
      "rewards/format_reward": 0.0,
      "step": 123
    },
    {
      "completion_length": 644.7916793823242,
      "epoch": 0.15121951219512195,
      "grad_norm": 0.36829161643981934,
      "kl": 0.0615234375,
      "learning_rate": 2.9760894439101857e-06,
      "loss": 0.0505,
      "reward": 0.3333333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3333333395421505,
      "rewards/format_reward": 0.0,
      "step": 124
    },
    {
      "completion_length": 760.6041870117188,
      "epoch": 0.1524390243902439,
      "grad_norm": 0.2317180037498474,
      "kl": 0.04949951171875,
      "learning_rate": 2.974940509826225e-06,
      "loss": 0.0206,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 125
    },
    {
      "completion_length": 704.7083587646484,
      "epoch": 0.15365853658536585,
      "grad_norm": 0.41940030455589294,
      "kl": 0.05035400390625,
      "learning_rate": 2.973764848086466e-06,
      "loss": -0.0164,
      "reward": 0.2708333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 126
    },
    {
      "completion_length": 610.9166717529297,
      "epoch": 0.1548780487804878,
      "grad_norm": 0.5854915976524353,
      "kl": 0.2664794921875,
      "learning_rate": 2.9725624799952824e-06,
      "loss": 0.1231,
      "reward": 0.5000000149011612,
      "reward_std": 0.28867513686418533,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.0,
      "step": 127
    },
    {
      "completion_length": 696.6875152587891,
      "epoch": 0.15609756097560976,
      "grad_norm": 3.0103659629821777,
      "kl": 0.138916015625,
      "learning_rate": 2.9713334273409965e-06,
      "loss": 0.0224,
      "reward": 0.3541666716337204,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.3541666716337204,
      "rewards/format_reward": 0.0,
      "step": 128
    },
    {
      "completion_length": 767.3958435058594,
      "epoch": 0.1573170731707317,
      "grad_norm": 2.9381563663482666,
      "kl": 0.0977783203125,
      "learning_rate": 2.9700777123954867e-06,
      "loss": 0.0159,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 129
    },
    {
      "completion_length": 614.4583587646484,
      "epoch": 0.15853658536585366,
      "grad_norm": 0.7002465128898621,
      "kl": 0.12139892578125,
      "learning_rate": 2.968795357913784e-06,
      "loss": 0.0806,
      "reward": 0.2708333395421505,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "completion_length": 763.2500305175781,
      "epoch": 0.1597560975609756,
      "grad_norm": 0.8434013724327087,
      "kl": 0.129638671875,
      "learning_rate": 2.9674863871336603e-06,
      "loss": 0.026,
      "reward": 0.1875,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 131
    },
    {
      "completion_length": 767.7500152587891,
      "epoch": 0.16097560975609757,
      "grad_norm": 0.5110030770301819,
      "kl": 0.11328125,
      "learning_rate": 2.9661508237752034e-06,
      "loss": 0.0196,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 132
    },
    {
      "completion_length": 707.7500152587891,
      "epoch": 0.16219512195121952,
      "grad_norm": 0.5542430281639099,
      "kl": 0.080322265625,
      "learning_rate": 2.9647886920403916e-06,
      "loss": 0.0225,
      "reward": 0.2708333395421505,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 133
    },
    {
      "completion_length": 558.6250152587891,
      "epoch": 0.16341463414634147,
      "grad_norm": 0.36637377738952637,
      "kl": 0.0582275390625,
      "learning_rate": 2.9634000166126534e-06,
      "loss": -0.0255,
      "reward": 0.5416666828095913,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.5416666828095913,
      "rewards/format_reward": 0.0,
      "step": 134
    },
    {
      "completion_length": 600.1250152587891,
      "epoch": 0.16463414634146342,
      "grad_norm": 2.5445566177368164,
      "kl": 0.0931396484375,
      "learning_rate": 2.9619848226564196e-06,
      "loss": 0.0446,
      "reward": 0.3958333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3958333358168602,
      "rewards/format_reward": 0.0,
      "step": 135
    },
    {
      "completion_length": 616.5625228881836,
      "epoch": 0.16585365853658537,
      "grad_norm": 0.5558161735534668,
      "kl": 0.05108642578125,
      "learning_rate": 2.9605431358166687e-06,
      "loss": 0.0278,
      "reward": 0.20833333395421505,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 136
    },
    {
      "completion_length": 701.7500152587891,
      "epoch": 0.16707317073170733,
      "grad_norm": 0.4788520932197571,
      "kl": 0.066986083984375,
      "learning_rate": 2.9590749822184602e-06,
      "loss": 0.0637,
      "reward": 0.3541666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3541666679084301,
      "rewards/format_reward": 0.0,
      "step": 137
    },
    {
      "completion_length": 631.6041793823242,
      "epoch": 0.16829268292682928,
      "grad_norm": 0.28127244114875793,
      "kl": 0.0634765625,
      "learning_rate": 2.9575803884664634e-06,
      "loss": 0.0104,
      "reward": 0.375,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.0,
      "step": 138
    },
    {
      "completion_length": 765.1458587646484,
      "epoch": 0.16951219512195123,
      "grad_norm": 0.3691443204879761,
      "kl": 0.0867919921875,
      "learning_rate": 2.9560593816444746e-06,
      "loss": 0.0504,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 139
    },
    {
      "completion_length": 727.3958435058594,
      "epoch": 0.17073170731707318,
      "grad_norm": 0.4014495015144348,
      "kl": 0.0753173828125,
      "learning_rate": 2.9545119893149243e-06,
      "loss": 0.0365,
      "reward": 0.16666667349636555,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.16666667349636555,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "completion_length": 707.2916870117188,
      "epoch": 0.1719512195121951,
      "grad_norm": 0.4650896191596985,
      "kl": 0.0860595703125,
      "learning_rate": 2.9529382395183812e-06,
      "loss": 0.0495,
      "reward": 0.2291666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 141
    },
    {
      "completion_length": 697.2708587646484,
      "epoch": 0.17317073170731706,
      "grad_norm": 0.2692044675350189,
      "kl": 0.0650634765625,
      "learning_rate": 2.9513381607730403e-06,
      "loss": 0.0175,
      "reward": 0.2500000074505806,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 142
    },
    {
      "completion_length": 650.0000152587891,
      "epoch": 0.174390243902439,
      "grad_norm": 0.6151372194290161,
      "kl": 0.072509765625,
      "learning_rate": 2.949711782074211e-06,
      "loss": 0.094,
      "reward": 0.31250001303851604,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.31250001303851604,
      "rewards/format_reward": 0.0,
      "step": 143
    },
    {
      "completion_length": 638.0833435058594,
      "epoch": 0.17560975609756097,
      "grad_norm": 0.2678242027759552,
      "kl": 0.0633544921875,
      "learning_rate": 2.948059132893786e-06,
      "loss": 0.0005,
      "reward": 0.2083333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 144
    },
    {
      "completion_length": 737.1041870117188,
      "epoch": 0.17682926829268292,
      "grad_norm": 0.5129300355911255,
      "kl": 0.0811767578125,
      "learning_rate": 2.9463802431797115e-06,
      "loss": 0.059,
      "reward": 0.22916667349636555,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.22916667349636555,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "completion_length": 759.1458587646484,
      "epoch": 0.17804878048780487,
      "grad_norm": 0.42329859733581543,
      "kl": 0.0982666015625,
      "learning_rate": 2.9446751433554426e-06,
      "loss": 0.0446,
      "reward": 0.25000000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 146
    },
    {
      "completion_length": 698.7916870117188,
      "epoch": 0.17926829268292682,
      "grad_norm": 0.4673228859901428,
      "kl": 0.07696533203125,
      "learning_rate": 2.942943864319392e-06,
      "loss": 0.0718,
      "reward": 0.18750000186264515,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 147
    },
    {
      "completion_length": 673.5208587646484,
      "epoch": 0.18048780487804877,
      "grad_norm": 0.39950287342071533,
      "kl": 0.08953857421875,
      "learning_rate": 2.941186437444372e-06,
      "loss": 0.0155,
      "reward": 0.29166667722165585,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.29166667722165585,
      "rewards/format_reward": 0.0,
      "step": 148
    },
    {
      "completion_length": 790.5833435058594,
      "epoch": 0.18170731707317073,
      "grad_norm": 0.3410983085632324,
      "kl": 0.1240234375,
      "learning_rate": 2.939402894577022e-06,
      "loss": 0.0171,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 149
    },
    {
      "completion_length": 696.9583435058594,
      "epoch": 0.18292682926829268,
      "grad_norm": 0.5283886194229126,
      "kl": 0.1400146484375,
      "learning_rate": 2.9375932680372358e-06,
      "loss": 0.0398,
      "reward": 0.2291666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "completion_length": 665.2500152587891,
      "epoch": 0.18414634146341463,
      "grad_norm": 0.3025929927825928,
      "kl": 0.60015869140625,
      "learning_rate": 2.935757590617574e-06,
      "loss": 0.018,
      "reward": 0.2291666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 151
    },
    {
      "completion_length": 793.3125152587891,
      "epoch": 0.18536585365853658,
      "grad_norm": 0.24417246878147125,
      "kl": 0.080078125,
      "learning_rate": 2.9338958955826685e-06,
      "loss": 0.0104,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 152
    },
    {
      "completion_length": 689.5000152587891,
      "epoch": 0.18658536585365854,
      "grad_norm": 0.5756121873855591,
      "kl": 0.091064453125,
      "learning_rate": 2.9320082166686226e-06,
      "loss": 0.0585,
      "reward": 0.27083333395421505,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 153
    },
    {
      "completion_length": 790.7916870117188,
      "epoch": 0.1878048780487805,
      "grad_norm": 0.3617158830165863,
      "kl": 0.2264404296875,
      "learning_rate": 2.9300945880823955e-06,
      "loss": 0.0628,
      "reward": 0.125,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 154
    },
    {
      "completion_length": 762.0833587646484,
      "epoch": 0.18902439024390244,
      "grad_norm": 0.4066513180732727,
      "kl": 0.139404296875,
      "learning_rate": 2.928155044501189e-06,
      "loss": 0.0637,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 155
    },
    {
      "completion_length": 722.9375152587891,
      "epoch": 0.1902439024390244,
      "grad_norm": 0.5707748532295227,
      "kl": 0.076263427734375,
      "learning_rate": 2.9261896210718106e-06,
      "loss": 0.049,
      "reward": 0.2708333358168602,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 156
    },
    {
      "completion_length": 806.2291870117188,
      "epoch": 0.19146341463414634,
      "grad_norm": 0.42460137605667114,
      "kl": 0.1236572265625,
      "learning_rate": 2.924198353410044e-06,
      "loss": 0.0185,
      "reward": 0.1875000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 157
    },
    {
      "completion_length": 652.0416870117188,
      "epoch": 0.1926829268292683,
      "grad_norm": 0.5879853367805481,
      "kl": 0.0853271484375,
      "learning_rate": 2.9221812776000003e-06,
      "loss": 0.0617,
      "reward": 0.3958333469927311,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3958333469927311,
      "rewards/format_reward": 0.0,
      "step": 158
    },
    {
      "completion_length": 710.0208587646484,
      "epoch": 0.19390243902439025,
      "grad_norm": 0.29009929299354553,
      "kl": 0.1103515625,
      "learning_rate": 2.9201384301934632e-06,
      "loss": 0.0397,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 159
    },
    {
      "completion_length": 787.9583587646484,
      "epoch": 0.1951219512195122,
      "grad_norm": 0.37455567717552185,
      "kl": 0.1287841796875,
      "learning_rate": 2.9180698482092302e-06,
      "loss": 0.017,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "completion_length": 703.0208435058594,
      "epoch": 0.19634146341463415,
      "grad_norm": 0.35033369064331055,
      "kl": 0.089599609375,
      "learning_rate": 2.9159755691324377e-06,
      "loss": 0.0031,
      "reward": 0.0625,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 161
    },
    {
      "completion_length": 658.5208587646484,
      "epoch": 0.1975609756097561,
      "grad_norm": 0.37733712792396545,
      "kl": 0.0885009765625,
      "learning_rate": 2.913855630913884e-06,
      "loss": 0.0108,
      "reward": 0.3125000074505806,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 162
    },
    {
      "completion_length": 696.5000152587891,
      "epoch": 0.19878048780487806,
      "grad_norm": 0.3383764326572418,
      "kl": 0.06280517578125,
      "learning_rate": 2.911710071969342e-06,
      "loss": 0.027,
      "reward": 0.2083333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333432674408,
      "rewards/format_reward": 0.0,
      "step": 163
    },
    {
      "completion_length": 675.0625152587891,
      "epoch": 0.2,
      "grad_norm": 0.5727590918540955,
      "kl": 0.0732421875,
      "learning_rate": 2.9095389311788626e-06,
      "loss": 0.0118,
      "reward": 0.3958333507180214,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.3958333507180214,
      "rewards/format_reward": 0.0,
      "step": 164
    },
    {
      "completion_length": 636.5833587646484,
      "epoch": 0.20121951219512196,
      "grad_norm": 0.5825509428977966,
      "kl": 0.06402587890625,
      "learning_rate": 2.9073422478860678e-06,
      "loss": 0.0833,
      "reward": 0.2916666679084301,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 165
    },
    {
      "completion_length": 676.5625152587891,
      "epoch": 0.20243902439024392,
      "grad_norm": 0.8432393074035645,
      "kl": 0.07525634765625,
      "learning_rate": 2.9051200618974418e-06,
      "loss": 0.0603,
      "reward": 0.25000000558793545,
      "reward_std": 0.25259073823690414,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 166
    },
    {
      "completion_length": 662.4375152587891,
      "epoch": 0.20365853658536584,
      "grad_norm": 0.3825041949748993,
      "kl": 0.064453125,
      "learning_rate": 2.9028724134816064e-06,
      "loss": 0.0568,
      "reward": 0.2291666753590107,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666753590107,
      "rewards/format_reward": 0.0,
      "step": 167
    },
    {
      "completion_length": 700.4375152587891,
      "epoch": 0.2048780487804878,
      "grad_norm": 0.5610911846160889,
      "kl": 0.09527587890625,
      "learning_rate": 2.9005993433685932e-06,
      "loss": 0.0842,
      "reward": 0.2083333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 168
    },
    {
      "completion_length": 760.0625152587891,
      "epoch": 0.20609756097560974,
      "grad_norm": 0.7904462218284607,
      "kl": 0.17626953125,
      "learning_rate": 2.8983008927491046e-06,
      "loss": 0.0572,
      "reward": 0.3125000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3125000037252903,
      "rewards/format_reward": 0.0,
      "step": 169
    },
    {
      "completion_length": 619.1041717529297,
      "epoch": 0.2073170731707317,
      "grad_norm": 0.46067285537719727,
      "kl": 0.120361328125,
      "learning_rate": 2.8959771032737673e-06,
      "loss": 0.0423,
      "reward": 0.1666666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "completion_length": 752.4166870117188,
      "epoch": 0.20853658536585365,
      "grad_norm": 0.5034551024436951,
      "kl": 0.169677734375,
      "learning_rate": 2.8936280170523784e-06,
      "loss": 0.0277,
      "reward": 0.2291666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 171
    },
    {
      "completion_length": 832.8750152587891,
      "epoch": 0.2097560975609756,
      "grad_norm": 0.5332885980606079,
      "kl": 0.234619140625,
      "learning_rate": 2.8912536766531423e-06,
      "loss": 0.0156,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 172
    },
    {
      "completion_length": 769.8750305175781,
      "epoch": 0.21097560975609755,
      "grad_norm": 0.6377100348472595,
      "kl": 0.21484375,
      "learning_rate": 2.8888541251018963e-06,
      "loss": 0.0295,
      "reward": 0.1875000074505806,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875000074505806,
      "rewards/format_reward": 0.0,
      "step": 173
    },
    {
      "completion_length": 784.6250305175781,
      "epoch": 0.2121951219512195,
      "grad_norm": 0.8216028213500977,
      "kl": 0.18408203125,
      "learning_rate": 2.8864294058813364e-06,
      "loss": 0.0078,
      "reward": 0.1666666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 174
    },
    {
      "completion_length": 593.2291793823242,
      "epoch": 0.21341463414634146,
      "grad_norm": 2.0351850986480713,
      "kl": 0.26904296875,
      "learning_rate": 2.883979562930225e-06,
      "loss": 0.0205,
      "reward": 0.3125000149011612,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3125000149011612,
      "rewards/format_reward": 0.0,
      "step": 175
    },
    {
      "completion_length": 675.2083587646484,
      "epoch": 0.2146341463414634,
      "grad_norm": 0.594922661781311,
      "kl": 0.0780029296875,
      "learning_rate": 2.8815046406425954e-06,
      "loss": 0.0132,
      "reward": 0.2083333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 176
    },
    {
      "completion_length": 686.6875305175781,
      "epoch": 0.21585365853658536,
      "grad_norm": 0.3503366708755493,
      "kl": 0.042449951171875,
      "learning_rate": 2.8790046838669493e-06,
      "loss": 0.0099,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 177
    },
    {
      "completion_length": 627.3958587646484,
      "epoch": 0.21707317073170732,
      "grad_norm": 0.6689103245735168,
      "kl": 0.063720703125,
      "learning_rate": 2.876479737905442e-06,
      "loss": 0.0028,
      "reward": 0.2708333358168602,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 178
    },
    {
      "completion_length": 600.2500152587891,
      "epoch": 0.21829268292682927,
      "grad_norm": 0.42488840222358704,
      "kl": 0.065643310546875,
      "learning_rate": 2.8739298485130627e-06,
      "loss": 0.0732,
      "reward": 0.2916666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2916666716337204,
      "rewards/format_reward": 0.0,
      "step": 179
    },
    {
      "completion_length": 636.4791870117188,
      "epoch": 0.21951219512195122,
      "grad_norm": 0.48857226967811584,
      "kl": 0.04949951171875,
      "learning_rate": 2.8713550618968034e-06,
      "loss": 0.027,
      "reward": 0.2291666716337204,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "completion_length": 660.4791870117188,
      "epoch": 0.22073170731707317,
      "grad_norm": 0.4850112497806549,
      "kl": 0.060791015625,
      "learning_rate": 2.8687554247148247e-06,
      "loss": 0.0623,
      "reward": 0.3750000149011612,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.3750000149011612,
      "rewards/format_reward": 0.0,
      "step": 181
    },
    {
      "completion_length": 669.1458587646484,
      "epoch": 0.22195121951219512,
      "grad_norm": 0.4661165177822113,
      "kl": 0.05609130859375,
      "learning_rate": 2.8661309840756093e-06,
      "loss": 0.0076,
      "reward": 0.27083333395421505,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 182
    },
    {
      "completion_length": 687.3541870117188,
      "epoch": 0.22317073170731708,
      "grad_norm": 0.46282872557640076,
      "kl": 0.06842041015625,
      "learning_rate": 2.863481787537105e-06,
      "loss": 0.0015,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 183
    },
    {
      "completion_length": 653.9583435058594,
      "epoch": 0.22439024390243903,
      "grad_norm": 0.21007929742336273,
      "kl": 0.05267333984375,
      "learning_rate": 2.8608078831058682e-06,
      "loss": -0.0003,
      "reward": 0.1666666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 184
    },
    {
      "completion_length": 534.1875228881836,
      "epoch": 0.22560975609756098,
      "grad_norm": 0.6113242506980896,
      "kl": 0.128173828125,
      "learning_rate": 2.8581093192361895e-06,
      "loss": 0.0359,
      "reward": 0.4166666865348816,
      "reward_std": 0.28867512941360474,
      "rewards/accuracy_reward": 0.4166666865348816,
      "rewards/format_reward": 0.0,
      "step": 185
    },
    {
      "completion_length": 697.0416870117188,
      "epoch": 0.22682926829268293,
      "grad_norm": 0.37834686040878296,
      "kl": 0.10772705078125,
      "learning_rate": 2.8553861448292185e-06,
      "loss": 0.0158,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 186
    },
    {
      "completion_length": 763.2708587646484,
      "epoch": 0.2280487804878049,
      "grad_norm": 0.3573489785194397,
      "kl": 0.118408203125,
      "learning_rate": 2.852638409232077e-06,
      "loss": 0.0881,
      "reward": 0.2500000111758709,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2500000111758709,
      "rewards/format_reward": 0.0,
      "step": 187
    },
    {
      "completion_length": 757.7292022705078,
      "epoch": 0.22926829268292684,
      "grad_norm": 1.2989840507507324,
      "kl": 0.101318359375,
      "learning_rate": 2.8498661622369637e-06,
      "loss": 0.0063,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 188
    },
    {
      "completion_length": 802.5833435058594,
      "epoch": 0.2304878048780488,
      "grad_norm": 0.23137597739696503,
      "kl": 0.1221923828125,
      "learning_rate": 2.8470694540802527e-06,
      "loss": 0.0615,
      "reward": 0.0625,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 189
    },
    {
      "completion_length": 837.6458587646484,
      "epoch": 0.23170731707317074,
      "grad_norm": 0.324842244386673,
      "kl": 0.115478515625,
      "learning_rate": 2.8442483354415836e-06,
      "loss": 0.0841,
      "reward": 0.2500000074505806,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 190
    },
    {
      "completion_length": 793.4583435058594,
      "epoch": 0.2329268292682927,
      "grad_norm": 0.3294341564178467,
      "kl": 0.0999755859375,
      "learning_rate": 2.841402857442942e-06,
      "loss": 0.0247,
      "reward": 0.2083333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 191
    },
    {
      "completion_length": 880.2083435058594,
      "epoch": 0.23414634146341465,
      "grad_norm": 0.18687376379966736,
      "kl": 0.1146240234375,
      "learning_rate": 2.8385330716477335e-06,
      "loss": 0.0211,
      "reward": 0.0416666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 192
    },
    {
      "completion_length": 795.9791717529297,
      "epoch": 0.23536585365853657,
      "grad_norm": 0.3349559009075165,
      "kl": 0.21075439453125,
      "learning_rate": 2.835639030059851e-06,
      "loss": 0.0246,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 193
    },
    {
      "completion_length": 854.7083587646484,
      "epoch": 0.23658536585365852,
      "grad_norm": 0.30049294233322144,
      "kl": 0.09716796875,
      "learning_rate": 2.8327207851227295e-06,
      "loss": 0.0445,
      "reward": 0.1458333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 194
    },
    {
      "completion_length": 889.4375152587891,
      "epoch": 0.23780487804878048,
      "grad_norm": 0.23594792187213898,
      "kl": 0.126708984375,
      "learning_rate": 2.829778389718398e-06,
      "loss": 0.0145,
      "reward": 0.2083333432674408,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2083333432674408,
      "rewards/format_reward": 0.0,
      "step": 195
    },
    {
      "completion_length": 827.1458587646484,
      "epoch": 0.23902439024390243,
      "grad_norm": 0.37217977643013,
      "kl": 0.1265869140625,
      "learning_rate": 2.826811897166519e-06,
      "loss": 0.0535,
      "reward": 0.1458333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 196
    },
    {
      "completion_length": 869.6458587646484,
      "epoch": 0.24024390243902438,
      "grad_norm": 0.33080142736434937,
      "kl": 0.1011962890625,
      "learning_rate": 2.8238213612234255e-06,
      "loss": 0.0478,
      "reward": 0.2291666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 197
    },
    {
      "completion_length": 832.7916870117188,
      "epoch": 0.24146341463414633,
      "grad_norm": 0.17182479798793793,
      "kl": 0.095703125,
      "learning_rate": 2.8208068360811445e-06,
      "loss": 0.0033,
      "reward": 0.2083333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 198
    },
    {
      "completion_length": 911.9166870117188,
      "epoch": 0.2426829268292683,
      "grad_norm": 0.17546705901622772,
      "kl": 0.10009765625,
      "learning_rate": 2.8177683763664137e-06,
      "loss": 0.0141,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 199
    },
    {
      "completion_length": 916.7292022705078,
      "epoch": 0.24390243902439024,
      "grad_norm": 0.23637273907661438,
      "kl": 0.0970458984375,
      "learning_rate": 2.8147060371396953e-06,
      "loss": 0.0396,
      "reward": 0.1250000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 200
    },
    {
      "completion_length": 853.1250305175781,
      "epoch": 0.2451219512195122,
      "grad_norm": 0.1615118533372879,
      "kl": 0.08892822265625,
      "learning_rate": 2.8116198738941766e-06,
      "loss": 0.0202,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 201
    },
    {
      "completion_length": 835.3750305175781,
      "epoch": 0.24634146341463414,
      "grad_norm": 0.6049842238426208,
      "kl": 0.10400390625,
      "learning_rate": 2.8085099425547627e-06,
      "loss": 0.1052,
      "reward": 0.25000000558793545,
      "reward_std": 0.25259073823690414,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 202
    },
    {
      "completion_length": 839.3958587646484,
      "epoch": 0.2475609756097561,
      "grad_norm": 0.328227698802948,
      "kl": 0.113037109375,
      "learning_rate": 2.8053762994770646e-06,
      "loss": 0.0349,
      "reward": 0.14583333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 203
    },
    {
      "completion_length": 829.1250305175781,
      "epoch": 0.24878048780487805,
      "grad_norm": 0.2759798467159271,
      "kl": 0.0972900390625,
      "learning_rate": 2.8022190014463794e-06,
      "loss": 0.0575,
      "reward": 0.14583333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 204
    },
    {
      "completion_length": 751.8541870117188,
      "epoch": 0.25,
      "grad_norm": 0.3387543857097626,
      "kl": 0.07476806640625,
      "learning_rate": 2.7990381056766585e-06,
      "loss": 0.0605,
      "reward": 0.2291666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 205
    },
    {
      "completion_length": 731.0000152587891,
      "epoch": 0.25121951219512195,
      "grad_norm": 0.3356465697288513,
      "kl": 0.087890625,
      "learning_rate": 2.795833669809471e-06,
      "loss": 0.0671,
      "reward": 0.3125000074505806,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 206
    },
    {
      "completion_length": 850.7916870117188,
      "epoch": 0.2524390243902439,
      "grad_norm": 0.3686448931694031,
      "kl": 0.07843017578125,
      "learning_rate": 2.7926057519129634e-06,
      "loss": 0.0478,
      "reward": 0.1041666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 207
    },
    {
      "completion_length": 767.7292022705078,
      "epoch": 0.25365853658536586,
      "grad_norm": 0.4388621151447296,
      "kl": 0.07598876953125,
      "learning_rate": 2.7893544104808017e-06,
      "loss": 0.0291,
      "reward": 0.2500000074505806,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 208
    },
    {
      "completion_length": 690.0000152587891,
      "epoch": 0.2548780487804878,
      "grad_norm": 0.26546475291252136,
      "kl": 0.0836181640625,
      "learning_rate": 2.7860797044311143e-06,
      "loss": 0.0494,
      "reward": 0.3333333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3333333432674408,
      "rewards/format_reward": 0.0,
      "step": 209
    },
    {
      "completion_length": 740.2916870117188,
      "epoch": 0.25609756097560976,
      "grad_norm": 0.35596325993537903,
      "kl": 0.1085205078125,
      "learning_rate": 2.7827816931054245e-06,
      "loss": 0.0692,
      "reward": 0.14583333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 210
    },
    {
      "completion_length": 738.7291870117188,
      "epoch": 0.2573170731707317,
      "grad_norm": 0.256867915391922,
      "kl": 0.0791015625,
      "learning_rate": 2.7794604362675733e-06,
      "loss": 0.0167,
      "reward": 0.14583333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 211
    },
    {
      "completion_length": 722.7500305175781,
      "epoch": 0.25853658536585367,
      "grad_norm": 0.23055137693881989,
      "kl": 0.0509033203125,
      "learning_rate": 2.7761159941026403e-06,
      "loss": 0.0188,
      "reward": 0.1875,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 212
    },
    {
      "completion_length": 727.6041870117188,
      "epoch": 0.2597560975609756,
      "grad_norm": 0.37032070755958557,
      "kl": 0.1087646484375,
      "learning_rate": 2.772748427215848e-06,
      "loss": 0.0334,
      "reward": 0.20833333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 213
    },
    {
      "completion_length": 703.1041870117188,
      "epoch": 0.26097560975609757,
      "grad_norm": 0.17007280886173248,
      "kl": 0.0704345703125,
      "learning_rate": 2.7693577966314664e-06,
      "loss": -0.0113,
      "reward": 0.2916666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2916666716337204,
      "rewards/format_reward": 0.0,
      "step": 214
    },
    {
      "completion_length": 693.5625152587891,
      "epoch": 0.2621951219512195,
      "grad_norm": 0.37101563811302185,
      "kl": 0.068603515625,
      "learning_rate": 2.7659441637917076e-06,
      "loss": 0.0401,
      "reward": 0.3541666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3541666716337204,
      "rewards/format_reward": 0.0,
      "step": 215
    },
    {
      "completion_length": 805.2291870117188,
      "epoch": 0.2634146341463415,
      "grad_norm": 0.402241975069046,
      "kl": 0.07598876953125,
      "learning_rate": 2.7625075905556117e-06,
      "loss": 0.0328,
      "reward": 0.25000000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 216
    },
    {
      "completion_length": 695.0833587646484,
      "epoch": 0.2646341463414634,
      "grad_norm": 0.4580467641353607,
      "kl": 0.0611572265625,
      "learning_rate": 2.7590481391979253e-06,
      "loss": 0.0555,
      "reward": 0.2500000074505806,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 217
    },
    {
      "completion_length": 798.3125152587891,
      "epoch": 0.2658536585365854,
      "grad_norm": 0.16156983375549316,
      "kl": 0.1004638671875,
      "learning_rate": 2.755565872407973e-06,
      "loss": 0.0047,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 218
    },
    {
      "completion_length": 831.1250305175781,
      "epoch": 0.26707317073170733,
      "grad_norm": 0.4522290527820587,
      "kl": 0.10052490234375,
      "learning_rate": 2.7520608532885228e-06,
      "loss": 0.0404,
      "reward": 0.16666666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 219
    },
    {
      "completion_length": 808.8333435058594,
      "epoch": 0.2682926829268293,
      "grad_norm": 0.480506956577301,
      "kl": 0.7591552734375,
      "learning_rate": 2.7485331453546407e-06,
      "loss": 0.0056,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 220
    },
    {
      "completion_length": 840.6041870117188,
      "epoch": 0.26951219512195124,
      "grad_norm": 0.25887441635131836,
      "kl": 0.1767578125,
      "learning_rate": 2.744982812532542e-06,
      "loss": 0.0209,
      "reward": 0.2291666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 221
    },
    {
      "completion_length": 826.4166870117188,
      "epoch": 0.2707317073170732,
      "grad_norm": 0.34200406074523926,
      "kl": 0.09765625,
      "learning_rate": 2.7414099191584305e-06,
      "loss": 0.0488,
      "reward": 0.4166666828095913,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.4166666828095913,
      "rewards/format_reward": 0.0,
      "step": 222
    },
    {
      "completion_length": 839.3750305175781,
      "epoch": 0.27195121951219514,
      "grad_norm": 0.4653070867061615,
      "kl": 0.12353515625,
      "learning_rate": 2.7378145299773337e-06,
      "loss": 0.0568,
      "reward": 0.1458333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 223
    },
    {
      "completion_length": 745.7083587646484,
      "epoch": 0.2731707317073171,
      "grad_norm": 0.823951005935669,
      "kl": 0.1021728515625,
      "learning_rate": 2.7341967101419303e-06,
      "loss": 0.0397,
      "reward": 0.2708333432674408,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 224
    },
    {
      "completion_length": 677.9166870117188,
      "epoch": 0.27439024390243905,
      "grad_norm": 0.2579309344291687,
      "kl": 0.12969970703125,
      "learning_rate": 2.730556525211368e-06,
      "loss": 0.0294,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 225
    },
    {
      "completion_length": 746.7500305175781,
      "epoch": 0.275609756097561,
      "grad_norm": 0.21033814549446106,
      "kl": 0.119140625,
      "learning_rate": 2.726894041150077e-06,
      "loss": 0.0054,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 226
    },
    {
      "completion_length": 825.9166717529297,
      "epoch": 0.27682926829268295,
      "grad_norm": 0.30933907628059387,
      "kl": 0.1221923828125,
      "learning_rate": 2.7232093243265727e-06,
      "loss": 0.0415,
      "reward": 0.25,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.0,
      "step": 227
    },
    {
      "completion_length": 732.5000152587891,
      "epoch": 0.2780487804878049,
      "grad_norm": 0.30718833208084106,
      "kl": 0.08837890625,
      "learning_rate": 2.7195024415122565e-06,
      "loss": 0.0786,
      "reward": 0.27083334140479565,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.27083334140479565,
      "rewards/format_reward": 0.0,
      "step": 228
    },
    {
      "completion_length": 721.2083587646484,
      "epoch": 0.27926829268292686,
      "grad_norm": 0.4732055366039276,
      "kl": 0.0865478515625,
      "learning_rate": 2.715773459880202e-06,
      "loss": 0.0548,
      "reward": 0.2500000111758709,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2500000111758709,
      "rewards/format_reward": 0.0,
      "step": 229
    },
    {
      "completion_length": 782.3125152587891,
      "epoch": 0.2804878048780488,
      "grad_norm": 0.4474899470806122,
      "kl": 0.1129150390625,
      "learning_rate": 2.7120224470039394e-06,
      "loss": 0.0423,
      "reward": 0.14583333767950535,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 230
    },
    {
      "completion_length": 711.0000152587891,
      "epoch": 0.2817073170731707,
      "grad_norm": 0.18610194325447083,
      "kl": 0.08013916015625,
      "learning_rate": 2.7082494708562316e-06,
      "loss": 0.0039,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 231
    },
    {
      "completion_length": 751.6458435058594,
      "epoch": 0.28292682926829266,
      "grad_norm": 0.3594924211502075,
      "kl": 0.0958251953125,
      "learning_rate": 2.7044545998078414e-06,
      "loss": 0.048,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 232
    },
    {
      "completion_length": 776.0416717529297,
      "epoch": 0.2841463414634146,
      "grad_norm": 0.48652586340904236,
      "kl": 0.11065673828125,
      "learning_rate": 2.7006379026262924e-06,
      "loss": 0.1083,
      "reward": 0.1666666716337204,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 233
    },
    {
      "completion_length": 760.1875152587891,
      "epoch": 0.28536585365853656,
      "grad_norm": 0.2556437849998474,
      "kl": 0.133544921875,
      "learning_rate": 2.696799448474625e-06,
      "loss": 0.006,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 234
    },
    {
      "completion_length": 783.4375305175781,
      "epoch": 0.2865853658536585,
      "grad_norm": 0.5801549553871155,
      "kl": 0.1484375,
      "learning_rate": 2.69293930691014e-06,
      "loss": 0.0255,
      "reward": 0.12500000186264515,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 235
    },
    {
      "completion_length": 705.3125076293945,
      "epoch": 0.28780487804878047,
      "grad_norm": 0.14213697612285614,
      "kl": 0.0775146484375,
      "learning_rate": 2.689057547883139e-06,
      "loss": 0.0075,
      "reward": 0.20833333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 236
    },
    {
      "completion_length": 656.5208587646484,
      "epoch": 0.2890243902439024,
      "grad_norm": 0.3858877420425415,
      "kl": 0.0679931640625,
      "learning_rate": 2.6851542417356605e-06,
      "loss": 0.0487,
      "reward": 0.29166667349636555,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.29166667349636555,
      "rewards/format_reward": 0.0,
      "step": 237
    },
    {
      "completion_length": 673.2083587646484,
      "epoch": 0.29024390243902437,
      "grad_norm": 0.6804270148277283,
      "kl": 0.1614990234375,
      "learning_rate": 2.6812294592001984e-06,
      "loss": 0.025,
      "reward": 0.0833333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 238
    },
    {
      "completion_length": 791.8125152587891,
      "epoch": 0.2914634146341463,
      "grad_norm": 0.41352158784866333,
      "kl": 0.090087890625,
      "learning_rate": 2.677283271398427e-06,
      "loss": 0.0205,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 239
    },
    {
      "completion_length": 819.4167022705078,
      "epoch": 0.2926829268292683,
      "grad_norm": 0.18944472074508667,
      "kl": 0.1143798828125,
      "learning_rate": 2.673315749839907e-06,
      "loss": 0.0049,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 240
    },
    {
      "completion_length": 851.4167022705078,
      "epoch": 0.2939024390243902,
      "grad_norm": 0.38409966230392456,
      "kl": 0.1602783203125,
      "learning_rate": 2.669326966420793e-06,
      "loss": 0.0238,
      "reward": 0.22916666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 241
    },
    {
      "completion_length": 756.0416946411133,
      "epoch": 0.2951219512195122,
      "grad_norm": 0.3536222577095032,
      "kl": 0.1505126953125,
      "learning_rate": 2.6653169934225295e-06,
      "loss": 0.0315,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 242
    },
    {
      "completion_length": 823.2708587646484,
      "epoch": 0.29634146341463413,
      "grad_norm": 0.4329302906990051,
      "kl": 0.150146484375,
      "learning_rate": 2.661285903510541e-06,
      "loss": 0.0401,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 243
    },
    {
      "completion_length": 843.6250305175781,
      "epoch": 0.2975609756097561,
      "grad_norm": 0.8410559296607971,
      "kl": 0.14971923828125,
      "learning_rate": 2.6572337697329145e-06,
      "loss": 0.0322,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 244
    },
    {
      "completion_length": 760.8958587646484,
      "epoch": 0.29878048780487804,
      "grad_norm": 0.6818459033966064,
      "kl": 0.077392578125,
      "learning_rate": 2.6531606655190777e-06,
      "loss": 0.0257,
      "reward": 0.16666667349636555,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.16666667349636555,
      "rewards/format_reward": 0.0,
      "step": 245
    },
    {
      "completion_length": 780.9166870117188,
      "epoch": 0.3,
      "grad_norm": 0.5315125584602356,
      "kl": 0.1397705078125,
      "learning_rate": 2.649066664678467e-06,
      "loss": 0.0354,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 246
    },
    {
      "completion_length": 705.9583587646484,
      "epoch": 0.30121951219512194,
      "grad_norm": 0.815244734287262,
      "kl": 0.122802734375,
      "learning_rate": 2.64495184139919e-06,
      "loss": 0.0385,
      "reward": 0.20833334513008595,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.20833334513008595,
      "rewards/format_reward": 0.0,
      "step": 247
    },
    {
      "completion_length": 759.0000152587891,
      "epoch": 0.3024390243902439,
      "grad_norm": 0.40733662247657776,
      "kl": 0.12432861328125,
      "learning_rate": 2.640816270246681e-06,
      "loss": 0.0206,
      "reward": 0.1041666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 248
    },
    {
      "completion_length": 769.0625152587891,
      "epoch": 0.30365853658536585,
      "grad_norm": 0.2662230432033539,
      "kl": 0.0831298828125,
      "learning_rate": 2.636660026162351e-06,
      "loss": 0.051,
      "reward": 0.2083333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 249
    },
    {
      "completion_length": 793.7291870117188,
      "epoch": 0.3048780487804878,
      "grad_norm": 0.3390912115573883,
      "kl": 0.063507080078125,
      "learning_rate": 2.6324831844622278e-06,
      "loss": 0.0483,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 250
    },
    {
      "completion_length": 649.5625152587891,
      "epoch": 0.30609756097560975,
      "grad_norm": 0.4368344843387604,
      "kl": 0.0704345703125,
      "learning_rate": 2.628285820835593e-06,
      "loss": 0.0146,
      "reward": 0.27083333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 251
    },
    {
      "completion_length": 739.4583587646484,
      "epoch": 0.3073170731707317,
      "grad_norm": 0.35745346546173096,
      "kl": 0.07891845703125,
      "learning_rate": 2.6240680113436096e-06,
      "loss": 0.0438,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 252
    },
    {
      "completion_length": 840.1042022705078,
      "epoch": 0.30853658536585366,
      "grad_norm": 0.23686417937278748,
      "kl": 0.081787109375,
      "learning_rate": 2.619829832417944e-06,
      "loss": 0.0185,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 253
    },
    {
      "completion_length": 701.2500152587891,
      "epoch": 0.3097560975609756,
      "grad_norm": 0.432927668094635,
      "kl": 0.0609130859375,
      "learning_rate": 2.6155713608593796e-06,
      "loss": 0.0465,
      "reward": 0.33333334140479565,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.33333334140479565,
      "rewards/format_reward": 0.0,
      "step": 254
    },
    {
      "completion_length": 705.1458587646484,
      "epoch": 0.31097560975609756,
      "grad_norm": 0.357907772064209,
      "kl": 0.071044921875,
      "learning_rate": 2.6112926738364267e-06,
      "loss": 0.0373,
      "reward": 0.1875000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 255
    },
    {
      "completion_length": 643.2916870117188,
      "epoch": 0.3121951219512195,
      "grad_norm": 0.43545010685920715,
      "kl": 0.05633544921875,
      "learning_rate": 2.606993848883924e-06,
      "loss": 0.0641,
      "reward": 0.1875000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 256
    },
    {
      "completion_length": 795.9583435058594,
      "epoch": 0.31341463414634146,
      "grad_norm": 0.3587723970413208,
      "kl": 0.10430908203125,
      "learning_rate": 2.6026749639016327e-06,
      "loss": 0.0376,
      "reward": 0.2291666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 257
    },
    {
      "completion_length": 747.0000152587891,
      "epoch": 0.3146341463414634,
      "grad_norm": 0.3565123677253723,
      "kl": 0.0775146484375,
      "learning_rate": 2.5983360971528252e-06,
      "loss": 0.0476,
      "reward": 0.27083334140479565,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.27083334140479565,
      "rewards/format_reward": 0.0,
      "step": 258
    },
    {
      "completion_length": 725.3958587646484,
      "epoch": 0.31585365853658537,
      "grad_norm": 0.36558693647384644,
      "kl": 0.0751953125,
      "learning_rate": 2.5939773272628674e-06,
      "loss": 0.0341,
      "reward": 0.33333333395421505,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.33333333395421505,
      "rewards/format_reward": 0.0,
      "step": 259
    },
    {
      "completion_length": 705.7500152587891,
      "epoch": 0.3170731707317073,
      "grad_norm": 0.3691394627094269,
      "kl": 0.057159423828125,
      "learning_rate": 2.5895987332177935e-06,
      "loss": 0.0472,
      "reward": 0.1250000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 260
    },
    {
      "completion_length": 668.5833587646484,
      "epoch": 0.3182926829268293,
      "grad_norm": 0.33139413595199585,
      "kl": 0.08819580078125,
      "learning_rate": 2.5852003943628746e-06,
      "loss": 0.04,
      "reward": 0.2708333432674408,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 261
    },
    {
      "completion_length": 736.2916870117188,
      "epoch": 0.3195121951219512,
      "grad_norm": 0.36281514167785645,
      "kl": 0.09375,
      "learning_rate": 2.5807823904011804e-06,
      "loss": 0.0208,
      "reward": 0.1041666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 262
    },
    {
      "completion_length": 750.3958435058594,
      "epoch": 0.3207317073170732,
      "grad_norm": 0.3322524428367615,
      "kl": 0.078460693359375,
      "learning_rate": 2.576344801392137e-06,
      "loss": 0.0509,
      "reward": 0.1666666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 263
    },
    {
      "completion_length": 743.6458587646484,
      "epoch": 0.32195121951219513,
      "grad_norm": 0.2838097810745239,
      "kl": 0.09014892578125,
      "learning_rate": 2.571887707750072e-06,
      "loss": 0.0202,
      "reward": 0.0833333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 264
    },
    {
      "completion_length": 790.8958587646484,
      "epoch": 0.3231707317073171,
      "grad_norm": 0.5115813612937927,
      "kl": 0.10302734375,
      "learning_rate": 2.5674111902427625e-06,
      "loss": 0.0609,
      "reward": 0.16666666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 265
    },
    {
      "completion_length": 649.3333435058594,
      "epoch": 0.32439024390243903,
      "grad_norm": 0.38355883955955505,
      "kl": 0.07867431640625,
      "learning_rate": 2.5629153299899673e-06,
      "loss": 0.0479,
      "reward": 0.2916666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 266
    },
    {
      "completion_length": 753.2083587646484,
      "epoch": 0.325609756097561,
      "grad_norm": 0.24421904981136322,
      "kl": 0.1646728515625,
      "learning_rate": 2.5584002084619593e-06,
      "loss": 0.0232,
      "reward": 0.3125000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.3125000037252903,
      "rewards/format_reward": 0.0,
      "step": 267
    },
    {
      "completion_length": 847.6041870117188,
      "epoch": 0.32682926829268294,
      "grad_norm": 0.2907714247703552,
      "kl": 0.1578369140625,
      "learning_rate": 2.5538659074780484e-06,
      "loss": 0.0194,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 268
    },
    {
      "completion_length": 800.2916870117188,
      "epoch": 0.3280487804878049,
      "grad_norm": 0.3864041864871979,
      "kl": 0.2236328125,
      "learning_rate": 2.549312509205097e-06,
      "loss": 0.0243,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 269
    },
    {
      "completion_length": 841.6875305175781,
      "epoch": 0.32926829268292684,
      "grad_norm": 0.464433491230011,
      "kl": 0.1407470703125,
      "learning_rate": 2.5447400961560355e-06,
      "loss": 0.0293,
      "reward": 0.1875000111758709,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000111758709,
      "rewards/format_reward": 0.0,
      "step": 270
    },
    {
      "completion_length": 682.1250152587891,
      "epoch": 0.3304878048780488,
      "grad_norm": 0.5541307330131531,
      "kl": 0.1240234375,
      "learning_rate": 2.5401487511883627e-06,
      "loss": 0.0339,
      "reward": 0.3125000111758709,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.3125000111758709,
      "rewards/format_reward": 0.0,
      "step": 271
    },
    {
      "completion_length": 647.3750305175781,
      "epoch": 0.33170731707317075,
      "grad_norm": 0.35826483368873596,
      "kl": 0.05914306640625,
      "learning_rate": 2.5355385575026464e-06,
      "loss": 0.0346,
      "reward": 0.3125000074505806,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 272
    },
    {
      "completion_length": 750.1041870117188,
      "epoch": 0.3329268292682927,
      "grad_norm": 0.5364739894866943,
      "kl": 0.1041259765625,
      "learning_rate": 2.5309095986410155e-06,
      "loss": 0.0145,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 273
    },
    {
      "completion_length": 648.8333587646484,
      "epoch": 0.33414634146341465,
      "grad_norm": 0.23630757629871368,
      "kl": 0.07232666015625,
      "learning_rate": 2.5262619584856456e-06,
      "loss": 0.0216,
      "reward": 0.2708333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 274
    },
    {
      "completion_length": 670.5416793823242,
      "epoch": 0.3353658536585366,
      "grad_norm": 0.8087655305862427,
      "kl": 0.057586669921875,
      "learning_rate": 2.52159572125724e-06,
      "loss": 0.0466,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 275
    },
    {
      "completion_length": 854.6250152587891,
      "epoch": 0.33658536585365856,
      "grad_norm": 0.5475993752479553,
      "kl": 0.1041259765625,
      "learning_rate": 2.5169109715135015e-06,
      "loss": 0.0211,
      "reward": 0.1458333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 276
    },
    {
      "completion_length": 707.7917022705078,
      "epoch": 0.3378048780487805,
      "grad_norm": 0.433810830116272,
      "kl": 0.129486083984375,
      "learning_rate": 2.512207794147603e-06,
      "loss": 0.0675,
      "reward": 0.25000000558793545,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 277
    },
    {
      "completion_length": 666.4166870117188,
      "epoch": 0.33902439024390246,
      "grad_norm": 0.29318520426750183,
      "kl": 0.1146240234375,
      "learning_rate": 2.507486274386647e-06,
      "loss": 0.0197,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 278
    },
    {
      "completion_length": 719.7500305175781,
      "epoch": 0.3402439024390244,
      "grad_norm": 0.7796963453292847,
      "kl": 0.091064453125,
      "learning_rate": 2.5027464977901206e-06,
      "loss": 0.0828,
      "reward": 0.22916666977107525,
      "reward_std": 0.28867512941360474,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 279
    },
    {
      "completion_length": 790.5000152587891,
      "epoch": 0.34146341463414637,
      "grad_norm": 0.6836724877357483,
      "kl": 0.10723876953125,
      "learning_rate": 2.4979885502483478e-06,
      "loss": 0.0502,
      "reward": 0.25000000931322575,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.25000000931322575,
      "rewards/format_reward": 0.0,
      "step": 280
    },
    {
      "completion_length": 770.7083435058594,
      "epoch": 0.3426829268292683,
      "grad_norm": 0.8554387092590332,
      "kl": 0.1102294921875,
      "learning_rate": 2.4932125179809316e-06,
      "loss": 0.0437,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 281
    },
    {
      "completion_length": 795.7083587646484,
      "epoch": 0.3439024390243902,
      "grad_norm": 0.2160601019859314,
      "kl": 0.1451416015625,
      "learning_rate": 2.4884184875351897e-06,
      "loss": 0.0149,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 282
    },
    {
      "completion_length": 797.9166870117188,
      "epoch": 0.34512195121951217,
      "grad_norm": 0.8815224766731262,
      "kl": 0.177734375,
      "learning_rate": 2.48360654578459e-06,
      "loss": 0.0519,
      "reward": 0.1250000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 283
    },
    {
      "completion_length": 728.9583587646484,
      "epoch": 0.3463414634146341,
      "grad_norm": 0.9917912483215332,
      "kl": 0.1558837890625,
      "learning_rate": 2.4787767799271725e-06,
      "loss": 0.0921,
      "reward": 0.16666666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 284
    },
    {
      "completion_length": 729.7708587646484,
      "epoch": 0.3475609756097561,
      "grad_norm": 0.2130976766347885,
      "kl": 0.12139892578125,
      "learning_rate": 2.473929277483972e-06,
      "loss": 0.0123,
      "reward": 0.2083333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 285
    },
    {
      "completion_length": 704.1666717529297,
      "epoch": 0.348780487804878,
      "grad_norm": 0.40542691946029663,
      "kl": 0.24169921875,
      "learning_rate": 2.4690641262974317e-06,
      "loss": 0.0438,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 286
    },
    {
      "completion_length": 683.3958435058594,
      "epoch": 0.35,
      "grad_norm": 0.20354585349559784,
      "kl": 0.124755859375,
      "learning_rate": 2.464181414529809e-06,
      "loss": 0.0049,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 287
    },
    {
      "completion_length": 756.5208435058594,
      "epoch": 0.35121951219512193,
      "grad_norm": 0.33828863501548767,
      "kl": 0.1773681640625,
      "learning_rate": 2.4592812306615812e-06,
      "loss": 0.0403,
      "reward": 0.0625,
      "reward_std": 0.10825318098068237,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 288
    },
    {
      "completion_length": 714.8750305175781,
      "epoch": 0.3524390243902439,
      "grad_norm": 0.46392297744750977,
      "kl": 0.139892578125,
      "learning_rate": 2.4543636634898398e-06,
      "loss": 0.0675,
      "reward": 0.14583333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 289
    },
    {
      "completion_length": 688.0625305175781,
      "epoch": 0.35365853658536583,
      "grad_norm": 0.21563249826431274,
      "kl": 0.09771728515625,
      "learning_rate": 2.4494288021266825e-06,
      "loss": 0.0222,
      "reward": 0.2083333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 290
    },
    {
      "completion_length": 746.9375152587891,
      "epoch": 0.3548780487804878,
      "grad_norm": 0.520687460899353,
      "kl": 0.17626953125,
      "learning_rate": 2.444476735997598e-06,
      "loss": 0.0792,
      "reward": 0.22916667349636555,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.22916667349636555,
      "rewards/format_reward": 0.0,
      "step": 291
    },
    {
      "completion_length": 756.0833435058594,
      "epoch": 0.35609756097560974,
      "grad_norm": 0.6303249597549438,
      "kl": 0.1348876953125,
      "learning_rate": 2.439507554839846e-06,
      "loss": 0.0638,
      "reward": 0.2500000074505806,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 292
    },
    {
      "completion_length": 673.1041870117188,
      "epoch": 0.3573170731707317,
      "grad_norm": 0.3091147840023041,
      "kl": 0.1329345703125,
      "learning_rate": 2.4345213487008296e-06,
      "loss": 0.0262,
      "reward": 0.1458333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 293
    },
    {
      "completion_length": 711.5625152587891,
      "epoch": 0.35853658536585364,
      "grad_norm": 0.33718910813331604,
      "kl": 0.1414794921875,
      "learning_rate": 2.4295182079364655e-06,
      "loss": 0.0217,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 294
    },
    {
      "completion_length": 765.2291870117188,
      "epoch": 0.3597560975609756,
      "grad_norm": 0.6034972071647644,
      "kl": 0.15216064453125,
      "learning_rate": 2.424498223209545e-06,
      "loss": -0.0056,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 295
    },
    {
      "completion_length": 728.7500152587891,
      "epoch": 0.36097560975609755,
      "grad_norm": 0.3447917401790619,
      "kl": 0.1527099609375,
      "learning_rate": 2.4194614854880937e-06,
      "loss": 0.0144,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 296
    },
    {
      "completion_length": 593.8958435058594,
      "epoch": 0.3621951219512195,
      "grad_norm": 0.7357714176177979,
      "kl": 0.09320068359375,
      "learning_rate": 2.4144080860437184e-06,
      "loss": 0.0385,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 297
    },
    {
      "completion_length": 693.7500305175781,
      "epoch": 0.36341463414634145,
      "grad_norm": 0.16408616304397583,
      "kl": 0.1358642578125,
      "learning_rate": 2.409338116449957e-06,
      "loss": 0.0048,
      "reward": 0.1875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 298
    },
    {
      "completion_length": 738.5625305175781,
      "epoch": 0.3646341463414634,
      "grad_norm": 0.31955286860466003,
      "kl": 0.070068359375,
      "learning_rate": 2.404251668580619e-06,
      "loss": 0.0227,
      "reward": 0.3125,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.3125,
      "rewards/format_reward": 0.0,
      "step": 299
    },
    {
      "completion_length": 694.3333587646484,
      "epoch": 0.36585365853658536,
      "grad_norm": 0.4365987479686737,
      "kl": 0.09368896484375,
      "learning_rate": 2.3991488346081183e-06,
      "loss": 0.0237,
      "reward": 0.3333333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3333333395421505,
      "rewards/format_reward": 0.0,
      "step": 300
    },
    {
      "completion_length": 686.6666870117188,
      "epoch": 0.3670731707317073,
      "grad_norm": 0.3122309744358063,
      "kl": 0.08984375,
      "learning_rate": 2.3940297070018048e-06,
      "loss": 0.0048,
      "reward": 0.2291666753590107,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2291666753590107,
      "rewards/format_reward": 0.0,
      "step": 301
    },
    {
      "completion_length": 643.3958587646484,
      "epoch": 0.36829268292682926,
      "grad_norm": 0.18671134114265442,
      "kl": 0.12109375,
      "learning_rate": 2.388894378526288e-06,
      "loss": 0.0082,
      "reward": 0.20833333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 302
    },
    {
      "completion_length": 660.7291870117188,
      "epoch": 0.3695121951219512,
      "grad_norm": 0.41376540064811707,
      "kl": 0.0968017578125,
      "learning_rate": 2.383742942239757e-06,
      "loss": -0.0228,
      "reward": 0.27083334140479565,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.27083334140479565,
      "rewards/format_reward": 0.0,
      "step": 303
    },
    {
      "completion_length": 584.3333511352539,
      "epoch": 0.37073170731707317,
      "grad_norm": 0.4374995827674866,
      "kl": 0.09735107421875,
      "learning_rate": 2.3785754914922923e-06,
      "loss": 0.0912,
      "reward": 0.3125000074505806,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 304
    },
    {
      "completion_length": 657.5208587646484,
      "epoch": 0.3719512195121951,
      "grad_norm": 0.651809811592102,
      "kl": 0.1390380859375,
      "learning_rate": 2.3733921199241755e-06,
      "loss": 0.0569,
      "reward": 0.16666667349636555,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.16666667349636555,
      "rewards/format_reward": 0.0,
      "step": 305
    },
    {
      "completion_length": 736.0833435058594,
      "epoch": 0.37317073170731707,
      "grad_norm": 0.5078076720237732,
      "kl": 0.09942626953125,
      "learning_rate": 2.3681929214641924e-06,
      "loss": 0.0645,
      "reward": 0.20833333767950535,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.20833333767950535,
      "rewards/format_reward": 0.0,
      "step": 306
    },
    {
      "completion_length": 640.3958587646484,
      "epoch": 0.374390243902439,
      "grad_norm": 0.34475183486938477,
      "kl": 0.1053466796875,
      "learning_rate": 2.362977990327931e-06,
      "loss": 0.0228,
      "reward": 0.3125000074505806,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 307
    },
    {
      "completion_length": 732.6458587646484,
      "epoch": 0.375609756097561,
      "grad_norm": 0.30445414781570435,
      "kl": 0.102294921875,
      "learning_rate": 2.357747421016073e-06,
      "loss": -0.0292,
      "reward": 0.1458333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 308
    },
    {
      "completion_length": 673.1458587646484,
      "epoch": 0.37682926829268293,
      "grad_norm": 0.5365584492683411,
      "kl": 0.0986328125,
      "learning_rate": 2.3525013083126835e-06,
      "loss": 0.0223,
      "reward": 0.16666666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 309
    },
    {
      "completion_length": 719.6666870117188,
      "epoch": 0.3780487804878049,
      "grad_norm": 0.2747936546802521,
      "kl": 0.1282958984375,
      "learning_rate": 2.34723974728349e-06,
      "loss": 0.004,
      "reward": 0.2291666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 310
    },
    {
      "completion_length": 763.9375305175781,
      "epoch": 0.37926829268292683,
      "grad_norm": 0.3625332713127136,
      "kl": 0.1343994140625,
      "learning_rate": 2.341962833274165e-06,
      "loss": 0.0037,
      "reward": 0.1250000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 311
    },
    {
      "completion_length": 712.4792022705078,
      "epoch": 0.3804878048780488,
      "grad_norm": 0.4034402072429657,
      "kl": 0.1416015625,
      "learning_rate": 2.336670661908592e-06,
      "loss": 0.0507,
      "reward": 0.1250000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 312
    },
    {
      "completion_length": 723.7083587646484,
      "epoch": 0.38170731707317074,
      "grad_norm": 0.34415197372436523,
      "kl": 0.143798828125,
      "learning_rate": 2.3313633290871373e-06,
      "loss": 0.0143,
      "reward": 0.18750000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 313
    },
    {
      "completion_length": 665.5000152587891,
      "epoch": 0.3829268292682927,
      "grad_norm": 0.35049009323120117,
      "kl": 0.21826171875,
      "learning_rate": 2.3260409309849103e-06,
      "loss": -0.0068,
      "reward": 0.3125000074505806,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 314
    },
    {
      "completion_length": 743.5000152587891,
      "epoch": 0.38414634146341464,
      "grad_norm": 0.33067408204078674,
      "kl": 0.11676025390625,
      "learning_rate": 2.3207035640500206e-06,
      "loss": 0.0205,
      "reward": 0.16666666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 315
    },
    {
      "completion_length": 793.5416870117188,
      "epoch": 0.3853658536585366,
      "grad_norm": 0.3643498420715332,
      "kl": 0.1380615234375,
      "learning_rate": 2.315351325001832e-06,
      "loss": 0.0563,
      "reward": 0.1875000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 316
    },
    {
      "completion_length": 741.0208587646484,
      "epoch": 0.38658536585365855,
      "grad_norm": 0.6269105076789856,
      "kl": 0.136962890625,
      "learning_rate": 2.3099843108292062e-06,
      "loss": 0.0089,
      "reward": 0.31250000558793545,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.31250000558793545,
      "rewards/format_reward": 0.0,
      "step": 317
    },
    {
      "completion_length": 773.5208587646484,
      "epoch": 0.3878048780487805,
      "grad_norm": 0.29589682817459106,
      "kl": 0.1302490234375,
      "learning_rate": 2.3046026187887498e-06,
      "loss": 0.0052,
      "reward": 0.2083333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 318
    },
    {
      "completion_length": 732.4583587646484,
      "epoch": 0.38902439024390245,
      "grad_norm": 0.23372584581375122,
      "kl": 0.154541015625,
      "learning_rate": 2.2992063464030482e-06,
      "loss": 0.0021,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 319
    },
    {
      "completion_length": 706.5625305175781,
      "epoch": 0.3902439024390244,
      "grad_norm": 0.49740323424339294,
      "kl": 0.1129150390625,
      "learning_rate": 2.293795591458901e-06,
      "loss": 0.0621,
      "reward": 0.20833334140479565,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.20833334140479565,
      "rewards/format_reward": 0.0,
      "step": 320
    },
    {
      "completion_length": 709.5625152587891,
      "epoch": 0.39146341463414636,
      "grad_norm": 0.5322088599205017,
      "kl": 0.1124267578125,
      "learning_rate": 2.288370452005547e-06,
      "loss": 0.0398,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 321
    },
    {
      "completion_length": 664.4791870117188,
      "epoch": 0.3926829268292683,
      "grad_norm": 0.2641255557537079,
      "kl": 0.0849609375,
      "learning_rate": 2.2829310263528907e-06,
      "loss": 0.0154,
      "reward": 0.29166667722165585,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.29166667722165585,
      "rewards/format_reward": 0.0,
      "step": 322
    },
    {
      "completion_length": 692.8541870117188,
      "epoch": 0.39390243902439026,
      "grad_norm": 0.5491902232170105,
      "kl": 0.119140625,
      "learning_rate": 2.2774774130697184e-06,
      "loss": 0.0434,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 323
    },
    {
      "completion_length": 682.6875152587891,
      "epoch": 0.3951219512195122,
      "grad_norm": 0.47783124446868896,
      "kl": 0.1650390625,
      "learning_rate": 2.2720097109819135e-06,
      "loss": -0.0201,
      "reward": 0.3333333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3333333432674408,
      "rewards/format_reward": 0.0,
      "step": 324
    },
    {
      "completion_length": 664.7291870117188,
      "epoch": 0.39634146341463417,
      "grad_norm": 0.6119130849838257,
      "kl": 0.136474609375,
      "learning_rate": 2.2665280191706656e-06,
      "loss": 0.0312,
      "reward": 0.27083334513008595,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.27083334513008595,
      "rewards/format_reward": 0.0,
      "step": 325
    },
    {
      "completion_length": 666.3541870117188,
      "epoch": 0.3975609756097561,
      "grad_norm": 1.065935730934143,
      "kl": 0.23486328125,
      "learning_rate": 2.2610324369706735e-06,
      "loss": 0.0298,
      "reward": 0.2500000037252903,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 326
    },
    {
      "completion_length": 621.3750305175781,
      "epoch": 0.39878048780487807,
      "grad_norm": 0.5678831934928894,
      "kl": 0.217498779296875,
      "learning_rate": 2.2555230639683464e-06,
      "loss": 0.012,
      "reward": 0.3541666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.3541666679084301,
      "rewards/format_reward": 0.0,
      "step": 327
    },
    {
      "completion_length": 640.1666870117188,
      "epoch": 0.4,
      "grad_norm": 0.7840041518211365,
      "kl": 0.300537109375,
      "learning_rate": 2.25e-06,
      "loss": 0.028,
      "reward": 0.2083333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 328
    },
    {
      "completion_length": 615.6875152587891,
      "epoch": 0.401219512195122,
      "grad_norm": 0.6310785412788391,
      "kl": 0.170654296875,
      "learning_rate": 2.2444633451500453e-06,
      "loss": 0.029,
      "reward": 0.3125000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3125000037252903,
      "rewards/format_reward": 0.0,
      "step": 329
    },
    {
      "completion_length": 596.6250305175781,
      "epoch": 0.4024390243902439,
      "grad_norm": 0.4667941927909851,
      "kl": 0.24072265625,
      "learning_rate": 2.2389131997491756e-06,
      "loss": 0.0102,
      "reward": 0.1875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 330
    },
    {
      "completion_length": 633.1666870117188,
      "epoch": 0.4036585365853659,
      "grad_norm": 0.5117998123168945,
      "kl": 0.29833984375,
      "learning_rate": 2.2333496643725505e-06,
      "loss": 0.0102,
      "reward": 0.2083333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 331
    },
    {
      "completion_length": 683.4583587646484,
      "epoch": 0.40487804878048783,
      "grad_norm": 0.2802461087703705,
      "kl": 0.1968994140625,
      "learning_rate": 2.2277728398379705e-06,
      "loss": 0.0068,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 332
    },
    {
      "completion_length": 676.5416870117188,
      "epoch": 0.4060975609756098,
      "grad_norm": 0.42859187722206116,
      "kl": 0.257568359375,
      "learning_rate": 2.2221828272040517e-06,
      "loss": 0.0141,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 333
    },
    {
      "completion_length": 679.9375152587891,
      "epoch": 0.4073170731707317,
      "grad_norm": 0.5404031872749329,
      "kl": 0.205810546875,
      "learning_rate": 2.2165797277683943e-06,
      "loss": 0.0633,
      "reward": 0.1875000111758709,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875000111758709,
      "rewards/format_reward": 0.0,
      "step": 334
    },
    {
      "completion_length": 781.7708435058594,
      "epoch": 0.40853658536585363,
      "grad_norm": 0.42941033840179443,
      "kl": 0.154052734375,
      "learning_rate": 2.2109636430657463e-06,
      "loss": 0.0279,
      "reward": 0.125,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 335
    },
    {
      "completion_length": 666.1041870117188,
      "epoch": 0.4097560975609756,
      "grad_norm": 0.2462126612663269,
      "kl": 0.12481689453125,
      "learning_rate": 2.2053346748661633e-06,
      "loss": 0.0161,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 336
    },
    {
      "completion_length": 750.0416870117188,
      "epoch": 0.41097560975609754,
      "grad_norm": 0.3934653699398041,
      "kl": 0.1331787109375,
      "learning_rate": 2.1996929251731665e-06,
      "loss": 0.0587,
      "reward": 0.27083334513008595,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.27083334513008595,
      "rewards/format_reward": 0.0,
      "step": 337
    },
    {
      "completion_length": 733.0416870117188,
      "epoch": 0.4121951219512195,
      "grad_norm": 0.3684786856174469,
      "kl": 0.1005859375,
      "learning_rate": 2.194038496221892e-06,
      "loss": 0.0455,
      "reward": 0.1875,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 338
    },
    {
      "completion_length": 785.0833435058594,
      "epoch": 0.41341463414634144,
      "grad_norm": 0.5052400231361389,
      "kl": 0.2265625,
      "learning_rate": 2.188371490477239e-06,
      "loss": 0.0143,
      "reward": 0.1666666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 339
    },
    {
      "completion_length": 819.2916870117188,
      "epoch": 0.4146341463414634,
      "grad_norm": 0.16904811561107635,
      "kl": 0.1165771484375,
      "learning_rate": 2.182692010632013e-06,
      "loss": 0.0265,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 340
    },
    {
      "completion_length": 833.1458587646484,
      "epoch": 0.41585365853658535,
      "grad_norm": 0.5802979469299316,
      "kl": 0.489501953125,
      "learning_rate": 2.177000159605065e-06,
      "loss": 0.0306,
      "reward": 0.1666666716337204,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 341
    },
    {
      "completion_length": 842.9583435058594,
      "epoch": 0.4170731707317073,
      "grad_norm": 0.3086927533149719,
      "kl": 0.15679931640625,
      "learning_rate": 2.1712960405394265e-06,
      "loss": 0.0353,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 342
    },
    {
      "completion_length": 811.6458587646484,
      "epoch": 0.41829268292682925,
      "grad_norm": 0.3507758677005768,
      "kl": 0.172119140625,
      "learning_rate": 2.1655797568004397e-06,
      "loss": 0.016,
      "reward": 0.27083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 343
    },
    {
      "completion_length": 721.6666870117188,
      "epoch": 0.4195121951219512,
      "grad_norm": 0.4427453577518463,
      "kl": 0.21728515625,
      "learning_rate": 2.1598514119738853e-06,
      "loss": 0.0388,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 344
    },
    {
      "completion_length": 800.8333587646484,
      "epoch": 0.42073170731707316,
      "grad_norm": 0.3072930574417114,
      "kl": 0.1744384765625,
      "learning_rate": 2.154111109864105e-06,
      "loss": 0.0075,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 345
    },
    {
      "completion_length": 796.1875305175781,
      "epoch": 0.4219512195121951,
      "grad_norm": 0.3828490972518921,
      "kl": 0.198974609375,
      "learning_rate": 2.1483589544921202e-06,
      "loss": 0.0087,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 346
    },
    {
      "completion_length": 728.1666870117188,
      "epoch": 0.42317073170731706,
      "grad_norm": 0.6082723140716553,
      "kl": 0.2261962890625,
      "learning_rate": 2.1425950500937493e-06,
      "loss": 0.0394,
      "reward": 0.08333333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 347
    },
    {
      "completion_length": 668.1875305175781,
      "epoch": 0.424390243902439,
      "grad_norm": 0.2896507978439331,
      "kl": 0.09716796875,
      "learning_rate": 2.1368195011177142e-06,
      "loss": 0.0106,
      "reward": 0.2708333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 348
    },
    {
      "completion_length": 746.9375305175781,
      "epoch": 0.42560975609756097,
      "grad_norm": 0.35949307680130005,
      "kl": 0.155517578125,
      "learning_rate": 2.1310324122237512e-06,
      "loss": 0.0223,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 349
    },
    {
      "completion_length": 629.3125305175781,
      "epoch": 0.4268292682926829,
      "grad_norm": 0.6106564998626709,
      "kl": 0.15380859375,
      "learning_rate": 2.125233888280715e-06,
      "loss": -0.0147,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 350
    },
    {
      "completion_length": 612.6041717529297,
      "epoch": 0.42804878048780487,
      "grad_norm": 15.078311920166016,
      "kl": 0.131439208984375,
      "learning_rate": 2.1194240343646732e-06,
      "loss": 0.0087,
      "reward": 0.22916667722165585,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.22916667722165585,
      "rewards/format_reward": 0.0,
      "step": 351
    },
    {
      "completion_length": 717.1250152587891,
      "epoch": 0.4292682926829268,
      "grad_norm": 0.6894350051879883,
      "kl": 0.10986328125,
      "learning_rate": 2.11360295575701e-06,
      "loss": -0.0302,
      "reward": 0.1666666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 352
    },
    {
      "completion_length": 721.9583587646484,
      "epoch": 0.4304878048780488,
      "grad_norm": 0.33386728167533875,
      "kl": 0.096435546875,
      "learning_rate": 2.1077707579425114e-06,
      "loss": 0.0203,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 353
    },
    {
      "completion_length": 747.2291870117188,
      "epoch": 0.4317073170731707,
      "grad_norm": 0.28224602341651917,
      "kl": 0.0738525390625,
      "learning_rate": 2.1019275466074585e-06,
      "loss": 0.0226,
      "reward": 0.0833333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 354
    },
    {
      "completion_length": 772.8541717529297,
      "epoch": 0.4329268292682927,
      "grad_norm": 0.33626002073287964,
      "kl": 0.0682373046875,
      "learning_rate": 2.0960734276377082e-06,
      "loss": 0.0396,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 355
    },
    {
      "completion_length": 678.0000152587891,
      "epoch": 0.43414634146341463,
      "grad_norm": 0.34307798743247986,
      "kl": 0.06781005859375,
      "learning_rate": 2.0902085071167774e-06,
      "loss": 0.0353,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 356
    },
    {
      "completion_length": 624.4375152587891,
      "epoch": 0.4353658536585366,
      "grad_norm": 0.5508283972740173,
      "kl": 0.1077880859375,
      "learning_rate": 2.0843328913239216e-06,
      "loss": -0.0028,
      "reward": 0.2708333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 357
    },
    {
      "completion_length": 639.1666870117188,
      "epoch": 0.43658536585365854,
      "grad_norm": 0.5479058623313904,
      "kl": 0.1180419921875,
      "learning_rate": 2.0784466867322037e-06,
      "loss": 0.0016,
      "reward": 0.2708333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 358
    },
    {
      "completion_length": 754.5833587646484,
      "epoch": 0.4378048780487805,
      "grad_norm": 0.3993290364742279,
      "kl": 0.09326171875,
      "learning_rate": 2.0725500000065715e-06,
      "loss": 0.0306,
      "reward": 0.1666666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 359
    },
    {
      "completion_length": 647.5000076293945,
      "epoch": 0.43902439024390244,
      "grad_norm": 0.5099925994873047,
      "kl": 0.126953125,
      "learning_rate": 2.0666429380019185e-06,
      "loss": -0.008,
      "reward": 0.14583333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 360
    },
    {
      "completion_length": 676.3958587646484,
      "epoch": 0.4402439024390244,
      "grad_norm": 0.7529412508010864,
      "kl": 0.1370849609375,
      "learning_rate": 2.060725607761153e-06,
      "loss": 0.0198,
      "reward": 0.2291666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 361
    },
    {
      "completion_length": 697.0833587646484,
      "epoch": 0.44146341463414634,
      "grad_norm": 0.4546225070953369,
      "kl": 0.13250732421875,
      "learning_rate": 2.0547981165132547e-06,
      "loss": 0.032,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 362
    },
    {
      "completion_length": 652.2916870117188,
      "epoch": 0.4426829268292683,
      "grad_norm": 0.6514831185340881,
      "kl": 0.206787109375,
      "learning_rate": 2.048860571671332e-06,
      "loss": -0.0108,
      "reward": 0.2291666753590107,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666753590107,
      "rewards/format_reward": 0.0,
      "step": 363
    },
    {
      "completion_length": 750.5416870117188,
      "epoch": 0.44390243902439025,
      "grad_norm": 0.21455013751983643,
      "kl": 0.158447265625,
      "learning_rate": 2.0429130808306767e-06,
      "loss": 0.0189,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 364
    },
    {
      "completion_length": 698.8125152587891,
      "epoch": 0.4451219512195122,
      "grad_norm": 0.5055814385414124,
      "kl": 0.2158203125,
      "learning_rate": 2.036955751766815e-06,
      "loss": -0.0251,
      "reward": 0.33333334513008595,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.33333334513008595,
      "rewards/format_reward": 0.0,
      "step": 365
    },
    {
      "completion_length": 804.2917022705078,
      "epoch": 0.44634146341463415,
      "grad_norm": 0.7169243097305298,
      "kl": 0.1954345703125,
      "learning_rate": 2.030988692433552e-06,
      "loss": 0.0025,
      "reward": 0.12500000186264515,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 366
    },
    {
      "completion_length": 698.5000305175781,
      "epoch": 0.4475609756097561,
      "grad_norm": 0.46428942680358887,
      "kl": 0.251953125,
      "learning_rate": 2.0250120109610155e-06,
      "loss": -0.0139,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 367
    },
    {
      "completion_length": 727.8541870117188,
      "epoch": 0.44878048780487806,
      "grad_norm": 0.5507683157920837,
      "kl": 0.1495361328125,
      "learning_rate": 2.019025815653701e-06,
      "loss": 0.056,
      "reward": 0.2708333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 368
    },
    {
      "completion_length": 787.9583435058594,
      "epoch": 0.45,
      "grad_norm": 0.8070011138916016,
      "kl": 0.291015625,
      "learning_rate": 2.0130302149885033e-06,
      "loss": 0.0307,
      "reward": 0.1458333358168602,
      "reward_std": 0.21650635451078415,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 369
    },
    {
      "completion_length": 760.0000152587891,
      "epoch": 0.45121951219512196,
      "grad_norm": 1.140531063079834,
      "kl": 0.31201171875,
      "learning_rate": 2.007025317612754e-06,
      "loss": 0.071,
      "reward": 0.2708333432674408,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 370
    },
    {
      "completion_length": 818.1250152587891,
      "epoch": 0.4524390243902439,
      "grad_norm": 1.2991398572921753,
      "kl": 0.506591796875,
      "learning_rate": 2.001011232342253e-06,
      "loss": 0.0218,
      "reward": 0.25,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.0,
      "step": 371
    },
    {
      "completion_length": 791.0000305175781,
      "epoch": 0.45365853658536587,
      "grad_norm": 2.349998712539673,
      "kl": 0.44775390625,
      "learning_rate": 1.994988068159294e-06,
      "loss": 0.1124,
      "reward": 0.27083334140479565,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.27083334140479565,
      "rewards/format_reward": 0.0,
      "step": 372
    },
    {
      "completion_length": 795.8125152587891,
      "epoch": 0.4548780487804878,
      "grad_norm": 3.5493900775909424,
      "kl": 1.0751953125,
      "learning_rate": 1.9889559342106926e-06,
      "loss": 0.0556,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 373
    },
    {
      "completion_length": 754.9583587646484,
      "epoch": 0.4560975609756098,
      "grad_norm": 2.3599045276641846,
      "kl": 0.70361328125,
      "learning_rate": 1.9829149398058068e-06,
      "loss": 0.0251,
      "reward": 0.2916666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 374
    },
    {
      "completion_length": 745.1041870117188,
      "epoch": 0.4573170731707317,
      "grad_norm": 1.07171630859375,
      "kl": 0.6044921875,
      "learning_rate": 1.976865194414555e-06,
      "loss": 0.0499,
      "reward": 0.1666666716337204,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 375
    },
    {
      "completion_length": 747.3125152587891,
      "epoch": 0.4585365853658537,
      "grad_norm": 0.6529096961021423,
      "kl": 0.49462890625,
      "learning_rate": 1.9708068076654364e-06,
      "loss": 0.0276,
      "reward": 0.2083333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333432674408,
      "rewards/format_reward": 0.0,
      "step": 376
    },
    {
      "completion_length": 771.2708587646484,
      "epoch": 0.45975609756097563,
      "grad_norm": 1.0113192796707153,
      "kl": 0.7041015625,
      "learning_rate": 1.9647398893435394e-06,
      "loss": 0.0246,
      "reward": 0.2500000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 377
    },
    {
      "completion_length": 767.6875152587891,
      "epoch": 0.4609756097560976,
      "grad_norm": 0.7399620413780212,
      "kl": 0.38330078125,
      "learning_rate": 1.9586645493885565e-06,
      "loss": 0.0224,
      "reward": 0.1875000074505806,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875000074505806,
      "rewards/format_reward": 0.0,
      "step": 378
    },
    {
      "completion_length": 695.4166870117188,
      "epoch": 0.46219512195121953,
      "grad_norm": 1.2159936428070068,
      "kl": 0.2763671875,
      "learning_rate": 1.9525808978927886e-06,
      "loss": 0.0397,
      "reward": 0.2291666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 379
    },
    {
      "completion_length": 797.3750305175781,
      "epoch": 0.4634146341463415,
      "grad_norm": 0.7549745440483093,
      "kl": 0.289306640625,
      "learning_rate": 1.946489045099152e-06,
      "loss": 0.0037,
      "reward": 0.1458333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 380
    },
    {
      "completion_length": 724.3958511352539,
      "epoch": 0.46463414634146344,
      "grad_norm": 1.1805777549743652,
      "kl": 0.513916015625,
      "learning_rate": 1.94038910139918e-06,
      "loss": 0.0383,
      "reward": 0.2708333395421505,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 381
    },
    {
      "completion_length": 721.7500152587891,
      "epoch": 0.4658536585365854,
      "grad_norm": 1.0325050354003906,
      "kl": 0.3204345703125,
      "learning_rate": 1.934281177331023e-06,
      "loss": 0.0466,
      "reward": 0.2291666716337204,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 382
    },
    {
      "completion_length": 734.6250152587891,
      "epoch": 0.46707317073170734,
      "grad_norm": 1.5376849174499512,
      "kl": 0.26513671875,
      "learning_rate": 1.928165383577445e-06,
      "loss": 0.0159,
      "reward": 0.20833333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.20833333767950535,
      "rewards/format_reward": 0.0,
      "step": 383
    },
    {
      "completion_length": 734.4583435058594,
      "epoch": 0.4682926829268293,
      "grad_norm": 0.4632957875728607,
      "kl": 0.13787841796875,
      "learning_rate": 1.9220418309638175e-06,
      "loss": 0.0171,
      "reward": 0.3125000074505806,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 384
    },
    {
      "completion_length": 639.8125152587891,
      "epoch": 0.4695121951219512,
      "grad_norm": 5.395277500152588,
      "kl": 0.377197265625,
      "learning_rate": 1.915910630456112e-06,
      "loss": 0.093,
      "reward": 0.18750000558793545,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 385
    },
    {
      "completion_length": 768.8958587646484,
      "epoch": 0.47073170731707314,
      "grad_norm": 5.406608581542969,
      "kl": 0.604736328125,
      "learning_rate": 1.909771893158889e-06,
      "loss": 0.0642,
      "reward": 0.1250000037252903,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 386
    },
    {
      "completion_length": 819.0625152587891,
      "epoch": 0.4719512195121951,
      "grad_norm": 2.1309621334075928,
      "kl": 0.806640625,
      "learning_rate": 1.9036257303132843e-06,
      "loss": 0.0717,
      "reward": 0.3333333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3333333358168602,
      "rewards/format_reward": 0.0,
      "step": 387
    },
    {
      "completion_length": 755.1875152587891,
      "epoch": 0.47317073170731705,
      "grad_norm": 1.5358734130859375,
      "kl": 0.800537109375,
      "learning_rate": 1.8974722532949929e-06,
      "loss": 0.0609,
      "reward": 0.22916666977107525,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 388
    },
    {
      "completion_length": 707.6250152587891,
      "epoch": 0.474390243902439,
      "grad_norm": 4.574571132659912,
      "kl": 0.9180908203125,
      "learning_rate": 1.8913115736122519e-06,
      "loss": 0.1522,
      "reward": 0.4166666716337204,
      "reward_std": 0.25259073823690414,
      "rewards/accuracy_reward": 0.4166666716337204,
      "rewards/format_reward": 0.0,
      "step": 389
    },
    {
      "completion_length": 856.3125152587891,
      "epoch": 0.47560975609756095,
      "grad_norm": 5.587050437927246,
      "kl": 1.64892578125,
      "learning_rate": 1.8851438029038191e-06,
      "loss": 0.1046,
      "reward": 0.2500000149011612,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 0.0,
      "step": 390
    },
    {
      "completion_length": 785.7708435058594,
      "epoch": 0.4768292682926829,
      "grad_norm": 1.2371622323989868,
      "kl": 0.8125,
      "learning_rate": 1.8789690529369492e-06,
      "loss": 0.0551,
      "reward": 0.37500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.0,
      "step": 391
    },
    {
      "completion_length": 876.8333587646484,
      "epoch": 0.47804878048780486,
      "grad_norm": 4.228133678436279,
      "kl": 1.3056640625,
      "learning_rate": 1.8727874356053706e-06,
      "loss": 0.049,
      "reward": 0.1875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 392
    },
    {
      "completion_length": 814.1458740234375,
      "epoch": 0.4792682926829268,
      "grad_norm": 0.6163298487663269,
      "kl": 0.67626953125,
      "learning_rate": 1.8665990629272555e-06,
      "loss": 0.0514,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 393
    },
    {
      "completion_length": 911.8333587646484,
      "epoch": 0.48048780487804876,
      "grad_norm": 0.9791918992996216,
      "kl": 0.5029296875,
      "learning_rate": 1.8604040470431908e-06,
      "loss": 0.0341,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 394
    },
    {
      "completion_length": 803.2500152587891,
      "epoch": 0.4817073170731707,
      "grad_norm": 0.4017774164676666,
      "kl": 0.25439453125,
      "learning_rate": 1.8542025002141474e-06,
      "loss": 0.026,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 395
    },
    {
      "completion_length": 844.6041870117188,
      "epoch": 0.48292682926829267,
      "grad_norm": 0.2321510910987854,
      "kl": 0.2178955078125,
      "learning_rate": 1.8479945348194423e-06,
      "loss": 0.0085,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 396
    },
    {
      "completion_length": 833.7916870117188,
      "epoch": 0.4841463414634146,
      "grad_norm": 0.2628988027572632,
      "kl": 0.1702880859375,
      "learning_rate": 1.8417802633547067e-06,
      "loss": -0.0006,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 397
    },
    {
      "completion_length": 797.4791870117188,
      "epoch": 0.4853658536585366,
      "grad_norm": 0.2977566421031952,
      "kl": 0.192138671875,
      "learning_rate": 1.8355597984298435e-06,
      "loss": 0.0037,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 398
    },
    {
      "completion_length": 883.6458435058594,
      "epoch": 0.4865853658536585,
      "grad_norm": 0.34783926606178284,
      "kl": 0.1744384765625,
      "learning_rate": 1.8293332527669897e-06,
      "loss": 0.0075,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 399
    },
    {
      "completion_length": 786.5208587646484,
      "epoch": 0.4878048780487805,
      "grad_norm": 0.262679785490036,
      "kl": 0.1478271484375,
      "learning_rate": 1.823100739198472e-06,
      "loss": 0.0362,
      "reward": 0.0833333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 400
    },
    {
      "completion_length": 850.3333435058594,
      "epoch": 0.48902439024390243,
      "grad_norm": 0.4966966509819031,
      "kl": 0.1131591796875,
      "learning_rate": 1.816862370664762e-06,
      "loss": 0.042,
      "reward": 0.18750000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 401
    },
    {
      "completion_length": 797.5208435058594,
      "epoch": 0.4902439024390244,
      "grad_norm": 0.2827613353729248,
      "kl": 0.0782470703125,
      "learning_rate": 1.8106182602124312e-06,
      "loss": 0.0408,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 402
    },
    {
      "completion_length": 818.7916870117188,
      "epoch": 0.49146341463414633,
      "grad_norm": 0.15073873102664948,
      "kl": 0.0850830078125,
      "learning_rate": 1.8043685209921002e-06,
      "loss": 0.0247,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 403
    },
    {
      "completion_length": 858.2083587646484,
      "epoch": 0.4926829268292683,
      "grad_norm": 0.4412778913974762,
      "kl": 0.082763671875,
      "learning_rate": 1.7981132662563906e-06,
      "loss": 0.0664,
      "reward": 0.1875000111758709,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.1875000111758709,
      "rewards/format_reward": 0.0,
      "step": 404
    },
    {
      "completion_length": 901.2083587646484,
      "epoch": 0.49390243902439024,
      "grad_norm": 0.15052859485149384,
      "kl": 0.07196044921875,
      "learning_rate": 1.7918526093578702e-06,
      "loss": 0.0036,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 405
    },
    {
      "completion_length": 882.8958587646484,
      "epoch": 0.4951219512195122,
      "grad_norm": 0.12602007389068604,
      "kl": 0.06585693359375,
      "learning_rate": 1.7855866637470027e-06,
      "loss": 0.0169,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 406
    },
    {
      "completion_length": 797.5000152587891,
      "epoch": 0.49634146341463414,
      "grad_norm": 0.37057146430015564,
      "kl": 0.07794189453125,
      "learning_rate": 1.7793155429700868e-06,
      "loss": 0.0239,
      "reward": 0.125,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 407
    },
    {
      "completion_length": 881.1667022705078,
      "epoch": 0.4975609756097561,
      "grad_norm": 0.22505596280097961,
      "kl": 0.06549072265625,
      "learning_rate": 1.7730393606672033e-06,
      "loss": 0.0026,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 408
    },
    {
      "completion_length": 784.9167022705078,
      "epoch": 0.49878048780487805,
      "grad_norm": 0.1624797284603119,
      "kl": 0.08319091796875,
      "learning_rate": 1.7667582305701528e-06,
      "loss": 0.0123,
      "reward": 0.2708333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 409
    },
    {
      "completion_length": 804.1458587646484,
      "epoch": 0.5,
      "grad_norm": 0.4143712818622589,
      "kl": 0.1273193359375,
      "learning_rate": 1.7604722665003958e-06,
      "loss": -0.0067,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 410
    },
    {
      "completion_length": 797.1458435058594,
      "epoch": 0.501219512195122,
      "grad_norm": 0.2382408231496811,
      "kl": 0.076171875,
      "learning_rate": 1.7541815823669903e-06,
      "loss": 0.0184,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 411
    },
    {
      "completion_length": 775.3125305175781,
      "epoch": 0.5024390243902439,
      "grad_norm": 0.2505840063095093,
      "kl": 0.06640625,
      "learning_rate": 1.7478862921645273e-06,
      "loss": 0.029,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 412
    },
    {
      "completion_length": 778.6875152587891,
      "epoch": 0.5036585365853659,
      "grad_norm": 0.2288188338279724,
      "kl": 0.08258056640625,
      "learning_rate": 1.7415865099710657e-06,
      "loss": 0.0368,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 413
    },
    {
      "completion_length": 757.6250152587891,
      "epoch": 0.5048780487804878,
      "grad_norm": 0.2969276010990143,
      "kl": 0.106689453125,
      "learning_rate": 1.735282349946064e-06,
      "loss": 0.0401,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 414
    },
    {
      "completion_length": 799.9375305175781,
      "epoch": 0.5060975609756098,
      "grad_norm": 0.39082932472229004,
      "kl": 0.08990478515625,
      "learning_rate": 1.7289739263283118e-06,
      "loss": 0.0154,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 415
    },
    {
      "completion_length": 793.1458587646484,
      "epoch": 0.5073170731707317,
      "grad_norm": 0.11890969425439835,
      "kl": 0.0556640625,
      "learning_rate": 1.7226613534338608e-06,
      "loss": 0.002,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 416
    },
    {
      "completion_length": 772.0625305175781,
      "epoch": 0.5085365853658537,
      "grad_norm": 1.2034021615982056,
      "kl": 0.1575927734375,
      "learning_rate": 1.716344745653952e-06,
      "loss": 0.0027,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 417
    },
    {
      "completion_length": 703.2083435058594,
      "epoch": 0.5097560975609756,
      "grad_norm": 0.18007124960422516,
      "kl": 0.0687255859375,
      "learning_rate": 1.7100242174529439e-06,
      "loss": 0.0219,
      "reward": 0.1666666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 418
    },
    {
      "completion_length": 878.1875,
      "epoch": 0.5109756097560976,
      "grad_norm": 0.054969798773527145,
      "kl": 0.0531005859375,
      "learning_rate": 1.7036998833662359e-06,
      "loss": 0.0021,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 419
    },
    {
      "completion_length": 764.4583435058594,
      "epoch": 0.5121951219512195,
      "grad_norm": 0.17391566932201385,
      "kl": 0.05712890625,
      "learning_rate": 1.6973718579981973e-06,
      "loss": 0.0022,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 420
    },
    {
      "completion_length": 735.4583587646484,
      "epoch": 0.5134146341463415,
      "grad_norm": 0.25800108909606934,
      "kl": 0.10198974609375,
      "learning_rate": 1.6910402560200854e-06,
      "loss": -0.0086,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 421
    },
    {
      "completion_length": 817.0833587646484,
      "epoch": 0.5146341463414634,
      "grad_norm": 0.1550700068473816,
      "kl": 0.07373046875,
      "learning_rate": 1.6847051921679702e-06,
      "loss": 0.0077,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 422
    },
    {
      "completion_length": 761.0416870117188,
      "epoch": 0.5158536585365854,
      "grad_norm": 0.37330228090286255,
      "kl": 0.060546875,
      "learning_rate": 1.6783667812406569e-06,
      "loss": 0.0109,
      "reward": 0.2500000074505806,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 423
    },
    {
      "completion_length": 811.5833435058594,
      "epoch": 0.5170731707317073,
      "grad_norm": 0.239409938454628,
      "kl": 0.04669189453125,
      "learning_rate": 1.672025138097601e-06,
      "loss": 0.0248,
      "reward": 0.1458333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 424
    },
    {
      "completion_length": 766.0625305175781,
      "epoch": 0.5182926829268293,
      "grad_norm": 0.359465092420578,
      "kl": 0.07049560546875,
      "learning_rate": 1.6656803776568307e-06,
      "loss": 0.0046,
      "reward": 0.31250000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.31250000558793545,
      "rewards/format_reward": 0.0,
      "step": 425
    },
    {
      "completion_length": 752.1666870117188,
      "epoch": 0.5195121951219512,
      "grad_norm": 0.26716890931129456,
      "kl": 0.0809326171875,
      "learning_rate": 1.6593326148928643e-06,
      "loss": 0.0265,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 426
    },
    {
      "completion_length": 841.3541870117188,
      "epoch": 0.5207317073170732,
      "grad_norm": 0.22908422350883484,
      "kl": 0.0615234375,
      "learning_rate": 1.652981964834623e-06,
      "loss": 0.0241,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 427
    },
    {
      "completion_length": 808.6666870117188,
      "epoch": 0.5219512195121951,
      "grad_norm": 0.40149325132369995,
      "kl": 0.0567626953125,
      "learning_rate": 1.6466285425633527e-06,
      "loss": 0.073,
      "reward": 0.2916666716337204,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2916666716337204,
      "rewards/format_reward": 0.0,
      "step": 428
    },
    {
      "completion_length": 733.6666870117188,
      "epoch": 0.5231707317073171,
      "grad_norm": 0.16452831029891968,
      "kl": 0.0648193359375,
      "learning_rate": 1.6402724632105323e-06,
      "loss": 0.0229,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 429
    },
    {
      "completion_length": 758.2708587646484,
      "epoch": 0.524390243902439,
      "grad_norm": 0.3680984377861023,
      "kl": 0.0736083984375,
      "learning_rate": 1.6339138419557916e-06,
      "loss": 0.0409,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 430
    },
    {
      "completion_length": 707.8333740234375,
      "epoch": 0.525609756097561,
      "grad_norm": 0.2060546576976776,
      "kl": 0.07415771484375,
      "learning_rate": 1.6275527940248218e-06,
      "loss": 0.0024,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 431
    },
    {
      "completion_length": 787.2917022705078,
      "epoch": 0.526829268292683,
      "grad_norm": 0.35094255208969116,
      "kl": 0.06036376953125,
      "learning_rate": 1.6211894346872887e-06,
      "loss": 0.0189,
      "reward": 0.1875000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 432
    },
    {
      "completion_length": 729.5000305175781,
      "epoch": 0.5280487804878049,
      "grad_norm": 0.37806692719459534,
      "kl": 0.0728759765625,
      "learning_rate": 1.614823879254744e-06,
      "loss": 0.015,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 433
    },
    {
      "completion_length": 797.8958587646484,
      "epoch": 0.5292682926829269,
      "grad_norm": 0.26446664333343506,
      "kl": 0.08892822265625,
      "learning_rate": 1.6084562430785336e-06,
      "loss": 0.0202,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 434
    },
    {
      "completion_length": 787.2916870117188,
      "epoch": 0.5304878048780488,
      "grad_norm": 0.16212700307369232,
      "kl": 0.04931640625,
      "learning_rate": 1.6020866415477108e-06,
      "loss": 0.002,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 435
    },
    {
      "completion_length": 816.5416870117188,
      "epoch": 0.5317073170731708,
      "grad_norm": 0.17697171866893768,
      "kl": 0.06512451171875,
      "learning_rate": 1.5957151900869425e-06,
      "loss": 0.0187,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 436
    },
    {
      "completion_length": 737.9583587646484,
      "epoch": 0.5329268292682927,
      "grad_norm": 0.4031173586845398,
      "kl": 0.0712890625,
      "learning_rate": 1.5893420041544193e-06,
      "loss": 0.0157,
      "reward": 0.0833333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 437
    },
    {
      "completion_length": 664.5208587646484,
      "epoch": 0.5341463414634147,
      "grad_norm": 0.3498460650444031,
      "kl": 0.07275390625,
      "learning_rate": 1.582967199239761e-06,
      "loss": 0.0381,
      "reward": 0.1875000111758709,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875000111758709,
      "rewards/format_reward": 0.0,
      "step": 438
    },
    {
      "completion_length": 685.9166870117188,
      "epoch": 0.5353658536585366,
      "grad_norm": 0.29588350653648376,
      "kl": 0.05401611328125,
      "learning_rate": 1.5765908908619258e-06,
      "loss": 0.0029,
      "reward": 0.20833333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 439
    },
    {
      "completion_length": 780.4375152587891,
      "epoch": 0.5365853658536586,
      "grad_norm": 0.1834474503993988,
      "kl": 0.0550537109375,
      "learning_rate": 1.5702131945671182e-06,
      "loss": 0.0259,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 440
    },
    {
      "completion_length": 805.9791717529297,
      "epoch": 0.5378048780487805,
      "grad_norm": 0.37172019481658936,
      "kl": 0.0511474609375,
      "learning_rate": 1.5638342259266904e-06,
      "loss": 0.0259,
      "reward": 0.229166679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.229166679084301,
      "rewards/format_reward": 0.0,
      "step": 441
    },
    {
      "completion_length": 769.2500152587891,
      "epoch": 0.5390243902439025,
      "grad_norm": 0.29101791977882385,
      "kl": 0.08740234375,
      "learning_rate": 1.5574541005350532e-06,
      "loss": 0.0185,
      "reward": 0.2083333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 442
    },
    {
      "completion_length": 790.9375305175781,
      "epoch": 0.5402439024390244,
      "grad_norm": 0.3134830892086029,
      "kl": 0.0799560546875,
      "learning_rate": 1.5510729340075781e-06,
      "loss": 0.0472,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 443
    },
    {
      "completion_length": 817.9375152587891,
      "epoch": 0.5414634146341464,
      "grad_norm": 0.20175106823444366,
      "kl": 0.05950927734375,
      "learning_rate": 1.544690841978504e-06,
      "loss": 0.0289,
      "reward": 0.22916666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 444
    },
    {
      "completion_length": 797.4583587646484,
      "epoch": 0.5426829268292683,
      "grad_norm": 0.2536517083644867,
      "kl": 0.0635986328125,
      "learning_rate": 1.5383079400988402e-06,
      "loss": 0.0429,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 445
    },
    {
      "completion_length": 769.6250305175781,
      "epoch": 0.5439024390243903,
      "grad_norm": 0.24389034509658813,
      "kl": 0.041748046875,
      "learning_rate": 1.5319243440342713e-06,
      "loss": 0.0026,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 446
    },
    {
      "completion_length": 729.1250152587891,
      "epoch": 0.5451219512195122,
      "grad_norm": 0.15414156019687653,
      "kl": 0.05645751953125,
      "learning_rate": 1.5255401694630625e-06,
      "loss": 0.0169,
      "reward": 0.2916666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 447
    },
    {
      "completion_length": 732.5833587646484,
      "epoch": 0.5463414634146342,
      "grad_norm": 0.18617714941501617,
      "kl": 0.06451416015625,
      "learning_rate": 1.5191555320739608e-06,
      "loss": 0.0197,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 448
    },
    {
      "completion_length": 707.9166870117188,
      "epoch": 0.5475609756097561,
      "grad_norm": 0.26069146394729614,
      "kl": 0.09326171875,
      "learning_rate": 1.5127705475641014e-06,
      "loss": 0.0271,
      "reward": 0.25,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.0,
      "step": 449
    },
    {
      "completion_length": 781.4375305175781,
      "epoch": 0.5487804878048781,
      "grad_norm": 0.30962687730789185,
      "kl": 0.058349609375,
      "learning_rate": 1.5063853316369081e-06,
      "loss": 0.0303,
      "reward": 0.0833333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 450
    },
    {
      "completion_length": 826.9791870117188,
      "epoch": 0.55,
      "grad_norm": 0.2549695670604706,
      "kl": 0.044921875,
      "learning_rate": 1.5e-06,
      "loss": 0.0012,
      "reward": 0.1875,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 451
    },
    {
      "completion_length": 749.2500152587891,
      "epoch": 0.551219512195122,
      "grad_norm": 0.4739878177642822,
      "kl": 0.094879150390625,
      "learning_rate": 1.4936146683630921e-06,
      "loss": 0.0026,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 452
    },
    {
      "completion_length": 781.7708587646484,
      "epoch": 0.552439024390244,
      "grad_norm": 0.35422784090042114,
      "kl": 0.05743408203125,
      "learning_rate": 1.4872294524358989e-06,
      "loss": 0.0262,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 453
    },
    {
      "completion_length": 799.0000152587891,
      "epoch": 0.5536585365853659,
      "grad_norm": 0.23160593211650848,
      "kl": 0.06402587890625,
      "learning_rate": 1.4808444679260396e-06,
      "loss": 0.0401,
      "reward": 0.29166666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.29166666977107525,
      "rewards/format_reward": 0.0,
      "step": 454
    },
    {
      "completion_length": 741.9166870117188,
      "epoch": 0.5548780487804879,
      "grad_norm": 0.2664400339126587,
      "kl": 0.0750732421875,
      "learning_rate": 1.4744598305369376e-06,
      "loss": 0.0072,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 455
    },
    {
      "completion_length": 787.8958435058594,
      "epoch": 0.5560975609756098,
      "grad_norm": 0.31548717617988586,
      "kl": 0.0703125,
      "learning_rate": 1.4680756559657292e-06,
      "loss": 0.0204,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 456
    },
    {
      "completion_length": 707.7500152587891,
      "epoch": 0.5573170731707318,
      "grad_norm": 0.24562372267246246,
      "kl": 0.0859375,
      "learning_rate": 1.4616920599011603e-06,
      "loss": 0.066,
      "reward": 0.14583333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 457
    },
    {
      "completion_length": 803.5833587646484,
      "epoch": 0.5585365853658537,
      "grad_norm": 0.10695140063762665,
      "kl": 0.06146240234375,
      "learning_rate": 1.4553091580214963e-06,
      "loss": 0.022,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 458
    },
    {
      "completion_length": 791.1666870117188,
      "epoch": 0.5597560975609757,
      "grad_norm": 1.8627831935882568,
      "kl": 0.1583251953125,
      "learning_rate": 1.4489270659924222e-06,
      "loss": 0.0064,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 459
    },
    {
      "completion_length": 722.2500305175781,
      "epoch": 0.5609756097560976,
      "grad_norm": 0.3127201497554779,
      "kl": 0.12152099609375,
      "learning_rate": 1.442545899464947e-06,
      "loss": 0.0593,
      "reward": 0.2916666828095913,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2916666828095913,
      "rewards/format_reward": 0.0,
      "step": 460
    },
    {
      "completion_length": 739.6666717529297,
      "epoch": 0.5621951219512196,
      "grad_norm": 0.32948949933052063,
      "kl": 0.0867919921875,
      "learning_rate": 1.4361657740733103e-06,
      "loss": 0.0307,
      "reward": 0.2083333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 461
    },
    {
      "completion_length": 805.7916870117188,
      "epoch": 0.5634146341463414,
      "grad_norm": 0.40065091848373413,
      "kl": 0.08294677734375,
      "learning_rate": 1.429786805432882e-06,
      "loss": 0.0536,
      "reward": 0.2291666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 462
    },
    {
      "completion_length": 769.0416870117188,
      "epoch": 0.5646341463414634,
      "grad_norm": 0.08529353886842728,
      "kl": 0.0762939453125,
      "learning_rate": 1.4234091091380743e-06,
      "loss": 0.0026,
      "reward": 0.1875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 463
    },
    {
      "completion_length": 744.3541870117188,
      "epoch": 0.5658536585365853,
      "grad_norm": 0.05589934438467026,
      "kl": 0.06561279296875,
      "learning_rate": 1.4170328007602395e-06,
      "loss": 0.0025,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 464
    },
    {
      "completion_length": 809.6250152587891,
      "epoch": 0.5670731707317073,
      "grad_norm": 0.2582624852657318,
      "kl": 0.1107177734375,
      "learning_rate": 1.4106579958455812e-06,
      "loss": 0.0048,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 465
    },
    {
      "completion_length": 851.2916870117188,
      "epoch": 0.5682926829268292,
      "grad_norm": 0.150581493973732,
      "kl": 0.1331787109375,
      "learning_rate": 1.4042848099130574e-06,
      "loss": 0.023,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 466
    },
    {
      "completion_length": 798.0000305175781,
      "epoch": 0.5695121951219512,
      "grad_norm": 0.3805694878101349,
      "kl": 0.2012939453125,
      "learning_rate": 1.3979133584522893e-06,
      "loss": 0.0203,
      "reward": 0.2291666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 467
    },
    {
      "completion_length": 759.9166870117188,
      "epoch": 0.5707317073170731,
      "grad_norm": 0.1965319663286209,
      "kl": 0.140869140625,
      "learning_rate": 1.391543756921467e-06,
      "loss": 0.0212,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 468
    },
    {
      "completion_length": 772.5000152587891,
      "epoch": 0.5719512195121951,
      "grad_norm": 0.22685562074184418,
      "kl": 0.1368408203125,
      "learning_rate": 1.3851761207452565e-06,
      "loss": 0.0527,
      "reward": 0.18750000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 469
    },
    {
      "completion_length": 779.4583587646484,
      "epoch": 0.573170731707317,
      "grad_norm": 0.32247573137283325,
      "kl": 0.0726318359375,
      "learning_rate": 1.3788105653127118e-06,
      "loss": -0.001,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 470
    },
    {
      "completion_length": 715.6875152587891,
      "epoch": 0.574390243902439,
      "grad_norm": 1.6217987537384033,
      "kl": 0.1097412109375,
      "learning_rate": 1.3724472059751785e-06,
      "loss": -0.0118,
      "reward": 0.2500000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 471
    },
    {
      "completion_length": 649.4166870117188,
      "epoch": 0.5756097560975609,
      "grad_norm": 0.563847541809082,
      "kl": 0.1817626953125,
      "learning_rate": 1.3660861580442087e-06,
      "loss": 0.0339,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 472
    },
    {
      "completion_length": 651.2083587646484,
      "epoch": 0.5768292682926829,
      "grad_norm": 0.2899157702922821,
      "kl": 0.11767578125,
      "learning_rate": 1.3597275367894676e-06,
      "loss": -0.0046,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 473
    },
    {
      "completion_length": 788.1666870117188,
      "epoch": 0.5780487804878048,
      "grad_norm": 0.21449121832847595,
      "kl": 0.099609375,
      "learning_rate": 1.3533714574366473e-06,
      "loss": 0.0009,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 474
    },
    {
      "completion_length": 697.9167022705078,
      "epoch": 0.5792682926829268,
      "grad_norm": 1.0301029682159424,
      "kl": 0.1512451171875,
      "learning_rate": 1.3470180351653773e-06,
      "loss": 0.0435,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 475
    },
    {
      "completion_length": 666.3541870117188,
      "epoch": 0.5804878048780487,
      "grad_norm": 0.24609684944152832,
      "kl": 0.1234130859375,
      "learning_rate": 1.3406673851071362e-06,
      "loss": 0.0099,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 476
    },
    {
      "completion_length": 747.3125305175781,
      "epoch": 0.5817073170731707,
      "grad_norm": 0.336141437292099,
      "kl": 0.1011962890625,
      "learning_rate": 1.3343196223431698e-06,
      "loss": 0.0403,
      "reward": 0.27083333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 477
    },
    {
      "completion_length": 739.8541717529297,
      "epoch": 0.5829268292682926,
      "grad_norm": 0.17715033888816833,
      "kl": 0.12158203125,
      "learning_rate": 1.3279748619023995e-06,
      "loss": -0.0141,
      "reward": 0.1250000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 478
    },
    {
      "completion_length": 720.0208587646484,
      "epoch": 0.5841463414634146,
      "grad_norm": 0.1463119238615036,
      "kl": 0.099853515625,
      "learning_rate": 1.3216332187593434e-06,
      "loss": 0.025,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 479
    },
    {
      "completion_length": 728.0000152587891,
      "epoch": 0.5853658536585366,
      "grad_norm": 0.3136350214481354,
      "kl": 0.109375,
      "learning_rate": 1.3152948078320297e-06,
      "loss": 0.0199,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 480
    },
    {
      "completion_length": 571.3958587646484,
      "epoch": 0.5865853658536585,
      "grad_norm": 0.3617924153804779,
      "kl": 0.1380615234375,
      "learning_rate": 1.3089597439799151e-06,
      "loss": -0.0017,
      "reward": 0.29166666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.29166666977107525,
      "rewards/format_reward": 0.0,
      "step": 481
    },
    {
      "completion_length": 700.8750305175781,
      "epoch": 0.5878048780487805,
      "grad_norm": 0.5338773131370544,
      "kl": 0.174072265625,
      "learning_rate": 1.3026281420018034e-06,
      "loss": 0.0548,
      "reward": 0.2291666753590107,
      "reward_std": 0.21650635451078415,
      "rewards/accuracy_reward": 0.2291666753590107,
      "rewards/format_reward": 0.0,
      "step": 482
    },
    {
      "completion_length": 738.7291870117188,
      "epoch": 0.5890243902439024,
      "grad_norm": 0.24385033547878265,
      "kl": 0.20263671875,
      "learning_rate": 1.2963001166337642e-06,
      "loss": 0.0208,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 483
    },
    {
      "completion_length": 815.1250305175781,
      "epoch": 0.5902439024390244,
      "grad_norm": 0.3287403881549835,
      "kl": 0.07940673828125,
      "learning_rate": 1.2899757825470568e-06,
      "loss": 0.0162,
      "reward": 0.14583333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 484
    },
    {
      "completion_length": 693.375,
      "epoch": 0.5914634146341463,
      "grad_norm": 0.32487422227859497,
      "kl": 0.10595703125,
      "learning_rate": 1.283655254346048e-06,
      "loss": 0.0409,
      "reward": 0.1250000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 485
    },
    {
      "completion_length": 780.4375305175781,
      "epoch": 0.5926829268292683,
      "grad_norm": 0.1882852166891098,
      "kl": 0.0699462890625,
      "learning_rate": 1.2773386465661395e-06,
      "loss": 0.0041,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 486
    },
    {
      "completion_length": 722.3958587646484,
      "epoch": 0.5939024390243902,
      "grad_norm": 0.392995148897171,
      "kl": 0.1416015625,
      "learning_rate": 1.2710260736716882e-06,
      "loss": 0.093,
      "reward": 0.10416666977107525,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 487
    },
    {
      "completion_length": 821.6041870117188,
      "epoch": 0.5951219512195122,
      "grad_norm": 0.1963929384946823,
      "kl": 0.0865478515625,
      "learning_rate": 1.264717650053936e-06,
      "loss": 0.0107,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 488
    },
    {
      "completion_length": 714.6875305175781,
      "epoch": 0.5963414634146341,
      "grad_norm": 0.3122914433479309,
      "kl": 0.087158203125,
      "learning_rate": 1.2584134900289346e-06,
      "loss": 0.0026,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 489
    },
    {
      "completion_length": 729.7500152587891,
      "epoch": 0.5975609756097561,
      "grad_norm": 0.15471084415912628,
      "kl": 0.1417236328125,
      "learning_rate": 1.2521137078354728e-06,
      "loss": 0.0249,
      "reward": 0.1666666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 490
    },
    {
      "completion_length": 761.5833587646484,
      "epoch": 0.598780487804878,
      "grad_norm": 0.2258668839931488,
      "kl": 0.07666015625,
      "learning_rate": 1.2458184176330102e-06,
      "loss": 0.0615,
      "reward": 0.1875000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 491
    },
    {
      "completion_length": 778.2500152587891,
      "epoch": 0.6,
      "grad_norm": 0.0527493879199028,
      "kl": 0.075408935546875,
      "learning_rate": 1.2395277334996047e-06,
      "loss": 0.0021,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 492
    },
    {
      "completion_length": 718.0208587646484,
      "epoch": 0.6012195121951219,
      "grad_norm": 0.2553408741950989,
      "kl": 0.09149169921875,
      "learning_rate": 1.2332417694298477e-06,
      "loss": 0.026,
      "reward": 0.2916666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 493
    },
    {
      "completion_length": 633.6875305175781,
      "epoch": 0.6024390243902439,
      "grad_norm": 0.3710525631904602,
      "kl": 0.113037109375,
      "learning_rate": 1.2269606393327968e-06,
      "loss": 0.0307,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 494
    },
    {
      "completion_length": 698.6458587646484,
      "epoch": 0.6036585365853658,
      "grad_norm": 0.3890061676502228,
      "kl": 0.08978271484375,
      "learning_rate": 1.2206844570299133e-06,
      "loss": -0.0019,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 495
    },
    {
      "completion_length": 777.4583587646484,
      "epoch": 0.6048780487804878,
      "grad_norm": 0.34140923619270325,
      "kl": 0.083251953125,
      "learning_rate": 1.2144133362529974e-06,
      "loss": 0.0505,
      "reward": 0.2083333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 496
    },
    {
      "completion_length": 684.7083587646484,
      "epoch": 0.6060975609756097,
      "grad_norm": 0.2975394129753113,
      "kl": 0.859619140625,
      "learning_rate": 1.2081473906421298e-06,
      "loss": 0.0422,
      "reward": 0.1666666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 497
    },
    {
      "completion_length": 783.2291870117188,
      "epoch": 0.6073170731707317,
      "grad_norm": 0.3838711977005005,
      "kl": 0.08880615234375,
      "learning_rate": 1.20188673374361e-06,
      "loss": 0.0502,
      "reward": 0.18750000186264515,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 498
    },
    {
      "completion_length": 704.0417022705078,
      "epoch": 0.6085365853658536,
      "grad_norm": 0.708202600479126,
      "kl": 0.109619140625,
      "learning_rate": 1.1956314790078998e-06,
      "loss": 0.0724,
      "reward": 0.2708333358168602,
      "reward_std": 0.21650635451078415,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 499
    },
    {
      "completion_length": 826.0208587646484,
      "epoch": 0.6097560975609756,
      "grad_norm": 0.133516326546669,
      "kl": 0.074951171875,
      "learning_rate": 1.189381739787569e-06,
      "loss": 0.003,
      "reward": 0.1875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 500
    },
    {
      "completion_length": 741.4583587646484,
      "epoch": 0.6109756097560975,
      "grad_norm": 0.6525834798812866,
      "kl": 0.107421875,
      "learning_rate": 1.1831376293352378e-06,
      "loss": -0.0266,
      "reward": 0.2083333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 501
    },
    {
      "completion_length": 734.4583435058594,
      "epoch": 0.6121951219512195,
      "grad_norm": 0.6010434031486511,
      "kl": 0.18511962890625,
      "learning_rate": 1.176899260801528e-06,
      "loss": -0.0396,
      "reward": 0.0833333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 502
    },
    {
      "completion_length": 752.4791870117188,
      "epoch": 0.6134146341463415,
      "grad_norm": 0.46545934677124023,
      "kl": 0.24658203125,
      "learning_rate": 1.1706667472330101e-06,
      "loss": 0.0462,
      "reward": 0.18750000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 503
    },
    {
      "completion_length": 654.6250152587891,
      "epoch": 0.6146341463414634,
      "grad_norm": 0.2879209518432617,
      "kl": 0.135498046875,
      "learning_rate": 1.1644402015701568e-06,
      "loss": -0.009,
      "reward": 0.12500000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000558793545,
      "rewards/format_reward": 0.0,
      "step": 504
    },
    {
      "completion_length": 735.1458587646484,
      "epoch": 0.6158536585365854,
      "grad_norm": 0.4487898647785187,
      "kl": 0.1240234375,
      "learning_rate": 1.158219736645294e-06,
      "loss": 0.0469,
      "reward": 0.125,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 505
    },
    {
      "completion_length": 713.8333587646484,
      "epoch": 0.6170731707317073,
      "grad_norm": 0.2627413868904114,
      "kl": 0.154541015625,
      "learning_rate": 1.152005465180558e-06,
      "loss": 0.0293,
      "reward": 0.18750000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 506
    },
    {
      "completion_length": 733.6458435058594,
      "epoch": 0.6182926829268293,
      "grad_norm": 0.5076727271080017,
      "kl": 0.1317138671875,
      "learning_rate": 1.145797499785853e-06,
      "loss": 0.0641,
      "reward": 0.1875000037252903,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 507
    },
    {
      "completion_length": 771.4375305175781,
      "epoch": 0.6195121951219512,
      "grad_norm": 0.21236741542816162,
      "kl": 0.17431640625,
      "learning_rate": 1.1395959529568088e-06,
      "loss": 0.0356,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 508
    },
    {
      "completion_length": 762.8750152587891,
      "epoch": 0.6207317073170732,
      "grad_norm": 0.32030531764030457,
      "kl": 0.097412109375,
      "learning_rate": 1.1334009370727446e-06,
      "loss": -0.0066,
      "reward": 0.2708333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 509
    },
    {
      "completion_length": 757.3333587646484,
      "epoch": 0.6219512195121951,
      "grad_norm": 0.3788541555404663,
      "kl": 0.103515625,
      "learning_rate": 1.127212564394629e-06,
      "loss": 0.0472,
      "reward": 0.2708333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 510
    },
    {
      "completion_length": 798.8750152587891,
      "epoch": 0.6231707317073171,
      "grad_norm": 0.3279261291027069,
      "kl": 0.2113037109375,
      "learning_rate": 1.1210309470630509e-06,
      "loss": 0.0485,
      "reward": 0.1875,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 511
    },
    {
      "completion_length": 768.1041717529297,
      "epoch": 0.624390243902439,
      "grad_norm": 0.24755020439624786,
      "kl": 0.2078857421875,
      "learning_rate": 1.1148561970961818e-06,
      "loss": 0.0215,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 512
    },
    {
      "completion_length": 879.3541870117188,
      "epoch": 0.625609756097561,
      "grad_norm": 0.2285972684621811,
      "kl": 0.15283203125,
      "learning_rate": 1.1086884263877486e-06,
      "loss": 0.0177,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 513
    },
    {
      "completion_length": 863.7500152587891,
      "epoch": 0.6268292682926829,
      "grad_norm": 0.22611166536808014,
      "kl": 0.18994140625,
      "learning_rate": 1.1025277467050079e-06,
      "loss": 0.022,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 514
    },
    {
      "completion_length": 905.5416717529297,
      "epoch": 0.6280487804878049,
      "grad_norm": 0.2169867306947708,
      "kl": 0.1121826171875,
      "learning_rate": 1.0963742696867162e-06,
      "loss": 0.0063,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 515
    },
    {
      "completion_length": 799.5208435058594,
      "epoch": 0.6292682926829268,
      "grad_norm": 0.5561642050743103,
      "kl": 0.20751953125,
      "learning_rate": 1.0902281068411114e-06,
      "loss": 0.0333,
      "reward": 0.2500000111758709,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000111758709,
      "rewards/format_reward": 0.0,
      "step": 516
    },
    {
      "completion_length": 864.6666717529297,
      "epoch": 0.6304878048780488,
      "grad_norm": 0.30151069164276123,
      "kl": 0.1224365234375,
      "learning_rate": 1.084089369543888e-06,
      "loss": 0.0238,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 517
    },
    {
      "completion_length": 787.5625152587891,
      "epoch": 0.6317073170731707,
      "grad_norm": 0.30315735936164856,
      "kl": 0.154052734375,
      "learning_rate": 1.077958169036183e-06,
      "loss": 0.0604,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 518
    },
    {
      "completion_length": 818.8958587646484,
      "epoch": 0.6329268292682927,
      "grad_norm": 0.42202794551849365,
      "kl": 0.1119384765625,
      "learning_rate": 1.0718346164225556e-06,
      "loss": 0.0306,
      "reward": 0.2291666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 519
    },
    {
      "completion_length": 713.4791717529297,
      "epoch": 0.6341463414634146,
      "grad_norm": 0.3794197142124176,
      "kl": 0.149658203125,
      "learning_rate": 1.0657188226689772e-06,
      "loss": 0.0387,
      "reward": 0.0416666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 520
    },
    {
      "completion_length": 770.3333587646484,
      "epoch": 0.6353658536585366,
      "grad_norm": 0.37589502334594727,
      "kl": 0.2100830078125,
      "learning_rate": 1.0596108986008203e-06,
      "loss": 0.0262,
      "reward": 0.27083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 521
    },
    {
      "completion_length": 765.9583435058594,
      "epoch": 0.6365853658536585,
      "grad_norm": 0.4906407296657562,
      "kl": 0.0966796875,
      "learning_rate": 1.0535109549008482e-06,
      "loss": 0.0371,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 522
    },
    {
      "completion_length": 866.7291870117188,
      "epoch": 0.6378048780487805,
      "grad_norm": 0.2712514102458954,
      "kl": 0.137939453125,
      "learning_rate": 1.0474191021072117e-06,
      "loss": 0.0004,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 523
    },
    {
      "completion_length": 735.9166717529297,
      "epoch": 0.6390243902439025,
      "grad_norm": 0.25394296646118164,
      "kl": 0.201171875,
      "learning_rate": 1.0413354506114434e-06,
      "loss": -0.0106,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 524
    },
    {
      "completion_length": 768.3958587646484,
      "epoch": 0.6402439024390244,
      "grad_norm": 0.45678046345710754,
      "kl": 0.0721435546875,
      "learning_rate": 1.0352601106564607e-06,
      "loss": 0.0462,
      "reward": 0.14583333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 525
    },
    {
      "completion_length": 769.2916870117188,
      "epoch": 0.6414634146341464,
      "grad_norm": 0.3751871585845947,
      "kl": 0.101318359375,
      "learning_rate": 1.0291931923345635e-06,
      "loss": 0.0632,
      "reward": 0.1458333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 526
    },
    {
      "completion_length": 839.4583435058594,
      "epoch": 0.6426829268292683,
      "grad_norm": 0.20345285534858704,
      "kl": 0.093505859375,
      "learning_rate": 1.0231348055854452e-06,
      "loss": 0.0268,
      "reward": 0.1666666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 527
    },
    {
      "completion_length": 679.5000152587891,
      "epoch": 0.6439024390243903,
      "grad_norm": 0.3024507164955139,
      "kl": 0.12255859375,
      "learning_rate": 1.0170850601941937e-06,
      "loss": 0.0031,
      "reward": 0.2500000074505806,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 528
    },
    {
      "completion_length": 755.3125305175781,
      "epoch": 0.6451219512195122,
      "grad_norm": 0.24449992179870605,
      "kl": 0.146484375,
      "learning_rate": 1.0110440657893074e-06,
      "loss": 0.001,
      "reward": 0.3333333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.3333333358168602,
      "rewards/format_reward": 0.0,
      "step": 529
    },
    {
      "completion_length": 656.3750152587891,
      "epoch": 0.6463414634146342,
      "grad_norm": 0.7463803887367249,
      "kl": 0.1195068359375,
      "learning_rate": 1.0050119318407061e-06,
      "loss": 0.0522,
      "reward": 0.1666666716337204,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 530
    },
    {
      "completion_length": 750.7292022705078,
      "epoch": 0.6475609756097561,
      "grad_norm": 0.31983718276023865,
      "kl": 0.1751708984375,
      "learning_rate": 9.98988767657747e-07,
      "loss": 0.0222,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 531
    },
    {
      "completion_length": 819.1250305175781,
      "epoch": 0.6487804878048781,
      "grad_norm": 0.30867284536361694,
      "kl": 0.1077880859375,
      "learning_rate": 9.929746823872462e-07,
      "loss": 0.0433,
      "reward": 0.2708333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333432674408,
      "rewards/format_reward": 0.0,
      "step": 532
    },
    {
      "completion_length": 820.8125152587891,
      "epoch": 0.65,
      "grad_norm": 0.14850081503391266,
      "kl": 0.0908203125,
      "learning_rate": 9.86969785011497e-07,
      "loss": 0.0037,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 533
    },
    {
      "completion_length": 781.8541870117188,
      "epoch": 0.651219512195122,
      "grad_norm": 0.4115312695503235,
      "kl": 0.142333984375,
      "learning_rate": 9.809741843462994e-07,
      "loss": 0.0185,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 534
    },
    {
      "completion_length": 723.375,
      "epoch": 0.6524390243902439,
      "grad_norm": 0.7542250752449036,
      "kl": 0.119873046875,
      "learning_rate": 9.749879890389848e-07,
      "loss": 0.0384,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 535
    },
    {
      "completion_length": 736.4583740234375,
      "epoch": 0.6536585365853659,
      "grad_norm": 0.4314693510532379,
      "kl": 0.11358642578125,
      "learning_rate": 9.690113075664488e-07,
      "loss": 0.0541,
      "reward": 0.14583333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 536
    },
    {
      "completion_length": 733.2916870117188,
      "epoch": 0.6548780487804878,
      "grad_norm": 0.4833289384841919,
      "kl": 0.10345458984375,
      "learning_rate": 9.630442482331853e-07,
      "loss": 0.0422,
      "reward": 0.1250000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 537
    },
    {
      "completion_length": 731.0208435058594,
      "epoch": 0.6560975609756098,
      "grad_norm": 0.26916182041168213,
      "kl": 0.0992431640625,
      "learning_rate": 9.57086919169323e-07,
      "loss": 0.0139,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 538
    },
    {
      "completion_length": 705.8333587646484,
      "epoch": 0.6573170731707317,
      "grad_norm": 0.3932608664035797,
      "kl": 0.1461181640625,
      "learning_rate": 9.511394283286686e-07,
      "loss": 0.0249,
      "reward": 0.1041666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 539
    },
    {
      "completion_length": 657.2916870117188,
      "epoch": 0.6585365853658537,
      "grad_norm": 0.6708987951278687,
      "kl": 0.14837646484375,
      "learning_rate": 9.452018834867454e-07,
      "loss": 0.0288,
      "reward": 0.27083333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 540
    },
    {
      "completion_length": 748.9583587646484,
      "epoch": 0.6597560975609756,
      "grad_norm": 0.3220924139022827,
      "kl": 0.1273193359375,
      "learning_rate": 9.392743922388469e-07,
      "loss": 0.0378,
      "reward": 0.18750000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 541
    },
    {
      "completion_length": 702.7708587646484,
      "epoch": 0.6609756097560976,
      "grad_norm": 0.596848726272583,
      "kl": 0.163330078125,
      "learning_rate": 9.333570619980818e-07,
      "loss": 0.0021,
      "reward": 0.2083333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 542
    },
    {
      "completion_length": 805.0625305175781,
      "epoch": 0.6621951219512195,
      "grad_norm": 0.32249799370765686,
      "kl": 0.1573486328125,
      "learning_rate": 9.27449999993429e-07,
      "loss": 0.0598,
      "reward": 0.1250000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 543
    },
    {
      "completion_length": 754.0208587646484,
      "epoch": 0.6634146341463415,
      "grad_norm": 0.3837894797325134,
      "kl": 0.07781982421875,
      "learning_rate": 9.215533132677969e-07,
      "loss": 0.0318,
      "reward": 0.2708333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 544
    },
    {
      "completion_length": 743.2291870117188,
      "epoch": 0.6646341463414634,
      "grad_norm": 0.39728447794914246,
      "kl": 0.09783935546875,
      "learning_rate": 9.156671086760788e-07,
      "loss": 0.0534,
      "reward": 0.2708333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 545
    },
    {
      "completion_length": 678.2500152587891,
      "epoch": 0.6658536585365854,
      "grad_norm": 0.6105949282646179,
      "kl": 0.1436767578125,
      "learning_rate": 9.097914928832228e-07,
      "loss": 0.1014,
      "reward": 0.3125000149011612,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.3125000149011612,
      "rewards/format_reward": 0.0,
      "step": 546
    },
    {
      "completion_length": 734.2916870117188,
      "epoch": 0.6670731707317074,
      "grad_norm": 0.39739248156547546,
      "kl": 0.162353515625,
      "learning_rate": 9.039265723622923e-07,
      "loss": -0.0294,
      "reward": 0.1875,
      "reward_std": 0.10825318098068237,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 547
    },
    {
      "completion_length": 672.8333587646484,
      "epoch": 0.6682926829268293,
      "grad_norm": 0.3068171441555023,
      "kl": 0.14227294921875,
      "learning_rate": 8.980724533925419e-07,
      "loss": 0.0142,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 548
    },
    {
      "completion_length": 733.5417022705078,
      "epoch": 0.6695121951219513,
      "grad_norm": 0.3112923204898834,
      "kl": 0.1214599609375,
      "learning_rate": 8.922292420574888e-07,
      "loss": 0.0314,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 549
    },
    {
      "completion_length": 782.3125152587891,
      "epoch": 0.6707317073170732,
      "grad_norm": 0.35543811321258545,
      "kl": 0.09661865234375,
      "learning_rate": 8.863970442429902e-07,
      "loss": 0.0083,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 550
    },
    {
      "completion_length": 701.5625152587891,
      "epoch": 0.6719512195121952,
      "grad_norm": 0.36748138070106506,
      "kl": 0.1202392578125,
      "learning_rate": 8.805759656353275e-07,
      "loss": 0.0404,
      "reward": 0.20833333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 551
    },
    {
      "completion_length": 641.0833587646484,
      "epoch": 0.6731707317073171,
      "grad_norm": 0.5275810360908508,
      "kl": 0.1756591796875,
      "learning_rate": 8.74766111719286e-07,
      "loss": -0.028,
      "reward": 0.2500000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 552
    },
    {
      "completion_length": 725.7917022705078,
      "epoch": 0.6743902439024391,
      "grad_norm": 0.4248809516429901,
      "kl": 0.11895751953125,
      "learning_rate": 8.689675877762487e-07,
      "loss": 0.0468,
      "reward": 0.22916667349636555,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.22916667349636555,
      "rewards/format_reward": 0.0,
      "step": 553
    },
    {
      "completion_length": 727.6666870117188,
      "epoch": 0.675609756097561,
      "grad_norm": 0.3157913088798523,
      "kl": 0.0596923828125,
      "learning_rate": 8.631804988822859e-07,
      "loss": 0.005,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 554
    },
    {
      "completion_length": 666.0000152587891,
      "epoch": 0.676829268292683,
      "grad_norm": 0.500182569026947,
      "kl": 0.1322021484375,
      "learning_rate": 8.574049499062509e-07,
      "loss": 0.0516,
      "reward": 0.47916668467223644,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.47916668467223644,
      "rewards/format_reward": 0.0,
      "step": 555
    },
    {
      "completion_length": 710.5,
      "epoch": 0.6780487804878049,
      "grad_norm": 0.48678165674209595,
      "kl": 0.1153564453125,
      "learning_rate": 8.516410455078793e-07,
      "loss": 0.0729,
      "reward": 0.1875000037252903,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 556
    },
    {
      "completion_length": 749.7291870117188,
      "epoch": 0.6792682926829269,
      "grad_norm": 0.5358501672744751,
      "kl": 0.41943359375,
      "learning_rate": 8.458888901358958e-07,
      "loss": 0.0387,
      "reward": 0.18750000186264515,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 557
    },
    {
      "completion_length": 677.2916717529297,
      "epoch": 0.6804878048780488,
      "grad_norm": 0.7909562587738037,
      "kl": 0.1656494140625,
      "learning_rate": 8.401485880261151e-07,
      "loss": -0.0007,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 558
    },
    {
      "completion_length": 699.1875305175781,
      "epoch": 0.6817073170731708,
      "grad_norm": 0.27314484119415283,
      "kl": 0.08013916015625,
      "learning_rate": 8.344202431995604e-07,
      "loss": 0.0267,
      "reward": 0.1041666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666716337204,
      "rewards/format_reward": 0.0,
      "step": 559
    },
    {
      "completion_length": 827.0416870117188,
      "epoch": 0.6829268292682927,
      "grad_norm": 0.27111247181892395,
      "kl": 0.1815185546875,
      "learning_rate": 8.287039594605737e-07,
      "loss": 0.0089,
      "reward": 0.0833333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 560
    },
    {
      "completion_length": 697.4583435058594,
      "epoch": 0.6841463414634147,
      "grad_norm": 0.5885205864906311,
      "kl": 0.1217041015625,
      "learning_rate": 8.229998403949348e-07,
      "loss": -0.0009,
      "reward": 0.0833333358168602,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 561
    },
    {
      "completion_length": 717.6041870117188,
      "epoch": 0.6853658536585366,
      "grad_norm": 0.6091434955596924,
      "kl": 0.158538818359375,
      "learning_rate": 8.173079893679873e-07,
      "loss": 0.0602,
      "reward": 0.16666666977107525,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 562
    },
    {
      "completion_length": 730.8333435058594,
      "epoch": 0.6865853658536586,
      "grad_norm": 0.5576154589653015,
      "kl": 0.199462890625,
      "learning_rate": 8.116285095227604e-07,
      "loss": 0.0377,
      "reward": 0.10416666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 563
    },
    {
      "completion_length": 740.4583587646484,
      "epoch": 0.6878048780487804,
      "grad_norm": 0.32833757996559143,
      "kl": 0.139892578125,
      "learning_rate": 8.05961503778108e-07,
      "loss": 0.0467,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 564
    },
    {
      "completion_length": 841.0208435058594,
      "epoch": 0.6890243902439024,
      "grad_norm": 0.25074008107185364,
      "kl": 0.156494140625,
      "learning_rate": 8.003070748268339e-07,
      "loss": 0.0242,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 565
    },
    {
      "completion_length": 745.0416870117188,
      "epoch": 0.6902439024390243,
      "grad_norm": 0.404649943113327,
      "kl": 0.3134765625,
      "learning_rate": 7.94665325133837e-07,
      "loss": 0.0202,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 566
    },
    {
      "completion_length": 710.0625152587891,
      "epoch": 0.6914634146341463,
      "grad_norm": 0.6244733929634094,
      "kl": 0.1131591796875,
      "learning_rate": 7.890363569342539e-07,
      "loss": 0.035,
      "reward": 0.14583333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 567
    },
    {
      "completion_length": 771.0625305175781,
      "epoch": 0.6926829268292682,
      "grad_norm": 0.40172481536865234,
      "kl": 0.1796875,
      "learning_rate": 7.834202722316054e-07,
      "loss": 0.0497,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 568
    },
    {
      "completion_length": 760.5208587646484,
      "epoch": 0.6939024390243902,
      "grad_norm": 1.2464784383773804,
      "kl": 0.16162109375,
      "learning_rate": 7.778171727959482e-07,
      "loss": 0.0101,
      "reward": 0.3333333469927311,
      "reward_std": 0.25259074568748474,
      "rewards/accuracy_reward": 0.3333333469927311,
      "rewards/format_reward": 0.0,
      "step": 569
    },
    {
      "completion_length": 817.3333740234375,
      "epoch": 0.6951219512195121,
      "grad_norm": 1.0639081001281738,
      "kl": 0.1851806640625,
      "learning_rate": 7.722271601620293e-07,
      "loss": 0.0169,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 570
    },
    {
      "completion_length": 719.6041870117188,
      "epoch": 0.6963414634146341,
      "grad_norm": 0.48873811960220337,
      "kl": 0.16259765625,
      "learning_rate": 7.6665033562745e-07,
      "loss": 0.0266,
      "reward": 0.18750000558793545,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 571
    },
    {
      "completion_length": 763.2500152587891,
      "epoch": 0.697560975609756,
      "grad_norm": 0.41101592779159546,
      "kl": 0.221435546875,
      "learning_rate": 7.610868002508248e-07,
      "loss": 0.0577,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 572
    },
    {
      "completion_length": 725.9375152587891,
      "epoch": 0.698780487804878,
      "grad_norm": 0.34577131271362305,
      "kl": 0.140380859375,
      "learning_rate": 7.555366548499551e-07,
      "loss": 0.0268,
      "reward": 0.3333333432674408,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3333333432674408,
      "rewards/format_reward": 0.0,
      "step": 573
    },
    {
      "completion_length": 705.7708587646484,
      "epoch": 0.7,
      "grad_norm": 0.4157039225101471,
      "kl": 0.153564453125,
      "learning_rate": 7.500000000000003e-07,
      "loss": 0.0312,
      "reward": 0.1666666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 574
    },
    {
      "completion_length": 783.4166717529297,
      "epoch": 0.7012195121951219,
      "grad_norm": 0.41202351450920105,
      "kl": 0.1884765625,
      "learning_rate": 7.444769360316534e-07,
      "loss": 0.0085,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 575
    },
    {
      "completion_length": 788.2291717529297,
      "epoch": 0.7024390243902439,
      "grad_norm": 0.27398866415023804,
      "kl": 0.2012939453125,
      "learning_rate": 7.389675630293269e-07,
      "loss": 0.0207,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 576
    },
    {
      "completion_length": 848.0833587646484,
      "epoch": 0.7036585365853658,
      "grad_norm": 0.9130232930183411,
      "kl": 0.272216796875,
      "learning_rate": 7.334719808293342e-07,
      "loss": 0.0822,
      "reward": 0.22916667349636555,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.22916667349636555,
      "rewards/format_reward": 0.0,
      "step": 577
    },
    {
      "completion_length": 775.3333435058594,
      "epoch": 0.7048780487804878,
      "grad_norm": 0.2764679789543152,
      "kl": 0.19317626953125,
      "learning_rate": 7.279902890180865e-07,
      "loss": 0.0245,
      "reward": 0.27083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 578
    },
    {
      "completion_length": 748.9375152587891,
      "epoch": 0.7060975609756097,
      "grad_norm": 0.39309820532798767,
      "kl": 0.23583984375,
      "learning_rate": 7.225225869302818e-07,
      "loss": 0.0414,
      "reward": 0.2500000074505806,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 579
    },
    {
      "completion_length": 764.8750305175781,
      "epoch": 0.7073170731707317,
      "grad_norm": 0.6356940865516663,
      "kl": 0.31982421875,
      "learning_rate": 7.1706897364711e-07,
      "loss": 0.0548,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 580
    },
    {
      "completion_length": 829.2500305175781,
      "epoch": 0.7085365853658536,
      "grad_norm": 0.42424410581588745,
      "kl": 0.451904296875,
      "learning_rate": 7.116295479944533e-07,
      "loss": 0.0306,
      "reward": 0.1666666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 581
    },
    {
      "completion_length": 688.9583587646484,
      "epoch": 0.7097560975609756,
      "grad_norm": 0.40363651514053345,
      "kl": 0.248291015625,
      "learning_rate": 7.062044085410991e-07,
      "loss": 0.0274,
      "reward": 0.2500000111758709,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000111758709,
      "rewards/format_reward": 0.0,
      "step": 582
    },
    {
      "completion_length": 772.9166870117188,
      "epoch": 0.7109756097560975,
      "grad_norm": 0.4686078131198883,
      "kl": 0.3818359375,
      "learning_rate": 7.007936535969516e-07,
      "loss": 0.0551,
      "reward": 0.1250000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 583
    },
    {
      "completion_length": 843.6041717529297,
      "epoch": 0.7121951219512195,
      "grad_norm": 0.4927625358104706,
      "kl": 0.486328125,
      "learning_rate": 6.9539738121125e-07,
      "loss": 0.0587,
      "reward": 0.1458333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 584
    },
    {
      "completion_length": 725.1458435058594,
      "epoch": 0.7134146341463414,
      "grad_norm": 0.39137816429138184,
      "kl": 0.45654296875,
      "learning_rate": 6.90015689170794e-07,
      "loss": 0.0429,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 585
    },
    {
      "completion_length": 737.5416717529297,
      "epoch": 0.7146341463414634,
      "grad_norm": 0.4557369351387024,
      "kl": 0.43115234375,
      "learning_rate": 6.846486749981684e-07,
      "loss": 0.0599,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 586
    },
    {
      "completion_length": 786.2291870117188,
      "epoch": 0.7158536585365853,
      "grad_norm": 0.5112011432647705,
      "kl": 0.21728515625,
      "learning_rate": 6.792964359499794e-07,
      "loss": 0.0369,
      "reward": 0.2500000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 587
    },
    {
      "completion_length": 701.7083587646484,
      "epoch": 0.7170731707317073,
      "grad_norm": 0.5337812900543213,
      "kl": 0.38330078125,
      "learning_rate": 6.739590690150903e-07,
      "loss": 0.0265,
      "reward": 0.3125000074505806,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.3125000074505806,
      "rewards/format_reward": 0.0,
      "step": 588
    },
    {
      "completion_length": 735.9166717529297,
      "epoch": 0.7182926829268292,
      "grad_norm": 0.6751530170440674,
      "kl": 0.295166015625,
      "learning_rate": 6.686366709128632e-07,
      "loss": 0.0502,
      "reward": 0.18750000558793545,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 589
    },
    {
      "completion_length": 722.7291717529297,
      "epoch": 0.7195121951219512,
      "grad_norm": 0.9492172598838806,
      "kl": 0.3935546875,
      "learning_rate": 6.633293380914087e-07,
      "loss": 0.0438,
      "reward": 0.0416666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 590
    },
    {
      "completion_length": 763.7916717529297,
      "epoch": 0.7207317073170731,
      "grad_norm": 3.102266788482666,
      "kl": 0.46484375,
      "learning_rate": 6.580371667258349e-07,
      "loss": 0.0828,
      "reward": 0.25000000558793545,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 591
    },
    {
      "completion_length": 722.6458435058594,
      "epoch": 0.7219512195121951,
      "grad_norm": 0.6278510093688965,
      "kl": 0.62548828125,
      "learning_rate": 6.527602527165099e-07,
      "loss": 0.0606,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 592
    },
    {
      "completion_length": 734.9166717529297,
      "epoch": 0.723170731707317,
      "grad_norm": 3.2532951831817627,
      "kl": 0.71875,
      "learning_rate": 6.474986916873168e-07,
      "loss": 0.0282,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 593
    },
    {
      "completion_length": 736.2291870117188,
      "epoch": 0.724390243902439,
      "grad_norm": 1.1925694942474365,
      "kl": 1.1259765625,
      "learning_rate": 6.422525789839273e-07,
      "loss": 0.1125,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 594
    },
    {
      "completion_length": 699.3750305175781,
      "epoch": 0.725609756097561,
      "grad_norm": 1.3285707235336304,
      "kl": 0.83203125,
      "learning_rate": 6.370220096720692e-07,
      "loss": 0.0772,
      "reward": 0.2500000074505806,
      "reward_std": 0.25259073823690414,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 595
    },
    {
      "completion_length": 785.8958587646484,
      "epoch": 0.7268292682926829,
      "grad_norm": 1.3592267036437988,
      "kl": 0.702880859375,
      "learning_rate": 6.318070785358074e-07,
      "loss": 0.0475,
      "reward": 0.1875,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 596
    },
    {
      "completion_length": 818.6250305175781,
      "epoch": 0.7280487804878049,
      "grad_norm": 0.9800528883934021,
      "kl": 0.740234375,
      "learning_rate": 6.266078800758249e-07,
      "loss": 0.0697,
      "reward": 0.1666666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 597
    },
    {
      "completion_length": 759.5833587646484,
      "epoch": 0.7292682926829268,
      "grad_norm": 1.697644591331482,
      "kl": 0.560546875,
      "learning_rate": 6.214245085077078e-07,
      "loss": 0.116,
      "reward": 0.2291666716337204,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 598
    },
    {
      "completion_length": 700.2083587646484,
      "epoch": 0.7304878048780488,
      "grad_norm": 1.0294965505599976,
      "kl": 0.46044921875,
      "learning_rate": 6.162570577602433e-07,
      "loss": 0.0201,
      "reward": 0.25000000186264515,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.0,
      "step": 599
    },
    {
      "completion_length": 713.5000305175781,
      "epoch": 0.7317073170731707,
      "grad_norm": 0.3935979902744293,
      "kl": 0.5078125,
      "learning_rate": 6.11105621473712e-07,
      "loss": 0.0543,
      "reward": 0.1250000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 600
    },
    {
      "completion_length": 673.8125152587891,
      "epoch": 0.7329268292682927,
      "grad_norm": 0.5782250165939331,
      "kl": 0.52001953125,
      "learning_rate": 6.059702929981952e-07,
      "loss": 0.0636,
      "reward": 0.1458333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 601
    },
    {
      "completion_length": 823.2708740234375,
      "epoch": 0.7341463414634146,
      "grad_norm": 0.8451728224754333,
      "kl": 0.563232421875,
      "learning_rate": 6.008511653918821e-07,
      "loss": 0.0566,
      "reward": 0.1458333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 602
    },
    {
      "completion_length": 726.6875305175781,
      "epoch": 0.7353658536585366,
      "grad_norm": 0.5462982654571533,
      "kl": 0.3779296875,
      "learning_rate": 5.957483314193813e-07,
      "loss": 0.0317,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 603
    },
    {
      "completion_length": 783.0833587646484,
      "epoch": 0.7365853658536585,
      "grad_norm": 1.1631156206130981,
      "kl": 0.42626953125,
      "learning_rate": 5.906618835500434e-07,
      "loss": 0.0192,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 604
    },
    {
      "completion_length": 778.4375152587891,
      "epoch": 0.7378048780487805,
      "grad_norm": 1.4091110229492188,
      "kl": 0.30291748046875,
      "learning_rate": 5.855919139562815e-07,
      "loss": 0.046,
      "reward": 0.14583333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 605
    },
    {
      "completion_length": 775.0417022705078,
      "epoch": 0.7390243902439024,
      "grad_norm": 0.440822571516037,
      "kl": 0.26068115234375,
      "learning_rate": 5.805385145119064e-07,
      "loss": 0.0313,
      "reward": 0.1250000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 606
    },
    {
      "completion_length": 671.7708587646484,
      "epoch": 0.7402439024390244,
      "grad_norm": 1.904105305671692,
      "kl": 0.28662109375,
      "learning_rate": 5.755017767904543e-07,
      "loss": 0.047,
      "reward": 0.33333334140479565,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.33333334140479565,
      "rewards/format_reward": 0.0,
      "step": 607
    },
    {
      "completion_length": 780.3333435058594,
      "epoch": 0.7414634146341463,
      "grad_norm": 0.3190377950668335,
      "kl": 0.38037109375,
      "learning_rate": 5.704817920635348e-07,
      "loss": 0.0597,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 608
    },
    {
      "completion_length": 771.0416870117188,
      "epoch": 0.7426829268292683,
      "grad_norm": 0.46209630370140076,
      "kl": 0.181304931640625,
      "learning_rate": 5.654786512991705e-07,
      "loss": 0.008,
      "reward": 0.22916666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 609
    },
    {
      "completion_length": 759.3541870117188,
      "epoch": 0.7439024390243902,
      "grad_norm": 0.63628089427948,
      "kl": 0.3106689453125,
      "learning_rate": 5.60492445160154e-07,
      "loss": 0.0077,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 610
    },
    {
      "completion_length": 677.2708587646484,
      "epoch": 0.7451219512195122,
      "grad_norm": 0.9306976795196533,
      "kl": 0.321533203125,
      "learning_rate": 5.555232640024021e-07,
      "loss": 0.04,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 611
    },
    {
      "completion_length": 804.4792022705078,
      "epoch": 0.7463414634146341,
      "grad_norm": 0.5185068845748901,
      "kl": 0.2357177734375,
      "learning_rate": 5.505711978733175e-07,
      "loss": 0.0248,
      "reward": 0.22916666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.22916666977107525,
      "rewards/format_reward": 0.0,
      "step": 612
    },
    {
      "completion_length": 758.1250152587891,
      "epoch": 0.7475609756097561,
      "grad_norm": 1.219456672668457,
      "kl": 0.3994140625,
      "learning_rate": 5.456363365101606e-07,
      "loss": 0.0406,
      "reward": 0.12500000186264515,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 613
    },
    {
      "completion_length": 764.6875,
      "epoch": 0.748780487804878,
      "grad_norm": 0.5506624579429626,
      "kl": 0.346923828125,
      "learning_rate": 5.407187693384191e-07,
      "loss": 0.0446,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 614
    },
    {
      "completion_length": 841.7291870117188,
      "epoch": 0.75,
      "grad_norm": 0.4663028120994568,
      "kl": 0.5244140625,
      "learning_rate": 5.358185854701909e-07,
      "loss": 0.0434,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 615
    },
    {
      "completion_length": 736.1250152587891,
      "epoch": 0.751219512195122,
      "grad_norm": 1.7369216680526733,
      "kl": 0.46728515625,
      "learning_rate": 5.309358737025682e-07,
      "loss": 0.0726,
      "reward": 0.16666666977107525,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 616
    },
    {
      "completion_length": 700.5833587646484,
      "epoch": 0.7524390243902439,
      "grad_norm": 1.2079142332077026,
      "kl": 0.2919921875,
      "learning_rate": 5.26070722516028e-07,
      "loss": 0.001,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 617
    },
    {
      "completion_length": 682.0625152587891,
      "epoch": 0.7536585365853659,
      "grad_norm": 1.0336294174194336,
      "kl": 0.47314453125,
      "learning_rate": 5.21223220072828e-07,
      "loss": 0.0613,
      "reward": 0.291666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.291666679084301,
      "rewards/format_reward": 0.0,
      "step": 618
    },
    {
      "completion_length": 775.0833435058594,
      "epoch": 0.7548780487804878,
      "grad_norm": 0.523438036441803,
      "kl": 0.475830078125,
      "learning_rate": 5.163934542154106e-07,
      "loss": 0.0507,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 619
    },
    {
      "completion_length": 778.3958587646484,
      "epoch": 0.7560975609756098,
      "grad_norm": 1.1852086782455444,
      "kl": 0.42236328125,
      "learning_rate": 5.115815124648103e-07,
      "loss": 0.0137,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 620
    },
    {
      "completion_length": 746.9375152587891,
      "epoch": 0.7573170731707317,
      "grad_norm": 0.4965358078479767,
      "kl": 0.30224609375,
      "learning_rate": 5.067874820190684e-07,
      "loss": 0.0103,
      "reward": 0.125,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 621
    },
    {
      "completion_length": 726.7708587646484,
      "epoch": 0.7585365853658537,
      "grad_norm": 1.1848160028457642,
      "kl": 0.218994140625,
      "learning_rate": 5.020114497516521e-07,
      "loss": 0.0243,
      "reward": 0.1250000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 622
    },
    {
      "completion_length": 641.1250305175781,
      "epoch": 0.7597560975609756,
      "grad_norm": 0.9911088347434998,
      "kl": 0.297119140625,
      "learning_rate": 4.972535022098795e-07,
      "loss": 0.0645,
      "reward": 0.3541666716337204,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.3541666716337204,
      "rewards/format_reward": 0.0,
      "step": 623
    },
    {
      "completion_length": 744.6875152587891,
      "epoch": 0.7609756097560976,
      "grad_norm": 0.4609319269657135,
      "kl": 0.427734375,
      "learning_rate": 4.925137256133533e-07,
      "loss": 0.0243,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 624
    },
    {
      "completion_length": 763.4583435058594,
      "epoch": 0.7621951219512195,
      "grad_norm": 0.66517573595047,
      "kl": 0.467041015625,
      "learning_rate": 4.877922058523971e-07,
      "loss": 0.0408,
      "reward": 0.1458333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 625
    },
    {
      "completion_length": 715.4166717529297,
      "epoch": 0.7634146341463415,
      "grad_norm": 0.4712808132171631,
      "kl": 0.4072265625,
      "learning_rate": 4.830890284864985e-07,
      "loss": 0.0235,
      "reward": 0.08333333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 626
    },
    {
      "completion_length": 793.2708435058594,
      "epoch": 0.7646341463414634,
      "grad_norm": 0.4655906558036804,
      "kl": 0.3056640625,
      "learning_rate": 4.784042787427605e-07,
      "loss": 0.0048,
      "reward": 0.25000000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 627
    },
    {
      "completion_length": 776.1458587646484,
      "epoch": 0.7658536585365854,
      "grad_norm": 0.6277633905410767,
      "kl": 0.35626220703125,
      "learning_rate": 4.7373804151435456e-07,
      "loss": 0.0219,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 628
    },
    {
      "completion_length": 726.7083587646484,
      "epoch": 0.7670731707317073,
      "grad_norm": 0.569290041923523,
      "kl": 0.1920166015625,
      "learning_rate": 4.6909040135898463e-07,
      "loss": 0.0098,
      "reward": 0.0833333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 629
    },
    {
      "completion_length": 709.2708587646484,
      "epoch": 0.7682926829268293,
      "grad_norm": 0.4517268240451813,
      "kl": 0.14959716796875,
      "learning_rate": 4.6446144249735345e-07,
      "loss": -0.0047,
      "reward": 0.12500000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000558793545,
      "rewards/format_reward": 0.0,
      "step": 630
    },
    {
      "completion_length": 717.3750152587891,
      "epoch": 0.7695121951219512,
      "grad_norm": 0.43553000688552856,
      "kl": 0.40234375,
      "learning_rate": 4.598512488116376e-07,
      "loss": 0.013,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 631
    },
    {
      "completion_length": 730.4791717529297,
      "epoch": 0.7707317073170732,
      "grad_norm": 1.2086296081542969,
      "kl": 0.31494140625,
      "learning_rate": 4.552599038439651e-07,
      "loss": -0.0012,
      "reward": 0.0833333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 632
    },
    {
      "completion_length": 733.2917022705078,
      "epoch": 0.7719512195121951,
      "grad_norm": 0.6143542528152466,
      "kl": 0.3182373046875,
      "learning_rate": 4.506874907949034e-07,
      "loss": 0.1023,
      "reward": 0.20833333767950535,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.20833333767950535,
      "rewards/format_reward": 0.0,
      "step": 633
    },
    {
      "completion_length": 678.5208435058594,
      "epoch": 0.7731707317073171,
      "grad_norm": 0.26520198583602905,
      "kl": 0.2115478515625,
      "learning_rate": 4.461340925219522e-07,
      "loss": 0.0171,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 634
    },
    {
      "completion_length": 742.1458587646484,
      "epoch": 0.774390243902439,
      "grad_norm": 1.6623936891555786,
      "kl": 0.263916015625,
      "learning_rate": 4.4159979153804064e-07,
      "loss": 0.0144,
      "reward": 0.1875,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 635
    },
    {
      "completion_length": 701.0000152587891,
      "epoch": 0.775609756097561,
      "grad_norm": 2.114635467529297,
      "kl": 0.265380859375,
      "learning_rate": 4.3708467001003305e-07,
      "loss": 0.0684,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 636
    },
    {
      "completion_length": 607.7708435058594,
      "epoch": 0.776829268292683,
      "grad_norm": 0.6238024830818176,
      "kl": 0.2293701171875,
      "learning_rate": 4.3258880975723777e-07,
      "loss": 0.0014,
      "reward": 0.1875000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 637
    },
    {
      "completion_length": 758.0208435058594,
      "epoch": 0.7780487804878049,
      "grad_norm": 0.3152739107608795,
      "kl": 0.3780517578125,
      "learning_rate": 4.2811229224992807e-07,
      "loss": 0.034,
      "reward": 0.2291666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 638
    },
    {
      "completion_length": 761.5208587646484,
      "epoch": 0.7792682926829269,
      "grad_norm": 0.7839704155921936,
      "kl": 0.313232421875,
      "learning_rate": 4.2365519860786316e-07,
      "loss": 0.0362,
      "reward": 0.16666666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 639
    },
    {
      "completion_length": 799.6041870117188,
      "epoch": 0.7804878048780488,
      "grad_norm": 0.5361413955688477,
      "kl": 0.4287109375,
      "learning_rate": 4.192176095988196e-07,
      "loss": 0.0589,
      "reward": 0.2916666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 640
    },
    {
      "completion_length": 735.3541870117188,
      "epoch": 0.7817073170731708,
      "grad_norm": 2.4874582290649414,
      "kl": 0.59814453125,
      "learning_rate": 4.147996056371258e-07,
      "loss": 0.0831,
      "reward": 0.1666666716337204,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 641
    },
    {
      "completion_length": 733.0416870117188,
      "epoch": 0.7829268292682927,
      "grad_norm": 0.5204437375068665,
      "kl": 0.459716796875,
      "learning_rate": 4.1040126678220656e-07,
      "loss": 0.0698,
      "reward": 0.1875000074505806,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875000074505806,
      "rewards/format_reward": 0.0,
      "step": 642
    },
    {
      "completion_length": 765.3333435058594,
      "epoch": 0.7841463414634147,
      "grad_norm": 2.1297390460968018,
      "kl": 0.4979248046875,
      "learning_rate": 4.060226727371327e-07,
      "loss": 0.0636,
      "reward": 0.4375000111758709,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.4375000111758709,
      "rewards/format_reward": 0.0,
      "step": 643
    },
    {
      "completion_length": 772.6458435058594,
      "epoch": 0.7853658536585366,
      "grad_norm": 1.1510241031646729,
      "kl": 0.65625,
      "learning_rate": 4.0166390284717475e-07,
      "loss": 0.0343,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 644
    },
    {
      "completion_length": 781.0833587646484,
      "epoch": 0.7865853658536586,
      "grad_norm": 0.7534167766571045,
      "kl": 0.418701171875,
      "learning_rate": 3.973250360983677e-07,
      "loss": 0.007,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 645
    },
    {
      "completion_length": 719.0416870117188,
      "epoch": 0.7878048780487805,
      "grad_norm": 1.2304871082305908,
      "kl": 0.332763671875,
      "learning_rate": 3.930061511160762e-07,
      "loss": 0.0146,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 646
    },
    {
      "completion_length": 709.9375305175781,
      "epoch": 0.7890243902439025,
      "grad_norm": 0.5384507775306702,
      "kl": 0.55712890625,
      "learning_rate": 3.8870732616357364e-07,
      "loss": 0.0499,
      "reward": 0.2500000074505806,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 647
    },
    {
      "completion_length": 767.0625305175781,
      "epoch": 0.7902439024390244,
      "grad_norm": 0.812372624874115,
      "kl": 0.51513671875,
      "learning_rate": 3.8442863914062065e-07,
      "loss": 0.0688,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 648
    },
    {
      "completion_length": 751.5208587646484,
      "epoch": 0.7914634146341464,
      "grad_norm": 2.1022772789001465,
      "kl": 0.59130859375,
      "learning_rate": 3.8017016758205597e-07,
      "loss": 0.0808,
      "reward": 0.2291666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 649
    },
    {
      "completion_length": 710.5416870117188,
      "epoch": 0.7926829268292683,
      "grad_norm": 0.880653440952301,
      "kl": 0.775390625,
      "learning_rate": 3.759319886563905e-07,
      "loss": 0.1213,
      "reward": 0.2291666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 650
    },
    {
      "completion_length": 827.4583587646484,
      "epoch": 0.7939024390243903,
      "grad_norm": 1.8597267866134644,
      "kl": 0.84130859375,
      "learning_rate": 3.7171417916440714e-07,
      "loss": 0.0522,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 651
    },
    {
      "completion_length": 762.9583587646484,
      "epoch": 0.7951219512195122,
      "grad_norm": 1.6760634183883667,
      "kl": 0.6256103515625,
      "learning_rate": 3.6751681553777236e-07,
      "loss": 0.0583,
      "reward": 0.2083333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 652
    },
    {
      "completion_length": 804.1041870117188,
      "epoch": 0.7963414634146342,
      "grad_norm": 2.001121759414673,
      "kl": 0.970703125,
      "learning_rate": 3.633399738376491e-07,
      "loss": 0.0509,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 653
    },
    {
      "completion_length": 808.7291870117188,
      "epoch": 0.7975609756097561,
      "grad_norm": 0.7535003423690796,
      "kl": 0.7021484375,
      "learning_rate": 3.5918372975331933e-07,
      "loss": 0.107,
      "reward": 0.27083334140479565,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.27083334140479565,
      "rewards/format_reward": 0.0,
      "step": 654
    },
    {
      "completion_length": 720.9791717529297,
      "epoch": 0.7987804878048781,
      "grad_norm": 2.3305118083953857,
      "kl": 0.77099609375,
      "learning_rate": 3.5504815860081056e-07,
      "loss": 0.0509,
      "reward": 0.18750000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 655
    },
    {
      "completion_length": 710.5416870117188,
      "epoch": 0.8,
      "grad_norm": 0.6330143213272095,
      "kl": 0.5576171875,
      "learning_rate": 3.5093333532153313e-07,
      "loss": 0.0654,
      "reward": 0.1875,
      "reward_std": 0.10825318098068237,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 656
    },
    {
      "completion_length": 725.5208587646484,
      "epoch": 0.801219512195122,
      "grad_norm": 2.6927578449249268,
      "kl": 0.47998046875,
      "learning_rate": 3.468393344809222e-07,
      "loss": 0.0393,
      "reward": 0.18750000186264515,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 657
    },
    {
      "completion_length": 753.1875152587891,
      "epoch": 0.802439024390244,
      "grad_norm": 1.1552072763442993,
      "kl": 0.3990478515625,
      "learning_rate": 3.4276623026708556e-07,
      "loss": 0.0074,
      "reward": 0.2708333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 658
    },
    {
      "completion_length": 719.8541870117188,
      "epoch": 0.8036585365853659,
      "grad_norm": 0.6939015984535217,
      "kl": 0.4619140625,
      "learning_rate": 3.3871409648945955e-07,
      "loss": 0.0611,
      "reward": 0.25000000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 659
    },
    {
      "completion_length": 698.0208587646484,
      "epoch": 0.8048780487804879,
      "grad_norm": 1.828070044517517,
      "kl": 0.2796630859375,
      "learning_rate": 3.346830065774706e-07,
      "loss": 0.0268,
      "reward": 0.2083333395421505,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 660
    },
    {
      "completion_length": 704.7083435058594,
      "epoch": 0.8060975609756098,
      "grad_norm": 0.8830689787864685,
      "kl": 0.266845703125,
      "learning_rate": 3.306730335792075e-07,
      "loss": 0.0239,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 661
    },
    {
      "completion_length": 773.2083587646484,
      "epoch": 0.8073170731707318,
      "grad_norm": 0.7095469832420349,
      "kl": 0.347412109375,
      "learning_rate": 3.266842501600934e-07,
      "loss": 0.0469,
      "reward": 0.2708333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 662
    },
    {
      "completion_length": 792.3541717529297,
      "epoch": 0.8085365853658537,
      "grad_norm": 0.7811137437820435,
      "kl": 0.50732421875,
      "learning_rate": 3.2271672860157324e-07,
      "loss": 0.0517,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 663
    },
    {
      "completion_length": 794.2291870117188,
      "epoch": 0.8097560975609757,
      "grad_norm": 2.9623236656188965,
      "kl": 0.4558868408203125,
      "learning_rate": 3.187705407998018e-07,
      "loss": 0.0768,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 664
    },
    {
      "completion_length": 706.1250152587891,
      "epoch": 0.8109756097560976,
      "grad_norm": 2.820039749145508,
      "kl": 0.53363037109375,
      "learning_rate": 3.148457582643398e-07,
      "loss": 0.0893,
      "reward": 0.1666666716337204,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 665
    },
    {
      "completion_length": 746.8125152587891,
      "epoch": 0.8121951219512196,
      "grad_norm": 0.8638564944267273,
      "kl": 0.40478515625,
      "learning_rate": 3.1094245211686106e-07,
      "loss": 0.0594,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 666
    },
    {
      "completion_length": 783.1041870117188,
      "epoch": 0.8134146341463414,
      "grad_norm": 2.670159339904785,
      "kl": 0.53094482421875,
      "learning_rate": 3.070606930898602e-07,
      "loss": 0.0804,
      "reward": 0.1666666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 667
    },
    {
      "completion_length": 756.1458587646484,
      "epoch": 0.8146341463414634,
      "grad_norm": 1.8002389669418335,
      "kl": 0.57421875,
      "learning_rate": 3.032005515253751e-07,
      "loss": 0.0794,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 668
    },
    {
      "completion_length": 727.2916870117188,
      "epoch": 0.8158536585365853,
      "grad_norm": 3.232239007949829,
      "kl": 0.45947265625,
      "learning_rate": 2.9936209737370727e-07,
      "loss": 0.0854,
      "reward": 0.2083333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 669
    },
    {
      "completion_length": 717.5000305175781,
      "epoch": 0.8170731707317073,
      "grad_norm": 0.515435516834259,
      "kl": 0.3421630859375,
      "learning_rate": 2.955454001921588e-07,
      "loss": 0.0164,
      "reward": 0.2500000037252903,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 670
    },
    {
      "completion_length": 655.5625152587891,
      "epoch": 0.8182926829268292,
      "grad_norm": 0.8738469481468201,
      "kl": 0.2606201171875,
      "learning_rate": 2.917505291437683e-07,
      "loss": 0.0134,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 671
    },
    {
      "completion_length": 756.4375305175781,
      "epoch": 0.8195121951219512,
      "grad_norm": 0.47037848830223083,
      "kl": 0.42626953125,
      "learning_rate": 2.879775529960603e-07,
      "loss": 0.0384,
      "reward": 0.20833333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 672
    },
    {
      "completion_length": 783.6041870117188,
      "epoch": 0.8207317073170731,
      "grad_norm": 3.444483995437622,
      "kl": 0.8154296875,
      "learning_rate": 2.842265401197982e-07,
      "loss": 0.0942,
      "reward": 0.29166667722165585,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.29166667722165585,
      "rewards/format_reward": 0.0,
      "step": 673
    },
    {
      "completion_length": 806.1250152587891,
      "epoch": 0.8219512195121951,
      "grad_norm": 0.7323744297027588,
      "kl": 0.742431640625,
      "learning_rate": 2.8049755848774337e-07,
      "loss": 0.0841,
      "reward": 0.20833333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 674
    },
    {
      "completion_length": 773.1041870117188,
      "epoch": 0.823170731707317,
      "grad_norm": 1.1480482816696167,
      "kl": 1.05078125,
      "learning_rate": 2.7679067567342766e-07,
      "loss": 0.108,
      "reward": 0.18750000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 675
    },
    {
      "completion_length": 740.7291870117188,
      "epoch": 0.824390243902439,
      "grad_norm": 2.3588831424713135,
      "kl": 0.88232421875,
      "learning_rate": 2.7310595884992354e-07,
      "loss": 0.0498,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 676
    },
    {
      "completion_length": 803.4166870117188,
      "epoch": 0.8256097560975609,
      "grad_norm": 0.8971549272537231,
      "kl": 0.83984375,
      "learning_rate": 2.6944347478863226e-07,
      "loss": 0.0681,
      "reward": 0.1875000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 677
    },
    {
      "completion_length": 672.1041870117188,
      "epoch": 0.8268292682926829,
      "grad_norm": 1.219985008239746,
      "kl": 0.769287109375,
      "learning_rate": 2.658032898580702e-07,
      "loss": 0.0398,
      "reward": 0.2500000111758709,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2500000111758709,
      "rewards/format_reward": 0.0,
      "step": 678
    },
    {
      "completion_length": 761.9375305175781,
      "epoch": 0.8280487804878048,
      "grad_norm": 1.3300340175628662,
      "kl": 0.8876953125,
      "learning_rate": 2.621854700226663e-07,
      "loss": 0.104,
      "reward": 0.1666666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 679
    },
    {
      "completion_length": 800.4166870117188,
      "epoch": 0.8292682926829268,
      "grad_norm": 3.330599784851074,
      "kl": 1.349609375,
      "learning_rate": 2.5859008084156986e-07,
      "loss": 0.1013,
      "reward": 0.14583333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 680
    },
    {
      "completion_length": 771.2708587646484,
      "epoch": 0.8304878048780487,
      "grad_norm": 1.2517083883285522,
      "kl": 0.480224609375,
      "learning_rate": 2.5501718746745766e-07,
      "loss": 0.0244,
      "reward": 0.12500000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000558793545,
      "rewards/format_reward": 0.0,
      "step": 681
    },
    {
      "completion_length": 777.7708587646484,
      "epoch": 0.8317073170731707,
      "grad_norm": 3.9392337799072266,
      "kl": 1.060546875,
      "learning_rate": 2.514668546453592e-07,
      "loss": 0.0589,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 682
    },
    {
      "completion_length": 786.2708435058594,
      "epoch": 0.8329268292682926,
      "grad_norm": 1.270626425743103,
      "kl": 0.720703125,
      "learning_rate": 2.4793914671147745e-07,
      "loss": 0.0595,
      "reward": 0.20833334140479565,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833334140479565,
      "rewards/format_reward": 0.0,
      "step": 683
    },
    {
      "completion_length": 780.4583587646484,
      "epoch": 0.8341463414634146,
      "grad_norm": 2.4505648612976074,
      "kl": 0.6016845703125,
      "learning_rate": 2.4443412759202745e-07,
      "loss": 0.0625,
      "reward": 0.27083334513008595,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.27083334513008595,
      "rewards/format_reward": 0.0,
      "step": 684
    },
    {
      "completion_length": 772.0833435058594,
      "epoch": 0.8353658536585366,
      "grad_norm": 2.4494102001190186,
      "kl": 0.7216796875,
      "learning_rate": 2.4095186080207505e-07,
      "loss": 0.0536,
      "reward": 0.14583333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 685
    },
    {
      "completion_length": 689.7500152587891,
      "epoch": 0.8365853658536585,
      "grad_norm": 3.0343613624572754,
      "kl": 0.493896484375,
      "learning_rate": 2.3749240944438845e-07,
      "loss": 0.0661,
      "reward": 0.14583333395421505,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 686
    },
    {
      "completion_length": 702.3125305175781,
      "epoch": 0.8378048780487805,
      "grad_norm": 1.3188613653182983,
      "kl": 0.396240234375,
      "learning_rate": 2.3405583620829268e-07,
      "loss": 0.0173,
      "reward": 0.18750000186264515,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 687
    },
    {
      "completion_length": 730.9375152587891,
      "epoch": 0.8390243902439024,
      "grad_norm": 1.1079875230789185,
      "kl": 0.32861328125,
      "learning_rate": 2.3064220336853398e-07,
      "loss": 0.0125,
      "reward": 0.2708333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2708333358168602,
      "rewards/format_reward": 0.0,
      "step": 688
    },
    {
      "completion_length": 691.3125152587891,
      "epoch": 0.8402439024390244,
      "grad_norm": 0.8602121472358704,
      "kl": 0.551025390625,
      "learning_rate": 2.272515727841527e-07,
      "loss": 0.0349,
      "reward": 0.2291666716337204,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 689
    },
    {
      "completion_length": 699.7708435058594,
      "epoch": 0.8414634146341463,
      "grad_norm": 1.7479124069213867,
      "kl": 0.4915771484375,
      "learning_rate": 2.2388400589735985e-07,
      "loss": 0.0263,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 690
    },
    {
      "completion_length": 663.7083435058594,
      "epoch": 0.8426829268292683,
      "grad_norm": 3.239636182785034,
      "kl": 0.284423828125,
      "learning_rate": 2.205395637324264e-07,
      "loss": 0.0536,
      "reward": 0.20833334513008595,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833334513008595,
      "rewards/format_reward": 0.0,
      "step": 691
    },
    {
      "completion_length": 659.7916870117188,
      "epoch": 0.8439024390243902,
      "grad_norm": 6.062473773956299,
      "kl": 0.7626953125,
      "learning_rate": 2.1721830689457583e-07,
      "loss": 0.113,
      "reward": 0.2291666753590107,
      "reward_std": 0.25259073823690414,
      "rewards/accuracy_reward": 0.2291666753590107,
      "rewards/format_reward": 0.0,
      "step": 692
    },
    {
      "completion_length": 719.8958587646484,
      "epoch": 0.8451219512195122,
      "grad_norm": 1.6376187801361084,
      "kl": 0.38720703125,
      "learning_rate": 2.1392029556888576e-07,
      "loss": 0.0381,
      "reward": 0.20833333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.20833333767950535,
      "rewards/format_reward": 0.0,
      "step": 693
    },
    {
      "completion_length": 823.1666717529297,
      "epoch": 0.8463414634146341,
      "grad_norm": 1.9714126586914062,
      "kl": 0.69970703125,
      "learning_rate": 2.1064558951919854e-07,
      "loss": 0.0414,
      "reward": 0.27083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.27083333395421505,
      "rewards/format_reward": 0.0,
      "step": 694
    },
    {
      "completion_length": 750.2916870117188,
      "epoch": 0.8475609756097561,
      "grad_norm": 2.118567705154419,
      "kl": 0.61474609375,
      "learning_rate": 2.0739424808703638e-07,
      "loss": 0.0811,
      "reward": 0.08333333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 695
    },
    {
      "completion_length": 758.2708587646484,
      "epoch": 0.848780487804878,
      "grad_norm": 1.5569214820861816,
      "kl": 0.415283203125,
      "learning_rate": 2.0416633019052882e-07,
      "loss": 0.0089,
      "reward": 0.1041666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 696
    },
    {
      "completion_length": 758.0625305175781,
      "epoch": 0.85,
      "grad_norm": 3.3884692192077637,
      "kl": 0.61474609375,
      "learning_rate": 2.0096189432334195e-07,
      "loss": 0.0744,
      "reward": 0.12500000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000558793545,
      "rewards/format_reward": 0.0,
      "step": 697
    },
    {
      "completion_length": 694.1041870117188,
      "epoch": 0.8512195121951219,
      "grad_norm": 0.6494787335395813,
      "kl": 0.60302734375,
      "learning_rate": 1.9778099855362085e-07,
      "loss": 0.0588,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 698
    },
    {
      "completion_length": 695.9375152587891,
      "epoch": 0.8524390243902439,
      "grad_norm": 1.7683531045913696,
      "kl": 0.3369140625,
      "learning_rate": 1.9462370052293544e-07,
      "loss": 0.0231,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 699
    },
    {
      "completion_length": 776.3333587646484,
      "epoch": 0.8536585365853658,
      "grad_norm": 0.5409860610961914,
      "kl": 0.4267578125,
      "learning_rate": 1.9149005744523757e-07,
      "loss": 0.0311,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 700
    },
    {
      "completion_length": 762.1250305175781,
      "epoch": 0.8548780487804878,
      "grad_norm": 0.9141250252723694,
      "kl": 0.82421875,
      "learning_rate": 1.8838012610582356e-07,
      "loss": 0.0723,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 701
    },
    {
      "completion_length": 789.7291870117188,
      "epoch": 0.8560975609756097,
      "grad_norm": 0.9757297039031982,
      "kl": 0.497161865234375,
      "learning_rate": 1.852939628603046e-07,
      "loss": 0.0526,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 702
    },
    {
      "completion_length": 757.1875152587891,
      "epoch": 0.8573170731707317,
      "grad_norm": 3.381772994995117,
      "kl": 0.356201171875,
      "learning_rate": 1.822316236335867e-07,
      "loss": 0.0197,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 703
    },
    {
      "completion_length": 828.5833435058594,
      "epoch": 0.8585365853658536,
      "grad_norm": 0.47453412413597107,
      "kl": 0.4620361328125,
      "learning_rate": 1.7919316391885593e-07,
      "loss": 0.0522,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 704
    },
    {
      "completion_length": 822.3750152587891,
      "epoch": 0.8597560975609756,
      "grad_norm": 1.0004751682281494,
      "kl": 0.5185546875,
      "learning_rate": 1.761786387765743e-07,
      "loss": 0.0494,
      "reward": 0.2708333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 705
    },
    {
      "completion_length": 730.0208435058594,
      "epoch": 0.8609756097560975,
      "grad_norm": 0.7801243662834167,
      "kl": 0.6240234375,
      "learning_rate": 1.731881028334808e-07,
      "loss": 0.0534,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 706
    },
    {
      "completion_length": 751.1667022705078,
      "epoch": 0.8621951219512195,
      "grad_norm": 1.9511408805847168,
      "kl": 0.4666748046875,
      "learning_rate": 1.7022161028160244e-07,
      "loss": 0.0207,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 707
    },
    {
      "completion_length": 747.3125152587891,
      "epoch": 0.8634146341463415,
      "grad_norm": 5.05046272277832,
      "kl": 0.394287109375,
      "learning_rate": 1.6727921487727095e-07,
      "loss": 0.0687,
      "reward": 0.31250000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.31250000558793545,
      "rewards/format_reward": 0.0,
      "step": 708
    },
    {
      "completion_length": 743.5625,
      "epoch": 0.8646341463414634,
      "grad_norm": 0.529568612575531,
      "kl": 0.326904296875,
      "learning_rate": 1.64360969940149e-07,
      "loss": 0.039,
      "reward": 0.2291666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 709
    },
    {
      "completion_length": 732.4791870117188,
      "epoch": 0.8658536585365854,
      "grad_norm": 0.8814006447792053,
      "kl": 0.47021484375,
      "learning_rate": 1.6146692835226669e-07,
      "loss": 0.0219,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 710
    },
    {
      "completion_length": 736.4166717529297,
      "epoch": 0.8670731707317073,
      "grad_norm": 0.6698026657104492,
      "kl": 0.392333984375,
      "learning_rate": 1.5859714255705843e-07,
      "loss": 0.0336,
      "reward": 0.16666666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 711
    },
    {
      "completion_length": 662.0625305175781,
      "epoch": 0.8682926829268293,
      "grad_norm": 1.3931370973587036,
      "kl": 0.399169921875,
      "learning_rate": 1.5575166455841677e-07,
      "loss": 0.0136,
      "reward": 0.1666666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 712
    },
    {
      "completion_length": 802.2083587646484,
      "epoch": 0.8695121951219512,
      "grad_norm": 1.0261173248291016,
      "kl": 0.261993408203125,
      "learning_rate": 1.5293054591974726e-07,
      "loss": 0.0599,
      "reward": 0.14583333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 713
    },
    {
      "completion_length": 778.1458435058594,
      "epoch": 0.8707317073170732,
      "grad_norm": 1.3354380130767822,
      "kl": 0.397705078125,
      "learning_rate": 1.501338377630362e-07,
      "loss": 0.0505,
      "reward": 0.18750000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 714
    },
    {
      "completion_length": 711.2500305175781,
      "epoch": 0.8719512195121951,
      "grad_norm": 1.3927069902420044,
      "kl": 0.1629638671875,
      "learning_rate": 1.473615907679229e-07,
      "loss": -0.0028,
      "reward": 0.18750000558793545,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 715
    },
    {
      "completion_length": 784.2291870117188,
      "epoch": 0.8731707317073171,
      "grad_norm": 2.6997110843658447,
      "kl": 0.197509765625,
      "learning_rate": 1.446138551707814e-07,
      "loss": 0.0568,
      "reward": 0.2500000037252903,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 716
    },
    {
      "completion_length": 750.4375305175781,
      "epoch": 0.874390243902439,
      "grad_norm": 32.169578552246094,
      "kl": 1.259765625,
      "learning_rate": 1.4189068076381078e-07,
      "loss": 0.0682,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 717
    },
    {
      "completion_length": 754.4792022705078,
      "epoch": 0.875609756097561,
      "grad_norm": 3.845160722732544,
      "kl": 0.26104736328125,
      "learning_rate": 1.3919211689413207e-07,
      "loss": 0.0618,
      "reward": 0.2916666828095913,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.2916666828095913,
      "rewards/format_reward": 0.0,
      "step": 718
    },
    {
      "completion_length": 664.5625305175781,
      "epoch": 0.8768292682926829,
      "grad_norm": 0.5239027142524719,
      "kl": 0.404541015625,
      "learning_rate": 1.365182124628949e-07,
      "loss": 0.0359,
      "reward": 0.25000000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.0,
      "step": 719
    },
    {
      "completion_length": 769.3333587646484,
      "epoch": 0.8780487804878049,
      "grad_norm": 1.6143977642059326,
      "kl": 0.484619140625,
      "learning_rate": 1.3386901592439071e-07,
      "loss": 0.0223,
      "reward": 0.1458333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 720
    },
    {
      "completion_length": 722.3541717529297,
      "epoch": 0.8792682926829268,
      "grad_norm": 0.4525848627090454,
      "kl": 0.17315673828125,
      "learning_rate": 1.3124457528517503e-07,
      "loss": 0.0333,
      "reward": 0.1250000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 721
    },
    {
      "completion_length": 731.5000152587891,
      "epoch": 0.8804878048780488,
      "grad_norm": 1.100823163986206,
      "kl": 0.58984375,
      "learning_rate": 1.2864493810319676e-07,
      "loss": 0.0164,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 722
    },
    {
      "completion_length": 735.6458435058594,
      "epoch": 0.8817073170731707,
      "grad_norm": 2.001176595687866,
      "kl": 0.326171875,
      "learning_rate": 1.260701514869379e-07,
      "loss": 0.0326,
      "reward": 0.29166667722165585,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.29166667722165585,
      "rewards/format_reward": 0.0,
      "step": 723
    },
    {
      "completion_length": 757.7083587646484,
      "epoch": 0.8829268292682927,
      "grad_norm": 0.9465669393539429,
      "kl": 0.4398193359375,
      "learning_rate": 1.2352026209455808e-07,
      "loss": 0.0664,
      "reward": 0.1458333395421505,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1458333395421505,
      "rewards/format_reward": 0.0,
      "step": 724
    },
    {
      "completion_length": 752.2500305175781,
      "epoch": 0.8841463414634146,
      "grad_norm": 1.317199945449829,
      "kl": 0.77099609375,
      "learning_rate": 1.209953161330507e-07,
      "loss": 0.0988,
      "reward": 0.2708333469927311,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333469927311,
      "rewards/format_reward": 0.0,
      "step": 725
    },
    {
      "completion_length": 752.4375152587891,
      "epoch": 0.8853658536585366,
      "grad_norm": 0.5425739288330078,
      "kl": 0.39697265625,
      "learning_rate": 1.1849535935740474e-07,
      "loss": 0.0161,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 726
    },
    {
      "completion_length": 827.1875152587891,
      "epoch": 0.8865853658536585,
      "grad_norm": 2.239192008972168,
      "kl": 0.5712890625,
      "learning_rate": 1.1602043706977538e-07,
      "loss": 0.0599,
      "reward": 0.10416666977107525,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 727
    },
    {
      "completion_length": 665.5416870117188,
      "epoch": 0.8878048780487805,
      "grad_norm": 1.820466160774231,
      "kl": 0.70849609375,
      "learning_rate": 1.1357059411866355e-07,
      "loss": 0.0401,
      "reward": 0.2291666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 728
    },
    {
      "completion_length": 753.875,
      "epoch": 0.8890243902439025,
      "grad_norm": 2.747743844985962,
      "kl": 0.5888671875,
      "learning_rate": 1.1114587489810352e-07,
      "loss": 0.0235,
      "reward": 0.1041666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666716337204,
      "rewards/format_reward": 0.0,
      "step": 729
    },
    {
      "completion_length": 718.8750305175781,
      "epoch": 0.8902439024390244,
      "grad_norm": 0.8431674838066101,
      "kl": 0.300994873046875,
      "learning_rate": 1.0874632334685808e-07,
      "loss": 0.0566,
      "reward": 0.1875,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 730
    },
    {
      "completion_length": 779.0208587646484,
      "epoch": 0.8914634146341464,
      "grad_norm": 3.0795443058013916,
      "kl": 0.773681640625,
      "learning_rate": 1.0637198294762152e-07,
      "loss": 0.0425,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 731
    },
    {
      "completion_length": 800.4791870117188,
      "epoch": 0.8926829268292683,
      "grad_norm": 0.7166914939880371,
      "kl": 0.5582275390625,
      "learning_rate": 1.0402289672623272e-07,
      "loss": 0.0654,
      "reward": 0.10416666977107525,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 732
    },
    {
      "completion_length": 778.3958587646484,
      "epoch": 0.8939024390243903,
      "grad_norm": 0.6740071177482605,
      "kl": 0.4248046875,
      "learning_rate": 1.0169910725089548e-07,
      "loss": 0.0049,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 733
    },
    {
      "completion_length": 676.5833587646484,
      "epoch": 0.8951219512195122,
      "grad_norm": 1.1036038398742676,
      "kl": 0.70068359375,
      "learning_rate": 9.940065663140663e-08,
      "loss": 0.0902,
      "reward": 0.12500000186264515,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 734
    },
    {
      "completion_length": 717.9583435058594,
      "epoch": 0.8963414634146342,
      "grad_norm": 0.5055205821990967,
      "kl": 0.14398193359375,
      "learning_rate": 9.71275865183936e-08,
      "loss": 0.0063,
      "reward": 0.2083333358168602,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 735
    },
    {
      "completion_length": 687.0833587646484,
      "epoch": 0.8975609756097561,
      "grad_norm": 3.46455717086792,
      "kl": 0.6573486328125,
      "learning_rate": 9.487993810255823e-08,
      "loss": 0.0914,
      "reward": 0.1875000074505806,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1875000074505806,
      "rewards/format_reward": 0.0,
      "step": 736
    },
    {
      "completion_length": 725.1250305175781,
      "epoch": 0.8987804878048781,
      "grad_norm": 0.7487300634384155,
      "kl": 0.57666015625,
      "learning_rate": 9.265775211393224e-08,
      "loss": 0.0516,
      "reward": 0.2708333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 737
    },
    {
      "completion_length": 766.2291870117188,
      "epoch": 0.9,
      "grad_norm": 1.0455366373062134,
      "kl": 0.75927734375,
      "learning_rate": 9.046106882113752e-08,
      "loss": 0.06,
      "reward": 0.18750000558793545,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 738
    },
    {
      "completion_length": 687.1041870117188,
      "epoch": 0.901219512195122,
      "grad_norm": 2.0819649696350098,
      "kl": 0.51953125,
      "learning_rate": 8.828992803065772e-08,
      "loss": 0.0089,
      "reward": 0.12500000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000558793545,
      "rewards/format_reward": 0.0,
      "step": 739
    },
    {
      "completion_length": 643.0208587646484,
      "epoch": 0.9024390243902439,
      "grad_norm": 2.3120899200439453,
      "kl": 0.48822021484375,
      "learning_rate": 8.614436908611617e-08,
      "loss": 0.0167,
      "reward": 0.2916666753590107,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2916666753590107,
      "rewards/format_reward": 0.0,
      "step": 740
    },
    {
      "completion_length": 755.9167022705078,
      "epoch": 0.9036585365853659,
      "grad_norm": 2.191702365875244,
      "kl": 0.646484375,
      "learning_rate": 8.402443086756273e-08,
      "loss": 0.0957,
      "reward": 0.16666666977107525,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 741
    },
    {
      "completion_length": 794.2500152587891,
      "epoch": 0.9048780487804878,
      "grad_norm": 0.6644926071166992,
      "kl": 0.5830078125,
      "learning_rate": 8.193015179076996e-08,
      "loss": 0.047,
      "reward": 0.18750000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 742
    },
    {
      "completion_length": 752.8958587646484,
      "epoch": 0.9060975609756098,
      "grad_norm": 0.8964568376541138,
      "kl": 0.432373046875,
      "learning_rate": 7.986156980653653e-08,
      "loss": 0.0522,
      "reward": 0.1250000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 743
    },
    {
      "completion_length": 801.6458587646484,
      "epoch": 0.9073170731707317,
      "grad_norm": 1.2347220182418823,
      "kl": 0.2987060546875,
      "learning_rate": 7.781872239999993e-08,
      "loss": 0.0081,
      "reward": 0.1458333432674408,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1458333432674408,
      "rewards/format_reward": 0.0,
      "step": 744
    },
    {
      "completion_length": 772.9583587646484,
      "epoch": 0.9085365853658537,
      "grad_norm": 3.951657295227051,
      "kl": 0.83935546875,
      "learning_rate": 7.580164658995603e-08,
      "loss": 0.1073,
      "reward": 0.1666666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 745
    },
    {
      "completion_length": 735.75,
      "epoch": 0.9097560975609756,
      "grad_norm": 0.8688337206840515,
      "kl": 0.4788818359375,
      "learning_rate": 7.381037892818959e-08,
      "loss": 0.0403,
      "reward": 0.16666666977107525,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 746
    },
    {
      "completion_length": 819.8125305175781,
      "epoch": 0.9109756097560976,
      "grad_norm": 1.4285547733306885,
      "kl": 1.01171875,
      "learning_rate": 7.184495549881131e-08,
      "loss": 0.0637,
      "reward": 0.08333333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 747
    },
    {
      "completion_length": 726.8750152587891,
      "epoch": 0.9121951219512195,
      "grad_norm": 1.8275103569030762,
      "kl": 0.52093505859375,
      "learning_rate": 6.990541191760418e-08,
      "loss": 0.0298,
      "reward": 0.3333333432674408,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.3333333432674408,
      "rewards/format_reward": 0.0,
      "step": 748
    },
    {
      "completion_length": 710.4166870117188,
      "epoch": 0.9134146341463415,
      "grad_norm": 0.7167033553123474,
      "kl": 0.5673828125,
      "learning_rate": 6.799178333137784e-08,
      "loss": 0.0471,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 749
    },
    {
      "completion_length": 811.5416870117188,
      "epoch": 0.9146341463414634,
      "grad_norm": 0.8002659678459167,
      "kl": 0.73193359375,
      "learning_rate": 6.610410441733156e-08,
      "loss": 0.0684,
      "reward": 0.0833333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.0833333358168602,
      "rewards/format_reward": 0.0,
      "step": 750
    },
    {
      "completion_length": 783.7291870117188,
      "epoch": 0.9158536585365854,
      "grad_norm": 1.8369277715682983,
      "kl": 0.44482421875,
      "learning_rate": 6.424240938242643e-08,
      "loss": 0.0205,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 751
    },
    {
      "completion_length": 792.6875152587891,
      "epoch": 0.9170731707317074,
      "grad_norm": 1.2638633251190186,
      "kl": 0.4395751953125,
      "learning_rate": 6.24067319627642e-08,
      "loss": 0.0185,
      "reward": 0.08333333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.08333333395421505,
      "rewards/format_reward": 0.0,
      "step": 752
    },
    {
      "completion_length": 750.7917022705078,
      "epoch": 0.9182926829268293,
      "grad_norm": 1.0106260776519775,
      "kl": 0.60400390625,
      "learning_rate": 6.059710542297824e-08,
      "loss": 0.0435,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 753
    },
    {
      "completion_length": 755.4375152587891,
      "epoch": 0.9195121951219513,
      "grad_norm": 1.0106931924819946,
      "kl": 0.4752197265625,
      "learning_rate": 5.8813562555628585e-08,
      "loss": 0.0563,
      "reward": 0.1041666679084301,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 754
    },
    {
      "completion_length": 696.3541870117188,
      "epoch": 0.9207317073170732,
      "grad_norm": 3.6543660163879395,
      "kl": 0.623291015625,
      "learning_rate": 5.7056135680607965e-08,
      "loss": 0.1148,
      "reward": 0.37500001303851604,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.37500001303851604,
      "rewards/format_reward": 0.0,
      "step": 755
    },
    {
      "completion_length": 713.9375,
      "epoch": 0.9219512195121952,
      "grad_norm": 1.7613054513931274,
      "kl": 0.49072265625,
      "learning_rate": 5.532485664455755e-08,
      "loss": 0.0768,
      "reward": 0.18750000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.0,
      "step": 756
    },
    {
      "completion_length": 740.7708587646484,
      "epoch": 0.9231707317073171,
      "grad_norm": 1.978395700454712,
      "kl": 0.74395751953125,
      "learning_rate": 5.3619756820288525e-08,
      "loss": 0.0677,
      "reward": 0.29166668094694614,
      "reward_std": 0.25259073823690414,
      "rewards/accuracy_reward": 0.29166668094694614,
      "rewards/format_reward": 0.0,
      "step": 757
    },
    {
      "completion_length": 724.7083435058594,
      "epoch": 0.9243902439024391,
      "grad_norm": 2.0498251914978027,
      "kl": 0.662353515625,
      "learning_rate": 5.194086710621404e-08,
      "loss": 0.0744,
      "reward": 0.1875000037252903,
      "reward_std": 0.18042196333408356,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 758
    },
    {
      "completion_length": 758.6666870117188,
      "epoch": 0.925609756097561,
      "grad_norm": 0.7953493595123291,
      "kl": 0.73193359375,
      "learning_rate": 5.0288217925789025e-08,
      "loss": 0.0538,
      "reward": 0.14583333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 759
    },
    {
      "completion_length": 685.9583435058594,
      "epoch": 0.926829268292683,
      "grad_norm": 1.0196458101272583,
      "kl": 0.30645751953125,
      "learning_rate": 4.86618392269596e-08,
      "loss": 0.051,
      "reward": 0.2291666716337204,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666716337204,
      "rewards/format_reward": 0.0,
      "step": 760
    },
    {
      "completion_length": 724.0416717529297,
      "epoch": 0.9280487804878049,
      "grad_norm": 0.9540734887123108,
      "kl": 0.635009765625,
      "learning_rate": 4.70617604816192e-08,
      "loss": 0.0645,
      "reward": 0.22916667722165585,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.22916667722165585,
      "rewards/format_reward": 0.0,
      "step": 761
    },
    {
      "completion_length": 729.1666870117188,
      "epoch": 0.9292682926829269,
      "grad_norm": 0.8325567841529846,
      "kl": 0.32177734375,
      "learning_rate": 4.54880106850758e-08,
      "loss": 0.0219,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 762
    },
    {
      "completion_length": 780.2708435058594,
      "epoch": 0.9304878048780488,
      "grad_norm": 1.175642728805542,
      "kl": 0.760986328125,
      "learning_rate": 4.394061835552554e-08,
      "loss": 0.0495,
      "reward": 0.1041666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 763
    },
    {
      "completion_length": 834.9583587646484,
      "epoch": 0.9317073170731708,
      "grad_norm": 0.5721205472946167,
      "kl": 0.5631103515625,
      "learning_rate": 4.2419611533536296e-08,
      "loss": 0.0408,
      "reward": 0.20833333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 764
    },
    {
      "completion_length": 781.1875152587891,
      "epoch": 0.9329268292682927,
      "grad_norm": 1.9766206741333008,
      "kl": 0.4287109375,
      "learning_rate": 4.0925017781539896e-08,
      "loss": 0.0507,
      "reward": 0.20833333767950535,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.20833333767950535,
      "rewards/format_reward": 0.0,
      "step": 765
    },
    {
      "completion_length": 777.0833587646484,
      "epoch": 0.9341463414634147,
      "grad_norm": 1.9794228076934814,
      "kl": 0.6573486328125,
      "learning_rate": 3.9456864183331557e-08,
      "loss": 0.0481,
      "reward": 0.0416666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 766
    },
    {
      "completion_length": 714.8750152587891,
      "epoch": 0.9353658536585366,
      "grad_norm": 1.3638488054275513,
      "kl": 0.39111328125,
      "learning_rate": 3.80151773435804e-08,
      "loss": -0.0019,
      "reward": 0.0416666679084301,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 767
    },
    {
      "completion_length": 674.1666870117188,
      "epoch": 0.9365853658536586,
      "grad_norm": 3.0777649879455566,
      "kl": 0.27197265625,
      "learning_rate": 3.659998338734671e-08,
      "loss": 0.0454,
      "reward": 0.1250000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 768
    },
    {
      "completion_length": 724.3541793823242,
      "epoch": 0.9378048780487804,
      "grad_norm": 1.6984995603561401,
      "kl": 0.4912109375,
      "learning_rate": 3.5211307959608475e-08,
      "loss": 0.0694,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 769
    },
    {
      "completion_length": 701.8750152587891,
      "epoch": 0.9390243902439024,
      "grad_norm": 0.6016904711723328,
      "kl": 0.2423095703125,
      "learning_rate": 3.3849176224796884e-08,
      "loss": 0.015,
      "reward": 0.1875000037252903,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 770
    },
    {
      "completion_length": 709.8958587646484,
      "epoch": 0.9402439024390243,
      "grad_norm": 2.3888635635375977,
      "kl": 0.70849609375,
      "learning_rate": 3.2513612866339916e-08,
      "loss": 0.0354,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 771
    },
    {
      "completion_length": 755.3750305175781,
      "epoch": 0.9414634146341463,
      "grad_norm": 0.9117576479911804,
      "kl": 0.60400390625,
      "learning_rate": 3.1204642086215817e-08,
      "loss": 0.0588,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 772
    },
    {
      "completion_length": 706.6666870117188,
      "epoch": 0.9426829268292682,
      "grad_norm": 1.184582233428955,
      "kl": 0.407470703125,
      "learning_rate": 2.992228760451349e-08,
      "loss": 0.0011,
      "reward": 0.1666666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.0,
      "step": 773
    },
    {
      "completion_length": 784.1667022705078,
      "epoch": 0.9439024390243902,
      "grad_norm": 2.701702356338501,
      "kl": 0.5849609375,
      "learning_rate": 2.8666572659003965e-08,
      "loss": 0.0352,
      "reward": 0.0416666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.0,
      "step": 774
    },
    {
      "completion_length": 787.3333587646484,
      "epoch": 0.9451219512195121,
      "grad_norm": 1.0537127256393433,
      "kl": 0.418212890625,
      "learning_rate": 2.743752000471761e-08,
      "loss": 0.0431,
      "reward": 0.2916666716337204,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2916666716337204,
      "rewards/format_reward": 0.0,
      "step": 775
    },
    {
      "completion_length": 813.5416870117188,
      "epoch": 0.9463414634146341,
      "grad_norm": 1.9663876295089722,
      "kl": 0.4765625,
      "learning_rate": 2.6235151913533595e-08,
      "loss": 0.0227,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 776
    },
    {
      "completion_length": 758.5208435058594,
      "epoch": 0.947560975609756,
      "grad_norm": 1.3560831546783447,
      "kl": 0.558349609375,
      "learning_rate": 2.50594901737749e-08,
      "loss": 0.0667,
      "reward": 0.10416666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.0,
      "step": 777
    },
    {
      "completion_length": 775.7083587646484,
      "epoch": 0.948780487804878,
      "grad_norm": 2.003946542739868,
      "kl": 0.5240478515625,
      "learning_rate": 2.3910556089814294e-08,
      "loss": 0.082,
      "reward": 0.20833333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 778
    },
    {
      "completion_length": 815.0625152587891,
      "epoch": 0.95,
      "grad_norm": 0.549389123916626,
      "kl": 0.6201171875,
      "learning_rate": 2.278837048168797e-08,
      "loss": 0.0434,
      "reward": 0.1666666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1666666679084301,
      "rewards/format_reward": 0.0,
      "step": 779
    },
    {
      "completion_length": 785.8750305175781,
      "epoch": 0.9512195121951219,
      "grad_norm": 2.0170860290527344,
      "kl": 0.4798583984375,
      "learning_rate": 2.1692953684718187e-08,
      "loss": 0.0692,
      "reward": 0.12500000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 780
    },
    {
      "completion_length": 719.3125152587891,
      "epoch": 0.9524390243902439,
      "grad_norm": 3.6892948150634766,
      "kl": 0.370849609375,
      "learning_rate": 2.0624325549144894e-08,
      "loss": 0.0566,
      "reward": 0.2500000074505806,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.0,
      "step": 781
    },
    {
      "completion_length": 701.4583587646484,
      "epoch": 0.9536585365853658,
      "grad_norm": 0.8499000668525696,
      "kl": 0.48046875,
      "learning_rate": 1.9582505439766028e-08,
      "loss": 0.0366,
      "reward": 0.1041666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666716337204,
      "rewards/format_reward": 0.0,
      "step": 782
    },
    {
      "completion_length": 654.6250152587891,
      "epoch": 0.9548780487804878,
      "grad_norm": 5.284634590148926,
      "kl": 0.404052734375,
      "learning_rate": 1.856751223558695e-08,
      "loss": 0.1073,
      "reward": 0.18750000558793545,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 783
    },
    {
      "completion_length": 681.1250305175781,
      "epoch": 0.9560975609756097,
      "grad_norm": 1.9143365621566772,
      "kl": 0.8544921875,
      "learning_rate": 1.7579364329477375e-08,
      "loss": 0.115,
      "reward": 0.2291666753590107,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2291666753590107,
      "rewards/format_reward": 0.0,
      "step": 784
    },
    {
      "completion_length": 670.1250152587891,
      "epoch": 0.9573170731707317,
      "grad_norm": 2.3851420879364014,
      "kl": 0.52734375,
      "learning_rate": 1.661807962783851e-08,
      "loss": 0.0351,
      "reward": 0.2291666679084301,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 785
    },
    {
      "completion_length": 734.6041870117188,
      "epoch": 0.9585365853658536,
      "grad_norm": 0.8727129101753235,
      "kl": 0.4429931640625,
      "learning_rate": 1.5683675550279943e-08,
      "loss": 0.0712,
      "reward": 0.1250000037252903,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.0,
      "step": 786
    },
    {
      "completion_length": 698.4791870117188,
      "epoch": 0.9597560975609756,
      "grad_norm": 2.6234490871429443,
      "kl": 0.8798828125,
      "learning_rate": 1.4776169029301234e-08,
      "loss": 0.0548,
      "reward": 0.37500000558793545,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.37500000558793545,
      "rewards/format_reward": 0.0,
      "step": 787
    },
    {
      "completion_length": 768.4166870117188,
      "epoch": 0.9609756097560975,
      "grad_norm": 0.8236937522888184,
      "kl": 1.14892578125,
      "learning_rate": 1.3895576509987685e-08,
      "loss": 0.0355,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 788
    },
    {
      "completion_length": 775.7500305175781,
      "epoch": 0.9621951219512195,
      "grad_norm": 0.9706014394760132,
      "kl": 0.458740234375,
      "learning_rate": 1.3041913949710715e-08,
      "loss": 0.0661,
      "reward": 0.1875000037252903,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.0,
      "step": 789
    },
    {
      "completion_length": 641.2083587646484,
      "epoch": 0.9634146341463414,
      "grad_norm": 2.8741683959960938,
      "kl": 0.36572265625,
      "learning_rate": 1.2215196817839447e-08,
      "loss": 0.0772,
      "reward": 0.27083334885537624,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.27083334885537624,
      "rewards/format_reward": 0.0,
      "step": 790
    },
    {
      "completion_length": 790.1666870117188,
      "epoch": 0.9646341463414634,
      "grad_norm": 1.9514979124069214,
      "kl": 0.64794921875,
      "learning_rate": 1.1415440095460083e-08,
      "loss": 0.0398,
      "reward": 0.14583333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.14583333395421505,
      "rewards/format_reward": 0.0,
      "step": 791
    },
    {
      "completion_length": 723.4375152587891,
      "epoch": 0.9658536585365853,
      "grad_norm": 2.7980079650878906,
      "kl": 1.38134765625,
      "learning_rate": 1.06426582751043e-08,
      "loss": 0.0774,
      "reward": 0.29166666977107525,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.29166666977107525,
      "rewards/format_reward": 0.0,
      "step": 792
    },
    {
      "completion_length": 750.5208740234375,
      "epoch": 0.9670731707317073,
      "grad_norm": 0.7902849316596985,
      "kl": 0.428466796875,
      "learning_rate": 9.896865360487451e-09,
      "loss": 0.0546,
      "reward": 0.18750000558793545,
      "reward_std": 0.14433757215738297,
      "rewards/accuracy_reward": 0.18750000558793545,
      "rewards/format_reward": 0.0,
      "step": 793
    },
    {
      "completion_length": 741.9791717529297,
      "epoch": 0.9682926829268292,
      "grad_norm": 0.7297914624214172,
      "kl": 0.4921875,
      "learning_rate": 9.178074866253605e-09,
      "loss": 0.0424,
      "reward": 0.1041666716337204,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.1041666716337204,
      "rewards/format_reward": 0.0,
      "step": 794
    },
    {
      "completion_length": 660.7291870117188,
      "epoch": 0.9695121951219512,
      "grad_norm": 2.9257161617279053,
      "kl": 0.7255859375,
      "learning_rate": 8.486299817731412e-09,
      "loss": 0.0601,
      "reward": 0.2916666679084301,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.2916666679084301,
      "rewards/format_reward": 0.0,
      "step": 795
    },
    {
      "completion_length": 660.2083435058594,
      "epoch": 0.9707317073170731,
      "grad_norm": 0.8202441930770874,
      "kl": 0.590087890625,
      "learning_rate": 7.821552750697958e-09,
      "loss": 0.0282,
      "reward": 0.2083333358168602,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2083333358168602,
      "rewards/format_reward": 0.0,
      "step": 796
    },
    {
      "completion_length": 750.8333587646484,
      "epoch": 0.9719512195121951,
      "grad_norm": 0.7230677604675293,
      "kl": 0.43408203125,
      "learning_rate": 7.1838457111516044e-09,
      "loss": 0.0408,
      "reward": 0.12500000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.0,
      "step": 797
    },
    {
      "completion_length": 703.8958587646484,
      "epoch": 0.973170731707317,
      "grad_norm": 1.1173596382141113,
      "kl": 0.443603515625,
      "learning_rate": 6.573190255093342e-09,
      "loss": 0.0003,
      "reward": 0.14583333767950535,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.0,
      "step": 798
    },
    {
      "completion_length": 736.7083740234375,
      "epoch": 0.974390243902439,
      "grad_norm": 1.5042527914047241,
      "kl": 0.568359375,
      "learning_rate": 5.989597448317785e-09,
      "loss": 0.0712,
      "reward": 0.20833334140479565,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.20833334140479565,
      "rewards/format_reward": 0.0,
      "step": 799
    },
    {
      "completion_length": 713.4375152587891,
      "epoch": 0.975609756097561,
      "grad_norm": 1.9430745840072632,
      "kl": 0.608642578125,
      "learning_rate": 5.433077866212999e-09,
      "loss": 0.0007,
      "reward": 0.29166667722165585,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.29166667722165585,
      "rewards/format_reward": 0.0,
      "step": 800
    },
    {
      "completion_length": 789.7917022705078,
      "epoch": 0.9768292682926829,
      "grad_norm": 1.0729237794876099,
      "kl": 0.7138671875,
      "learning_rate": 4.903641593567654e-09,
      "loss": 0.072,
      "reward": 0.20833333395421505,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.20833333395421505,
      "rewards/format_reward": 0.0,
      "step": 801
    },
    {
      "completion_length": 639.8541870117188,
      "epoch": 0.9780487804878049,
      "grad_norm": 1.2329329252243042,
      "kl": 0.2347412109375,
      "learning_rate": 4.401298224389338e-09,
      "loss": 0.0134,
      "reward": 0.20833334140479565,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.20833334140479565,
      "rewards/format_reward": 0.0,
      "step": 802
    },
    {
      "completion_length": 719.3750305175781,
      "epoch": 0.9792682926829268,
      "grad_norm": 1.0990557670593262,
      "kl": 0.3681640625,
      "learning_rate": 3.926056861730532e-09,
      "loss": 0.0401,
      "reward": 0.16666666977107525,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.16666666977107525,
      "rewards/format_reward": 0.0,
      "step": 803
    },
    {
      "completion_length": 682.5625305175781,
      "epoch": 0.9804878048780488,
      "grad_norm": 0.3879597783088684,
      "kl": 0.48486328125,
      "learning_rate": 3.4779261175232334e-09,
      "loss": 0.0504,
      "reward": 0.1041666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.0,
      "step": 804
    },
    {
      "completion_length": 667.8333435058594,
      "epoch": 0.9817073170731707,
      "grad_norm": 2.3464221954345703,
      "kl": 0.2630615234375,
      "learning_rate": 3.0569141124234256e-09,
      "loss": 0.0625,
      "reward": 0.1458333358168602,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.0,
      "step": 805
    },
    {
      "completion_length": 792.3750152587891,
      "epoch": 0.9829268292682927,
      "grad_norm": 2.1872403621673584,
      "kl": 0.75390625,
      "learning_rate": 2.6630284756635204e-09,
      "loss": 0.0494,
      "reward": 0.02083333395421505,
      "reward_std": 0.03608439117670059,
      "rewards/accuracy_reward": 0.02083333395421505,
      "rewards/format_reward": 0.0,
      "step": 806
    },
    {
      "completion_length": 656.9583435058594,
      "epoch": 0.9841463414634146,
      "grad_norm": 4.716466426849365,
      "kl": 0.3326416015625,
      "learning_rate": 2.2962763449141387e-09,
      "loss": 0.072,
      "reward": 0.3541666716337204,
      "reward_std": 0.18042195588350296,
      "rewards/accuracy_reward": 0.3541666716337204,
      "rewards/format_reward": 0.0,
      "step": 807
    },
    {
      "completion_length": 699.7916717529297,
      "epoch": 0.9853658536585366,
      "grad_norm": 1.0867366790771484,
      "kl": 0.4921875,
      "learning_rate": 1.9566643661550478e-09,
      "loss": 0.0286,
      "reward": 0.2291666679084301,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2291666679084301,
      "rewards/format_reward": 0.0,
      "step": 808
    },
    {
      "completion_length": 730.2708587646484,
      "epoch": 0.9865853658536585,
      "grad_norm": 1.8666948080062866,
      "kl": 0.46484375,
      "learning_rate": 1.6441986935545884e-09,
      "loss": 0.023,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 809
    },
    {
      "completion_length": 771.6666870117188,
      "epoch": 0.9878048780487805,
      "grad_norm": 2.122633695602417,
      "kl": 0.48486328125,
      "learning_rate": 1.3588849893579336e-09,
      "loss": 0.0223,
      "reward": 0.0625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.0,
      "step": 810
    },
    {
      "completion_length": 772.0208435058594,
      "epoch": 0.9890243902439024,
      "grad_norm": 1.248072862625122,
      "kl": 0.41845703125,
      "learning_rate": 1.1007284237850025e-09,
      "loss": 0.0442,
      "reward": 0.06250000186264515,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 811
    },
    {
      "completion_length": 777.9375305175781,
      "epoch": 0.9902439024390244,
      "grad_norm": 1.3395018577575684,
      "kl": 0.4093017578125,
      "learning_rate": 8.697336749358687e-10,
      "loss": 0.0125,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 812
    },
    {
      "completion_length": 623.6458587646484,
      "epoch": 0.9914634146341463,
      "grad_norm": 3.9482204914093018,
      "kl": 0.322265625,
      "learning_rate": 6.659049287071617e-10,
      "loss": 0.0476,
      "reward": 0.27083334140479565,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.27083334140479565,
      "rewards/format_reward": 0.0,
      "step": 813
    },
    {
      "completion_length": 754.0416717529297,
      "epoch": 0.9926829268292683,
      "grad_norm": 1.1337177753448486,
      "kl": 0.54931640625,
      "learning_rate": 4.892458787154608e-10,
      "loss": 0.0632,
      "reward": 0.06250000186264515,
      "reward_std": 0.10825317353010178,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.0,
      "step": 814
    },
    {
      "completion_length": 757.7291717529297,
      "epoch": 0.9939024390243902,
      "grad_norm": 0.8009763360023499,
      "kl": 0.427734375,
      "learning_rate": 3.397597262300156e-10,
      "loss": 0.032,
      "reward": 0.2708333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2708333395421505,
      "rewards/format_reward": 0.0,
      "step": 815
    },
    {
      "completion_length": 706.5208587646484,
      "epoch": 0.9951219512195122,
      "grad_norm": 3.026621103286743,
      "kl": 0.70458984375,
      "learning_rate": 2.1744918011595837e-10,
      "loss": 0.046,
      "reward": 0.2083333395421505,
      "reward_std": 0.07216878235340118,
      "rewards/accuracy_reward": 0.2083333395421505,
      "rewards/format_reward": 0.0,
      "step": 816
    },
    {
      "completion_length": 699.5416717529297,
      "epoch": 0.9963414634146341,
      "grad_norm": 3.1237189769744873,
      "kl": 0.54345703125,
      "learning_rate": 1.2231645678401072e-10,
      "loss": 0.1144,
      "reward": 0.2916666753590107,
      "reward_std": 0.14433756470680237,
      "rewards/accuracy_reward": 0.2916666753590107,
      "rewards/format_reward": 0.0,
      "step": 817
    },
    {
      "completion_length": 773.9791870117188,
      "epoch": 0.9975609756097561,
      "grad_norm": 1.6994426250457764,
      "kl": 0.4207763671875,
      "learning_rate": 5.436328015101522e-11,
      "loss": 0.0196,
      "reward": 0.1875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 0.0,
      "step": 818
    },
    {
      "completion_length": 756.7500305175781,
      "epoch": 0.998780487804878,
      "grad_norm": 0.8073121905326843,
      "kl": 0.183837890625,
      "learning_rate": 1.359088160846067e-11,
      "loss": 0.0083,
      "reward": 0.125,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.125,
      "rewards/format_reward": 0.0,
      "step": 819
    },
    {
      "completion_length": 847.75,
      "epoch": 1.0,
      "grad_norm": 3.6782705783843994,
      "kl": 0.670654296875,
      "learning_rate": 0.0,
      "loss": 0.0836,
      "reward": 0.29166667349636555,
      "reward_std": 0.21650634706020355,
      "rewards/accuracy_reward": 0.29166667349636555,
      "rewards/format_reward": 0.0,
      "step": 820
    },
    {
      "epoch": 1.0,
      "step": 820,
      "total_flos": 0.0,
      "train_loss": 0.032500523863215745,
      "train_runtime": 33634.9613,
      "train_samples_per_second": 0.39,
      "train_steps_per_second": 0.024
    }
  ],
  "logging_steps": 1,
  "max_steps": 820,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}