{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9984,
  "eval_steps": 100,
  "global_step": 468,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 637.8268123626709,
      "epoch": 0.010666666666666666,
      "grad_norm": 0.9318448901176453,
      "kl": 0.00011245012283325195,
      "learning_rate": 3.1914893617021275e-07,
      "loss": 0.0,
      "reward": 0.6035714592784643,
      "reward_std": 0.3709465142339468,
      "rewards/accuracy_reward": 0.6017857443541288,
      "rewards/format_reward": 0.001785714365541935,
      "step": 5
    },
    {
      "completion_length": 611.45181350708,
      "epoch": 0.021333333333333333,
      "grad_norm": 6.491418838500977,
      "kl": 0.0001920461654663086,
      "learning_rate": 6.382978723404255e-07,
      "loss": 0.0,
      "reward": 0.617857176065445,
      "reward_std": 0.36645040661096573,
      "rewards/accuracy_reward": 0.617857176065445,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "completion_length": 604.3607414245605,
      "epoch": 0.032,
      "grad_norm": 9.168869018554688,
      "kl": 0.00031398534774780275,
      "learning_rate": 9.574468085106384e-07,
      "loss": 0.0,
      "reward": 0.6482143150642514,
      "reward_std": 0.34973760321736336,
      "rewards/accuracy_reward": 0.6482143150642514,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "completion_length": 617.7803817749024,
      "epoch": 0.042666666666666665,
      "grad_norm": 1.5698275566101074,
      "kl": 0.0008591651916503906,
      "learning_rate": 1.276595744680851e-06,
      "loss": 0.0,
      "reward": 0.6303571715950966,
      "reward_std": 0.35366948917508123,
      "rewards/accuracy_reward": 0.6303571715950966,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "completion_length": 639.5375259399414,
      "epoch": 0.05333333333333334,
      "grad_norm": 0.858055055141449,
      "kl": 0.002974271774291992,
      "learning_rate": 1.5957446808510639e-06,
      "loss": 0.0001,
      "reward": 0.6125000298023224,
      "reward_std": 0.3530873417854309,
      "rewards/accuracy_reward": 0.6125000298023224,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "completion_length": 664.7107475280761,
      "epoch": 0.064,
      "grad_norm": 1.1970467567443848,
      "kl": 0.005088996887207031,
      "learning_rate": 1.9148936170212767e-06,
      "loss": 0.0002,
      "reward": 0.6553571727126837,
      "reward_std": 0.33203957192599776,
      "rewards/accuracy_reward": 0.6553571727126837,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 632.7196670532227,
      "epoch": 0.07466666666666667,
      "grad_norm": 1.0202564001083374,
      "kl": 0.996532678604126,
      "learning_rate": 2.2340425531914894e-06,
      "loss": 0.0398,
      "reward": 0.7000000279396772,
      "reward_std": 0.3037954304367304,
      "rewards/accuracy_reward": 0.7000000279396772,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "completion_length": 640.1286018371582,
      "epoch": 0.08533333333333333,
      "grad_norm": 0.5117438435554504,
      "kl": 0.002056884765625,
      "learning_rate": 2.553191489361702e-06,
      "loss": 0.0001,
      "reward": 0.7250000268220902,
      "reward_std": 0.2653495166450739,
      "rewards/accuracy_reward": 0.7250000268220902,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "completion_length": 607.9250259399414,
      "epoch": 0.096,
      "grad_norm": 0.3817993402481079,
      "kl": 0.002574253082275391,
      "learning_rate": 2.872340425531915e-06,
      "loss": 0.0001,
      "reward": 0.7392857447266579,
      "reward_std": 0.24080881737172605,
      "rewards/accuracy_reward": 0.7392857447266579,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "completion_length": 628.1535995483398,
      "epoch": 0.10666666666666667,
      "grad_norm": 0.7749062180519104,
      "kl": 0.004121017456054687,
      "learning_rate": 2.9996241442585123e-06,
      "loss": 0.0002,
      "reward": 0.6964286014437675,
      "reward_std": 0.300217243283987,
      "rewards/accuracy_reward": 0.6964286014437675,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "completion_length": 600.3839553833008,
      "epoch": 0.11733333333333333,
      "grad_norm": 0.5632015466690063,
      "kl": 0.003362083435058594,
      "learning_rate": 2.9973279301399446e-06,
      "loss": 0.0001,
      "reward": 0.7178571719676257,
      "reward_std": 0.27802310809493064,
      "rewards/accuracy_reward": 0.7178571719676257,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "completion_length": 585.9303833007813,
      "epoch": 0.128,
      "grad_norm": 1.2080248594284058,
      "kl": 0.004734039306640625,
      "learning_rate": 2.992947502998804e-06,
      "loss": 0.0002,
      "reward": 0.7750000357627869,
      "reward_std": 0.25223284475505353,
      "rewards/accuracy_reward": 0.7750000357627869,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "completion_length": 591.1464614868164,
      "epoch": 0.13866666666666666,
      "grad_norm": 1.7432464361190796,
      "kl": 0.013347434997558593,
      "learning_rate": 2.9864889601923268e-06,
      "loss": 0.0005,
      "reward": 0.7267857432365418,
      "reward_std": 0.2874183960258961,
      "rewards/accuracy_reward": 0.7267857432365418,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "completion_length": 588.239315032959,
      "epoch": 0.14933333333333335,
      "grad_norm": 0.28103670477867126,
      "kl": 0.004166412353515625,
      "learning_rate": 2.977961291721137e-06,
      "loss": 0.0002,
      "reward": 0.7875000283122062,
      "reward_std": 0.23745907917618753,
      "rewards/accuracy_reward": 0.7875000283122062,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "completion_length": 554.2375259399414,
      "epoch": 0.16,
      "grad_norm": 0.26267099380493164,
      "kl": 0.004508209228515625,
      "learning_rate": 2.9673763677155655e-06,
      "loss": 0.0002,
      "reward": 0.7767857410013675,
      "reward_std": 0.21466484777629374,
      "rewards/accuracy_reward": 0.7767857410013675,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "completion_length": 553.0071640014648,
      "epoch": 0.17066666666666666,
      "grad_norm": 0.34145233035087585,
      "kl": 0.0057952880859375,
      "learning_rate": 2.9547489219129666e-06,
      "loss": 0.0002,
      "reward": 0.8285714581608772,
      "reward_std": 0.20639798790216446,
      "rewards/accuracy_reward": 0.8285714581608772,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "completion_length": 585.5857421875,
      "epoch": 0.18133333333333335,
      "grad_norm": 0.240849107503891,
      "kl": 0.0038000106811523437,
      "learning_rate": 2.9400965311490175e-06,
      "loss": 0.0002,
      "reward": 0.7660714596509933,
      "reward_std": 0.24059830717742442,
      "rewards/accuracy_reward": 0.7660714596509933,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "completion_length": 558.014315032959,
      "epoch": 0.192,
      "grad_norm": 0.42303234338760376,
      "kl": 0.004909515380859375,
      "learning_rate": 2.9234395908915565e-06,
      "loss": 0.0002,
      "reward": 0.7142857456579804,
      "reward_std": 0.2469261337071657,
      "rewards/accuracy_reward": 0.7142857456579804,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "completion_length": 565.8589538574219,
      "epoch": 0.20266666666666666,
      "grad_norm": 0.36556175351142883,
      "kl": 0.004494476318359375,
      "learning_rate": 2.904801286851009e-06,
      "loss": 0.0002,
      "reward": 0.7517857383936644,
      "reward_std": 0.2242392159998417,
      "rewards/accuracy_reward": 0.7517857383936644,
      "rewards/format_reward": 0.0,
      "step": 95
    },
    {
      "completion_length": 520.4696662902832,
      "epoch": 0.21333333333333335,
      "grad_norm": 0.2530968487262726,
      "kl": 0.00856475830078125,
      "learning_rate": 2.884207562706925e-06,
      "loss": 0.0003,
      "reward": 0.8125000301748514,
      "reward_std": 0.18475012369453908,
      "rewards/accuracy_reward": 0.8125000301748514,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "epoch": 0.21333333333333335,
      "eval_completion_length": 547.995567590332,
      "eval_kl": 0.00711322021484375,
      "eval_loss": 0.00028467908850871027,
      "eval_reward": 0.6861428862035275,
      "eval_reward_std": 0.270268753862381,
      "eval_rewards/accuracy_reward": 0.6860857433497906,
      "eval_rewards/format_reward": 5.7142859697341916e-05,
      "eval_runtime": 28593.9461,
      "eval_samples_per_second": 0.175,
      "eval_steps_per_second": 0.013,
      "step": 100
    },
    {
      "completion_length": 549.5107398986817,
      "epoch": 0.224,
      "grad_norm": 0.2235504686832428,
      "kl": 0.004962539672851563,
      "learning_rate": 2.8616870839955444e-06,
      "loss": 0.0002,
      "reward": 0.7964286103844642,
      "reward_std": 0.26562733463943006,
      "rewards/accuracy_reward": 0.7964286103844642,
      "rewards/format_reward": 0.0,
      "step": 105
    },
    {
      "completion_length": 581.1750267028808,
      "epoch": 0.23466666666666666,
      "grad_norm": 0.4279918968677521,
      "kl": 0.005224609375,
      "learning_rate": 2.837271198208662e-06,
      "loss": 0.0002,
      "reward": 0.7785714581608772,
      "reward_std": 0.20086282528936864,
      "rewards/accuracy_reward": 0.7785714581608772,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "completion_length": 532.0053810119629,
      "epoch": 0.24533333333333332,
      "grad_norm": 0.6788883805274963,
      "kl": 0.0057430267333984375,
      "learning_rate": 2.8109938911593322e-06,
      "loss": 0.0002,
      "reward": 0.7767857426777482,
      "reward_std": 0.20144498273730277,
      "rewards/accuracy_reward": 0.7767857426777482,
      "rewards/format_reward": 0.0,
      "step": 115
    },
    {
      "completion_length": 544.6714546203614,
      "epoch": 0.256,
      "grad_norm": 0.23265813291072845,
      "kl": 0.006147003173828125,
      "learning_rate": 2.7828917396751474e-06,
      "loss": 0.0002,
      "reward": 0.7696428894996643,
      "reward_std": 0.20026272870600223,
      "rewards/accuracy_reward": 0.7696428894996643,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "completion_length": 518.471452331543,
      "epoch": 0.26666666666666666,
      "grad_norm": 0.41492336988449097,
      "kl": 0.00635986328125,
      "learning_rate": 2.753003860684943e-06,
      "loss": 0.0003,
      "reward": 0.8375000298023224,
      "reward_std": 0.20442306995391846,
      "rewards/accuracy_reward": 0.8375000298023224,
      "rewards/format_reward": 0.0,
      "step": 125
    },
    {
      "completion_length": 570.2303848266602,
      "epoch": 0.2773333333333333,
      "grad_norm": 0.4288278818130493,
      "kl": 0.006764602661132812,
      "learning_rate": 2.721371856769793e-06,
      "loss": 0.0003,
      "reward": 0.7160714600235224,
      "reward_std": 0.2673064887523651,
      "rewards/accuracy_reward": 0.7160714600235224,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "completion_length": 494.16431350708007,
      "epoch": 0.288,
      "grad_norm": 0.5349715352058411,
      "kl": 0.008311080932617187,
      "learning_rate": 2.688039758254093e-06,
      "loss": 0.0003,
      "reward": 0.7910714618861675,
      "reward_std": 0.22742781266570092,
      "rewards/accuracy_reward": 0.7910714618861675,
      "rewards/format_reward": 0.0,
      "step": 135
    },
    {
      "completion_length": 503.5232357025146,
      "epoch": 0.2986666666666667,
      "grad_norm": 0.9469221830368042,
      "kl": 0.011474609375,
      "learning_rate": 2.65305396191733e-06,
      "loss": 0.0005,
      "reward": 0.8053571686148644,
      "reward_std": 0.22859212197363377,
      "rewards/accuracy_reward": 0.8053571686148644,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "completion_length": 526.3750251770019,
      "epoch": 0.30933333333333335,
      "grad_norm": 0.4655410945415497,
      "kl": 0.016347885131835938,
      "learning_rate": 2.61646316641186e-06,
      "loss": 0.0007,
      "reward": 0.7767857382073998,
      "reward_std": 0.21031193807721138,
      "rewards/accuracy_reward": 0.7767857382073998,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "completion_length": 536.4660957336425,
      "epoch": 0.32,
      "grad_norm": 0.36182740330696106,
      "kl": 0.020062255859375,
      "learning_rate": 2.5783183044765715e-06,
      "loss": 0.0008,
      "reward": 0.7517857436090708,
      "reward_std": 0.2297923181205988,
      "rewards/accuracy_reward": 0.7517857436090708,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "completion_length": 514.8803802490235,
      "epoch": 0.33066666666666666,
      "grad_norm": 0.7956529855728149,
      "kl": 0.03543167114257813,
      "learning_rate": 2.5386724720408135e-06,
      "loss": 0.0014,
      "reward": 0.7517857443541288,
      "reward_std": 0.26203566156327723,
      "rewards/accuracy_reward": 0.7517857443541288,
      "rewards/format_reward": 0.0,
      "step": 155
    },
    {
      "completion_length": 549.3571701049805,
      "epoch": 0.3413333333333333,
      "grad_norm": 0.29695257544517517,
      "kl": 0.062085723876953124,
      "learning_rate": 2.49758085431725e-06,
      "loss": 0.0025,
      "reward": 0.7285714589059353,
      "reward_std": 0.269120267406106,
      "rewards/accuracy_reward": 0.7285714589059353,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "completion_length": 589.5339569091797,
      "epoch": 0.352,
      "grad_norm": 0.41898104548454285,
      "kl": 0.13505706787109376,
      "learning_rate": 2.455100648986533e-06,
      "loss": 0.0054,
      "reward": 0.6857143165543675,
      "reward_std": 0.32202624566853044,
      "rewards/accuracy_reward": 0.6857143165543675,
      "rewards/format_reward": 0.0,
      "step": 165
    },
    {
      "completion_length": 627.4214576721191,
      "epoch": 0.3626666666666667,
      "grad_norm": 0.2985088527202606,
      "kl": 0.1629364013671875,
      "learning_rate": 2.4112909865807053e-06,
      "loss": 0.0065,
      "reward": 0.6410714607685805,
      "reward_std": 0.2777946576476097,
      "rewards/accuracy_reward": 0.6410714607685805,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "completion_length": 567.5768112182617,
      "epoch": 0.37333333333333335,
      "grad_norm": 0.49856701493263245,
      "kl": 0.115283203125,
      "learning_rate": 2.366212848176164e-06,
      "loss": 0.0046,
      "reward": 0.7089286031201482,
      "reward_std": 0.2502220422029495,
      "rewards/accuracy_reward": 0.7089286031201482,
      "rewards/format_reward": 0.0,
      "step": 175
    },
    {
      "completion_length": 528.6750221252441,
      "epoch": 0.384,
      "grad_norm": 0.6558699607849121,
      "kl": 0.18359222412109374,
      "learning_rate": 2.319928980510752e-06,
      "loss": 0.0073,
      "reward": 0.6535714527592063,
      "reward_std": 0.28206231258809566,
      "rewards/accuracy_reward": 0.6535714527592063,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "completion_length": 571.7518112182618,
      "epoch": 0.39466666666666667,
      "grad_norm": 6.485378742218018,
      "kl": 0.34942626953125,
      "learning_rate": 2.272503808643123e-06,
      "loss": 0.014,
      "reward": 0.4696428783237934,
      "reward_std": 0.27855144031345846,
      "rewards/accuracy_reward": 0.4696428783237934,
      "rewards/format_reward": 0.0,
      "step": 185
    },
    {
      "completion_length": 542.8089500427246,
      "epoch": 0.4053333333333333,
      "grad_norm": 1.8310225009918213,
      "kl": 0.4147308349609375,
      "learning_rate": 2.2240033462759628e-06,
      "loss": 0.0166,
      "reward": 0.46071431171149013,
      "reward_std": 0.36014051400125024,
      "rewards/accuracy_reward": 0.46071431171149013,
      "rewards/format_reward": 0.0,
      "step": 190
    },
    {
      "completion_length": 626.9553817749023,
      "epoch": 0.416,
      "grad_norm": 17.148595809936523,
      "kl": 0.4925048828125,
      "learning_rate": 2.1744951038678905e-06,
      "loss": 0.0197,
      "reward": 0.4000000203028321,
      "reward_std": 0.3870592150837183,
      "rewards/accuracy_reward": 0.4000000203028321,
      "rewards/format_reward": 0.0,
      "step": 195
    },
    {
      "completion_length": 525.5928825378418,
      "epoch": 0.4266666666666667,
      "grad_norm": 38.49786376953125,
      "kl": 0.23444061279296874,
      "learning_rate": 2.124047994661941e-06,
      "loss": 0.0094,
      "reward": 0.6821428872644901,
      "reward_std": 0.3179911646991968,
      "rewards/accuracy_reward": 0.6821428872644901,
      "rewards/format_reward": 0.0,
      "step": 200
    },
    {
      "epoch": 0.4266666666666667,
      "eval_completion_length": 528.541595147705,
      "eval_kl": 0.42865986328125,
      "eval_loss": 0.017097920179367065,
      "eval_reward": 0.662742884466052,
      "eval_reward_std": 0.2557599967300892,
      "eval_rewards/accuracy_reward": 0.662742884466052,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 28459.3155,
      "eval_samples_per_second": 0.176,
      "eval_steps_per_second": 0.013,
      "step": 200
    },
    {
      "completion_length": 509.92859649658203,
      "epoch": 0.43733333333333335,
      "grad_norm": 4.442358493804932,
      "kl": 0.09366302490234375,
      "learning_rate": 2.072732238761434e-06,
      "loss": 0.0037,
      "reward": 0.7660714574158192,
      "reward_std": 0.21246148198843,
      "rewards/accuracy_reward": 0.7660714574158192,
      "rewards/format_reward": 0.0,
      "step": 205
    },
    {
      "completion_length": 497.5357364654541,
      "epoch": 0.448,
      "grad_norm": 0.5748523473739624,
      "kl": 0.085107421875,
      "learning_rate": 2.0206192653867536e-06,
      "loss": 0.0034,
      "reward": 0.796428595483303,
      "reward_std": 0.18161089681088924,
      "rewards/accuracy_reward": 0.796428595483303,
      "rewards/format_reward": 0.0,
      "step": 210
    },
    {
      "completion_length": 614.8643173217773,
      "epoch": 0.45866666666666667,
      "grad_norm": 0.3233760893344879,
      "kl": 0.15386199951171875,
      "learning_rate": 1.967781613449095e-06,
      "loss": 0.0062,
      "reward": 0.6392857398837805,
      "reward_std": 0.26852017305791376,
      "rewards/accuracy_reward": 0.6392857398837805,
      "rewards/format_reward": 0.0,
      "step": 215
    },
    {
      "completion_length": 575.8321681976319,
      "epoch": 0.4693333333333333,
      "grad_norm": 0.6134320497512817,
      "kl": 0.16066131591796876,
      "learning_rate": 1.9142928305795637e-06,
      "loss": 0.0064,
      "reward": 0.6446428839117289,
      "reward_std": 0.3076914418488741,
      "rewards/accuracy_reward": 0.6446428839117289,
      "rewards/format_reward": 0.0,
      "step": 220
    },
    {
      "completion_length": 529.201806640625,
      "epoch": 0.48,
      "grad_norm": 14.935003280639648,
      "kl": 0.27866592407226565,
      "learning_rate": 1.8602273707541886e-06,
      "loss": 0.0111,
      "reward": 0.6982143182307482,
      "reward_std": 0.29295355789363386,
      "rewards/accuracy_reward": 0.6982143182307482,
      "rewards/format_reward": 0.0,
      "step": 225
    },
    {
      "completion_length": 599.1643127441406,
      "epoch": 0.49066666666666664,
      "grad_norm": 29.34033203125,
      "kl": 1.2954010009765624,
      "learning_rate": 1.8056604906573418e-06,
      "loss": 0.0518,
      "reward": 0.605357170291245,
      "reward_std": 0.30737774074077606,
      "rewards/accuracy_reward": 0.605357170291245,
      "rewards/format_reward": 0.0,
      "step": 230
    },
    {
      "completion_length": 577.0696716308594,
      "epoch": 0.5013333333333333,
      "grad_norm": 33.51802062988281,
      "kl": 0.703057861328125,
      "learning_rate": 1.7506681449278226e-06,
      "loss": 0.0281,
      "reward": 0.571428601257503,
      "reward_std": 0.32992145605385303,
      "rewards/accuracy_reward": 0.571428601257503,
      "rewards/format_reward": 0.0,
      "step": 235
    },
    {
      "completion_length": 501.30717010498046,
      "epoch": 0.512,
      "grad_norm": 7.524632930755615,
      "kl": 0.3678741455078125,
      "learning_rate": 1.6953268804334257e-06,
      "loss": 0.0147,
      "reward": 0.589285738952458,
      "reward_std": 0.28248333670198916,
      "rewards/accuracy_reward": 0.589285738952458,
      "rewards/format_reward": 0.0,
      "step": 240
    },
    {
      "completion_length": 462.3607376098633,
      "epoch": 0.5226666666666666,
      "grad_norm": 6.015294075012207,
      "kl": 0.3237091064453125,
      "learning_rate": 1.6397137297211436e-06,
      "loss": 0.0129,
      "reward": 0.5589285986497998,
      "reward_std": 0.3269613076001406,
      "rewards/accuracy_reward": 0.5589285986497998,
      "rewards/format_reward": 0.0,
      "step": 245
    },
    {
      "completion_length": 638.7928871154785,
      "epoch": 0.5333333333333333,
      "grad_norm": 1.8317821025848389,
      "kl": 0.807421875,
      "learning_rate": 1.5839061037913395e-06,
      "loss": 0.0323,
      "reward": 0.3250000160187483,
      "reward_std": 0.34638786166906355,
      "rewards/accuracy_reward": 0.3250000160187483,
      "rewards/format_reward": 0.0,
      "step": 250
    },
    {
      "completion_length": 636.7161018371582,
      "epoch": 0.544,
      "grad_norm": 0.6774631142616272,
      "kl": 0.53349609375,
      "learning_rate": 1.527981684345115e-06,
      "loss": 0.0213,
      "reward": 0.35714287366718056,
      "reward_std": 0.3179732210934162,
      "rewards/accuracy_reward": 0.35714287366718056,
      "rewards/format_reward": 0.0,
      "step": 255
    },
    {
      "completion_length": 558.9875259399414,
      "epoch": 0.5546666666666666,
      "grad_norm": 2.738410472869873,
      "kl": 0.205511474609375,
      "learning_rate": 1.4720183156548855e-06,
      "loss": 0.0082,
      "reward": 0.6000000305473805,
      "reward_std": 0.3251969013363123,
      "rewards/accuracy_reward": 0.6000000305473805,
      "rewards/format_reward": 0.0,
      "step": 260
    },
    {
      "completion_length": 545.3893096923828,
      "epoch": 0.5653333333333334,
      "grad_norm": 17.57588005065918,
      "kl": 1.5068023681640625,
      "learning_rate": 1.4160938962086612e-06,
      "loss": 0.0603,
      "reward": 0.6089285997673869,
      "reward_std": 0.28422979824244976,
      "rewards/accuracy_reward": 0.6089285997673869,
      "rewards/format_reward": 0.0,
      "step": 265
    },
    {
      "completion_length": 526.2607391357421,
      "epoch": 0.576,
      "grad_norm": 38.554412841796875,
      "kl": 1.0892745971679687,
      "learning_rate": 1.3602862702788567e-06,
      "loss": 0.0436,
      "reward": 0.6267857417464257,
      "reward_std": 0.32045440524816515,
      "rewards/accuracy_reward": 0.6267857417464257,
      "rewards/format_reward": 0.0,
      "step": 270
    },
    {
      "completion_length": 517.2678787231446,
      "epoch": 0.5866666666666667,
      "grad_norm": 4.718568325042725,
      "kl": 1.1258697509765625,
      "learning_rate": 1.3046731195665748e-06,
      "loss": 0.045,
      "reward": 0.6464285977184773,
      "reward_std": 0.320261836796999,
      "rewards/accuracy_reward": 0.6464285977184773,
      "rewards/format_reward": 0.0,
      "step": 275
    },
    {
      "completion_length": 514.8375244140625,
      "epoch": 0.5973333333333334,
      "grad_norm": 4.270685195922852,
      "kl": 1.4042236328125,
      "learning_rate": 1.2493318550721775e-06,
      "loss": 0.0562,
      "reward": 0.6982143145054579,
      "reward_std": 0.2838581532239914,
      "rewards/accuracy_reward": 0.6982143145054579,
      "rewards/format_reward": 0.0,
      "step": 280
    },
    {
      "completion_length": 534.7946655273438,
      "epoch": 0.608,
      "grad_norm": 9.498967170715332,
      "kl": 0.8575332641601563,
      "learning_rate": 1.1943395093426585e-06,
      "loss": 0.0343,
      "reward": 0.6910714596509934,
      "reward_std": 0.2867689304053783,
      "rewards/accuracy_reward": 0.6910714596509934,
      "rewards/format_reward": 0.0,
      "step": 285
    },
    {
      "completion_length": 529.5875297546387,
      "epoch": 0.6186666666666667,
      "grad_norm": 8.131922721862793,
      "kl": 0.6070526123046875,
      "learning_rate": 1.1397726292458115e-06,
      "loss": 0.0243,
      "reward": 0.675000030733645,
      "reward_std": 0.2807727467268705,
      "rewards/accuracy_reward": 0.675000030733645,
      "rewards/format_reward": 0.0,
      "step": 290
    },
    {
      "completion_length": 557.6500236511231,
      "epoch": 0.6293333333333333,
      "grad_norm": 2.224785089492798,
      "kl": 1.1581512451171876,
      "learning_rate": 1.085707169420437e-06,
      "loss": 0.0463,
      "reward": 0.6142857430502773,
      "reward_std": 0.2811264578253031,
      "rewards/accuracy_reward": 0.6142857430502773,
      "rewards/format_reward": 0.0,
      "step": 295
    },
    {
      "completion_length": 529.6071685791015,
      "epoch": 0.64,
      "grad_norm": 8.0582857131958,
      "kl": 1.2655914306640625,
      "learning_rate": 1.0322183865509054e-06,
      "loss": 0.0506,
      "reward": 0.6750000296160579,
      "reward_std": 0.28525091484189036,
      "rewards/accuracy_reward": 0.6750000296160579,
      "rewards/format_reward": 0.0,
      "step": 300
    },
    {
      "epoch": 0.64,
      "eval_completion_length": 562.8381398986817,
      "eval_kl": 0.95684150390625,
      "eval_loss": 0.03832858428359032,
      "eval_reward": 0.5700857400953769,
      "eval_reward_std": 0.30191256090998647,
      "eval_rewards/accuracy_reward": 0.5700857400953769,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 30482.0361,
      "eval_samples_per_second": 0.164,
      "eval_steps_per_second": 0.012,
      "step": 300
    },
    {
      "completion_length": 557.5018104553222,
      "epoch": 0.6506666666666666,
      "grad_norm": 2.934140682220459,
      "kl": 0.6237777709960938,
      "learning_rate": 9.793807346132464e-07,
      "loss": 0.025,
      "reward": 0.6339285995811224,
      "reward_std": 0.3295004416257143,
      "rewards/accuracy_reward": 0.6339285995811224,
      "rewards/format_reward": 0.0,
      "step": 305
    },
    {
      "completion_length": 507.2857383728027,
      "epoch": 0.6613333333333333,
      "grad_norm": 1.5333462953567505,
      "kl": 0.18207855224609376,
      "learning_rate": 9.272677612385667e-07,
      "loss": 0.0073,
      "reward": 0.6017857391387225,
      "reward_std": 0.2927430454641581,
      "rewards/accuracy_reward": 0.6017857391387225,
      "rewards/format_reward": 0.0,
      "step": 310
    },
    {
      "completion_length": 535.4660949707031,
      "epoch": 0.672,
      "grad_norm": 0.8268552422523499,
      "kl": 0.2217437744140625,
      "learning_rate": 8.759520053380591e-07,
      "loss": 0.0089,
      "reward": 0.5303571715950965,
      "reward_std": 0.30413119196891786,
      "rewards/accuracy_reward": 0.5303571715950965,
      "rewards/format_reward": 0.0,
      "step": 315
    },
    {
      "completion_length": 587.7893112182617,
      "epoch": 0.6826666666666666,
      "grad_norm": 0.8802245259284973,
      "kl": 0.28748779296875,
      "learning_rate": 8.255048961321088e-07,
      "loss": 0.0115,
      "reward": 0.5250000260770321,
      "reward_std": 0.32775397710502147,
      "rewards/accuracy_reward": 0.5250000260770321,
      "rewards/format_reward": 0.0,
      "step": 320
    },
    {
      "completion_length": 568.7143135070801,
      "epoch": 0.6933333333333334,
      "grad_norm": 0.42495808005332947,
      "kl": 0.20731658935546876,
      "learning_rate": 7.759966537240373e-07,
      "loss": 0.0083,
      "reward": 0.6357143180444836,
      "reward_std": 0.32221881337463854,
      "rewards/accuracy_reward": 0.6357143180444836,
      "rewards/format_reward": 0.0,
      "step": 325
    },
    {
      "completion_length": 572.9053848266601,
      "epoch": 0.704,
      "grad_norm": 0.8112100958824158,
      "kl": 0.2259490966796875,
      "learning_rate": 7.274961913568773e-07,
      "loss": 0.009,
      "reward": 0.6071428848430515,
      "reward_std": 0.28955445289611814,
      "rewards/accuracy_reward": 0.6071428848430515,
      "rewards/format_reward": 0.0,
      "step": 330
    },
    {
      "completion_length": 585.521450805664,
      "epoch": 0.7146666666666667,
      "grad_norm": 2.5413522720336914,
      "kl": 0.1725799560546875,
      "learning_rate": 6.800710194892484e-07,
      "loss": 0.0069,
      "reward": 0.6946428865194321,
      "reward_std": 0.27458811886608603,
      "rewards/accuracy_reward": 0.6946428865194321,
      "rewards/format_reward": 0.0,
      "step": 335
    },
    {
      "completion_length": 578.700025177002,
      "epoch": 0.7253333333333334,
      "grad_norm": 2.077190637588501,
      "kl": 0.23962783813476562,
      "learning_rate": 6.33787151823836e-07,
      "loss": 0.0096,
      "reward": 0.6517857488244772,
      "reward_std": 0.29764223508536813,
      "rewards/accuracy_reward": 0.6517857488244772,
      "rewards/format_reward": 0.0,
      "step": 340
    },
    {
      "completion_length": 517.0357376098633,
      "epoch": 0.736,
      "grad_norm": 1.8005852699279785,
      "kl": 0.1731414794921875,
      "learning_rate": 5.887090134192947e-07,
      "loss": 0.0069,
      "reward": 0.7303571674972773,
      "reward_std": 0.20812651440501212,
      "rewards/accuracy_reward": 0.7303571674972773,
      "rewards/format_reward": 0.0,
      "step": 345
    },
    {
      "completion_length": 537.6071670532226,
      "epoch": 0.7466666666666667,
      "grad_norm": 5.837470531463623,
      "kl": 0.22538909912109376,
      "learning_rate": 5.448993510134669e-07,
      "loss": 0.009,
      "reward": 0.7285714605823159,
      "reward_std": 0.23610220104455948,
      "rewards/accuracy_reward": 0.7285714605823159,
      "rewards/format_reward": 0.0,
      "step": 350
    },
    {
      "completion_length": 511.03573455810545,
      "epoch": 0.7573333333333333,
      "grad_norm": 10.607796669006348,
      "kl": 0.2825325012207031,
      "learning_rate": 5.024191456827498e-07,
      "loss": 0.0113,
      "reward": 0.7357143178582192,
      "reward_std": 0.24888310953974724,
      "rewards/accuracy_reward": 0.7357143178582192,
      "rewards/format_reward": 0.0,
      "step": 355
    },
    {
      "completion_length": 528.4375274658203,
      "epoch": 0.768,
      "grad_norm": 17.805084228515625,
      "kl": 0.409503173828125,
      "learning_rate": 4.6132752795918667e-07,
      "loss": 0.0164,
      "reward": 0.6857143182307481,
      "reward_std": 0.28699738159775734,
      "rewards/accuracy_reward": 0.6857143182307481,
      "rewards/format_reward": 0.0,
      "step": 360
    },
    {
      "completion_length": 519.4303817749023,
      "epoch": 0.7786666666666666,
      "grad_norm": 7.785521507263184,
      "kl": 0.34308319091796874,
      "learning_rate": 4.2168169552342905e-07,
      "loss": 0.0137,
      "reward": 0.6928571717813611,
      "reward_std": 0.29954983331263063,
      "rewards/accuracy_reward": 0.6928571717813611,
      "rewards/format_reward": 0.0,
      "step": 365
    },
    {
      "completion_length": 504.0107368469238,
      "epoch": 0.7893333333333333,
      "grad_norm": 4.727066516876221,
      "kl": 0.7092597961425782,
      "learning_rate": 3.8353683358814046e-07,
      "loss": 0.0285,
      "reward": 0.7000000275671482,
      "reward_std": 0.23943399637937546,
      "rewards/accuracy_reward": 0.7000000275671482,
      "rewards/format_reward": 0.0,
      "step": 370
    },
    {
      "completion_length": 517.7428833007813,
      "epoch": 0.8,
      "grad_norm": 7.284985542297363,
      "kl": 0.2571014404296875,
      "learning_rate": 3.469460380826697e-07,
      "loss": 0.0103,
      "reward": 0.673214310593903,
      "reward_std": 0.2639926388859749,
      "rewards/accuracy_reward": 0.673214310593903,
      "rewards/format_reward": 0.0,
      "step": 375
    },
    {
      "completion_length": 543.2714515686035,
      "epoch": 0.8106666666666666,
      "grad_norm": 14.041029930114746,
      "kl": 0.28242645263671873,
      "learning_rate": 3.119602417459075e-07,
      "loss": 0.0113,
      "reward": 0.6732143096625804,
      "reward_std": 0.25163274370133876,
      "rewards/accuracy_reward": 0.6732143096625804,
      "rewards/format_reward": 0.0,
      "step": 380
    },
    {
      "completion_length": 546.032169342041,
      "epoch": 0.8213333333333334,
      "grad_norm": 15.783819198608398,
      "kl": 0.20561065673828124,
      "learning_rate": 2.786281432302071e-07,
      "loss": 0.0082,
      "reward": 0.7339286010712385,
      "reward_std": 0.234727381169796,
      "rewards/accuracy_reward": 0.7339286010712385,
      "rewards/format_reward": 0.0,
      "step": 385
    },
    {
      "completion_length": 546.478592300415,
      "epoch": 0.832,
      "grad_norm": 6.679286003112793,
      "kl": 0.18747406005859374,
      "learning_rate": 2.46996139315057e-07,
      "loss": 0.0075,
      "reward": 0.7375000283122063,
      "reward_std": 0.2588605497032404,
      "rewards/accuracy_reward": 0.7375000283122063,
      "rewards/format_reward": 0.0,
      "step": 390
    },
    {
      "completion_length": 531.3750255584716,
      "epoch": 0.8426666666666667,
      "grad_norm": 4.5977044105529785,
      "kl": 0.1937103271484375,
      "learning_rate": 2.1710826032485286e-07,
      "loss": 0.0077,
      "reward": 0.7357143165543676,
      "reward_std": 0.2378486678004265,
      "rewards/accuracy_reward": 0.7357143165543676,
      "rewards/format_reward": 0.0,
      "step": 395
    },
    {
      "completion_length": 595.6964569091797,
      "epoch": 0.8533333333333334,
      "grad_norm": 4.166606903076172,
      "kl": 0.2564910888671875,
      "learning_rate": 1.8900610884066817e-07,
      "loss": 0.0103,
      "reward": 0.6285714594647288,
      "reward_std": 0.2736343163996935,
      "rewards/accuracy_reward": 0.6285714594647288,
      "rewards/format_reward": 0.0,
      "step": 400
    },
    {
      "epoch": 0.8533333333333334,
      "eval_completion_length": 569.8527975585938,
      "eval_kl": 0.20817554931640625,
      "eval_loss": 0.008296786807477474,
      "eval_reward": 0.6110285988628864,
      "eval_reward_std": 0.28674686477184297,
      "eval_rewards/accuracy_reward": 0.6110285988628864,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 29873.0033,
      "eval_samples_per_second": 0.167,
      "eval_steps_per_second": 0.012,
      "step": 400
    },
    {
      "completion_length": 605.528596496582,
      "epoch": 0.864,
      "grad_norm": 49.79132843017578,
      "kl": 0.341680908203125,
      "learning_rate": 1.627288017913383e-07,
      "loss": 0.0137,
      "reward": 0.5571428839117288,
      "reward_std": 0.3719855587929487,
      "rewards/accuracy_reward": 0.5571428839117288,
      "rewards/format_reward": 0.0,
      "step": 405
    },
    {
      "completion_length": 637.5250267028808,
      "epoch": 0.8746666666666667,
      "grad_norm": 12.56246566772461,
      "kl": 0.39511566162109374,
      "learning_rate": 1.3831291600445573e-07,
      "loss": 0.0158,
      "reward": 0.5392857391387225,
      "reward_std": 0.3052775662392378,
      "rewards/accuracy_reward": 0.5392857391387225,
      "rewards/format_reward": 0.0,
      "step": 410
    },
    {
      "completion_length": 631.8607391357422,
      "epoch": 0.8853333333333333,
      "grad_norm": 6.395442962646484,
      "kl": 0.45283203125,
      "learning_rate": 1.1579243729307487e-07,
      "loss": 0.0181,
      "reward": 0.49107145331799984,
      "reward_std": 0.36503970213234427,
      "rewards/accuracy_reward": 0.49107145331799984,
      "rewards/format_reward": 0.0,
      "step": 415
    },
    {
      "completion_length": 611.7053848266602,
      "epoch": 0.896,
      "grad_norm": 1.6087334156036377,
      "kl": 0.3666015625,
      "learning_rate": 9.519871314899092e-08,
      "loss": 0.0147,
      "reward": 0.6000000327825546,
      "reward_std": 0.3550443138927221,
      "rewards/accuracy_reward": 0.6000000327825546,
      "rewards/format_reward": 0.0,
      "step": 420
    },
    {
      "completion_length": 613.4946708679199,
      "epoch": 0.9066666666666666,
      "grad_norm": 1.6784651279449463,
      "kl": 0.3438507080078125,
      "learning_rate": 7.656040910844358e-08,
      "loss": 0.0138,
      "reward": 0.5946428859606385,
      "reward_std": 0.3528768301010132,
      "rewards/accuracy_reward": 0.5946428859606385,
      "rewards/format_reward": 0.0,
      "step": 425
    },
    {
      "completion_length": 603.3143119812012,
      "epoch": 0.9173333333333333,
      "grad_norm": 4.043834209442139,
      "kl": 0.3444915771484375,
      "learning_rate": 5.990346885098235e-08,
      "loss": 0.0138,
      "reward": 0.5964285979047418,
      "reward_std": 0.3912374936044216,
      "rewards/accuracy_reward": 0.5964285979047418,
      "rewards/format_reward": 0.0,
      "step": 430
    },
    {
      "completion_length": 575.2339546203614,
      "epoch": 0.928,
      "grad_norm": 4.404404163360596,
      "kl": 0.30106658935546876,
      "learning_rate": 4.5251078087033493e-08,
      "loss": 0.012,
      "reward": 0.6696428902447223,
      "reward_std": 0.3243862982839346,
      "rewards/accuracy_reward": 0.6696428902447223,
      "rewards/format_reward": 0.0,
      "step": 435
    },
    {
      "completion_length": 582.6393119812012,
      "epoch": 0.9386666666666666,
      "grad_norm": 2.593308448791504,
      "kl": 0.383941650390625,
      "learning_rate": 3.262363228443427e-08,
      "loss": 0.0154,
      "reward": 0.6035714616999031,
      "reward_std": 0.29665699824690817,
      "rewards/accuracy_reward": 0.6035714616999031,
      "rewards/format_reward": 0.0,
      "step": 440
    },
    {
      "completion_length": 627.2125282287598,
      "epoch": 0.9493333333333334,
      "grad_norm": 5.177957057952881,
      "kl": 0.3490997314453125,
      "learning_rate": 2.2038708278862952e-08,
      "loss": 0.014,
      "reward": 0.5839285971596837,
      "reward_std": 0.30867667235434054,
      "rewards/accuracy_reward": 0.5839285971596837,
      "rewards/format_reward": 0.0,
      "step": 445
    },
    {
      "completion_length": 576.0964553833007,
      "epoch": 0.96,
      "grad_norm": 3.4815688133239746,
      "kl": 0.2956207275390625,
      "learning_rate": 1.3511039807673209e-08,
      "loss": 0.0118,
      "reward": 0.6589286010712385,
      "reward_std": 0.3096484154462814,
      "rewards/accuracy_reward": 0.6589286010712385,
      "rewards/format_reward": 0.0,
      "step": 450
    },
    {
      "completion_length": 531.355379486084,
      "epoch": 0.9706666666666667,
      "grad_norm": 2.63545298576355,
      "kl": 0.261767578125,
      "learning_rate": 7.0524970011963675e-09,
      "loss": 0.0105,
      "reward": 0.7125000340864063,
      "reward_std": 0.28212962336838243,
      "rewards/accuracy_reward": 0.7125000340864063,
      "rewards/format_reward": 0.0,
      "step": 455
    },
    {
      "completion_length": 569.5285957336425,
      "epoch": 0.9813333333333333,
      "grad_norm": 1.8794143199920654,
      "kl": 0.2968017578125,
      "learning_rate": 2.6720698600553595e-09,
      "loss": 0.0119,
      "reward": 0.6303571775555611,
      "reward_std": 0.3074988707900047,
      "rewards/accuracy_reward": 0.6303571775555611,
      "rewards/format_reward": 0.0,
      "step": 460
    },
    {
      "completion_length": 612.3214569091797,
      "epoch": 0.992,
      "grad_norm": 3.266709327697754,
      "kl": 0.39468994140625,
      "learning_rate": 3.7585574148779613e-10,
      "loss": 0.0158,
      "reward": 0.5517857383936644,
      "reward_std": 0.3217977944761515,
      "rewards/accuracy_reward": 0.5517857383936644,
      "rewards/format_reward": 0.0,
      "step": 465
    },
    {
      "completion_length": 626.7321701049805,
      "epoch": 0.9984,
      "kl": 0.32488250732421875,
      "reward": 0.604166692122817,
      "reward_std": 0.3102082473536332,
      "rewards/accuracy_reward": 0.604166692122817,
      "rewards/format_reward": 0.0,
      "step": 468,
      "total_flos": 0.0,
      "train_loss": 0.00286464851636153,
      "train_runtime": 7716.3929,
      "train_samples_per_second": 0.972,
      "train_steps_per_second": 0.061
    }
  ],
  "logging_steps": 5,
  "max_steps": 468,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}