{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 469,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 614.4910888671875,
      "epoch": 0.0021321961620469083,
      "grad_norm": 0.653412401676178,
      "grpo_loss": -0.05488306283950806,
      "kl": 0.0,
      "learning_rate": 6.382978723404255e-08,
      "loss": 0.0,
      "reward": 0.3571428656578064,
      "reward_std": 0.3737320303916931,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "completion_length": 649.857177734375,
      "epoch": 0.0042643923240938165,
      "grad_norm": 0.5258876085281372,
      "grpo_loss": 0.047233134508132935,
      "kl": 0.0,
      "learning_rate": 1.276595744680851e-07,
      "loss": 0.0,
      "reward": 0.3571428656578064,
      "reward_std": 0.42088788747787476,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 0.0,
      "step": 2
    },
    {
      "completion_length": 606.1875,
      "epoch": 0.006396588486140725,
      "grad_norm": 7.712527751922607,
      "grpo_loss": 0.016701802611351013,
      "kl": 6.341934204101562e-05,
      "learning_rate": 1.9148936170212765e-07,
      "loss": 0.0,
      "reward": 0.3035714328289032,
      "reward_std": 0.3098275065422058,
      "rewards/accuracy_reward": 0.3035714328289032,
      "rewards/format_reward": 0.0,
      "step": 3
    },
    {
      "completion_length": 624.0625,
      "epoch": 0.008528784648187633,
      "grad_norm": 0.40128573775291443,
      "grpo_loss": 2.2761523723602295e-06,
      "kl": 4.601478576660156e-05,
      "learning_rate": 2.553191489361702e-07,
      "loss": 0.0,
      "reward": 0.4285714626312256,
      "reward_std": 0.42494088411331177,
      "rewards/accuracy_reward": 0.4285714626312256,
      "rewards/format_reward": 0.0,
      "step": 4
    },
    {
      "completion_length": 535.2232666015625,
      "epoch": 0.010660980810234541,
      "grad_norm": 1.0285664796829224,
      "grpo_loss": -0.01669881120324135,
      "kl": 6.628036499023438e-05,
      "learning_rate": 3.1914893617021275e-07,
      "loss": 0.0,
      "reward": 0.330357164144516,
      "reward_std": 0.3667682409286499,
      "rewards/accuracy_reward": 0.330357164144516,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "completion_length": 565.2142944335938,
      "epoch": 0.01279317697228145,
      "grad_norm": 1.2028206586837769,
      "grpo_loss": -0.05488109588623047,
      "kl": 5.555152893066406e-05,
      "learning_rate": 3.829787234042553e-07,
      "loss": 0.0,
      "reward": 0.4285714626312256,
      "reward_std": 0.42670944333076477,
      "rewards/accuracy_reward": 0.4285714626312256,
      "rewards/format_reward": 0.0,
      "step": 6
    },
    {
      "completion_length": 547.4107666015625,
      "epoch": 0.014925373134328358,
      "grad_norm": 1.7286690473556519,
      "grpo_loss": -0.18294166028499603,
      "kl": 7.390975952148438e-05,
      "learning_rate": 4.468085106382979e-07,
      "loss": 0.0,
      "reward": 0.4375000298023224,
      "reward_std": 0.33336058259010315,
      "rewards/accuracy_reward": 0.4285714626312256,
      "rewards/format_reward": 0.008928571827709675,
      "step": 7
    },
    {
      "completion_length": 589.6160888671875,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.6688471436500549,
      "grpo_loss": 1.0952353477478027e-06,
      "kl": 5.459785461425781e-05,
      "learning_rate": 5.106382978723404e-07,
      "loss": 0.0,
      "reward": 0.392857164144516,
      "reward_std": 0.42565688490867615,
      "rewards/accuracy_reward": 0.3839285969734192,
      "rewards/format_reward": 0.008928571827709675,
      "step": 8
    },
    {
      "completion_length": 583.9017944335938,
      "epoch": 0.019189765458422176,
      "grad_norm": 0.6561264991760254,
      "grpo_loss": 0.10020516812801361,
      "kl": 5.5789947509765625e-05,
      "learning_rate": 5.74468085106383e-07,
      "loss": 0.0,
      "reward": 0.5089285969734192,
      "reward_std": 0.40017586946487427,
      "rewards/accuracy_reward": 0.5089285969734192,
      "rewards/format_reward": 0.0,
      "step": 9
    },
    {
      "completion_length": 564.0625,
      "epoch": 0.021321961620469083,
      "grad_norm": 2.1838438510894775,
      "grpo_loss": 0.04723589867353439,
      "kl": 7.343292236328125e-05,
      "learning_rate": 6.382978723404255e-07,
      "loss": 0.0,
      "reward": 0.3035714328289032,
      "reward_std": 0.3835168778896332,
      "rewards/accuracy_reward": 0.2946428656578064,
      "rewards/format_reward": 0.008928571827709675,
      "step": 10
    },
    {
      "completion_length": 601.5982666015625,
      "epoch": 0.023454157782515993,
      "grad_norm": 0.7729598879814148,
      "grpo_loss": 1.9669532775878906e-06,
      "kl": 7.82012939453125e-05,
      "learning_rate": 7.021276595744681e-07,
      "loss": 0.0,
      "reward": 0.3392857313156128,
      "reward_std": 0.38351690769195557,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.0,
      "step": 11
    },
    {
      "completion_length": 557.0535888671875,
      "epoch": 0.0255863539445629,
      "grad_norm": 0.7991155385971069,
      "grpo_loss": -0.11808028072118759,
      "kl": 6.031990051269531e-05,
      "learning_rate": 7.659574468085106e-07,
      "loss": 0.0,
      "reward": 0.4821428656578064,
      "reward_std": 0.4198353588581085,
      "rewards/accuracy_reward": 0.4821428656578064,
      "rewards/format_reward": 0.0,
      "step": 12
    },
    {
      "completion_length": 557.8035888671875,
      "epoch": 0.02771855010660981,
      "grad_norm": 0.7683190703392029,
      "grpo_loss": 9.611248970031738e-07,
      "kl": 5.555152893066406e-05,
      "learning_rate": 8.297872340425533e-07,
      "loss": 0.0,
      "reward": 0.455357164144516,
      "reward_std": 0.3334502577781677,
      "rewards/accuracy_reward": 0.455357164144516,
      "rewards/format_reward": 0.0,
      "step": 13
    },
    {
      "completion_length": 591.169677734375,
      "epoch": 0.029850746268656716,
      "grad_norm": 0.54457026720047,
      "grpo_loss": 0.04723565652966499,
      "kl": 0.00010728836059570312,
      "learning_rate": 8.936170212765958e-07,
      "loss": 0.0,
      "reward": 0.4017857313156128,
      "reward_std": 0.35989412665367126,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 0.008928571827709675,
      "step": 14
    },
    {
      "completion_length": 607.357177734375,
      "epoch": 0.031982942430703626,
      "grad_norm": 0.44422805309295654,
      "grpo_loss": 0.11808501183986664,
      "kl": 6.389617919921875e-05,
      "learning_rate": 9.574468085106384e-07,
      "loss": 0.0,
      "reward": 0.3750000298023224,
      "reward_std": 0.2989003658294678,
      "rewards/accuracy_reward": 0.3750000298023224,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "completion_length": 602.857177734375,
      "epoch": 0.03411513859275053,
      "grad_norm": 0.5179570913314819,
      "grpo_loss": -0.054881345480680466,
      "kl": 6.580352783203125e-05,
      "learning_rate": 1.0212765957446809e-06,
      "loss": 0.0,
      "reward": 0.3660714328289032,
      "reward_std": 0.357073038816452,
      "rewards/accuracy_reward": 0.3660714328289032,
      "rewards/format_reward": 0.0,
      "step": 16
    },
    {
      "completion_length": 634.8035888671875,
      "epoch": 0.03624733475479744,
      "grad_norm": 0.45830434560775757,
      "grpo_loss": -0.07317615300416946,
      "kl": 5.650520324707031e-05,
      "learning_rate": 1.0851063829787236e-06,
      "loss": 0.0,
      "reward": 0.3839285969734192,
      "reward_std": 0.4347257912158966,
      "rewards/accuracy_reward": 0.3839285969734192,
      "rewards/format_reward": 0.0,
      "step": 17
    },
    {
      "completion_length": 576.4732666015625,
      "epoch": 0.03837953091684435,
      "grad_norm": 0.5573252439498901,
      "grpo_loss": -0.07317613065242767,
      "kl": 6.246566772460938e-05,
      "learning_rate": 1.148936170212766e-06,
      "loss": 0.0,
      "reward": 0.6339285969734192,
      "reward_std": 0.4847923815250397,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 18
    },
    {
      "completion_length": 632.6964721679688,
      "epoch": 0.04051172707889126,
      "grad_norm": 0.6752464771270752,
      "grpo_loss": 2.041459083557129e-06,
      "kl": 5.7697296142578125e-05,
      "learning_rate": 1.2127659574468085e-06,
      "loss": 0.0,
      "reward": 0.330357164144516,
      "reward_std": 0.3501092493534088,
      "rewards/accuracy_reward": 0.3214285969734192,
      "rewards/format_reward": 0.008928571827709675,
      "step": 19
    },
    {
      "completion_length": 641.8125,
      "epoch": 0.042643923240938165,
      "grad_norm": 1.7546889781951904,
      "grpo_loss": -0.054879434406757355,
      "kl": 0.00014495849609375,
      "learning_rate": 1.276595744680851e-06,
      "loss": 0.0,
      "reward": 0.2946428656578064,
      "reward_std": 0.4111029803752899,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.008928571827709675,
      "step": 20
    },
    {
      "completion_length": 589.169677734375,
      "epoch": 0.04477611940298507,
      "grad_norm": 0.4196793735027313,
      "grpo_loss": 0.054885976016521454,
      "kl": 7.677078247070312e-05,
      "learning_rate": 1.3404255319148935e-06,
      "loss": 0.0,
      "reward": 0.3839285969734192,
      "reward_std": 0.42379865050315857,
      "rewards/accuracy_reward": 0.3839285969734192,
      "rewards/format_reward": 0.0,
      "step": 21
    },
    {
      "completion_length": 655.5714721679688,
      "epoch": 0.046908315565031986,
      "grad_norm": 0.41036805510520935,
      "grpo_loss": 2.6362554308434483e-06,
      "kl": 8.821487426757812e-05,
      "learning_rate": 1.4042553191489362e-06,
      "loss": 0.0,
      "reward": 0.3392857313156128,
      "reward_std": 0.3501092493534088,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.0,
      "step": 22
    },
    {
      "completion_length": 641.6875,
      "epoch": 0.04904051172707889,
      "grad_norm": 3.576270580291748,
      "grpo_loss": 1.2814998626708984e-05,
      "kl": 0.0001163482666015625,
      "learning_rate": 1.4680851063829787e-06,
      "loss": 0.0,
      "reward": 0.3750000298023224,
      "reward_std": 0.3450036942958832,
      "rewards/accuracy_reward": 0.3750000298023224,
      "rewards/format_reward": 0.0,
      "step": 23
    },
    {
      "completion_length": 653.0267944335938,
      "epoch": 0.0511727078891258,
      "grad_norm": 1.1303625106811523,
      "grpo_loss": 1.1533498764038086e-05,
      "kl": 0.0001163482666015625,
      "learning_rate": 1.5319148936170212e-06,
      "loss": 0.0,
      "reward": 0.3660714328289032,
      "reward_std": 0.42785170674324036,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 0.008928571827709675,
      "step": 24
    },
    {
      "completion_length": 640.3928833007812,
      "epoch": 0.053304904051172705,
      "grad_norm": 5.7301344871521,
      "grpo_loss": 4.673004150390625e-05,
      "kl": 0.0005645751953125,
      "learning_rate": 1.5957446808510639e-06,
      "loss": 0.0,
      "reward": 0.3482142984867096,
      "reward_std": 0.33627134561538696,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.008928571827709675,
      "step": 25
    },
    {
      "completion_length": 709.2500610351562,
      "epoch": 0.05543710021321962,
      "grad_norm": 0.3460543155670166,
      "grpo_loss": 0.016702793538570404,
      "kl": 0.00013637542724609375,
      "learning_rate": 1.6595744680851066e-06,
      "loss": 0.0,
      "reward": 0.3035714328289032,
      "reward_std": 0.29598960280418396,
      "rewards/accuracy_reward": 0.3035714328289032,
      "rewards/format_reward": 0.0,
      "step": 26
    },
    {
      "completion_length": 467.8660888671875,
      "epoch": 0.057569296375266525,
      "grad_norm": 0.8011674880981445,
      "grpo_loss": -0.07316015660762787,
      "kl": 0.0003757476806640625,
      "learning_rate": 1.723404255319149e-06,
      "loss": 0.0,
      "reward": 0.5089285969734192,
      "reward_std": 0.393301784992218,
      "rewards/accuracy_reward": 0.5089285969734192,
      "rewards/format_reward": 0.0,
      "step": 27
    },
    {
      "completion_length": 659.2767944335938,
      "epoch": 0.05970149253731343,
      "grad_norm": 0.3461753726005554,
      "grpo_loss": -0.1336020827293396,
      "kl": 0.0002613067626953125,
      "learning_rate": 1.7872340425531915e-06,
      "loss": 0.0,
      "reward": 0.3035714328289032,
      "reward_std": 0.3432351350784302,
      "rewards/accuracy_reward": 0.3035714328289032,
      "rewards/format_reward": 0.0,
      "step": 28
    },
    {
      "completion_length": 722.732177734375,
      "epoch": 0.06183368869936034,
      "grad_norm": 0.4500427842140198,
      "grpo_loss": -0.05487678945064545,
      "kl": 0.000286102294921875,
      "learning_rate": 1.851063829787234e-06,
      "loss": 0.0,
      "reward": 0.3660714328289032,
      "reward_std": 0.3737320005893707,
      "rewards/accuracy_reward": 0.3660714328289032,
      "rewards/format_reward": 0.0,
      "step": 29
    },
    {
      "completion_length": 598.294677734375,
      "epoch": 0.06396588486140725,
      "grad_norm": 1.4335441589355469,
      "grpo_loss": -0.01668836548924446,
      "kl": 0.0007171630859375,
      "learning_rate": 1.9148936170212767e-06,
      "loss": 0.0,
      "reward": 0.3035714328289032,
      "reward_std": 0.37258976697921753,
      "rewards/accuracy_reward": 0.3035714328289032,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 576.669677734375,
      "epoch": 0.06609808102345416,
      "grad_norm": 0.637279748916626,
      "grpo_loss": -0.100145623087883,
      "kl": 0.000812530517578125,
      "learning_rate": 1.978723404255319e-06,
      "loss": 0.0,
      "reward": 0.330357164144516,
      "reward_std": 0.3501092493534088,
      "rewards/accuracy_reward": 0.330357164144516,
      "rewards/format_reward": 0.0,
      "step": 31
    },
    {
      "completion_length": 590.0357666015625,
      "epoch": 0.06823027718550106,
      "grad_norm": 1.1126184463500977,
      "grpo_loss": -0.01668006181716919,
      "kl": 0.00135040283203125,
      "learning_rate": 2.0425531914893617e-06,
      "loss": 0.0001,
      "reward": 0.3660714328289032,
      "reward_std": 0.33336055278778076,
      "rewards/accuracy_reward": 0.3660714328289032,
      "rewards/format_reward": 0.0,
      "step": 32
    },
    {
      "completion_length": 550.5892944335938,
      "epoch": 0.07036247334754797,
      "grad_norm": 0.5474586486816406,
      "grpo_loss": 6.0905444115633145e-05,
      "kl": 0.00099945068359375,
      "learning_rate": 2.1063829787234044e-06,
      "loss": 0.0,
      "reward": 0.3214285969734192,
      "reward_std": 0.3206649422645569,
      "rewards/accuracy_reward": 0.3214285969734192,
      "rewards/format_reward": 0.0,
      "step": 33
    },
    {
      "completion_length": 610.5892944335938,
      "epoch": 0.07249466950959488,
      "grad_norm": 0.6375254988670349,
      "grpo_loss": 0.10025076568126678,
      "kl": 0.008544921875,
      "learning_rate": 2.170212765957447e-06,
      "loss": 0.0003,
      "reward": 0.3750000298023224,
      "reward_std": 0.3903909921646118,
      "rewards/accuracy_reward": 0.3750000298023224,
      "rewards/format_reward": 0.0,
      "step": 34
    },
    {
      "completion_length": 633.2589721679688,
      "epoch": 0.07462686567164178,
      "grad_norm": 0.5401435494422913,
      "grpo_loss": -0.05477993190288544,
      "kl": 0.002655029296875,
      "learning_rate": 2.2340425531914894e-06,
      "loss": 0.0001,
      "reward": 0.3392857313156128,
      "reward_std": 0.35407260060310364,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "completion_length": 586.9732666015625,
      "epoch": 0.0767590618336887,
      "grad_norm": 0.5796173810958862,
      "grpo_loss": 0.05496525764465332,
      "kl": 0.00262451171875,
      "learning_rate": 2.297872340425532e-06,
      "loss": 0.0001,
      "reward": 0.5446428656578064,
      "reward_std": 0.4081922173500061,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 36
    },
    {
      "completion_length": 654.9017944335938,
      "epoch": 0.07889125799573561,
      "grad_norm": 0.5337525010108948,
      "grpo_loss": -0.0982007384300232,
      "kl": 0.0093994140625,
      "learning_rate": 2.3617021276595748e-06,
      "loss": 0.0004,
      "reward": 0.4107142984867096,
      "reward_std": 0.41005048155784607,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 0.0,
      "step": 37
    },
    {
      "completion_length": 615.4285888671875,
      "epoch": 0.08102345415778252,
      "grad_norm": 0.40850356221199036,
      "grpo_loss": -0.07306132465600967,
      "kl": 0.002777099609375,
      "learning_rate": 2.425531914893617e-06,
      "loss": 0.0001,
      "reward": 0.3482142984867096,
      "reward_std": 0.38642770051956177,
      "rewards/accuracy_reward": 0.3482142984867096,
      "rewards/format_reward": 0.0,
      "step": 38
    },
    {
      "completion_length": 645.3928833007812,
      "epoch": 0.08315565031982942,
      "grad_norm": 2.3292202949523926,
      "grpo_loss": 4.441291093826294e-05,
      "kl": 0.010009765625,
      "learning_rate": 2.4893617021276598e-06,
      "loss": 0.0004,
      "reward": 0.5,
      "reward_std": 0.4099607765674591,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "completion_length": 647.3303833007812,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.3452330231666565,
      "grpo_loss": -0.016505777835845947,
      "kl": 0.0033111572265625,
      "learning_rate": 2.553191489361702e-06,
      "loss": 0.0001,
      "reward": 0.535714328289032,
      "reward_std": 0.3580358922481537,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "completion_length": 680.4107666015625,
      "epoch": 0.08742004264392324,
      "grad_norm": 0.4223065972328186,
      "grpo_loss": 0.0549992173910141,
      "kl": 0.003387451171875,
      "learning_rate": 2.6170212765957447e-06,
      "loss": 0.0001,
      "reward": 0.4017857313156128,
      "reward_std": 0.3679104745388031,
      "rewards/accuracy_reward": 0.4017857313156128,
      "rewards/format_reward": 0.0,
      "step": 41
    },
    {
      "completion_length": 524.732177734375,
      "epoch": 0.08955223880597014,
      "grad_norm": 0.6519092321395874,
      "grpo_loss": -0.07294157892465591,
      "kl": 0.01025390625,
      "learning_rate": 2.680851063829787e-06,
      "loss": 0.0004,
      "reward": 0.5,
      "reward_std": 0.44742143154144287,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 42
    },
    {
      "completion_length": 653.4910888671875,
      "epoch": 0.09168443496801706,
      "grad_norm": 0.4341220557689667,
      "grpo_loss": 0.01703798398375511,
      "kl": 0.00616455078125,
      "learning_rate": 2.7446808510638297e-06,
      "loss": 0.0002,
      "reward": 0.4910714626312256,
      "reward_std": 0.30004259943962097,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 43
    },
    {
      "completion_length": 603.5535888671875,
      "epoch": 0.09381663113006397,
      "grad_norm": 1.1075998544692993,
      "grpo_loss": -0.05463188886642456,
      "kl": 0.010498046875,
      "learning_rate": 2.8085106382978724e-06,
      "loss": 0.0004,
      "reward": 0.625,
      "reward_std": 0.38351690769195557,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 44
    },
    {
      "completion_length": 577.9017944335938,
      "epoch": 0.09594882729211088,
      "grad_norm": 0.45368558168411255,
      "grpo_loss": 0.000235043466091156,
      "kl": 0.007110595703125,
      "learning_rate": 2.872340425531915e-06,
      "loss": 0.0003,
      "reward": 0.4821428656578064,
      "reward_std": 0.3490566909313202,
      "rewards/accuracy_reward": 0.4732142984867096,
      "rewards/format_reward": 0.008928571827709675,
      "step": 45
    },
    {
      "completion_length": 617.7410888671875,
      "epoch": 0.09808102345415778,
      "grad_norm": 0.4727841019630432,
      "grpo_loss": 0.07361437380313873,
      "kl": 0.0216064453125,
      "learning_rate": 2.9361702127659574e-06,
      "loss": 0.0009,
      "reward": 0.4375000298023224,
      "reward_std": 0.3638574182987213,
      "rewards/accuracy_reward": 0.4375000298023224,
      "rewards/format_reward": 0.0,
      "step": 46
    },
    {
      "completion_length": 638.0892944335938,
      "epoch": 0.10021321961620469,
      "grad_norm": 0.23905974626541138,
      "grpo_loss": 0.017111822962760925,
      "kl": 0.00933837890625,
      "learning_rate": 3e-06,
      "loss": 0.0004,
      "reward": 0.5714285969734192,
      "reward_std": 0.31264859437942505,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 47
    },
    {
      "completion_length": 640.9732666015625,
      "epoch": 0.1023454157782516,
      "grad_norm": 1.3742345571517944,
      "grpo_loss": 0.0002704933285713196,
      "kl": 0.029296875,
      "learning_rate": 2.9999584343954855e-06,
      "loss": 0.0012,
      "reward": 0.5,
      "reward_std": 0.3501092493534088,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 48
    },
    {
      "completion_length": 633.0892944335938,
      "epoch": 0.1044776119402985,
      "grad_norm": 0.33841472864151,
      "grpo_loss": -0.11769130825996399,
      "kl": 0.01080322265625,
      "learning_rate": 2.999833739885541e-06,
      "loss": 0.0004,
      "reward": 0.5535714626312256,
      "reward_std": 0.4169245660305023,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 49
    },
    {
      "completion_length": 516.9642944335938,
      "epoch": 0.10660980810234541,
      "grad_norm": 0.32168567180633545,
      "grpo_loss": 0.10090474039316177,
      "kl": 0.0263671875,
      "learning_rate": 2.9996259233808368e-06,
      "loss": 0.0011,
      "reward": 0.6696428656578064,
      "reward_std": 0.38351690769195557,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "completion_length": 669.3839721679688,
      "epoch": 0.10874200426439233,
      "grad_norm": 0.7801251411437988,
      "grpo_loss": -0.07295064628124237,
      "kl": 0.044677734375,
      "learning_rate": 2.9993349963987307e-06,
      "loss": 0.0018,
      "reward": 0.5535714626312256,
      "reward_std": 0.31379079818725586,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 51
    },
    {
      "completion_length": 662.794677734375,
      "epoch": 0.11087420042643924,
      "grad_norm": 0.7691836953163147,
      "grpo_loss": -0.1333162784576416,
      "kl": 0.03076171875,
      "learning_rate": 2.9989609750626313e-06,
      "loss": 0.0012,
      "reward": 0.455357164144516,
      "reward_std": 0.3195226788520813,
      "rewards/accuracy_reward": 0.455357164144516,
      "rewards/format_reward": 0.0,
      "step": 52
    },
    {
      "completion_length": 632.3928833007812,
      "epoch": 0.11300639658848614,
      "grad_norm": 5.331604957580566,
      "grpo_loss": 0.10157421231269836,
      "kl": 0.0255126953125,
      "learning_rate": 2.998503880101102e-06,
      "loss": 0.001,
      "reward": 0.5267857313156128,
      "reward_std": 0.3679104745388031,
      "rewards/accuracy_reward": 0.5267857313156128,
      "rewards/format_reward": 0.0,
      "step": 53
    },
    {
      "completion_length": 635.169677734375,
      "epoch": 0.11513859275053305,
      "grad_norm": 0.25747939944267273,
      "grpo_loss": 0.0004050950810778886,
      "kl": 0.006378173828125,
      "learning_rate": 2.9979637368467146e-06,
      "loss": 0.0003,
      "reward": 0.625,
      "reward_std": 0.3028636872768402,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 54
    },
    {
      "completion_length": 655.8035888671875,
      "epoch": 0.11727078891257996,
      "grad_norm": 26.88686752319336,
      "grpo_loss": 0.0002674385905265808,
      "kl": 0.59765625,
      "learning_rate": 2.9973405752346425e-06,
      "loss": 0.0237,
      "reward": 0.5446428656578064,
      "reward_std": 0.37664279341697693,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "completion_length": 681.5178833007812,
      "epoch": 0.11940298507462686,
      "grad_norm": 0.22809027135372162,
      "grpo_loss": 0.00032356235897168517,
      "kl": 0.00927734375,
      "learning_rate": 2.9966344298010057e-06,
      "loss": 0.0004,
      "reward": 0.5535714626312256,
      "reward_std": 0.29016801714897156,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 56
    },
    {
      "completion_length": 597.0357666015625,
      "epoch": 0.12153518123667377,
      "grad_norm": 0.9132996797561646,
      "grpo_loss": 0.047636114060878754,
      "kl": 0.032470703125,
      "learning_rate": 2.9958453396809527e-06,
      "loss": 0.0013,
      "reward": 0.5803571939468384,
      "reward_std": 0.2931685149669647,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 57
    },
    {
      "completion_length": 621.4107666015625,
      "epoch": 0.12366737739872068,
      "grad_norm": 0.29949259757995605,
      "grpo_loss": -0.015969865024089813,
      "kl": 0.01531982421875,
      "learning_rate": 2.994973348606494e-06,
      "loss": 0.0006,
      "reward": 0.535714328289032,
      "reward_std": 0.316791296005249,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 58
    },
    {
      "completion_length": 587.0178833007812,
      "epoch": 0.1257995735607676,
      "grad_norm": 0.3156592547893524,
      "grpo_loss": 0.0001606196165084839,
      "kl": 0.0283203125,
      "learning_rate": 2.994018504904078e-06,
      "loss": 0.0011,
      "reward": 0.598214328289032,
      "reward_std": 0.26945602893829346,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 59
    },
    {
      "completion_length": 538.6517944335938,
      "epoch": 0.1279317697228145,
      "grad_norm": 0.2868943214416504,
      "grpo_loss": 0.00023730844259262085,
      "kl": 0.0250244140625,
      "learning_rate": 2.9929808614919113e-06,
      "loss": 0.001,
      "reward": 0.7053571939468384,
      "reward_std": 0.25976085662841797,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "completion_length": 679.1875,
      "epoch": 0.1300639658848614,
      "grad_norm": 0.15767447650432587,
      "grpo_loss": 8.637533755972981e-05,
      "kl": 0.004791259765625,
      "learning_rate": 2.9918604758770298e-06,
      "loss": 0.0002,
      "reward": 0.5446428656578064,
      "reward_std": 0.24874404072761536,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 61
    },
    {
      "completion_length": 593.4464721679688,
      "epoch": 0.13219616204690832,
      "grad_norm": 0.20824034512043,
      "grpo_loss": 0.016763929277658463,
      "kl": 0.006134033203125,
      "learning_rate": 2.9906574101521067e-06,
      "loss": 0.0002,
      "reward": 0.5803571939468384,
      "reward_std": 0.3431454598903656,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 62
    },
    {
      "completion_length": 582.4107666015625,
      "epoch": 0.13432835820895522,
      "grad_norm": 0.32729294896125793,
      "grpo_loss": -0.09963612258434296,
      "kl": 0.017333984375,
      "learning_rate": 2.9893717309920135e-06,
      "loss": 0.0007,
      "reward": 0.6517857313156128,
      "reward_std": 0.26267164945602417,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.008928571827709675,
      "step": 63
    },
    {
      "completion_length": 706.3303833007812,
      "epoch": 0.13646055437100213,
      "grad_norm": 0.17250199615955353,
      "grpo_loss": -0.10007388144731522,
      "kl": 0.00469970703125,
      "learning_rate": 2.9880035096501265e-06,
      "loss": 0.0002,
      "reward": 0.4107142984867096,
      "reward_std": 0.3334502577781677,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 0.0,
      "step": 64
    },
    {
      "completion_length": 508.1875305175781,
      "epoch": 0.13859275053304904,
      "grad_norm": 0.18875280022621155,
      "grpo_loss": 0.10045573860406876,
      "kl": 0.01495361328125,
      "learning_rate": 2.9865528219543747e-06,
      "loss": 0.0006,
      "reward": 0.7589285969734192,
      "reward_std": 0.3254339098930359,
      "rewards/accuracy_reward": 0.7589285969734192,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "completion_length": 589.0625,
      "epoch": 0.14072494669509594,
      "grad_norm": 0.2253061980009079,
      "grpo_loss": 0.0006140768527984619,
      "kl": 0.024658203125,
      "learning_rate": 2.9850197483030394e-06,
      "loss": 0.001,
      "reward": 0.6875000596046448,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 66
    },
    {
      "completion_length": 553.5714721679688,
      "epoch": 0.14285714285714285,
      "grad_norm": 0.26960617303848267,
      "grpo_loss": 0.0003652747254818678,
      "kl": 0.020263671875,
      "learning_rate": 2.9834043736602984e-06,
      "loss": 0.0008,
      "reward": 0.7142857313156128,
      "reward_std": 0.2792409062385559,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 67
    },
    {
      "completion_length": 590.5357666015625,
      "epoch": 0.14498933901918976,
      "grad_norm": 0.18081212043762207,
      "grpo_loss": 0.055027492344379425,
      "kl": 0.01373291015625,
      "learning_rate": 2.9817067875515166e-06,
      "loss": 0.0005,
      "reward": 0.598214328289032,
      "reward_std": 0.351251482963562,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 68
    },
    {
      "completion_length": 573.75,
      "epoch": 0.14712153518123666,
      "grad_norm": 0.14580072462558746,
      "grpo_loss": 9.892135858535767e-05,
      "kl": 0.010009765625,
      "learning_rate": 2.9799270840582835e-06,
      "loss": 0.0004,
      "reward": 0.7321428656578064,
      "reward_std": 0.24001170694828033,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 69
    },
    {
      "completion_length": 549.3125,
      "epoch": 0.14925373134328357,
      "grad_norm": 0.16547168791294098,
      "grpo_loss": 0.0005362048395909369,
      "kl": 0.0089111328125,
      "learning_rate": 2.978065361813203e-06,
      "loss": 0.0004,
      "reward": 0.8660714626312256,
      "reward_std": 0.15152154862880707,
      "rewards/accuracy_reward": 0.8660714626312256,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "completion_length": 627.2410888671875,
      "epoch": 0.1513859275053305,
      "grad_norm": 0.2011519968509674,
      "grpo_loss": -0.04696042090654373,
      "kl": 0.00927734375,
      "learning_rate": 2.97612172399442e-06,
      "loss": 0.0004,
      "reward": 0.4821428656578064,
      "reward_std": 0.31961238384246826,
      "rewards/accuracy_reward": 0.4821428656578064,
      "rewards/format_reward": 0.0,
      "step": 71
    },
    {
      "completion_length": 707.5267944335938,
      "epoch": 0.1535181236673774,
      "grad_norm": 0.5303527116775513,
      "grpo_loss": 0.047381460666656494,
      "kl": 0.0166015625,
      "learning_rate": 2.9740962783199107e-06,
      "loss": 0.0007,
      "reward": 0.4910714626312256,
      "reward_std": 0.3098274767398834,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 72
    },
    {
      "completion_length": 553.2678833007812,
      "epoch": 0.15565031982942432,
      "grad_norm": 0.19617433845996857,
      "grpo_loss": 0.0549967885017395,
      "kl": 0.00518798828125,
      "learning_rate": 2.971989137041507e-06,
      "loss": 0.0002,
      "reward": 0.6696428656578064,
      "reward_std": 0.3748742640018463,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 73
    },
    {
      "completion_length": 639.7410888671875,
      "epoch": 0.15778251599147122,
      "grad_norm": 0.18848314881324768,
      "grpo_loss": -0.0726790651679039,
      "kl": 0.00927734375,
      "learning_rate": 2.969800416938676e-06,
      "loss": 0.0004,
      "reward": 0.5892857313156128,
      "reward_std": 0.25570783019065857,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 74
    },
    {
      "completion_length": 618.9285888671875,
      "epoch": 0.15991471215351813,
      "grad_norm": 0.15183982253074646,
      "grpo_loss": 0.0001978054642677307,
      "kl": 0.006591796875,
      "learning_rate": 2.967530239312051e-06,
      "loss": 0.0003,
      "reward": 0.5446428656578064,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "completion_length": 628.6428833007812,
      "epoch": 0.16204690831556504,
      "grad_norm": 0.17083314061164856,
      "grpo_loss": 0.0003498228034004569,
      "kl": 0.0093994140625,
      "learning_rate": 2.9651787299767046e-06,
      "loss": 0.0004,
      "reward": 0.5089285969734192,
      "reward_std": 0.3029533922672272,
      "rewards/accuracy_reward": 0.5089285969734192,
      "rewards/format_reward": 0.0,
      "step": 76
    },
    {
      "completion_length": 645.6964721679688,
      "epoch": 0.16417910447761194,
      "grad_norm": 0.1758926510810852,
      "grpo_loss": 0.04745073243975639,
      "kl": 0.01544189453125,
      "learning_rate": 2.9627460192551807e-06,
      "loss": 0.0006,
      "reward": 0.6517857313156128,
      "reward_std": 0.2027304321527481,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 77
    },
    {
      "completion_length": 548.2857666015625,
      "epoch": 0.16631130063965885,
      "grad_norm": 0.35794636607170105,
      "grpo_loss": 0.00010507553815841675,
      "kl": 0.0294189453125,
      "learning_rate": 2.960232241970268e-06,
      "loss": 0.0012,
      "reward": 0.7321428656578064,
      "reward_std": 0.3235756754875183,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 78
    },
    {
      "completion_length": 602.4107666015625,
      "epoch": 0.16844349680170576,
      "grad_norm": 0.40275028347969055,
      "grpo_loss": 0.00044640994747169316,
      "kl": 0.018310546875,
      "learning_rate": 2.957637537437529e-06,
      "loss": 0.0007,
      "reward": 0.5535714626312256,
      "reward_std": 0.20846228301525116,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 79
    },
    {
      "completion_length": 645.7589721679688,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.15431800484657288,
      "grpo_loss": 0.0002495869994163513,
      "kl": 0.00836181640625,
      "learning_rate": 2.9549620494575816e-06,
      "loss": 0.0003,
      "reward": 0.5892857313156128,
      "reward_std": 0.2430122047662735,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "completion_length": 612.6339721679688,
      "epoch": 0.17270788912579957,
      "grad_norm": 0.1417791098356247,
      "grpo_loss": 0.0002077631652355194,
      "kl": 0.00592041015625,
      "learning_rate": 2.952205926308125e-06,
      "loss": 0.0002,
      "reward": 0.6339285969734192,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 81
    },
    {
      "completion_length": 476.14288330078125,
      "epoch": 0.17484008528784648,
      "grad_norm": 0.16035664081573486,
      "grpo_loss": 0.00023966887965798378,
      "kl": 0.0093994140625,
      "learning_rate": 2.9493693207357266e-06,
      "loss": 0.0004,
      "reward": 0.7767857313156128,
      "reward_std": 0.22626349329948425,
      "rewards/accuracy_reward": 0.7767857313156128,
      "rewards/format_reward": 0.0,
      "step": 82
    },
    {
      "completion_length": 633.6785888671875,
      "epoch": 0.17697228144989338,
      "grad_norm": 0.22332347929477692,
      "grpo_loss": 0.05593012273311615,
      "kl": 0.01171875,
      "learning_rate": 2.946452389947353e-06,
      "loss": 0.0005,
      "reward": 0.5535714626312256,
      "reward_std": 0.33741360902786255,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 83
    },
    {
      "completion_length": 681.1428833007812,
      "epoch": 0.1791044776119403,
      "grad_norm": 0.16065801680088043,
      "grpo_loss": 0.01748570427298546,
      "kl": 0.00799560546875,
      "learning_rate": 2.9434552956016594e-06,
      "loss": 0.0003,
      "reward": 0.625,
      "reward_std": 0.33336058259010315,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 84
    },
    {
      "completion_length": 675.5357666015625,
      "epoch": 0.1812366737739872,
      "grad_norm": 0.17096219956874847,
      "grpo_loss": 0.04753614217042923,
      "kl": 0.009521484375,
      "learning_rate": 2.9403782038000303e-06,
      "loss": 0.0004,
      "reward": 0.4910714626312256,
      "reward_std": 0.35989415645599365,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "completion_length": 626.3482666015625,
      "epoch": 0.18336886993603413,
      "grad_norm": 0.16287542879581451,
      "grpo_loss": -0.07303746789693832,
      "kl": 0.006500244140625,
      "learning_rate": 2.9372212850773744e-06,
      "loss": 0.0003,
      "reward": 0.6428571939468384,
      "reward_std": 0.25279706716537476,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 86
    },
    {
      "completion_length": 617.419677734375,
      "epoch": 0.18550106609808104,
      "grad_norm": 0.2341645210981369,
      "grpo_loss": -0.11798445880413055,
      "kl": 0.006378173828125,
      "learning_rate": 2.9339847143926705e-06,
      "loss": 0.0003,
      "reward": 0.6071428656578064,
      "reward_std": 0.28338363766670227,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 87
    },
    {
      "completion_length": 664.5,
      "epoch": 0.18763326226012794,
      "grad_norm": 0.14599959552288055,
      "grpo_loss": 0.10042649507522583,
      "kl": 0.004547119140625,
      "learning_rate": 2.9306686711192755e-06,
      "loss": 0.0002,
      "reward": 0.535714328289032,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 88
    },
    {
      "completion_length": 574.3928833007812,
      "epoch": 0.18976545842217485,
      "grad_norm": 0.19283778965473175,
      "grpo_loss": 0.1338942050933838,
      "kl": 0.008056640625,
      "learning_rate": 2.92727333903498e-06,
      "loss": 0.0003,
      "reward": 0.7946428656578064,
      "reward_std": 0.15839563310146332,
      "rewards/accuracy_reward": 0.7946428656578064,
      "rewards/format_reward": 0.0,
      "step": 89
    },
    {
      "completion_length": 562.8125,
      "epoch": 0.19189765458422176,
      "grad_norm": 0.21768853068351746,
      "grpo_loss": 0.05516081675887108,
      "kl": 0.006927490234375,
      "learning_rate": 2.923798906311825e-06,
      "loss": 0.0003,
      "reward": 0.7767857313156128,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.7767857313156128,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "completion_length": 590.607177734375,
      "epoch": 0.19402985074626866,
      "grad_norm": 0.18265320360660553,
      "grpo_loss": 0.11892025172710419,
      "kl": 0.00775146484375,
      "learning_rate": 2.920245565505673e-06,
      "loss": 0.0003,
      "reward": 0.455357164144516,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.455357164144516,
      "rewards/format_reward": 0.0,
      "step": 91
    },
    {
      "completion_length": 556.732177734375,
      "epoch": 0.19616204690831557,
      "grad_norm": 0.21578750014305115,
      "grpo_loss": -0.04708147421479225,
      "kl": 0.00531005859375,
      "learning_rate": 2.916613513545535e-06,
      "loss": 0.0002,
      "reward": 0.6875000596046448,
      "reward_std": 0.3069167137145996,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 92
    },
    {
      "completion_length": 562.2232666015625,
      "epoch": 0.19829424307036247,
      "grad_norm": 1.9854023456573486,
      "grpo_loss": -0.07264155149459839,
      "kl": 0.07373046875,
      "learning_rate": 2.912902951722658e-06,
      "loss": 0.003,
      "reward": 0.6696428656578064,
      "reward_std": 0.29307880997657776,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 93
    },
    {
      "completion_length": 563.5535888671875,
      "epoch": 0.20042643923240938,
      "grad_norm": 0.19654212892055511,
      "grpo_loss": 0.0001397952437400818,
      "kl": 0.00885009765625,
      "learning_rate": 2.909114085679369e-06,
      "loss": 0.0004,
      "reward": 0.5625,
      "reward_std": 0.31961238384246826,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 94
    },
    {
      "completion_length": 564.732177734375,
      "epoch": 0.2025586353944563,
      "grad_norm": 0.306581050157547,
      "grpo_loss": -0.07296334207057953,
      "kl": 0.0213623046875,
      "learning_rate": 2.9052471253976785e-06,
      "loss": 0.0009,
      "reward": 0.7142857313156128,
      "reward_std": 0.26654526591300964,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 95
    },
    {
      "completion_length": 484.5000305175781,
      "epoch": 0.2046908315565032,
      "grad_norm": 0.0992506816983223,
      "grpo_loss": -0.0729239359498024,
      "kl": 0.00653076171875,
      "learning_rate": 2.9013022851876413e-06,
      "loss": 0.0003,
      "reward": 0.8928571939468384,
      "reward_std": 0.1514318436384201,
      "rewards/accuracy_reward": 0.8928571939468384,
      "rewards/format_reward": 0.0,
      "step": 96
    },
    {
      "completion_length": 566.6875,
      "epoch": 0.2068230277185501,
      "grad_norm": 0.16707083582878113,
      "grpo_loss": 6.328150629997253e-05,
      "kl": 0.006195068359375,
      "learning_rate": 2.897279783675483e-06,
      "loss": 0.0002,
      "reward": 0.7410714626312256,
      "reward_std": 0.25279706716537476,
      "rewards/accuracy_reward": 0.7410714626312256,
      "rewards/format_reward": 0.0,
      "step": 97
    },
    {
      "completion_length": 658.6964721679688,
      "epoch": 0.208955223880597,
      "grad_norm": 0.13153815269470215,
      "grpo_loss": 0.00013254210352897644,
      "kl": 0.006195068359375,
      "learning_rate": 2.893179843791478e-06,
      "loss": 0.0002,
      "reward": 0.5803571939468384,
      "reward_std": 0.2418699413537979,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.008928571827709675,
      "step": 98
    },
    {
      "completion_length": 604.9553833007812,
      "epoch": 0.21108742004264391,
      "grad_norm": 0.12218047678470612,
      "grpo_loss": -0.047130413353443146,
      "kl": 0.005096435546875,
      "learning_rate": 2.8890026927576e-06,
      "loss": 0.0002,
      "reward": 0.6428571939468384,
      "reward_std": 0.19471408426761627,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 99
    },
    {
      "completion_length": 474.1160888671875,
      "epoch": 0.21321961620469082,
      "grad_norm": 0.25340017676353455,
      "grpo_loss": 0.10033953934907913,
      "kl": 0.011962890625,
      "learning_rate": 2.8847485620749263e-06,
      "loss": 0.0005,
      "reward": 0.7321428656578064,
      "reward_std": 0.2724565267562866,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "epoch": 0.21321961620469082,
      "eval_completion_length": 609.0062072558906,
      "eval_grpo_loss": 0.00038164247871754457,
      "eval_kl": 0.014738820231379792,
      "eval_loss": 0.0005614032270386815,
      "eval_reward": 0.5379963731422973,
      "eval_reward_std": 0.29159833647953437,
      "eval_rewards/accuracy_reward": 0.5379393215948781,
      "eval_rewards/format_reward": 5.7051577173863736e-05,
      "eval_runtime": 4026.4829,
      "eval_samples_per_second": 1.242,
      "eval_steps_per_second": 0.011,
      "step": 100
    },
    {
      "completion_length": 607.4464721679688,
      "epoch": 0.21535181236673773,
      "grad_norm": 0.10197024047374725,
      "grpo_loss": 0.00038136146031320095,
      "kl": 0.00634765625,
      "learning_rate": 2.8804176875108078e-06,
      "loss": 0.0003,
      "reward": 0.6071428656578064,
      "reward_std": 0.14164696633815765,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.008928571827709675,
      "step": 101
    },
    {
      "completion_length": 553.2142944335938,
      "epoch": 0.21748400852878466,
      "grad_norm": 0.15817181766033173,
      "grpo_loss": 0.0006425325991585851,
      "kl": 0.01019287109375,
      "learning_rate": 2.876010309085804e-06,
      "loss": 0.0004,
      "reward": 0.6875000596046448,
      "reward_std": 0.1928558498620987,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 102
    },
    {
      "completion_length": 554.8214721679688,
      "epoch": 0.21961620469083157,
      "grad_norm": 0.3014904260635376,
      "grpo_loss": 0.000414554524468258,
      "kl": 0.01361083984375,
      "learning_rate": 2.8715266710603797e-06,
      "loss": 0.0005,
      "reward": 0.6696428656578064,
      "reward_std": 0.2625819444656372,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 103
    },
    {
      "completion_length": 684.0089721679688,
      "epoch": 0.22174840085287847,
      "grad_norm": 0.1303500086069107,
      "grpo_loss": 0.04732581973075867,
      "kl": 0.00628662109375,
      "learning_rate": 2.8669670219213677e-06,
      "loss": 0.0003,
      "reward": 0.4821428656578064,
      "reward_std": 0.216478630900383,
      "rewards/accuracy_reward": 0.4821428656578064,
      "rewards/format_reward": 0.0,
      "step": 104
    },
    {
      "completion_length": 553.3660888671875,
      "epoch": 0.22388059701492538,
      "grad_norm": 0.24533118307590485,
      "grpo_loss": 0.047773294150829315,
      "kl": 0.00714111328125,
      "learning_rate": 2.862331614368199e-06,
      "loss": 0.0003,
      "reward": 0.660714328289032,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 105
    },
    {
      "completion_length": 601.3125,
      "epoch": 0.2260127931769723,
      "grad_norm": 0.17875026166439056,
      "grpo_loss": 0.0001519266952527687,
      "kl": 0.00836181640625,
      "learning_rate": 2.857620705298896e-06,
      "loss": 0.0003,
      "reward": 0.6875000596046448,
      "reward_std": 0.26654526591300964,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 106
    },
    {
      "completion_length": 592.0625,
      "epoch": 0.2281449893390192,
      "grad_norm": 0.19455592334270477,
      "grpo_loss": 9.074594709090889e-05,
      "kl": 0.00518798828125,
      "learning_rate": 2.8528345557958363e-06,
      "loss": 0.0002,
      "reward": 0.6785714626312256,
      "reward_std": 0.222300186753273,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 107
    },
    {
      "completion_length": 607.6339721679688,
      "epoch": 0.2302771855010661,
      "grad_norm": 0.15129579603672028,
      "grpo_loss": 0.00017056881915777922,
      "kl": 0.00799560546875,
      "learning_rate": 2.847973431111284e-06,
      "loss": 0.0003,
      "reward": 0.6160714626312256,
      "reward_std": 0.22626349329948425,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 108
    },
    {
      "completion_length": 575.4107666015625,
      "epoch": 0.232409381663113,
      "grad_norm": 0.23053210973739624,
      "grpo_loss": 9.338387462776154e-05,
      "kl": 0.0098876953125,
      "learning_rate": 2.843037600652686e-06,
      "loss": 0.0004,
      "reward": 0.6071428656578064,
      "reward_std": 0.21542608737945557,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 109
    },
    {
      "completion_length": 634.8392944335938,
      "epoch": 0.2345415778251599,
      "grad_norm": 0.1780773103237152,
      "grpo_loss": 0.0002103373408317566,
      "kl": 0.0050048828125,
      "learning_rate": 2.8380273379677463e-06,
      "loss": 0.0002,
      "reward": 0.6517857313156128,
      "reward_std": 0.29598960280418396,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "completion_length": 601.3035888671875,
      "epoch": 0.23667377398720682,
      "grad_norm": 0.22908470034599304,
      "grpo_loss": -0.01645032875239849,
      "kl": 0.0106201171875,
      "learning_rate": 2.8329429207292592e-06,
      "loss": 0.0004,
      "reward": 0.6696428656578064,
      "reward_std": 0.31970205903053284,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 111
    },
    {
      "completion_length": 700.0714721679688,
      "epoch": 0.23880597014925373,
      "grad_norm": 0.1368950456380844,
      "grpo_loss": 0.0002479627728462219,
      "kl": 0.00531005859375,
      "learning_rate": 2.827784630719728e-06,
      "loss": 0.0002,
      "reward": 0.6160714626312256,
      "reward_std": 0.26945602893829346,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 112
    },
    {
      "completion_length": 594.7053833007812,
      "epoch": 0.24093816631130063,
      "grad_norm": 0.11902152001857758,
      "grpo_loss": 0.00034128251718357205,
      "kl": 0.005340576171875,
      "learning_rate": 2.8225527538157413e-06,
      "loss": 0.0002,
      "reward": 0.6428571939468384,
      "reward_std": 0.24697549641132355,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 113
    },
    {
      "completion_length": 556.8035888671875,
      "epoch": 0.24307036247334754,
      "grad_norm": 2.0720760822296143,
      "grpo_loss": 0.1181984543800354,
      "kl": 0.05810546875,
      "learning_rate": 2.8172475799721353e-06,
      "loss": 0.0023,
      "reward": 0.7678571939468384,
      "reward_std": 0.3040059506893158,
      "rewards/accuracy_reward": 0.7678571939468384,
      "rewards/format_reward": 0.0,
      "step": 114
    },
    {
      "completion_length": 609.9375,
      "epoch": 0.24520255863539445,
      "grad_norm": 0.45657312870025635,
      "grpo_loss": 0.00024568289518356323,
      "kl": 0.00994873046875,
      "learning_rate": 2.8118694032059203e-06,
      "loss": 0.0004,
      "reward": 0.5714285969734192,
      "reward_std": 0.29016801714897156,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 115
    },
    {
      "completion_length": 606.669677734375,
      "epoch": 0.24733475479744135,
      "grad_norm": 0.14336203038692474,
      "grpo_loss": 0.00017342311912216246,
      "kl": 0.00799560546875,
      "learning_rate": 2.806418521579987e-06,
      "loss": 0.0003,
      "reward": 0.6071428656578064,
      "reward_std": 0.2193893939256668,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 116
    },
    {
      "completion_length": 561.6607666015625,
      "epoch": 0.24946695095948826,
      "grad_norm": 0.21323221921920776,
      "grpo_loss": 0.00015168637037277222,
      "kl": 0.00628662109375,
      "learning_rate": 2.8008952371865888e-06,
      "loss": 0.0003,
      "reward": 0.7142857313156128,
      "reward_std": 0.3795536160469055,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 117
    },
    {
      "completion_length": 570.419677734375,
      "epoch": 0.2515991471215352,
      "grad_norm": 0.1808198243379593,
      "grpo_loss": 0.00019172427710145712,
      "kl": 0.00872802734375,
      "learning_rate": 2.795299856130599e-06,
      "loss": 0.0004,
      "reward": 0.5535714626312256,
      "reward_std": 0.3069167137145996,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 118
    },
    {
      "completion_length": 625.3660888671875,
      "epoch": 0.2537313432835821,
      "grad_norm": 8.979061126708984,
      "grpo_loss": 0.10089191794395447,
      "kl": 0.181640625,
      "learning_rate": 2.789632688512545e-06,
      "loss": 0.0073,
      "reward": 0.5,
      "reward_std": 0.2625819444656372,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 119
    },
    {
      "completion_length": 719.3482666015625,
      "epoch": 0.255863539445629,
      "grad_norm": 0.17377711832523346,
      "grpo_loss": 0.016832787543535233,
      "kl": 0.008544921875,
      "learning_rate": 2.783894048411425e-06,
      "loss": 0.0003,
      "reward": 0.4642857313156128,
      "reward_std": 0.22917428612709045,
      "rewards/accuracy_reward": 0.4642857313156128,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "completion_length": 520.7857666015625,
      "epoch": 0.2579957356076759,
      "grad_norm": 0.1450602114200592,
      "grpo_loss": 0.047628480941057205,
      "kl": 0.00958251953125,
      "learning_rate": 2.7780842538672983e-06,
      "loss": 0.0004,
      "reward": 0.7410714626312256,
      "reward_std": 0.2499759942293167,
      "rewards/accuracy_reward": 0.7410714626312256,
      "rewards/format_reward": 0.0,
      "step": 121
    },
    {
      "completion_length": 588.1339721679688,
      "epoch": 0.2601279317697228,
      "grad_norm": 0.24374298751354218,
      "grpo_loss": 0.00011707842350006104,
      "kl": 0.0140380859375,
      "learning_rate": 2.772203626863661e-06,
      "loss": 0.0006,
      "reward": 0.6071428656578064,
      "reward_std": 0.31670159101486206,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 122
    },
    {
      "completion_length": 539.0982666015625,
      "epoch": 0.2622601279317697,
      "grad_norm": 0.6004351377487183,
      "grpo_loss": 0.00021716207265853882,
      "kl": 0.01055908203125,
      "learning_rate": 2.766252493309603e-06,
      "loss": 0.0004,
      "reward": 0.6875000596046448,
      "reward_std": 0.2822414040565491,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 123
    },
    {
      "completion_length": 640.2232666015625,
      "epoch": 0.26439232409381663,
      "grad_norm": 0.15684480965137482,
      "grpo_loss": 0.0734117180109024,
      "kl": 0.004180908203125,
      "learning_rate": 2.7602311830217408e-06,
      "loss": 0.0002,
      "reward": 0.5446428656578064,
      "reward_std": 0.2692219913005829,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.008928571827709675,
      "step": 124
    },
    {
      "completion_length": 594.2767944335938,
      "epoch": 0.26652452025586354,
      "grad_norm": 0.15290771424770355,
      "grpo_loss": -0.046935826539993286,
      "kl": 0.00457763671875,
      "learning_rate": 2.754140029705945e-06,
      "loss": 0.0002,
      "reward": 0.5178571939468384,
      "reward_std": 0.28725728392601013,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 125
    },
    {
      "completion_length": 613.8839721679688,
      "epoch": 0.26865671641791045,
      "grad_norm": 0.2813848555088043,
      "grpo_loss": -0.016470883041620255,
      "kl": 0.00555419921875,
      "learning_rate": 2.7479793709388413e-06,
      "loss": 0.0002,
      "reward": 0.7053571939468384,
      "reward_std": 0.22521094977855682,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 126
    },
    {
      "completion_length": 580.6339721679688,
      "epoch": 0.27078891257995735,
      "grad_norm": 0.2711651027202606,
      "grpo_loss": 0.11836685985326767,
      "kl": 0.009033203125,
      "learning_rate": 2.741749548149105e-06,
      "loss": 0.0004,
      "reward": 0.8303571939468384,
      "reward_std": 0.2960793077945709,
      "rewards/accuracy_reward": 0.8303571939468384,
      "rewards/format_reward": 0.0,
      "step": 127
    },
    {
      "completion_length": 619.5357666015625,
      "epoch": 0.27292110874200426,
      "grad_norm": 0.18601906299591064,
      "grpo_loss": 0.04743209481239319,
      "kl": 0.006134033203125,
      "learning_rate": 2.735450906598535e-06,
      "loss": 0.0002,
      "reward": 0.7053571939468384,
      "reward_std": 0.3206648826599121,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 128
    },
    {
      "completion_length": 593.7142944335938,
      "epoch": 0.27505330490405117,
      "grad_norm": 0.30626460909843445,
      "grpo_loss": 0.000193767249584198,
      "kl": 0.00970458984375,
      "learning_rate": 2.7290837953629244e-06,
      "loss": 0.0004,
      "reward": 0.6785714626312256,
      "reward_std": 0.3795536160469055,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 129
    },
    {
      "completion_length": 705.8928833007812,
      "epoch": 0.2771855010660981,
      "grad_norm": 0.17629599571228027,
      "grpo_loss": 0.00018097274005413055,
      "kl": 0.006988525390625,
      "learning_rate": 2.722648567312709e-06,
      "loss": 0.0003,
      "reward": 0.5,
      "reward_std": 0.2625819444656372,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "completion_length": 622.5982666015625,
      "epoch": 0.279317697228145,
      "grad_norm": 0.22440823912620544,
      "grpo_loss": 0.11841889470815659,
      "kl": 0.00799560546875,
      "learning_rate": 2.716145579093415e-06,
      "loss": 0.0003,
      "reward": 0.6071428656578064,
      "reward_std": 0.30004259943962097,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 131
    },
    {
      "completion_length": 673.0625,
      "epoch": 0.2814498933901919,
      "grad_norm": 0.17421214282512665,
      "grpo_loss": 0.055158667266368866,
      "kl": 0.00592041015625,
      "learning_rate": 2.70957519110589e-06,
      "loss": 0.0002,
      "reward": 0.455357164144516,
      "reward_std": 0.33336055278778076,
      "rewards/accuracy_reward": 0.455357164144516,
      "rewards/format_reward": 0.0,
      "step": 132
    },
    {
      "completion_length": 579.982177734375,
      "epoch": 0.2835820895522388,
      "grad_norm": 0.21082566678524017,
      "grpo_loss": 0.04748784005641937,
      "kl": 0.007293701171875,
      "learning_rate": 2.702937767486333e-06,
      "loss": 0.0003,
      "reward": 0.535714328289032,
      "reward_std": 0.34323516488075256,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 133
    },
    {
      "completion_length": 648.9107666015625,
      "epoch": 0.2857142857142857,
      "grad_norm": 0.16841772198677063,
      "grpo_loss": 0.017035316675901413,
      "kl": 0.00640869140625,
      "learning_rate": 2.6962336760861105e-06,
      "loss": 0.0003,
      "reward": 0.6339285969734192,
      "reward_std": 0.2891155183315277,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 134
    },
    {
      "completion_length": 579.5,
      "epoch": 0.2878464818763326,
      "grad_norm": 0.15011237561702728,
      "grpo_loss": 0.00023609399795532227,
      "kl": 0.004119873046875,
      "learning_rate": 2.689463288451372e-06,
      "loss": 0.0002,
      "reward": 0.6696428656578064,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 135
    },
    {
      "completion_length": 542.25,
      "epoch": 0.2899786780383795,
      "grad_norm": 0.23496368527412415,
      "grpo_loss": 0.0005470961332321167,
      "kl": 0.00994873046875,
      "learning_rate": 2.6826269798024567e-06,
      "loss": 0.0004,
      "reward": 0.6696428656578064,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 136
    },
    {
      "completion_length": 660.232177734375,
      "epoch": 0.2921108742004264,
      "grad_norm": 0.1643698662519455,
      "grpo_loss": 0.05503631383180618,
      "kl": 0.0052490234375,
      "learning_rate": 2.6757251290131003e-06,
      "loss": 0.0002,
      "reward": 0.6517857313156128,
      "reward_std": 0.33450281620025635,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 137
    },
    {
      "completion_length": 597.4375,
      "epoch": 0.2942430703624733,
      "grad_norm": 0.3305177092552185,
      "grpo_loss": -0.07286404073238373,
      "kl": 0.0123291015625,
      "learning_rate": 2.6687581185894363e-06,
      "loss": 0.0005,
      "reward": 0.6339285969734192,
      "reward_std": 0.46417009830474854,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 138
    },
    {
      "completion_length": 606.5803833007812,
      "epoch": 0.29637526652452023,
      "grad_norm": 0.15607373416423798,
      "grpo_loss": -0.09995691478252411,
      "kl": 0.0059814453125,
      "learning_rate": 2.6617263346487987e-06,
      "loss": 0.0002,
      "reward": 0.5803571939468384,
      "reward_std": 0.22812172770500183,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 139
    },
    {
      "completion_length": 636.4017944335938,
      "epoch": 0.29850746268656714,
      "grad_norm": 0.4233054518699646,
      "grpo_loss": -0.04692365229129791,
      "kl": 0.007720947265625,
      "learning_rate": 2.6546301668983207e-06,
      "loss": 0.0003,
      "reward": 0.6160714626312256,
      "reward_std": 0.36280491948127747,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "completion_length": 654.5625,
      "epoch": 0.3006396588486141,
      "grad_norm": 0.30437514185905457,
      "grpo_loss": -0.04689915478229523,
      "kl": 0.006378173828125,
      "learning_rate": 2.6474700086133384e-06,
      "loss": 0.0003,
      "reward": 0.5803571939468384,
      "reward_std": 0.3363610506057739,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 141
    },
    {
      "completion_length": 645.0625,
      "epoch": 0.302771855010661,
      "grad_norm": 0.16769744455814362,
      "grpo_loss": 0.00032076984643936157,
      "kl": 0.0069580078125,
      "learning_rate": 2.640246256615596e-06,
      "loss": 0.0003,
      "reward": 0.5535714626312256,
      "reward_std": 0.2803831696510315,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 142
    },
    {
      "completion_length": 600.5892944335938,
      "epoch": 0.3049040511727079,
      "grad_norm": 0.185498908162117,
      "grpo_loss": 0.0009974241256713867,
      "kl": 0.0123291015625,
      "learning_rate": 2.632959311251251e-06,
      "loss": 0.0005,
      "reward": 0.5535714626312256,
      "reward_std": 0.27641981840133667,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 143
    },
    {
      "completion_length": 667.2678833007812,
      "epoch": 0.3070362473347548,
      "grad_norm": 0.18703410029411316,
      "grpo_loss": 7.99037516117096e-05,
      "kl": 0.00555419921875,
      "learning_rate": 2.625609576368689e-06,
      "loss": 0.0002,
      "reward": 0.5446428656578064,
      "reward_std": 0.33354002237319946,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 144
    },
    {
      "completion_length": 594.232177734375,
      "epoch": 0.3091684434968017,
      "grad_norm": 0.17264287173748016,
      "grpo_loss": 0.0005372203886508942,
      "kl": 0.0086669921875,
      "learning_rate": 2.6181974592961413e-06,
      "loss": 0.0003,
      "reward": 0.7142857313156128,
      "reward_std": 0.23613807559013367,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "completion_length": 636.4910888671875,
      "epoch": 0.31130063965884863,
      "grad_norm": 0.34566304087638855,
      "grpo_loss": -0.047039397060871124,
      "kl": 0.0096435546875,
      "learning_rate": 2.610723370819111e-06,
      "loss": 0.0004,
      "reward": 0.6160714626312256,
      "reward_std": 0.29713183641433716,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 146
    },
    {
      "completion_length": 693.5000610351562,
      "epoch": 0.31343283582089554,
      "grad_norm": 0.1667724996805191,
      "grpo_loss": -0.09995532780885696,
      "kl": 0.005706787109375,
      "learning_rate": 2.6031877251576055e-06,
      "loss": 0.0002,
      "reward": 0.5892857313156128,
      "reward_std": 0.24195964634418488,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 147
    },
    {
      "completion_length": 625.357177734375,
      "epoch": 0.31556503198294245,
      "grad_norm": 0.2609766125679016,
      "grpo_loss": 0.0003863796591758728,
      "kl": 0.006622314453125,
      "learning_rate": 2.5955909399431797e-06,
      "loss": 0.0003,
      "reward": 0.5178571939468384,
      "reward_std": 0.21824714541435242,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 148
    },
    {
      "completion_length": 539.3482666015625,
      "epoch": 0.31769722814498935,
      "grad_norm": 0.20627616345882416,
      "grpo_loss": 0.04802427440881729,
      "kl": 0.0081787109375,
      "learning_rate": 2.5879334361957955e-06,
      "loss": 0.0003,
      "reward": 0.6964285969734192,
      "reward_std": 0.32366544008255005,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 149
    },
    {
      "completion_length": 622.8214721679688,
      "epoch": 0.31982942430703626,
      "grad_norm": 0.1990203708410263,
      "grpo_loss": 0.00011615455150604248,
      "kl": 0.01055908203125,
      "learning_rate": 2.5802156383004816e-06,
      "loss": 0.0004,
      "reward": 0.7142857313156128,
      "reward_std": 0.24997597932815552,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "completion_length": 603.875,
      "epoch": 0.32196162046908317,
      "grad_norm": 0.2065436840057373,
      "grpo_loss": -0.047058749943971634,
      "kl": 0.01055908203125,
      "learning_rate": 2.572437973983818e-06,
      "loss": 0.0004,
      "reward": 0.625,
      "reward_std": 0.33159202337265015,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 151
    },
    {
      "completion_length": 576.2232666015625,
      "epoch": 0.32409381663113007,
      "grad_norm": 0.2540753185749054,
      "grpo_loss": -0.04715569317340851,
      "kl": 0.007293701171875,
      "learning_rate": 2.5646008742902307e-06,
      "loss": 0.0003,
      "reward": 0.7321428656578064,
      "reward_std": 0.23208504915237427,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 152
    },
    {
      "completion_length": 596.4910888671875,
      "epoch": 0.326226012793177,
      "grad_norm": 0.31783199310302734,
      "grpo_loss": 0.00040250414167530835,
      "kl": 0.0098876953125,
      "learning_rate": 2.556704773558101e-06,
      "loss": 0.0004,
      "reward": 0.598214328289032,
      "reward_std": 0.2625819444656372,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 153
    },
    {
      "completion_length": 665.375,
      "epoch": 0.3283582089552239,
      "grad_norm": 0.19316673278808594,
      "grpo_loss": -0.0164567232131958,
      "kl": 0.0067138671875,
      "learning_rate": 2.5487501093956955e-06,
      "loss": 0.0003,
      "reward": 0.5178571939468384,
      "reward_std": 0.4376365542411804,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 154
    },
    {
      "completion_length": 576.1517944335938,
      "epoch": 0.3304904051172708,
      "grad_norm": 0.2708187997341156,
      "grpo_loss": -0.046256136149168015,
      "kl": 0.0133056640625,
      "learning_rate": 2.540737322656915e-06,
      "loss": 0.0005,
      "reward": 0.6517857313156128,
      "reward_std": 0.38756993412971497,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 155
    },
    {
      "completion_length": 585.6160888671875,
      "epoch": 0.3326226012793177,
      "grad_norm": 0.3584955930709839,
      "grpo_loss": 0.00033188471570611,
      "kl": 0.0206298828125,
      "learning_rate": 2.532666857416858e-06,
      "loss": 0.0008,
      "reward": 0.6517857313156128,
      "reward_std": 0.2586185932159424,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 156
    },
    {
      "completion_length": 408.6875305175781,
      "epoch": 0.3347547974413646,
      "grad_norm": 0.290502667427063,
      "grpo_loss": 0.0005320041673257947,
      "kl": 0.0177001953125,
      "learning_rate": 2.524539160947213e-06,
      "loss": 0.0007,
      "reward": 0.8303571939468384,
      "reward_std": 0.2637241780757904,
      "rewards/accuracy_reward": 0.8303571939468384,
      "rewards/format_reward": 0.0,
      "step": 157
    },
    {
      "completion_length": 634.544677734375,
      "epoch": 0.3368869936034115,
      "grad_norm": 0.30897942185401917,
      "grpo_loss": 0.0007045269012451172,
      "kl": 0.0135498046875,
      "learning_rate": 2.5163546836914705e-06,
      "loss": 0.0005,
      "reward": 0.598214328289032,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 158
    },
    {
      "completion_length": 519.1160888671875,
      "epoch": 0.3390191897654584,
      "grad_norm": 0.17009268701076508,
      "grpo_loss": 0.00013849325478076935,
      "kl": 0.00860595703125,
      "learning_rate": 2.5081138792399558e-06,
      "loss": 0.0003,
      "reward": 0.8392857313156128,
      "reward_std": 0.20564121007919312,
      "rewards/accuracy_reward": 0.8392857313156128,
      "rewards/format_reward": 0.0,
      "step": 159
    },
    {
      "completion_length": 649.6785888671875,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.20049817860126495,
      "grpo_loss": 0.00011303433711873367,
      "kl": 0.0062255859375,
      "learning_rate": 2.499817204304694e-06,
      "loss": 0.0002,
      "reward": 0.392857164144516,
      "reward_std": 0.2153363674879074,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "completion_length": 584.1607666015625,
      "epoch": 0.34328358208955223,
      "grad_norm": 0.32430610060691833,
      "grpo_loss": 0.0003687143325805664,
      "kl": 0.01513671875,
      "learning_rate": 2.491465118694097e-06,
      "loss": 0.0006,
      "reward": 0.6696428656578064,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 161
    },
    {
      "completion_length": 661.607177734375,
      "epoch": 0.34541577825159914,
      "grad_norm": 0.11526723206043243,
      "grpo_loss": 0.00017088651657104492,
      "kl": 0.0135498046875,
      "learning_rate": 2.483058085287483e-06,
      "loss": 0.0005,
      "reward": 0.598214328289032,
      "reward_std": 0.1928558498620987,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 162
    },
    {
      "completion_length": 644.357177734375,
      "epoch": 0.34754797441364604,
      "grad_norm": 0.16817770898342133,
      "grpo_loss": 0.055027663707733154,
      "kl": 0.00567626953125,
      "learning_rate": 2.474596570009417e-06,
      "loss": 0.0002,
      "reward": 0.6517857313156128,
      "reward_std": 0.23904886841773987,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 163
    },
    {
      "completion_length": 600.8928833007812,
      "epoch": 0.34968017057569295,
      "grad_norm": 0.2603262662887573,
      "grpo_loss": -0.04644265025854111,
      "kl": 0.021728515625,
      "learning_rate": 2.4660810418038984e-06,
      "loss": 0.0009,
      "reward": 0.6964285969734192,
      "reward_std": 0.2862047255039215,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 164
    },
    {
      "completion_length": 667.419677734375,
      "epoch": 0.35181236673773986,
      "grad_norm": 0.1652158796787262,
      "grpo_loss": 7.62939453125e-05,
      "kl": 0.0089111328125,
      "learning_rate": 2.4575119726083624e-06,
      "loss": 0.0004,
      "reward": 0.6339285969734192,
      "reward_std": 0.2389591485261917,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 165
    },
    {
      "completion_length": 694.2142944335938,
      "epoch": 0.35394456289978676,
      "grad_norm": 0.16855111718177795,
      "grpo_loss": 0.00020201876759529114,
      "kl": 0.01031494140625,
      "learning_rate": 2.448889837327531e-06,
      "loss": 0.0004,
      "reward": 0.5803571939468384,
      "reward_std": 0.35187777876853943,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 166
    },
    {
      "completion_length": 568.6160888671875,
      "epoch": 0.35607675906183367,
      "grad_norm": 0.21794408559799194,
      "grpo_loss": 0.01685473695397377,
      "kl": 0.006439208984375,
      "learning_rate": 2.440215113807091e-06,
      "loss": 0.0003,
      "reward": 0.6071428656578064,
      "reward_std": 0.40203410387039185,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 167
    },
    {
      "completion_length": 555.6160888671875,
      "epoch": 0.3582089552238806,
      "grad_norm": 0.2721169888973236,
      "grpo_loss": -0.0713144987821579,
      "kl": 0.01556396484375,
      "learning_rate": 2.4314882828072126e-06,
      "loss": 0.0006,
      "reward": 0.723214328289032,
      "reward_std": 0.21251529455184937,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 168
    },
    {
      "completion_length": 576.1339721679688,
      "epoch": 0.3603411513859275,
      "grad_norm": 0.22736425697803497,
      "grpo_loss": 0.05517146736383438,
      "kl": 0.0185546875,
      "learning_rate": 2.422709827975903e-06,
      "loss": 0.0007,
      "reward": 0.7678571939468384,
      "reward_std": 0.27350908517837524,
      "rewards/accuracy_reward": 0.7678571939468384,
      "rewards/format_reward": 0.0,
      "step": 169
    },
    {
      "completion_length": 586.4910888671875,
      "epoch": 0.3624733475479744,
      "grad_norm": 0.21218353509902954,
      "grpo_loss": 0.18360236287117004,
      "kl": 0.01043701171875,
      "learning_rate": 2.413880235822205e-06,
      "loss": 0.0004,
      "reward": 0.6517857313156128,
      "reward_std": 0.31961238384246826,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "completion_length": 623.1339721679688,
      "epoch": 0.3646055437100213,
      "grad_norm": 0.19818057119846344,
      "grpo_loss": -0.1175161674618721,
      "kl": 0.01446533203125,
      "learning_rate": 2.4049999956892328e-06,
      "loss": 0.0006,
      "reward": 0.6964285969734192,
      "reward_std": 0.3069167137145996,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 171
    },
    {
      "completion_length": 565.6875,
      "epoch": 0.36673773987206826,
      "grad_norm": 0.18358924984931946,
      "grpo_loss": 0.055470846593379974,
      "kl": 0.01068115234375,
      "learning_rate": 2.396069599727051e-06,
      "loss": 0.0004,
      "reward": 0.6785714626312256,
      "reward_std": 0.23322731256484985,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 172
    },
    {
      "completion_length": 644.7410888671875,
      "epoch": 0.36886993603411516,
      "grad_norm": 0.09578749537467957,
      "grpo_loss": 5.925947334617376e-05,
      "kl": 0.00555419921875,
      "learning_rate": 2.3870895428654033e-06,
      "loss": 0.0002,
      "reward": 0.6160714626312256,
      "reward_std": 0.1986774057149887,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 173
    },
    {
      "completion_length": 561.4553833007812,
      "epoch": 0.37100213219616207,
      "grad_norm": 0.1927262395620346,
      "grpo_loss": 0.05529294162988663,
      "kl": 0.0096435546875,
      "learning_rate": 2.3780603227862782e-06,
      "loss": 0.0004,
      "reward": 0.7053571939468384,
      "reward_std": 0.3276287019252777,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 174
    },
    {
      "completion_length": 581.4642944335938,
      "epoch": 0.373134328358209,
      "grad_norm": 0.2750888466835022,
      "grpo_loss": -0.047094203531742096,
      "kl": 0.014404296875,
      "learning_rate": 2.3689824398963307e-06,
      "loss": 0.0006,
      "reward": 0.723214328289032,
      "reward_std": 0.29713186621665955,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 175
    },
    {
      "completion_length": 658.9910888671875,
      "epoch": 0.3752665245202559,
      "grad_norm": 0.1884259432554245,
      "grpo_loss": 0.0011673122644424438,
      "kl": 0.01190185546875,
      "learning_rate": 2.3598563972991476e-06,
      "loss": 0.0005,
      "reward": 0.4732142984867096,
      "reward_std": 0.21251530945301056,
      "rewards/accuracy_reward": 0.4732142984867096,
      "rewards/format_reward": 0.0,
      "step": 176
    },
    {
      "completion_length": 614.5089721679688,
      "epoch": 0.3773987206823028,
      "grad_norm": 0.1662159264087677,
      "grpo_loss": 0.0007153626647777855,
      "kl": 0.0152587890625,
      "learning_rate": 2.350682700767365e-06,
      "loss": 0.0006,
      "reward": 0.7321428656578064,
      "reward_std": 0.16130642592906952,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 177
    },
    {
      "completion_length": 615.1875,
      "epoch": 0.3795309168443497,
      "grad_norm": 0.2590639293193817,
      "grpo_loss": 0.00103016197681427,
      "kl": 0.0233154296875,
      "learning_rate": 2.3414618587146394e-06,
      "loss": 0.0009,
      "reward": 0.5625,
      "reward_std": 0.3599838316440582,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 178
    },
    {
      "completion_length": 653.2142944335938,
      "epoch": 0.3816631130063966,
      "grad_norm": 0.1929996758699417,
      "grpo_loss": -0.0464971587061882,
      "kl": 0.0118408203125,
      "learning_rate": 2.3321943821674686e-06,
      "loss": 0.0005,
      "reward": 0.5625,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 179
    },
    {
      "completion_length": 566.169677734375,
      "epoch": 0.3837953091684435,
      "grad_norm": 0.17937371134757996,
      "grpo_loss": 0.0013021007180213928,
      "kl": 0.01483154296875,
      "learning_rate": 2.3228807847368733e-06,
      "loss": 0.0006,
      "reward": 0.6428571939468384,
      "reward_std": 0.34323516488075256,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "completion_length": 613.7410888671875,
      "epoch": 0.3859275053304904,
      "grad_norm": 0.08274456858634949,
      "grpo_loss": 0.01679709553718567,
      "kl": 0.0079345703125,
      "learning_rate": 2.3135215825899298e-06,
      "loss": 0.0003,
      "reward": 0.7321428656578064,
      "reward_std": 0.20846226811408997,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 181
    },
    {
      "completion_length": 590.2232666015625,
      "epoch": 0.3880597014925373,
      "grad_norm": 0.22545363008975983,
      "grpo_loss": 0.0002574026584625244,
      "kl": 0.00885009765625,
      "learning_rate": 2.304117294421165e-06,
      "loss": 0.0004,
      "reward": 0.5803571939468384,
      "reward_std": 0.25976085662841797,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 182
    },
    {
      "completion_length": 572.4464721679688,
      "epoch": 0.39019189765458423,
      "grad_norm": 0.2379264086484909,
      "grpo_loss": 0.0008170232176780701,
      "kl": 0.01513671875,
      "learning_rate": 2.294668441423809e-06,
      "loss": 0.0006,
      "reward": 0.625,
      "reward_std": 0.27236682176589966,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 183
    },
    {
      "completion_length": 597.5267944335938,
      "epoch": 0.39232409381663114,
      "grad_norm": 0.19797514379024506,
      "grpo_loss": 0.00013130396837368608,
      "kl": 0.00946044921875,
      "learning_rate": 2.2851755472609112e-06,
      "loss": 0.0004,
      "reward": 0.6428571939468384,
      "reward_std": 0.19180330634117126,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 184
    },
    {
      "completion_length": 635.2142944335938,
      "epoch": 0.39445628997867804,
      "grad_norm": 0.15277421474456787,
      "grpo_loss": 0.04737038165330887,
      "kl": 0.006622314453125,
      "learning_rate": 2.2756391380363176e-06,
      "loss": 0.0003,
      "reward": 0.4910714626312256,
      "reward_std": 0.29016801714897156,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 185
    },
    {
      "completion_length": 602.2767944335938,
      "epoch": 0.39658848614072495,
      "grad_norm": 0.8596810698509216,
      "grpo_loss": 0.11840670555830002,
      "kl": 0.0244140625,
      "learning_rate": 2.2660597422655136e-06,
      "loss": 0.001,
      "reward": 0.6517857313156128,
      "reward_std": 0.17223355174064636,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 186
    },
    {
      "completion_length": 629.0714721679688,
      "epoch": 0.39872068230277186,
      "grad_norm": 0.23284487426280975,
      "grpo_loss": -0.014619708061218262,
      "kl": 0.0194091796875,
      "learning_rate": 2.2564378908463343e-06,
      "loss": 0.0008,
      "reward": 0.455357164144516,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.455357164144516,
      "rewards/format_reward": 0.0,
      "step": 187
    },
    {
      "completion_length": 644.25,
      "epoch": 0.40085287846481876,
      "grad_norm": 0.16274768114089966,
      "grpo_loss": 0.0004352938267402351,
      "kl": 0.01171875,
      "learning_rate": 2.24677411702954e-06,
      "loss": 0.0005,
      "reward": 0.6339285969734192,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 188
    },
    {
      "completion_length": 567.6875,
      "epoch": 0.40298507462686567,
      "grad_norm": 0.3285521864891052,
      "grpo_loss": 0.07440680265426636,
      "kl": 0.031494140625,
      "learning_rate": 2.237068956389266e-06,
      "loss": 0.0013,
      "reward": 0.660714328289032,
      "reward_std": 0.3069167137145996,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 189
    },
    {
      "completion_length": 634.5625,
      "epoch": 0.4051172707889126,
      "grad_norm": 0.2191726267337799,
      "grpo_loss": 0.00019488348334562033,
      "kl": 0.0205078125,
      "learning_rate": 2.227322946793337e-06,
      "loss": 0.0008,
      "reward": 0.6339285969734192,
      "reward_std": 0.29307880997657776,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 190
    },
    {
      "completion_length": 517.9910888671875,
      "epoch": 0.4072494669509595,
      "grad_norm": 0.2723895311355591,
      "grpo_loss": 0.05497168004512787,
      "kl": 0.0152587890625,
      "learning_rate": 2.21753662837346e-06,
      "loss": 0.0006,
      "reward": 0.7500000596046448,
      "reward_std": 0.4014078378677368,
      "rewards/accuracy_reward": 0.7500000596046448,
      "rewards/format_reward": 0.0,
      "step": 191
    },
    {
      "completion_length": 660.4107666015625,
      "epoch": 0.4093816631130064,
      "grad_norm": 0.343417763710022,
      "grpo_loss": -0.015305643901228905,
      "kl": 0.035888671875,
      "learning_rate": 2.20771054349529e-06,
      "loss": 0.0014,
      "reward": 0.6517857313156128,
      "reward_std": 0.3334502577781677,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 192
    },
    {
      "completion_length": 505.982177734375,
      "epoch": 0.4115138592750533,
      "grad_norm": 0.2592337131500244,
      "grpo_loss": -0.01446167379617691,
      "kl": 0.0235595703125,
      "learning_rate": 2.197845236728371e-06,
      "loss": 0.0009,
      "reward": 0.7678571939468384,
      "reward_std": 0.305774450302124,
      "rewards/accuracy_reward": 0.7678571939468384,
      "rewards/format_reward": 0.0,
      "step": 193
    },
    {
      "completion_length": 547.2589721679688,
      "epoch": 0.4136460554371002,
      "grad_norm": 0.1847042590379715,
      "grpo_loss": 0.00015539950982201844,
      "kl": 0.0157470703125,
      "learning_rate": 2.187941254815956e-06,
      "loss": 0.0006,
      "reward": 0.7589285969734192,
      "reward_std": 0.216478630900383,
      "rewards/accuracy_reward": 0.7589285969734192,
      "rewards/format_reward": 0.0,
      "step": 194
    },
    {
      "completion_length": 629.1428833007812,
      "epoch": 0.4157782515991471,
      "grad_norm": 0.22321003675460815,
      "grpo_loss": 0.055582527071237564,
      "kl": 0.0302734375,
      "learning_rate": 2.1779991466447045e-06,
      "loss": 0.0012,
      "reward": 0.5535714626312256,
      "reward_std": 0.3471984565258026,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 195
    },
    {
      "completion_length": 696.7142944335938,
      "epoch": 0.417910447761194,
      "grad_norm": 0.1626308560371399,
      "grpo_loss": 0.047444313764572144,
      "kl": 0.01507568359375,
      "learning_rate": 2.168019463214266e-06,
      "loss": 0.0006,
      "reward": 0.4017857313156128,
      "reward_std": 0.3294869661331177,
      "rewards/accuracy_reward": 0.4017857313156128,
      "rewards/format_reward": 0.0,
      "step": 196
    },
    {
      "completion_length": 620.294677734375,
      "epoch": 0.4200426439232409,
      "grad_norm": 0.17187148332595825,
      "grpo_loss": 0.0013135150074958801,
      "kl": 0.020751953125,
      "learning_rate": 2.1580027576067387e-06,
      "loss": 0.0008,
      "reward": 0.4821428656578064,
      "reward_std": 0.3264864683151245,
      "rewards/accuracy_reward": 0.4821428656578064,
      "rewards/format_reward": 0.0,
      "step": 197
    },
    {
      "completion_length": 561.7767944335938,
      "epoch": 0.42217484008528783,
      "grad_norm": 0.1164230927824974,
      "grpo_loss": 0.0005880985409021378,
      "kl": 0.010986328125,
      "learning_rate": 2.1479495849560225e-06,
      "loss": 0.0004,
      "reward": 0.6428571939468384,
      "reward_std": 0.1779654175043106,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 198
    },
    {
      "completion_length": 601.8125,
      "epoch": 0.42430703624733473,
      "grad_norm": 0.27141571044921875,
      "grpo_loss": 0.10048054158687592,
      "kl": 0.019287109375,
      "learning_rate": 2.1378605024170477e-06,
      "loss": 0.0008,
      "reward": 0.6517857313156128,
      "reward_std": 0.2862944006919861,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 199
    },
    {
      "completion_length": 759.8839721679688,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.13523803651332855,
      "grpo_loss": 0.001226574881002307,
      "kl": 0.01348876953125,
      "learning_rate": 2.127736069134901e-06,
      "loss": 0.0005,
      "reward": 0.4375000298023224,
      "reward_std": 0.3460562527179718,
      "rewards/accuracy_reward": 0.4375000298023224,
      "rewards/format_reward": 0.0,
      "step": 200
    },
    {
      "epoch": 0.42643923240938164,
      "eval_completion_length": 611.8816388894955,
      "eval_grpo_loss": -0.0015939308651721107,
      "eval_kl": 0.022449334970297524,
      "eval_loss": 0.0009027738706208766,
      "eval_reward": 0.5542560714883165,
      "eval_reward_std": 0.28316991823835497,
      "eval_rewards/accuracy_reward": 0.5542560714883165,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 4037.199,
      "eval_samples_per_second": 1.238,
      "eval_steps_per_second": 0.011,
      "step": 200
    },
    {
      "completion_length": 531.4375,
      "epoch": 0.42857142857142855,
      "grad_norm": 0.21513928472995758,
      "grpo_loss": 0.11847112327814102,
      "kl": 0.015869140625,
      "learning_rate": 2.117576846213835e-06,
      "loss": 0.0006,
      "reward": 0.723214328289032,
      "reward_std": 0.3363610804080963,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 201
    },
    {
      "completion_length": 666.0892944335938,
      "epoch": 0.43070362473347545,
      "grad_norm": 0.18399223685264587,
      "grpo_loss": 0.00018279292271472514,
      "kl": 0.02001953125,
      "learning_rate": 2.107383396686173e-06,
      "loss": 0.0008,
      "reward": 0.5,
      "reward_std": 0.23499584197998047,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 202
    },
    {
      "completion_length": 628.7767944335938,
      "epoch": 0.43283582089552236,
      "grad_norm": 0.18898162245750427,
      "grpo_loss": -0.0157584547996521,
      "kl": 0.0218505859375,
      "learning_rate": 2.0971562854811047e-06,
      "loss": 0.0009,
      "reward": 0.4196428656578064,
      "reward_std": 0.265492707490921,
      "rewards/accuracy_reward": 0.4196428656578064,
      "rewards/format_reward": 0.0,
      "step": 203
    },
    {
      "completion_length": 642.169677734375,
      "epoch": 0.4349680170575693,
      "grad_norm": 0.14317277073860168,
      "grpo_loss": -0.09992219507694244,
      "kl": 0.01165771484375,
      "learning_rate": 2.0868960793933745e-06,
      "loss": 0.0005,
      "reward": 0.5446428656578064,
      "reward_std": 0.34032437205314636,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 204
    },
    {
      "completion_length": 529.7589721679688,
      "epoch": 0.43710021321961623,
      "grad_norm": 0.1738615781068802,
      "grpo_loss": 0.00030606670770794153,
      "kl": 0.01385498046875,
      "learning_rate": 2.076603347051875e-06,
      "loss": 0.0006,
      "reward": 0.8214285969734192,
      "reward_std": 0.25279706716537476,
      "rewards/accuracy_reward": 0.8214285969734192,
      "rewards/format_reward": 0.0,
      "step": 205
    },
    {
      "completion_length": 556.75,
      "epoch": 0.43923240938166314,
      "grad_norm": 0.19676941633224487,
      "grpo_loss": -0.047070156782865524,
      "kl": 0.015869140625,
      "learning_rate": 2.0662786588881275e-06,
      "loss": 0.0006,
      "reward": 0.6517857313156128,
      "reward_std": 0.27641984820365906,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 206
    },
    {
      "completion_length": 650.7232666015625,
      "epoch": 0.44136460554371004,
      "grad_norm": 0.1258775293827057,
      "grpo_loss": 6.173212022986263e-05,
      "kl": 0.00579833984375,
      "learning_rate": 2.0559225871046738e-06,
      "loss": 0.0002,
      "reward": 0.6160714626312256,
      "reward_std": 0.1986774057149887,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 207
    },
    {
      "completion_length": 639.232177734375,
      "epoch": 0.44349680170575695,
      "grad_norm": 0.18094466626644135,
      "grpo_loss": -0.046934500336647034,
      "kl": 0.014892578125,
      "learning_rate": 2.045535705643358e-06,
      "loss": 0.0006,
      "reward": 0.625,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 208
    },
    {
      "completion_length": 609.0892944335938,
      "epoch": 0.44562899786780386,
      "grad_norm": 0.17628911137580872,
      "grpo_loss": -0.04628950357437134,
      "kl": 0.0152587890625,
      "learning_rate": 2.0351185901535227e-06,
      "loss": 0.0006,
      "reward": 0.6160714626312256,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 209
    },
    {
      "completion_length": 580.0982666015625,
      "epoch": 0.44776119402985076,
      "grad_norm": 0.17487889528274536,
      "grpo_loss": -0.09924271702766418,
      "kl": 0.0172119140625,
      "learning_rate": 2.0246718179601055e-06,
      "loss": 0.0007,
      "reward": 0.723214328289032,
      "reward_std": 0.21137306094169617,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 210
    },
    {
      "completion_length": 667.8214721679688,
      "epoch": 0.44989339019189767,
      "grad_norm": 0.11795714497566223,
      "grpo_loss": 0.00025861337780952454,
      "kl": 0.01043701171875,
      "learning_rate": 2.014195968031641e-06,
      "loss": 0.0004,
      "reward": 0.5178571939468384,
      "reward_std": 0.2902577519416809,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 211
    },
    {
      "completion_length": 549.5714721679688,
      "epoch": 0.4520255863539446,
      "grad_norm": 0.15185706317424774,
      "grpo_loss": 0.11831355839967728,
      "kl": 0.00970458984375,
      "learning_rate": 2.003691620948176e-06,
      "loss": 0.0004,
      "reward": 0.7678571939468384,
      "reward_std": 0.1791076362133026,
      "rewards/accuracy_reward": 0.7678571939468384,
      "rewards/format_reward": 0.0,
      "step": 212
    },
    {
      "completion_length": 685.2142944335938,
      "epoch": 0.4541577825159915,
      "grad_norm": 0.2295156866312027,
      "grpo_loss": 0.04919052869081497,
      "kl": 0.037353515625,
      "learning_rate": 1.993159358869093e-06,
      "loss": 0.0015,
      "reward": 0.5535714626312256,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 213
    },
    {
      "completion_length": 569.1160888671875,
      "epoch": 0.4562899786780384,
      "grad_norm": 0.17347709834575653,
      "grpo_loss": 0.01694546267390251,
      "kl": 0.0196533203125,
      "learning_rate": 1.9825997655008457e-06,
      "loss": 0.0008,
      "reward": 0.6875000596046448,
      "reward_std": 0.24874402582645416,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 214
    },
    {
      "completion_length": 482.8035888671875,
      "epoch": 0.4584221748400853,
      "grad_norm": 0.2306608110666275,
      "grpo_loss": 0.048458606004714966,
      "kl": 0.015869140625,
      "learning_rate": 1.9720134260646093e-06,
      "loss": 0.0006,
      "reward": 0.6160714626312256,
      "reward_std": 0.24883373081684113,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 215
    },
    {
      "completion_length": 658.9107666015625,
      "epoch": 0.4605543710021322,
      "grad_norm": 0.1076991856098175,
      "grpo_loss": 0.0008707195520401001,
      "kl": 0.01287841796875,
      "learning_rate": 1.9614009272638483e-06,
      "loss": 0.0005,
      "reward": 0.5089285969734192,
      "reward_std": 0.16818054020404816,
      "rewards/accuracy_reward": 0.5089285969734192,
      "rewards/format_reward": 0.0,
      "step": 216
    },
    {
      "completion_length": 619.625,
      "epoch": 0.4626865671641791,
      "grad_norm": 0.16215959191322327,
      "grpo_loss": 0.000715792179107666,
      "kl": 0.01007080078125,
      "learning_rate": 1.9507628572518003e-06,
      "loss": 0.0004,
      "reward": 0.5178571939468384,
      "reward_std": 0.27350908517837524,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 217
    },
    {
      "completion_length": 536.8125,
      "epoch": 0.464818763326226,
      "grad_norm": 0.2339649200439453,
      "grpo_loss": 0.0006662532687187195,
      "kl": 0.025634765625,
      "learning_rate": 1.9400998055988797e-06,
      "loss": 0.001,
      "reward": 0.7678571939468384,
      "reward_std": 0.29713183641433716,
      "rewards/accuracy_reward": 0.7678571939468384,
      "rewards/format_reward": 0.0,
      "step": 218
    },
    {
      "completion_length": 617.169677734375,
      "epoch": 0.4669509594882729,
      "grad_norm": 0.16902314126491547,
      "grpo_loss": 0.07331579923629761,
      "kl": 0.01611328125,
      "learning_rate": 1.9294123632600037e-06,
      "loss": 0.0006,
      "reward": 0.6517857313156128,
      "reward_std": 0.31088003516197205,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 219
    },
    {
      "completion_length": 625.9642944335938,
      "epoch": 0.4690831556503198,
      "grad_norm": 0.21332789957523346,
      "grpo_loss": -0.046973273158073425,
      "kl": 0.0225830078125,
      "learning_rate": 1.9187011225418415e-06,
      "loss": 0.0009,
      "reward": 0.6071428656578064,
      "reward_std": 0.2862047255039215,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 220
    },
    {
      "completion_length": 558.9732666015625,
      "epoch": 0.47121535181236673,
      "grad_norm": 0.205609530210495,
      "grpo_loss": -0.046633653342723846,
      "kl": 0.01953125,
      "learning_rate": 1.907966677069986e-06,
      "loss": 0.0008,
      "reward": 0.6964285969734192,
      "reward_std": 0.24883373081684113,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 221
    },
    {
      "completion_length": 585.7589721679688,
      "epoch": 0.47334754797441364,
      "grad_norm": 0.21356123685836792,
      "grpo_loss": -0.07238468527793884,
      "kl": 0.031494140625,
      "learning_rate": 1.8972096217560587e-06,
      "loss": 0.0013,
      "reward": 0.6428571939468384,
      "reward_std": 0.27059829235076904,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 222
    },
    {
      "completion_length": 660.482177734375,
      "epoch": 0.47547974413646055,
      "grad_norm": 0.13936030864715576,
      "grpo_loss": -0.04697471857070923,
      "kl": 0.0184326171875,
      "learning_rate": 1.886430552764735e-06,
      "loss": 0.0007,
      "reward": 0.5892857313156128,
      "reward_std": 0.1986774057149887,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 223
    },
    {
      "completion_length": 591.4375,
      "epoch": 0.47761194029850745,
      "grad_norm": 0.13793712854385376,
      "grpo_loss": 0.01739496923983097,
      "kl": 0.01116943359375,
      "learning_rate": 1.8756300674807064e-06,
      "loss": 0.0004,
      "reward": 0.6785714626312256,
      "reward_std": 0.20158818364143372,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 224
    },
    {
      "completion_length": 647.232177734375,
      "epoch": 0.47974413646055436,
      "grad_norm": 0.12633898854255676,
      "grpo_loss": 0.00012505054473876953,
      "kl": 0.005950927734375,
      "learning_rate": 1.8648087644755727e-06,
      "loss": 0.0002,
      "reward": 0.6696428656578064,
      "reward_std": 0.24874402582645416,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 225
    },
    {
      "completion_length": 603.3392944335938,
      "epoch": 0.48187633262260127,
      "grad_norm": 0.25130102038383484,
      "grpo_loss": 0.00014068186283111572,
      "kl": 0.01904296875,
      "learning_rate": 1.8539672434746695e-06,
      "loss": 0.0008,
      "reward": 0.5625,
      "reward_std": 0.3098274767398834,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 226
    },
    {
      "completion_length": 631.232177734375,
      "epoch": 0.4840085287846482,
      "grad_norm": 0.18931794166564941,
      "grpo_loss": 0.00030638277530670166,
      "kl": 0.025390625,
      "learning_rate": 1.8431061053238275e-06,
      "loss": 0.001,
      "reward": 0.6428571939468384,
      "reward_std": 0.3098274767398834,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 227
    },
    {
      "completion_length": 606.982177734375,
      "epoch": 0.4861407249466951,
      "grad_norm": 0.13589723408222198,
      "grpo_loss": -0.04714548587799072,
      "kl": 0.00933837890625,
      "learning_rate": 1.832225951956079e-06,
      "loss": 0.0004,
      "reward": 0.785714328289032,
      "reward_std": 0.20564121007919312,
      "rewards/accuracy_reward": 0.785714328289032,
      "rewards/format_reward": 0.0,
      "step": 228
    },
    {
      "completion_length": 618.125,
      "epoch": 0.488272921108742,
      "grad_norm": 0.16579659283161163,
      "grpo_loss": 0.017814993858337402,
      "kl": 0.0208740234375,
      "learning_rate": 1.8213273863582918e-06,
      "loss": 0.0008,
      "reward": 0.5714285969734192,
      "reward_std": 0.23208506405353546,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 229
    },
    {
      "completion_length": 622.4910888671875,
      "epoch": 0.4904051172707889,
      "grad_norm": 0.1541467159986496,
      "grpo_loss": 0.00014001131057739258,
      "kl": 0.01806640625,
      "learning_rate": 1.8104110125377569e-06,
      "loss": 0.0007,
      "reward": 0.6428571939468384,
      "reward_std": 0.2223002016544342,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 230
    },
    {
      "completion_length": 669.0267944335938,
      "epoch": 0.4925373134328358,
      "grad_norm": 0.3476913571357727,
      "grpo_loss": 0.04737507551908493,
      "kl": 0.031494140625,
      "learning_rate": 1.7994774354887095e-06,
      "loss": 0.0013,
      "reward": 0.660714328289032,
      "reward_std": 0.20960453152656555,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 231
    },
    {
      "completion_length": 584.1607666015625,
      "epoch": 0.4946695095948827,
      "grad_norm": 0.103955939412117,
      "grpo_loss": 0.0015392377972602844,
      "kl": 0.0244140625,
      "learning_rate": 1.7885272611588032e-06,
      "loss": 0.001,
      "reward": 0.6964285969734192,
      "reward_std": 0.16818051040172577,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 232
    },
    {
      "completion_length": 626.6964721679688,
      "epoch": 0.4968017057569296,
      "grad_norm": 0.12988092005252838,
      "grpo_loss": 0.0003137141466140747,
      "kl": 0.007598876953125,
      "learning_rate": 1.7775610964155246e-06,
      "loss": 0.0003,
      "reward": 0.7142857313156128,
      "reward_std": 0.2862047255039215,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 233
    },
    {
      "completion_length": 644.7857666015625,
      "epoch": 0.4989339019189765,
      "grad_norm": 0.1723473221063614,
      "grpo_loss": 0.00012041255831718445,
      "kl": 0.018310546875,
      "learning_rate": 1.7665795490125628e-06,
      "loss": 0.0007,
      "reward": 0.4910714626312256,
      "reward_std": 0.2459229677915573,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 234
    },
    {
      "completion_length": 565.25,
      "epoch": 0.5010660980810234,
      "grad_norm": 0.19746161997318268,
      "grpo_loss": -0.046281278133392334,
      "kl": 0.0291748046875,
      "learning_rate": 1.7555832275561267e-06,
      "loss": 0.0012,
      "reward": 0.6428571939468384,
      "reward_std": 0.31670159101486206,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 235
    },
    {
      "completion_length": 683.0535888671875,
      "epoch": 0.5031982942430704,
      "grad_norm": 0.10217569023370743,
      "grpo_loss": 0.00012935325503349304,
      "kl": 0.00860595703125,
      "learning_rate": 1.7445727414712143e-06,
      "loss": 0.0003,
      "reward": 0.4196428656578064,
      "reward_std": 0.21251529455184937,
      "rewards/accuracy_reward": 0.4196428656578064,
      "rewards/format_reward": 0.0,
      "step": 236
    },
    {
      "completion_length": 677.669677734375,
      "epoch": 0.5053304904051172,
      "grad_norm": 0.1365184187889099,
      "grpo_loss": -0.04525519907474518,
      "kl": 0.0155029296875,
      "learning_rate": 1.7335487009678392e-06,
      "loss": 0.0006,
      "reward": 0.535714328289032,
      "reward_std": 0.2625819146633148,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 237
    },
    {
      "completion_length": 601.4732666015625,
      "epoch": 0.5074626865671642,
      "grad_norm": 0.20106498897075653,
      "grpo_loss": 7.900974014773965e-05,
      "kl": 0.0167236328125,
      "learning_rate": 1.7225117170072113e-06,
      "loss": 0.0007,
      "reward": 0.625,
      "reward_std": 0.222300186753273,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 238
    },
    {
      "completion_length": 712.4375610351562,
      "epoch": 0.509594882729211,
      "grad_norm": 0.17602279782295227,
      "grpo_loss": -0.046808093786239624,
      "kl": 0.0203857421875,
      "learning_rate": 1.7114624012678788e-06,
      "loss": 0.0008,
      "reward": 0.6160714626312256,
      "reward_std": 0.2862046957015991,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 239
    },
    {
      "completion_length": 604.25,
      "epoch": 0.511727078891258,
      "grad_norm": 0.15483473241329193,
      "grpo_loss": 0.05522661283612251,
      "kl": 0.018798828125,
      "learning_rate": 1.7004013661118244e-06,
      "loss": 0.0008,
      "reward": 0.598214328289032,
      "reward_std": 0.25279706716537476,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 240
    },
    {
      "completion_length": 617.6339721679688,
      "epoch": 0.5138592750533049,
      "grad_norm": 0.1640341877937317,
      "grpo_loss": 0.0003432449884712696,
      "kl": 0.01220703125,
      "learning_rate": 1.689329224550533e-06,
      "loss": 0.0005,
      "reward": 0.598214328289032,
      "reward_std": 0.19471408426761627,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 241
    },
    {
      "completion_length": 643.857177734375,
      "epoch": 0.5159914712153518,
      "grad_norm": 0.16077256202697754,
      "grpo_loss": -0.04704210162162781,
      "kl": 0.0341796875,
      "learning_rate": 1.6782465902110138e-06,
      "loss": 0.0014,
      "reward": 0.4821428656578064,
      "reward_std": 0.21251529455184937,
      "rewards/accuracy_reward": 0.4821428656578064,
      "rewards/format_reward": 0.0,
      "step": 242
    },
    {
      "completion_length": 558.3125,
      "epoch": 0.5181236673773987,
      "grad_norm": 0.22869129478931427,
      "grpo_loss": -0.09804785996675491,
      "kl": 0.0262451171875,
      "learning_rate": 1.6671540773017952e-06,
      "loss": 0.0011,
      "reward": 0.7410714626312256,
      "reward_std": 0.25570783019065857,
      "rewards/accuracy_reward": 0.7410714626312256,
      "rewards/format_reward": 0.0,
      "step": 243
    },
    {
      "completion_length": 624.982177734375,
      "epoch": 0.5202558635394456,
      "grad_norm": 0.1790386587381363,
      "grpo_loss": 0.11985684931278229,
      "kl": 0.02099609375,
      "learning_rate": 1.6560523005788832e-06,
      "loss": 0.0008,
      "reward": 0.625,
      "reward_std": 0.21542608737945557,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 244
    },
    {
      "completion_length": 654.6517944335938,
      "epoch": 0.5223880597014925,
      "grad_norm": 0.16443993151187897,
      "grpo_loss": -0.04706743732094765,
      "kl": 0.0181884765625,
      "learning_rate": 1.6449418753116917e-06,
      "loss": 0.0007,
      "reward": 0.598214328289032,
      "reward_std": 0.35698333382606506,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 245
    },
    {
      "completion_length": 542.6785888671875,
      "epoch": 0.5245202558635395,
      "grad_norm": 0.1444680094718933,
      "grpo_loss": -0.016090432181954384,
      "kl": 0.012939453125,
      "learning_rate": 1.6338234172489443e-06,
      "loss": 0.0005,
      "reward": 0.660714328289032,
      "reward_std": 0.1583956480026245,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 246
    },
    {
      "completion_length": 595.5625,
      "epoch": 0.5266524520255863,
      "grad_norm": 0.26646825671195984,
      "grpo_loss": 0.000985555350780487,
      "kl": 0.019775390625,
      "learning_rate": 1.6226975425845488e-06,
      "loss": 0.0008,
      "reward": 0.5803571939468384,
      "reward_std": 0.2695457637310028,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 247
    },
    {
      "completion_length": 574.3928833007812,
      "epoch": 0.5287846481876333,
      "grad_norm": 0.19544407725334167,
      "grpo_loss": -0.04526256024837494,
      "kl": 0.026611328125,
      "learning_rate": 1.6115648679234468e-06,
      "loss": 0.0011,
      "reward": 0.6428571939468384,
      "reward_std": 0.2666349411010742,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 248
    },
    {
      "completion_length": 597.419677734375,
      "epoch": 0.5309168443496801,
      "grad_norm": 0.19683720171451569,
      "grpo_loss": -0.07291500270366669,
      "kl": 0.0235595703125,
      "learning_rate": 1.6004260102474408e-06,
      "loss": 0.0009,
      "reward": 0.660714328289032,
      "reward_std": 0.2774723768234253,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 249
    },
    {
      "completion_length": 636.5535888671875,
      "epoch": 0.5330490405117271,
      "grad_norm": 0.20401231944561005,
      "grpo_loss": -0.11792036145925522,
      "kl": 0.0220947265625,
      "learning_rate": 1.5892815868810014e-06,
      "loss": 0.0009,
      "reward": 0.5892857313156128,
      "reward_std": 0.32075461745262146,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 250
    },
    {
      "completion_length": 670.4017944335938,
      "epoch": 0.535181236673774,
      "grad_norm": 0.24213992059230804,
      "grpo_loss": 0.00031362520530819893,
      "kl": 0.01458740234375,
      "learning_rate": 1.5781322154570548e-06,
      "loss": 0.0006,
      "reward": 0.535714328289032,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 251
    },
    {
      "completion_length": 579.4732666015625,
      "epoch": 0.5373134328358209,
      "grad_norm": 0.1810131072998047,
      "grpo_loss": 0.0024060215801000595,
      "kl": 0.02685546875,
      "learning_rate": 1.5669785138827513e-06,
      "loss": 0.0011,
      "reward": 0.598214328289032,
      "reward_std": 0.24883373081684113,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 252
    },
    {
      "completion_length": 581.419677734375,
      "epoch": 0.5394456289978679,
      "grad_norm": 0.5175444483757019,
      "grpo_loss": 0.00015514915867242962,
      "kl": 0.01904296875,
      "learning_rate": 1.5558211003052223e-06,
      "loss": 0.0008,
      "reward": 0.7142857313156128,
      "reward_std": 0.16526976227760315,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 253
    },
    {
      "completion_length": 624.4553833007812,
      "epoch": 0.5415778251599147,
      "grad_norm": 0.21199892461299896,
      "grpo_loss": -0.047045424580574036,
      "kl": 0.0361328125,
      "learning_rate": 1.5446605930773215e-06,
      "loss": 0.0014,
      "reward": 0.6428571939468384,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 254
    },
    {
      "completion_length": 590.6607666015625,
      "epoch": 0.5437100213219617,
      "grad_norm": 0.24763309955596924,
      "grpo_loss": 0.0017136521637439728,
      "kl": 0.032958984375,
      "learning_rate": 1.5334976107233556e-06,
      "loss": 0.0013,
      "reward": 0.6785714626312256,
      "reward_std": 0.29598960280418396,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 255
    },
    {
      "completion_length": 547.5535888671875,
      "epoch": 0.5458422174840085,
      "grad_norm": 0.2570716440677643,
      "grpo_loss": -0.046061620116233826,
      "kl": 0.052490234375,
      "learning_rate": 1.5223327719048027e-06,
      "loss": 0.0021,
      "reward": 0.7053571939468384,
      "reward_std": 0.40317636728286743,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 256
    },
    {
      "completion_length": 700.4910888671875,
      "epoch": 0.5479744136460555,
      "grad_norm": 0.1590230017900467,
      "grpo_loss": 0.0008169859647750854,
      "kl": 0.042724609375,
      "learning_rate": 1.511166695386031e-06,
      "loss": 0.0017,
      "reward": 0.4107142984867096,
      "reward_std": 0.2891155183315277,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 0.0,
      "step": 257
    },
    {
      "completion_length": 563.8035888671875,
      "epoch": 0.5501066098081023,
      "grad_norm": 0.22808882594108582,
      "grpo_loss": -0.1334923654794693,
      "kl": 0.047119140625,
      "learning_rate": 1.5e-06,
      "loss": 0.0019,
      "reward": 0.7142857313156128,
      "reward_std": 0.3363610506057739,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 258
    },
    {
      "completion_length": 588.4464721679688,
      "epoch": 0.5522388059701493,
      "grad_norm": 0.1522502601146698,
      "grpo_loss": -0.053403325378894806,
      "kl": 0.0177001953125,
      "learning_rate": 1.4888333046139694e-06,
      "loss": 0.0007,
      "reward": 0.535714328289032,
      "reward_std": 0.219389408826828,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 259
    },
    {
      "completion_length": 598.5714721679688,
      "epoch": 0.5543710021321961,
      "grad_norm": 0.14512988924980164,
      "grpo_loss": 0.0016735941171646118,
      "kl": 0.02001953125,
      "learning_rate": 1.477667228095197e-06,
      "loss": 0.0008,
      "reward": 0.6071428656578064,
      "reward_std": 0.24592295289039612,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 260
    },
    {
      "completion_length": 675.044677734375,
      "epoch": 0.5565031982942431,
      "grad_norm": 0.23608967661857605,
      "grpo_loss": 0.057517096400260925,
      "kl": 0.052001953125,
      "learning_rate": 1.4665023892766447e-06,
      "loss": 0.0021,
      "reward": 0.4375000298023224,
      "reward_std": 0.3206648826599121,
      "rewards/accuracy_reward": 0.4375000298023224,
      "rewards/format_reward": 0.0,
      "step": 261
    },
    {
      "completion_length": 610.3839721679688,
      "epoch": 0.55863539445629,
      "grad_norm": 0.17539887130260468,
      "grpo_loss": -0.06992875784635544,
      "kl": 0.03564453125,
      "learning_rate": 1.4553394069226786e-06,
      "loss": 0.0014,
      "reward": 0.6785714626312256,
      "reward_std": 0.24478070437908173,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 262
    },
    {
      "completion_length": 561.732177734375,
      "epoch": 0.5607675906183369,
      "grad_norm": 0.18649928271770477,
      "grpo_loss": 0.13406673073768616,
      "kl": 0.021728515625,
      "learning_rate": 1.444178899694778e-06,
      "loss": 0.0009,
      "reward": 0.7500000596046448,
      "reward_std": 0.25279706716537476,
      "rewards/accuracy_reward": 0.7500000596046448,
      "rewards/format_reward": 0.0,
      "step": 263
    },
    {
      "completion_length": 585.482177734375,
      "epoch": 0.5628997867803838,
      "grad_norm": 0.1891089528799057,
      "grpo_loss": -0.1000761166214943,
      "kl": 0.03466796875,
      "learning_rate": 1.4330214861172487e-06,
      "loss": 0.0014,
      "reward": 0.7589285969734192,
      "reward_std": 0.29899007081985474,
      "rewards/accuracy_reward": 0.7589285969734192,
      "rewards/format_reward": 0.0,
      "step": 264
    },
    {
      "completion_length": 666.8660888671875,
      "epoch": 0.5650319829424307,
      "grad_norm": 0.37016505002975464,
      "grpo_loss": -0.014646738767623901,
      "kl": 0.047607421875,
      "learning_rate": 1.4218677845429455e-06,
      "loss": 0.0019,
      "reward": 0.5178571939468384,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 265
    },
    {
      "completion_length": 611.1339721679688,
      "epoch": 0.5671641791044776,
      "grad_norm": 0.15229058265686035,
      "grpo_loss": 0.05506877228617668,
      "kl": 0.01202392578125,
      "learning_rate": 1.410718413118999e-06,
      "loss": 0.0005,
      "reward": 0.660714328289032,
      "reward_std": 0.2724565267562866,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 266
    },
    {
      "completion_length": 507.8035888671875,
      "epoch": 0.5692963752665245,
      "grad_norm": 0.20434416830539703,
      "grpo_loss": 0.0012201443314552307,
      "kl": 0.0206298828125,
      "learning_rate": 1.3995739897525592e-06,
      "loss": 0.0008,
      "reward": 0.8392857313156128,
      "reward_std": 0.27059829235076904,
      "rewards/accuracy_reward": 0.8392857313156128,
      "rewards/format_reward": 0.0,
      "step": 267
    },
    {
      "completion_length": 576.5267944335938,
      "epoch": 0.5714285714285714,
      "grad_norm": 0.19015087187290192,
      "grpo_loss": -0.047088075429201126,
      "kl": 0.0302734375,
      "learning_rate": 1.3884351320765534e-06,
      "loss": 0.0012,
      "reward": 0.7053571939468384,
      "reward_std": 0.2803831696510315,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 268
    },
    {
      "completion_length": 554.9107666015625,
      "epoch": 0.5735607675906184,
      "grad_norm": 0.19130241870880127,
      "grpo_loss": 0.0005413442850112915,
      "kl": 0.0400390625,
      "learning_rate": 1.3773024574154515e-06,
      "loss": 0.0016,
      "reward": 0.5892857313156128,
      "reward_std": 0.27236682176589966,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 269
    },
    {
      "completion_length": 586.3928833007812,
      "epoch": 0.5756929637526652,
      "grad_norm": 0.1519208699464798,
      "grpo_loss": -0.04699365422129631,
      "kl": 0.0186767578125,
      "learning_rate": 1.3661765827510562e-06,
      "loss": 0.0007,
      "reward": 0.535714328289032,
      "reward_std": 0.219389408826828,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 270
    },
    {
      "completion_length": 733.2410888671875,
      "epoch": 0.5778251599147122,
      "grad_norm": 0.12186726182699203,
      "grpo_loss": 0.00036180391907691956,
      "kl": 0.0245361328125,
      "learning_rate": 1.3550581246883084e-06,
      "loss": 0.001,
      "reward": 0.3660714328289032,
      "reward_std": 0.2124256044626236,
      "rewards/accuracy_reward": 0.3660714328289032,
      "rewards/format_reward": 0.0,
      "step": 271
    },
    {
      "completion_length": 600.6428833007812,
      "epoch": 0.579957356076759,
      "grad_norm": 0.15703478455543518,
      "grpo_loss": 0.00036340951919555664,
      "kl": 0.033203125,
      "learning_rate": 1.3439476994211171e-06,
      "loss": 0.0013,
      "reward": 0.598214328289032,
      "reward_std": 0.2459229677915573,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 272
    },
    {
      "completion_length": 608.2232666015625,
      "epoch": 0.582089552238806,
      "grad_norm": 0.17682023346424103,
      "grpo_loss": 0.017652347683906555,
      "kl": 0.026123046875,
      "learning_rate": 1.3328459226982051e-06,
      "loss": 0.001,
      "reward": 0.6696428656578064,
      "reward_std": 0.36280491948127747,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 273
    },
    {
      "completion_length": 669.1875,
      "epoch": 0.5842217484008528,
      "grad_norm": 0.13847747445106506,
      "grpo_loss": 0.00015268649440258741,
      "kl": 0.0142822265625,
      "learning_rate": 1.3217534097889865e-06,
      "loss": 0.0006,
      "reward": 0.5446428656578064,
      "reward_std": 0.31670159101486206,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 274
    },
    {
      "completion_length": 672.4285888671875,
      "epoch": 0.5863539445628998,
      "grad_norm": 0.15076103806495667,
      "grpo_loss": 0.0006213184678927064,
      "kl": 0.0167236328125,
      "learning_rate": 1.310670775449467e-06,
      "loss": 0.0007,
      "reward": 0.535714328289032,
      "reward_std": 0.31961238384246826,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 275
    },
    {
      "completion_length": 762.8839721679688,
      "epoch": 0.5884861407249466,
      "grad_norm": 0.12425374239683151,
      "grpo_loss": 0.0010431259870529175,
      "kl": 0.014892578125,
      "learning_rate": 1.2995986338881757e-06,
      "loss": 0.0006,
      "reward": 0.4910714626312256,
      "reward_std": 0.2430122047662735,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 276
    },
    {
      "completion_length": 478.9285888671875,
      "epoch": 0.5906183368869936,
      "grad_norm": 0.15039986371994019,
      "grpo_loss": -0.04706709831953049,
      "kl": 0.0186767578125,
      "learning_rate": 1.2885375987321222e-06,
      "loss": 0.0007,
      "reward": 0.8392857313156128,
      "reward_std": 0.1928558498620987,
      "rewards/accuracy_reward": 0.8392857313156128,
      "rewards/format_reward": 0.0,
      "step": 277
    },
    {
      "completion_length": 575.3660888671875,
      "epoch": 0.5927505330490405,
      "grad_norm": 0.14196723699569702,
      "grpo_loss": 0.05593723803758621,
      "kl": 0.015869140625,
      "learning_rate": 1.2774882829927886e-06,
      "loss": 0.0006,
      "reward": 0.6875000596046448,
      "reward_std": 0.20960454642772675,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 278
    },
    {
      "completion_length": 699.4910888671875,
      "epoch": 0.5948827292110874,
      "grad_norm": 0.18591664731502533,
      "grpo_loss": 0.05759921297430992,
      "kl": 0.037841796875,
      "learning_rate": 1.2664512990321611e-06,
      "loss": 0.0015,
      "reward": 0.5625,
      "reward_std": 0.35407260060310364,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 279
    },
    {
      "completion_length": 637.1785888671875,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.16850171983242035,
      "grpo_loss": 0.00017352678696624935,
      "kl": 0.020263671875,
      "learning_rate": 1.2554272585287862e-06,
      "loss": 0.0008,
      "reward": 0.598214328289032,
      "reward_std": 0.23208506405353546,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 280
    },
    {
      "completion_length": 593.5,
      "epoch": 0.5991471215351812,
      "grad_norm": 0.1354869157075882,
      "grpo_loss": 0.0011160299181938171,
      "kl": 0.0133056640625,
      "learning_rate": 1.244416772443874e-06,
      "loss": 0.0005,
      "reward": 0.5178571939468384,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 281
    },
    {
      "completion_length": 542.4910888671875,
      "epoch": 0.6012793176972282,
      "grad_norm": 0.19948160648345947,
      "grpo_loss": 0.0181841142475605,
      "kl": 0.038818359375,
      "learning_rate": 1.2334204509874373e-06,
      "loss": 0.0016,
      "reward": 0.7142857313156128,
      "reward_std": 0.305864155292511,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 282
    },
    {
      "completion_length": 567.794677734375,
      "epoch": 0.603411513859275,
      "grad_norm": 0.14735370874404907,
      "grpo_loss": 0.0002326183021068573,
      "kl": 0.018798828125,
      "learning_rate": 1.2224389035844757e-06,
      "loss": 0.0008,
      "reward": 0.6517857313156128,
      "reward_std": 0.22812172770500183,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 283
    },
    {
      "completion_length": 590.7678833007812,
      "epoch": 0.605543710021322,
      "grad_norm": 0.18942324817180634,
      "grpo_loss": -0.10003073513507843,
      "kl": 0.0299072265625,
      "learning_rate": 1.2114727388411973e-06,
      "loss": 0.0012,
      "reward": 0.598214328289032,
      "reward_std": 0.3235756754875183,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 284
    },
    {
      "completion_length": 562.1607666015625,
      "epoch": 0.6076759061833689,
      "grad_norm": 0.23257029056549072,
      "grpo_loss": -0.05442488566040993,
      "kl": 0.050537109375,
      "learning_rate": 1.2005225645112906e-06,
      "loss": 0.002,
      "reward": 0.5178571939468384,
      "reward_std": 0.26654526591300964,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 285
    },
    {
      "completion_length": 545.0089721679688,
      "epoch": 0.6098081023454158,
      "grad_norm": 0.16995713114738464,
      "grpo_loss": 0.0012850586790591478,
      "kl": 0.0242919921875,
      "learning_rate": 1.1895889874622432e-06,
      "loss": 0.001,
      "reward": 0.6785714626312256,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 286
    },
    {
      "completion_length": 630.0357666015625,
      "epoch": 0.6119402985074627,
      "grad_norm": 0.15201345086097717,
      "grpo_loss": 0.04734975844621658,
      "kl": 0.02099609375,
      "learning_rate": 1.1786726136417083e-06,
      "loss": 0.0008,
      "reward": 0.598214328289032,
      "reward_std": 0.20855198800563812,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 287
    },
    {
      "completion_length": 562.3928833007812,
      "epoch": 0.6140724946695096,
      "grad_norm": 0.14891578257083893,
      "grpo_loss": -0.04348957538604736,
      "kl": 0.024169921875,
      "learning_rate": 1.1677740480439218e-06,
      "loss": 0.001,
      "reward": 0.8214285969734192,
      "reward_std": 0.24697549641132355,
      "rewards/accuracy_reward": 0.8214285969734192,
      "rewards/format_reward": 0.0,
      "step": 288
    },
    {
      "completion_length": 716.5000610351562,
      "epoch": 0.6162046908315565,
      "grad_norm": 0.14191721379756927,
      "grpo_loss": 0.0008554987725801766,
      "kl": 0.030029296875,
      "learning_rate": 1.1568938946761726e-06,
      "loss": 0.0012,
      "reward": 0.4821428656578064,
      "reward_std": 0.18889252841472626,
      "rewards/accuracy_reward": 0.4821428656578064,
      "rewards/format_reward": 0.0,
      "step": 289
    },
    {
      "completion_length": 646.1875,
      "epoch": 0.6183368869936035,
      "grad_norm": 0.18612991273403168,
      "grpo_loss": -0.047120094299316406,
      "kl": 0.05126953125,
      "learning_rate": 1.1460327565253308e-06,
      "loss": 0.0021,
      "reward": 0.6785714626312256,
      "reward_std": 0.27641981840133667,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 290
    },
    {
      "completion_length": 532.75,
      "epoch": 0.6204690831556503,
      "grad_norm": 0.15706709027290344,
      "grpo_loss": -0.046887509524822235,
      "kl": 0.019287109375,
      "learning_rate": 1.1351912355244273e-06,
      "loss": 0.0008,
      "reward": 0.8125000596046448,
      "reward_std": 0.23613810539245605,
      "rewards/accuracy_reward": 0.8125000596046448,
      "rewards/format_reward": 0.0,
      "step": 291
    },
    {
      "completion_length": 528.8125,
      "epoch": 0.6226012793176973,
      "grad_norm": 0.26372042298316956,
      "grpo_loss": -0.05313573777675629,
      "kl": 0.046875,
      "learning_rate": 1.1243699325192936e-06,
      "loss": 0.0019,
      "reward": 0.6785714626312256,
      "reward_std": 0.3501092493534088,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 292
    },
    {
      "completion_length": 582.669677734375,
      "epoch": 0.6247334754797441,
      "grad_norm": 0.14917142689228058,
      "grpo_loss": 0.0013749670470133424,
      "kl": 0.0247802734375,
      "learning_rate": 1.1135694472352654e-06,
      "loss": 0.001,
      "reward": 0.6696428656578064,
      "reward_std": 0.1986774057149887,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 293
    },
    {
      "completion_length": 600.6428833007812,
      "epoch": 0.6268656716417911,
      "grad_norm": 0.24822071194648743,
      "grpo_loss": 0.0018174201250076294,
      "kl": 0.0390625,
      "learning_rate": 1.1027903782439413e-06,
      "loss": 0.0016,
      "reward": 0.6875000596046448,
      "reward_std": 0.2989003360271454,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 294
    },
    {
      "completion_length": 689.8482666015625,
      "epoch": 0.6289978678038379,
      "grad_norm": 0.13699181377887726,
      "grpo_loss": 0.0017366781830787659,
      "kl": 0.01446533203125,
      "learning_rate": 1.092033322930014e-06,
      "loss": 0.0006,
      "reward": 0.4910714626312256,
      "reward_std": 0.2223002016544342,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 295
    },
    {
      "completion_length": 597.2232666015625,
      "epoch": 0.6311300639658849,
      "grad_norm": 0.1713307648897171,
      "grpo_loss": -0.07294347882270813,
      "kl": 0.0286865234375,
      "learning_rate": 1.0812988774581588e-06,
      "loss": 0.0012,
      "reward": 0.723214328289032,
      "reward_std": 0.3264864683151245,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 296
    },
    {
      "completion_length": 509.5625305175781,
      "epoch": 0.6332622601279317,
      "grad_norm": 0.22946512699127197,
      "grpo_loss": 0.11910638958215714,
      "kl": 0.02734375,
      "learning_rate": 1.0705876367399966e-06,
      "loss": 0.0011,
      "reward": 0.6875000596046448,
      "reward_std": 0.2666349411010742,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 297
    },
    {
      "completion_length": 631.6964721679688,
      "epoch": 0.6353944562899787,
      "grad_norm": 0.29484447836875916,
      "grpo_loss": 0.05576039105653763,
      "kl": 0.07080078125,
      "learning_rate": 1.0599001944011204e-06,
      "loss": 0.0028,
      "reward": 0.6160714626312256,
      "reward_std": 0.2538496255874634,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 298
    },
    {
      "completion_length": 634.0714721679688,
      "epoch": 0.6375266524520256,
      "grad_norm": 0.5733016729354858,
      "grpo_loss": 0.0023387782275676727,
      "kl": 0.0213623046875,
      "learning_rate": 1.0492371427482002e-06,
      "loss": 0.0009,
      "reward": 0.7767857313156128,
      "reward_std": 0.16526976227760315,
      "rewards/accuracy_reward": 0.7767857313156128,
      "rewards/format_reward": 0.0,
      "step": 299
    },
    {
      "completion_length": 587.0982666015625,
      "epoch": 0.6396588486140725,
      "grad_norm": 0.15578092634677887,
      "grpo_loss": 0.0022323429584503174,
      "kl": 0.0264892578125,
      "learning_rate": 1.0385990727361518e-06,
      "loss": 0.0011,
      "reward": 0.7053571939468384,
      "reward_std": 0.3029533922672272,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 300
    },
    {
      "epoch": 0.6396588486140725,
      "eval_completion_length": 607.3742406680561,
      "eval_grpo_loss": -0.0026763521267401317,
      "eval_kl": 0.03561235616763179,
      "eval_loss": 0.0014403663808479905,
      "eval_reward": 0.5617298292466246,
      "eval_reward_std": 0.2739596153362491,
      "eval_rewards/accuracy_reward": 0.5617298292466246,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 4024.0485,
      "eval_samples_per_second": 1.243,
      "eval_steps_per_second": 0.011,
      "step": 300
    },
    {
      "completion_length": 598.4375,
      "epoch": 0.6417910447761194,
      "grad_norm": 0.13117839395999908,
      "grpo_loss": -0.04589998349547386,
      "kl": 0.018310546875,
      "learning_rate": 1.027986573935391e-06,
      "loss": 0.0007,
      "reward": 0.6428571939468384,
      "reward_std": 0.23208506405353546,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 301
    },
    {
      "completion_length": 584.4642944335938,
      "epoch": 0.6439232409381663,
      "grad_norm": 0.17518894374370575,
      "grpo_loss": 0.04966043680906296,
      "kl": 0.04052734375,
      "learning_rate": 1.0174002344991544e-06,
      "loss": 0.0016,
      "reward": 0.598214328289032,
      "reward_std": 0.14861077070236206,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 302
    },
    {
      "completion_length": 561.0892944335938,
      "epoch": 0.6460554371002132,
      "grad_norm": 0.15005171298980713,
      "grpo_loss": 0.00026489607989788055,
      "kl": 0.0213623046875,
      "learning_rate": 1.0068406411309068e-06,
      "loss": 0.0009,
      "reward": 0.6428571939468384,
      "reward_std": 0.19585634768009186,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 303
    },
    {
      "completion_length": 598.232177734375,
      "epoch": 0.6481876332622601,
      "grad_norm": 0.19444920122623444,
      "grpo_loss": 0.00011016204371117055,
      "kl": 0.03564453125,
      "learning_rate": 9.96308379051824e-07,
      "loss": 0.0014,
      "reward": 0.5267857313156128,
      "reward_std": 0.30409565567970276,
      "rewards/accuracy_reward": 0.5267857313156128,
      "rewards/format_reward": 0.0,
      "step": 304
    },
    {
      "completion_length": 541.9464721679688,
      "epoch": 0.650319829424307,
      "grad_norm": 0.19593097269535065,
      "grpo_loss": -0.07303665578365326,
      "kl": 0.040771484375,
      "learning_rate": 9.858040319683595e-07,
      "loss": 0.0016,
      "reward": 0.6339285969734192,
      "reward_std": 0.2989003658294678,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 305
    },
    {
      "completion_length": 662.419677734375,
      "epoch": 0.652452025586354,
      "grad_norm": 0.156465083360672,
      "grpo_loss": 0.0031750944908708334,
      "kl": 0.04833984375,
      "learning_rate": 9.753281820398952e-07,
      "loss": 0.0019,
      "reward": 0.660714328289032,
      "reward_std": 0.25570783019065857,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 306
    },
    {
      "completion_length": 623.2232666015625,
      "epoch": 0.6545842217484008,
      "grad_norm": 0.16462744772434235,
      "grpo_loss": 0.04866427183151245,
      "kl": 0.03271484375,
      "learning_rate": 9.648814098464774e-07,
      "loss": 0.0013,
      "reward": 0.6785714626312256,
      "reward_std": 0.29713183641433716,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 307
    },
    {
      "completion_length": 568.5535888671875,
      "epoch": 0.6567164179104478,
      "grad_norm": 0.20251184701919556,
      "grpo_loss": 0.0002517774701118469,
      "kl": 0.0267333984375,
      "learning_rate": 9.544642943566426e-07,
      "loss": 0.0011,
      "reward": 0.5446428656578064,
      "reward_std": 0.30004259943962097,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 308
    },
    {
      "completion_length": 512.169677734375,
      "epoch": 0.6588486140724946,
      "grad_norm": 0.20472615957260132,
      "grpo_loss": 0.00059002626221627,
      "kl": 0.04638671875,
      "learning_rate": 9.440774128953266e-07,
      "loss": 0.0018,
      "reward": 0.723214328289032,
      "reward_std": 0.21251530945301056,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 309
    },
    {
      "completion_length": 630.8839721679688,
      "epoch": 0.6609808102345416,
      "grad_norm": 0.15194885432720184,
      "grpo_loss": 0.00018625587108545005,
      "kl": 0.032958984375,
      "learning_rate": 9.337213411118722e-07,
      "loss": 0.0013,
      "reward": 0.6517857313156128,
      "reward_std": 0.23904886841773987,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 310
    },
    {
      "completion_length": 678.7142944335938,
      "epoch": 0.6631130063965884,
      "grad_norm": 0.14824587106704712,
      "grpo_loss": -0.014349933713674545,
      "kl": 0.0247802734375,
      "learning_rate": 9.233966529481255e-07,
      "loss": 0.001,
      "reward": 0.4910714626312256,
      "reward_std": 0.2459229677915573,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 311
    },
    {
      "completion_length": 564.0267944335938,
      "epoch": 0.6652452025586354,
      "grad_norm": 0.16684581339359283,
      "grpo_loss": 0.0020749643445014954,
      "kl": 0.0279541015625,
      "learning_rate": 9.131039206066261e-07,
      "loss": 0.0011,
      "reward": 0.7500000596046448,
      "reward_std": 0.2724565267562866,
      "rewards/accuracy_reward": 0.7500000596046448,
      "rewards/format_reward": 0.0,
      "step": 312
    },
    {
      "completion_length": 641.0089721679688,
      "epoch": 0.6673773987206824,
      "grad_norm": 0.17203587293624878,
      "grpo_loss": 0.001197556615807116,
      "kl": 0.033447265625,
      "learning_rate": 9.028437145188962e-07,
      "loss": 0.0013,
      "reward": 0.5803571939468384,
      "reward_std": 0.3391821086406708,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 313
    },
    {
      "completion_length": 550.5178833007812,
      "epoch": 0.6695095948827292,
      "grad_norm": 0.16829533874988556,
      "grpo_loss": 0.00012534562847577035,
      "kl": 0.031982421875,
      "learning_rate": 8.926166033138266e-07,
      "loss": 0.0013,
      "reward": 0.5892857313156128,
      "reward_std": 0.21251530945301056,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 314
    },
    {
      "completion_length": 680.6785888671875,
      "epoch": 0.6716417910447762,
      "grad_norm": 0.36079153418540955,
      "grpo_loss": 0.0016356073319911957,
      "kl": 0.031005859375,
      "learning_rate": 8.82423153786165e-07,
      "loss": 0.0012,
      "reward": 0.5446428656578064,
      "reward_std": 0.29016801714897156,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 315
    },
    {
      "completion_length": 610.3482666015625,
      "epoch": 0.673773987206823,
      "grad_norm": 0.1447429209947586,
      "grpo_loss": 0.001389749813824892,
      "kl": 0.033447265625,
      "learning_rate": 8.72263930865099e-07,
      "loss": 0.0013,
      "reward": 0.5714285969734192,
      "reward_std": 0.2124256193637848,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 316
    },
    {
      "completion_length": 716.5982666015625,
      "epoch": 0.67590618336887,
      "grad_norm": 0.12955766916275024,
      "grpo_loss": 0.0010962896049022675,
      "kl": 0.010009765625,
      "learning_rate": 8.62139497582953e-07,
      "loss": 0.0004,
      "reward": 0.5892857313156128,
      "reward_std": 0.29598960280418396,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 317
    },
    {
      "completion_length": 766.857177734375,
      "epoch": 0.6780383795309168,
      "grad_norm": 0.14128656685352325,
      "grpo_loss": 0.0012699184007942677,
      "kl": 0.02978515625,
      "learning_rate": 8.520504150439773e-07,
      "loss": 0.0012,
      "reward": 0.3571428656578064,
      "reward_std": 0.3028636872768402,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 0.0,
      "step": 318
    },
    {
      "completion_length": 644.1428833007812,
      "epoch": 0.6801705756929638,
      "grad_norm": 0.15934458374977112,
      "grpo_loss": 0.0012877359986305237,
      "kl": 0.04541015625,
      "learning_rate": 8.419972423932613e-07,
      "loss": 0.0018,
      "reward": 0.5178571939468384,
      "reward_std": 0.29307880997657776,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 319
    },
    {
      "completion_length": 571.5803833007812,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.1932043880224228,
      "grpo_loss": 0.00025262683629989624,
      "kl": 0.0269775390625,
      "learning_rate": 8.319805367857344e-07,
      "loss": 0.0011,
      "reward": 0.6875000596046448,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 320
    },
    {
      "completion_length": 548.4553833007812,
      "epoch": 0.6844349680170576,
      "grad_norm": 0.13966359198093414,
      "grpo_loss": 0.07430977374315262,
      "kl": 0.0159912109375,
      "learning_rate": 8.220008533552953e-07,
      "loss": 0.0006,
      "reward": 0.6696428656578064,
      "reward_std": 0.2625819444656372,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 321
    },
    {
      "completion_length": 656.044677734375,
      "epoch": 0.6865671641791045,
      "grad_norm": 0.25401002168655396,
      "grpo_loss": 0.00247298926115036,
      "kl": 0.043701171875,
      "learning_rate": 8.120587451840439e-07,
      "loss": 0.0017,
      "reward": 0.625,
      "reward_std": 0.2891154885292053,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 322
    },
    {
      "completion_length": 615.544677734375,
      "epoch": 0.6886993603411514,
      "grad_norm": 0.16577310860157013,
      "grpo_loss": -0.05358615145087242,
      "kl": 0.00927734375,
      "learning_rate": 8.021547632716291e-07,
      "loss": 0.0004,
      "reward": 0.4910714626312256,
      "reward_std": 0.29713183641433716,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 323
    },
    {
      "completion_length": 623.0982666015625,
      "epoch": 0.6908315565031983,
      "grad_norm": 0.12740829586982727,
      "grpo_loss": 0.05615002661943436,
      "kl": 0.0255126953125,
      "learning_rate": 7.9228945650471e-07,
      "loss": 0.001,
      "reward": 0.5714285969734192,
      "reward_std": 0.2389591485261917,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 324
    },
    {
      "completion_length": 638.9375,
      "epoch": 0.6929637526652452,
      "grad_norm": 0.6456794738769531,
      "grpo_loss": -0.045602887868881226,
      "kl": 0.0302734375,
      "learning_rate": 7.824633716265399e-07,
      "loss": 0.0012,
      "reward": 0.6071428656578064,
      "reward_std": 0.1820184290409088,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 325
    },
    {
      "completion_length": 610.3035888671875,
      "epoch": 0.6950959488272921,
      "grad_norm": 0.12833137810230255,
      "grpo_loss": 0.00032634875969961286,
      "kl": 0.0096435546875,
      "learning_rate": 7.72677053206663e-07,
      "loss": 0.0004,
      "reward": 0.6517857313156128,
      "reward_std": 0.2418699413537979,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 326
    },
    {
      "completion_length": 561.544677734375,
      "epoch": 0.697228144989339,
      "grad_norm": 0.19920487701892853,
      "grpo_loss": -0.04694785922765732,
      "kl": 0.0240478515625,
      "learning_rate": 7.629310436107342e-07,
      "loss": 0.001,
      "reward": 0.6785714626312256,
      "reward_std": 0.3472881615161896,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 327
    },
    {
      "completion_length": 600.1517944335938,
      "epoch": 0.6993603411513859,
      "grad_norm": 0.13510799407958984,
      "grpo_loss": 0.0018375907093286514,
      "kl": 0.0257568359375,
      "learning_rate": 7.5322588297046e-07,
      "loss": 0.001,
      "reward": 0.5446428656578064,
      "reward_std": 0.21251529455184937,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 328
    },
    {
      "completion_length": 540.9285888671875,
      "epoch": 0.7014925373134329,
      "grad_norm": 0.17828156054019928,
      "grpo_loss": 0.00013902665523346514,
      "kl": 0.02587890625,
      "learning_rate": 7.43562109153666e-07,
      "loss": 0.001,
      "reward": 0.8125000596046448,
      "reward_std": 0.14746850728988647,
      "rewards/accuracy_reward": 0.8125000596046448,
      "rewards/format_reward": 0.0,
      "step": 329
    },
    {
      "completion_length": 651.9910888671875,
      "epoch": 0.7036247334754797,
      "grad_norm": 0.1514289230108261,
      "grpo_loss": 0.05768703296780586,
      "kl": 0.028564453125,
      "learning_rate": 7.339402577344863e-07,
      "loss": 0.0011,
      "reward": 0.6160714626312256,
      "reward_std": 0.2360483705997467,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 330
    },
    {
      "completion_length": 638.544677734375,
      "epoch": 0.7057569296375267,
      "grad_norm": 0.16231295466423035,
      "grpo_loss": 0.05624303221702576,
      "kl": 0.044921875,
      "learning_rate": 7.243608619636828e-07,
      "loss": 0.0018,
      "reward": 0.6964285969734192,
      "reward_std": 0.29307880997657776,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 331
    },
    {
      "completion_length": 593.3303833007812,
      "epoch": 0.7078891257995735,
      "grad_norm": 0.1985163390636444,
      "grpo_loss": 0.001253962516784668,
      "kl": 0.03271484375,
      "learning_rate": 7.14824452739089e-07,
      "loss": 0.0013,
      "reward": 0.6428571939468384,
      "reward_std": 0.3433248698711395,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 332
    },
    {
      "completion_length": 563.4642944335938,
      "epoch": 0.7100213219616205,
      "grad_norm": 0.1781807243824005,
      "grpo_loss": -0.04701165482401848,
      "kl": 0.0274658203125,
      "learning_rate": 7.053315585761911e-07,
      "loss": 0.0011,
      "reward": 0.723214328289032,
      "reward_std": 0.2794203460216522,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 333
    },
    {
      "completion_length": 621.7142944335938,
      "epoch": 0.7121535181236673,
      "grad_norm": 0.16407182812690735,
      "grpo_loss": 0.00021706148982048035,
      "kl": 0.0255126953125,
      "learning_rate": 6.958827055788351e-07,
      "loss": 0.001,
      "reward": 0.6428571939468384,
      "reward_std": 0.3127382695674896,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 334
    },
    {
      "completion_length": 651.419677734375,
      "epoch": 0.7142857142857143,
      "grad_norm": 0.1288588047027588,
      "grpo_loss": 0.0014822334051132202,
      "kl": 0.044921875,
      "learning_rate": 6.864784174100703e-07,
      "loss": 0.0018,
      "reward": 0.5089285969734192,
      "reward_std": 0.1820184290409088,
      "rewards/accuracy_reward": 0.5089285969734192,
      "rewards/format_reward": 0.0,
      "step": 335
    },
    {
      "completion_length": 654.6964721679688,
      "epoch": 0.7164179104477612,
      "grad_norm": 0.14913132786750793,
      "grpo_loss": 0.00022607296705245972,
      "kl": 0.0203857421875,
      "learning_rate": 6.771192152631274e-07,
      "loss": 0.0008,
      "reward": 0.5714285969734192,
      "reward_std": 0.21251530945301056,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 336
    },
    {
      "completion_length": 635.4910888671875,
      "epoch": 0.7185501066098081,
      "grad_norm": 0.20662419497966766,
      "grpo_loss": -0.04704027250409126,
      "kl": 0.02880859375,
      "learning_rate": 6.678056178325319e-07,
      "loss": 0.0012,
      "reward": 0.535714328289032,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 337
    },
    {
      "completion_length": 680.0,
      "epoch": 0.720682302771855,
      "grad_norm": 0.1413746476173401,
      "grpo_loss": -0.11560186743736267,
      "kl": 0.0296630859375,
      "learning_rate": 6.585381412853612e-07,
      "loss": 0.0012,
      "reward": 0.5892857313156128,
      "reward_std": 0.3177541494369507,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 338
    },
    {
      "completion_length": 607.1339721679688,
      "epoch": 0.7228144989339019,
      "grad_norm": 0.21040762960910797,
      "grpo_loss": 0.0747302919626236,
      "kl": 0.0380859375,
      "learning_rate": 6.49317299232635e-07,
      "loss": 0.0015,
      "reward": 0.5803571939468384,
      "reward_std": 0.3954068720340729,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 339
    },
    {
      "completion_length": 647.3214721679688,
      "epoch": 0.7249466950959488,
      "grad_norm": 0.1362115442752838,
      "grpo_loss": -0.01653050072491169,
      "kl": 0.0269775390625,
      "learning_rate": 6.401436027008529e-07,
      "loss": 0.0011,
      "reward": 0.7321428656578064,
      "reward_std": 0.2684035003185272,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 340
    },
    {
      "completion_length": 732.0089721679688,
      "epoch": 0.7270788912579957,
      "grad_norm": 0.13052456080913544,
      "grpo_loss": 0.0003261137753725052,
      "kl": 0.018798828125,
      "learning_rate": 6.310175601036695e-07,
      "loss": 0.0007,
      "reward": 0.5803571939468384,
      "reward_std": 0.23613807559013367,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 341
    },
    {
      "completion_length": 587.8839721679688,
      "epoch": 0.7292110874200426,
      "grad_norm": 0.20251357555389404,
      "grpo_loss": 0.001199830323457718,
      "kl": 0.047607421875,
      "learning_rate": 6.219396772137218e-07,
      "loss": 0.0019,
      "reward": 0.6339285969734192,
      "reward_std": 0.2833836078643799,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 342
    },
    {
      "completion_length": 658.625,
      "epoch": 0.7313432835820896,
      "grad_norm": 0.19635452330112457,
      "grpo_loss": 9.168311953544617e-05,
      "kl": 0.0289306640625,
      "learning_rate": 6.129104571345968e-07,
      "loss": 0.0012,
      "reward": 0.5803571939468384,
      "reward_std": 0.1986774057149887,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 343
    },
    {
      "completion_length": 741.2767944335938,
      "epoch": 0.7334754797441365,
      "grad_norm": 0.11911910772323608,
      "grpo_loss": -0.01411936804652214,
      "kl": 0.021240234375,
      "learning_rate": 6.039304002729494e-07,
      "loss": 0.0009,
      "reward": 0.4464285969734192,
      "reward_std": 0.1750546246767044,
      "rewards/accuracy_reward": 0.4464285969734192,
      "rewards/format_reward": 0.0,
      "step": 344
    },
    {
      "completion_length": 647.1517944335938,
      "epoch": 0.7356076759061834,
      "grad_norm": 0.15977312624454498,
      "grpo_loss": 0.005686178803443909,
      "kl": 0.058349609375,
      "learning_rate": 5.950000043107681e-07,
      "loss": 0.0023,
      "reward": 0.5446428656578064,
      "reward_std": 0.259671151638031,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 345
    },
    {
      "completion_length": 570.1339721679688,
      "epoch": 0.7377398720682303,
      "grad_norm": 0.1959923654794693,
      "grpo_loss": 0.002989797620102763,
      "kl": 0.052490234375,
      "learning_rate": 5.861197641777953e-07,
      "loss": 0.0021,
      "reward": 0.6785714626312256,
      "reward_std": 0.22626349329948425,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 346
    },
    {
      "completion_length": 546.8482666015625,
      "epoch": 0.7398720682302772,
      "grad_norm": 0.15611131489276886,
      "grpo_loss": 0.13409610092639923,
      "kl": 0.035400390625,
      "learning_rate": 5.772901720240971e-07,
      "loss": 0.0014,
      "reward": 0.7767857313156128,
      "reward_std": 0.2389591485261917,
      "rewards/accuracy_reward": 0.7767857313156128,
      "rewards/format_reward": 0.0,
      "step": 347
    },
    {
      "completion_length": 677.794677734375,
      "epoch": 0.7420042643923241,
      "grad_norm": 0.11709578335285187,
      "grpo_loss": 0.1003992035984993,
      "kl": 0.0177001953125,
      "learning_rate": 5.685117171927879e-07,
      "loss": 0.0007,
      "reward": 0.5,
      "reward_std": 0.20555149018764496,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 348
    },
    {
      "completion_length": 671.2142944335938,
      "epoch": 0.744136460554371,
      "grad_norm": 0.14730335772037506,
      "grpo_loss": 0.0018871352076530457,
      "kl": 0.044677734375,
      "learning_rate": 5.597848861929091e-07,
      "loss": 0.0018,
      "reward": 0.6160714626312256,
      "reward_std": 0.21542608737945557,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 349
    },
    {
      "completion_length": 588.6785888671875,
      "epoch": 0.746268656716418,
      "grad_norm": 0.4064622223377228,
      "grpo_loss": 0.00033608078956604004,
      "kl": 0.044189453125,
      "learning_rate": 5.511101626724693e-07,
      "loss": 0.0018,
      "reward": 0.5535714626312256,
      "reward_std": 0.3206648826599121,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 350
    },
    {
      "completion_length": 570.4375,
      "epoch": 0.7484008528784648,
      "grad_norm": 0.1400219202041626,
      "grpo_loss": 0.05647098273038864,
      "kl": 0.02197265625,
      "learning_rate": 5.42488027391638e-07,
      "loss": 0.0009,
      "reward": 0.6339285969734192,
      "reward_std": 0.2331376075744629,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 351
    },
    {
      "completion_length": 632.4732666015625,
      "epoch": 0.7505330490405118,
      "grad_norm": 0.19489286839962006,
      "grpo_loss": -0.046175189316272736,
      "kl": 0.05419921875,
      "learning_rate": 5.339189581961024e-07,
      "loss": 0.0022,
      "reward": 0.6875000596046448,
      "reward_std": 0.3264864683151245,
      "rewards/accuracy_reward": 0.6875000596046448,
      "rewards/format_reward": 0.0,
      "step": 352
    },
    {
      "completion_length": 650.6160888671875,
      "epoch": 0.7526652452025586,
      "grad_norm": 0.1552044153213501,
      "grpo_loss": 0.12257284671068192,
      "kl": 0.0279541015625,
      "learning_rate": 5.254034299905825e-07,
      "loss": 0.0011,
      "reward": 0.5535714626312256,
      "reward_std": 0.24883373081684113,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 353
    },
    {
      "completion_length": 590.6428833007812,
      "epoch": 0.7547974413646056,
      "grad_norm": 0.1753339320421219,
      "grpo_loss": 0.11830225586891174,
      "kl": 0.03076171875,
      "learning_rate": 5.169419147125176e-07,
      "loss": 0.0012,
      "reward": 0.7946428656578064,
      "reward_std": 0.23613807559013367,
      "rewards/accuracy_reward": 0.7946428656578064,
      "rewards/format_reward": 0.0,
      "step": 354
    },
    {
      "completion_length": 533.4553833007812,
      "epoch": 0.7569296375266524,
      "grad_norm": 0.23187799751758575,
      "grpo_loss": 0.0014726445078849792,
      "kl": 0.04736328125,
      "learning_rate": 5.085348813059028e-07,
      "loss": 0.0019,
      "reward": 0.6785714626312256,
      "reward_std": 0.25279709696769714,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 355
    },
    {
      "completion_length": 637.8392944335938,
      "epoch": 0.7590618336886994,
      "grad_norm": 0.1692523956298828,
      "grpo_loss": 0.11937975883483887,
      "kl": 0.01556396484375,
      "learning_rate": 5.001827956953066e-07,
      "loss": 0.0006,
      "reward": 0.535714328289032,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.535714328289032,
      "rewards/format_reward": 0.0,
      "step": 356
    },
    {
      "completion_length": 628.4642944335938,
      "epoch": 0.7611940298507462,
      "grad_norm": 0.15542620420455933,
      "grpo_loss": 0.0015178961912170053,
      "kl": 0.05419921875,
      "learning_rate": 4.918861207600445e-07,
      "loss": 0.0022,
      "reward": 0.598214328289032,
      "reward_std": 0.20564121007919312,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 357
    },
    {
      "completion_length": 535.4642944335938,
      "epoch": 0.7633262260127932,
      "grad_norm": 0.1944689154624939,
      "grpo_loss": 0.1363382488489151,
      "kl": 0.054443359375,
      "learning_rate": 4.8364531630853e-07,
      "loss": 0.0022,
      "reward": 0.6964285969734192,
      "reward_std": 0.31670159101486206,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 358
    },
    {
      "completion_length": 563.6607666015625,
      "epoch": 0.7654584221748401,
      "grad_norm": 0.17013201117515564,
      "grpo_loss": 0.0014447793364524841,
      "kl": 0.041015625,
      "learning_rate": 4.754608390527869e-07,
      "loss": 0.0016,
      "reward": 0.7142857313156128,
      "reward_std": 0.16130642592906952,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 359
    },
    {
      "completion_length": 588.7142944335938,
      "epoch": 0.767590618336887,
      "grad_norm": 0.26594096422195435,
      "grpo_loss": 0.01679377630352974,
      "kl": 0.035888671875,
      "learning_rate": 4.6733314258314206e-07,
      "loss": 0.0014,
      "reward": 0.5535714626312256,
      "reward_std": 0.31379079818725586,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 360
    },
    {
      "completion_length": 599.4464721679688,
      "epoch": 0.7697228144989339,
      "grad_norm": 0.7181060314178467,
      "grpo_loss": 0.00020268559455871582,
      "kl": 0.0322265625,
      "learning_rate": 4.59262677343085e-07,
      "loss": 0.0013,
      "reward": 0.5625,
      "reward_std": 0.25976085662841797,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 361
    },
    {
      "completion_length": 578.7678833007812,
      "epoch": 0.7718550106609808,
      "grad_norm": 0.1754969209432602,
      "grpo_loss": 0.05639899522066116,
      "kl": 0.053955078125,
      "learning_rate": 4.512498906043046e-07,
      "loss": 0.0022,
      "reward": 0.7142857313156128,
      "reward_std": 0.2902577519416809,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 362
    },
    {
      "completion_length": 676.5178833007812,
      "epoch": 0.7739872068230277,
      "grad_norm": 0.1559028923511505,
      "grpo_loss": -0.05473152920603752,
      "kl": 0.04443359375,
      "learning_rate": 4.432952264418995e-07,
      "loss": 0.0018,
      "reward": 0.5,
      "reward_std": 0.2193893939256668,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 363
    },
    {
      "completion_length": 670.5267944335938,
      "epoch": 0.7761194029850746,
      "grad_norm": 0.14586284756660461,
      "grpo_loss": 0.1182800680398941,
      "kl": 0.0235595703125,
      "learning_rate": 4.3539912570976967e-07,
      "loss": 0.0009,
      "reward": 0.598214328289032,
      "reward_std": 0.3098275065422058,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 364
    },
    {
      "completion_length": 514.3125,
      "epoch": 0.7782515991471215,
      "grad_norm": 0.15664242208003998,
      "grpo_loss": 0.0003577885509002954,
      "kl": 0.03173828125,
      "learning_rate": 4.275620260161819e-07,
      "loss": 0.0013,
      "reward": 0.7589285969734192,
      "reward_std": 0.2891154885292053,
      "rewards/accuracy_reward": 0.7589285969734192,
      "rewards/format_reward": 0.0,
      "step": 365
    },
    {
      "completion_length": 647.982177734375,
      "epoch": 0.7803837953091685,
      "grad_norm": 0.15858404338359833,
      "grpo_loss": 0.05947595462203026,
      "kl": 0.052734375,
      "learning_rate": 4.1978436169951883e-07,
      "loss": 0.0021,
      "reward": 0.5178571939468384,
      "reward_std": 0.16235896944999695,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 366
    },
    {
      "completion_length": 689.1875610351562,
      "epoch": 0.7825159914712153,
      "grad_norm": 0.1707778126001358,
      "grpo_loss": -0.11788696050643921,
      "kl": 0.02734375,
      "learning_rate": 4.1206656380420467e-07,
      "loss": 0.0011,
      "reward": 0.4910714626312256,
      "reward_std": 0.2902577519416809,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 367
    },
    {
      "completion_length": 612.2053833007812,
      "epoch": 0.7846481876332623,
      "grad_norm": 0.1905473917722702,
      "grpo_loss": 0.05517677217721939,
      "kl": 0.038330078125,
      "learning_rate": 4.044090600568202e-07,
      "loss": 0.0015,
      "reward": 0.5178571939468384,
      "reward_std": 0.33450281620025635,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 368
    },
    {
      "completion_length": 597.8125,
      "epoch": 0.7867803837953091,
      "grad_norm": 0.2364969551563263,
      "grpo_loss": 0.00435774028301239,
      "kl": 0.054443359375,
      "learning_rate": 3.9681227484239497e-07,
      "loss": 0.0022,
      "reward": 0.5803571939468384,
      "reward_std": 0.3264864683151245,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 369
    },
    {
      "completion_length": 680.4910888671875,
      "epoch": 0.7889125799573561,
      "grad_norm": 0.11851251870393753,
      "grpo_loss": -0.05272572487592697,
      "kl": 0.0164794921875,
      "learning_rate": 3.892766291808893e-07,
      "loss": 0.0007,
      "reward": 0.5714285969734192,
      "reward_std": 0.26363450288772583,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 370
    },
    {
      "completion_length": 576.919677734375,
      "epoch": 0.7910447761194029,
      "grad_norm": 0.14334093034267426,
      "grpo_loss": 0.04737274721264839,
      "kl": 0.031494140625,
      "learning_rate": 3.818025407038582e-07,
      "loss": 0.0013,
      "reward": 0.660714328289032,
      "reward_std": 0.3069167137145996,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 371
    },
    {
      "completion_length": 675.6160888671875,
      "epoch": 0.7931769722814499,
      "grad_norm": 0.16912008821964264,
      "grpo_loss": -0.04576794058084488,
      "kl": 0.03173828125,
      "learning_rate": 3.74390423631311e-07,
      "loss": 0.0013,
      "reward": 0.4732142984867096,
      "reward_std": 0.27641984820365906,
      "rewards/accuracy_reward": 0.4732142984867096,
      "rewards/format_reward": 0.0,
      "step": 372
    },
    {
      "completion_length": 625.6607666015625,
      "epoch": 0.7953091684434968,
      "grad_norm": 0.20823583006858826,
      "grpo_loss": -0.09997741132974625,
      "kl": 0.05322265625,
      "learning_rate": 3.6704068874874916e-07,
      "loss": 0.0021,
      "reward": 0.5625,
      "reward_std": 0.27641984820365906,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 373
    },
    {
      "completion_length": 630.6517944335938,
      "epoch": 0.7974413646055437,
      "grad_norm": 0.28135645389556885,
      "grpo_loss": 0.004614211618900299,
      "kl": 0.0693359375,
      "learning_rate": 3.597537433844046e-07,
      "loss": 0.0028,
      "reward": 0.5267857313156128,
      "reward_std": 0.3029533922672272,
      "rewards/accuracy_reward": 0.5267857313156128,
      "rewards/format_reward": 0.0,
      "step": 374
    },
    {
      "completion_length": 537.8928833007812,
      "epoch": 0.7995735607675906,
      "grad_norm": 0.1784980595111847,
      "grpo_loss": 0.0028876028954982758,
      "kl": 0.03759765625,
      "learning_rate": 3.525299913866615e-07,
      "loss": 0.0015,
      "reward": 0.7500000596046448,
      "reward_std": 0.1554848700761795,
      "rewards/accuracy_reward": 0.7500000596046448,
      "rewards/format_reward": 0.0,
      "step": 375
    },
    {
      "completion_length": 590.2142944335938,
      "epoch": 0.8017057569296375,
      "grad_norm": 0.24804876744747162,
      "grpo_loss": 0.12203040719032288,
      "kl": 0.04443359375,
      "learning_rate": 3.4536983310167975e-07,
      "loss": 0.0018,
      "reward": 0.6964285969734192,
      "reward_std": 0.357073038816452,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 376
    },
    {
      "completion_length": 588.1785888671875,
      "epoch": 0.8038379530916845,
      "grad_norm": 0.23009361326694489,
      "grpo_loss": -0.07306977361440659,
      "kl": 0.04443359375,
      "learning_rate": 3.382736653512016e-07,
      "loss": 0.0018,
      "reward": 0.6428571939468384,
      "reward_std": 0.26267164945602417,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 377
    },
    {
      "completion_length": 628.3125,
      "epoch": 0.8059701492537313,
      "grad_norm": 0.13790591061115265,
      "grpo_loss": 0.0014116030652076006,
      "kl": 0.0240478515625,
      "learning_rate": 3.312418814105638e-07,
      "loss": 0.001,
      "reward": 0.6071428656578064,
      "reward_std": 0.1820184290409088,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 378
    },
    {
      "completion_length": 552.4553833007812,
      "epoch": 0.8081023454157783,
      "grad_norm": 0.18349257111549377,
      "grpo_loss": -0.09748789668083191,
      "kl": 0.04052734375,
      "learning_rate": 3.242748709868999e-07,
      "loss": 0.0016,
      "reward": 0.660714328289032,
      "reward_std": 0.2822414040565491,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 379
    },
    {
      "completion_length": 673.5535888671875,
      "epoch": 0.8102345415778252,
      "grad_norm": 0.17732153832912445,
      "grpo_loss": 0.0002803046954795718,
      "kl": 0.0341796875,
      "learning_rate": 3.173730201975439e-07,
      "loss": 0.0014,
      "reward": 0.3839285969734192,
      "reward_std": 0.27641984820365906,
      "rewards/accuracy_reward": 0.3839285969734192,
      "rewards/format_reward": 0.0,
      "step": 380
    },
    {
      "completion_length": 615.857177734375,
      "epoch": 0.8123667377398721,
      "grad_norm": 0.22036485373973846,
      "grpo_loss": 0.07462641596794128,
      "kl": 0.032958984375,
      "learning_rate": 3.1053671154862833e-07,
      "loss": 0.0013,
      "reward": 0.660714328289032,
      "reward_std": 0.2873469889163971,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 381
    },
    {
      "completion_length": 657.3392944335938,
      "epoch": 0.814498933901919,
      "grad_norm": 0.14535248279571533,
      "grpo_loss": -0.10004384815692902,
      "kl": 0.021728515625,
      "learning_rate": 3.037663239138895e-07,
      "loss": 0.0009,
      "reward": 0.5535714626312256,
      "reward_std": 0.2792409062385559,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 382
    },
    {
      "completion_length": 566.0803833007812,
      "epoch": 0.8166311300639659,
      "grad_norm": 0.2126779854297638,
      "grpo_loss": 0.002174288034439087,
      "kl": 0.05517578125,
      "learning_rate": 2.970622325136669e-07,
      "loss": 0.0022,
      "reward": 0.6964285969734192,
      "reward_std": 0.22626349329948425,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 383
    },
    {
      "completion_length": 602.5178833007812,
      "epoch": 0.8187633262260128,
      "grad_norm": 0.3173239827156067,
      "grpo_loss": -0.13343751430511475,
      "kl": 0.03759765625,
      "learning_rate": 2.904248088941102e-07,
      "loss": 0.0015,
      "reward": 0.7500000596046448,
      "reward_std": 0.2862047255039215,
      "rewards/accuracy_reward": 0.7500000596046448,
      "rewards/format_reward": 0.0,
      "step": 384
    },
    {
      "completion_length": 625.3035888671875,
      "epoch": 0.8208955223880597,
      "grad_norm": 0.13514848053455353,
      "grpo_loss": 0.0001852833229349926,
      "kl": 0.0230712890625,
      "learning_rate": 2.8385442090658554e-07,
      "loss": 0.0009,
      "reward": 0.5446428656578064,
      "reward_std": 0.1692330688238144,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 385
    },
    {
      "completion_length": 539.857177734375,
      "epoch": 0.8230277185501066,
      "grad_norm": 0.21709002554416656,
      "grpo_loss": -0.04592399671673775,
      "kl": 0.04248046875,
      "learning_rate": 2.773514326872911e-07,
      "loss": 0.0017,
      "reward": 0.7589285969734192,
      "reward_std": 0.22926399111747742,
      "rewards/accuracy_reward": 0.7589285969734192,
      "rewards/format_reward": 0.0,
      "step": 386
    },
    {
      "completion_length": 547.044677734375,
      "epoch": 0.8251599147121536,
      "grad_norm": 0.20077502727508545,
      "grpo_loss": 0.0027071647346019745,
      "kl": 0.036865234375,
      "learning_rate": 2.7091620463707565e-07,
      "loss": 0.0015,
      "reward": 0.6428571939468384,
      "reward_std": 0.27527761459350586,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 387
    },
    {
      "completion_length": 675.4732666015625,
      "epoch": 0.8272921108742004,
      "grad_norm": 0.23532140254974365,
      "grpo_loss": -0.047013476490974426,
      "kl": 0.042724609375,
      "learning_rate": 2.6454909340146526e-07,
      "loss": 0.0017,
      "reward": 0.6339285969734192,
      "reward_std": 0.27641984820365906,
      "rewards/accuracy_reward": 0.6339285969734192,
      "rewards/format_reward": 0.0,
      "step": 388
    },
    {
      "completion_length": 579.1339721679688,
      "epoch": 0.8294243070362474,
      "grad_norm": 0.17333650588989258,
      "grpo_loss": 0.0011966601014137268,
      "kl": 0.02001953125,
      "learning_rate": 2.5825045185089576e-07,
      "loss": 0.0008,
      "reward": 0.5535714626312256,
      "reward_std": 0.27236682176589966,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 389
    },
    {
      "completion_length": 663.6428833007812,
      "epoch": 0.8315565031982942,
      "grad_norm": 0.15946130454540253,
      "grpo_loss": 0.11816628277301788,
      "kl": 0.0296630859375,
      "learning_rate": 2.5202062906115886e-07,
      "loss": 0.0012,
      "reward": 0.5625,
      "reward_std": 0.3236653804779053,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 390
    },
    {
      "completion_length": 727.0982666015625,
      "epoch": 0.8336886993603412,
      "grad_norm": 0.1320721060037613,
      "grpo_loss": -0.04703760892152786,
      "kl": 0.0279541015625,
      "learning_rate": 2.458599702940551e-07,
      "loss": 0.0011,
      "reward": 0.5803571939468384,
      "reward_std": 0.3098274767398834,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 391
    },
    {
      "completion_length": 661.625,
      "epoch": 0.835820895522388,
      "grad_norm": 0.21743802726268768,
      "grpo_loss": -0.0698922798037529,
      "kl": 0.039794921875,
      "learning_rate": 2.3976881697825945e-07,
      "loss": 0.0016,
      "reward": 0.5803571939468384,
      "reward_std": 0.25685009360313416,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 392
    },
    {
      "completion_length": 641.2053833007812,
      "epoch": 0.837953091684435,
      "grad_norm": 3.930511236190796,
      "grpo_loss": -0.07228344678878784,
      "kl": 0.030517578125,
      "learning_rate": 2.337475066903973e-07,
      "loss": 0.0012,
      "reward": 0.625,
      "reward_std": 0.3098274767398834,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 393
    },
    {
      "completion_length": 551.8303833007812,
      "epoch": 0.8400852878464818,
      "grad_norm": 0.19481515884399414,
      "grpo_loss": 0.00013812871475238353,
      "kl": 0.019287109375,
      "learning_rate": 2.2779637313633867e-07,
      "loss": 0.0008,
      "reward": 0.6964285969734192,
      "reward_std": 0.21251529455184937,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 394
    },
    {
      "completion_length": 632.1964721679688,
      "epoch": 0.8422174840085288,
      "grad_norm": 0.28915032744407654,
      "grpo_loss": -0.04566855728626251,
      "kl": 0.04443359375,
      "learning_rate": 2.2191574613270176e-07,
      "loss": 0.0018,
      "reward": 0.660714328289032,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 395
    },
    {
      "completion_length": 733.4732666015625,
      "epoch": 0.8443496801705757,
      "grad_norm": 0.258179247379303,
      "grpo_loss": -0.11782819032669067,
      "kl": 0.0400390625,
      "learning_rate": 2.16105951588575e-07,
      "loss": 0.0016,
      "reward": 0.598214328289032,
      "reward_std": 0.316791296005249,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 396
    },
    {
      "completion_length": 607.3839721679688,
      "epoch": 0.8464818763326226,
      "grad_norm": 0.3129705488681793,
      "grpo_loss": -0.045982833951711655,
      "kl": 0.052001953125,
      "learning_rate": 2.103673114874552e-07,
      "loss": 0.0021,
      "reward": 0.598214328289032,
      "reward_std": 0.27536728978157043,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 397
    },
    {
      "completion_length": 571.7589721679688,
      "epoch": 0.8486140724946695,
      "grad_norm": 0.6058499217033386,
      "grpo_loss": 0.05869365483522415,
      "kl": 0.046142578125,
      "learning_rate": 2.047001438694015e-07,
      "loss": 0.0018,
      "reward": 0.7678571939468384,
      "reward_std": 0.36782076954841614,
      "rewards/accuracy_reward": 0.7678571939468384,
      "rewards/format_reward": 0.0,
      "step": 398
    },
    {
      "completion_length": 669.75,
      "epoch": 0.8507462686567164,
      "grad_norm": 0.19755227863788605,
      "grpo_loss": -0.1000501960515976,
      "kl": 0.031005859375,
      "learning_rate": 1.9910476281341155e-07,
      "loss": 0.0012,
      "reward": 0.5178571939468384,
      "reward_std": 0.24988628923892975,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 399
    },
    {
      "completion_length": 650.6964721679688,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.24487632513046265,
      "grpo_loss": 0.0018350593745708466,
      "kl": 0.0625,
      "learning_rate": 1.9358147842001328e-07,
      "loss": 0.0025,
      "reward": 0.5178571939468384,
      "reward_std": 0.1514318436384201,
      "rewards/accuracy_reward": 0.5178571939468384,
      "rewards/format_reward": 0.0,
      "step": 400
    },
    {
      "epoch": 0.8528784648187633,
      "eval_completion_length": 611.5137771752695,
      "eval_grpo_loss": 0.002802959766321497,
      "eval_kl": 0.049074215629992014,
      "eval_loss": 0.0019418156007304788,
      "eval_reward": 0.5530294624570841,
      "eval_reward_std": 0.28017472720945985,
      "eval_rewards/accuracy_reward": 0.5530294624570841,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 4037.4288,
      "eval_samples_per_second": 1.238,
      "eval_steps_per_second": 0.011,
      "step": 400
    },
    {
      "completion_length": 519.6964721679688,
      "epoch": 0.8550106609808102,
      "grad_norm": 0.25118526816368103,
      "grpo_loss": -0.04701920598745346,
      "kl": 0.04052734375,
      "learning_rate": 1.8813059679408012e-07,
      "loss": 0.0016,
      "reward": 0.785714328289032,
      "reward_std": 0.24592295289039612,
      "rewards/accuracy_reward": 0.785714328289032,
      "rewards/format_reward": 0.0,
      "step": 401
    },
    {
      "completion_length": 615.7232666015625,
      "epoch": 0.8571428571428571,
      "grad_norm": 0.2910488545894623,
      "grpo_loss": 0.00017983283032663167,
      "kl": 0.0478515625,
      "learning_rate": 1.827524200278648e-07,
      "loss": 0.0019,
      "reward": 0.6428571939468384,
      "reward_std": 0.18889254331588745,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 402
    },
    {
      "completion_length": 614.8035888671875,
      "epoch": 0.8592750533049041,
      "grad_norm": 0.31290826201438904,
      "grpo_loss": -0.045705024152994156,
      "kl": 0.02880859375,
      "learning_rate": 1.7744724618425868e-07,
      "loss": 0.0012,
      "reward": 0.6517857313156128,
      "reward_std": 0.3277184069156647,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 403
    },
    {
      "completion_length": 629.3928833007812,
      "epoch": 0.8614072494669509,
      "grad_norm": 0.35251688957214355,
      "grpo_loss": -0.11518432199954987,
      "kl": 0.04443359375,
      "learning_rate": 1.7221536928027232e-07,
      "loss": 0.0018,
      "reward": 0.625,
      "reward_std": 0.3334502577781677,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 404
    },
    {
      "completion_length": 609.9375,
      "epoch": 0.8635394456289979,
      "grad_norm": 0.4341069757938385,
      "grpo_loss": -0.046183593571186066,
      "kl": 0.03466796875,
      "learning_rate": 1.6705707927074104e-07,
      "loss": 0.0014,
      "reward": 0.6696428656578064,
      "reward_std": 0.3264864683151245,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 405
    },
    {
      "completion_length": 611.6517944335938,
      "epoch": 0.8656716417910447,
      "grad_norm": 0.4119378626346588,
      "grpo_loss": -0.013815686106681824,
      "kl": 0.03759765625,
      "learning_rate": 1.6197266203225441e-07,
      "loss": 0.0015,
      "reward": 0.5625,
      "reward_std": 0.3127382695674896,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 406
    },
    {
      "completion_length": 590.1875,
      "epoch": 0.8678038379530917,
      "grad_norm": 0.3130354583263397,
      "grpo_loss": 0.0482063889503479,
      "kl": 0.041748046875,
      "learning_rate": 1.5696239934731394e-07,
      "loss": 0.0017,
      "reward": 0.6160714626312256,
      "reward_std": 0.3235756754875183,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 407
    },
    {
      "completion_length": 489.95538330078125,
      "epoch": 0.8699360341151386,
      "grad_norm": 1.4019256830215454,
      "grpo_loss": 0.0030681490898132324,
      "kl": 0.07373046875,
      "learning_rate": 1.520265688887162e-07,
      "loss": 0.0029,
      "reward": 0.6785714626312256,
      "reward_std": 0.3737320303916931,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 408
    },
    {
      "completion_length": 634.6785888671875,
      "epoch": 0.8720682302771855,
      "grad_norm": 0.7867122888565063,
      "grpo_loss": 0.002787616103887558,
      "kl": 0.049560546875,
      "learning_rate": 1.4716544420416366e-07,
      "loss": 0.002,
      "reward": 0.6785714626312256,
      "reward_std": 0.3265761733055115,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 409
    },
    {
      "completion_length": 695.1964721679688,
      "epoch": 0.8742004264392325,
      "grad_norm": 0.35332122445106506,
      "grpo_loss": 0.0004137594369240105,
      "kl": 0.032470703125,
      "learning_rate": 1.4237929470110406e-07,
      "loss": 0.0013,
      "reward": 0.5803571939468384,
      "reward_std": 0.26267164945602417,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 410
    },
    {
      "completion_length": 672.4285888671875,
      "epoch": 0.8763326226012793,
      "grad_norm": 0.5385404229164124,
      "grpo_loss": 0.07336309552192688,
      "kl": 0.042236328125,
      "learning_rate": 1.3766838563180117e-07,
      "loss": 0.0017,
      "reward": 0.5446428656578064,
      "reward_std": 0.31379079818725586,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 411
    },
    {
      "completion_length": 591.0267944335938,
      "epoch": 0.8784648187633263,
      "grad_norm": 1.787886619567871,
      "grpo_loss": -0.053161606192588806,
      "kl": 0.2265625,
      "learning_rate": 1.330329780786324e-07,
      "loss": 0.0091,
      "reward": 0.6696428656578064,
      "reward_std": 0.3177541494369507,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 412
    },
    {
      "completion_length": 596.6339721679688,
      "epoch": 0.8805970149253731,
      "grad_norm": 0.8436120748519897,
      "grpo_loss": 0.01977485604584217,
      "kl": 0.045654296875,
      "learning_rate": 1.2847332893962078e-07,
      "loss": 0.0018,
      "reward": 0.7142857313156128,
      "reward_std": 0.2862047255039215,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 413
    },
    {
      "completion_length": 618.9464721679688,
      "epoch": 0.8827292110874201,
      "grad_norm": 0.6383178234100342,
      "grpo_loss": 0.0014204904437065125,
      "kl": 0.054931640625,
      "learning_rate": 1.23989690914196e-07,
      "loss": 0.0022,
      "reward": 0.6160714626312256,
      "reward_std": 0.2724565267562866,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 414
    },
    {
      "completion_length": 668.4732666015625,
      "epoch": 0.8848614072494669,
      "grad_norm": 0.4944930076599121,
      "grpo_loss": 0.0009383819997310638,
      "kl": 0.04052734375,
      "learning_rate": 1.1958231248919232e-07,
      "loss": 0.0016,
      "reward": 0.5446428656578064,
      "reward_std": 0.3519674837589264,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 415
    },
    {
      "completion_length": 652.8035888671875,
      "epoch": 0.8869936034115139,
      "grad_norm": 0.739689826965332,
      "grpo_loss": 0.001788435154594481,
      "kl": 0.05615234375,
      "learning_rate": 1.1525143792507397e-07,
      "loss": 0.0022,
      "reward": 0.625,
      "reward_std": 0.24988627433776855,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 416
    },
    {
      "completion_length": 621.5357666015625,
      "epoch": 0.8891257995735607,
      "grad_norm": 1.2304091453552246,
      "grpo_loss": 0.0009033530950546265,
      "kl": 0.1025390625,
      "learning_rate": 1.1099730724240037e-07,
      "loss": 0.0041,
      "reward": 0.660714328289032,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 417
    },
    {
      "completion_length": 579.1964721679688,
      "epoch": 0.8912579957356077,
      "grad_norm": 1.6807209253311157,
      "grpo_loss": 0.0036621836479753256,
      "kl": 0.07275390625,
      "learning_rate": 1.0682015620852214e-07,
      "loss": 0.0029,
      "reward": 0.6964285969734192,
      "reward_std": 0.2666349411010742,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 418
    },
    {
      "completion_length": 513.3928833007812,
      "epoch": 0.8933901918976546,
      "grad_norm": 7.738749027252197,
      "grpo_loss": 0.12100201845169067,
      "kl": 0.470703125,
      "learning_rate": 1.0272021632451745e-07,
      "loss": 0.0188,
      "reward": 0.6160714626312256,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 419
    },
    {
      "completion_length": 633.0892944335938,
      "epoch": 0.8955223880597015,
      "grad_norm": 0.5201921463012695,
      "grpo_loss": 0.12260356545448303,
      "kl": 0.046142578125,
      "learning_rate": 9.869771481235851e-08,
      "loss": 0.0019,
      "reward": 0.5535714626312256,
      "reward_std": 0.24883373081684113,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 420
    },
    {
      "completion_length": 627.919677734375,
      "epoch": 0.8976545842217484,
      "grad_norm": 1.1365468502044678,
      "grpo_loss": 0.12431085109710693,
      "kl": 0.054443359375,
      "learning_rate": 9.475287460232174e-08,
      "loss": 0.0022,
      "reward": 0.6428571939468384,
      "reward_std": 0.2832939326763153,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 421
    },
    {
      "completion_length": 687.1428833007812,
      "epoch": 0.8997867803837953,
      "grad_norm": 0.8904653191566467,
      "grpo_loss": 0.0030726641416549683,
      "kl": 0.07666015625,
      "learning_rate": 9.088591432063109e-08,
      "loss": 0.0031,
      "reward": 0.4464285969734192,
      "reward_std": 0.24883374571800232,
      "rewards/accuracy_reward": 0.4464285969734192,
      "rewards/format_reward": 0.0,
      "step": 422
    },
    {
      "completion_length": 702.982177734375,
      "epoch": 0.9019189765458422,
      "grad_norm": 0.5287430882453918,
      "grpo_loss": 0.0010234508663415909,
      "kl": 0.03564453125,
      "learning_rate": 8.709704827734244e-08,
      "loss": 0.0014,
      "reward": 0.4732142984867096,
      "reward_std": 0.21251530945301056,
      "rewards/accuracy_reward": 0.4732142984867096,
      "rewards/format_reward": 0.0,
      "step": 423
    },
    {
      "completion_length": 582.6607666015625,
      "epoch": 0.9040511727078892,
      "grad_norm": 1.1305034160614014,
      "grpo_loss": 0.0023384217638522387,
      "kl": 0.0654296875,
      "learning_rate": 8.33864864544654e-08,
      "loss": 0.0026,
      "reward": 0.660714328289032,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.660714328289032,
      "rewards/format_reward": 0.0,
      "step": 424
    },
    {
      "completion_length": 522.7589721679688,
      "epoch": 0.906183368869936,
      "grad_norm": 1.2338292598724365,
      "grpo_loss": 0.0016110788565129042,
      "kl": 0.08740234375,
      "learning_rate": 7.975443449432712e-08,
      "loss": 0.0035,
      "reward": 0.7410714626312256,
      "reward_std": 0.23904886841773987,
      "rewards/accuracy_reward": 0.7410714626312256,
      "rewards/format_reward": 0.0,
      "step": 425
    },
    {
      "completion_length": 662.3392944335938,
      "epoch": 0.908315565031983,
      "grad_norm": 1.3673124313354492,
      "grpo_loss": 0.004744287580251694,
      "kl": 0.06884765625,
      "learning_rate": 7.620109368817479e-08,
      "loss": 0.0028,
      "reward": 0.6428571939468384,
      "reward_std": 0.3069167137145996,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 426
    },
    {
      "completion_length": 576.375,
      "epoch": 0.9104477611940298,
      "grad_norm": 1.6284408569335938,
      "grpo_loss": -0.09632082283496857,
      "kl": 0.0859375,
      "learning_rate": 7.272666096502017e-08,
      "loss": 0.0034,
      "reward": 0.5625,
      "reward_std": 0.35116177797317505,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 427
    },
    {
      "completion_length": 515.625,
      "epoch": 0.9125799573560768,
      "grad_norm": 1.218694806098938,
      "grpo_loss": 0.0009350869804620743,
      "kl": 0.0625,
      "learning_rate": 6.933132888072452e-08,
      "loss": 0.0025,
      "reward": 0.7321428656578064,
      "reward_std": 0.29713186621665955,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 428
    },
    {
      "completion_length": 610.9642944335938,
      "epoch": 0.9147121535181236,
      "grad_norm": 1.6818090677261353,
      "grpo_loss": 0.002560252556577325,
      "kl": 0.0849609375,
      "learning_rate": 6.601528560732978e-08,
      "loss": 0.0034,
      "reward": 0.7142857313156128,
      "reward_std": 0.3040059804916382,
      "rewards/accuracy_reward": 0.7142857313156128,
      "rewards/format_reward": 0.0,
      "step": 429
    },
    {
      "completion_length": 561.5714721679688,
      "epoch": 0.9168443496801706,
      "grad_norm": 1.855384349822998,
      "grpo_loss": 0.005668744444847107,
      "kl": 0.10302734375,
      "learning_rate": 6.277871492262593e-08,
      "loss": 0.0041,
      "reward": 0.7410714626312256,
      "reward_std": 0.36685794591903687,
      "rewards/accuracy_reward": 0.7410714626312256,
      "rewards/format_reward": 0.0,
      "step": 430
    },
    {
      "completion_length": 650.4285888671875,
      "epoch": 0.9189765458422174,
      "grad_norm": 1.2501542568206787,
      "grpo_loss": -0.07116611301898956,
      "kl": 0.049560546875,
      "learning_rate": 5.962179619996966e-08,
      "loss": 0.002,
      "reward": 0.4910714626312256,
      "reward_std": 0.2862047255039215,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 431
    },
    {
      "completion_length": 606.2410888671875,
      "epoch": 0.9211087420042644,
      "grad_norm": 1.4896941184997559,
      "grpo_loss": -0.014261804521083832,
      "kl": 0.1328125,
      "learning_rate": 5.654470439834058e-08,
      "loss": 0.0053,
      "reward": 0.6696428656578064,
      "reward_std": 0.29016801714897156,
      "rewards/accuracy_reward": 0.6696428656578064,
      "rewards/format_reward": 0.0,
      "step": 432
    },
    {
      "completion_length": 596.0625,
      "epoch": 0.9232409381663113,
      "grad_norm": 1.6670781373977661,
      "grpo_loss": -0.0711505264043808,
      "kl": 0.126953125,
      "learning_rate": 5.3547610052647246e-08,
      "loss": 0.0051,
      "reward": 0.7321428656578064,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.7321428656578064,
      "rewards/format_reward": 0.0,
      "step": 433
    },
    {
      "completion_length": 752.544677734375,
      "epoch": 0.9253731343283582,
      "grad_norm": 1.245611310005188,
      "grpo_loss": 0.007793694734573364,
      "kl": 0.0771484375,
      "learning_rate": 5.063067926427345e-08,
      "loss": 0.0031,
      "reward": 0.4732142984867096,
      "reward_std": 0.3363610506057739,
      "rewards/accuracy_reward": 0.4732142984867096,
      "rewards/format_reward": 0.0,
      "step": 434
    },
    {
      "completion_length": 634.5535888671875,
      "epoch": 0.9275053304904051,
      "grad_norm": 1.4781724214553833,
      "grpo_loss": 0.0024833083152770996,
      "kl": 0.07666015625,
      "learning_rate": 4.7794073691875e-08,
      "loss": 0.0031,
      "reward": 0.5714285969734192,
      "reward_std": 0.3098274767398834,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 435
    },
    {
      "completion_length": 675.8392944335938,
      "epoch": 0.929637526652452,
      "grad_norm": 1.7764922380447388,
      "grpo_loss": 0.10629189014434814,
      "kl": 0.130859375,
      "learning_rate": 4.5037950542418327e-08,
      "loss": 0.0052,
      "reward": 0.6071428656578064,
      "reward_std": 0.23208504915237427,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 436
    },
    {
      "completion_length": 595.0267944335938,
      "epoch": 0.9317697228144989,
      "grad_norm": 0.9373055696487427,
      "grpo_loss": 0.0014045275747776031,
      "kl": 0.0751953125,
      "learning_rate": 4.236246256247084e-08,
      "loss": 0.003,
      "reward": 0.6517857313156128,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 437
    },
    {
      "completion_length": 622.2053833007812,
      "epoch": 0.9339019189765458,
      "grad_norm": 1.3948901891708374,
      "grpo_loss": 0.003673933446407318,
      "kl": 0.09716796875,
      "learning_rate": 3.976775802973232e-08,
      "loss": 0.0039,
      "reward": 0.6785714626312256,
      "reward_std": 0.3206649422645569,
      "rewards/accuracy_reward": 0.6785714626312256,
      "rewards/format_reward": 0.0,
      "step": 438
    },
    {
      "completion_length": 641.232177734375,
      "epoch": 0.9360341151385928,
      "grad_norm": 1.9123766422271729,
      "grpo_loss": -0.036110423505306244,
      "kl": 0.16015625,
      "learning_rate": 3.7253980744819335e-08,
      "loss": 0.0064,
      "reward": 0.5535714626312256,
      "reward_std": 0.3265761733055115,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 439
    },
    {
      "completion_length": 649.419677734375,
      "epoch": 0.9381663113006397,
      "grad_norm": 1.4534177780151367,
      "grpo_loss": 0.003873981535434723,
      "kl": 0.162109375,
      "learning_rate": 3.482127002329532e-08,
      "loss": 0.0065,
      "reward": 0.6428571939468384,
      "reward_std": 0.27933064103126526,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 440
    },
    {
      "completion_length": 701.0089721679688,
      "epoch": 0.9402985074626866,
      "grad_norm": 3.5775556564331055,
      "grpo_loss": 0.12740851938724518,
      "kl": 0.13671875,
      "learning_rate": 3.246976068794938e-08,
      "loss": 0.0055,
      "reward": 0.5089285969734192,
      "reward_std": 0.28338366746902466,
      "rewards/accuracy_reward": 0.5089285969734192,
      "rewards/format_reward": 0.0,
      "step": 441
    },
    {
      "completion_length": 630.2410888671875,
      "epoch": 0.9424307036247335,
      "grad_norm": 2.2547216415405273,
      "grpo_loss": 0.0050017498433589935,
      "kl": 0.2080078125,
      "learning_rate": 3.019958306132409e-08,
      "loss": 0.0083,
      "reward": 0.5,
      "reward_std": 0.40713968873023987,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.0,
      "step": 442
    },
    {
      "completion_length": 545.8839721679688,
      "epoch": 0.9445628997867804,
      "grad_norm": 0.666123628616333,
      "grpo_loss": 0.010975733399391174,
      "kl": 0.09765625,
      "learning_rate": 2.8010862958493143e-08,
      "loss": 0.0039,
      "reward": 0.6964285969734192,
      "reward_std": 0.2153363674879074,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 443
    },
    {
      "completion_length": 654.375,
      "epoch": 0.9466950959488273,
      "grad_norm": 4.539513111114502,
      "grpo_loss": 0.07739832252264023,
      "kl": 0.28125,
      "learning_rate": 2.5903721680089088e-08,
      "loss": 0.0112,
      "reward": 0.5535714626312256,
      "reward_std": 0.27933061122894287,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 0.0,
      "step": 444
    },
    {
      "completion_length": 606.5892944335938,
      "epoch": 0.9488272921108742,
      "grad_norm": 1.9015904664993286,
      "grpo_loss": -0.04408063739538193,
      "kl": 0.09375,
      "learning_rate": 2.3878276005580014e-08,
      "loss": 0.0037,
      "reward": 0.7767857313156128,
      "reward_std": 0.30004262924194336,
      "rewards/accuracy_reward": 0.7767857313156128,
      "rewards/format_reward": 0.0,
      "step": 445
    },
    {
      "completion_length": 636.732177734375,
      "epoch": 0.9509594882729211,
      "grad_norm": 0.7139691710472107,
      "grpo_loss": -0.09958329796791077,
      "kl": 0.0654296875,
      "learning_rate": 2.1934638186797418e-08,
      "loss": 0.0026,
      "reward": 0.6517857313156128,
      "reward_std": 0.31670159101486206,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 446
    },
    {
      "completion_length": 550.25,
      "epoch": 0.9530916844349681,
      "grad_norm": 0.9299170970916748,
      "grpo_loss": -0.0699908658862114,
      "kl": 0.1328125,
      "learning_rate": 2.0072915941716176e-08,
      "loss": 0.0053,
      "reward": 0.5714285969734192,
      "reward_std": 0.3235757052898407,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 447
    },
    {
      "completion_length": 547.2857666015625,
      "epoch": 0.9552238805970149,
      "grad_norm": 1.3427575826644897,
      "grpo_loss": 0.0044424869120121,
      "kl": 0.0791015625,
      "learning_rate": 1.8293212448483476e-08,
      "loss": 0.0032,
      "reward": 0.6964285969734192,
      "reward_std": 0.24301216006278992,
      "rewards/accuracy_reward": 0.6964285969734192,
      "rewards/format_reward": 0.0,
      "step": 448
    },
    {
      "completion_length": 607.3392944335938,
      "epoch": 0.9573560767590619,
      "grad_norm": 1933856.75,
      "grpo_loss": 0.0023131519556045532,
      "kl": 27648.0,
      "learning_rate": 1.6595626339701407e-08,
      "loss": 1106.2889,
      "reward": 0.6428571939468384,
      "reward_std": 0.27641984820365906,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 449
    },
    {
      "completion_length": 599.5803833007812,
      "epoch": 0.9594882729211087,
      "grad_norm": 0.7657116055488586,
      "grpo_loss": 0.11946123838424683,
      "kl": 0.1064453125,
      "learning_rate": 1.498025169696049e-08,
      "loss": 0.0043,
      "reward": 0.6160714626312256,
      "reward_std": 0.3098275065422058,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 450
    },
    {
      "completion_length": 558.044677734375,
      "epoch": 0.9616204690831557,
      "grad_norm": 1.7871955633163452,
      "grpo_loss": 0.0009799189865589142,
      "kl": 0.10888671875,
      "learning_rate": 1.344717804562534e-08,
      "loss": 0.0044,
      "reward": 0.7053571939468384,
      "reward_std": 0.23904886841773987,
      "rewards/accuracy_reward": 0.7053571939468384,
      "rewards/format_reward": 0.0,
      "step": 451
    },
    {
      "completion_length": 568.357177734375,
      "epoch": 0.9637526652452025,
      "grad_norm": 93.76356506347656,
      "grpo_loss": 0.013948827981948853,
      "kl": 1.28125,
      "learning_rate": 1.1996490349873657e-08,
      "loss": 0.0513,
      "reward": 0.6428571939468384,
      "reward_std": 0.2695457339286804,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.0,
      "step": 452
    },
    {
      "completion_length": 672.2678833007812,
      "epoch": 0.9658848614072495,
      "grad_norm": 2.1468069553375244,
      "grpo_loss": -0.04484168440103531,
      "kl": 0.1865234375,
      "learning_rate": 1.062826900798647e-08,
      "loss": 0.0075,
      "reward": 0.5803571939468384,
      "reward_std": 0.3708212673664093,
      "rewards/accuracy_reward": 0.5803571939468384,
      "rewards/format_reward": 0.0,
      "step": 453
    },
    {
      "completion_length": 598.8303833007812,
      "epoch": 0.9680170575692963,
      "grad_norm": 1.5067918300628662,
      "grpo_loss": -0.00438106432557106,
      "kl": 0.16796875,
      "learning_rate": 9.34258984789338e-09,
      "loss": 0.0067,
      "reward": 0.6160714626312256,
      "reward_std": 0.3363610506057739,
      "rewards/accuracy_reward": 0.6160714626312256,
      "rewards/format_reward": 0.0,
      "step": 454
    },
    {
      "completion_length": 666.1964721679688,
      "epoch": 0.9701492537313433,
      "grad_norm": 1.465742588043213,
      "grpo_loss": 0.009304620325565338,
      "kl": 0.1376953125,
      "learning_rate": 8.139524122970254e-09,
      "loss": 0.0055,
      "reward": 0.4285714626312256,
      "reward_std": 0.20264071226119995,
      "rewards/accuracy_reward": 0.4285714626312256,
      "rewards/format_reward": 0.0,
      "step": 455
    },
    {
      "completion_length": 611.8839721679688,
      "epoch": 0.9722814498933902,
      "grad_norm": 0.6134836077690125,
      "grpo_loss": -0.06908067315816879,
      "kl": 0.072265625,
      "learning_rate": 7.019138508088552e-09,
      "loss": 0.0029,
      "reward": 0.6517857313156128,
      "reward_std": 0.27641984820365906,
      "rewards/accuracy_reward": 0.6517857313156128,
      "rewards/format_reward": 0.0,
      "step": 456
    },
    {
      "completion_length": 609.9285888671875,
      "epoch": 0.9744136460554371,
      "grad_norm": 2.488274097442627,
      "grpo_loss": -0.09877821803092957,
      "kl": 0.125,
      "learning_rate": 5.98149509592244e-09,
      "loss": 0.005,
      "reward": 0.5267857313156128,
      "reward_std": 0.26945602893829346,
      "rewards/accuracy_reward": 0.5267857313156128,
      "rewards/format_reward": 0.0,
      "step": 457
    },
    {
      "completion_length": 549.5535888671875,
      "epoch": 0.976545842217484,
      "grad_norm": 26.11075210571289,
      "grpo_loss": 0.21839597821235657,
      "kl": 0.8359375,
      "learning_rate": 5.026651393506387e-09,
      "loss": 0.0335,
      "reward": 0.7410714626312256,
      "reward_std": 0.316791296005249,
      "rewards/accuracy_reward": 0.7410714626312256,
      "rewards/format_reward": 0.0,
      "step": 458
    },
    {
      "completion_length": 562.25,
      "epoch": 0.9786780383795309,
      "grad_norm": 5.394749164581299,
      "grpo_loss": 0.0021801290567964315,
      "kl": 0.3359375,
      "learning_rate": 4.154660319047543e-09,
      "loss": 0.0134,
      "reward": 0.723214328289032,
      "reward_std": 0.22926399111747742,
      "rewards/accuracy_reward": 0.723214328289032,
      "rewards/format_reward": 0.0,
      "step": 459
    },
    {
      "completion_length": 687.1517944335938,
      "epoch": 0.9808102345415778,
      "grad_norm": 5.675504207611084,
      "grpo_loss": 0.06382696330547333,
      "kl": 0.341796875,
      "learning_rate": 3.3655701989944164e-09,
      "loss": 0.0137,
      "reward": 0.4821428656578064,
      "reward_std": 0.26363447308540344,
      "rewards/accuracy_reward": 0.4821428656578064,
      "rewards/format_reward": 0.0,
      "step": 460
    },
    {
      "completion_length": 585.3125,
      "epoch": 0.9829424307036247,
      "grad_norm": 2.350153684616089,
      "grpo_loss": 0.0019125863909721375,
      "kl": 0.1630859375,
      "learning_rate": 2.659424765357521e-09,
      "loss": 0.0065,
      "reward": 0.598214328289032,
      "reward_std": 0.2331376075744629,
      "rewards/accuracy_reward": 0.598214328289032,
      "rewards/format_reward": 0.0,
      "step": 461
    },
    {
      "completion_length": 660.732177734375,
      "epoch": 0.9850746268656716,
      "grad_norm": 5.494519233703613,
      "grpo_loss": -0.0505463182926178,
      "kl": 0.2412109375,
      "learning_rate": 2.0362631532856445e-09,
      "loss": 0.0096,
      "reward": 0.4910714626312256,
      "reward_std": 0.3402346670627594,
      "rewards/accuracy_reward": 0.4910714626312256,
      "rewards/format_reward": 0.0,
      "step": 462
    },
    {
      "completion_length": 677.8928833007812,
      "epoch": 0.9872068230277186,
      "grad_norm": 1.2300056219100952,
      "grpo_loss": -0.049303088337183,
      "kl": 0.1044921875,
      "learning_rate": 1.4961198988979185e-09,
      "loss": 0.0042,
      "reward": 0.5892857313156128,
      "reward_std": 0.3501092493534088,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 463
    },
    {
      "completion_length": 562.107177734375,
      "epoch": 0.9893390191897654,
      "grad_norm": 1.0299087762832642,
      "grpo_loss": -0.09899073094129562,
      "kl": 0.12890625,
      "learning_rate": 1.0390249373686823e-09,
      "loss": 0.0052,
      "reward": 0.7589285969734192,
      "reward_std": 0.26945602893829346,
      "rewards/accuracy_reward": 0.7589285969734192,
      "rewards/format_reward": 0.0,
      "step": 464
    },
    {
      "completion_length": 614.982177734375,
      "epoch": 0.9914712153518124,
      "grad_norm": 1.0833293199539185,
      "grpo_loss": 0.006685294210910797,
      "kl": 0.123046875,
      "learning_rate": 6.650036012693095e-10,
      "loss": 0.0049,
      "reward": 0.5625,
      "reward_std": 0.2360483705997467,
      "rewards/accuracy_reward": 0.5625,
      "rewards/format_reward": 0.0,
      "step": 465
    },
    {
      "completion_length": 575.0089721679688,
      "epoch": 0.9936034115138592,
      "grad_norm": 0.6512173414230347,
      "grpo_loss": 0.07838420569896698,
      "kl": 0.10595703125,
      "learning_rate": 3.7407661916349924e-10,
      "loss": 0.0042,
      "reward": 0.6071428656578064,
      "reward_std": 0.346056193113327,
      "rewards/accuracy_reward": 0.6071428656578064,
      "rewards/format_reward": 0.0,
      "step": 466
    },
    {
      "completion_length": 612.419677734375,
      "epoch": 0.9957356076759062,
      "grad_norm": 3.9021387100219727,
      "grpo_loss": 0.020383043214678764,
      "kl": 0.228515625,
      "learning_rate": 1.6626011445919353e-10,
      "loss": 0.0091,
      "reward": 0.5892857313156128,
      "reward_std": 0.22917428612709045,
      "rewards/accuracy_reward": 0.5892857313156128,
      "rewards/format_reward": 0.0,
      "step": 467
    },
    {
      "completion_length": 642.1160888671875,
      "epoch": 0.997867803837953,
      "grad_norm": 0.9386311769485474,
      "grpo_loss": 0.009107634425163269,
      "kl": 0.1259765625,
      "learning_rate": 4.156560451462621e-11,
      "loss": 0.005,
      "reward": 0.5714285969734192,
      "reward_std": 0.18483950197696686,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 0.0,
      "step": 468
    },
    {
      "completion_length": 644.1759033203125,
      "epoch": 1.0,
      "grad_norm": 1.1123640537261963,
      "grpo_loss": 0.005602791905403137,
      "kl": 0.10107421875,
      "learning_rate": 0.0,
      "loss": 0.004,
      "reward": 0.625,
      "reward_std": 0.2999529242515564,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.0,
      "step": 469
    },
    {
      "epoch": 1.0,
      "step": 469,
      "total_flos": 0.0,
      "train_loss": 0.0,
      "train_runtime": 1.7067,
      "train_samples_per_second": 4394.419,
      "train_steps_per_second": 274.798
    }
  ],
  "logging_steps": 1,
  "max_steps": 469,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "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": 16,
  "trial_name": null,
  "trial_params": null
}