{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 10,
  "global_step": 938,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 121.971875,
      "epoch": 0.010666666666666666,
      "grad_norm": 0.156667098402977,
      "kl": 2.0313262939453126e-05,
      "learning_rate": 1.0638297872340426e-05,
      "loss": 0.001,
      "reward": 0.0125,
      "reward_std": 0.025,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.003125,
      "step": 10
    },
    {
      "completion_length": 122.521875,
      "epoch": 0.021333333333333333,
      "grad_norm": 0.0012713409960269928,
      "kl": 0.00021836161613464355,
      "learning_rate": 2.1276595744680852e-05,
      "loss": 0.0051,
      "reward": 0.015625,
      "reward_std": 0.025966878235340118,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "completion_length": 117.5125,
      "epoch": 0.032,
      "grad_norm": 0.002654253738000989,
      "kl": 0.0003068089485168457,
      "learning_rate": 3.1914893617021275e-05,
      "loss": -0.0002,
      "reward": 0.00625,
      "reward_std": 0.007216878235340118,
      "rewards/accuracy_reward": 0.00625,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 118.871875,
      "epoch": 0.042666666666666665,
      "grad_norm": 0.00353299081325531,
      "kl": 0.000412750244140625,
      "learning_rate": 4.2553191489361704e-05,
      "loss": 0.0055,
      "reward": 0.009375,
      "reward_std": 0.01875,
      "rewards/accuracy_reward": 0.00625,
      "rewards/format_reward": 0.003125,
      "step": 40
    },
    {
      "completion_length": 121.046875,
      "epoch": 0.05333333333333334,
      "grad_norm": 0.003619612194597721,
      "kl": 0.0004070043563842773,
      "learning_rate": 5.319148936170213e-05,
      "loss": 0.0,
      "reward": 0.0,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "completion_length": 119.4375,
      "epoch": 0.064,
      "grad_norm": 0.11174867302179337,
      "kl": 0.00045168399810791016,
      "learning_rate": 6.382978723404255e-05,
      "loss": 0.0064,
      "reward": 0.01875,
      "reward_std": 0.0375,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.003125,
      "step": 60
    },
    {
      "completion_length": 119.05625,
      "epoch": 0.07466666666666667,
      "grad_norm": 0.006828859448432922,
      "kl": 0.0011888980865478516,
      "learning_rate": 7.446808510638297e-05,
      "loss": 0.0012,
      "reward": 0.0125,
      "reward_std": 0.025,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "completion_length": 120.28125,
      "epoch": 0.08533333333333333,
      "grad_norm": 0.0064537739381194115,
      "kl": 0.0019659996032714844,
      "learning_rate": 8.510638297872341e-05,
      "loss": 0.0028,
      "reward": 0.0125,
      "reward_std": 0.025,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "completion_length": 117.559375,
      "epoch": 0.096,
      "grad_norm": 0.09068689495325089,
      "kl": 0.0025023460388183595,
      "learning_rate": 9.574468085106382e-05,
      "loss": 0.003,
      "reward": 0.021875,
      "reward_std": 0.04375,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.003125,
      "step": 90
    },
    {
      "completion_length": 117.68125,
      "epoch": 0.10666666666666667,
      "grad_norm": 0.16541939973831177,
      "kl": 0.00291900634765625,
      "learning_rate": 0.00010638297872340425,
      "loss": 0.0008,
      "reward": 0.021875,
      "reward_std": 0.03125,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "completion_length": 116.971875,
      "epoch": 0.11733333333333333,
      "grad_norm": 0.07206544280052185,
      "kl": 0.0038990020751953126,
      "learning_rate": 0.00011702127659574467,
      "loss": 0.0026,
      "reward": 0.015625,
      "reward_std": 0.03125,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.003125,
      "step": 110
    },
    {
      "completion_length": 114.996875,
      "epoch": 0.128,
      "grad_norm": 0.02286006510257721,
      "kl": 0.007346725463867188,
      "learning_rate": 0.0001276595744680851,
      "loss": 0.0076,
      "reward": 0.025,
      "reward_std": 0.05,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.009375,
      "step": 120
    },
    {
      "completion_length": 119.315625,
      "epoch": 0.13866666666666666,
      "grad_norm": 0.015629781410098076,
      "kl": 0.008090972900390625,
      "learning_rate": 0.00013829787234042552,
      "loss": 0.0011,
      "reward": 0.009375,
      "reward_std": 0.01875,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "completion_length": 121.821875,
      "epoch": 0.14933333333333335,
      "grad_norm": 0.15498439967632294,
      "kl": 0.006272506713867187,
      "learning_rate": 0.00014893617021276593,
      "loss": -0.0012,
      "reward": 0.021875,
      "reward_std": 0.03846687823534012,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.003125,
      "step": 140
    },
    {
      "completion_length": 121.409375,
      "epoch": 0.16,
      "grad_norm": 0.18756870925426483,
      "kl": 0.00465240478515625,
      "learning_rate": 0.00015957446808510637,
      "loss": 0.0012,
      "reward": 0.021875,
      "reward_std": 0.03846687823534012,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.003125,
      "step": 150
    },
    {
      "completion_length": 118.26875,
      "epoch": 0.17066666666666666,
      "grad_norm": 0.011626984924077988,
      "kl": 0.01092681884765625,
      "learning_rate": 0.00017021276595744682,
      "loss": -0.0011,
      "reward": 0.021875,
      "reward_std": 0.03318375647068024,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.003125,
      "step": 160
    },
    {
      "completion_length": 119.25,
      "epoch": 0.18133333333333335,
      "grad_norm": 0.00764912273734808,
      "kl": 0.00976104736328125,
      "learning_rate": 0.0001808510638297872,
      "loss": 0.0045,
      "reward": 0.021875,
      "reward_std": 0.04375,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "completion_length": 115.35,
      "epoch": 0.192,
      "grad_norm": 0.0785018652677536,
      "kl": 0.014077377319335938,
      "learning_rate": 0.00019148936170212765,
      "loss": 0.0037,
      "reward": 0.025,
      "reward_std": 0.04471687823534012,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.015625,
      "step": 180
    },
    {
      "completion_length": 108.6125,
      "epoch": 0.20266666666666666,
      "grad_norm": 0.13107918202877045,
      "kl": 0.039361572265625,
      "learning_rate": 0.00020212765957446807,
      "loss": 0.0412,
      "reward": 0.11875,
      "reward_std": 0.18907372057437896,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.096875,
      "step": 190
    },
    {
      "completion_length": 89.915625,
      "epoch": 0.21333333333333335,
      "grad_norm": 0.19012346863746643,
      "kl": 0.08895263671875,
      "learning_rate": 0.0002127659574468085,
      "loss": 0.1321,
      "reward": 0.46875,
      "reward_std": 0.41404569447040557,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.453125,
      "step": 200
    },
    {
      "completion_length": 47.74375,
      "epoch": 0.224,
      "grad_norm": 0.4668453335762024,
      "kl": 0.26416015625,
      "learning_rate": 0.0002234042553191489,
      "loss": 0.0712,
      "reward": 0.871875,
      "reward_std": 0.19805223047733306,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.853125,
      "step": 210
    },
    {
      "completion_length": 45.15625,
      "epoch": 0.23466666666666666,
      "grad_norm": 0.21052278578281403,
      "kl": 0.3112213134765625,
      "learning_rate": 0.00023404255319148934,
      "loss": 0.0464,
      "reward": 0.890625,
      "reward_std": 0.11346687823534012,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.878125,
      "step": 220
    },
    {
      "completion_length": 57.2875,
      "epoch": 0.24533333333333332,
      "grad_norm": 0.16618619859218597,
      "kl": 0.254522705078125,
      "learning_rate": 0.00024468085106382976,
      "loss": 0.0589,
      "reward": 0.834375,
      "reward_std": 0.12261751294136047,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.821875,
      "step": 230
    },
    {
      "completion_length": 68.153125,
      "epoch": 0.256,
      "grad_norm": 0.17739807069301605,
      "kl": 0.214471435546875,
      "learning_rate": 0.0002553191489361702,
      "loss": 0.1375,
      "reward": 0.659375,
      "reward_std": 0.28527562469244006,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.65,
      "step": 240
    },
    {
      "completion_length": 52.709375,
      "epoch": 0.26666666666666666,
      "grad_norm": 0.09843996912240982,
      "kl": 0.2847900390625,
      "learning_rate": 0.0002659574468085106,
      "loss": 0.1085,
      "reward": 0.834375,
      "reward_std": 0.290549997985363,
      "rewards/accuracy_reward": 0.034375,
      "rewards/format_reward": 0.8,
      "step": 250
    },
    {
      "completion_length": 51.55625,
      "epoch": 0.2773333333333333,
      "grad_norm": 0.1133696436882019,
      "kl": 0.276953125,
      "learning_rate": 0.00027659574468085103,
      "loss": 0.0437,
      "reward": 0.903125,
      "reward_std": 0.1361730858683586,
      "rewards/accuracy_reward": 0.025,
      "rewards/format_reward": 0.878125,
      "step": 260
    },
    {
      "completion_length": 55.046875,
      "epoch": 0.288,
      "grad_norm": 0.14536090195178986,
      "kl": 0.2501953125,
      "learning_rate": 0.0002872340425531915,
      "loss": 0.0588,
      "reward": 0.878125,
      "reward_std": 0.13846687823534012,
      "rewards/accuracy_reward": 0.00625,
      "rewards/format_reward": 0.871875,
      "step": 270
    },
    {
      "completion_length": 55.853125,
      "epoch": 0.2986666666666667,
      "grad_norm": 0.1799221634864807,
      "kl": 0.3143310546875,
      "learning_rate": 0.00029787234042553186,
      "loss": 0.0609,
      "reward": 0.90625,
      "reward_std": 0.18080126941204072,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.884375,
      "step": 280
    },
    {
      "completion_length": 59.85,
      "epoch": 0.30933333333333335,
      "grad_norm": 0.10688479989767075,
      "kl": 0.20706787109375,
      "learning_rate": 0.0002999925930442553,
      "loss": 0.0522,
      "reward": 0.815625,
      "reward_std": 0.2959165498614311,
      "rewards/accuracy_reward": 0.065625,
      "rewards/format_reward": 0.75,
      "step": 290
    },
    {
      "completion_length": 64.625,
      "epoch": 0.32,
      "grad_norm": 0.03851361572742462,
      "kl": 0.201220703125,
      "learning_rate": 0.00029996250354024344,
      "loss": 0.0815,
      "reward": 0.8625,
      "reward_std": 0.21301814764738083,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.85,
      "step": 300
    },
    {
      "completion_length": 57.95,
      "epoch": 0.33066666666666666,
      "grad_norm": 0.23480646312236786,
      "kl": 0.221240234375,
      "learning_rate": 0.0002999092731927958,
      "loss": 0.0292,
      "reward": 0.921875,
      "reward_std": 0.15895397514104842,
      "rewards/accuracy_reward": 0.0375,
      "rewards/format_reward": 0.884375,
      "step": 310
    },
    {
      "completion_length": 64.196875,
      "epoch": 0.3413333333333333,
      "grad_norm": 0.1151675432920456,
      "kl": 0.20123291015625,
      "learning_rate": 0.0002998329102159332,
      "loss": 0.0491,
      "reward": 0.83125,
      "reward_std": 0.19258119761943818,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.8125,
      "step": 320
    },
    {
      "completion_length": 70.628125,
      "epoch": 0.352,
      "grad_norm": 0.1377689391374588,
      "kl": 0.1906005859375,
      "learning_rate": 0.0002997334263932927,
      "loss": 0.0841,
      "reward": 0.846875,
      "reward_std": 0.21890811175107955,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.83125,
      "step": 330
    },
    {
      "completion_length": 61.54375,
      "epoch": 0.3626666666666667,
      "grad_norm": 0.0947548896074295,
      "kl": 0.21240234375,
      "learning_rate": 0.0002996108370763087,
      "loss": 0.062,
      "reward": 0.88125,
      "reward_std": 0.13713996410369872,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.865625,
      "step": 340
    },
    {
      "completion_length": 60.109375,
      "epoch": 0.37333333333333335,
      "grad_norm": 0.14599719643592834,
      "kl": 0.2236083984375,
      "learning_rate": 0.0002994651611818448,
      "loss": 0.0408,
      "reward": 0.928125,
      "reward_std": 0.18282372057437896,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.9,
      "step": 350
    },
    {
      "completion_length": 62.90625,
      "epoch": 0.384,
      "grad_norm": 0.3738599121570587,
      "kl": 0.2464111328125,
      "learning_rate": 0.00029929642118927394,
      "loss": 0.0753,
      "reward": 0.834375,
      "reward_std": 0.20676814764738083,
      "rewards/accuracy_reward": 0.025,
      "rewards/format_reward": 0.809375,
      "step": 360
    },
    {
      "completion_length": 70.1625,
      "epoch": 0.39466666666666667,
      "grad_norm": 2.8762810230255127,
      "kl": 0.88681640625,
      "learning_rate": 0.00029910464313701013,
      "loss": 0.2053,
      "reward": 0.640625,
      "reward_std": 0.38192625939846037,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.63125,
      "step": 370
    },
    {
      "completion_length": 43.578125,
      "epoch": 0.4053333333333333,
      "grad_norm": 0.9355350136756897,
      "kl": 2.06865234375,
      "learning_rate": 0.0002988898566184902,
      "loss": 0.2631,
      "reward": 0.725,
      "reward_std": 0.32462068647146225,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.725,
      "step": 380
    },
    {
      "completion_length": 44.19375,
      "epoch": 0.416,
      "grad_norm": 0.709173858165741,
      "kl": 3.47939453125,
      "learning_rate": 0.0002986520947776074,
      "loss": 0.3225,
      "reward": 0.6125,
      "reward_std": 0.3950331017374992,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.603125,
      "step": 390
    },
    {
      "completion_length": 54.775,
      "epoch": 0.4266666666666667,
      "grad_norm": 0.6549698114395142,
      "kl": 4.3202392578125,
      "learning_rate": 0.0002983913943035968,
      "loss": 0.3808,
      "reward": 0.66875,
      "reward_std": 0.3901100158691406,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.65,
      "step": 400
    },
    {
      "completion_length": 53.871875,
      "epoch": 0.43733333333333335,
      "grad_norm": 0.01826515607535839,
      "kl": 2.477734375,
      "learning_rate": 0.00029810779542537355,
      "loss": 0.2661,
      "reward": 0.79375,
      "reward_std": 0.22999776750802994,
      "rewards/accuracy_reward": 0.025,
      "rewards/format_reward": 0.76875,
      "step": 410
    },
    {
      "completion_length": 49.434375,
      "epoch": 0.448,
      "grad_norm": 0.5134692192077637,
      "kl": 2.07587890625,
      "learning_rate": 0.0002978013419053255,
      "loss": 0.2091,
      "reward": 0.771875,
      "reward_std": 0.26785253882408144,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.771875,
      "step": 420
    },
    {
      "completion_length": 59.475,
      "epoch": 0.45866666666666667,
      "grad_norm": 0.7835673689842224,
      "kl": 2.516943359375,
      "learning_rate": 0.00029747208103256,
      "loss": 0.2312,
      "reward": 0.740625,
      "reward_std": 0.31220938116312025,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.728125,
      "step": 430
    },
    {
      "completion_length": 58.815625,
      "epoch": 0.4693333333333333,
      "grad_norm": 0.021143430843949318,
      "kl": 2.1959716796875,
      "learning_rate": 0.0002971200636156068,
      "loss": 0.2386,
      "reward": 0.796875,
      "reward_std": 0.2231356605887413,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.796875,
      "step": 440
    },
    {
      "completion_length": 55.696875,
      "epoch": 0.48,
      "grad_norm": 1.8231980800628662,
      "kl": 2.9314697265625,
      "learning_rate": 0.00029674534397457745,
      "loss": 0.3506,
      "reward": 0.796875,
      "reward_std": 0.2616912335157394,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.775,
      "step": 450
    },
    {
      "completion_length": 49.778125,
      "epoch": 0.49066666666666664,
      "grad_norm": 0.5252532362937927,
      "kl": 1.47607421875,
      "learning_rate": 0.00029634797993278333,
      "loss": 0.2026,
      "reward": 0.89375,
      "reward_std": 0.11293471753597259,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.88125,
      "step": 460
    },
    {
      "completion_length": 57.903125,
      "epoch": 0.5013333333333333,
      "grad_norm": 0.10718824714422226,
      "kl": 1.683837890625,
      "learning_rate": 0.000295928032807813,
      "loss": 0.1887,
      "reward": 0.859375,
      "reward_std": 0.1423343911767006,
      "rewards/accuracy_reward": 0.00625,
      "rewards/format_reward": 0.853125,
      "step": 470
    },
    {
      "completion_length": 63.734375,
      "epoch": 0.512,
      "grad_norm": 0.32101932168006897,
      "kl": 2.9671875,
      "learning_rate": 0.00029548556740206994,
      "loss": 0.3254,
      "reward": 0.79375,
      "reward_std": 0.29874250292778015,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.784375,
      "step": 480
    },
    {
      "completion_length": 66.253125,
      "epoch": 0.5226666666666666,
      "grad_norm": 0.7132259011268616,
      "kl": 2.6101806640625,
      "learning_rate": 0.0002950206519927731,
      "loss": 0.2574,
      "reward": 0.728125,
      "reward_std": 0.3086773693561554,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.709375,
      "step": 490
    },
    {
      "completion_length": 64.765625,
      "epoch": 0.5333333333333333,
      "grad_norm": 0.7293491959571838,
      "kl": 3.3251220703125,
      "learning_rate": 0.00029453335832142075,
      "loss": 0.3315,
      "reward": 0.75625,
      "reward_std": 0.27030970752239225,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.746875,
      "step": 500
    },
    {
      "completion_length": 61.140625,
      "epoch": 0.544,
      "grad_norm": 0.4900813400745392,
      "kl": 1.6069091796875,
      "learning_rate": 0.0002940237615827202,
      "loss": 0.162,
      "reward": 0.86875,
      "reward_std": 0.21899680644273758,
      "rewards/accuracy_reward": 0.053125,
      "rewards/format_reward": 0.815625,
      "step": 510
    },
    {
      "completion_length": 59.621875,
      "epoch": 0.5546666666666666,
      "grad_norm": 0.4984245002269745,
      "kl": 1.695751953125,
      "learning_rate": 0.00029349194041298435,
      "loss": 0.2075,
      "reward": 0.903125,
      "reward_std": 0.16081304997205734,
      "rewards/accuracy_reward": 0.034375,
      "rewards/format_reward": 0.86875,
      "step": 520
    },
    {
      "completion_length": 62.209375,
      "epoch": 0.5653333333333334,
      "grad_norm": 0.25215986371040344,
      "kl": 1.5575439453125,
      "learning_rate": 0.0002929379768779971,
      "loss": 0.1648,
      "reward": 0.890625,
      "reward_std": 0.17983439117670058,
      "rewards/accuracy_reward": 0.040625,
      "rewards/format_reward": 0.85,
      "step": 530
    },
    {
      "completion_length": 65.865625,
      "epoch": 0.576,
      "grad_norm": 0.1489488184452057,
      "kl": 2.0063720703125,
      "learning_rate": 0.0002923619564603501,
      "loss": 0.187,
      "reward": 0.78125,
      "reward_std": 0.2043856605887413,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.75,
      "step": 540
    },
    {
      "completion_length": 68.68125,
      "epoch": 0.5866666666666667,
      "grad_norm": 0.24991311132907867,
      "kl": 1.0421142578125,
      "learning_rate": 0.00029176396804625135,
      "loss": 0.0977,
      "reward": 0.909375,
      "reward_std": 0.1441847175359726,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.8875,
      "step": 550
    },
    {
      "completion_length": 64.2,
      "epoch": 0.5973333333333334,
      "grad_norm": 0.7193971872329712,
      "kl": 2.2302490234375,
      "learning_rate": 0.00029114410391180946,
      "loss": 0.2166,
      "reward": 0.834375,
      "reward_std": 0.2048343911767006,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.81875,
      "step": 560
    },
    {
      "completion_length": 66.01875,
      "epoch": 0.608,
      "grad_norm": 0.59996098279953,
      "kl": 2.652294921875,
      "learning_rate": 0.0002905024597087945,
      "loss": 0.2907,
      "reward": 0.815625,
      "reward_std": 0.20596464574337006,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.7875,
      "step": 570
    },
    {
      "completion_length": 69.009375,
      "epoch": 0.6186666666666667,
      "grad_norm": 0.32363754510879517,
      "kl": 0.75888671875,
      "learning_rate": 0.0002898391344498775,
      "loss": 0.112,
      "reward": 0.896875,
      "reward_std": 0.14761751294136047,
      "rewards/accuracy_reward": 0.034375,
      "rewards/format_reward": 0.8625,
      "step": 580
    },
    {
      "completion_length": 58.490625,
      "epoch": 0.6293333333333333,
      "grad_norm": 0.5817243456840515,
      "kl": 3.2273193359375,
      "learning_rate": 0.0002891542304933521,
      "loss": 0.3775,
      "reward": 0.796875,
      "reward_std": 0.24620190411806106,
      "rewards/accuracy_reward": 0.00625,
      "rewards/format_reward": 0.790625,
      "step": 590
    },
    {
      "completion_length": 56.890625,
      "epoch": 0.64,
      "grad_norm": 0.561817467212677,
      "kl": 1.345703125,
      "learning_rate": 0.00028844785352733924,
      "loss": 0.1409,
      "reward": 0.884375,
      "reward_std": 0.1315855011343956,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.8625,
      "step": 600
    },
    {
      "completion_length": 54.61875,
      "epoch": 0.6506666666666666,
      "grad_norm": 0.41451311111450195,
      "kl": 1.3966552734375,
      "learning_rate": 0.00028772011255347873,
      "loss": 0.1476,
      "reward": 0.890625,
      "reward_std": 0.16838996410369872,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.871875,
      "step": 610
    },
    {
      "completion_length": 58.15,
      "epoch": 0.6613333333333333,
      "grad_norm": 0.38927924633026123,
      "kl": 2.2388427734375,
      "learning_rate": 0.00028697111987010865,
      "loss": 0.2576,
      "reward": 0.871875,
      "reward_std": 0.1995512694120407,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.84375,
      "step": 620
    },
    {
      "completion_length": 57.20625,
      "epoch": 0.672,
      "grad_norm": 0.09751415997743607,
      "kl": 0.793994140625,
      "learning_rate": 0.0002862009910549369,
      "loss": 0.0629,
      "reward": 0.9375,
      "reward_std": 0.10386751294136047,
      "rewards/accuracy_reward": 0.01875,
      "rewards/format_reward": 0.91875,
      "step": 630
    },
    {
      "completion_length": 65.846875,
      "epoch": 0.6826666666666666,
      "grad_norm": 0.2675510048866272,
      "kl": 2.466015625,
      "learning_rate": 0.0002854098449472061,
      "loss": 0.2627,
      "reward": 0.79375,
      "reward_std": 0.2520918682217598,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.778125,
      "step": 640
    },
    {
      "completion_length": 62.98125,
      "epoch": 0.6933333333333334,
      "grad_norm": 0.15855202078819275,
      "kl": 1.8398193359375,
      "learning_rate": 0.00028459780362935527,
      "loss": 0.177,
      "reward": 0.91875,
      "reward_std": 0.15879059880971907,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.90625,
      "step": 650
    },
    {
      "completion_length": 62.6,
      "epoch": 0.704,
      "grad_norm": 0.12087615579366684,
      "kl": 2.0813720703125,
      "learning_rate": 0.0002837649924081816,
      "loss": 0.1866,
      "reward": 0.90625,
      "reward_std": 0.19479155987501146,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.875,
      "step": 660
    },
    {
      "completion_length": 65.925,
      "epoch": 0.7146666666666667,
      "grad_norm": 0.39411771297454834,
      "kl": 1.392919921875,
      "learning_rate": 0.00028291153979550387,
      "loss": 0.2015,
      "reward": 0.915625,
      "reward_std": 0.1775405988097191,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.8875,
      "step": 670
    },
    {
      "completion_length": 64.378125,
      "epoch": 0.7253333333333334,
      "grad_norm": 1.1659783124923706,
      "kl": 2.8165283203125,
      "learning_rate": 0.00028203757748833174,
      "loss": 0.3109,
      "reward": 0.778125,
      "reward_std": 0.21169123351573943,
      "rewards/accuracy_reward": 0.003125,
      "rewards/format_reward": 0.775,
      "step": 680
    },
    {
      "completion_length": 57.334375,
      "epoch": 0.736,
      "grad_norm": 0.27627384662628174,
      "kl": 1.0085693359375,
      "learning_rate": 0.0002811432403485437,
      "loss": 0.1226,
      "reward": 0.859375,
      "reward_std": 0.11540063470602036,
      "rewards/accuracy_reward": 0.009375,
      "rewards/format_reward": 0.85,
      "step": 690
    },
    {
      "completion_length": 54.6,
      "epoch": 0.7466666666666667,
      "grad_norm": 0.4506663382053375,
      "kl": 2.3274169921875,
      "learning_rate": 0.00028022866638207624,
      "loss": 0.2726,
      "reward": 0.853125,
      "reward_std": 0.2143363133072853,
      "rewards/accuracy_reward": 0.025,
      "rewards/format_reward": 0.828125,
      "step": 700
    },
    {
      "completion_length": 64.646875,
      "epoch": 0.7573333333333333,
      "grad_norm": 0.24161870777606964,
      "kl": 1.10751953125,
      "learning_rate": 0.00027929399671762793,
      "loss": 0.1497,
      "reward": 0.878125,
      "reward_std": 0.18096464574337007,
      "rewards/accuracy_reward": 0.04375,
      "rewards/format_reward": 0.834375,
      "step": 710
    },
    {
      "completion_length": 65.128125,
      "epoch": 0.768,
      "grad_norm": 0.22652657330036163,
      "kl": 1.7567138671875,
      "learning_rate": 0.00027833937558488183,
      "loss": 0.1692,
      "reward": 0.865625,
      "reward_std": 0.19575843811035157,
      "rewards/accuracy_reward": 0.059375,
      "rewards/format_reward": 0.80625,
      "step": 720
    },
    {
      "completion_length": 76.259375,
      "epoch": 0.7786666666666666,
      "grad_norm": 0.46417316794395447,
      "kl": 3.4798828125,
      "learning_rate": 0.0002773649502922495,
      "loss": 0.3618,
      "reward": 0.7125,
      "reward_std": 0.31879488229751585,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.7,
      "step": 730
    },
    {
      "completion_length": 64.928125,
      "epoch": 0.7893333333333333,
      "grad_norm": 0.8743041753768921,
      "kl": 2.166650390625,
      "learning_rate": 0.00027637087120413933,
      "loss": 0.2562,
      "reward": 0.840625,
      "reward_std": 0.2851921945810318,
      "rewards/accuracy_reward": 0.0375,
      "rewards/format_reward": 0.803125,
      "step": 740
    },
    {
      "completion_length": 58.028125,
      "epoch": 0.8,
      "grad_norm": 0.18655003607273102,
      "kl": 1.73994140625,
      "learning_rate": 0.000275357291717754,
      "loss": 0.191,
      "reward": 0.909375,
      "reward_std": 0.19460364878177644,
      "rewards/accuracy_reward": 0.053125,
      "rewards/format_reward": 0.85625,
      "step": 750
    },
    {
      "completion_length": 60.803125,
      "epoch": 0.8106666666666666,
      "grad_norm": 0.04459076747298241,
      "kl": 1.7782470703125,
      "learning_rate": 0.0002743243682394195,
      "loss": 0.2117,
      "reward": 0.83125,
      "reward_std": 0.16213996410369874,
      "rewards/accuracy_reward": 0.025,
      "rewards/format_reward": 0.80625,
      "step": 760
    },
    {
      "completion_length": 56.203125,
      "epoch": 0.8213333333333334,
      "grad_norm": 0.04220689460635185,
      "kl": 1.7406494140625,
      "learning_rate": 0.00027327226016044963,
      "loss": 0.1999,
      "reward": 0.878125,
      "reward_std": 0.1423343911767006,
      "rewards/accuracy_reward": 0.034375,
      "rewards/format_reward": 0.84375,
      "step": 770
    },
    {
      "completion_length": 58.375,
      "epoch": 0.832,
      "grad_norm": 0.3807085156440735,
      "kl": 1.8222412109375,
      "learning_rate": 0.00027220112983255087,
      "loss": 0.2296,
      "reward": 0.903125,
      "reward_std": 0.20482564270496367,
      "rewards/accuracy_reward": 0.034375,
      "rewards/format_reward": 0.86875,
      "step": 780
    },
    {
      "completion_length": 63.378125,
      "epoch": 0.8426666666666667,
      "grad_norm": 0.01206011138856411,
      "kl": 2.458740234375,
      "learning_rate": 0.00027111114254276913,
      "loss": 0.3096,
      "reward": 0.84375,
      "reward_std": 0.2114198923110962,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.821875,
      "step": 790
    },
    {
      "completion_length": 59.7625,
      "epoch": 0.8533333333333334,
      "grad_norm": 0.40591439604759216,
      "kl": 1.378076171875,
      "learning_rate": 0.00027000246648798456,
      "loss": 0.1403,
      "reward": 0.934375,
      "reward_std": 0.14083535224199295,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.903125,
      "step": 800
    },
    {
      "completion_length": 62.284375,
      "epoch": 0.864,
      "grad_norm": 0.27511999011039734,
      "kl": 2.2107177734375,
      "learning_rate": 0.0002688752727489565,
      "loss": 0.2636,
      "reward": 0.8875,
      "reward_std": 0.21739855110645295,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.859375,
      "step": 810
    },
    {
      "completion_length": 65.265625,
      "epoch": 0.8746666666666667,
      "grad_norm": 0.2582601010799408,
      "kl": 2.3897705078125,
      "learning_rate": 0.00026772973526392453,
      "loss": 0.2965,
      "reward": 0.83125,
      "reward_std": 0.2494538262486458,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.803125,
      "step": 820
    },
    {
      "completion_length": 54.865625,
      "epoch": 0.8853333333333333,
      "grad_norm": 0.23494267463684082,
      "kl": 2.6015625,
      "learning_rate": 0.0002665660308017671,
      "loss": 0.252,
      "reward": 0.9,
      "reward_std": 0.23950843811035155,
      "rewards/accuracy_reward": 0.04375,
      "rewards/format_reward": 0.85625,
      "step": 830
    },
    {
      "completion_length": 55.503125,
      "epoch": 0.896,
      "grad_norm": 0.20798054337501526,
      "kl": 1.5889892578125,
      "learning_rate": 0.000265384338934725,
      "loss": 0.1996,
      "reward": 0.9375,
      "reward_std": 0.20120493620634078,
      "rewards/accuracy_reward": 0.0625,
      "rewards/format_reward": 0.875,
      "step": 840
    },
    {
      "completion_length": 59.909375,
      "epoch": 0.9066666666666666,
      "grad_norm": 0.23807695508003235,
      "kl": 1.656982421875,
      "learning_rate": 0.00026418484201069055,
      "loss": 0.194,
      "reward": 0.840625,
      "reward_std": 0.17524680644273757,
      "rewards/accuracy_reward": 0.034375,
      "rewards/format_reward": 0.80625,
      "step": 850
    },
    {
      "completion_length": 56.74375,
      "epoch": 0.9173333333333333,
      "grad_norm": 0.21559438109397888,
      "kl": 0.813427734375,
      "learning_rate": 0.00026296772512507025,
      "loss": 0.1054,
      "reward": 0.884375,
      "reward_std": 0.13916241526603698,
      "rewards/accuracy_reward": 0.021875,
      "rewards/format_reward": 0.8625,
      "step": 860
    },
    {
      "completion_length": 62.390625,
      "epoch": 0.928,
      "grad_norm": 0.1291944831609726,
      "kl": 1.9663330078125,
      "learning_rate": 0.0002617331760922218,
      "loss": 0.2316,
      "reward": 0.85625,
      "reward_std": 0.15685684233903885,
      "rewards/accuracy_reward": 0.0125,
      "rewards/format_reward": 0.84375,
      "step": 870
    },
    {
      "completion_length": 56.009375,
      "epoch": 0.9386666666666666,
      "grad_norm": 1.045857548713684,
      "kl": 1.652001953125,
      "learning_rate": 0.0002604813854164726,
      "loss": 0.1616,
      "reward": 0.9375,
      "reward_std": 0.16336943507194518,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.909375,
      "step": 880
    },
    {
      "completion_length": 64.446875,
      "epoch": 0.9493333333333334,
      "grad_norm": 0.33091413974761963,
      "kl": 3.235400390625,
      "learning_rate": 0.0002592125462627231,
      "loss": 0.3973,
      "reward": 0.796875,
      "reward_std": 0.2716366216540337,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.78125,
      "step": 890
    },
    {
      "completion_length": 59.0,
      "epoch": 0.96,
      "grad_norm": 0.25974419713020325,
      "kl": 1.752197265625,
      "learning_rate": 0.00025792685442663877,
      "loss": 0.1938,
      "reward": 0.89375,
      "reward_std": 0.1826515957713127,
      "rewards/accuracy_reward": 0.028125,
      "rewards/format_reward": 0.865625,
      "step": 900
    },
    {
      "completion_length": 57.4125,
      "epoch": 0.9706666666666667,
      "grad_norm": 0.2569887936115265,
      "kl": 2.5720703125,
      "learning_rate": 0.00025662450830443733,
      "loss": 0.3213,
      "reward": 0.846875,
      "reward_std": 0.22065922170877456,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.815625,
      "step": 910
    },
    {
      "completion_length": 51.140625,
      "epoch": 0.9813333333333333,
      "grad_norm": 0.18798935413360596,
      "kl": 1.6124267578125,
      "learning_rate": 0.0002553057088622736,
      "loss": 0.2214,
      "reward": 0.925,
      "reward_std": 0.1477062076330185,
      "rewards/accuracy_reward": 0.015625,
      "rewards/format_reward": 0.909375,
      "step": 920
    },
    {
      "completion_length": 53.921875,
      "epoch": 0.992,
      "grad_norm": 0.8309330940246582,
      "kl": 1.651806640625,
      "learning_rate": 0.0002539706596052286,
      "loss": 0.1893,
      "reward": 0.909375,
      "reward_std": 0.16504059880971908,
      "rewards/accuracy_reward": 0.03125,
      "rewards/format_reward": 0.878125,
      "step": 930
    }
  ],
  "logging_steps": 10,
  "max_steps": 2811,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}